Filet vs tapestry vs C2C: which chart for which picture
These three techniques all end up as a grid of squares you work from, which makes them look interchangeable on a chart. They are not. Each one has a different number of states per square, and that number decides what kind of picture survives the conversion.
Filet crochet: two states per square
Filet is a mesh. Every square is either solid (a filled block) or open (an empty space). There is no third option, which means two colors, one of which is usually the background.
What that is excellent for: silhouettes, lettering, borders, geometric patterns, single-color logos. Anything whose shape carries the information.
What it cannot do: faces. A face needs shading, and shading needs intermediate tones. Filet has no intermediate tones, so the face becomes a dark blob with a lighter blob inside. People try anyway, and the result usually looks like a stencil of someone who has been mildly surprised.
Filet does have one thing going for it that nothing else matches: it is fast, and the open mesh makes even a large piece feel light. A filet curtain or a table runner is a very different object from a dense tapestry panel of the same size.
The filet generator takes a photo or a grid and outputs the mesh →
Tapestry crochet: many colors, all at once
Tapestry carries multiple colors per row. You hold the inactive colors behind or inside the work and bring each one forward when its square comes up. Practically, most pieces use somewhere between three and a dozen colors, with more possible at the cost of more strands to manage.
What that is excellent for: portraits, illustrated panels, anything with gradients or fine detail. This is the technique that can actually do a face.
What it costs you: every color is a strand. Six colors means six strands running along the row, tangling, needing to be turned at the end of each row. Rows get slow. A large multi-color tapestry is a long project, and the difficulty is not in the chart — it is in the handling.
A specific trap: carrying a color across many stitches (called floating) makes the back thick and can show through on the front if the floats are too tight. Long floats need to be caught, and catching them adds work per stitch.
C2C: diagonal blocks
Corner-to-corner works in diagonal squares, each one a small block of stitches. The chart is a grid of those blocks, and the diagonal grain is visible in the finished object.
What that is excellent for: pixel art, game sprites, simple cartoon shapes, bold lettering. The blocky diagonal texture suits images that are already made of blocks.
What it fights: smooth curves. A circle in C2C becomes a staircase. At large block counts the staircase is fine; at small counts it is the whole shape and the curve is gone. If the picture depends on a smooth outline — a wine glass, a crescent moon at small size — C2C will make it look like a rendering error.
Picking, in one table of questions
- Is the information in the outline? A silhouette, a logo, a word, a border. Use filet. It will be faster and lighter than anything else.
- Is the information in the shading? A face, a photo, a landscape with depth. Use tapestry. Accept that it will be slow.
- Is the image already made of large blocks? A game sprite, a pixel character, chunky lettering. Use C2C and lean into the texture.
If a picture needs both an outline and shading — most portraits do — the answer is tapestry with a deliberate outline color, not filet with a hope that the shading shows up.
One more thing before you convert
Whichever you pick, decide the finished size first and work backwards to stitch count. The same photo at 80 stitches and at 200 stitches are two different charts, not one chart at two scales. A converter will give you either; only one of them will be worth the hours.
All three generators on this site run in your browser. Your photo never leaves the device.