How Revision Rounds Really Work

Revisions are the part of a rendering project that clients understand least and worry about most. Everyone has heard a story about a job that spiraled into endless rounds of changes, and everyone wants to know, before they sign, how many rounds they get and what happens when they run out. Those are the right questions. But the honest answer is more useful than a number, because the two things that actually decide how the review stage goes are what counts as a revision in the first place, and how you give your feedback.

Get those two right and revisions are a short, finite, calm stage at the end of the project. Get them wrong and the same allowance evaporates into a fog of scattered notes and nobody can say where it went. Here is how the stage really works, drawn from how it actually plays out on real projects.

Two stages, two rounds each

Most studios quote you a round count and leave it there. The structure underneath is more specific, and worth understanding, because it is where a lot of the value hides. Revisions come in at two separate stages of the project, with two rounds available at each. There are two rounds at the white-model stage, where you are adjusting framing, composition, and camera angle before any color exists, and two rounds at the photoreal stage, where you are refining lighting, materials, and atmosphere on the finished image.

When a studio says two rounds out loud, it usually means the photoreal rounds, because those are the ones on the beautiful images everyone pictures. The white-model rounds are included too, and they are real, and they are the ones almost nobody mentions. That matters because the cheapest, most powerful changes you can make to an image, moving the camera, recomposing the frame, happen in those under-advertised early rounds. The reason the split exists is technical, and a producer on a Montego Bay resort project put it to the client exactly: once the color stage begins, the camera angle and the lighting are locked, a daylight scene stays daylight, and only the finishes and furniture remain open, across two rounds. The stage structure is not bureaucracy. It is a map of what is still changeable when.

A correction is not a revision

This is the single most important distinction in the whole stage, and the one that keeps it sane. A correction is us fixing something that does not match the brief. If a material is wrong, a dimension is off, or something we built does not match the files you gave us, that is corrected as a matter of course and it does not count against your rounds. A revision is a refinement you request after approval: a new angle you have decided you want, a material swapped for a different one, a lighting change. That is a genuine change of direction, and it is what the rounds are for.

The rule we hold ourselves to is that we tell you which one a request is before we start it, so there are never surprises on the invoice. You can watch this happen in the real exchanges. On one project, a producer flagged that a stone placement the client wanted would be an additional revision, and said so up front before touching it. On another, when a landscaping change came in, the team told the client plainly that small tweaks like that they would do free of charge, of course. And on a third, the client themselves acknowledged the line, telling us to check on the cost of the additional revision that the landscaping change on the front view would require. Nobody was ambushed, because the category of the request was named out loud before any work happened. That transparency is what stops the correction-versus-revision line from becoming an argument.

Rounds are a finite, counted budget

Rounds are real, which means they are also finite, and a good studio tracks them openly rather than springing the limit on you at the end. On a residential project that ran across many images, our team kept a shared status sheet for every render and used it to warn the client, well before the finish line, that the project was going to need close to all of the remaining revisions. That is the opposite of a gotcha. It is the studio showing you the fuel gauge so you can spend what is left on the changes that matter most. On another project, a producer told the client plainly that a particular adjustment would count as one revision round and that two rounds still remained. The balance is not a secret held until the end. It is a running number you are told as you spend it.

There is give in the system for the right reasons. When a project has recurring potential, a further photoreal round is available as a goodwill concession, and we have made changes free of charge as a courtesy more than once to keep a relationship strong. But goodwill is a decision, not a default, and you can see the studio thinking about it in the open: on one job, a producer noted that this was already the second round of white models and said they would check with the team about whether there would be a charge. The allowance is generous and it is tracked. Both things are true, and knowing they are both true is what lets you plan your feedback instead of rationing it in fear.

Speed can cost a round too, and it is worth deciding that trade out loud. On a job that had to be expedited to hit a hard deadline, the plan was explicitly to skip a round of revisions, because there was no time to run it before the final delivery, and giving up that round was part of what paid for the rush. A compressed schedule and a full set of rounds are not always available at the same time. Better to choose which one you need than to discover the trade halfway through.

How feedback should arrive

Here is the lever you actually control, and it matters more than the round count: how your feedback reaches us. The single biggest predictor of a smooth review stage is consolidated feedback through one channel, one clear set of notes per round, ideally filtered through one person on your side. The best clients organize this themselves. On a project with around fifty images in play, the client’s project lead set up a single review call for the express purpose, in his words, of getting consolidated feedback in one quick hit, so our team could take all the information and keep rolling. On another, the client offered to pick his images and send one email with three or four attachments rather than a trickle of messages. That habit alone does more to protect your rounds than anything we can do on our side. The strongest version of it is a single person who owns the funnel. One client described their own role on the project in exactly those terms: their owner was very particular but not always good at putting what he wanted into words, so they acted as the liaison, pinning him down to specifics and handing us one clear brief instead of a moving target. On a job with several stakeholders, that person is worth their weight in rounds.

The channel matters less than the consolidation. On one project the client asked whether she could give her comments as a marked-up PDF instead of using our review software, because she found it easier to gather her own team’s notes in one place first. The answer was whichever is easier for you, as long as all the information lands together, because we consolidate everything into one review for our team regardless. We are channel-agnostic and consolidation-insistent, because a round only stays a single round if it arrives as a single, settled set of notes. That is also why, before spending a delivery, our producers will often ask you to confirm that a given comment is your only comment: if it is, we move to final, and if it is not, we would rather have the rest now than burn a round on a partial list. This hardens into a standard step before the last round. As the next round becomes the final version, we ask for all remaining feedback first, because anything that lands after the final delivery starts a new one. It is not a trap. It is the single moment where gathering up everything you have left saves you a round you would otherwise spend on an afterthought.

How it goes wrong, and how the process protects you

The failure mode is the mirror image of all this: scattered notes from several people, arriving at different times, sometimes contradicting each other. It is almost always a client-side habit rather than a studio problem, and clients often see it in themselves. On a fast-moving development, one stakeholder admitted the direction coming from their side could seem conflicting, wanting the facade brighter but the sun in the back at the same time, and reasoned their way to the fix, that each rendering almost needed to be treated as its own thing. When feedback pulls in two directions, no number of rounds is enough, because each round spends itself reconciling the contradiction instead of improving the image.

Some clients use the round structure deliberately to keep their own side in line. One design-firm principal asked us to write the revision terms into the quote for exactly that reason, so he could put them in front of his partners and tell them, in his words, that they all needed to behave or it was going to cost them. He was candid about the chaos he was guarding against: a principal who would sit there all day saying move this, shift that. A defined, finite round is not only a limit on the studio. It is a lever that forces a client’s own stakeholders to consolidate before their indecision turns expensive.

The structural protection against all of this is the white-model gate. Approving the camera and composition in gray, before any color, means your expensive photoreal rounds are never spent fixing a framing problem you could have caught for free. On that same development, earlier batches had been produced by skipping the white models to save time, and for the new scope the team walked the process back to the full sequence: model, then a round of white models to pick the cameras, a second if needed, and only then color. It is no accident that a client on a high-stakes townhouse project told us he wanted to get it right within the rounds because, as he put it, they had one shot at these renderings. The process exists so that one shot is enough.

What to take into your next project

Two habits will do more for your review stage than any contract clause. Nominate one person to gather and send your team’s feedback as a single, consolidated set per round, and treat the white-model previews as the moment to get framing exactly right, not an afterthought. Do those two things and the rounds you are given will be more than enough. Skip them, and the most generous allowance in the industry will still feel tight.

For why the camera gets locked first, read why we white-model every angle before adding color, and for where this stage sits in the whole project, see the architectural visualization process, stage by stage. For a finished set that came through its review cleanly, look at Austin Surf Club. If you have a project coming up, that is a conversation we are happy to have.