Crystal Dreams II excluded from DemoDVD, unless hackable?
category: general [glöplog]
I thought I'd post this here in case someone capable considers the effort worth their while. Big kudos to anyone who bothers ;)
Here's a quote of the issue from demodvd.org:
"Unless I can work a miracle, the following demos are going to be excluded:
Crystal Dream II: Inits 67Hz video mode that evades my 60Hz tweaking. Entire demo is written in custom protected-mode extender that defies debugging. Stupid Borland Pascal CRT unit suffers from speed issue, preventing demo from running on any machine faster than a P166. (Unpacking the demo, fixing the Pascal speed issue, and repacking the demo results in the extender complaining about too many relocations.) Standalone converter drops frames to match 60Hz, which just ruins every effect because they're all full-framerate."
I guess if anyone is both a fan of old demos and is an old cracker as well, then this might not be so difficult?
Here's a quote of the issue from demodvd.org:
"Unless I can work a miracle, the following demos are going to be excluded:
Crystal Dream II: Inits 67Hz video mode that evades my 60Hz tweaking. Entire demo is written in custom protected-mode extender that defies debugging. Stupid Borland Pascal CRT unit suffers from speed issue, preventing demo from running on any machine faster than a P166. (Unpacking the demo, fixing the Pascal speed issue, and repacking the demo results in the extender complaining about too many relocations.) Standalone converter drops frames to match 60Hz, which just ruins every effect because they're all full-framerate."
I guess if anyone is both a fan of old demos and is an old cracker as well, then this might not be so difficult?
perhaps he should try a TSR to fix the Borland Pascal CRT thing. it seems to work for most prods.
(i'll try and remember to check if it works when I get home :)
(i'll try and remember to check if it works when I get home :)
How about actually running the demo on a machine less than P2?
:)
(i have a celeron300a for old demos and a cyrix p150+ for ancient demos).
:)
(i have a celeron300a for old demos and a cyrix p150+ for ancient demos).
if all hacking fails, they might still record it with a good camera off a good monitor, one could make an exception in such cases. still better than nothing.
I believe warp is right.The tsr proggy works in most cases.
I talked to Trixter about this (I was trying to help out, since we have some nifty video-gear here at the TG-office :)
Some quotes from Trixter about CD2:
==[ cut cut ]==
Me:
You're in luck, both the analogue and encoders I have
should handle this. The pascal stuff I have a fix for,
and I also have a machine that can run this demo > >
Trixter:
It's more complicated than that. Sure, the VGA->NTSC converters will handle it, but they drop frames to map the 67Hz video mode CD2 uses to 60Hz NTSC. The same would happen if you were using PAL -- even MORE frames would be dropped to map it to 50Hz PAL. The end result is that
the video looks jerky -- scrollers aren't smooth, etc.
No analog converter can do it properly.
(note: Things like the "Vectorplasma" will look awfull
since it uses the previously drawn frame to render the
effect)
As for the Pascal stuff, it cannot be fixed since when I unpack the exe and fix it and repack it, CD2 knows it has been changed and complains about too many relocations. There's a memory-resident fix that doesn't involve patching, but it doesn't work.
The REAL solution has to come from Vogue and Mr. H. I am going to email them after New Year's to see if they can at least patch the video mode to something other than what they're using. That's the most important part.
==[ cut cut ]==
I don't know what the progression is on the CD2-issue, but
I'll talk with Trixter about it again.
Some quotes from Trixter about CD2:
==[ cut cut ]==
Me:
You're in luck, both the analogue and encoders I have
should handle this. The pascal stuff I have a fix for,
and I also have a machine that can run this demo > >
Trixter:
It's more complicated than that. Sure, the VGA->NTSC converters will handle it, but they drop frames to map the 67Hz video mode CD2 uses to 60Hz NTSC. The same would happen if you were using PAL -- even MORE frames would be dropped to map it to 50Hz PAL. The end result is that
the video looks jerky -- scrollers aren't smooth, etc.
No analog converter can do it properly.
(note: Things like the "Vectorplasma" will look awfull
since it uses the previously drawn frame to render the
effect)
As for the Pascal stuff, it cannot be fixed since when I unpack the exe and fix it and repack it, CD2 knows it has been changed and complains about too many relocations. There's a memory-resident fix that doesn't involve patching, but it doesn't work.
The REAL solution has to come from Vogue and Mr. H. I am going to email them after New Year's to see if they can at least patch the video mode to something other than what they're using. That's the most important part.
==[ cut cut ]==
I don't know what the progression is on the CD2-issue, but
I'll talk with Trixter about it again.
yep. what gloom said :)
I'm sorry I didn't explain the issue more with my first post. I'd been following the demodvd project so to me it seemed like there was enough information to go by, but really I left some information out.
Here's my thought. Trixter got stumped when CD2 detected that the EXE had changed and refused to run. In my mind, at this point, preventing CD2 from noticing that the EXE has changed should be very analagous to (and probably easier than) preventing a copy-protected game from noticing that it has been modified (as is/was sometimes required for cracks).. so an experienced cracker/trainer should be able to handle this issue. The strange dos extender might limit the set of tools available, but some people are very good at ploughing through the raw code so I don't think it should be too much of a roadblock if someone works on it.
I'm guessing the plan after this is dealt with would be to tweak the timing of the effects, and to also either set a 60Hz graphics mode or dump all the gfx to the harddrive.
I'm sorry I didn't explain the issue more with my first post. I'd been following the demodvd project so to me it seemed like there was enough information to go by, but really I left some information out.
Here's my thought. Trixter got stumped when CD2 detected that the EXE had changed and refused to run. In my mind, at this point, preventing CD2 from noticing that the EXE has changed should be very analagous to (and probably easier than) preventing a copy-protected game from noticing that it has been modified (as is/was sometimes required for cracks).. so an experienced cracker/trainer should be able to handle this issue. The strange dos extender might limit the set of tools available, but some people are very good at ploughing through the raw code so I don't think it should be too much of a roadblock if someone works on it.
I'm guessing the plan after this is dealt with would be to tweak the timing of the effects, and to also either set a 60Hz graphics mode or dump all the gfx to the harddrive.
.. and of course it's possible that effects might need to be tweaked individually and this could get tedious.. and in some cases (such as plasma), effects will have to run their course at about 10% slower than the original, and get chopped at the end to synch to music, etc. I'm still just speculating though. The last time I watched this demo was more than 5 years ago and I'm not involved with the demodvd project.
(BTW, I apologise for calling it "Crystal Dreams II" in the topic instead of "Crystal Dream II". Pouet.net's entry has it spelled incorrectly too though, so I don't feel too ashamed :))
Last time that I could watch CD2 was on a pII-266 after starting 'bremze', a small tsr prog for slowing down the computer. But unfortunately this program takes no effect on my pIII-450 (actually used, I had never more than one computer at a time).
Could someone post a URL for the Pascal CRT-bug TSR again? I remember seeing it once but can't find it now, and I'd like to try it out.
Meanwhile, we now have an EXE for CD2 that will supposedly work on fast PCs, haven't heard the results yet. Now we're working on the 67->60Hz issue, and we're getting help there too. We just might be able to pull it off after all, thanks to the support from you guys.
Meanwhile, we now have an EXE for CD2 that will supposedly work on fast PCs, haven't heard the results yet. Now we're working on the 67->60Hz issue, and we're getting help there too. We just might be able to pull it off after all, thanks to the support from you guys.
I found the Pascal CRT-bug TSR at http://www.gwdg.de/~uconrad1/programe.htm
The direct link to the file is ftp://ftp.gwdg.de/pub/zoologie/programming/tp7p5fix.zip
The direct link to the file is ftp://ftp.gwdg.de/pub/zoologie/programming/tp7p5fix.zip
Thanks. As has been said, the TSR fix doesn't work with CD2.. but I just wanted to try it myself to make sure.
I would be grateful if someone had found a way to capture screens from Crystal Dream II. A legendary demo, but I would just hate putting a video-captured screenshot into Freax.
And what is Freax?
<Stupid Borland Pascal CRT unit suffers from speed issue, preventing demo from running on any machine faster than a P166.>
at least works fine with my p233mmx
at least works fine with my p233mmx
macaw: Give the URL http://www.freax.hu a try and you'll know ...
personally, I didn't have success with that link.. but did with http://www.freax.hu/welcome.html
Hmm... I've got a monochrome videocard + monitor and SoftICE skills, but no dos pc. I could put together my old 486/66, but I don't know if it works...
Duh, buying a used 386 with a crap mono monitor costs some $ 10.
And disabling internal cache in your bios is free. :)
Oh, and to follow up on the CD2 dilemma, Trixter posted a news item on www.demodvd.org.
Oh, and to follow up on the CD2 dilemma, Trixter posted a news item on www.demodvd.org.