WinUAE 4.9.1 beta series

[DZh]

Starting with 4.9.0 in 'PAL: High Res Laced' mode when the 'Remove Interlace Artifacts' is enabled, the screen shifted on 1 line up.

[Toni Wilen]

Originally Posted by DZh

I tried various options - OCS, all ECS, AGA but result is approximately the same. Also tried JIT and cycle-exact. Before beta 4 it works.

It might depend on previous screen mode. Can you duplicate it if you copy all required files to floppy image and boot from it?

[DZh]

Originally Posted by Zarnal

Can you upload somewhere your version please. I have tried with the version provided by Hexaae and I am unable to reproduce with many CE and JIT modes ?

I have the same version as hexaae.

[DZh]

Originally Posted by Toni Wilen

It might depend on previous screen mode. Can you duplicate it if you copy all required files to floppy image and boot from it?

I tried boot from floppy - all fine, without garbage.

[Toni Wilen]

I also can't duplicate it. Does it happen if you use your usual config (that has the issue) but boot without startup-sequence?

Do you run it from WB? Does it still happen if you change WB screen mode to something different? (or if RTG, not RTG or vice versa)

EDIT: does the glitch disappear if you resize window (forcing display refresh)?

[DZh]

1 Run from WB - problem. When window resized and refreshed - Ok. Next run - problem.
2 Run from CLI - Ok.
2 Run from non-WB - Ok.

[Toni Wilen]

Originally Posted by DZh

1 Run from WB - problem. When window resized and refreshed - Ok. Next run - problem.
2 Run from CLI - Ok.
2 Run from non-WB - Ok.

Which WB screen mode? RTG or native? Try changing it and see if it changes the corruption.

EDIT: I found a way to duplicate it (by filled line output buffer with random garbage every scanline, exactly same lines appear corrupted as in your screenshot)

[DZh]

I don't use RTG never. Any attempt to change at least one parameter in ScreenMode ends with crash. No, I know for sure that this is the ancient problem of my WB. But the last winuae.7z removes glitch in PR3.62.

[Toni Wilen]

https://download.abime.net/winuae/fi...nuae_4910b5.7z
https://download.abime.net/winuae/fi...ae64_4910b5.7z

Final: next week end.

Beta 5:

- Filtered lores had wrong colors if host graphics mode was 16-bit.
- Horizontal DIWSTOP = last possible horizontal position (+1 more would cause always on hdiw) didn't ignore DIWHIGH AGA H0/H1 bits, if either (or both) were set, always on diw was detected.
- Disk not in ADKCON FAST mode and normal 2us MFM disk in drive: disk index was unreliable.
- Some normal programmed WB modes unnecessarily enabled slower cycle-based custom chipset emulation.
- Another HBLANK fix, if previous line ends in border closed state and without hblank start, next line also starts in border closed state. Programmed HBLANK + AGA only, ECS Denise CSYNC blanking always resets border state when HSYNC starts.
- Previous A1000 3 lores pixel earlier change was not accurate, it is OCS Denise feature (including A1000 Denise) and it isn't hblank end difference but for some only bitplane data is not allowed until first 3 pixels, border works normally (Previous test used black border color which hide the 3 lores pixel border stripe). This is now fixed.
- In some situations old bitplane data was not properly cleared when number of planes decreased mid scanline (including change hidden in invisible part of scanline). b4 number of bitplane change side-effect.
- Adjusted overscan+/extreme positioning.

[DZh]

#101:
http://eab.abime.net/attachment.php?...1&d=1642332712 http://eab.abime.net/attachment.php?...1&d=1642332718

[Toni Wilen]

Originally Posted by DZh

#101:
http://eab.abime.net/attachment.php?...1&d=1642332712 http://eab.abime.net/attachment.php?...1&d=1642332718

I can't duplicate. Last line changes because artifact removal requires line "pairs" but interlace has normally odd number of lines but top does not change.

Do you have some centering or some filter option enabled?

[DZh]

My settings:

http://eab.abime.net/attachment.php?...1&d=1642349550 http://eab.abime.net/attachment.php?...1&d=1642349555

[DZh]

As you can see from my settings, the image is cling to the borders of the screen. Therefore, you can easily see the lack of one upper line in this example with the enabled Remove interlace artifacts.

[DZh]

http://eab.abime.net/attachment.php?...1&d=1642355161 http://eab.abime.net/attachment.php?...1&d=1642355165

[Toni Wilen]

Originally Posted by DZh

As you can see from my settings, the image is cling to the borders of the screen. Therefore, you can easily see the lack of one upper line in this example with the enabled Remove interlace artifacts.

I managed to duplicate this few days ago (It does not happen always in all situations)

I almost decided to consider this a "feature" in 4.9.1 but then I noticed sprites (mouse cursor) don't move vertically with the bitplane when enabling the option. Which can be become annoying.. I'll try to fix this but there isn't much time left..

[DZh]

By the way, I also noticed this (mouse cursor don't move vertically with the bitplane) when I tried to make screenshots, but did not take it into account. Well, that you noticed it.

[Toni Wilen]

Originally Posted by DZh

By the way, I also noticed this (mouse cursor don't move vertically with the bitplane) when I tried to make screenshots, but did not take it into account. Well, that you noticed it.

Does https://download.abime.net/winuae/files/b/winuae.7z fix the bitplane shift? Without introducing anything unexpected

Sprite issue seems to be mainly odd/even line difference, does not affect much in interlaced modes. Probably not going to fix this until 4.9.1 is out.

[DZh]

Remove interlace artifacts (RIA) is now working well as in 4.4.0. But after manipulations in the GUI, the cursor loses its position in the emulator. In full screen mode don't loses and shifted on 1 pixel when changed RIA.

[Toni Wilen]

winuae.7z updated. Should fix sprite vertical alignment with remove interlace artifacts enabled. (Make sure you get updated winuae.7z, sometimes old cached version gets downloaded)

[DZh]

In full screen mode, the cursor moves along with the bitplane (normal). In the window mode, when RIA changes (and enable and disable), the cursor rises to several tens of pixels up.