What File Size Should Client Proofs Be?

A proofing gallery exists to answer one question. Which photographs does the client want finished and delivered. That is the entire job.

Everything else a photograph file can be good for belongs to a different step. Printing is a delivery concern. Re-editing is your concern. Archiving is a storage concern and usually not the client's at all. None of those happen during a proofing round.

Once the job is that narrow, file weight stops being a quality question and becomes a fitness question. The right size is the smallest one that lets the client answer the question confidently. Anything past that point is weight the client carries for no return.

Watch a client work through a gallery and they are doing four things, no more.

• Keep or cut. Is this frame worth finishing at all, or is it a moment they do not care about.
• Pick the best of a cluster. Three near-identical frames of the same pose, one of which has the expression they want.
• React to the whole picture. The crop, the light, the way they look, whether their mother is in it.
• Flag a change. The exit sign, the stray strap, the cousin blinking in the back row.

Every one of those is made at viewing size. A client choosing between two smiles is comparing faces on a screen they are holding at arm's length. A client spotting an exit sign is looking at the frame, not into it.

So the honest test for a proof is whether it supports those four judgements. A file sized for fast viewing supports all four. That is why proofing galleries across the industry publish a lighter file than the one they deliver, and why the practice is not a compromise.

A larger proof does not change which smile is better. It changes how long the client waits to see either of them.

There are three costs, and they all land on the client rather than on you, which is why they are easy to miss. The first is time. A gallery that takes a while to become viewable is a gallery that gets closed, and a closed gallery is a proofing round that stalls for a week. The second is data. Clients open these links on phones, on plans with limits, and a set that is heavier than it needs to be is a set that some of them will postpone until they are on wifi, which often means never. The third is commercial. A proof large enough to be useful as a finished file is a file you have released before the selection is made and, frequently, before the invoice is settled.

Set against that, the benefit of extra weight during proofing is detail the client is not looking for at a stage where they are not looking for it.

Critical sharpness. A proof sized for fast viewing cannot tell anyone whether focus landed on the near eye or drifted to the ear behind it.

At viewing scale, three different faults look the same. A slight focus miss reads as soft. A trace of subject motion reads as soft. Compression smoothing a fine texture reads as soft. You cannot separate them from a proof, and neither can your client. If a frame is marginally out of focus, a light proof will quietly hide it, and the first person to discover the problem will be whoever opens the delivered file.

The consequence is a workflow rule rather than a file size rule. Cull for technical faults yourself, at full magnification, before anything reaches the gallery. Focus, motion, eyes open. That pass is not the client's job and a proofing gallery is not the instrument for it. Sending a doubtful frame to a client because you were not sure is how a doubtful frame gets chosen.

There is a second version of this worth naming. Some commercial work asks the client to check technical quality as part of approval, for instance product, architectural, reproduction or catalogue work where a reviewer is checking detail rather than choosing favourites. That review is a different task from a proofing round. Send those few frames at full size, directly, and keep the gallery for the choosing.

Nothing above argues that clients should get small files. It argues that the large files belong at a different moment, in a different container, for a different reason.

The final handoff is that moment. It happens once, at the end, to a client who is expecting it, who has chosen what is in it, and who is far more willing to wait for a download they asked for than for a gallery they were only browsing. That is the right place to spend a client's patience and a client's bandwidth.

Separating the two lets each one be good at its own job. The gallery stays light, fast and permanent, so the client can go back to it a year later from a phone. The delivery is heavy and temporary, because a big file only has to exist while it is being collected. Trying to make one file do both jobs produces a gallery that is slow for the whole of its life in exchange for a convenience the client uses once.

Size the proof for the largest screen that will judge it, and not for paper. Paper gets its own file in the handoff.

Judge the result by behaviour rather than by numbers. Open your own gallery on a phone, on a cellular connection, not on the studio wifi next to the router. If the first photograph is there before you have finished thinking about it, the file is right. If you notice a wait, it is not.

Keep the whole set consistent. A gallery is only as fast as its heaviest frames, and one oversized export in a set of three hundred is the one the client happens to land on. This is a good argument for letting the platform do the sizing rather than exporting by hand, because a tool applies the same treatment to every photograph and a person does not.

Then do the arithmetic, because per-photo weight is only half the picture. What actually travels is the weight of a photograph multiplied by how many of them the client scrolls through.

A gallery is loaded progressively rather than all at once, so a client never waits for the full figure. The table is there to make the multiplication visible. Doubling the weight of a proof doubles that column, and the column is what a tight cull shrinks before any export setting does.

The sizing decision is made for you. Photos are compressed in the browser as they upload, into proof JPEGs of about 0.6 MB each, and those galleries stay hosted for as long as the account exists. There are no storage tiers to choose between and no per-gallery fees.

Full-resolution delivery is deliberately a separate step, which is the separation this page argues for. You upload a full-resolution ZIP for a gallery, it stays live for 3 days per handoff, and you can re-upload it any time the client needs it again. The proofing gallery itself does not expire when that window closes.

The ceilings are written down rather than hidden because they are the reason the price is flat. The service hosts JPEGs. It does not warehouse RAW files or master video, and a proofing gallery is not an archive. If you work in Lightroom Classic, the two way plugin publishes proofs straight from the catalog and pulls the client's picks back into it, so the sizing, the upload and the selection all happen without a manual export in the middle.

Every figure on this page is a fact about this product rather than an external statistic, so the sources are this product's own pages and its code. No third-party number is quoted here.

• clientgallery.io pricing and plan page, clientgallery.io. Cited for: one plan at $10 a month or $100 a year, no storage tiers, no per-gallery fees, hosted galleries that do not expire. Stated from the product itself, confirmed 11 September 2026.
• clientgallery.io product behaviour: proof compression in the browser at about 0.6 MB a photo, hosting for as long as the account exists, and the fair-use ceilings of 1,000 photos per gallery, 250 GB of proofs per paid account and 5 GB on a free account. Stated from the product itself, confirmed 11 September 2026.
• clientgallery.io full-resolution delivery limits, read in the product code on 11 September 2026. Cited for: a ZIP that stays live 3 days per handoff, and 100 GB of standing ZIPs per account.
• How to Send Full-Resolution Photos to a Client. Checked 11 September 2026. Cited for: how the full-resolution handoff works alongside a proofing gallery.
• Client Photo Proofing: The Complete Guide. Checked 11 September 2026. Cited for: what a proofing round is for and how clients mark their picks.