How do I photograph a vial so that someone can actually tell me what they are looking at?

Almost every "is this normal?" post I read is unanswerable because the photo is a bright yellow blur taken under a kitchen light with a flash reflection in the middle of the vial. I have been guilty of this. I would like to know how to take a photograph that carries real information, and equally what a photograph fundamentally cannot show, so I stop asking questions that no image can answer.

Things I suspect are going wrong:

• Phone cameras apply noise reduction and HDR aggressively. I assume that erases exactly the faint haze I am trying to show.
• Flash reflects off curved glass and blows out the region of interest.
• There is no scale reference, so a "large" particle could be 0.2 mm or 3 mm.
• The white balance under domestic lighting makes everything look yellow, which matters if the question is about discolouration.

What is the actual protocol? And what should I write alongside the photo so that the answer does not have to start with five clarifying questions?

Research use only.

The protocol is short and the metadata matters more than the photograph. Both below.

Lighting and background

1. Matte black background for haze and light-scattering particles, matte white for dark particles and colour. Take both, always. A sheet of black card and a sheet of white printer paper. Not glossy — gloss gives you reflections you will spend the whole shot fighting.
2. One point source, off to the side, roughly 45 to 90 degrees from the camera axis. A small torch or a phone torch on a second device. You are photographing scattered light, which is visible off-axis and invisible on-axis. This is the single most important item on the list.
3. Flash off. Always. On-axis flash on curved glass produces a specular blowout exactly where the meniscus is.
4. Room lights off or dimmed. Mixed lighting wrecks white balance and fills in the shadows you need for contrast.

Camera settings

1. Lock focus and exposure manually by tapping and holding on the vial. Autoexposure will meter for the black card and blow out the vial.
2. Turn off HDR, night mode, and any "photographic style" or beautification. These stack frames and denoise, and faint uniform haze is exactly what a denoiser interprets as noise and removes. If your phone has a pro or raw mode, use it — a single unprocessed frame is far more informative than a computationally cleaned one.
3. Set white balance manually if you can, or include something known-white in the frame so the colour can be judged. This is essential if the question is about discolouration and irrelevant otherwise.
4. Get close but respect the minimum focus distance. Most phones focus at 8 to 10 cm; closer than that and you get a soft image that looks like haze. If you need more magnification, shoot from the minimum focus distance and crop rather than moving closer.
5. Brace the phone. Against a book, a mug, anything. Low light plus handheld equals motion blur that mimics haze.

Composition

• Include a scale reference in at least one frame. The vial's own graduations if it has them, otherwise something with a known dimension beside it. Without scale, no one can tell a 0.2 mm speck from a 3 mm fragment, and those have completely different explanations.
• One frame of the label and lot number, separately. Legible. This is often the frame that answers the question, because the excipients and the fill mass are on it.
• One frame of the septum if the question involves particulate, so a cored stopper can be seen.
• Do not crop the vial out of context. A tight crop on a particle loses the fill level, the meniscus and the cake remnant, all of which carry information.

Short video beats stills for two things

Motility and dispersion behaviour cannot be photographed. A ten-second clip of a gentle swirl followed by the flow coming to rest distinguishes a fibre from an aggregate from a motile organism in a way no still image can, because the diagnostic information is in the motion. If your question is "what is this particle", shoot the clip.

What to write alongside — the part that matters most

Most of these threads are unanswerable for textual reasons, not photographic ones. Six fields, and a post that includes them will get a real answer:

• Compound and nominal fill mass.
• Diluent, exactly which one, and the volume added.
• Date and time of reconstitution, and elapsed time since.
• Storage temperature and location, including how many times it has been out.
• Number of stopper entries and the needle gauge used.
• Whether it was ever shaken, foamed, frozen, or left at room temperature, and for how long.

What a photograph cannot show, ever

Be clear about this so you do not over-read an answer:

• Chemical degradation below the visibility threshold. Deamidation, oxidation and isomerisation are invisible. A vial can have lost a meaningful fraction of intact peptide and photograph perfectly.
• Potency or content. No image speaks to how much peptide is in the vial. That is a content assay, and the independent services — Janoshik, Medutest, PeptideMeter, VendorInvestigate — exist for exactly that.
• Endotoxin. Invisible, and unaffected by anything you can see or filter.
• Low-level microbial contamination. Visible turbidity needs somewhere around 10^6 organisms per millilitre. A photograph of a clear vial excludes heavy growth and nothing more.

So the honest scope of a good photograph is: it can identify visible particulate, it can distinguish cake morphologies, it can show a meniscus film, and it can show gross discolouration or turbidity. Those are worth having. Anything else is a question for an assay or a clinician, and no amount of photographic technique changes that.

Adding the decision half of the question, since knowing how to photograph a vial is only useful if you know what you would do with the answer.

Discard without further discussion:

• Any liquid or smearing in what should be a dry cake.
• Yellow, brown or grey discolouration of cake or solution.
• Discrete motile specks, or turbidity you can see without a dark background and a torch.
• A meniscus ring or film that appeared during storage rather than at reconstitution.
• Any suspicion of microbial growth. This is the one where filtration is a trap rather than a solution.
• A stopper with visible coring damage, or a crimp that rotates freely.

Worth asking about:

• Cake morphology you cannot classify — glossy versus matte is the discriminator and it is genuinely hard to judge in your own photographs.
• A single fibre or a single dark speck in otherwise clean solution.
• Faint haze you are not sure is there. If you are unsure, photograph it against black with an off-axis torch and ask; this is exactly the case images can settle.
• Reconstitution that took much longer than your baseline.

Not worth a thread:

• Bubbles after reconstitution. They clear. Stand the vial cold and undisturbed.
• A cake stuck to the wall, off-centre, cracked, domed or dimpled. Geometry, not texture.
• A cake so thin you thought the vial was empty. Do the volume arithmetic instead — a 10 mg neat fill is around a millimetre of porous solid and is nearly invisible.
• No cold pack in an ambient shipment of lyophilised powder. Expected and normal.

The general rule that covers the ambiguous middle: a vial is worth the price of a vial, and the failure modes you cannot see are the ones that matter. If the analysis is taking more than a few minutes and the material is not unusually expensive, the answer is to discard and to fix whichever upstream habit produced the finding.

One practical trick that improves haze photographs more than any camera setting, since the accepted answer covers everything else.

Photograph two vials side by side. One is the vial in question, the other is a known-good reference — a freshly reconstituted vial, or a vial of plain diluent at the same fill level in an identical container. Faint uniform haze is nearly impossible to judge in isolation because there is no reference for what "clear" looks like through that particular glass under that particular light, but it is obvious in a paired comparison. The human eye is very good at detecting differences and very bad at judging absolutes, and a side-by-side turns an absolute judgement into a difference one.

Same trick for cake morphology: a photograph of a questionable cake beside a known-good cake from the same supplier, under the same raking light, settles the matte-versus-glossy question instantly.

Two refinements:

• Match the fill level. Different liquid depths give different path lengths and therefore different apparent haze, which will mislead you.
• Use identical containers. Glass varies in clarity, and a vial from a different supplier is not a valid control.

Keeping one vial of plain bacteriostatic water at the same fill volume as a permanent optical reference costs almost nothing and makes every subsequent inspection quantitative in the only sense available to you.