Problematic AGA demos.

[Toni Wilen]

Originally Posted by amilo3438

ComaDose (40k Intro) by Passion -> http://janeway.exotica.org.uk/release.php?id=11008 (A1200 basic)(pic.1,2,3) ... when part (pic.1) ends (pic.2) it quick shows/appears a strange screen (pic.3) !

Not A1200 confirmed. Vector object means the usual. (I am still not interested in anything else than chipset problems)

Lazy (40k Intro) by Fake -> http://janeway.exotica.org.uk/release.php?id=18710 (A1200 basic)(pic.4) ... lower part of screen! (when CE=OFF then is fine ... and also if 4Mb Fast)

Very similar but worse looking glitches on real A1200.

Party Intro (40k Intro) by In-sect -> http://janeway.exotica.org.uk/release.php?id=11010 (A1200 basic)(pic.5) ... on face guard shading part strange red dots appear!

Stupid demo. Forgets to set FMODE register and only works if booted to AGA WB mode.. (At the moment I am not interested in finding or transferring WB3.0 disk)

Rose (40k Intro) by Chryseis & Hoodlum -> http://janeway.exotica.org.uk/release.php?id=10694 (A1200 basic)(no pic) ... on cube part sometimes glitches! (perhaps CPU speed issue)

A1200 confirmed.

Symptom (40k Intro) by Freezers -> http://janeway.exotica.org.uk/release.php?id=18704 (A1200 basic)(pic.6) ... glitches! (when CE=OFF then is fine) ... but it glitches also on 4Mb Fast!

Not A1200 confirmed. Too fast CPU most likely because demo immediately switches to following part.

[amilo3438]

The Tranquilizer (40k Intro) by Smellon -> http://janeway.exotica.org.uk/release.php?id=23797 (A1200 basic)(pic.1) ... glitches! (when CE=OFF then is fine)

Yaontalama (40k Intro) by Mg Design -> http://janeway.exotica.org.uk/release.php?id=10659 (A1200 basic)(pic.2) ... strange dots !

----------------------------------------------------------------------------------------------------------
Well, I think then that will be all from 1995 year, except following still need to be tested:

1995 AGA Demo -> http://janeway.exotica.org.uk/search...mat=&more=show (pages 5-8 not tested)

1995 AGA 4k Intro -> http://janeway.exotica.org.uk/search...mat=&more=show (not tested)
----------------------------------------------------------------------------------------------------------

My one year volunteering has just expired!

Cheers

[Zilog]

Hi,
this demo is Rift by The Black Lotus, is desing for aga/060 but i go only with 68040; with 68060 generate error.

I use Workbench 3.1, with Mu680x0Libs, naturally use only 68060.library and i delete 68040.library, i try with jit and with mmu without jit naturally.

Error:


Download Demo
https://www.scene.org/file.php?file=...l.zip&fileinfo

YouTube Demo
[ Show youtube player ]

[Snake79]

Originally Posted by Zilog

Hi,
this demo is Rift by The Black Lotus, is desing for aga/060 but i go only with 68040; with 68060 generate error.

I use Workbench 3.1, with Mu680x0Libs, naturally use only 68060.library and i delete 68040.library, i try with jit and with mmu without jit naturally.

I am using Quickstart A4000 (68040) config + Motherboard Fast (64MB), mounting Workbench 3.1 HDF or ADF (from Amiga Forever 2012), adding Rift Directory, booting, and it is working perfectly fine... PS. Great Demo!

[Vot]

Toni,

Your a machine. I get tired reading these posts let alone, verifying and bugfixing them. I always wondered although, would the decapping of the custom chips help you find all of these corner cases. Like this. http://www.visual6502.org/JSSim/ they decapped, photographed then vectorized the silicon so it could be simulated..

If so the community should all chip in to have it done.. That project i think had the chips to do it, but no schedule...

Vot

Originally Posted by Toni Wilen

A1200 confirmed.



Uses some strange programmed display mode that is sort of PAL compatible but also sets BEAMCON0 bits that no one has set previously ("DUAL" bit which apparently does something strange).

It looks even worse in older versions.



A1200 confirmed. (right mouse button exists so it isn't completely hung)



A1200 confirmed. Hangs with "FU"



A1200 confirmed. (Finally glitch that looks like CPU problem but is actually correct behavior! Not 100% identical glitch but mostly same and in same positions)

Bonus:

White bar in "Shade cluster" part is fixed. (Lots of requirements to trigger it: AGA + at least 1 sprite in horizontal border + bitplane active in same scanline + BPLCON4 XOR bits non-zero)

[Zilog]

Originally Posted by Snake79

I am using Quickstart A4000 (68040) config + Motherboard Fast (64MB), mounting Workbench 3.1 HDF or ADF (from Amiga Forever 2012), adding Rift Directory, booting, and it is working perfectly fine... PS. Great Demo!

Yes also to me function with 68040, the problem if is use 68060. I post for understand if a problem depend from emulator.

[Toni Wilen]

Originally Posted by Zilog

Yes also to me function with 68040, the problem if is use 68060. I post for understand if a problem depend from emulator.

Nothing to do with emulation. You need dummy 68040.library (small few hundred bytes file, redirects to 68060.library) because 3.1 C:SetPatch does not know nothing about 68060 or 68060.library. Or download newer SetPatch from Aminet.

[Zilog]

Originally Posted by Toni Wilen

Nothing to do with emulation. You need dummy 68040.library (small few hundred bytes file, redirects to 68060.library) because 3.1 C:SetPatch does not know nothing about 68060 or 68060.library. Or download newer SetPatch from Aminet.

Tony Thanks a lot! Now function perfectly.

[amilo3438]

JFI...

Checking Raiden demo from this thread -> http://eab.abime.net/showpost.php?p=838632&postcount=64

found that scrolling text glitches after WinUAE2700 version -> see on pic.2 below (but if CE=OFF seems fine and also if CE=ON with +4MB Fast)


EDIT: Weird, it seems to happens only if the disk is inserted after WB 3.0 boot screen ?! (but also it doesnt glitch in WinUAE2700!)
EDIT2: Seems it doesnt happen with WB 3.1!

(p.s. WinUAE config. was Quickstart A1200 base model)

[amilo3438]

Originally Posted by amilo3438

EDIT2:

Found temporary solution to run "Olympe (Trackmo) by Gods" on newer WinUAE versions -> see pic.2 !

After choosing QS1200 (Basic) config and changing the "CPU Emulation speed" to "Fastest possible" it runs fine.

(Weird, it refuse to run in "Approximate A500/A1200 or cycle-exact" mode even with CE=OFF ?!)

JFI ... the demo Olympe/Gods:

In Approximate A500/1200 mode (mode when CE=OFF) by default it stops at track 77 on Olymp rotation part, but ...

if in config file value of cpu_throttle >=-626.0 till -461.0 or (>=066.0 till 5000.0) it passes to next part. (see on pic. below)
(so it doesnt pass only if cpu_throttle= <-626.0 or >-461.0 till 066.0)

Starting from -464.0 till -461.0, it waits more and more time at last small rotation Olymp, before pass to next part.
(Note: if cpu_throttle >=-460.0 (till 066.0) it waits forever)

Also, starting from -561.0 till -626.0, it waits more and more time at first bigger rotation Olymp ... but it doesnt wait at small rotation Olymp at all.
(Note: if cpu_throttle <=-627 it waits forever)

-----------------------------------------------------------------------------------------------------------------
EDIT: Below are attached nice WinUAE A1200 configurations made for an old PC that seems to work just fine !!!
-> (The configs are standard QS A1200 but with CE=OFF + D3D + Legacy VSync + Fastest possible or Approximate)

... works fine on my slow old desktop PC and seems to be enough compatible! (f.e. Olympe/Gods demo works)
-----------------------------------------------------------------------------------------------------------------
BTW it seems Fastest possible % slider doesnt work right ... there is huge difference between -10% and 0% ?! (in below file is set to -000.1 what is -0.01%)

[Zilog]

Tony is interest for you at demo that function without jitter and not function with jitter?

[amilo3438]

I was playing with "most likely CPU emulator timing" issues demos mentioned on the 1st page of these threed.

My goal was to try to find a CPU Custom Freq on which most of problematic demos might work correctly.

I managed to find some interesting CPU Custom Freq, so here below are my findings:


(Note: used A1200 basic configuration (as f.ex. attached below) and all tests were done in WinUAE 3.1.0 beta18!)

------------------------------
CPU Freq = Custom 12.737830 ... (SysInfo 4.0 = 1.05 , 14.90Mhz -> just for comparison!)
------------------------------
STONE_ARTS-Bjarne => the lowest freq for demo to not glitch!(otherwise it glitches in picture rotation part)
MirageUk-SubliminalXtc => the lowest freq for demo to not glitch!(otherwise it shows strange line in circle)

------------------------------
CPU Freq = Custom 12.737829 ... (SysInfo 4.0 = 1.00 , 14.90Mhz -> just for comparison!)
------------------------------
Gods-Olympe => the highest freq for demo to work! (otherwise it stops after loading at track 77)
CHRYSEIS&POSEIDON-botm => the highest freq for demo to not glitch! (otherwise it glitches on PHOTO$ part)
DAMAGE-Uquala => the highest freq for demo to not glitch! (otherwise it glitches on KOLORKI plasma part)
US-YulBrynnerwSkinh => the highest freq for demo to not glitch! (otherwise it glitches on "pink" rotation part)


NOTE: Did you noticed how close are above mentioned Custom frequences! (the difference is only 0.000001)

------------------------------
CPU Freq = Custom 12.171703
------------------------------
C_LOUS-Cosmicglide => the highest freq for demo to not glitch! (otherwise it glitches on on ring rotation part)


p.s.
Yeah, I know, CPU Custom Freq is working in async to chipset in comparrison to 4x(A1200) that works in sync to chipset, but anyway,
hope that this findings perhaps might still be helpful in further enhancing of A1200 emulation, so I decided to publish them here.


Cheers!

[amilo3438]

further ... regarding other problematic demos not mentioned in above post that might work or not:

demos that works: (probably never been problematic!?)

Access_AGA => 4x(A1200) seems to work OK! (or any of above mentioned Custom freq)
DARKNESS-Dogtro => 4x(A1200) with 4MB fast seems to work OK! (or same with any of above mentioned Custom freq)
Freezers-Lech => 4x(A1200) with 4MB fast seems to work OK! (or same with any of above mentioned Custom freq)
NOVA-Soho => 4x(A1200) seems to work OK! (or any of above mentioned Custom freq)
STONE_ARTS-Vidar => 4x(A1200) with 4MB fast seems to work OK! (or same with any of above mentioned Custom freq)
QUARTEX-40kPartyWatEenGoreKop => 4x(A1200) seems to work OK! (or any of above mentioned Custom freq)(but not sure does it ends same on a real amiga too!?)


demos that only works without glitches if CE=OFF: (but usually running faster than they should)

INFECT-Harmagedon => (otherwise after a while only black screen with music)
Insane-Despair => (otherwise it stops after a while at text part)
SHD-Revelation => (otherwise it glitches on plasma part)
TRIUMPH-Invention => (otherwise it glitches) ... (tested in WinUAE 3.2.0.b11 -> no problem with +4MB RAM)
X_DREAM-Borntro => (otherwise it glitches)
ZOMO-Speed => (otherwise it glitches)


and for these demos seems nothing helps:

C_LOUS-Mustard => (it still glitches) ... (tested in WinUAE 3.2.0.b11 -> no problem with +4MB RAM)
MYSTIC-Vital => (it stops after a while on "green" part and after of some 22sec waiting with black screen it continues ... if CE=OFF it behaves worse)
Noxious-BeyondBelief => (it stops at cubes rotation part ... or if CE=OFF never reach to cube part) ... (tested in WinUAE 3.2.0.b11 -> no problem with +4MB RAM)
Oops-Anniversary => (it glitches on upper left screen part) ... (tested in WinUAE 3.2.0.b11 -> no problem with +4MB RAM)



EDIT:

demos that only works if CE=OFF and also "More compatible"=OFF:

grv-beut => (otherwise it stops after cube rotation part) ... seems fixed! (tested in WinUAE 3.2.0.b11)


in addition... possibly not AGA demos:

Bliss-Out (Demo) by Ukonx => seems not AGA demo! (works fine on A500 ECS 1mb)
-> http://eab.abime.net/showpost.php?p=...&postcount=509)


EDIT 2

demos that needs correct KS version to work without problems: (http://eab.abime.net/showpost.php?p=937202&postcount=24)

The Breath of the Muse (Demo) by Chryseis & Poseidon => needs KS 3.0 !!! (otherwise after loading will show a guru)
http://janeway.exotica.org.uk/release.php?id=10095 (Released: 1994-04-24 on Saturne Party 2 1994)

Twisted -Wild- (Trackmo) by Polka Brothers => needs KS 3.1 !!! (otherwise glitches on disk-C and stops - tested with WinUAE 3.1.0 beta 21)
http://janeway.exotica.org.uk/release.php?id=10549 (Released: 1994-12-28 on The Party 1994)

[amilo3438]

Just to mention something regarding demo Twisted by Polka Brothers -> http://janeway.exotica.org.uk/release.php?id=10549

seems that it needs KS 3.1, otherwise will glitch on disk-C and stop (as shown in attached pic)

(tested with WinUAE 3.1.0 beta 21)

EDIT

The glitching starts after pic.2 on disk-C !

EDIT 2

According to this [ Show youtube player ] (Captured on an Amiga 1200)

starting at 4:04, seems it also glitches but differently (hmm, perhaps KS3.0 A1200 ?!)

(Btw., starting at 1:50 there is no any glitches that appears in emulation?! -> perhaps emu CPU timing problem, right?)


p.s.
Uh, sorry... seems the problem belongs more to "Problematic AGA demos" thread! (so please remove it there if needed)

p.s.2
Just to say... with this finding I finished with testing demos for WinUAE 3.1.0 release!
(Hope that nothing fundamental will change until the final version. )

[amilo3438]

JFI ... updated post #93 - additional test (see text in red -> http://eab.abime.net/showpost.php?p=...9&postcount=93)

p.s. New test done with WinUAE 3.2.0.b11 (Note: old test done with some version before a WinUAE 3.1.0 release!)

EDIT:

Originally Posted by amilo3438

and for these demos seems nothing helps:
MYSTIC-Vital => (it stops after a while on "green" part and after of some 22sec waiting with black screen it continues ... if CE=OFF it behaves worse)

this demo MYSTIC-Vital behaves very weird... there are 2 parts with a black screen where music stops, but it will continue after waiting a while till the end!

[amilo3438]

Originally Posted by amilo3438

JOnboCken by candle -> http://www.pouet.net/prod.php?which=62418

A1200 (no-expanded)(2710b8) ... HALT1 !!!

found more links:

http://aminet.net/package/demo/aga/CDL-JOnboCken
http://www.m4rko.com/candle/?n=JOnboCken

it says on aminet -> Requirements: 020+, 2MB chip (AGA), >=14MHz (A1200)

but it still does not work! (WinUAE 3.2.0.b11)

[amilo3438]

tested winuae_3200b20

(noticed a new option in Chipset tab -> "Cycle-exact (DMA/Memory accesses)")

tested with settings A1200 quickstart basic and...

Cycle-exact (Full) = ON/OFF
Cycle-exact (DMA/Memory accesses) = ON


Have tested again all demos from post #92:

STONE_ARTS-Bjarne => it does not glitches in picture rotation part anymore!.....(Cycle-exact (Full) = ON)
MirageUk-SubliminalXtc => it does not shows strange line in circle anymore!......(Cycle-exact (Full) = ON/OFF)
Gods-Olympe -> it does not stops after loading at track 77 anymore!..............(Cycle-exact (Full) = OFF)
CHRYSEIS&POSEIDON-botm => it does not glitches on PHOTO$ part anymore!...(Cycle-exact (Full) = OFF)
DAMAGE-Uquala => it does not glitches on KOLORKI plasma part anymore!........(Cycle-exact (Full) = OFF)
US-YulBrynnerwSkinh -> it does not glitches on "pink" rotation part anymore!...(Cycle-exact (Full) = OFF)
C_LOUS-Cosmicglide -> it does not glitches on on ring rotation part anyomore! (Cycle-exact (Full) = OFF)


And some problematic demos from post #93:

INFECT-Harmagedon => no after a while only black screen with music! (Cycle-exact (Full) = OFF)
Insane-Despair => it does not stops after a while at text part!...........(Cycle-exact (Full) = OFF)
SHD-Revelation => it still glitches on plasma part!............................(Cycle-exact (Full) = OFF - does not help!)
X_DREAM-Borntro => it still glitches!..............................................(Cycle-exact (Full) = OFF - does not help!)
ZOMO-Speed => it still glitches!....................................................(Cycle-exact (Full) = OFF - does not help!)


Though CPU timings seems still not quite exact, anyway can be noticed some significant improvement in emulation! (Thx)

[vagrant]

Have been supplied a very rare trackmo from exolon called "Dreams 3" by TRSI.
It always freeze half way through demo when I assume it is trying to access disk 2.
Have tried many different config including quickstart A1200 basic and itterations.
The demo does not even start without CE enabled.

[Toni Wilen]

Originally Posted by ED-209

Have been supplied a very rare trackmo from exolon called "Dreams 3" by TRSI.
It always freeze half way through demo when I assume it is trying to access disk 2.
Have tried many different config including quickstart A1200 basic and itterations.
The demo does not even start without CE enabled.

Looks like image is corrupted. "CHECKSUM-ERROR at track 78. O:ED9E N:EE33" says Amiga DMS.

[vagrant]

Oh no.. that is a shame
Thanks for checking it out Toni