Diffusion Dither verses FM Screen

[ReflexBlueHorror]

Just wondering for a 2400dpi greyscale image, what the visual quality difference would be between using 10μm or 20μm FM screen verses converting the image to a 2400dpi bitmap with diffusion dither?

 

[Slammer]

If memory serves, as it frequently doesn´t these days. DD was or is used to give pictures with a low bit depth the illusion of a higher bit depth via interpolation. Compaired to FM screen DD comes over as a bit clunky, maybe the Printplanets brothels and cisterns can confirm.

 

[Bret Hesler]

You will probably get similar effect to first generation FM screening, but will lose the ability to apply plate curves. Using RIP based screening will allow you to use second generation screening which will most likely print with less graininess, and be able to apply press calibration curves/compensation.

 

[gordo]

Just wondering for a 2400dpi greyscale image, what the visual quality difference would be between using 10μm or 20μm FM screen verses converting the image to a 2400dpi bitmap with diffusion dither?

All FM screens can be called diffusion dithers. Where they differ is in how the pixels are placed to render tones. Some (“first order” “disordered diffusion dither”) place the dots pseudo randomly and use more or less pixels to simulate tones. Some (“second order” “ordered diffusion dither”) use pseudo random individual pixels to build larger structures to simulate tones.

The diffusion dither in PShop is a very simple FM screen. Just like the AM screen is in PShop. The FM screen in a modern RIP uses a very sophisticated FM screen design.

You can screen a greyscale image in PShop using it’s diffusion’s dither and print it but it’s a first order screen and be quite noisy/grainy in flat screen tint builds.

the difference between 10 and 20 microns will be more about detail rendering and the finer process control required with the finer sceen

So you can do it but I wouldn’t.

 

[Stephen Marsh]

You will really need to get the gain/TVI correct if doing this in Photoshop, lightening the shadows more so than the highlights. I’d go with a proper first or second order RIP algorithm over Photoshop if given the choice.


Stephen Marsh