Problematic Demos. - Page 45

amilo3438 · 24 November 2021, 20:24

Great thanks!

What about CEN Demo (Demo) by Rare on a real ECS A500 test?

Currently in the emu:
- ECS 0.5M chip only (glitches in the emu)
- ECS 0.5M+0.5M (no glitches only if subpixel display is enabled)
- ECS 1M chip only (no glitches)

Zarnal · 25 November 2021, 20:21

Quote:

Originally Posted by amilo3438

Great thanks!

What about CEN Demo (Demo) by Rare on a real ECS A500 test?

Currently in the emu:
- ECS 0.5M chip only (glitches in the emu)
- ECS 0.5M+0.5M (no glitches only if subpixel display is enabled)
- ECS 1M chip only (no glitches)

See post #877.

No glitches on real A500 ECS 512+512.

Subpixel mode is only A1200 related ( if I don't write nonsense ).

amilo3438 · 25 November 2021, 20:36

Quote:

Originally Posted by Zarnal

See post #877.

No glitches on real A500 ECS 512+512.

Subpixel mode is only A1200 related ( if I don't write nonsense ).

Yes I know but [A500 ECS 512] was not yet tested on a real A500!?
(on A500 OCS 512 in the emu it is working fine)

Regarding subpixel mode in ECS, yes is a little weird as it is suppose to be only for AGA!?
EDIT: In winuae.7z (2021.11.20) was working fine for [A500 ECS 512+512] but in next beta 41 (RC2) is broken again if subpixel mode is disabled!?

redblade · 25 November 2021, 22:39

Wow that Cen DEMO looks like a nice 16 colour font in the logo

Toni Wilen · 27 November 2021, 17:27

Cen is now properly fixed.

Optimized bitplane emulation mode didn't properly check if bitplane pointer is valid in ECS Agnus 0.5M+0.5M config where Agnus sees 1M chip but CPU sees 512k chip ram and 512k of "slow" RAM. Optimized mode (which uses direct memory access) read from blank 512k region, every 16th line was still drawn correctly because it has copper color changes which force non-optimized mode.

It also worked in subpixel mode because it always bypasses optimized mode (which can't handle subpixel mode).

amilo3438 · 27 November 2021, 17:53

Great, thanks!

Only unknown part left is what happens on a real [A500 512Mb chip only] amiga with this demo? (in the emu it is missing the vector objects if ECS - in OCS its fine)

Toni Wilen · 27 November 2021, 18:51

OCS Agnus 0.5M: vector object visible (works accidentally, bitplane pointer is out of 512k chip range but because OCS Agnus has max 512k support, pointer still points to chip ram and luckily this position does not have anything important that would crash the demo or cause graphics corruption...)

ECS Agnus 0.5M: vector object disappears, also garbage appears on screen because bitplane pointer points to upper 0.5M chip which ECS Agnus but without connected memory it returns noise. This is not yet emulated because it isn't simply random data.

amilo3438 · 28 November 2021, 15:44

Fill 'Em All (Demo) by Vertigo

[A500 ECS 512+512] The text part is shown in ECS in pink background where in OCS is in black background (as on pic http://kestra.exotica.org.uk/files/s...3000/1146e.png).

Please find the file in the Zone! (vertigo_.uss)

Toni Wilen · 28 November 2021, 16:26

Quote:

Originally Posted by amilo3438

Fill 'Em All (Demo) by Vertigo

[A500 ECS 512+512] The text part is shown in ECS in pink background where in OCS is in black background (as on pic http://kestra.exotica.org.uk/files/s...3000/1146e.png).

Please find the file in the Zone! (vertigo_.uss)

A500 ECS confirmed.

Code:

00049028 2d7c 0000 c03f 0058      move.l #$0000c03f,(a6,$0058) == $00dff058

Buggy code that accidentally worked until ECS introduced new blitter register at $dff05a..

amilo3438 · 30 November 2021, 15:45

Minigolf Muzak (Demo) by The Link

[A500 ECS 512+512] Small glitches near the top of the screen if ECS! (in OCS is fine)

Toni Wilen · 30 November 2021, 18:48

Quote:

Originally Posted by amilo3438

Minigolf Muzak (Demo) by The Link

[A500 ECS 512+512] Small glitches near the top of the screen if ECS! (in OCS is fine)

ECS A500 confirmed.

Toni Wilen · 13 March 2022, 18:04

New feature: refresh/bitplane conflict cycles are now accurately emulated. winuae.7z only. No betas yet, 68000 IPL testing still to do.

For example these demos now have fully (*) accurate glitches if ECS Agnus:

http://janeway.exotica.org.uk/release.php?id=6029 (Only single conflict)
http://janeway.exotica.org.uk/release.php?id=2219 (Multiple conflict lines)
http://janeway.exotica.org.uk/release.php?id=19588 (Whole visible display! Everything!)

Note that glitches can change depending on unused memory contents, memory config and chipset model.

WARNING: music is also very glitchy in last two because they have too many conflict cycles (Paula can't sent audio DMA requests to Agnus if first refresh cycle conflicts). Technically emulation can easily keep audio intact in this situation but it would not be correct behavior.

Undocumented hardware feature post coming later that explains what is going on when refresh cycles conflict and why it looks like that. It is quite complex but there is nothing random.

*) there is one missing feature which can affect line's brightness or colors which is display device side-effect (not Amiga hardware generated) which may or may not be implemented in the future. It does not affect programs that have black background.

Zarnal · 16 March 2022, 12:11

Quote:

Originally Posted by Toni Wilen

No betas yet, 68000 IPL testing still to do.

I did not understand very well. Are the 68000 major IPL changes you mentioned already present in winuae.7z (and you just have to test to see if they haven't broken anything else) or are the major IPL changes not yet included in winuae.7z ?

Mclane · 16 March 2022, 13:00

I am sure he's saying there is the usual test zip up but just no official beta released as yet

So if you grab a test version you will have the changes..

Toni Wilen · 16 March 2022, 21:07

Quote:

Originally Posted by Zarnal

I did not understand very well. Are the 68000 major IPL changes you mentioned already present in winuae.7z (and you just have to test to see if they haven't broken anything else) or are the major IPL changes not yet included in winuae.7z ?

IPL changes are not yet done, testing software and hardware is still not fully done. Soon. Some related updates are already included.

Photon · 30 March 2022, 01:13

I was hoping to record Celebrandil's Progress, Peace, and Passion demo from NS+FLT MD 3 tonight (for his birthday!), but I have to give up.

There's a WHDLoad patch, and I know this particular part from the Megademo disk was the stopping point on my real Amiga back in the day. It's a nice one with a nice tune which I think is the one we in Phenomena later blasted out the car speakers going to a party from Stockholm to Gothenburg back then.

I believe Celebrandil had an A1000, but none of the disks I've tried even boot under 1.2/1.3/noslow/A1000 emu in WinUAE. So I'm posting it here to maybe be able to record it.

a/b · 30 March 2022, 16:55

This worked for me:
1. download part 15 exe (NorthStar-Progress) from janeway
2. start it in winuae
3. enter debugger during depacking, type il (enable exception breakpopints)
4. type g to continue
5. demo will crash after depacking (illegal instruction), you are in debugger now
6. type g 35000 (basically restart)

ross · 30 March 2022, 17:21

Or start "NorthStar-Progress" with KS1.2

(the first fake jmp on irq3 is good here)

amilo3438 · 30 March 2022, 19:04

Quote:

Originally Posted by ross

Or start "NorthStar-Progress" with KS1.2

(the first fake jmp on irq3 is good here)

On latest winuae.7z (2022.03.27) works also with KS1.3 OCS 512!

(P.s. Still cannot find enough time for a more demo tests.)

AmiClassic · 30 March 2022, 20:54

Quote:

Originally Posted by amilo3438

On latest winuae.7z (2022.03.27) works also with KS1.3 OCS 512!

Where find this release?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problematic GVP A530 Turbo	Mogsy	support.Hardware	22	13 November 2012 19:12
Problematic thread?	prowler	project.EAB	5	28 October 2011 00:06
Problematic Holiday Lemmings 94 demo disk	MethodGit	support.Games	19	08 November 2010 03:56
Problematic slaves list	keropi	project.Killergorilla's WHD packs	44	28 November 2008 12:46
Looking for two demos...	eternalcrusader	request.Demos	3	01 August 2006 08:57

24 November 2021, 20:24	#881
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,501	Great thanks! What about CEN Demo (Demo) by Rare on a real ECS A500 test? Currently in the emu: - ECS 0.5M chip only (glitches in the emu) - ECS 0.5M+0.5M (no glitches only if subpixel display is enabled) - ECS 1M chip only (no glitches)

25 November 2021, 22:39	#884
redblade Zone Friend Join Date: Mar 2004 Location: Middle Earth Age: 40 Posts: 2,127	Wow that Cen DEMO looks like a nice 16 colour font in the logo

27 November 2021, 17:27	#885
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,502	Cen is now properly fixed. Optimized bitplane emulation mode didn't properly check if bitplane pointer is valid in ECS Agnus 0.5M+0.5M config where Agnus sees 1M chip but CPU sees 512k chip ram and 512k of "slow" RAM. Optimized mode (which uses direct memory access) read from blank 512k region, every 16th line was still drawn correctly because it has copper color changes which force non-optimized mode. It also worked in subpixel mode because it always bypasses optimized mode (which can't handle subpixel mode).

27 November 2021, 17:53	#886
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,501	Great, thanks! Only unknown part left is what happens on a real [A500 512Mb chip only] amiga with this demo? (in the emu it is missing the vector objects if ECS - in OCS its fine)

27 November 2021, 18:51	#887
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,502	OCS Agnus 0.5M: vector object visible (works accidentally, bitplane pointer is out of 512k chip range but because OCS Agnus has max 512k support, pointer still points to chip ram and luckily this position does not have anything important that would crash the demo or cause graphics corruption...) ECS Agnus 0.5M: vector object disappears, also garbage appears on screen because bitplane pointer points to upper 0.5M chip which ECS Agnus but without connected memory it returns noise. This is not yet emulated because it isn't simply random data.

28 November 2021, 15:44	#888
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,501	Fill 'Em All (Demo) by Vertigo [A500 ECS 512+512] The text part is shown in ECS in pink background where in OCS is in black background (as on pic http://kestra.exotica.org.uk/files/s...3000/1146e.png). Please find the file in the Zone! (vertigo_.uss)

30 November 2021, 15:45	#890
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,501	Minigolf Muzak (Demo) by The Link [A500 ECS 512+512] Small glitches near the top of the screen if ECS! (in OCS is fine)

13 March 2022, 18:04	#892
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,502	New feature: refresh/bitplane conflict cycles are now accurately emulated. winuae.7z only. No betas yet, 68000 IPL testing still to do. For example these demos now have fully () accurate glitches if ECS Agnus: http://janeway.exotica.org.uk/release.php?id=6029 (Only single conflict) http://janeway.exotica.org.uk/release.php?id=2219 (Multiple conflict lines) http://janeway.exotica.org.uk/release.php?id=19588 (Whole visible display! Everything!) Note that glitches can change depending on unused memory contents, memory config and chipset model. WARNING: music is also very glitchy in last two because they have too many conflict cycles (Paula can't sent audio DMA requests to Agnus if first refresh cycle conflicts). Technically emulation can easily keep audio intact in this situation but it would not be correct behavior. Undocumented hardware feature post coming later that explains what is going on when refresh cycles conflict and why it looks like that. It is quite complex but there is nothing random. ) there is one missing feature which can affect line's brightness or colors which is display device side-effect (not Amiga hardware generated) which may or may not be implemented in the future. It does not affect programs that have black background.

16 March 2022, 13:00	#894
Mclane Old retro god. Join Date: Apr 2002 Location: Northolt, West London Age: 62 Posts: 857	I am sure he's saying there is the usual test zip up but just no official beta released as yet So if you grab a test version you will have the changes..

30 March 2022, 01:13	#896
Photon Moderator Join Date: Nov 2004 Location: Eksjö / Sweden Posts: 5,602	I was hoping to record Celebrandil's Progress, Peace, and Passion demo from NS+FLT MD 3 tonight (for his birthday!), but I have to give up. There's a WHDLoad patch, and I know this particular part from the Megademo disk was the stopping point on my real Amiga back in the day. It's a nice one with a nice tune which I think is the one we in Phenomena later blasted out the car speakers going to a party from Stockholm to Gothenburg back then. I believe Celebrandil had an A1000, but none of the disks I've tried even boot under 1.2/1.3/noslow/A1000 emu in WinUAE. So I'm posting it here to maybe be able to record it.

30 March 2022, 16:55	#897
a/b Registered User Join Date: Jun 2016 Location: europe Posts: 1,039	This worked for me: 1. download part 15 exe (NorthStar-Progress) from janeway 2. start it in winuae 3. enter debugger during depacking, type il (enable exception breakpopints) 4. type g to continue 5. demo will crash after depacking (illegal instruction), you are in debugger now 6. type g 35000 (basically restart)

30 March 2022, 17:21	#898
ross Defendit numerus Join Date: Mar 2017 Location: Crossing the Rubicon Age: 53 Posts: 4,468	Or start "NorthStar-Progress" with KS1.2 (the first fake jmp on irq3 is good here)

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)