Late enough that I was not booking a flight to Spain. Late enough that third shift in Buford was not moving out of the way. I was going to miss the thing I was writing about.

I could still build the guide.

Drone pilots had real questions. Could a normal ND filter protect the camera? Where should the aircraft sit during the important minutes? How early should it launch? What happens when the best-looking field fills with people or falls inside controlled airspace?

I had two days to answer them.

The guide went live on August 10. Roughly 48 hours later, the publisher dashboard showed 758 views. Search supplied 62 percent of the traffic. Google alone supplied 60 percent. Nearly three out of every four readers were on a phone.

New subscribers: zero.

That number bothered me more than the 758 impressed me.

The article had done what I asked it to do. It answered a timely question, ranked quickly and reached far beyond my email list. Then those readers took the answer and left.

Search had done its part. I had not.

Discovery worked. Conversion did not.

This case study is free because the lesson belongs in the open. Search can find a useful page. It cannot tell a new reader why the rest of a publication is worth following.

That second job was mine.

I am leaving the workbench open: the numbers, the decisions, the wrong turns and the checklist I will carry into the next timed story.

The eclipse guide split the result cleanly. Discovery worked. Conversion did not.

Here is the dashboard without the victory lap.

The first 48 hours

Total views: 758
Search traffic: 62%
Google traffic: 60%
Direct traffic: 34%
Email traffic: 4%
Mobile web: 74%
Desktop web: 22%
Email recipients: 92
Email opens: 27
Open rate: 29.35%
Link clicks: 2
Click-through rate: 7.41%
Engagement rate: 14.81%
Likes: 7
Public comments: 7
Restacks: 3
New subscribers: 0
Revenue: $0

Those percentages work out to about 455 visits from Google and about 561 readers on mobile.

My email list received 92 copies. Twenty-seven people opened the email. Only 4 percent of total traffic came through email. Google brought a crowd several times larger than the list I already had.

Direct traffic accounted for 34 percent. Some of that may have come from X, Notes, copied links or apps that stripped the referral information. I did not use separate tracked links, so I cannot honestly divide that bucket after the fact.

The numbers are a snapshot, not a lifetime verdict. The article was only about two days old and the eclipse was driving a short, hard burst of interest. I cannot claim the page had earned a permanent search position.

I can claim this: SEO put the article in front of people who had never heard of Drone Drifter.

Then I gave them no clear reason to remember the name.

What worked

The title answered the search

The title did not try to be mysterious:

How to Film the August 12, 2026 Solar Eclipse with a Drone (Without Damaging Your Camera)

It named the event, the tool, the date and the risk. A pilot searching the night before knew exactly what the page promised.

I am leaving that title and URL alone. When a page is already getting search traffic, changing the address for cosmetic reasons is not clever. It breaks old links and forces Google to relearn the page.

The guide dealt with decisions, not spectacle

The useful part was not telling people that an eclipse would look dramatic. Everybody already knew that.

The guide dealt with choices a pilot had to make before takeoff:

• A normal ND filter controls exposure. It is not automatically a safe solar filter.

• A wide drone camera is better at showing the changing ground than filling the frame with the Sun.

• The launch point still needs an airspace check.

• An eclipse crowd is still a crowd.

• Battery planning has to leave room for a return and landing after the shot.

Those decisions gave the article a reason to exist after the headline wore off.

The place kept it from becoming a manual

The guide could have turned into a list of settings copied from camera sites.

Instead, it came back to the ground. Streetlights waking up. Water losing its shine. A church dome holding the last hard light. A town changing color while everyone looked in the other direction.

That is the part that belongs to Drone Drifter.

The deadline cut the soft parts

An event with a fixed date does not care how long I want to polish a sentence. The deadline forced every section to answer a question or leave.

That pressure helped the article.

It also hid the missing conversion path until the traffic arrived.

What failed

The article solved the problem and ended the relationship

A search reader arrived with one job: film the eclipse without hurting the camera or breaking a rule.

The article did that job.

Then it ended with a broad invitation to subscribe for more stories. That invitation could have appeared under almost any article on the internet. It did not connect the eclipse guide to a specific thing the reader would receive next.

Useful is not the same as memorable.

I never built the bridge to the rest of Drone Drifter

Someone landing from Google did not know about the small-town flight logs, airspace lessons, courthouse stories or field decisions behind the footage.

The guide needed one short bridge, not a sales speech:

Drone Drifter follows what happens before and after the shot: practical flight planning, small-town history and the decisions that turn drone footage into a story.

That tells a new reader what remains after eclipse day.

Mobile readers had to work too hard

Seventy-four percent of the audience arrived on mobile.

Several section labels were bold paragraphs instead of real headings. Long checklists were harder to scan than they needed to be. The information was present, but presence is not enough on a phone.

A mobile reader should be able to find camera safety, settings, airspace and battery strategy with a thumb and half a second of attention.

I measured promotion badly

The direct bucket was 34 percent. That is too large to shrug off and too vague to learn from.

Future X posts, Notes and email links need separate tracking. Otherwise a successful promotion disappears into “direct” and I end up guessing which platform did the work.

Event traffic borrowed urgency from the calendar

The date created demand I did not own.

That does not make the traffic worthless. It means the follow-up had to be ready before the interest disappeared. I published the guide, promoted it and only then started thinking about the permanent version.

Next time, the afterword and evergreen update get planned with the original article.

What I will change without damaging what worked

I am not rebuilding a page that Google already found. Changes should be controlled.

1. Keep the existing title and URL.

2. Turn every major section into a real H2 and every subsection into an H3.

3. Convert settings, mistakes and preflight items into actual lists.

4. Add two or three internal links where they naturally answer the next question.

5. Use one strong subscription invitation near the end, not buttons scattered through the article.

6. Make that invitation explain what Drone Drifter publishes after the event passes.

7. Use tracked links for X, Notes and future email promotion.

8. Update time-sensitive details while keeping the permanent safety and planning sections intact.

One CTA is enough if it says something.

My event-article flight plan

This is the repeatable part. Eclipse, launch, demolition, storm front, festival or courthouse reopening—the date changes, but the work around it does not.

Before writing

• Name the exact question a reader will type.

• Identify the decision only field experience can improve.

• Separate hard safety facts from personal judgment.

• Choose the permanent lesson that survives the date.

• Decide what the reader should open next.

Before publishing

• Put the event, date and practical promise in the title.

• Write an SEO description that finishes the promise without repeating the title.

• Set a clean URL before publication.

• Use real headings and short lists for mobile readers.

• Add one internal link to the next useful field note.

• Prepare separate tracked links for Notes, X and email.

• Write one specific subscription invitation.

• Draft the post-event update before the event happens.

During the first 48 hours

• Record total views.

• Record search, direct, Substack and email shares.

• Record mobile versus desktop traffic.

• Record opens, clicks, likes, comments and restacks.

• Record free and paid subscribers attributed to the post.

• Save the snapshot before changing the title, CTA or structure.

After the event

• Remove expired urgency.

• Keep safety guidance and reusable planning sections.

• Update the introduction for the next event.

• Link the original guide to the afterword without creating two pages that chase the same search.

• Compare subscriber conversion before and after the revision.

Copyable brief for the next timed story

Event:

Publication deadline:

Reader’s immediate question:

Decision this article helps them make:

Safety or legal risk that needs a primary source:

Field detail only I can add:

What remains useful after the date passes:

Article the reader should open next:

One-sentence subscription promise:

Tracked Notes link:

Tracked X link:

Tracked email link:

First 48-hour snapshot:

Seven-day snapshot:

Post-event revision date:

What counts as a win next time

I am not choosing a conversion percentage because it looks good in a marketing article. I do not have enough data for that.

The next test has a simpler job: produce a measurable, nonzero subscriber result and preserve the search traffic.

One conversion gives me a baseline. More than one gives me something to compare. Zero tells me the bridge is still missing.

The eclipse guide was not a failure. It answered real questions. It may have kept someone from treating an ND filter like solar protection. It showed that a small publication can reach well beyond its subscriber list when the page is useful and the title is clear.

But attention is not an audience.

Seven hundred fifty-eight people came through the door. I showed them the eclipse and forgot to show them what came next.

This free case study is the correction.

The numbers are open. So are the wrong turns and the field plan.

Related work

• How to Film the August 12, 2026 Solar Eclipse with a Drone

Analytics note: Figures above come from the Substack publisher dashboard snapshot taken August 12, 2026, roughly 48 hours after publication. Percentages may not sum to 100 because the dashboard groups devices and sources separately and rounds displayed values.

Subscribe to Drone Drifter for field-tested drone guides, flight logs, small-town history and honest after-action reports. I will show you the flight plan and the result.

I load the vehicle, point it toward another state and spend more hours watching highway lines than I care to count. The drone rides with me because work gets me to the place, but the camera gives me a reason to look around once I am there.

That became Drone Drifter.

Why I carry a drone

I did not start traveling across the USA with a drone because I had a perfect route mapped out. There was no production crew waiting at the next exit. I had work, a little down time and an aircraft small enough to fit beside everything else.

The first flights changed how I looked at a stop.

A courthouse square can feel ordinary from the sidewalk. From above, the streets show why the town grew where it did. A church that disappears behind trees from the road suddenly owns the hill. Old factories, rivers, farm fields and empty railroad lines start explaining one another.

The drone slows me down.

Places between the destinations

Big landmarks are easy to find. Search engines have already told everybody where to stand.

I keep getting pulled toward the places in between. Decatur, Texas. Galena, Illinois. A marina on Lemon Bay. An Orthodox church far from the part of the world where its story began. A restaurant at the end of a long workday. Sometimes a place earns a full article. Sometimes the clip is the whole story.

That is fine.

I want Drone Drifter to feel like the road really feels. Some days give me clean light and enough time for three batteries. Other days give me wind, a locked gate and ten minutes before I need to leave. The camera does not owe me a shot because I drove a long way.

Flying around work

The job comes first. That rule keeps the travel honest.

I fly when the schedule, airspace and location allow it. I check weather, temporary flight restrictions and nearby airports before launch. If people are moving through the area, the flight path changes or the drone stays in its case.

There are days when I see a place from the highway and know the light is right. I also know I cannot stop. That one hurts.

Another town eventually appears.

What I am trying to show

Drone footage can make every place look polished. Smooth movement, music and a sunset can hide the cracked parking lot just outside the frame.

I am more interested in the place itself.

I want the old brick beside the new roof. I want the water tower, church dome, courthouse clock and road that keeps trucks moving through town. If a building is weathered, let it be weathered. That is part of why I stopped.

The best flights give people another view of somewhere they already know. A local sees the roofline differently. Somebody who moved away remembers the street. A stranger adds the place to a future drive.

That is enough reason to keep launching.

What you will find here

Flight Log holds the road stops and drone views. These are the towns, churches, restaurants, waterfronts and odd places that made me pull over.

Airspace covers rules, equipment and policy when those things affect the aircraft in my case. I read the documents because a loud headline can make a legal drone sound illegal by lunch.

The essays sit closer to the person behind the controller. Travel changes what I notice. Faith, fatigue, work and long roads come with me whether they appear in the shot or not.

I am not chasing one narrow version of drone content. The drone is the thread. The road decides what it runs through next.

Why subscribe

Social platforms are good at showing a clip for eight seconds and burying it by dinner. This publication gives the flight a home.

Subscribers get the story behind the shot, the route, what worked and what went wrong. You will also get practical drone guides when rules or equipment change. No launch-day shouting. No copied press release pretending to be a field report.

If that sounds like your kind of trip, come along.

Where should I stop next?

Tell me about the place people pass without seeing. Give me the town square, church, roadside restaurant, strange monument or piece of water that looks different at sunrise.

I may already be heading that direction for work.

Subscribe to Drone Drifter:

https://www.dronedrifter.com/subscribe

The drone in my case feels modern. Its family tree reaches back more than a century.

Britain tested a small radio-controlled aircraft called the Aerial Target in March 1917. Dr. Archibald Low designed its remote-control equipment.

The first controlled flight

Imperial War Museums describes the Aerial Target as the first drone to fly under control. The United States followed with the Kettering Bug in 1918. Neither aircraft entered operational service during World War I, but both proved that a pilotless aircraft could leave the ground.

Where the word drone came from

Britain built radio-controlled DH.82B Queen Bee aircraft in the 1930s as targets for anti-aircraft training. The Imperial War Museums and Royal Air Force Museum both connect the word drone to the Queen Bee name.

The Queen Bee gave gunners a moving target without putting a pilot in the line of fire. Its spruce-and-plywood fuselage could float after the aircraft came down in water.

The United States watched. By 1939, the Radioplane OQ-2 had become the first mass-produced UAV in America. The factory supplied another strange footnote. Photographer David Conover met Norma Jeane Dougherty while she worked at Radioplane. Those photographs helped start the career of Marilyn Monroe.

From targets to cameras

For decades, unmanned aircraft stayed tied to target practice, surveillance and war. The aircraft grew larger. Radio links improved. Autopilots stopped behaving like science-fair projects.

Then cameras, batteries, GPS receivers and small electric motors shrank.

That brought the technology off military runways and into cases carried by photographers, roof inspectors, farmers, firefighters and people like me who pull off a road because a town looks different from above.

The history is not a straight line. It is a shelf full of odd machines, failed tests and useful ideas that waited years for the rest of the hardware to catch up.

Large drones changed civilian work

NASA helped move large unmanned aircraft into science work. Its Global Hawk aircraft flew high-altitude missions that could last about 24 hours and cover roughly 8,500 nautical miles. Researchers used unmanned aircraft to study storms, wildfires and the atmosphere without putting a crew inside a long, risky flight.

That scale feels far away from a folding camera drone. The same bones remain: a ground station, a radio link, navigation, sensors and an aircraft doing work where a person does not need to sit.

The small-drone turn

Consumer drones became useful when several technologies matured together. Brushless motors became reliable. Lithium batteries carried enough energy. GPS held position. Tiny inertial sensors kept the aircraft level, and digital cameras stopped looking like security footage.

Flying changed.

Older radio-control pilots had to manage the aircraft every second. A modern camera drone can hover while I frame a courthouse or wait for a car to leave the shot. The software carries part of the workload. It does not carry responsibility.

That distinction matters. A hundred years of engineering can hold the drone in place. The pilot still decides whether it should be there.

Sources

Imperial War Museums, A Brief History of Drones:

https://www.iwm.org.uk/history/a-brief-history-of-drones

Smithsonian National Air and Space Museum, The Secret History of Drones:

https://airandspace.si.edu/air-and-space-quarterly/issue-12/secret-history-of-drones

NASA Global Hawk:

https://www.nasa.gov/aeronautics/global-hawk-uas/

If you want more drone history mixed with what I see from the road, subscribe. Old machines explain more about the one in my case than most product launches do.

https://www.dronedrifter.com/subscribe

This was a quick drone pass over Lighthouse Grill and Stump Pass Marina in Englewood, Florida. The water on Lemon Bay was flat. The big American flag had enough wind to keep moving, and the tiki roof gave the frame its color.

A small place from the road. A different place from the air.

What caught my eye

The marina sits in a narrow piece of Florida where almost everything points back to the water. Boats stack into dry storage. Docks run toward Lemon Bay. The Lighthouse Grill sits under its namesake tower with the flag high enough to pull your eye into the shot.

From ground level, you see a restaurant and a working marina. The drone shows how close the whole property sits to the bay. One turn of the aircraft catches the roofline, dock slips and open water in the same frame.

That is why I stopped.

Where this is

Lighthouse Grill and Stump Pass Marina are at 260 Maryland Avenue in Englewood. Boaters know the location as Channel Marker 17A on Lemon Bay. Drivers reach it from Placida Road by way of Maryland Avenue.

The marina connects boat traffic on Lemon Bay with the Intracoastal Waterway and the Gulf. Stump Pass Beach State Park sits nearby at the south end of Manasota Key, where the bay side runs into mangroves and the Gulf side opens into white sand.

The restaurant and marina are their own stop. The state park is a separate destination. Both make more sense once you see the water around them.

A good stop by road or boat

This is one of those Florida places built around arrival. You can come off the road, walk under the tiki roof and watch boats work through the marina. You can also arrive from Lemon Bay and tie the afternoon to a meal.

The official Lighthouse Grill page lists live music nightly and gives the boat location as Channel Marker 17A. The marina describes dry storage, a fuel dock, ship store and waterfront dining. It is not a staged waterfront made for a brochure. Boats still need fuel. Forklifts still move hulls. People are coming and going.

The camera likes that movement.

Flying the shoreline

Waterfront flying needs patience. Flags tell me what the wind is doing above the roofline. Boat traffic changes the shot without warning. People move through the docks, and a clean composition can become a bad flight path in seconds.

I keep the aircraft away from people and active boats. I check airspace, weather and temporary restrictions before launch. Wildlife gets room too. Manatees and coastal birds use this part of Florida; a low drone is a visitor, not the main event.

The short pass worked because I did not ask it to become a movie. The water stayed quiet. The flag moved. The marina filled the frame.

That was enough.

Plan your stop

Lighthouse Grill:

https://www.lighthouse-grill.com/location/lighthouse-grill/

Stump Pass Marina:

https://stumppassmarina.com/

Stump Pass Beach State Park:

https://www.floridastateparks.org/parks-and-trails/stump-pass-beach-state-park

If you like seeing the places most people pass from the road or pull up to by boat, subscribe. More flights are waiting down the road.

https://www.dronedrifter.com/subscribe

Time fixed some of that. I am fixing the rest.

The road has cured me of spec-sheet fever. A drone that takes ten minutes to unpack stays in the case when the light is disappearing. A smaller aircraft that launches in two minutes gets the shot. Battery condition, legal airspace and whether I can replace a prop in the next town matter more than another line on a feature chart.

What changed after 2025

Buying a drone in 2026 takes more homework than it did a year ago. The aircraft may be sitting on a store shelf while its app support, FCC authorization or replacement-battery supply tells a different story.

DJI still has a huge installed base, but new foreign-produced models face FCC equipment-authorization limits. Existing authorized DJI drones can still be used, marketed, imported and sold. I covered that change and the August 2026 status here:

That makes the exact model important. “DJI drone” is too broad. Check the model number, its FCC authorization and whether the seller is offering a U.S. unit or a gray-market import. A low price can become expensive once the controller will not activate, warranty service disappears or batteries are impossible to find.

Buy for the flight you will make

Start with four questions.

Where will you fly most often? What will you carry on foot? Do you need a vertical social-media frame or a camera built for grading later? Will the drone earn money?

My travel kit needs to fit beside everything else in the vehicle. I want fast setup, batteries that do not eat the whole charger and a controller I can read in Georgia sun. Someone inspecting roofs needs different hardware. A real-estate pilot may care more about repeatable moves and fast delivery. A mapper needs positioning, payload support and long days of battery management.

Write down the work before opening a product page. The answer usually removes half the choices.

Rules belong in the buying decision

A sub-250-gram drone is not a blanket exemption from FAA rules. The registration break applies when the aircraft weighs 0.55 pounds or less and is flown only under the recreational exception. Every drone flown under Part 107 must be registered, regardless of weight.

Registered drones must comply with Remote ID unless they are flown under a permitted exception, such as inside a FRIA. Recreational pilots also need to pass TRUST and carry proof while flying. Controlled airspace may require authorization through LAANC or DroneZone.

These are buying details. A Remote ID module adds cost and another battery. Prop guards, lights and accessories can push a small aircraft over 250 grams. A “no registration” sales pitch can fall apart as soon as the flight supports a business.

My used-drone check

Used drones can be a good buy. I check four things before money changes hands.

First, I ask the seller to power up the aircraft, controller and every battery. Swollen packs walk away with the deal.

Second, I inspect the gimbal during startup. It should finish its movement without grinding, shaking or parking crooked.

Third, I confirm the exact model, serial number, Remote ID status and account-binding status. A drone still tied to somebody else’s account can turn a bargain into a desk ornament.

Fourth, I price replacement props and batteries before I price the drone. An older aircraft with two dying batteries is not cheaper than a current kit.

I also look at the case. Dust, bent latches and props thrown loose in a compartment tell their own story. Owners who care for the small pieces usually care for the aircraft.

DJI and the FCC question

I would not ground an authorized DJI drone because of a headline. The FCC action is prospective. It blocks covered new equipment authorizations; it did not make previously authorized aircraft illegal to fly.

I would check before buying a newly released model. I would also keep proof of purchase, firmware records and the FCC information tied to that exact aircraft. Regulation can move faster than a retailer changes a product page.

For pilots who depend on one platform for paid work, a second aircraft deserves thought. It does not need to be a political purchase. It needs to finish the job when a battery, app update, controller or supply chain refuses to cooperate.

What I would carry today

For road travel, I still favor the smallest aircraft that can handle the wind and camera work I expect. The drone has to leave the case. Portability wins a lot of real flights.

For paid work, I would buy around the deliverable and keep room in the budget for batteries, insurance, storage cards, props and a controller hood. The aircraft price is the first receipt, not the full cost.

Do not buy a rumor. Do not buy a model because a countdown video made the box look heroic. Buy the aircraft you can legally fly, service and replace.

Sources

FAA drone registration:

https://www.faa.gov/uas/getting_started/register_drone

FAA Remote ID:

https://www.faa.gov/uas/getting_started/remote_id

FAA TRUST information:

https://www.faa.gov/uas/recreational_flyers/knowledge_test_updates

FCC Covered List update:

https://docs.fcc.gov/public/attachments/DA-26-673A1.pdf

If you want more road-tested drone notes without the launch-day noise, subscribe. I will keep writing about what survives the first battery, the first windy field and the first long drive.

https://www.dronedrifter.com/subscribe

This is where the DJI drone ban stands as of August 13, 2026. Here is what that means in practical terms.

On December 22, 2025, the FCC added foreign-produced drones and critical drone components to its Covered List. DJI and Autel equipment named in Section 1709 of the 2025 defense law were included.

The FCC Covered List controls equipment authorizations for new models. The authorized drone already in your case remains usable.

Previously authorized DJI drones can still be used. Models that already received FCC equipment authorization may also continue to be marketed, imported and sold under that authorization. The restriction hits new equipment authorizations. A new covered model cannot legally enter the U.S. market without one.

That difference matters. Your Mini, Mavic, Air, Avata or Matrice did not become illegal to fly because the calendar turned over. FAA rules still apply exactly as they did before: registration when required, Remote ID, airspace authorization and the operating rules tied to your flight.

The harder question sits on the store shelf. Existing authorized models can remain available, but the pipeline for new DJI models is blocked unless the equipment qualifies for an exemption or the FCC changes its position. That can narrow selection and keep older hardware on sale longer than pilots expect.

The FCC has carved out several categories. Certain Blue UAS systems and products that qualify as domestic end products received temporary treatment through January 1, 2027. Foreign-produced toy drones were removed from the Covered List in June 2026. Those changes did not clear DJI’s consumer and professional aircraft.

DJI challenged the decision on January 21, 2026. The FCC opened ET Docket 26-22 and continued taking public input after DJI submitted a security assessment covering two drone systems, controllers and applications. Comments on that filing are due August 28, 2026.

I still think the U.S. needs real competition before regulators close the door on the equipment pilots already depend on. A policy can push domestic manufacturing without pretending a replacement fleet appears overnight. Farmers, roof inspectors, filmmakers and small drone businesses buy what works at a price they can carry.

For now, I am not grounding my authorized DJI drones. I am checking FCC authorization before buying a newly released model, keeping firmware and flight records in order, and watching ET Docket 26-22 instead of repeating headlines that leave out the part about existing aircraft.

That is the practical answer. Keep flying legally. Check the exact model before you buy. Read the FCC documents yourself.

Sources:

FCC Covered List, updated July 7, 2026: https://docs.fcc.gov/public/attachments/DA-26-673A1.pdf

FCC fact sheet on foreign-produced UAS: https://docs.fcc.gov/public/attachments/DOC-416839A1.pdf

FCC notice on DJI’s challenge and August 28 comment deadline: https://docs.fcc.gov/public/attachments/DA-26-592A1.pdf

If you want the next policy update without the panic language, subscribe. I will keep reading the orders before I tell you what they mean.

Subscribe now

If you’re planning to film the August 12, 2026 solar eclipse with a drone, the most important decision isn’t your camera settings. It’s where you point the camera.

Are you filming the Sun... or what the eclipse does to the landscape?

On August 12, totality crosses Greenland, Iceland, northern Russia, the Atlantic, Spain and western Portugal, with less than two minutes in most places and about 2 minutes 18 seconds at the maximum. Parts of the northern United States and much of Europe will see a partial eclipse.

Once the drone is up, nothing you forgot gets easier.

The short version: use the drone for the land. Keep the Sun out of frame unless you have a photographic solar filter made to fit that camera safely. An ND filter doesn’t become a solar filter because the number printed on it is big. Lock the white balance, test the exposure and launch early enough that you aren’t still moving when the light falls out of the sky.

Before You Leave Home

• Batteries charged

• Firmware updated

• Airspace checked

• Solar filter packed

• Weather checked

• Backup landing zone identified

• White balance locked

• Manual exposure tested

• Extra SD card packed

• Observer lined up, if available

Plan the eclipse drone flight before the Sun is up

There’s a part of the morning when the whole place belongs to you for about twenty minutes.

No glare yet. No bad decisions yet. The sky hasn’t decided what it’s going to do.

Walk the takeoff point, then walk the backup landing zone too, because the first one always looks better from the car than it does twenty minutes later with a drone coming home on 22 percent and three people standing where you planned to put it down. Check wind at the height you intend to fly. Trees at your feet can make a windy ridge look calm.

In Spain, totality comes late in the evening, shortly before sunset. A clear view at noon proves nothing. Check the Sun’s altitude and direction for your exact spot and exact minute. Then go there at roughly the same time on another day.

Before eclipse day, I want four things settled

• One legal takeoff point and a landing zone I can reach without crossing a crowd

• The route to the shot, including the gimbal angle at the end of it

• Camera settings saved as a note or screenshot

• A return point that leaves real battery in reserve

Best Places to Film the August 12, 2026 Solar Eclipse

Buford, Georgia, sits outside the partial eclipse, so this one has to be a road trip. Unfortunately, third shift has other ideas.

The word best changes with the person carrying the gear. Duration points toward Greenland and the North Atlantic. A higher Sun points toward Iceland. For a place you can reach, scout and find a road out of, I would look hardest at inland northern Spain.

Inland northern Spain

León gets totality from about 8:28 to 8:30 p.m. local time. Zaragoza gets roughly a minute around 8:29 p.m. Both sit under the path, with roads, lodging and open country nearby.

Spain’s weather agency, AEMET, studied cloud cover for the same August dates and evening hours from 2010 through 2025. Inland Castilla y León and the Ebro valley had clear or lightly clouded skies about 50 to 70 percent of the time. The Cantabrian coast, including northern Galicia, Asturias, Cantabria and the Basque Country, came in closer to 30 to 50 percent.

Those are old skies, not a promise for Wednesday.

The best climate record in the world still loses to tomorrow morning’s satellite image. Let history narrow your search. Let the forecast choose your field.

The practical choice is open ground inland with a clear west-northwest horizon, outside town crowds and outside any drone geo-zone. Spain’s Sun will be low. One ridge in the wrong place can eat the whole two minutes.

Western Iceland

Reykjavík enters totality around 5:48 p.m. and comes out about a minute later. The Sun sits higher than it does in Spain, which gives buildings and mountains less chance to block it. Icelandic cloud can end the argument before it starts, so I would keep more than one location available and make the final move from the short forecast.

Western Iceland also asks more from a drone pilot. Wind, protected land and fast-changing visibility can turn a beautiful launch point into a bad one before the battery is warm.

Greenland, northern Russia and the North Atlantic

The longest totality, up to about 2 minutes 18 seconds, falls near the middle of the eclipse path across Greenland, northern Russia and the North Atlantic. Access gets hard quickly. Roads disappear. Weather support thins out and a drone becomes another piece of equipment that has to survive the trip.

For an expedition already going there, the duration is hard to beat. For a first eclipse flight, Spain gives you more ways to recover when the plan bends.

Partial eclipse locations

Most of Europe sees at least a partial eclipse. London gets about 91 percent coverage, Dublin about 94 percent and Madrid about 99 percent, though Madrid sits just outside totality. In the United States, the partial phase reaches parts of the North, from Alaska across northern states and into North Carolina.

Wherever I went, I would check the official forecast the night before and again that morning. Climate maps choose a region. The last satellite image chooses the field.

Can drones film a solar eclipse safely?

Yes. The margin gets thin when you point the camera at the Sun.

We don’t launch into the Sun. We never have.

DJI warns against leaving the Mini 5 Pro camera pointed at an intense light source such as the Sun because it can damage the sensor. That warning matters during the partial phases, when the uncovered part of the Sun is still hard, direct light.

An ND1000 is an exposure tool. A solar filter is safety equipment. Keep those two sentences separate.

If the Sun will be in frame during a partial phase, use a purpose-made photographic solar filter over the front of the lens, fitted to that exact camera and attached without loading the gimbal, fouling its movement or changing the drone’s balance. Eclipse glasses taped over the lens don’t count. Neither does a stack of ordinary ND filters.

Can’t find a proper solar filter made for the drone? Put the solar camera on the ground. The drone can face away from the Sun and film the changing light across the land.

That’s the shot I would choose anyway.

A Mini 5 Pro carries a wide camera, so the Sun will be a small coin in a large frame while a long lens on the ground does the solar disc better and the drone catches what can’t be seen from a tripod: water going dark, streetlights waking up and a line of shadow moving across open country.

One more distinction matters. ISO 12312-2 covers filters made for direct solar viewing with your eyes. It isn’t a camera-filter certification. Wear compliant eclipse glasses when looking at the partial eclipse and buy a solar filter intended for photographic optics when pointing a camera at it.

Best drone solar eclipse camera settings

Settings lists get confident fast. ISO 100. Shutter at 1/1000. Focus at infinity. Done.

Except those settings mix two different shots.

For wide video of the ground, I would start here:

• Record 4K at 24 or 30 frames per second.

• Start at ISO 100. Raise it only when the ground needs it.

• Try 1/50 at 24 fps or 1/60 at 30 fps for normal-looking motion.

• Lock the white balance before launch.

• Focus on distant ground, check it, then lock focus if the camera allows it.

• Turn on the histogram and overexposure warning.

Manual exposure records the fall of light. Auto exposure fights it. If I want the eclipse to look like an eclipse, I use manual and accept that the ground will get dark. Someone who needs detail in the buildings below may prefer a controlled auto setting. Make that choice during rehearsal, when nobody is counting down in your ear.

A shutter speed of 1/1000 or faster can work for a still photograph of the Sun through a proper solar filter. It makes poor blanket advice for drone video. Motion gets sharp and twitchy, and the ground can disappear into black.

Direct solar photographs need their own test, with the correct filter on the camera while the drone is still on the ground, pointed at the uneclipsed Sun long enough to check the histogram and bracket the exposure if the aircraft supports it. Filter strength and cloud cover decide the final numbers.

Don’t discover them at launch.

Where to position a drone for eclipse footage

The sky was doing something orange and violent behind them.

The shot was the ground.

Coastlines are good because moving light shows up on water. Open fields give the shadow room to travel. A town gives you windows and streetlights. Church domes, towers and ridgelines hold the last light in a way flat ground doesn’t.

Altitude helps until it doesn’t. Go too high and the foreground flattens out. The frame loses the one thing that tells a viewer how big the event felt.

Spain and Portugal add a low evening Sun to the problem, which means the trees, roofs and ridges you ignored at noon may matter more than another fifty metres of altitude when totality arrives shortly before sunset. There is no useful Europe-wide 10-degree rule. The angle changes with the location and minute. Check yours.

I aim a little off-axis and give the light somewhere to move. The Sun doesn’t need to sit in the middle of the frame like a target.

Battery strategy for the August 12 solar eclipse

The eclipse runs for hours. The Mini 5 Pro doesn’t.

DJI rates its standard battery for up to 36 minutes under test conditions, a number that starts shrinking as soon as wind, cold, repositioning, battery age and the reserve you should have kept all along enter the flight. The useful number is the one in your own flight log from a similar day.

Launch eight to ten minutes before totality or maximum eclipse. Get into position, stop spending battery on pretty circles and let the aircraft sit where you planned it.

I keep a second battery ready. I don’t build the main shot around changing it. Two minutes is too short for that kind of optimism.

Set the return point before takeoff. Land when the shot is finished, while you still have choices. A low-battery warning is a lousy director.

Solar eclipse drone laws, airspace and crowds

Check the airspace before you charge anything. Check it again near flight time.

Most of the path of totality is in Europe, where EASA open-category flights generally have to remain in visual line of sight, stay below 120 metres from the nearest point of the surface and avoid assemblies of people, with each country free to add its own geo-zones and local limits.

The United States gets a partial eclipse. FAA rules still require visual line of sight, and many flights over people aren’t allowed. Temporary Flight Restrictions appear through NOTAMs and can include drones.

An eclipse crowd is still a crowd. A sub-250-gram aircraft doesn’t give you a pass to hang over it.

Use an observer if you can. Totality pulls your eyes upward. The controller pulls them down. Someone needs to keep watching the aircraft and the air around it while the world turns the wrong color.

What I actually fly

I carry a DJI Mini 5 Pro most places, with a Neo for low work. That’s enough.

For this flight, I would put the Mini 5 Pro in the main slot. Its camera controls and longer battery give me more room. The Neo can grab a low establishing shot before or after the eclipse. I wouldn’t make it responsible for the two-minute window.

Heavy gear won’t rescue a bad position. The road teaches that lesson every week. Put the camera in the right place, know what it’s pointed at and be ready before the light goes.

Common mistakes when filming a solar eclipse with a drone

• Treating an ordinary ND filter as a solar filter

• Pointing bare glass at the Sun during a partial phase

• Copying settings from a different camera and filter

• Letting auto white balance erase the color change

• Flying so high that the foreground loses its shape

• Staring at the screen and losing visual line of sight

• Planning around the battery time printed on the box

• Launching over people because everyone else is looking up

The machinery recurs. People burn sensors the same way every eclipse, thinking one quick shot won’t matter.

It does.

One flight, then the light is gone

Restraint wins this one.

Do less than you think.

Point away from the obvious.

Let the land carry the story while the sky changes above it.

Find the spot early.

Hold the route.

Leave the white balance alone.

When the light begins to fade, let it fade.

Two minutes later the color starts coming back.

The crowd cheers.

Streetlights blink off.

Your props slow to a stop over a landscape that looks almost ordinary again.

Almost.

Because the best eclipse footage isn’t really about the Sun.

It’s about what the changing light revealed about the world beneath it.

Frequently asked questions

Do I need a solar filter for drone eclipse filming?

You need a purpose-made photographic solar filter whenever the Sun is in the camera frame during a partial phase. If the drone is filming the ground with the Sun safely out of frame, you don’t need one.

Is an ND1000 filter enough for a solar eclipse?

Don’t assume it is. Standard ND filters control exposure. They aren’t necessarily made to reduce solar energy safely across the required wavelengths. Use a filter sold for solar photography.

What are the best drone solar photography settings?

For ground footage, start with ISO 100, locked white balance and manual exposure. At 24 fps, try 1/50. At 30 fps, try 1/60. Test at the location before eclipse day. A filtered solar-disc photograph needs separate settings based on the filter.

How long is totality on August 12, 2026?

The maximum is about 2 minutes 18 seconds. Most locations inside the path get less than two minutes. Your exact time depends on where you stand.

How long should the drone stay in the air?

Launch roughly eight to ten minutes before totality or maximum eclipse and land after the planned shot, with a conservative reserve. Use the flight time from your own logs rather than the manufacturer’s test maximum.

Where can I practice?

Practice at the real location and close to the same time of day. A sunset run will show the low solar angle in Spain or Portugal. It won’t copy the sudden darkness of totality, so test exposure changes separately.

Sources and safety references

-NASA: Total Solar Eclipse on August 12, 2026

- DJI Mini 5 Pro User Manual

- Nikon: How to Photograph a Solar Eclipse

- ISO: What ISO 12312-2 Covers

- EASA: Understanding the Open Category

- FAA: Temporary Flight Restrictions for Drones

Either name fits. It is a 30-foot human eye planted on a lawn in downtown Dallas, surrounded by buildings that seem almost ordinary beside it. The blue iris belongs to artist Tony Tasset. The red veins spread across its surface. From the street, the sculpture looks like something that rolled loose from one of the skyscrapers above and came to rest on Main Street.

I went there to film it from the air.

The evening began with dinner alongside my creative group. Afterward, I grabbed my drone and walked over to The Eye. The first few minutes went exactly as planned. I launched, established my position and started collecting footage—left pan, right pan, gimbal up, gimbal down.

Fan-freaking-tastic.

Then I tried descending while raising the gimbal. I held the stick down a beat too long, and the drone did exactly what I had unintentionally told it to do.

It landed.

Right on the lawn.

Behind the fence.

An Eye With a Past

Before I explain how I got the drone back, The Eye deserves a proper introduction.

Tony Tasset created the sculpture as a monumental replica of his own eye. It is made from fiberglass, resin and steel, and its scale turns a familiar part of the human body into something surreal.

The sculpture made its public debut in Chicago’s Pritzker Park in 2010 after being commissioned for the Chicago Loop Alliance’s Art Loop program. Three years later, it arrived at its current home across from The Joule hotel in downtown Dallas.

Tasset has connected the image with everything from the eye of a pharaoh and the Masonic eye to Big Brother, God and human consciousness. He has also resisted assigning it one fixed interpretation. The work can mean whatever the person standing before it brings to the encounter.

That openness is part of its power. Humanity has attached ideas of knowledge, authority, protection and surveillance to the eye for thousands of years. Tasset simply made one too large to ignore.

It has since become one of the most recognizable sights on Main Street. People photograph it, practice yoga beneath its gaze and even hold wedding receptions on the surrounding lawn. The Dallas Art Fair’s closing celebration is appropriately called the Eye Ball.

On the night I filmed it, however, none of that was on my mind.

My drone was stranded on the other side of the fence.

“Re-calibrate Compass”

I pressed the launch control.

Nothing.

I tried again.

Still nothing.

Then a message appeared across the top of my screen:

Re-calibrate compass.

Shit.

At first, I thought the downtown buildings and trees might have blocked the connection. I moved closer, hoping the signal would return. It didn’t. The drone wanted a compass recalibration, and it could not perform one while sitting motionless on the grass.

There was only one solution: I had to physically retrieve it.

I found a man at a nearby bar and asked how someone might get over the fence to recover a stranded drone. He told me that The Joule owned the property.

Fine.

I walked next door, found a security guard and explained exactly what had happened.

“My drone landed on your property. Can I go get it?”

He looked at me and said, “Oh. I was wondering what that was.”

Thankfully, he let me retrieve it.

Relief does not quite describe that moment. The drone was hot when I picked it up. I shut down both the aircraft and controller and decided I had carried enough gear through downtown Dallas for one night.

That should have been the end of the story.

It wasn’t.

The Scooter That Wouldn’t Unlock

I found a Lime scooter and tried to rent it for the trip back. I scanned it and thought the rental had started, but the scooter refused to unlock. After another attempt, the app finally explained why: I was outside its operating hours.

Great.

I left the scooter and started walking. A few steps later, my chest went cold.

My drone bag.

I had walked away from it.

I turned around and moved quickly toward the unattended bag. A man nearby noticed me approaching it with purpose. A police officer had apparently been watching it too. When he realized it belonged to me, he said he was glad I had returned because someone had left a drone bag sitting there and nobody knew whose it was.

That someone was me.

My fellow creators stayed with me through the entire ordeal, and we walked back to the hotel together. It was only about ten minutes away. I was full from dinner, carrying all my gear again and soaked through with nervous sweat.

The finished footage looks calm.

The person operating the drone was anything but.

What the Flight Taught Me

Every expedition teaches me something. This one delivered two lessons in quick succession.

First, dense urban surroundings can complicate a flight. Concrete, steel, trees and restricted access leave little room for correcting a mistake. A landing that would be a minor inconvenience in an open field becomes a recovery operation when it happens behind a fence.

Second, small control inputs matter. Hold the descent command a little too long and the aircraft may land regardless of what you intended.

That was the entire education: one hot drone, one scooter that wouldn’t unlock, one temporarily abandoned bag and a shirt soaked through by the time I reached the hotel.

Cheap tuition—if you are the one reading about it instead of living it.

The Eye itself remained perfectly still throughout the whole episode. Fifteen feet of blue iris, fixed on the lawn, watching me turn a short filming session into a downtown recovery mission.

Maybe that is what the sculpture means to me now.

It saw everything.

Another landmark flight from the road: Mount Vernon From Above.

If you like field notes where the footage looks calmer than the flight, subscribe. I have more stories from the air coming.

Subscribe now

If you like firsthand flight stories and the history behind the places, subscribe. More films and field notes are coming.

I didn’t arrive in Mount Vernon with a complicated plan. I wanted to fly, look around and see what caught my attention.

Subscribe now

The weather made the decision easy. It was mild, not too hot, and the wind was sitting in that comfortable range where the drone feels steady without the air feeling dead.

So I launched the mini 5 pro near the Fifth District Appellate Courthouse and started watching the town move beneath me. Cars passed through the intersections. People continued with their day. I was enjoying myself, and one flight naturally became another.

Watch DD-0001 below, then continue reading to discover the stories behind the places in the film.

A City at the Crossroads

The traffic is what first held my attention. Mount Vernon is not a large city, but it has always been a crossroads. Zadok Casey founded it in 1817, and the historic Vandalia Road helped turn the settlement into a stopping point for travelers and commerce across southern Illinois.

Jefferson County was created two years later, and Mount Vernon became the county seat. During the nineteenth century, the city also served the southern division of the Illinois Supreme Court. That brought lawyers from around the state, including Abraham Lincoln, to argue cases here.

You can drive through town without thinking about any of that. From the air, the older landmarks interrupt the ordinary pattern of streets, roofs and traffic. That is what made me stop and keep filming.

The Fifth District Appellate Courthouse

The first landmark in the film is the Fifth District Appellate Courthouse at 14th and Main Streets.

This is not the current Jefferson County Courthouse on South 10th Street. It is the older white Greek Revival building with the raised courtroom and two exterior stairways. Once you see those stairs, there is no mistaking it.

I made several passes around it. From ground level, trees and neighboring buildings break up the view. Once the drone climbed, the symmetry became obvious.

The courthouse was completed in 1856 for the southern division of the Illinois Supreme Court. The courtroom occupied the upper floor, reached by those twin stairways.

Before he became president, Abraham Lincoln argued and won a tax case in Mount Vernon. I like that the connection is not trapped behind museum glass. The building still works. The Fifth District Appellate Court continues to hear cases there.

The white façade catches the light. Cars continue along Main Street. The courthouse stays planted in the middle of it all, nearly two centuries of Illinois legal history surrounded by an ordinary weekday.

That contrast was enough for me. I kept circling.

Veterans Memorial Park

After the courthouse, I moved over to Veterans Memorial Park at 800 South 27th Street.

The park wraps around a six-acre lake with fountains, islands, footbridges and a gazebo. A one-mile trail follows the water. Playgrounds, courts and shelters fill the rest of the grounds.

From above, it looked calm. The history underneath it was not what I expected.

The grounds once held a small city zoo with lions, monkeys and zebras. They were also home to the Mount Vernon Kings, a minor-league baseball team whose park hosted league All-Star Games in 1949 and 1951. The club played its final season in 1954.

The park carries a different kind of history today.

At the entrance stands the Veterans Memorial, dedicated on December 7, 2022. Bricks along the approach record the names, ranks and military service of individual veterans.

A few minutes away, children can play, people can walk the lake and traffic can continue beyond the trees. The memorial is not separated from everyday life. It sits inside it.

That was what I wanted the second half of the film to hold: remembrance at the entrance, then water, movement and a town carrying on.

What the Drone Reveals

I did not think of the courthouse and park as a matched pair when I launched. The connection appeared later, while I was reviewing the footage.

The courthouse preserves memory through a building that is still doing its original kind of work.

The park does it differently. Its memorial shares the grounds with walking paths, basketball courts, playgrounds, fishing and fountains. One place keeps history inside its walls. The other lets daily life move around it.

Not every pass worked.

When I filmed the Lincoln statue, my stick inputs became jerky. The orbit lost the smoothness I wanted, and I could see the mistake as soon as I watched it back.

I refuse to use automated flight paths for these films. Automation could draw a cleaner circle, but it would also remove the part I am trying to improve: flying the aircraft myself, reading the wind and learning how small movements translate through the camera.

DD-0001 has no narration or natural location sound. The first music track fades with the courthouse. A new track begins when Veterans Memorial Park appears. It separates the two locations without breaking the journey.

The finished footage is calmer than every moment behind the controller. That gap between the flight I planned and the flight I actually flew is the reason I keep going back up.

Follow me on X: @Dronedrifterx2
Explore more films and field notes: DroneDrifter.com

Sources and further reading

• City of Mount Vernon: About Mount Vernon

• Illinois Courts: Fifth District Appellate Court

• Society of Architectural Historians: Appellate Courthouse

• University of Illinois: The Appellate Courthouse

• City of Mount Vernon: Veterans Memorial Park

Finishers stood around in the heat, medals still on, paper cups in hand. Nobody in a hurry to leave. An empty stroller parked by the cones. Same parking lot the freeway commuters blow past every morning, and for one Saturday morning it belonged to the people who ran it.

See the run and the finish line first:

And the church behind the whole scene:

If you like these small places and the people who fill them, subscribe. I’ll keep putting the drone up.

Subscribe now

Kids crossed the line in gold laurel crowns. A red FINISH flag, a folding table, a timing clock on a tripod, and the inner belt arching over all of it with downtown sitting right behind. Same patch of grass I shot empty for the church piece. Packed this time.

The church story behind the finish line:

Next stop, the after party:

If this is your kind of ground-level history with a drone in the air, subscribe. More stops are coming.

Subscribe now

Coming off I-90 onto the West 14th ramp the road bends, and there it is. Three domes and a row of gold crosses, sitting maybe a stone’s throw off the off-ramp like the highway department drew the merge lane around it and kept going. Most people clock it for half a second at sixty and never think about it again. Tens of thousands of cars a day. A church older than the road, parked at the road’s front door.

That’s Annunciation Greek Orthodox, 2187 West 14th. The first Greek church in Cleveland, and for eighteen years the only one.

Get a drone up over it and you see what the drivers never do. The domes are gold, a cross standing on each. The yellow brick towers wedged in tight against the inner belt, and past all of it, flat on the horizon, Lake Erie.

The Greeks didn’t start on this side of the river. The first ones showed up in the 1890s and packed into a few blocks downtown off Bolivar Road, a stretch they called Greek Town, between Ontario and East 9th. Coffeehouses. Rented rooms. When a priest passed through he’d say the liturgy wherever there was space, somebody’s apartment, or up the hill at St. Theodosius, the Russian church with the onion domes you’ve seen in The Deer Hunter. For a while the only regular clergy they had was Father George Scarpas, who rode the train in from Pittsburgh once a month and rode it back.

The first real Greek Orthodox service in Cleveland had a roof over it and a movie screen under it. Christmas Eve, in a hall above a theater at Ontario and Bolivar. That’s where it started. Above a movie house.

By 1912 the congregation had organized itself as the Pan-Hellenic Union and scraped together $12,000. Fifteen families. They crossed the river and bought two mansions on West 14th, then called Jennings Avenue, that had belonged to a pair of industrialists named Lamson and a partner named Sessions. Cleveland money, the kind of names that ended up bolted to factory walls. The Greeks tore the mansions down and started building a church on the lot.

They incorporated in February 1913. The building wasn’t done for years. They held the liturgy in it anyway, unfinished, starting in 1919. Weddings and funerals were still happening in people’s living rooms into the 1930s because the church couldn’t hold them yet.

Then the icons.

In 1924 a priest named John Zografos took over the parish. He stayed four years. In those four years he painted every icon in the building. All 85 of them. The Christ in the dome, the saints down the walls, the screen standing between the people and the altar, every one of them came off the brush of one man who was also running the parish and starting the first Greek school in the city. They’re still up there. Walk in today and you’re looking at one man’s four years of work, a century later, nobody having to repaint it.

That’s the part that gets me. In the Orthodox way of seeing it, an icon is a window. You look at the saint and the saint looks back. You don’t sign one the way a painter signs a canvas. Zografos spent four years putting those faces on the walls so people could look through them to whatever’s on the other side. He died, the priests after him died, and the windows stayed open.

The neighborhood around the church came apart. By the 1960s the old estates on West 14th were gone and the state ran Interstate 90 straight through Tremont. The freeway split the neighborhood and dumped an on-ramp at the church’s front step. The money moved to the suburbs. SS. Constantine and Helen went up in Cleveland Heights, St. Demetrios out in Rocky River. Both trace back to this one corner. The parish on West 14th stayed put. Back in the thirties the mortgage nearly went under, and a man named Antony Hoty stepped in and pulled it off the foreclosure block. They kept the doors open through all of it. That’s why they still call it the Mother Church, the one every Greek parish in town descends from.

There’s something fitting in the name. The Annunciation is March 25, the morning the angel Gabriel tells Mary she’s going to carry Christ. It’s the hinge the whole story turns on, the very start of it. They named the first Greek church in Cleveland after a beginning, and it became the root the rest grew out of.

For most of a hundred years the view out the front door was an on-ramp and not much else. That’s finally turning. Cleveland Metroparks is putting a four-acre park on the ground in front of the church, tied into the new bridge work down by the river. Grass and trees where there’s been concrete and merge lanes since Eisenhower was president. And every Memorial Day weekend the lot fills for the Tremont Greek Fest, lamb off the spit and loukoumades fried to order and a line out the door, the one weekend a year the rest of Cleveland walks through the doors instead of glancing at the roof from the highway.

Next time you come off the belt and the road bends toward West 14th, give it the half second you’ve got. Look up. Fifteen families. Twelve thousand dollars. One priest with a paintbrush and four years on the scaffold. It’s all still up there, holding the same corner it held before the road existed, and it isn’t going anywhere the freeway can put it.

If places like this are your kind of stop—the history hiding beside the road, seen from the ground and the air—subscribe to Drone Drifter. I’m always looking for the next one.

Subscribe now

Another park that changed shape from the air: North Jetty at Casey Key.

Six pickleball courts on one side. Soccer fields on the other. And a deer highway running right down the middle.

Carter Woods Park in Grand Chute, Wisconsin hides a wild corridor most people walk right past.

The parking lot won't tell you. Drive in off Bluemound and you get the usual town-park stuff. A softball diamond. A playground. Picnic tables, and pickleball courts that fill up with regulars by 8 in the morning. Looks like every other suburban park in the Fox Cities.

Then there's the middle.

Down the center of Carter Woods runs a strip of real woods, with dried-out stream beds that feed Mudd Creek. Bridges cross it. Footpaths wind through it. On foot it's all leaf litter and downed trunks, the light coming through green. And the deer are almost always in there. They hang back, far enough to keep one eye on you.

You can't read the shape of it from the ground. So I put the drone up.

Not for long. There's a National Guard armory sitting right next to the park, and that changes how you fly. You keep it short and tight. You keep the bird close, and you don't linger over anything with a fence and a flag. I got my shots and brought it back down.

Even in that short window, from a couple hundred feet, the whole place flips. Two ball diamonds, red infields set in a sea of green. I dropped low and skimmed the open field, tree shadows stretched halfway across the grass. What the camera caught from up there is a green seam cutting straight through the suburb, and it doesn't stop at the park line. Follow that tree line west and it threads through the Butte des Morts Country Club and runs to the Fox River. The deer use it like a road. So do the hawks nesting on the cell towers.

The pickleball place is a wildlife corridor wearing a town park.

A patch of wild woods right alongside a National Guard armory. Same fence line, middle of the Fox Cities. The deer don't seem to mind the neighbors.

Footage is up top. If you want more of these, overlooked parks and small-town corners shot from a couple hundred feet, subscribe to the Drone Drifter.

Subscribe

The rain started the second I had the drone up. A Gulf drizzle, the kind that blows in sideways off the water for ten minutes and quits before you've decided whether to pack up. My niece was in town. She'd brought her boyfriend, and my second nephew came along too, so the four of us stood out on the rocks at the south tip of Casey Key watching me fly a drone into weather I'd normally land it for.

Nobody wanted to leave. That's the thing about the North Jetty.

It sits at the very bottom of Casey Key, a skinny barrier island about eight miles long that runs down the coast just north of Venice. The rocks hold Venice Inlet in place and help control the sand moving along the shore. Practical, ugly, the kind of thing nobody flies across the country to see.

Then the rock did something useful in a different way. The jetty and the sandbar beside it can shape a rideable wave when Gulf conditions line up. On a coast that spends plenty of days nearly flat, that is enough to draw surfers. They built a wall to manage the inlet and ended up with a wave. I think about that more than I should. You set out to fix one small problem and the world hands you back something it never occurred to you to ask for.

From two hundred feet up, the inlet goes dark green where it drops off. You can see the exact seam where the calm intracoastal water meets the chop of the open gulf, sharp as a chalk line. Two guys fished the rocks below us with cut bait. Not moving. Barely talking. A dolphin worked the inlet for a minute, then gone.

And under all of it, in the dark band of sand the low tide leaves behind, the shark teeth.

People drive here for them. Black fossilized teeth, washed out of old marine sediment and pushed back onto the beach by waves. Find one and you are holding a piece of an ocean that existed long before the shoreline under your feet. My niece found two before the rain came. Didn’t make a thing of it, just dropped them in her pocket and kept her eyes down on the sand.

The drizzle passed. The light went soft and gray instead of gold. Better that way.

We stayed until the lot emptied out. Best afternoon I’ve had in a while, and I didn’t get one clean sunset shot to show for it.

Another flight from the same stretch of Florida coast: Stump Pass Marina From Above.

If you like quiet places, imperfect weather and the story behind the footage, subscribe. More flights are coming.

Subscribe now

All of it sitting on top of a strange piece of history. On August 1, 1901, this was open prairie. By the next morning, 13,000 people were living in tents while they waited for land, and the place picked up the name Ragtown. When the town lots began selling on August 6, Hobart had 2,530 residents. Now the historic Kiowa County Courthouse sits in the middle of the square, with a tile mosaic laid into the ground nearby.

Catch it from the air at the right hour and the whole place glows the same red as the dirt it grew out of.

A small town that still knows how to fill its own square.

If you like small towns, local history and the view from a couple hundred feet, subscribe. More flights are coming.

Another courthouse landmark from above: Mount Vernon, Illinois.

Subscribe now

The red-stone clubhouse was built in 1925 for the original Clinton Country Club. It came back to life in 2007 as White Dog Hill, sitting high enough above the prairie that the sunset seems to take its time getting there.

Eat there once and you are already planning the drive back.

This was my last flight over the place. Every other time I came up here I flew it alone. This time I had folks with me I can truly call friends. Makes a difference, looking down at a spot you love and knowing the people standing next to you actually matter.

The Mini 5 Pro caught one more sunset off that hill.

Good way to shut it down.

My earlier White Dog Hill stops:

Another Clinton landmark from above:

If you like the places worth pulling off the highway for, subscribe. I am still looking for the next one.

Subscribe now

Three hundred fifteen acres is a lot of park.

The front side is easy to understand from the parking lot. There is a 23,000-square-foot recreation center, a competition pool, football fields, basketball and tennis courts, a skate park, disc golf, cricket and a dog park. It looks like Charlotte County gathered every kind of recreation it could fit and gave each one a piece of the property.

From the air, that busy side becomes geometry. The blue pool sits beside the pale roofs. Fields and courts break the grass into rectangles. Paths connect one use to the next. Then the lines loosen as the camera moves toward the back of the park.

That is where the place changes. Ann & Chuck Dever was formerly called Oyster Creek Regional Park. Of its 315.6 acres, Charlotte County identifies 50 acres for active use and 120 acres as environmental land, with another 36 acres north of the creek. The numbers make sense once the drone is high enough. Recreation occupies one section. The rest starts folding into pine, scrub and water.

The Oyster Creek Waterfront Trail connects this park with Cedar Point Environmental Park. It follows the creek and passes beneath State Road 775, tying the two properties together without sending walkers back onto the road. From above, that green connection matters more than any single field or court.

The park was built in 2007. The recreation center arrived in 2019. Both are substantial, but neither one explains the whole place. The interesting part is the seam between the developed side and the older Florida behind it.

That was the reason for coming back. I did not need another version of the same flight. I needed to go farther and let the park show its second half.

If you like return flights, overlooked Florida places and the story behind the footage, subscribe. More field notes are coming.

Another Englewood flight from the same stretch of coast: Lighthouse Grill and Stump Pass Marina.

Subscribe now

Why the Same Story?

Convergence, Not Influence

I want to be clear about something. There’s no evidence that Shotaro Ishinomori knew anything about the Ottoman devshirme, the Janissary corps, Skanderbeg, Ballaban Badera, Sokollu Mehmed Pasha, or any of the history I’ve described. Zero evidence. I looked. I don’t think he was drawing on Ottoman history. I don’t think there’s a chain of influence connecting Ishinomori’s studio in 1970 to the Sultan’s court in 1450.

This isn’t an influence argument. It’s a convergence argument. And the convergence argument is stronger.

Two completely unrelated systems of imperial subject-creation, separated by five centuries and six thousand miles, produced the same story. The same sequence: seizure, erasure, remaking, deployment, escape, rebellion. The same spectrum of outcomes: loyalty, failed resistance, internal subversion, defection, revolt, seizure of power. The same climax: the state trying to destroy what it made when the creation proves uncontrollable.

The stories converge because both systems begin with the same act: treating a person as raw material. The details are different. The structure is close enough to expose the same weakness.

The Logic

Any system that (1) seizes people from outside itself, (2) erases their prior identity, and (3) remakes them as instruments of its own power contains a built-in vulnerability. The remaking might be incomplete. The instrument might retain enough of its prior self to refuse. And the system can’t easily hedge against this, because the same process that makes the instrument powerful is the process that makes the instrument dangerous if it turns. A Janissary trained to be the finest soldier in Europe is also, if he defects, the finest soldier fighting against the empire. A cyborg built to be Shocker’s most lethal weapon is also, if he escapes, the most lethal threat Shocker faces.

The system can’t produce weak weapons. Weak weapons are useless. But strong weapons, if they turn, are existential threats. The better the system is at its job, the worse the consequences when the job fails.

This is a structural feature of empires that manufacture soldiers from stolen subjects. Not a bug. A load-bearing design flaw. The devshirme had it. Shocker has it. Any system built on this model will have it, because the vulnerability is baked into the logic.

The Image That Doesn’t Erase

Political and structural analysis gets you most of the way through this parallel. The mechanical argument holds. But the Neomartyrs break the mechanical explanation.

Structural analysis can explain Skanderbeg. He defected before the final step, retained memory of who he was, had the military capability to act on it. You can explain that as incomplete conditioning. You can explain Sokollu as ethnic solidarity, or the ordinary human drive to preserve what you came from. Patrona Halil, the Köprülüs, Kabakçı Mustafa. All explainable in political and psychological terms.

The Neomartyrs are harder. The process finished. The conditioning completed. There’s no political incentive to recant publicly. It means death. There’s no escape, no war. Just a person who had been living as something the system made them, and one day decided they weren’t, and said so out loud.

Aristotle argued in De Anima that the soul is the form of the body, not a substance added to it but the body’s organizing principle. The rational soul, the nous, isn’t a product of conditioning. It’s the thing doing the processing, the thing that orients and decides. You can alter circumstances. You can layer new habits over old ones. But the form persists. It goes quiet. It waits.

Plato’s argument is related but sharper. The soul has a natural orientation toward the Good. Not the good defined for it by the Sultan, not the good Shocker’s scientists approved. The Good itself, which no institution gets to assign. Coercion can redirect the soul’s attention, make it look the other way, train it to perform against its own nature. But the appetite for its true end doesn’t get trained out. The allegory of the cave makes it concrete: a person who has glimpsed the light, even briefly, even in some half-remembered childhood before the officials arrived on horseback, can’t be permanently satisfied with shadows.

Start with Part I: The Collection.

Then read Part II: The Weapon.

The Eastern Church synthesized both of these and went further. The eikon isn’t only a philosophical concept. It’s a theological reality. God made every person in his image, and that image is what makes resistance possible at the deepest level. Not political will, not military training. The image. The thing that no production line can reach, that no decade of conditioning can fully overwrite, that asserts itself in the Neomartyr who walks out into a public square and says “I am Christian” knowing exactly what comes next.

Ottoman history supplies cases that test that claim. Across centuries of forced recruitment and identity change, it still produced defectors, internal subversion and public reversals. Not because every subject resisted. Most did not. The point is that the process never guaranteed complete ownership of the person inside it.

Ishinomori’s Postwar Context

Ishinomori was born in 1938 in Tome, Miyagi Prefecture. He was seven years old when Japan surrendered. He grew up in the rubble of an empire that had, within living memory, done its own version of this same thing: conscripted its citizens, indoctrinated them in the ideology of kokutai (the national body), remade civilians into soldiers who were expected to die for the Emperor.

I’m not going to overclaim this. I don’t know what Ishinomori was thinking about when he created Skull Man and Kamen Rider. I don’t have diary entries or interviews where he says “I based Shocker on the Imperial Japanese military” or “I was thinking about the devshirme.” But he was a Japanese man born in the late 1930s, writing in the early 1970s, and the thing he kept writing about, the story he told over and over, was: a state kidnaps a person and remakes him into a weapon, and the weapon escapes.

Japan had direct experience with that machinery. The whole country did. Every family that sent a son to the Imperial Army knew some version of it. The state takes your child. The state remakes your child. The state sends your child to die.

That context helps explain the convergence. Ishinomori and the Ottoman devshirme produced the same narrative because they were both describing systems of imperial subject-creation. Different empires, different centuries, different methods. Same logic. And the same logic produces the same story, because the story isn’t about any particular empire. The story is about what happens when power tries to eat people and turn them into instruments, and what it looks like when one of them gets free.

The Weapon Remembers

November 28, 1443. Krujë. Skanderbeg raises the double-headed eagle of his family over the fortress walls. He took the city with a forged Ottoman letter and a lifetime of Ottoman military training. His brother’s killer, maybe. His father’s pride, maybe. A warlord who would spend twenty-five years bleeding the empire that built him. They called him Iskander Bey, Lord Alexander, and the name they gave him became the name that haunted them.

1557. Istanbul. Sokollu Mehmed Pasha, Grand Vizier of the Ottoman Empire, born a Serbian Orthodox boy in Herzegovina, signs the order restoring the Serbian Patriarchate of Peć. He installs his own cousin as Patriarch. He does this using the full authority of the office the devshirme gave him. The imposed identity funding the survival of the erased one.

1558. Krujë again. Ballaban Badera, born Michael, son of Mark, dies outside the walls his brother defends from within. A hunter named Gjergj Aleksi shoots him through the neck. The conversion held. Michael never came back. The system won that one.

September 1730. Istanbul. Twelve thousand Albanian Janissaries storm the capital. Patrona Halil forces a Sultan off his throne. Two months later, the new Sultan has him killed. The creation was too dangerous to leave alive, even after it did what the state wanted.

June 15, 1826. Istanbul. Mahmud II fires on the Janissary barracks. Six thousand dead. The empire dismantling the army it spent four centuries building out of stolen children, because the children grew too strong.

1803. Constantinople. A man named Demetrius, living as a Muslim for years or decades, walks out into a public space and announces that he is Christian. He knows what happens next. The Ottoman court executes him. The Orthodox Church declares him a saint. There was no tactical calculation in what he did. No escape route, no guerrilla campaign, no twenty-five-year war. Just the image asserting itself, past all the conditioning, past everything the system spent years installing.

April 3, 1971. A Japanese television screen. A college student named Takeshi Hongo, surgically altered into a grasshopper cyborg by a global terrorist organization called Shocker, escapes before the brainwashing can take. He looks at the body they gave him, this monstrous rebuilt thing, and he decides to use it against them. He puts on the mask. He fights.

Different centuries. Different continents. Different bodies carrying the marks of different empires.

The same crack in the system. The place where the remaking failed. The ten percent that didn’t take. The will that survived the surgery, the conversion, the renaming, the decade of training, all of it designed to produce an obedient weapon, and somehow the person inside held on.

Five hundred years of Ottoman history couldn’t close that crack. Fifty years of Kamen Rider franchise storytelling keeps returning to it, episode after episode, series after series, because the story refuses to be finished. The weapon keeps remembering. The weapon keeps turning.

And the empire, whatever empire, in whatever century, on whatever continent, keeps learning the same thing the hard way: you can remake the body. You can bury the name. You can kill the god and teach a new language and train the hands to hold a sword for your wars.

The Church Fathers said you can’t destroy the eikon. The Ottoman Empire tested that for four centuries and never succeeded. Ishinomori wrote the same answer in manga and then in a children’s television show.

You can’t always get the last ten percent. And that’s enough.

That's the series. If it earned your time, share Part I — that's the entry point, and it's free. If you've been reading along without subscribing, now's the time. And if you want to support work like this, a paid subscription keeps it going.

Subscribe now

God Bless. — George

Now hold all of that. Keep the devshirme in your head. Keep Skanderbeg on the walls of Krujë and Ballaban dying outside them. Keep the eight boys tortured for a year. Keep Sokollu restoring the Serbian Patriarchate with Ottoman money. Keep the 6,000 dead in Istanbul in 1826.

And open a manga from 1970.

Ishinomori’s Architecture: From Skull Man to Kamen Rider

Skull Man: The Unfiltered Version

Shotaro Ishinomori published Skull Man as a hundred-page one-shot in Weekly Shōnen Magazine in 1970. It sold over 1.5 million copies. The protagonist is Tatsuo Kagura, eighteen years old, an orphan whose parents were murdered. He takes the name Skull Man. His only companion is Garo, an artificial creature. He pursues vengeance against the people who destroyed his family.

That description makes him sound like a hero. He isn’t.

Skull Man kills innocents. He kills indiscriminately. Ishinomori designed him to frighten children. The press in the story’s world censors his existence because the public can’t handle knowing he’s out there. He is manga’s first anti-hero by most credible accounts, and the reason he’s the first is that nobody before Ishinomori had the nerve to build a protagonist this ugly. A person made monstrous by what was done to him, who turns that monstrousness outward without much caring where it lands.

The weapon is loose. And it doesn’t care who’s standing in the way.

This is the honest version of the escaped-weapon story. The version where the system’s product, once freed, isn’t a savior. He’s damaged. Dangerous. Shaped entirely by the violence that made him. Skull Man is what Skanderbeg looks like if you strip away five centuries of national mythology and look at the warlord underneath: a man who burned a quarter-century of his life making war, trained to kill by the people he was killing, incapable of becoming anything other than what the system built him to be.

The Transition

In 1970, Toei Company producer Toru Hirayama came to Ishinomori looking for a TV superhero. Ishinomori pitched Skull Man.

Toei said no. Too dark. Too violent. A hundred pages of an anti-hero slaughtering people wasn’t going to work as a children’s show airing on Saturday mornings.

So they softened it. The skull motif became a grasshopper. Ishinomori’s son reportedly picked the insect. The indiscriminate violence became directed heroism. The moral ambiguity got scrubbed out for broadcast.

On April 3, 1971, Kamen Rider premiered on Japanese television.

The underlying machinery was untouched. A person is seized. Remade into a weapon. Escapes before the process is complete. Turns against the system that made him. The structure survived the transition from Skull Man to Kamen Rider perfectly intact. What changed was the moral packaging. The escaped weapon became a good guy.

This mirrors something in the Ottoman parallel that I think is worth flagging: the same softening happened to Skanderbeg over centuries of Albanian folk memory. The warlord who fought a brutal guerrilla campaign for twenty-five years became a clean national hero, a symbol of freedom, a figure on stamps and statues. Ishinomori told the raw version in 1970 and the mythologized version in 1971. Albanian culture did the same thing over five hundred years. Different timescales. Same process. The escaped weapon gets cleaned up for the audience that needs him.

Kamen Rider: The Hero

Takeshi Hongo is a college student. Shocker, a global terrorist organization, kidnaps him and surgically alters him into a cyborg. The Japanese term is kaizō ningen, “modified human.” They graft animal and insect DNA into his body alongside cybernetic enhancements. The conversion happens while Hongo is unconscious.

The brainwashing doesn’t.

Dr. Hiroshi Midorikawa, a scientist working inside Shocker, intervenes before the final step. He gets Hongo out. Later, when Shocker tries the brainwashing again, Hongo uses self-hypnosis to resist. The body is Shocker’s work. The mind stays his.

In Ishinomori’s manga, the version he drew himself, Hongo is wrecked by this. He hates what he’s become. He looks at his cyborg body and sees a monster, something Shocker built, something that isn’t fully human anymore. He carries that self-loathing for a long time before he accepts who and what he is and decides to use Shocker’s modifications against them.

The gap between physical transformation and mental subjugation is Hongo’s entire story. That gap is where his freedom lives. It’s a thin space. Everything Hongo is as a hero exists because of one incomplete step in a process designed to produce an obedient weapon.

Sound familiar?

Shocker: The Antagonist as Empire

Shocker wants to conquer the world. The method: kidnap human beings, surgically modify them into mutant cyborgs called Kaijin, brainwash them, and deploy them. The organization’s scientists are themselves coerced, kidnapped specialists forced to work under threat of death. When a scientist stops being useful, Shocker kills them. When a scientist tries to escape, Shocker kills them.

The Kamen Rider Spirits manga connects Shocker’s founders to former Nazis. The fictional organization is rooted in the specific history of twentieth-century totalitarianism. The Great Leader of Shocker’s stated endgame is the conversion of the entire human race into Kaijin, a single unified entity under his control.

Shocker doesn’t want to rule people. It wants to replace them. Every human being is raw material waiting to be processed. The person who walks into a Shocker lab ceases to exist. What walks out is an instrument. A tool. Property.

The devshirme, with surgical tables instead of circumcision rites and cybernetic implants instead of Turkish language immersion.

The Mirror: Structural Parallels

The Production Cycle

Lay the two systems next to each other and walk through them step by step.

Seizure. Ottoman officials ride into a Christian village and take boys from their families. Shocker operatives kidnap Takeshi Hongo off a street.

Erasure. The Ottoman boy is separated from his parents, his church, his community. He is told to forget his name. Hongo is placed on an operating table, unconscious, his body opened and rewritten without his knowledge or consent.

Remaking. The boy is converted to Islam, given a Turkish name, trained in a new language, acculturated through years of labor in Anatolian villages, then military drilling in Istanbul. Hongo is implanted with insect DNA and cybernetic systems, his body rebuilt into a weapon.

Deployment. The boy enters the Janissary corps. He fights the Sultan’s wars. Hongo is intended to enter Shocker’s Kaijin army. He is meant to fight the Great Leader’s wars.

The key fact in both cases: the subject isn’t killed. He’s replaced. The original person is overwritten with a new one. That’s what makes the escaped weapon so dangerous to the system. Shocker knows exactly what it put inside Hongo, because Shocker put it there. The Ottoman state knew exactly what its Janissaries were capable of, because the state spent a decade building them.

When the product escapes, the manufacturer knows its specifications. That’s terrifying.

What the System Produces

I laid out nine Ottoman figures in Part I. Each one represents a different outcome of the devshirme, a different answer to the question of what happens when you steal a child and remake him. Each one maps onto something in Ishinomori’s world.

Skanderbeg defected. He’s Takeshi Hongo, the escaped hero who fights back.

Ballaban Badera stayed loyal. He’s the Kaijin, the completed product sent to kill the one who got away. He fought his own father across a siege wall and never flinched, or if he flinched we have no record of it.

The Neomartyrs completed the conversion, lived inside it for years, then reversed it at the cost of their lives. They’re the Kaijin who, deep into service, somehow remembered who they were before the table. No escape route, no war. Just the image coming back through.

The eight Christian boys tried to resist from inside and failed. They’re the unnamed captives in Shocker’s labs, the ones the audience never sees, the ones whose resistance ended in a room with no witnesses.

Sokollu Mehmed Pasha subverted from within. He’s Dr. Midorikawa, the insider who uses his position to sabotage the machine. Sokollu restored the Serbian Patriarchate with Ottoman authority. Midorikawa broke Hongo out of Shocker using Shocker’s own infrastructure.

Patrona Halil seized the throne. He deposed a Sultan. This is the anxiety that runs underneath the entire Kamen Rider franchise, the one that never quite gets spoken aloud: the modified human isn’t just a rogue. He’s a rival. He could replace his makers.

Kabakçı Mustafa refused to be replaced. When Selim III tried to build a new army, the old one tore him off his throne. Shocker can’t tolerate Kamen Rider’s existence for the same reason. A superior product, walking around free, is proof that the system’s control has failed.

The Dahijas went rogue and burned everything down. Their tyranny in Serbia triggered a national uprising that eventually ripped a piece of the empire away. This is what happens when Shocker’s methods produce consequences the organization can’t contain. Chaos multiplying beyond the point of retrieval.

The Köprülü family ran the empire. Six Grand Viziers from a single Albanian devshirme line. The stolen children surpassed their masters. This is the thing the Kamen Rider franchise accidentally keeps proving: the hero is always stronger than the organization. Always. Every single series. The modified human surpasses the modifier. The weapon outperforms the arsenal that produced it.

That’s nine outcomes and nine answers to the same question. The point isn’t that Ottoman history and Ishinomori’s fiction are identical. It’s that both expose the same weakness in a system that remakes people as tools: the stronger the tool becomes, the greater the danger if control fails.

The Incomplete Conversion

Skanderbeg served the Ottoman military for roughly two decades before he defected. Whatever identity erasure the devshirme performed on George Kastrioti, it held for twenty years. Then it didn’t. Something remained, or something came back. He raised his family’s banner over Krujë. He used a name the Ottomans gave him, Skanderbeg, “Lord Alexander,” to wage war against the Ottomans. He couldn’t shed the name. He couldn’t shed the military training. He could only point it all in the other direction.

Hongo can’t undo the surgery. His body is Shocker’s work permanently. The grasshopper cyborg form, the enhanced strength, the combat capabilities. All of it stays. He carries Shocker’s modifications for the rest of his life. What he can do, the only thing he can do, is refuse the purpose those modifications were designed to serve.

The hero in both cases is defined by an incomplete process. The devshirme or Shocker did ninety percent of the work. The body was remade, the skills were installed, the power was granted. The last ten percent, the part where the person inside gets erased and replaced with obedience, didn’t take.

Contrast Ballaban. The conversion completed. He fought his brother. He besieged a city defended by his father. Whatever was left of Michael of Badera, it wasn’t enough. The hero and the villain in this story came off the same production line. Same system, same process, same inputs. The only variable is whether that final step, the one that kills the person and replaces them with a tool, finishes.

That’s a specific kind of heroism. Not the heroism of the pure, the untouched, the chosen one born special. The heroism of the partially converted. The hero is always, to some degree, a product of the enemy. He carries the enemy’s work in his body. He fights using skills the enemy gave him. His power comes from the people he’s fighting against. And his freedom, the narrow gap between what was done to his body and what wasn’t done to his mind, is everything.

The Name and the Mask

George Kastrioti was renamed Iskander Bey. After his defection he was known by both names. The Ottoman name stuck. It’s the name Albania remembers him by: Skanderbeg. The imposed identity became inseparable from the person. He couldn’t go back to being George Kastrioti, a minor nobleman’s son from a province the Ottomans swallowed. The devshirme made him Skanderbeg, and Skanderbeg is who he was when he fought the devshirme.

Hongo becomes Kamen Rider. Literally: Masked Rider. The mask is Shocker’s work. The grasshopper cyborg form is what they built him to be. But Hongo wears it to fight them. The mark of captivity becomes a weapon of resistance. He can’t take it off. He can’t go back to being a college student. The transformation is permanent. So he turns it around.

Sokollu Mehmed Pasha kept his Ottoman name for his entire life. He never reclaimed whatever Serbian name he was born with. He used the authority that the Ottoman name gave him, the Grand Vizierate, the full weight of imperial power, to restore the Serbian Patriarchate. The imposed identity became the instrument of preserving the erased one.

In all of these cases, the person can’t return to who they were before the system took them. The process is irreversible. What they can do is repurpose the new identity. Turn it against its creators. The mask becomes the weapon. The Ottoman title becomes a rebel’s banner. The Grand Vizier’s seal restores a church the empire tried to bury.

The State Destroys What It Creates

Selim III tried to replace the Janissaries in 1807. Kabakçı Mustafa and the Janissaries deposed him. Nineteen years later, Mahmud II succeeded where Selim failed, but only by surrounding the barracks with cannon and firing until the problem was solved. Six thousand dead. The corps abolished.

The Dahijas, rogue Janissaries in Serbia, provoked the First Serbian Uprising by overreaching so badly that an entire nation rose against them. The empire’s products, operating outside any command structure, triggered the dismemberment of the empire’s territory.

Shocker pours resources into hunting Kamen Rider. Every episode, every arc. The organization sends Kaijin after Kaijin, devotes operational capacity, sacrifices personnel. Because it has to. It built Hongo. It knows what he can do. An escaped product walking around free, using Shocker’s own technology against Shocker, is an existential problem precisely because of how good the technology is.

Any system that manufactures weapons out of human beings will eventually face the possibility that those weapons turn. The Janissary history shows every version of that turning: defection, subversion, revolt, rogue action, outright seizure of power. The Ottoman state tried integration, accommodation, reform, and finally extermination. Shocker tries the same progression across the franchise. None of it works. The weapon keeps fighting.

The Conscience Inside the Machine

Dr. Midorikawa saved Hongo. He was a Shocker scientist who broke ranks, who looked at the operating table and the unconscious college student strapped to it and decided he couldn’t let the last step happen. He got Hongo out. It probably cost him his life.

Sokollu Mehmed Pasha spent fourteen years as Grand Vizier doing something quieter but structurally identical. He worked inside the system and used the system’s resources to protect the thing the system was designed to destroy. He restored a church. He appointed his Christian relatives to positions of power. He never left. He never raised a banner. He just bent the machinery, degree by degree, toward a purpose its builders never intended.

Ottoman administrators sometimes looked the other way when Balkan families tried to hide their sons from the devshirme. They let circumcision ruses work. They lost paperwork. The system was never airtight. There were always people inside it who cracked, who couldn’t stomach what the machine required of them.

Skanderbeg didn’t defect alone. Three hundred Albanians left with him at Niš. Skull Man has Garo, his artificial companion, the one loyal figure in a hostile world. The escapee rarely operates in isolation. The system produces not just individual defectors but clusters of dissent, small networks of people who recognize each other and move together.

The machine leaks. That’s structural too.

Start with Part I if you came in here: The Collection.

If this series is earning your time, subscribe and keep going. And tell me which figure from Part I maps closest to Hongo in your reading.

Continue with Part III: The Image That Doesn’t Erase, where the argument moves from political machinery to the person the machinery cannot fully replace.

Subscribe now

An empire steals a child.

It takes him from his family, strips his name, buries his language, kills his god and gives him a new one. It trains him to fight, to kill, to obey. It pours years into making him the sharpest blade in its arsenal. Then it points him at its enemies and says: Go!

And sometimes the blade turns around.

This is one of the oldest stories that power tells about itself, and one of the oldest stories that the powerless tell about surviving it. You can find it in Albanian folk memory, where a stolen nobleman’s son named George Kastrioti burned twenty-five years of his life fighting the empire that raised him. You can find it in a 1970 Japanese manga called Skull Man, where an orphan remade by violence becomes a walking nightmare that frightens children. You can find it the following year, April 3, 1971, when a TV show called Kamen Rider premiered on Japanese television and told a cleaned-up version of the same story to millions of kids eating breakfast.

I found no evidence that Ishinomori drew on Ottoman history when he created Skull Man or Kamen Rider. The connection I see isn’t influence. It’s convergence.

Systems that seize people and remake them as weapons also create the possibility of the weapon that turns. The story repeats because the machinery repeats. An empire can bury a name, install a new language and train the hands. The harder question is whether it can finish the erasure.

There’s a frame for this that political analysis misses. The boys taken by the devshirme didn’t come from a secular tradition. They came from the Eastern Church. And the Eastern Church had been thinking about this exact problem for over a thousand years before the Ottomans arrived. What happens when power tries to unmake a person. The Greek Fathers called it the eikon. The image. Genesis 1:26, in the Greek of the Septuagint: kat’ eikona hēmeteron, “according to our image.” Not a metaphor. An ontological fact about every human person. The way a coin carries the face of the emperor, the Church Fathers said, every person carries the image of God. You can bury the name. You can stamp over the face. But Basil of Caesarea and Gregory of Nyssa were clear on the result: you can’t destroy the image. The Ottoman Empire spent four centuries running an empirical test on that claim.

The Devshirme: Empire as Production Line

The Ottoman word was devshirme. It means “collection.” The bureaucratic flatness of the term is worth sitting with for a second, because what it describes is the systematic kidnapping of children.

Starting roughly under Sultan Bayazid I in the late fourteenth century and reaching its full form by the 1400s, the devshirme worked like this: every four or five years, Ottoman officials rode into the rural Christian provinces of the Balkans and Anatolia and took boys. One boy per forty households. Ages seven to twenty, though the younger ones were preferred. The boys were separated from their parents, their churches, their villages. They were told to forget.

That was step one.

Step two was conversion. The boys were circumcised, given Muslim names, and forced to convert to Islam. Their birth names were buried. Step three was assimilation: the boys were shipped to Turkish villages in Anatolia and hired out as agricultural laborers for three to five years, learning Turkish, absorbing Islamic practices, forgetting whatever language they’d spoken at home. Step four was the recall to Istanbul for another three to five years of labor and training. Step five was deployment: the finished product entered the Janissary infantry corps, the Sultan’s personal military force, or was placed in palace schools to be trained as an administrator.

The whole process took a decade or more. By the end, the Christian boy from a Balkan village had been replaced. His name, his language, his religion, his family, his community, his sense of who he was before the officials came through town on horseback. All of it erased and overwritten. What came out the other end was a kul, a slave of the Sultan, loyal to no one and nothing except the Ottoman state.

That was the idea, anyway.

The Janissary Paradox

The Janissaries became the most feared fighting force in Europe and the Middle East for roughly two centuries. The Sultan’s personal guard. The shock troops of imperial expansion. The institution that held the empire together.

And they were all stolen children.

This is the paradox at the center of the devshirme, and it’s the same paradox that powers Ishinomori’s fiction: the system’s most capable instruments were the very people it had seized and remade. The empire elevated its captives into positions of extraordinary power. Janissaries could rise to the highest offices of the state. They accumulated wealth, political influence, institutional authority. By the seventeenth century they were deposing Sultans.

Some historians have complicated the picture, arguing that the devshirme was also a vehicle for social mobility, that families sometimes volunteered their sons because the Janissary path offered wealth and status otherwise impossible for a Christian peasant. This is true. It’s also true that a kidnapped boy given a sword and a title is still a kidnapped boy. Elevation and enslavement aren’t opposites. They can live in the same body at the same time, and the tension between them is exactly where the story gets interesting.

The Spectrum of Resistance

The devshirme didn’t produce one type of outcome. It produced a full range of them, from total loyalty to outright rebellion, with every shade of compromise and subversion in between. Each of these outcomes has a counterpart in Ishinomori’s fictional world. I want to walk through the major ones, because the breadth of the parallel is where the real argument lives.

Skanderbeg: The Hero Who Broke Free

George Kastrioti was the son of an Albanian lord. He was taken into Ottoman service around 1414, possibly as early as age four, certainly by his late teens. The Ottomans renamed him Iskander, after Alexander the Great, and added the title Bey: Lord. He trained as a Janissary. He was good at it. He rose through the Ottoman military ranks and commanded troops in the Sultan’s campaigns.

Then, on a November night in 1443, during the Battle of Niš, Skanderbeg deserted.

He took three hundred Albanian soldiers with him. He rode south to the fortress of Krujë, presented a forged letter from Sultan Murad II to the garrison commander, and took the city. He raised the double-headed eagle of his family’s banner over the walls. On November 28, 1443, he proclaimed Albanian independence.

He would fight the Ottoman Empire for the next twenty-five years, until his death in 1468. He founded the League of Lezhë in 1444. Albanian chieftains who’d spent generations fighting each other joined his coalition and held off repeated Ottoman invasions. He won battles he should have lost, against armies that massively outnumbered his, using terrain and guerrilla tactics he’d learned while serving the very empire he was now fighting.

The name he reclaimed when he raised that banner was the one he was baptized with. George. In the Orthodox tradition, a baptismal name isn’t administrative. You’re named for a saint, placed under that saint’s protection, expected to imitate that life. George Kastrioti was named for the soldier-martyr whose icon hangs in churches across the Christian East: the one who kills the dragon. The Ottomans gave him Alexander. He took George back. Whether he thought of it in those theological terms I can’t say. But the act maps onto something the Church Fathers would have recognized immediately: a person reasserting the identity that no empire has the authority to revoke.

Everything the Ottomans gave him, he turned against them. Their military training. Their strategic doctrine. His understanding of their logistics and their weaknesses. Skanderbeg is the weapon that remembered who it was before. He is Albania’s national hero to this day.

Ballaban Badera: The Weapon That Didn’t Escape

Skanderbeg’s story has a mirror, and it’s uglier.

A boy named Michael was born Catholic in the village of Badera, in the Mat region of Albania. Son of Mark (some sources say Milosh) and Helena. Taken through the devshirme. Given the name Ballaban. He rose through the Ottoman ranks just as Skanderbeg had, and by the 1460s he’d reached the rank of Pasha under Sultan Mehmed II.

Here’s where it turns. Michael’s brother, Constantine, and Michael’s father, Mark, both served as soldiers under Skanderbeg’s command. They were fighting against Ottoman armies led by their own son and brother.

Ballaban commanded Ottoman forces against Skanderbeg at the Battle of Vajkal in April 1465. He came back as commanding general of the Second Siege of Krujë in 1466. He kept the fortress under siege for nearly a year. He died there, shot through the neck by an arquebus fired by Gjergj Aleksi, a city defender who was a hunter in civilian life.

Father and son on opposite sides of a siege wall. Brother against brother. The devshirme’s logic made flesh.

Ballaban is what happens when the conversion finishes. He is the completed product, the one who didn’t remember or didn’t let himself remember. In Ishinomori’s terms, Ballaban is every Kaijin that Kamen Rider has to fight: a former human, fully converted, sent back to destroy the one who escaped. The system made both Skanderbeg and Ballaban. The only difference was whether the final step held.

The Neomartyrs: The Conversion That Still Turned

The Orthodox Church documented a category of person who doesn’t appear in most political histories of the Ottoman period but is essential to any full accounting of what the devshirme did. They called them Neomartyrs. New Martyrs.

These were men and women, and some of them began as boys taken through forced conversion, who had been living as Muslims for years or decades. The conversion appeared to have held. Then they recanted.

Publicly. In front of witnesses. Knowing the penalty.

The Orthodox Church canonized several hundred of them between the fifteenth and nineteenth centuries. The scholar Nomikos Michael Vaporis documented them in Witnesses for Christ. The wave of public reconversions was especially concentrated between 1730 and 1820, the same decades when the Janissary Corps was at its most politically unstable. One of them, Demetrius the Neomartyr, was executed in 1803. They weren’t political revolutionaries. They weren’t military commanders. They were people who had apparently completed the conversion process and then, by their own choice, reversed it at the cost of their lives.

They washed out the stain of forced apostasy with an atoning death. That’s the theological category. That’s how the Church understood what they were doing.

In Ishinomori’s terms, the Neomartyrs are the strangest entry in the gallery. Skanderbeg escaped before the final step. Ballaban never looked back. The eight Christian boys resisted from inside and failed immediately. The Neomartyrs don’t fit any of those slots. They’re the completed product, the ones the process appeared to finish, who years or decades later reasserted the identity the system had buried. In Kamen Rider terms, they’re the Kaijin who, after full conversion and years of service, remembered the name on their baptismal certificate and said it out loud in public. No escape. No war. Just the image asserting itself, past everything the system spent years installing.

The Eight Christian Youths: The Resistance That Failed

Not every act of resistance produced a hero. Under Mehmed II, eight Christian boys taken into the Sultan’s personal service formed a pact to assassinate him. Their stated reason, preserved in the historical record: “None other than our great sorrow for our fathers and dear friends.”

The plot was discovered. Mehmed had the boys tortured over the course of a year, then beheaded.

No escape. No rebellion. Just eight boys who couldn’t bear what had been done to them and tried to strike at the source, and failed. The will to resist existed. The capacity to succeed didn’t. These are the stories the hero narrative requires us to forget, the ones who fought the system from inside and lost. In Ishinomori’s world, they’re the nameless captives in Shocker’s labs who resisted the brainwashing but never made it out.

Sokollu Mehmed Pasha: The Subversive Who Stayed

Sokollu Mehmed Pasha was born into a Serbian Orthodox family in Herzegovina. Taken through the devshirme as a boy. He never defected, never rebelled, never raised a banner or forged a letter. He did something stranger.

He climbed.

By 1565 he’d become Grand Vizier, the highest office in the Ottoman Empire below the Sultan himself. He held the position for fourteen years, until his assassination in 1579. And during that time he used the full weight of imperial authority to do something the devshirme was specifically designed to prevent: he preserved the identity the system had tried to erase.

In 1557, Sokollu secured the restoration of the Serbian Patriarchate of Peć, the institutional heart of Serbian Orthodoxy, and installed his own cousin Makarije Sokolović as Patriarch. He appointed both Muslim and Christian relatives to positions of power throughout the empire.

He never stopped being an Ottoman Grand Vizier. He never reclaimed his Serbian name. He operated entirely within the system. He used the tools the system gave him to protect the thing the system was built to destroy. If Skanderbeg is Kamen Rider, Sokollu is Dr. Midorikawa: the conscience inside the machine.

Patrona Halil: The Creation That Seized the Throne

In September 1730, an Albanian Janissary named Patrona Halil led a force of 12,000 Janissaries, most of them Albanian, in an open revolt in Istanbul. They forced the abdication of Sultan Ahmed III. They killed his Grand Vizier, Nevşehirli Damad Ibrahim Pasha. They ended the Tulip Period, an era of cultural flourishing and elite extravagance that the Janissaries saw as corrupt. They installed Mahmud I in Ahmed’s place.

Two months later, Mahmud I invited Patrona Halil to a meeting and had him killed.

Patrona Halil didn’t escape the system. He seized it. The creation deposed the creator. A stolen Albanian child, processed through the devshirme, trained as a Janissary, became powerful enough to remove a Sultan from his throne. The empire’s own production line had built something it couldn’t control.

Kabakçı Mustafa: The Creation That Refused to Be Replaced

In 1807, Sultan Selim III attempted to modernize the Ottoman military through the Nizam-ı Cedid, the “New Order,” which aimed to train a new European-style army that would eventually replace the Janissaries entirely.

The Janissaries saw this for what it was: their own obsolescence.

A Janissary officer named Kabakçı Mustafa led the revolt. The rebels forced Selim’s abdication, installed Mustafa IV, and demanded the abolition of every reformist institution Selim had created. The creation refused to be scrapped. The tool informed its maker that it wouldn’t be replaced by a newer model, and backed the refusal with violence.

The Dahijas: Rogue Weapons Catalyzing Revolution

By 1801, renegade Janissary commanders called the Dahijas had seized power in the Sanjak of Smederevo, roughly modern central Serbia. They ruled independently of both the Sultan and any Ottoman chain of command. They suspended Serbian rights, confiscated land, imposed forced labor.

In January 1804, the Dahijas ordered what became known as the Slaughter of the Knezes: the execution of somewhere between 70 and 150 Serbian village leaders and Orthodox priests, in a single coordinated purge designed to prevent a rebellion.

It caused one instead.

The First Serbian Uprising began on February 14, 1804. It was the start of the Serbian Revolution, the process that would eventually produce an independent Serbian state. The Dahijas were captured and beheaded by August of that year.

Janissaries, gone completely rogue, with no imperial authority behind them, committed an atrocity so severe it triggered a national liberation movement that eventually tore a piece of the empire away. The system’s products, uncontrolled, catalyzed the system’s dismemberment.

The Köprülü Family: The Weapons That Ruled

The Köprülü family came from Roshnik, a village near Berat in Albania. Taken through the devshirme. They produced six Grand Viziers between the mid-1600s and the early 1700s. Köprülü Mehmed Pasha, the first of them, served as Grand Vizier from 1656 to 1661 with a degree of autonomy no Grand Vizier before him had held. The Sultan granted him extraordinary powers, political rule without interference.

The Köprülü era represents the Ottoman Empire at its peak geographic expansion. The largest version of the empire, its greatest territorial reach across Europe, Asia Minor, and Africa, was administered by a family of stolen Albanian children.

The devshirme’s final paradox. The weapons didn’t just serve. They ran the whole operation.

The Auspicious Incident: The State Destroys Its Creation

By the seventeenth century, the devshirme had largely been abandoned as a recruitment method. The Janissary Corps had become entrenched, hereditary, politically untouchable, and increasingly resistant to any reform that might weaken their position.

On June 15, 1826, Sultan Mahmud II ended the argument. He surrounded the Janissary barracks in Istanbul with loyalist troops and artillery. He opened fire. Approximately 6,000 Janissaries were killed. The survivors were exiled, imprisoned, or executed in the following weeks. The corps was abolished. Mahmud II called it the “Auspicious Incident.” The Janissaries didn’t get to name it.

Read against everything above, the Auspicious Incident is the empire’s answer to every variation of the problem the devshirme created: escape, subversion, revolt, rogue action, seizure of power. When the system can no longer manage what it made, it reaches for the only tool left. Extermination.

Continue with Part II: The Weapon, where Shotaro Ishinomori builds the same machine in Skull Man and Kamen Rider.

Read Part II: The Weapon

If this is your first time here, subscribe below and continue to Part II. If someone else would want this argument, send them the entry point. That’s how the series grows.

Subscribe now