How do I take good progress photos?

Consistency beats everything: same light, same time of day, same distance, same poses, same state of fed-ness. An inconsistent series produces apparent changes that are lighting and hydration, and those are indistinguishable from real ones.

Are my progress photos uploaded anywhere?

No. Progress photos are the one category of data explicitly excluded from sync — they never leave the device and never reach a server. A copy is written to an album in your own photo library, which is your storage rather than the app's.

Why does the app save photos to my camera roll?

Because iOS and Android both erase everything an app stores when the app is deleted. A photo that lived only in app storage is guaranteed to be lost on the next reinstall, and a photo series is the one record you genuinely cannot reconstruct afterwards.

How often should I take progress photos?

Every one to two weeks for the record, and compare across eight to twelve weeks. Visible change over a fortnight is mostly lighting and water; over three months it is real.

Can I lock my progress photos?

Yes, behind your device's biometric authentication. They are also never uploaded, so the only copies that exist are on your phone and in your own photo library.

Progress Photos: The Only Record That Survives Your Memory

Where the photos actually live

Two copies, and neither of them is on a server.

In app storage, as the working copy the gallery reads.

In a "Base Progress" album in your own photo library, as the archive.

The second one exists for a specific reason. iOS and Android both destroy an app's entire sandbox when the app is deleted — a photo stored only there is guaranteed to be lost on the next reinstall, which is exactly what kept happening before the archive existed.

The photo library is not part of that sandbox. Deleting the app does not touch it. It is your album, not the app's — if you have cloud photo sync turned on, those photos sync the way the rest of your camera roll does, because that is your setting.

The filename carries the date and the pose: base-2026-08-25-most-muscular.jpg. That is deliberate. If the database is ever gone as well — a reinstall with no account, a phone restored from nothing — the album is still a complete, self-describing archive that the gallery can be rebuilt from. A folder of IMG_4471.jpg could not.

Progress photos are never uploaded and never sent to a server. They are the one category of data explicitly excluded from sync. The full disclosure.

Consistency is the whole thing

An inconsistent photo series is worse than no series, because it produces apparent changes that are lighting, hydration and time of day — and those are indistinguishable from real ones.

The variables that matter, in rough order of how much damage they do:

Variable	Why it matters
Light	The largest single factor. Overhead light versus front light changes apparent definition enormously
Time of day	Morning fasted and evening fed are different bodies on the same day
Distance and angle	Closer looks bigger. A slight downward angle changes everything
Pose	Same poses, held the same way, same tension
Clothing	Same, and minimal
Pump	Post-session and pre-session are different photos

The post-workout prompt exists because post-session is the most repeatable moment in a week: same time of day, same state, same place, same everything if you shoot in the same corner of the same gym.

The pose set

A fixed set rather than a free-form gallery, because a comparison needs the same input twice:

Front relaxed — the baseline
Back relaxed
Side relaxed — the one that shows the waist
Front double biceps — the one people take
Most muscular

Relaxed poses are the informative ones. Flexed poses are the satisfying ones. Both are in the set, and the comparison view defaults to relaxed for a reason: it is much harder to unconsciously flex harder over time in a relaxed pose than in a flexed one, which is the most common way a photo series lies to its owner.

The timescale

Take them every one to two weeks. Compare across eight to twelve.

Visible change over a fortnight is mostly lighting and water. Over three months it is real. The comparison view puts two dates side by side rather than showing a scrolling gallery, because a gallery invites the fortnightly comparison and the fortnightly comparison is noise.

What photos catch that the numbers do not

Regional distribution. Trend weight up four pounds tells you the mass arrived. It does not tell you where.

Recomposition at stable weight. The specific case where the scale is flat and the waist is flat and something is nevertheless happening. Slow, real, and completely invisible in numbers over a short window.

The thing you cannot measure. Posture, structure, how the tissue actually sits. Not quantifiable and not nothing.

They also correct the memory problem, which is the underrated one. Your sense of what you looked like three months ago is not a record. It is a reconstruction, and it is systematically biased.

What the screen does not do

No AI body fat estimate from the photo. The error on photo-based estimation is wide and lighting-dependent, and a confident wrong number about your body composition is worse than no number.

No sharing. There is no feed, no export-to-social, no comparison against anyone else.

No before-and-after template. That is a marketing artefact and this is a training log.

No progress score. A photo is qualitative. Quantifying it would be inventing precision.

Locking them

Photos can be locked behind your device's biometric authentication. Combined with never being uploaded, that means the only copies that exist are on your phone and in your own photo library, and the app's view of them requires your face or your fingerprint.

What to do next

Take the first set today, in the light you can reproduce, and set a fortnightly reminder. The value of a photo series is entirely in its length, and the only way to have a two-year one is to start.