Photograph Material References So Texture and Shine Can Actually Be Compared
Photograph Material References So Texture and Shine Can Actually Be Compared
You photograph a swatch of cloth and a square of foil on the same table. The cloth frame shows a crisp weave. The foil frame shows a flat gray rectangle. Side by side, the pair "proves" that the cloth has surface and the foil doesn't.
What it actually documents is that the cloth got side light and the foil got light from a position where its shine bounced away from the lens. Same materials, different photographs, opposite conclusion, and nothing on the sheet tells you which of those facts you are looking at.
A material reference becomes readable when the view and the illumination that reveal the surface are part of the record. That's the whole job. Nothing about it requires a studio, and almost none of it is about the camera.
Decide what the picture has to answer
"Texture" is three or four separate questions wearing one coat.
Color relationship is how two samples sit against each other — warmer, cooler, lighter, darker as a pair.
Surface relief is height variation: weave, grain, fiber, embossing, nap, crinkle. You see it because tiny shadows fall in the low spots.
Gloss is how concentrated the reflection is. A glossy surface throws back a tight, defined reflection of whatever light it's facing. A matte surface spreads that same energy into a broad, soft glow. Gloss is not brightness. A dull sample under a hard lamp can measure brighter than a shiny one photographed away from its reflection angle.
Translucency is light passing through the material from behind, which reveals internal structure, thin spots, and fiber density.
These four respond to different setups. A single handsome photograph will not explain all of them, and trying to make it do so is how the accidental-variable problem starts.
So name the treatment question in a form a photograph can settle. "Does the foil read as distinctly glossier than the coated paper at the size the treatment will show them?" is answerable. "Which one looks more premium?" is a design judgment. The camera has no opinion on it, and a reference sheet that pretends otherwise is really a mood board wearing a lab coat.
Give the samples an identity, then hold everything else still
Use ordinary, replaceable material. A second sheet of the same paper, a spare square of cloth, a fresh piece of foil. If the sample is the actual irreplaceable thing — the only print, the client's one prototype — this exercise is the wrong tool for it, and that isn't a technicality you can talk your way around with careful handling. Substitute, or do something else.
Label every sample with something that appears in the frame: a card reading A, B, C. Record which face you photographed, which way it's oriented, and its approximate size. Note the orientation out loud, because weave direction, nap direction, and the grain of an embossed paper all do something different when they turn ninety degrees, and six weeks later you will not remember which way B was facing.
Then decide what you are not testing, and freeze it.
Camera position — tripod, on a marked spot. Focal length, framing, and the distance to the sample. Exposure in manual. White balance fixed, not auto. Whatever processing you apply, apply identically to every frame in the same comparison, or to none.
That last point matters more than it sounds. If auto-exposure is running, the darker cloth and the brighter paper will each get their own correction, and the brightness difference you see in the sheet is the camera's decision, not the material's. Manual exposure across the whole comparison is what makes brightness comparable at all.
Keep the originals. The selected, processed, resized frame is the deliverable; the untouched capture is the evidence. You will want to go back.
What each light position actually shows
The Canadian Conservation Institute's page on lighting techniques for photographing heritage objects — the HTML transcript, which is the part available here; the video and its illustrations haven't been watched — lays out normal, specular, raking, and transmitted illumination as distinct relationships between light, surface, and viewer. The useful thing to borrow is that each one reveals a different property. The comparison discipline below is a working method for a commercial treatment, not something the page is proposing.
Frontal or normal light — source near the camera axis, camera looking straight at the sample. This is your baseline, and it's the view that shows color relationships and overall form. It flattens fine relief, because light arriving straight on fills the micro-shadows. On a glossy sample it can produce a large flare, because the mirror direction of a light sitting at the camera points straight back at the camera.
Raking light — the source brought down almost parallel to the surface, from the side. Now the light catches the tops of fibers and casts shadows in the valleys, and relief appears. Change the direction and you're lighting fibers from a different side, which is why orientation has to be recorded. Note what raking light does to a glossy sample: the mirror direction points off to the side, away from the lens, so almost no reflected energy reaches the camera. The sample goes dark. It hasn't stopped being glossy; you've simply pointed the reflection somewhere else.
A specular or sheen view — put the source where its reflection can reach the lens. The simplest version is the light right at the camera. The cleaner version is a matched pair of angles: source 30° off the surface normal on one side, camera 30° off the normal on the other. Mirror angles are mirror angles. If the two don't match, you're photographing the edge of the highlight rather than its center.
Two warnings about this view. First, what you're photographing on a shiny surface is largely a picture of your light source, so the size and shape of the highlight says as much about the lamp as about the material — hold the source identical across samples or the comparison is a comparison of your equipment. Second, a shiny sample reflects the room: your body, a window, a doorway, the edge of the board. Before reading a dark band in the foil as a property of the foil, check what it's reflecting. And look at both faces of any foil you use; if they differ, that difference belongs in the record rather than being averaged away.
Transmitted light — source behind the sample, sample lifted off any opaque backing. Include it only when translucency is the question. If you make a set of paper, cloth, and foil because you want to know about sheen and weave depth, no sample raises a translucency question, and no transmitted frame should be made. Record why it's absent, so a later reader doesn't mistake a decision for an oversight.
One more constraint: within any single comparison, keep the geometry identical for every sample. Across different views, geometry is allowed to change, because that's the point of having views — but each change gets written down.
A comparison that lies, and its remake
Nobody has photographed these samples. This is a setup to run, not a session being reported, and the numbers are a plan.
Three samples on a neutral board with a ruler lying in frame: A, a sheet of matte inkjet paper. B, a square of textured cloth. C, a square of kitchen foil. All replaceable, all labeled. Tripod fixed, manual exposure, one lamp on a rail so its distance stays constant as you move it. The question this set is meant to answer: at the treatment's reading size — imagine a four-centimeter inset on a page — does C read as distinctly glossier than A, and does B's weave survive?
The mistake. Photograph C with the lamp raked across the board, nearly parallel to it, from the left. Photograph A with the lamp at the camera. Lay the two frames side by side.
A comes back bright and even. C comes back dark, with a long smeared gradient across it. The sheet now appears to say: the foil is darker and less reflective than the paper.
That's wrong twice over. The foil is the shinier material, and "darker" was a decision about where the lamp stood. Both errors have the same cause. With the lamp at the lens, C's mirror direction points back at the camera and it would have flared into a bright patch. With the lamp raked, C's mirror direction points away to the side, so almost nothing it reflects reaches the lens. Meanwhile A, which scatters light in every direction, looks roughly the same under both arrangements. That asymmetry is the trap: the matte sample is nearly indifferent to exactly the change that transforms the glossy one, so a comparison that changes the light between samples will always seem to show a material difference.
The remake. Same camera position, same lens, same manual exposure, same white balance, same lamp at the same distance. Now the lamp sits at the camera axis for both frames — or, if you prefer the cleaner gloss read, source and camera are at matched angles on opposite sides of the normal, identically for both samples.
A now shows a broad, low-contrast gradient with no defined edge. C shows a small, hard-edged reflection with real intensity behind it. The difference is now attributable to the samples, because the geometry held still. That pair goes into the sheet, both frames captioned with the same light position.
The earlier pair doesn't have to disappear. Keep it in the working folder with a note saying what was wrong with it and what replaced it. A labeled wrong pair teaches the distinction faster than a correct pair in isolation, and it's honest about how the set was built.
Then the cloth. A third view, lamp raked across B, camera above, shows the weave. Before declaring victory, shrink the frame to the size the treatment will actually use it at and look again. If the weave vanishes at four centimeters, the sheet has told you something real and useful — that the weave isn't a usable treatment detail at that size — and a flattering close-up would have hidden it.
The picture that isn't a comparison
Somewhere in the shoot you'll make a frame that just looks good: the foil crumpled and catching a window, the cloth in a soft falloff, a hard shadow splitting the paper. Fine. Keep it, and label it. An expressive frame can propose a look, which is a perfectly good thing for a treatment to have.
What it can't do is sit in a controlled comparison as though its light and view were comparable to the others. Once that happens, the sheet has one frame doing persuasive work and the rest doing evidential work, and the reader has no way to tell which is which.
What travels with the image
A selected frame without its conditions is a picture of a thing that happens to be gray. Keep the sample ID, the view, the light relationship, the distance, the exposure, and any deliberate processing together with the file. A caption block like this does it in a form nobody has to decode later:
A · matte paper · overall · light at camera, 30 cm · ruler in frame
f/8, 1/125, ISO 200, WB 5200K, no processing
A · matte paper · sheen · mirror pair: light 30° left of normal,
camera 30° right, 40 cm · same exposure as C sheen
B · cloth · relief · light raked 80° from normal, camera left, 45 cm
C · foil · sheen · mirror pair: light 30° left of normal,
camera 30° right, 40 cm · same exposure as A sheen
C · foil · sheen (bracket) · same geometry, −2 stops, highlight
structure only — not a substitute for the comparison frame
— · no transmitted view: no sample raises a translucency question
Two details in there are worth defending. The bracket exists because a fixed exposure shared across samples with very different reflectance will clip the glossy one — and that clipping is itself information, but it erases the highlight's shape. The bracket pulls the exposure down instead, so the reflection's edge and internal structure survive; the blown-out version stays where it is, in the comparison frame. And the bracket is labeled, because quietly swapping a re-metered frame into the comparison converts a material fact into a camera decision.
And the scale reference stays in frame rather than being assumed. If the ruler gets cropped out, add a labeled detail view paired with the overall frame instead. A standalone crop erases scale and orientation, which are two of the few things the reference was supposed to preserve.
If a frame in the set comes back different for a reason you can't name, go back to the setup. Don't grade the images until the samples appear to agree — at that point you're not making a reference, you're making a claim.
What this set cannot establish
It cannot certify color. With no target and no profiling, the frames show relative relationships under one arrangement — useful for comparing two samples, useless as a color standard.
It cannot predict what the product will look like on a shoot. It shows one surface under one set of conditions. That's the honest scope of a reference sheet, and it's enough, as long as nobody quietly upgrades it into a lighting plan.
It cannot identify the material. A photograph of a surface isn't a specification.
It cannot cover fragile or irreplaceable objects. Handling requirements, ultraviolet and infrared imaging, radiography, conservation documentation — that's a different discipline with different training, and the heritage-imaging material it comes from is being borrowed here only for its visible-light relationships. None of it transfers into a kitchen-table comparison of replaceable samples.
And an effect that doesn't appear in a frame doesn't prove the material lacks it. If the foil shows no strong highlight from a given angle, that means this angle doesn't show it. Angle the lamp differently and it will. Which is the entire argument for keeping the light position in the caption.
The set you actually want
Two or three samples, one surface question, one identity system, one frozen baseline and one deliberately varied relationship per view. An overall frame for context. A relief view if relief matters. A sheen view if gloss does. A transmitted view only if translucency is genuinely in question. Captions that name the change and the conditions, originals retained, and a stated list of what the setup leaves unsettled.
The test isn't whether the sheet looks authoritative. It's whether someone who wasn't in the room can point at any frame and say which property it shows, what was held fixed, and what changed — and then say what the photograph still doesn't tell them.
Frequently asked questions
Why can photographing a cloth and foil on the same table produce a false material claim?
The cloth frame may show a crisp weave because it got side light, while the foil frame shows a flat gray rectangle because it got light from a position where its shine bounced away from the lens. Same materials, different photographs, opposite conclusion, and nothing on the sheet tells you which fact you are looking at. A material reference becomes readable when the view and the illumination that reveal the surface are part of the record.
What separate questions can texture stand in for?
Color relationship, surface relief, gloss, and translucency. They respond to different setups, and one handsome photograph will not explain all of them. Gloss is how concentrated the reflection is, not brightness; a dull sample under a hard lamp can measure brighter than a shiny one photographed away from its reflection angle. Translucency is light passing through the material from behind. Name the treatment question in a form a photograph can settle.
How do raking, frontal, and sheen views differ for material samples?
Frontal or normal light near the camera axis shows color relationships and overall form, but it flattens fine relief and can produce a large flare on a glossy sample. Raking light brought down almost parallel to the surface casts shadows in valleys and shows relief, but on a glossy sample the mirror direction points away from the lens and the sample goes dark. A sheen or specular view puts the source where its reflection can reach the lens; matched angles, source 30 degrees off normal on one side and camera 30 degrees off on the other, show the center of the highlight. The highlight largely pictures the light source, so hold the source identical across samples.
What does the foil-versus-paper mistake show, and how is it remade?
Photographing foil with the lamp raked from the left and matte paper with the lamp at the camera makes foil look dark and smeared while paper looks bright and even, falsely suggesting the foil is darker and less reflective. The foil is the shinier material, and the error comes from where the lamp stood: with the lamp at the lens, the foil would have flared; with the lamp raked, almost nothing it reflects reaches the lens. The remake holds camera position, lens, manual exposure, white balance, and lamp distance, and uses the same light position for both samples. The foil then shows a small, hard-edged reflection, while the paper shows a broad, low-contrast gradient.
What can a material reference set not establish?
It cannot certify color without a target and profiling; it shows relative relationships under one arrangement. It cannot predict what the product will look like on a shoot, identify the material, or cover fragile or irreplaceable objects, which is a different discipline. An effect that does not appear in a frame does not prove the material lacks it; it may just be an angle that does not show it. If a frame comes back different for a reason you cannot name, go back to the setup rather than grading the images until the samples appear to agree.