Build a Colour Scheme from a Photo | Tinta

The short answer

A photo gives you a palette — the colours that were found. It does not give you a scheme — a set of colours that agree well enough to actually design with. Tinta produces both: a subject-aware palette read from the picture, and harmony suggestions when what you need is a coordinated set rather than isolated samples.

The subject-aware step is what makes the scheme worth anything. If the palette is the beige wall, the harmony suggestions are an elegant application of the wrong colour. If the palette is the rust of the jacket you were actually looking at, the same suggestions are immediately usable.

Harmony suggestions and every mount style sit behind Tinta Pro. If you only want the short version, the Tinta product page has it; if you would rather skip the argument and try it, Tinta is on the App Store.

Why the searcher is here

"Build a colour scheme from a photo" is almost never asked by someone who wants a tutorial on colour theory. It is asked by someone who has already tried the obvious method and found it disappointing. The disappointing part is usually one of two things:

• The eyedropper gave them the background. In most photographs the largest flat area is a wall, a sky, a table — and it is rarely the colour that made them stop and look. Palette extraction is the layer that fixes this, and it matters to the scheme more than anything that happens afterwards.
• They have four colours and no idea what to do with them. A row of swatches is not a system. Something has to decide which colour leads, which support it, and which relationship holds them together.

Both are failures of selection, not of colour theory. Nobody needs a primer in harmony. They need the colours that were actually in the picture, ordered and grouped so the relationships become visible.

A palette and a scheme are different problems

Worth stating precisely, because the two words get used interchangeably and the tools built for them are not interchangeable.

A palette is an inventory: the rust of a jacket, the beige of a wall, the green of a hedge, the grey of the pavement — a list of what the camera saw. A scheme is a selection with a structure: one dominant colour, something to support it, something to punctuate it.

The standard route from the first to the second is to pick by eye and hope. That fails for a boring reason: the eye is poor at judging whether two colours sit close enough to work together, and worse at judging the same pair on two screens. Whether a pair works is a measurable relationship in colour space, and re-picking by eye re-rolls the dice each time.

Working from a photograph gives you something eyeballing does not: a starting set that was already composed. The light and materials in a real scene tend to produce colours that sit together, because the same light produced all of them. That is the real argument for starting from a photo — and it holds only for the colours that were part of the subject. Take the palette from the wrong region of the frame and the composition argument evaporates: a beige wall, a beige shadow and a beige highlight is not a palette, it is one colour three times.

What subject-weighting changes about the scheme

Subject-aware clustering is described as pulling colour from what you actually shot rather than the biggest grey area in the frame. For a palette that is the difference between the rust and the beige. For a scheme the consequences compound:

• Dominant versus accent. How much area a colour occupies is the usual proxy for how much weight it should carry in a design. Weighting toward the subject is what lets the rust of the jacket take the dominant position while a smaller green acts as the accent — the reverse of what ranking by raw area gives you.
• The relationships survive. Colour relationships only mean anything between distinct colours. A palette of the background produces harmonies between variants of a single hue, which are formally valid and practically useless.
• There is no repair step further up. Harmony suggestions operate on the colours they are given. The scheme is exactly as good as the palette underneath it.

The chips arrive draggable, so you can rearrange them until the card reads the way you meant. Worth doing before treating a suggested scheme as settled — the order is a proposal, not a verdict.

Harmony suggestions: what they are and what they are not

Tinta offers harmony suggestions "for when you want a scheme, not just a sample", and they are a Pro feature. The free tier extracts a palette from any photo, uses the core frame styles, and exports a card — so you can judge the extraction before paying for the suggestions.

Here is the honest boundary. The published material states that harmony suggestions exist and that they are the right tool when you want a scheme; it does not document which specific relationship rules produce them — complementary, triadic, analogous or otherwise. If you need certainty that your scheme is built on a named rule, verify that in the app before committing the palette to a design. What is documented is solid: every swatch carries its exact HEX, so whichever scheme you settle on is expressed as exact values rather than approximations.

That exactness is the part that survives contact with the rest of your workflow. A colour eyeballed off an image is approximate in a way that compounds — write it into a stylesheet, export a layout, hand it to a developer, and every downstream artefact inherits the approximation. An exact HEX pasted once is exact everywhere after. The dedicated page on the values covers what they are for.

One thing to be clear about: a HEX is an RGB value, not a spot colour. Tinta makes no claim about matching Pantone or any other library, so a HEX from a card tells you what goes in a stylesheet, not what to order from a paint counter.

Twelve mounts, one scheme

A scheme is easy to lose on the way out of the app. Twelve editorial mount styles — instant, film strip, arch window, diptych, spread, deckle edge and more — reframe the same colours without touching them: one swipe changes the whole mood, same photo, a new frame. Swiping the style reel changes the whole look in place, and tapping any element fine-tunes it where it sits rather than sending you back to a settings screen.

This matters more for a scheme than for a palette. A palette shown as a row of flat rectangles has no hierarchy; the same palette mounted as an arch window does, because the frame tells the eye what leads. Cards collect into a light-table gallery that can be kept warm or cool, minimal or vintage — which is where a set of schemes starts to read as a collection rather than a folder of loose swatches. The export page covers carrying the result out.

Every mount style is a Pro unlock; the free tier gives you the core frames.

What Tinta does differently

The differentiator is not the harmony engine. It is that the whole pipeline — read the subject, express it as exact values, suggest a structure, mount it, export it — runs entirely on your device, with no account, no ads, no tracking SDKs and no cloud. Your photos never leave the phone, and the place a photo was taken is resolved through an offline gazetteer, so no location reaches a server either.

For a tool whose entire input is your camera roll, that is not a footnote. It is the reason a palette of your own clothes, your own walls and your own family is something you would actually make.

Two honest limits. First, subject-weighting is a heuristic: in a photograph where the subject genuinely is the flat background, it will not rescue the palette for you. Second, on-device means there is no cloud to sync through — the export is the artefact, and there is no account to sign into on a second device. Pro is $1.99 monthly or $11.99 yearly in the US, both auto-renewing with a 7-day free trial for new subscribers, unlocking all mount styles, harmony suggestions, custom typography, and watermark-free export up to 4K, lossless PNG, P3 wide-gamut. You can cancel in iOS Settings → Apple ID → Subscriptions, and restore a purchase from the paywall.

The decision

If what you need is a scheme you can design with, the question is not which app has the longest feature list. It is whether the palette underneath is the colour you actually noticed, and whether the values are exact enough to build on. Tinta answers both, and the mounts are what make the result presentable rather than merely correct.

If you need one hex value, an eyedropper is faster — and the page on getting hex codes explains why it is also usually wrong. If you already have the colours and just want them on a card, you are paying for an extraction step you do not need.

Otherwise: Tinta is on the App Store. Use the free tier first, so you can see whether subject-aware extraction agrees with your eye before you spend anything on Pro.