Iterate, Iterate, Iterate
Apparently the real workflow is admitting that the workflow keeps changing.
That sounds annoying because it is. It is also useful. These posts started in 2018 as a way to document the moving target: apps, backups, exports, culling, and the recurring question of where the archive actually lives.
So yes: iterate, iterate, iterate.
The difference this year is that the editor is no longer the whole question. The editor matters, but the more important question now is whether the archive stays portable, backed up, and publishable without being trapped inside one company’s app, cloud library, or subscription plan.
Photomator Was Close
I still love Photomator. The desktop/mobile workflow is genuinely good. Moving between the Mac, iPad, and Apple Photos library with so little friction solved a real problem. The auto-masking is fast: it gets out of the way before the adjustment becomes a whole event.
I also still like the idea of Photomator, especially after Apple acquired Pixelmator. The first photo editor I used seriously was Aperture, so some stubborn part of my brain is still waiting for the Aperture reincarnation. That has not happened yet.
The problem was not that Photomator became bad. It did not. It is a good app. The problem was that the results stopped feeling strong enough to make it the center of the whole workflow.
The Capture One Detour
After that, I tried Capture One. I signed up for the trial, liked a lot of what I saw, and then, in a truly elegant piece of timing, realized about a week into a yearly subscription I had already paid for that it was not the right fit.
Some of this is taste, some of it is irritation, and some of it is the ongoing daydream that if I buy the Hasselblad X2D II 100C, some locked portion of my photographic destiny will open up and explain itself. It will not.
Capture One made one useful thing clear: I had been over-weighting file syncing as a requirement. If an editor did not have some version of a sync story, I dismissed it. Once that rule loosened, other options became possible.
Reprocessing the Catalog, Again
There is a ridiculous pleasure in re-processing the catalog of selects every time the software changes.
It is also a pain in the ass.
The 2025 Photomator move already forced this realization. Losing old Lightroom edits sounded awful until I treated it as a chance to revisit the images with my current eye. Taste changes. What I liked ten years ago is not always what I like now.
That part is worthwhile. Re-editing old work is a way to see your taste changing in public. The practical problem is different: when edits live inside app-specific systems, switching costs become archive costs. I do not want the archive to depend on my memory of which app held which truth.
Why DxO PhotoLab 9 Stuck
With the syncing requirement relaxed, I tried DxO PhotoLab 9.
Fast-forward to now: I am using it exclusively for photo processing.
The AI masking is not as fast as Photomator’s, but it works. I need it to help me dodge and burn without turning the edit into a little software negotiation.
The part that stuck is the RAW processing. DxO’s camera and lens profiles, the DxO Modules, feel like boring technical work that shows up in the file instead of the marketing copy. Noise reduction, lens correction, color, and detail all feel less like I am fighting the RAW file into shape.
The multi-export workflow helps too. One processed file can move out in the sizes and formats I need without making export management a second hobby. DxO fits because it makes the working file look good and lets me get finished images out cleanly.
Sidecars Changed the Trust Model
The bigger change is not the slider layout. It is the sidecar files.
DxO stores edits in .dop sidecar files beside the original RAW files. They are still DxO-specific. I am not pretending some other editor can open a .dop file and magically recreate the edit.
The point is that the instructions live next to the source file.
That changes how much I trust the archive. With Photomator, edits lived in a specific iCloud folder, and in my experience that was not always as reliable as I wanted. The archive felt too dependent on the invisible plumbing of one app and one cloud folder.
With DxO, the RAW and the edit sidecar travel together. If I move the folder, the edit instructions move with it. If I back up the RAW, I back up the .dop file beside it. If I restore a folder later, I am restoring the working context, not just the original negative and a vague memory.
This is nerd housekeeping, sure, but not for its own sake. It means the archive is easier to understand later, and I am less likely to re-process everything because the edit state lived somewhere too clever for its own good.
Cold Storage Got Simpler
The backup side changed too.
I moved the RAW archive from Google Cloud to S3 Glacier Deep Archive. Right now, it is costing a little over $5/month for almost 6TB of backups.
That is absurdly cheap, with the obvious caveat that Glacier Deep Archive is not a fast working drive. I am not using it as an active photo server. It is cold storage. The point is that if something truly dumb happens locally, the files that should survive have a durable offsite copy.
This is where archive discipline matters more than tool preference. The RAWs matter. The .dop files matter. The exported selects matter. The active working experience should be pleasant, but the archive should survive my preferences changing again.
The Archive Has a Public Home Again
The other piece this year is that the selected work has a public home again.
I used Claude Code to build a custom version of Sam Becker’s EXIF Photo Blog and now have it running at photo.banast.as.
Claude Code is not the point here. It was useful, but the important part is that the archive has a controlled public surface again. The selected images do not only live inside Instagram, Apple Photos, Lightroom, DxO, or some old archive system I stopped maintaining three workflows ago.
This matters to me more than it probably should. A public photo archive is not just a gallery. It is a canonical place for the work to land outside the editor and outside the social feed. It connects back to why I keep building these little personal systems: I like having my own infrastructure for the things I care about.
Current Snapshot
So this is the 2026 version.
Photomator was close. Capture One was an expensive detour. DxO PhotoLab 9 is the editor that stuck. The RAW files now stay with their .dop sidecars. The cold archive sits in S3 Glacier Deep Archive. The selected images have a public home at photo.banast.as.
This is not final truth. None of these workflow posts are. They are versioned snapshots of what worked at a given moment, with the compromises still visible.
The useful shift is that the workflow is finally less about chasing the perfect app and more about keeping the archive legible: RAW files, sidecars, cold storage, and a place to publish.