2D Post-Production
The render is post-produced — and there's an honest line
The pipeline ends in Photoshop — and so does the course, honestly. A professional render is almost always post-produced: entourage, sky, colour grading and compositing turn a flat raw render into a legible image, which is craft, not cheating. This unit sets the foundations right — raster versus vector, resolution and colour, the file formats and why a layered master beats saving everything as JPEG — and then draws the line where presenting well becomes lying.
Learning objectives
By the end of this lesson, you will be able to — mapped to the course outcomes for Computer Graphics:
Distinguish raster from vector, and explain why a raster cannot simply be enlarged.
Choose the right file format, and keep a layered master rather than saving everything as JPEG.
Post-produce a render non-destructively (layers, masks, adjustment layers, smart objects).
Stay on the honest side of the ethics line between presenting well and misrepresenting the space.
Raster vs vector
Raster is a fixed grid of pixels that pixelates when enlarged; vector is maths that scales infinitely. You cannot simply make a raster bigger — upsampling invents pixels rather than recovering detail. And resolution and colour only make sense relative to output.[1, 2]
Pixels versus maths
A RASTER image is a fixed grid of pixels (Photoshop, JPEG, PNG, and every render output); a VECTOR image is described mathematically as paths (Illustrator, SVG, PDF linework, CAD), so it scales to any size with no loss. A floor plan is vector — which is why it plots crisply at any scale. An interior board mixes both: the render is raster, but the annotations, logo and scale bar should be vector.[1]
File formats & the working file
Choose the format by the job — JPEG for screen photos, PNG for transparency, TIFF for print, PDF for boards, PSD for the layered working file, SVG for vector. Saving everything as JPEG degrades it on every re-save.[2]
Choose by the job
JPEG is lossy and small — photos/renders for screen, no transparency. PNG is lossless with transparency — logos, cut-out entourage, sharp graphics. TIFF is lossless — print masters and archives. PDF is a vector+raster container — boards and sharing (but can still hold low-res images). PSD is the layered WORKING file. SVG is vector — icons, diagrams, scale bars.[2]
The render is post-produced
The raw render is a starting point: pros composite entourage, sky, grading and render passes on top. The vital discipline is non-destructive editing — layers, masks, adjustment layers and smart objects, so nothing is thrown away.[3]
The raw render is a starting point
'The raw render straight out of the software is the final image' is a MYTH. A polished visual is almost always POST-PRODUCED: entourage (people, plants, styling) composited in as cut-outs; sky/background/view added; colour grading, contrast and white balance set; lens and atmosphere effects; and render PASSES (beauty, ambient occlusion, reflection, Z-depth) blended for control. This is normal, expected craft — finishing and communication, not cheating.[3]
The ethics line
Style, grade and add entourage freely — but faking impossible light, hiding a real column or inventing a view misrepresents the built space. The test: would the client feel deceived standing in the finished room?[4, 5]
Presenting well vs lying
Post-production is legitimate communication — a render is a representation, and enhancing legibility, mood and appeal is honest craft. The line is crossed when it MISREPRESENTS the space as it will be built: fake light that cannot physically exist, hiding a real structural column or duct, inventing a view through a window that faces a wall, or showing impossible/unspecified materials. Presenting well is telling the truth beautifully; lying is telling a beautiful thing that isn't true.[4]
At a glance
| Aspect | The fact | The folklore |
|---|---|---|
| Raster vs vector | Pixels (fixed, pixelate when enlarged) vs maths (scale infinitely) | The same thing |
| 'Just make it bigger in Photoshop' | Upsampling invents pixels — it can't recover detail | Adds real detail |
| Saving everything as JPEG | Lossy, degrades on every re-save, no transparency | Fine for a working file |
| The raw render | A starting point — pros post-produce it | The final deliverable |
| Editing | Non-destructive (mask, adjustment layers, smart objects) | Erase and flatten |
| Post-production's limit | Enhance freely; never misrepresent the built space | Anything goes if it looks good |
Key terms
Raster = a fixed grid of pixels (photos, renders) that pixelate when enlarged; vector = maths/paths (SVG, CAD, PDF linework) that scale infinitely.
Enlarging a raster by inventing new pixels via interpolation — it guesses, it never recovers detail that wasn't captured.
Screens are additive RGB (sRGB); print is subtractive CMYK with a smaller gamut — screen colours can shift on print without colour management.
Editing that preserves the original: layers, masks (hide, don't delete), adjustment layers and smart objects — the mark of a competent operator.
Entourage = composited people/plants/styling; passes = separate render outputs (AO, reflection, Z-depth) blended in post for control.
Would the client feel deceived standing in the finished room? Enhancement that survives the as-built comparison is fair; enhancement that collapses is misrepresentation.
Study task
Take one raw render and post-produce it non-destructively: add entourage (a person, a plant), a sky or view, and a colour grade — using layers, masks and adjustment layers so nothing is baked in. Keep the layered master as a PSD/TIFF and export a JPEG copy only as the final deliverable. Then write the one honest paragraph this course is really about: everything you enhanced that is fair communication, and the line you would refuse to cross — the fake light, the hidden column, the invented view — measured against the test, ‘would the client feel deceived standing in the finished room?’ Presenting the truth beautifully, and knowing where beauty becomes a lie, is the note the whole course ends on.
Self-assessment
1. Why can't you 'just make a small image bigger' in Photoshop?
2. Why is 'just save everything as JPEG' a bad habit?
3. What is true about a professional interior render?
4. Where does post-production cross the ethics line?
Recap
References & further reading
- [1]Adobe Photoshop Help — 'Image size and resolution' (resampling) and raster-vs-vector fundamentals; PPI vs DPI and screen-vs-print resolution. https://helpx.adobe.com/photoshop/using/image-size-resolution.html
- [2]Adobe — colour management and soft-proofing; file-format references (ISO JPEG, W3C PNG/SVG) for lossy vs lossless and vector. https://helpx.adobe.com/photoshop/using/color-management.html
- [3]Archviz post-production practice — render passes/AOVs, entourage compositing, and non-destructive editing (layers, masks, adjustment layers, smart objects). https://helpx.adobe.com/photoshop/using/nondestructive-editing.html
- [4]The ethics of visualisation — presenting well vs misrepresenting the built space; the 'as-built' honest test (professional-integrity framing).
- [5]GIMP and Krita (open-source, GPL), Photopea (free in-browser, opens PSD) and Affinity Photo (perpetual licence) — the skills transfer beyond Photoshop. https://www.gimp.org/
Further reading
- Adobe Photoshop Classroom in a Book (current release), Adobe Press.
- Deke McClelland, Photoshop CS Bible (Wiley) — fundamentals (note: the 'CS' era predates current Photoshop).
- Bruce Fraser, Chris Murphy & Fred Bunting, Real World Color Management (Peachpit).
Sources gathered and fact-checked June 2026. Published values vary by source, sample and method — treat as indicative and confirm against the cited standard before structural use.
The author
Amogh N P
Architect, interior designer, and creative polymath. Studio Matrx began in his notebooks — his vision of design made honest, useful, and open to everyone. Its Academy is written and taught in his memory, and free, forever.
More about Amogh →