What Files a Commercial Project Needs for Renderings

The fastest way to lose a week on a commercial visualization job is to start it with an incomplete package and find out during production rather than at kickoff. The reconciliation work happens either way. Starting without the inputs simply moves it later, onto the critical path, and usually under deadline.

This is a plain checklist of what a studio needs, why each item matters, what happens when it is missing, and which substitutes are acceptable. It is written for the person on the project side assembling the handover, which is often a project coordinator rather than the architect.

The short version

If you read nothing else: send the model or drawing set, a survey or boundary, the material direction, and photographs of the real site from every approach that matters. Those four cover most of the risk. Everything below is detail on why, and what to do when one of them does not exist yet.

1. The model or drawing set

The single most important input, and the one whose quality determines whether the first days go into production or into reconstruction.

Accepted formats: .dwg, .dxf, .rvt, .skp, .3dm, .pln and PDF. A native 3D model is best. A 2D set is workable. A PDF-only set is workable but slowest, because geometry has to be rebuilt rather than imported.

What matters more than format: internal consistency. A set where the plan, the elevations and the sections disagree is the normal condition of a live commercial project, and reconciling those disagreements is real work that has to happen before any image exists. Flagging known inconsistencies at handover is far cheaper than having them discovered mid-production.

If it does not exist yet: massing can be produced from a plan and a height. That is genuinely enough for early feasibility and zoning conversations, and it is the correct deliverable at that stage anyway.

What to include with the set

Floor plans for every distinct level, not just the typical. Elevations for every facade, including the ones nobody photographs, because they appear in aerials. At least one section, which resolves floor-to-floor heights faster than any other drawing. A roof plan if rooftop equipment or amenity is visible from any likely viewpoint. And the drawing revision number, so that everyone can tell later which version an image was built from.

2. Survey, boundary and levels

The input most often assumed to be optional and most often responsible for images that look subtly wrong.

A boundary tells the studio where the building actually sits on the parcel. Levels tell them what the ground does. A site that slopes two metres across its length produces a completely different eye-level view than a flat one, and approximating that is guesswork that shows.

What happens without it: the building is placed plausibly rather than accurately. In an aerial this is often invisible. At eye level from the sidewalk it is frequently obvious to anyone who knows the site, which includes exactly the audience at a planning hearing.

Acceptable substitutes: a civil site plan with spot levels, or a topographic survey. Public mapping data is a last resort and should be flagged as such in anything going to a hearing.

3. Site photography

The cheapest input to supply, the most frequently skipped, and the one with the largest effect on whether a final image looks like it belongs on that parcel.

Photographs do two jobs. They let the studio build accurate context rather than generic surroundings. And they let the camera be matched to a real viewpoint, which is what makes a before and after comparison credible.

What to shoot: every approach a person would actually use. The main street frontage. The corner conditions. The view from directly across the road. The adjacent buildings that will appear in frame. If entitlement is contested, also the view from the properties most likely to object, because that image will be requested eventually and it is cheaper to plan for it than to add it later.

How to shoot it: a phone is fine. Overcast light is better than harsh sun. Shoot wide, shoot level, and note roughly where you stood. Perfect photography is not the point; coverage is.

4. Materials and finish direction

Materials can be provisional. They cannot be absent.

A studio given no material direction will choose, and the choice will be generic because there is nothing else to work from. That produces a first draft that generates a full round of revisions which were entirely avoidable.

What helps: a schedule if one exists, product references or manufacturer names, precedent images of buildings with the intended feel, and any decisions that are already fixed even if the rest is not. "The base is dark brick and the upper floors are a light metal panel, glazing tint undecided" is enormously more useful than silence.

For interiors: flooring, wall finish, ceiling treatment and lighting intent, plus the furniture standard. Interior renders are judged at close range and generic furniture is the most common reason an interior reads as cheap.

5. Branding, signage and tenant identity

The item most often sent late, and the one where late arrival is most expensive.

Signage appears in most exterior views. Adding it after a set has been produced is not an edit, it is a production pass across every affected image. If the asset has a name, a logo, a wayfinding standard or known anchor tenants, send them at kickoff.

If tenants are unknown: say so, and agree what the signage will show. Generic placeholder, blank, or indicative. All three are legitimate; discovering the assumption at review is not.

6. The brief itself

Not a file, but the input that most changes the outcome. Four lines cover it.

Who is the audience for each image. Equity, hearing, broker, tenant, internal. This changes camera, treatment and level of polish more than any other instruction.

What each image has to prove. Address quality, capacity, circulation, finish level.

Your actual dates. Not a duration. The date the board packet is due.

How settled the design is. Honesty here lets a studio schedule and price properly, and it prevents the awkward conversation where a facade change arrives after final lighting.

What happens when inputs are missing

A realistic handover sequence

Projects that run smoothly tend to do the same four things.

Send everything at once, even if incomplete. A partial package delivered in one handover is far better than a complete one delivered across three weeks, because the studio can plan around known gaps but cannot plan around unknown ones.

Name what is missing and when it will exist. "Glazing tint decided by the twelfth" lets production sequence around it.

Nominate one point of contact. Feedback arriving from three people over four days produces contradictory revisions and burns any reasonable-revisions allowance.

Confirm the drawing revision. Write it down. When a question arises three months later about why an image shows something the current set does not, this is the answer.

What the studio should tell you in return

The obligation runs both ways, and a studio that does not volunteer these is worth pressing.

What they will assume where information is absent. What they need next and by when. Which of your inputs is weakest and what improving it would change. And what they cannot do with what they have, stated before production rather than after.

On our side the published turnaround is first visuals in 48 to 72 hours and a plan set in 5 to 7 days, with reasonable revisions included at no extra charge. Those figures assume a workable input package. When the package is thin, the honest answer is that the first days go into reconstruction, and we would rather say that at kickoff than absorb it silently and miss a date.

One thing worth stating plainly

None of this input preparation makes a project more likely to be approved. Visuals are the material used in entitlement and approval presentations; no studio obtains approvals and none can guarantee them. What good inputs do is narrower and more reliable: they make the images accurate, they keep the schedule honest, and they prevent the category of rework that consumes budgets without anybody deciding to spend it.

Inputs by asset class

The core list is universal. What sits on top of it changes by building type, and briefing the wrong extras wastes time on both sides.

Office and mixed-use

Core and circulation drawings matter more than for any other type, because every interior view depends on them. Add the amenity program, the ground floor tenancy intent, and the furniture standard if speculative suites are in scope. If the ground floor is active retail, treat it as a second building type and brief it separately.

Industrial and logistics

The site plan with dimensions is the primary document rather than the elevations. Add the dock and trailer layout, clear height, column grid, truck circulation and the direction of the nearest interchange. If the project is build-to-suit, send whatever is known about the tenant operation: racking, equipment, vehicle mix, shift pattern. If speculative, state the target tenant profile instead, because it decides whether interiors read as empty and flexible or fitted and specific.

Retail

Signage and tenant identity move from optional to central. Add the tenant mix if known, the signage standard, parking and pedestrian flow, and whether evening lighting is part of the story. Retail is judged on activity more than on architecture, so the entourage brief matters unusually much.

Hospitality and amenity-heavy assets

The most input-hungry category. Interior finish schedules, lighting intent, FF&E direction and the operator standard if there is one. These are interior-weighted and judged at close range, so generic substitutions are visible immediately.

Common handover problems and how to avoid them

The model is federated across consultants. Normal on commercial projects, and workable, but say so. A studio expecting one file and receiving five linked models from four consultants loses a day just establishing what is authoritative.

The set is a mix of revisions. Plans from one issue, elevations from another. This is the single most common source of geometry that does not reconcile, and it is invisible until production. Check the revision block on every drawing before sending.

Nobody can confirm the material palette. Usually because the decision genuinely has not been made. Say that, propose a working assumption, and agree it in writing. Silence produces a generic default and a wasted round.

The site photographs are from a marketing shoot. Beautiful, dramatically lit, and shot from angles nobody walks. Useful for mood, close to useless for context matching. Plain, level, overcast phone photographs from real approaches are worth more.

Inputs arrive in pieces over three weeks. The most expensive pattern of all, because production cannot be scheduled around it. One handover with declared gaps beats three handovers with none.

A checklist you can send internally

Copy this into an email to whoever is assembling the package.

Model or drawing set in .dwg, .dxf, .rvt, .skp, .3dm, .pln or PDF, with the revision noted. Floor plans for every distinct level. Elevations for every facade. At least one section. Roof plan if rooftop elements are visible. Survey, boundary or civil plan with levels. Site photographs from every real approach, plus from any property likely to object. Material and finish direction, provisional is fine. Interior finish and furniture standard if interiors are in scope. Branding, signage and tenant identity. A one-paragraph brief naming the audience and the date for each image. And a single named point of contact for feedback.

Why this is worth the hour it takes

Assembling this package properly takes somebody roughly an hour. The alternative is not that the work disappears, it is that the same reconciliation happens later, performed by people who do not know the project, using assumptions, during the production window that was supposed to produce images.

That is where schedule loss comes from on commercial visualization jobs. Almost never from rendering time. Almost always from starting production on a package that could not support it.

Frequently asked questions

What file formats do rendering studios accept?

Commonly .dwg, .dxf, .rvt, .skp, .3dm, .pln and PDF. A native 3D model is fastest. A PDF-only set is workable but adds days because geometry has to be rebuilt.

Can renderings start without a survey?

Yes, but the building gets placed by estimate rather than accurately. On sloping or urban sites this shows at eye level, which is exactly where a planning audience looks.

Why does the studio want site photographs?

To build accurate context instead of generic surroundings, and to match cameras to real viewpoints so that existing versus proposed comparisons are credible.

What if materials are not decided yet?

Send whatever is fixed and say what is not. Provisional direction is far better than silence, which forces the studio to choose generically and generates an avoidable revision round.

When should signage and branding be sent?

At kickoff. Signage appears in most exterior views, so adding it later is a production pass across the whole set rather than a small edit.