Output options explained

What each setting does, and when to change it. If you just want a good result, the defaults are chosen for logos and illustrations: pick what the file is made for and export.

1. File formats: PDF, AI, EPS, SVG and DXF

All five formats are true vector files: they describe shapes with points and curves, so they stay sharp at any size. They differ in what else they can carry. The export menu notes anything a format can't keep for your artwork before you save.

Format	Best for	Keeps	Doesn't keep
PDF	Print, sending to a printer or another designer	Everything: gradients, full shading, soft highlights, CMYK and named spot colours	Nothing
AI	Opening in Adobe Illustrator	The same as PDF (it is a PDF-based Illustrator file)	Nothing
EPS	Older print workflows and software that asks for EPS	Gradients, CMYK and spot colours	Soft highlights (EPS has no transparency); complex shading becomes one gradient per area
SVG	Websites, apps, Figma, Inkscape	Gradients and soft highlights	CMYK and spot colours (SVG is RGB); complex shading becomes one gradient per area
DXF	Vinyl cutters, laser cutters, plotters, CNC and CAD	The outlines, one layer per colour	Gradients (they become flat colours), highlights and print colours

PDF

A single page exactly the size of your artwork, with nothing else on it. Complex shading is drawn as fine flat-colour tiles clipped to each shape, which every program can open and print. Soft highlights use PDF transparency. Named spot colours are written as real separations, so a printer sees them as their own inks.

AI

The same content as the PDF, saved with an .ai extension so Illustrator opens it as editable artwork. Everything stays an ordinary path or group: shaded areas arrive as a clipped group of tiles rather than an Illustrator gradient mesh.

EPS

Encapsulated PostScript, for print workflows that still ask for it. Files with gradients use PostScript Level 3 smooth shading; files without stay at Level 2 for the widest compatibility. Spot colours are declared so separations work.

SVG

The format of the web. Sizes are in pixels, matching your image, and gradients are written as SVG gradients with names unique to the image, so two traced SVGs on one page never clash. Use Copy SVG code in the export menu to paste the markup straight into code.

DXF

For machines that follow paths rather than print them. PathSnapper writes a classic R12 DXF, which almost every cutter and CAD program reads. Curves are converted to closely spaced straight segments (within 0.05 px of the curve), because many cutters handle these more reliably than splines. Each colour goes on its own layer, named after the colour (for example FILL_1D2A5C), so you can cut colour by colour. Choose Made for: Cutting and plotting first so the shapes are cut-friendly.

2. Made for: print, web or cutting

Made for, on the Output step, sets sensible defaults for where the file is going. You can still change any individual setting afterwards.

Print and design: offers PDF first and keeps gradients and shading.
Web and screens: offers SVG first and keeps gradients.
Cutting and plotting: offers DXF first and makes everything cuttable: gradients become flat colours, and thin lines become outlined shapes instead of strokes, because a cutter can only follow edges.

3. Print colours: RGB, CMYK and spot colours

Screens mix red, green and blue light (RGB); printing presses mix cyan, magenta, yellow and black ink (CMYK). Set Print colours on the Output step to choose which PDF, AI and EPS files use.

RGB keeps colours exactly as they look on screen. Right for anything digital, and for printers that convert colour themselves.
CMYK writes ink values. PathSnapper converts each colour for you, or you can type the exact CMYK values your brand guidelines give, per colour, under the swatch.
Spot colours (for example PANTONE 2768 C) are named inks printed as their own colour. Type the name under Spot colour for a swatch, or load a swatch library and match your colours to it. Named spot colours are always kept as spot colours in PDF, AI and EPS, whichever mode you choose.

SVG and DXF don't carry print colours: SVG is always RGB, and DXF has no colour model a printer would use.

4. Shapes: stacked layers, cut-outs and shared edges

Under Advanced, Shapes decides how areas of different colours fit together.

Stacked layers: a shape that sits completely inside another colour is drawn on top of it, and the colour around it stays whole underneath. A red circle in the middle of a blue square is a full circle on a full square. Colours that meet side by side, without one surrounding the other, are not extended under each other: they meet at their shared edge, just as with cut-outs. A circle hanging over the edge of the square still takes a bite out of it. A hole is also kept open where something you left out (such as the background) shows through it, so a logo on a left-out background often looks the same either way. Stacked files are simpler to recolour, and suit layered vinyl, where a top colour is laid over a full piece of the colour below. On press the top shape knocks out the one beneath it (nothing is set to overprint), so the print is the same as with cut-outs.
Cut-outs: shapes never overlap. The square has a circular hole where the circle sits, like pieces of a jigsaw. This is what you want for vinyl, screen printing by layer, or anything where each colour is made separately.
Automatic: cut-outs for one-colour artwork, stacked layers for everything else.

Shared edges

Whichever you choose, PathSnapper traces every boundary once and shares it between the shapes on either side. When you move a point on an edge where two colours meet, both shapes change together, so you never get a hairline gap or an overlap to tidy up afterwards. In Outline view you can see it: where two colours meet there is a single line. This matters most with cut-outs, where every edge is shared. See Output: curves, points and export for it in action.

Drop shadows in an image with a transparent background become one grey shape of their own, beside the shapes that cast them, instead of a dark fringe along their edges. To drop the shadow, tick Leave out for its colour on the Colours step.

Artwork cut out of its background (white paper made transparent, as many older logo files are) keeps its pale edge pixels, the anti-aliasing against the paper. PathSnapper recognises the colour it was cut from and reads those edges against it, so thin hairlines and small text trace as faithfully as they would on the paper, and the transparent parts stay transparent in the result.

Leave background out, on the Colours step, drops the background colour so the artwork stands on its own. Enclosed areas of the background colour that are part of the design (the white parts of a roundel, say) are kept.

5. Curves: complexity, smoothing and points

Curve complexity

How closely each curve must follow the edge in your image. Towards More detail, curves hug every bump of the original, using more points. Towards Fewer points, curves are smoother and lighter and easier to edit, but small details are simplified. For clean logos the default is right; for soft or noisy images, fewer points usually looks better.

After fitting, PathSnapper checks its curves against your image: it draws them, compares the drawing with the pixels (including the soft, blended pixels along every edge) and adjusts points and handles until the two match, removing points that do not earn their place. Exact shapes (true circles, tidied polygons and straight edges) are left as they are. This is what keeps thin, soft lettering smooth. Very large or very detailed images (over about 1.5 megapixels or 3,000 points) skip this step, as their edges are already placed precisely.

Path smoothing

How sharp a turn has to be to stay a corner. Towards Follow image, more turns are kept as sharp corners. Towards Smoother, gentle corners become flowing curves, which suits script lettering and organic shapes.

Nodes at curve extremes

Puts a point at the top, bottom, left and right of every curve, as a designer would draw it. Curves with points at their extremes are much easier to adjust by hand, so leave this on unless you need the fewest possible points.

Straight lines

Lines within 1.5° of horizontal or vertical are made exactly horizontal or vertical, so edges that should be straight are.

6. Tidy circles, polygons and stars

Traced edges are never perfect, so a circle in a logo comes out as a nearly-round curve. With Tidy circles, polygons and stars on, PathSnapper recognises shapes that are meant to be geometric and makes them exact: circles and ellipses become true ones, circles that share a centre (rings and targets) get one exact centre, and rectangles, rounded rectangles, polygons and stars get straight sides and true corners. Turn it off for hand-drawn artwork where the wobble is part of the character.

Square ends (on for the Geometric & icons artwork type) is for thin bars that end square: small geometric lettering, icons. Where a bar is only a few pixels across, the image's blur rounds the two corners of its end into one, and it would trace as a round end. With Square ends on, an end that stands out past the bar's straight sides by no more than that rounding is drawn square, meeting the sides at corners. An end that is truly round stands out by half the bar's width and stays round. Leave it off for lettering with rounded ends, serifs or script, where it would square what should be curved.

7. Thin lines as strokes

With Thin lines as strokes on, lines no wider than Widest stroke line (under Advanced) become single stroked paths with a line width, instead of long thin filled shapes. A stroke needs far fewer points than an outline and keeps an even width when you edit it. A thin outline drawn round a shape in a colour of its own (a dark keyline round a gradient, say) becomes a stroke that follows the shape. Lines that taper are kept as filled shapes, and so is lettering or brushwork drawn with a pen that varies in width, even where a short stretch of it is even. Turn it off for cutting, where every line must be an outline the blade can follow (Made for: Cutting and plotting does this for you).

8. Gradients, glows and shading

With Gradients and glows on, smooth colour ramps in your image become real gradients instead of bands of flat colour, and soft highlights become radial glows. Turn it off for flat artwork you want to stay strictly flat.

Some artwork has complex shading that one gradient can't describe: chrome, 3D lettering, a lit sphere. When PathSnapper finds it, the Shading panel lets you choose:

Smooth shading: few points, and the full shading is kept in PDF and AI. SVG and EPS get the best single gradient for each area, and DXF a flat colour.
Bands: the shading is split into flat-colour bands, each its own shape. More points, but every format shows exactly the same thing, and each band can be recoloured or cut.

Turn on Show in the Shading panel to outline the shaded areas on your result.

9. Stepped edges

When a small image is enlarged, its edges turn into stair-steps, and a faithful trace keeps them. Stepped edges (under Advanced) smooths them away:

Smooth automatically: measures the size of the steps in your image and smooths exactly those. Right almost always.
Leave as traced: no smoothing, for pixel art and deliberately blocky designs.
Smooth small, medium or large steps (2, 4 or 6 px): forces a step size when the automatic setting misjudges it.

10. Artwork type

A starting point for curve complexity, path smoothing and square ends, chosen on the Image step and changeable on the Output step. Changing any of them yourself switches the type to Custom. Settings belong to each image: every open image keeps its own as you switch between them, and a newly opened image starts from the defaults, with its type detected for it (a photo gets Photo, lettering Script, and so on).

Logo & illustration: the balanced default.
Geometric & icons: tighter fitting, sharper corners and square ends, for crisp geometric marks and lettering.
Script & lettering: smoother curves and fewer corners, for flowing type.
Detailed artwork: the closest fit, keeping fine detail.
Photo: turns a photo into a poster-style illustration. Small specks merge into larger areas, up to 24 colours are kept with nothing left out, smooth areas become gradients, and stepped edges are smoothed.

11. Colour settings: number of colours, speck size and merge

These are on the Colours step and decide the colour map that everything is traced from.

Number of colours: how many colours to keep. Fewer joins similar shades; drag to All to keep every colour found.
Speck size: the smallest area that is kept. Anything smaller is absorbed by the colour around it, which removes noise and compression artefacts.
Merge: how different two shades must be to stay separate colours. Raise it if anti-aliasing or a blurry edge has produced extra in-between colours.

You can also merge, recolour, remove or add colours by hand. See Colours: choose, change and merge them.

12. Working size and AI clean-up

Working size, on the Image step, is the largest size PathSnapper traces at: full size up to 2600 px, or 2000, 1400 or 1000 px. Larger keeps fine detail; smaller is faster and gives fewer points, which suits simple logos supplied at a large size.

Small images that are soft or made of thin strokes (small print, script lettering, line art; under 600 px on the long side) have their outlines traced at two to four times their size and scaled back, so thin lettering and small curves come out smooth instead of faceted, and strokes a pixel wide are kept. Colours, painting and thin lines are still worked out at the image's own size, and the result is the same size as the image. Crisp small images of broad shapes do not need this and are traced as they are. It happens by itself; there is no setting.

Thin ink keeps its colour. Text or lines about a pixel wide have almost no pixel fully covered, so their colour in the image is paler than the ink they were drawn with. Such a colour is filled with its darkest pixels' colour instead of the average, and its outlines are placed to match, so small dark text stays dark rather than coming out grey. The colour map still shows the colours as they appear in the image, and the selected colour says which paler colour it was found as; you can change it like any other. Ink thinner than the pixels can show (hairlines at the smallest sizes) can still come out lighter than it was drawn.

Clean up with AI removes JPEG blocks, blur and noise from low-quality images before they are traced, using the open Real-ESRGAN model running on your own device. It enlarges the image four times while cleaning it, and keeps it at twice its size, so small lettering and fine detail have more pixels to be traced from: in testing on small and soft artwork it removed about 40% of the tracing error. It helps most with small, blurry, blocky or enlarged images; PathSnapper tells you if an image already looks clean enough not to need it.

13. Size and units

Exports keep the size of the image you traced, one pixel to one unit:

PDF, AI and EPS: one pixel becomes one point (1/72 inch). A 960 × 540 px image exports at 960 × 540 pt, about 339 × 191 mm.
SVG: sizes are in pixels, and the artwork scales freely in the browser.
DXF: coordinates are in pixels without units, so set the size you want when you import it into your cutter or CAD software.

Because the files are vectors, scaling them up or down in your design software loses nothing.

14. Reading the result numbers

The Result and export panel on the Output step describes the vector you are about to export:

Colours: colours used in the vector.
Contours: separate outlines, each shape's edge and any holes.
Anchors: points in total. Fewer anchors mean smoother, lighter files that are easier to edit.
Curves: segments that are curves rather than straight lines.
Strokes: thin lines kept as strokes.
Gradients: gradients and glows in the file.
Paths: drawing objects in the SVG.
File size: the size of the SVG.

Change a setting and the numbers update as the vector is traced again, so you can see straight away what each setting costs or saves.

15. Fidelity and the honesty check

Every trace gets a fidelity score: how closely the vector matches your image. PathSnapper draws the result back at the image's size and compares the two, area by area. Fine detail hides small errors at that size (a stroke a fraction of a pixel off looks almost the same once shrunk), so the score allows for it, more the thinner the strokes: an area passes only when it is very likely at least 90% right.

When an area falls short, the honesty check says so the first time it happens for an image, and lists each area with its score and why: usually strokes thinner than about a pixel and a half (small text, hairlines), or an image without enough detail. Show outlines those areas on the result; the Fidelity panel reopens the check at any time. Where strokes are too thin for the image to show their shape (under about 3 px across: in testing, every area with thicker strokes traced faithfully), the check says how much larger an image of the same artwork would need to be, in pixels. A larger original is the way to a faithful copy; Clean up with AI can help, but cannot recover detail the image never had. Nothing stops you exporting: the check tells you, it does not decide for you.

Your own choices are not counted as errors: output colours you changed, pixels you painted and colours you left out are compared as you meant them. A photo traced as flat colour is never a faithful copy, so for the Photo artwork type the score shows without the check opening by itself. The score is measured when the vector is traced; point edits made afterwards are not included.