Display register experiment problems...

Jherek Carnelia · 21 April 2011, 22:49

Well, I'm experimenting changing the display window registers to create 'wipe' type of effects and I think I've bashed up against the limit of my knowledge

It pretty much works, and thanks to some earlier tips, I've fixed (bodged, probably), a major display glitch I was having - but there are some niggly little glitches and jerking in the display that I have no idea how to fix.

It would be brilliant if one of you gurus could take a look at my efforts and give me a clue as to what I need to look at.

I suspect that it may be due to my lack of knowledge in how the display is written and some sort of timing issue, although to be honest it may well be just my crappy coding! I dunno anymore.

I kinda feel like I've just given up, and in a way I have, but I really just can't think of what to do next, and a little input of expert knowledge could help me move on to the next level.

Cheers chaps.

Leffmann · 22 April 2011, 16:50

Can't get it to assemble with my old include files, DOSCALL and other macros are missing. These include files always were a complete mess, and there are several sets floating around with various modifications. What assembler, include files and tab size are you using? Maybe you could just describe the exact effect you're looking for, and someone might know a good way of achieveing it.

Jherek Carnelia · 22 April 2011, 19:42

Bah! Can't even upload something that works

Okay, I've attached an executable (with gfx files) that hopefully will work, and show what I'm trying to do.

I've also attached my includes if anyone wants to try and assemble with them

Lonewolf10 · 22 April 2011, 20:11

Just tried your executable version with various setups and couldn't get it to work. What is your exact configuration (e.g. OS version, CPU, RAM etc.)?

It's loading (I can see available memory dropping and not going back up again) and appears to do nothing.

Edit: Am looking at your source code now and trying to assemble it. However it is missing include "utilsV1.1.s" (unless that is a typo). Tried with "utilsV1.0.s". Some routines appear to be missing (e.g. waitline_to_end) when assembling with case sensitive off. With case sensitive on multiple symbols are defined twice.

Regards,
Lonewolf10

Toni Wilen · 22 April 2011, 22:08

My crystal ball says it is DIWHIGH calculation problem, I didn't bother to check it because the code is unnecessarily complex and has too much code duplication, sorry

Better use simple WORD sized displaystart and displaystop variables and only calculate DIWSTRT/DIWSTOP/DIWHIGH values in UpdateDisplayRegisters. (no unnecessary bitshifting or masking anywhere else)

Also make sure you update all above registers at the end of field (but before line zero!) because DIWSTRT/DIWSTOP are "immediate", for example if you are going to set vertical start to line 256, write 0 to DIWSTRT vertical position during line zero will immediately match, following write to DIWHIGH won't "cancel" it.

Jherek Carnelia · 22 April 2011, 22:58

@LoneWolf

What a plonker I am

i attached the wrong file. I've attached the right one now...
The config is a basic A4000, nothing fancy. Attached as well.

@Toni

Hmm. I need to think about what you suggest - I don't fully understand it

(You're right that there is a lot of duplication - I don't want to remove that redundancy until it works, and then I'll simplify it - if I ever get it working, that is!)

Lonewolf10 · 23 April 2011, 01:39

The good news is that with your A4000 config I was able to run your executable

A quick rundown of what I found (numbers indicate picture number):

1 - no glitch
2 (machine) - "glitches" at what I think is your waiting point ~line 250? *
3 - as 2
4 - glitches 1/4 way through opening - horizontal calc out of sync for 1VBL. You need to check your DIWSTRT & DIWSTOP values.
5 - glitches on opening - horizontal calc out of sync for 1VBL. Same as 4.

* I believe it's your waiting point, as the scanlines below this point are 1 pixel wider/smaller on either side than the scanlines above it.

You are clearly using WinUAE (or UAE). Did you try going to the "display" option and adjusting the frame rate? I ran your program a few times noting roughly when it glitches. Then I ran it again, adjusting the FPS down to 1FPS just before the glitches occur. That is how I can see what is happening.

Regards,
Lonewolf10

Jherek Carnelia · 23 April 2011, 13:32

Quote:

Originally Posted by Lonewolf10

The good news is that with your A4000 config I was able to run your executable

A quick rundown of what I found (numbers indicate picture number):

1 - no glitch
2 (machine) - "glitches" at what I think is your waiting point ~line 250? *
3 - as 2
4 - glitches 1/4 way through opening - horizontal calc out of sync for 1VBL. You need to check your DIWSTRT & DIWSTOP values.
5 - glitches on opening - horizontal calc out of sync for 1VBL. Same as 4.

* I believe it's your waiting point, as the scanlines below this point are 1 pixel wider/smaller on either side than the scanlines above it.

You are clearly using WinUAE (or UAE). Did you try going to the "display" option and adjusting the frame rate? I ran your program a few times noting roughly when it glitches. Then I ran it again, adjusting the FPS down to 1FPS just before the glitches occur. That is how I can see what is happening.

Regards,
Lonewolf10

Thanks Lonewolf. Good tip about the FPS slider. Back to the drawing board, I guess.

Lonewolf10 · 23 April 2011, 16:17

Quote:

Originally Posted by Jherek Carnelia

Thanks Lonewolf. Good tip about the FPS slider. Back to the drawing board, I guess.

No problem. I only discovered it myself a few months back. Very useful for debugging glitches.

I have spent a few hours looking at your code and couldn't find the cause of the glitch.

Regards,
Lonewolf10

Jherek Carnelia · 23 April 2011, 19:33

Quote:

Originally Posted by Lonewolf10

I have spent a few hours looking at your code and couldn't find the cause of the glitch.

Regards,
Lonewolf10

Wow, Thanks for trying. It's a very annoying thing

Lonewolf10 · 07 May 2011, 15:00

Quote:

Originally Posted by Jherek Carnelia

Wow, Thanks for trying. It's a very annoying thing

I agree.

Did you manage to fix it in the end?

Regards,
Lonewolf10

Jherek Carnelia · 07 May 2011, 22:40

No.

I've put it aside and moved on to other things. I'll go back to it later.

21 April 2011, 22:49	#1
Jherek Carnelia Dazed and Confused Join Date: Dec 2001 Location: portsmouth/uk Posts: 242	Display register experiment problems... Well, I'm experimenting changing the display window registers to create 'wipe' type of effects and I think I've bashed up against the limit of my knowledge It pretty much works, and thanks to some earlier tips, I've fixed (bodged, probably), a major display glitch I was having - but there are some niggly little glitches and jerking in the display that I have no idea how to fix. It would be brilliant if one of you gurus could take a look at my efforts and give me a clue as to what I need to look at. I suspect that it may be due to my lack of knowledge in how the display is written and some sort of timing issue, although to be honest it may well be just my crappy coding! I dunno anymore. I kinda feel like I've just given up, and in a way I have, but I really just can't think of what to do next, and a little input of expert knowledge could help me move on to the next level. Cheers chaps. Last edited by Jherek Carnelia; 29 April 2011 at 11:31.

22 April 2011, 16:50	#2
Leffmann Join Date: Jul 2008 Location: Sweden Posts: 2,269	Can't get it to assemble with my old include files, DOSCALL and other macros are missing. These include files always were a complete mess, and there are several sets floating around with various modifications. What assembler, include files and tab size are you using? Maybe you could just describe the exact effect you're looking for, and someone might know a good way of achieveing it. Last edited by Leffmann; 22 April 2011 at 16:56.

22 April 2011, 19:42	#3
Jherek Carnelia Dazed and Confused Join Date: Dec 2001 Location: portsmouth/uk Posts: 242	Bah! Can't even upload something that works Okay, I've attached an executable (with gfx files) that hopefully will work, and show what I'm trying to do. I've also attached my includes if anyone wants to try and assemble with them Last edited by Jherek Carnelia; 29 April 2011 at 11:31.

22 April 2011, 20:11	#4
Lonewolf10 AMOS Extensions Developer Join Date: Jun 2007 Location: near Cambridge, UK Age: 44 Posts: 1,924	Just tried your executable version with various setups and couldn't get it to work. What is your exact configuration (e.g. OS version, CPU, RAM etc.)? It's loading (I can see available memory dropping and not going back up again) and appears to do nothing. Edit: Am looking at your source code now and trying to assemble it. However it is missing include "utilsV1.1.s" (unless that is a typo). Tried with "utilsV1.0.s". Some routines appear to be missing (e.g. waitline_to_end) when assembling with case sensitive off. With case sensitive on multiple symbols are defined twice. Regards, Lonewolf10 Last edited by Lonewolf10; 22 April 2011 at 20:31.

22 April 2011, 22:58	#6
Jherek Carnelia Dazed and Confused Join Date: Dec 2001 Location: portsmouth/uk Posts: 242	@LoneWolf What a plonker I am i attached the wrong file. I've attached the right one now... The config is a basic A4000, nothing fancy. Attached as well. @Toni Hmm. I need to think about what you suggest - I don't fully understand it (You're right that there is a lot of duplication - I don't want to remove that redundancy until it works, and then I'll simplify it - if I ever get it working, that is!) Last edited by Jherek Carnelia; 29 April 2011 at 11:31.

22 April 2011, 22:08	#5
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,505	My crystal ball says it is DIWHIGH calculation problem, I didn't bother to check it because the code is unnecessarily complex and has too much code duplication, sorry Better use simple WORD sized displaystart and displaystop variables and only calculate DIWSTRT/DIWSTOP/DIWHIGH values in UpdateDisplayRegisters. (no unnecessary bitshifting or masking anywhere else) Also make sure you update all above registers at the end of field (but before line zero!) because DIWSTRT/DIWSTOP are "immediate", for example if you are going to set vertical start to line 256, write 0 to DIWSTRT vertical position during line zero will immediately match, following write to DIWHIGH won't "cancel" it.

23 April 2011, 01:39	#7
Lonewolf10 AMOS Extensions Developer Join Date: Jun 2007 Location: near Cambridge, UK Age: 44 Posts: 1,924	The good news is that with your A4000 config I was able to run your executable A quick rundown of what I found (numbers indicate picture number): 1 - no glitch 2 (machine) - "glitches" at what I think is your waiting point ~line 250? * 3 - as 2 4 - glitches 1/4 way through opening - horizontal calc out of sync for 1VBL. You need to check your DIWSTRT & DIWSTOP values. 5 - glitches on opening - horizontal calc out of sync for 1VBL. Same as 4. * I believe it's your waiting point, as the scanlines below this point are 1 pixel wider/smaller on either side than the scanlines above it. You are clearly using WinUAE (or UAE). Did you try going to the "display" option and adjusting the frame rate? I ran your program a few times noting roughly when it glitches. Then I ran it again, adjusting the FPS down to 1FPS just before the glitches occur. That is how I can see what is happening. Regards, Lonewolf10 Last edited by Lonewolf10; 23 April 2011 at 01:44. Reason: Added explaination of "waiting point".

07 May 2011, 22:40	#12
Jherek Carnelia Dazed and Confused Join Date: Dec 2001 Location: portsmouth/uk Posts: 242	No. I've put it aside and moved on to other things. I'll go back to it later.

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

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Amiga CD32 Display problems	Chiptune	support.Hardware	16	17 February 2011 05:13
Display Problems on A1200	campbl2	support.Hardware	6	26 May 2010 21:12
Display problems	ginge73	support.Hardware	8	20 July 2009 10:21
Display Problems with WinUAE 0.9.92	Lorday	support.WinUAE	5	19 February 2005 00:08
Having problems with wunuae display	Ryan15UK_Dex	New to Emulation or Amiga scene	1	13 March 2003 19:38