Major Issue: Inaccurate Pressure Processing

[cestarian]

Was pretty sure there was another topic about this but couldn't find it. I was happy that we got a new custom resolution option but less happy that the one bug that truly bothers me in the program wasn't fixed

The problem is that when a stroke is lagging behind the cursor (i.e. the stroke is drawn, but the program has yet to draw it to canvas, this happens for like 99% of all strokes I make); if the pen (or mouse) is then released, then for the remaining parts of the stroke it will read pressure as the lowest possible (something like 1%)

The same problem occurs everytime a stroke is started in-motion (i.e. if I transition from doing the stroke in the air to doing it on the active tablet surface as is fairly common, this sort of effect will take place at the start of the stroke)

The only way to avoid this issue is to stop for a second at the start and end of each stroke,, I think that the pressure sensitivity is being read continuously instead of being saved to memory so that the stroke can be accurately processed, and that there is a delay before the pressure can be read. at the start of the stroke or something along those lines.

strangely also, possibly unrelated, at the start of every stroke there is low opacity for at least 1 step (possibly more) even if that stroke is drawin with the mouse with pressure sensitivity disabled. This effect becomes significantly more pronounced when the opacity pressure sensitivity is enabled thanks to the issue mentioned above.

This is what I would consider the most serious bug in the program right now and I am hoping to see it prioritized for the next update

[support]

I think I know whats you mean. It's happening because programme should simulate pen pressure after the pen release. Actually, you don't have smooth pressure at the fast strokes, and the tip is blunt - we smoothly enter to the stroke but usually exit much faster. And the tip is different with the speed.

We think maybe to make it depending on speed (hardcoded)... there are much time spent before to find the way to realize it reasonably

[Urist]

I think I experience something similar to this. I created a thread about my issue here: http://www.paintstormstudio.com/forum/v ... f=6&t=4267

See the YouTube video there for a demonstration of the behaviour. It's also very easy to reproduce those results using a mouse. Again, if this is intended behaviour I'd rather just be able to completely disable stabilization and rely on third party programs or just work more carefully at different zoom levels depending on the detail required.

[cestarian]

I think I know whats you mean. It's happening because programme should simulate pen pressure after the pen release. Actually, you don't have smooth pressure at the fast strokes, and the tip is blunt - we smoothly enter to the stroke but usually exit much faster. And the tip is different with the speed.

We think maybe to make it depending on speed (hardcoded)... there are much time spent before to find the way to realize it reasonably

Yup that's essentially what I'm talking about, or so it looks like; but remember I said the same thing happens with a mouse also. I think using a mouse or something that has only 100% pressure and 0% pressure options is the best way to troubleshoot and try to fix this so I recommend you test it that way (with a mouse) as you try to come up with solutions before testing it with pressure sensitive devices

I must admit I know very little about memory programming, I imagine this could be fixed by either storing the physical strokes input data in RAM and read the processing data from ram instead of directly, or by processing the hw data and software data on separate threads so that a whole thread gets devoted to delivering accurate input data, but these are the best ideas I can come up with and I don't even honestly know if it'd work. But I have full faith in your abilities to figure something out You can also always look up krita's and mypaint's codes to see how they process it for clues.

[cestarian]

Update 2.10 was nice, still holding my breath for this to be fixed tho, it's the single worst bug imo.

If I wasn't clear about the issue, it's the tapering at the beginning/end of each stroke. Especially at the end.