Polargraph is straight forward to build and operate, but there are a few “gotchas” that come up.  This page might help with some of the more technical things.

1. Density

My “render pixels…” drawings are coming out all empty.  I can see a good preview in the controller, but on the page?  Just blank lines man! What gives?


My drawings are coming out solid colour! There’s no variation in the shading! Come on, sort this out!


I’m interested in the relationship between pixel resolution and pixel depth, maybe you can explain a bit about it?

This is most common when using a really small grid size.  Imagine each pixel you’re trying to draw as being square filled in with a squiggly line.  With a pen of a certain size, you can maybe fit in 10 wiggles before the pixel is completely full.  That’s cool, you can use that to express 10 different levels of density.

Now, most people take umbrage at the low resolution of the pixelated preview image in the controller, (they didn’t buy a retina display to look at this blocky rubbish!) and push the grid size right down.  Dangerous move! Now each pixel can only fit in a maximum of two wiggles, and all of a sudden you can only express two different density levels.  Your beautiful smooth graduations are starting to look awful like sharp black and white now! Disaster! Anything vaguely grey has gone completely.

With a reasonably small pixel, and a reasonably thick pen, it isn’t hard to find yourself losing a lot of detail to this phenomenon.

The truth is that you can have either maximum resolution OR maximum dynamic range, but not both. There is a sweet spot for any given combination of pen and grid size, but you’ll need to test to find it.

My main advice is that if you are looking for very fine, high resolution output, you have constructed the wrong tool.


  • Increase the size of your grid.
  • Use a smaller pen.  Do a proper pen width test at your desired grid size to figure out how dense you want your darkest pixels to be.

2. Pixel styles

I’ve tried ‘variable frequency square” and ‘variable size square’, all very nice, whoop, but ‘solid square’ gives me no detail at all! It’s just solid – what’s the point of that? I think you’re stupid.

Yes solid square renders all pixels at maximum density, with the exception of masked pixels, which can be skipped entirely.  This was designed for situations where you have created colour separations, where each separation has no shading of its own, and the impression of density it gives is supplied by the colour of pen you use instead.

Masking like this – Colour all the pixels you wish to hide a hideous green colour, then load that image into the controller (use PNG format), click “choose mask colour” on the input tab and click a bit of that horrible green.  Now select your image area as usual and you will see that the masked colour is missing from the preview.  This is a primitive kind of chroma-keying, and works with all pixel styles, but has special usefulness in “solid square”.