Emulating Pioneer LaserActive (Mega-LD) games

Moderator: Mask of Destiny

Post Reply
Nemesis
Very interested
Posts: 791
Joined: Wed Nov 07, 2007 1:09 am
Location: Sydney, Australia

Post by Nemesis » Thu Apr 19, 2012 12:01 am

Anyways, if this project needs any hardware or software I have quite a bit that can be loaned out no problem.
Wow, that's quite a collection! I'm trying to get this audio track issue ironed out right now, but once I'm done and have the Space Berserker videos online, I'll send you a message when I'm ready to accept loans of disks for dumping purposes.

In terms of hardware, I don't need any hardware loaned at this point, but one thing that does need to be done is to dump the bios from the NEC PAC-N1/PAC-N10. I don't have one in my collection, but I also don't really have the equipment to do it either. Not sure who would be up to the task, but I suspect the controller IC we're currently reverse-engineering the interface to will be the same in the NEC PAC, meaning there's probably no real testing that needs to be done on the PAC-N10 itself, we just need the bios dump, and combined with the LDROM2 game dumps, and the notes from reverse-engineering the PAC-S10, someone should be able to come along later and emulate the system if they put the time in.
Also, I had a thought... It might be better to tap the LD video before it even reaches the chip where it gets mixed with PAC video. I'll look at the schematic again, but we can probably wire up a system to tap video way before the composite out jack on the back. Might help with quality etc.
Intercepting the video earlier in the pipeline is a good idea. I have a second LaserActive unit for parts that I can experiment on. I'll take a look and see if there's a good point to grab the signal. If it comes out cleaner, I can use that instead.
I have heard that when LD video (movie or game) is run through a PAC, the video is run in a different (lower quality) mode than when playback is done without a PAC (movies only of course) Maybe someone else can chime in on more details concerning this, but if its true, then maybe we can look into modding a system specifically for capturing this video.
That might be only when the digital buffer is enabled? Most games turn the buffer on, so the video signal is converted to a digital image inside the LaserActive, and the video output comes from that. I have noticed the quality doesn't seem to be as good with the buffer on, but I don't have to enable the buffer when I play the video track.


One thing I've been thinking about over the last few days is how to capture video data in the "overscan" regions of the video. I noticed with Space Berserker, when messing around with some overlay bios settings, I triggered an invalid video mode which caused the video data in the overscan regions to be visible, and I noticed the overscan region contains time codes. The time code data could be very useful to capture. I'm going to try and figure out a way to get a stable image with the overscan regions fully visible, so I can rip the time code data along with the frame data.

Nemesis
Very interested
Posts: 791
Joined: Wed Nov 07, 2007 1:09 am
Location: Sydney, Australia

Post by Nemesis » Thu Apr 19, 2012 12:04 am

I should add, I might have to take some time out soon to repair my PAC-S10. I took the cover off a few weeks ago, and while none of the capacitors had leaked yet, a few are bulging. I have a PAC-S1 in reserve which looks ok, but it really needs the caps replaced too, and I don't want to lose either of the units, so I'll probably refurbish my PAC units before I start ripping these disks on-mass. I don't want a dead PAC interrupting the ripping process, or worse, writing off my LaserActive unit.

BlueBMW
Newbie
Posts: 2
Joined: Tue Apr 17, 2012 4:15 pm
Location: Knoxville, TN

Post by BlueBMW » Thu Apr 19, 2012 12:56 am

Nemesis wrote:I should add, I might have to take some time out soon to repair my PAC-S10. I took the cover off a few weeks ago, and while none of the capacitors had leaked yet, a few are bulging. I have a PAC-S1 in reserve which looks ok, but it really needs the caps replaced too, and I don't want to lose either of the units, so I'll probably refurbish my PAC units before I start ripping these disks on-mass. I don't want a dead PAC interrupting the ripping process, or worse, writing off my LaserActive unit.
If you need the info, here's posting where I have some charts listing the caps etc for the pacs. http://nfggames.com/forum2/index.php?topic=4434.0 and http://nfggames.com/forum2/index.php?topic=4390.0

Do your caps sooner rather than later. The worse the leakage gets, the harder it gets. Tasco's pac that I have here had a lot more leakage than it first appeared to have. Lots of pads lifted and there was lots of corrosion to clean up. Its a lot easier if you can avoid all the extra cleanup / repair work.

Regarding the NEC pac bios dumps... those pacs dont have an easily removed bios chip like the Sega PACs do... I'm sure its stored on one of the chips, but they're all surface mount, so it may be tricky to desolder then read out the chip. I know it can be done, just a little riskier. Its too bad that NEC pacs are so expensive.


EDIT:

Did some looking over the schematic.... looks like we might want to look at the LD video inputs at the BA7230LS chip on the RGBB board. That chip looks to be the final encoder that combines the LD video with the RGB overlay from the PAC.

Schematic: http://www.cyberroach.com/new_laseracti ... age059.gif

Datasheet for the BA7230LS: http://pdf1.alldatasheet.com/datasheet- ... 230LS.html


ALSO:

Someone mentioned to me an idea.... if a program can be written for the Sega PAC to control the LD... couldn't a homebrew Laseractive game be written say to utilize the Arcade Laserdiscs from games like Dragons Lair? Kind of rare discs, but would be an interesting project....

Xavier
Interested
Posts: 16
Joined: Wed Sep 30, 2009 4:21 am

Post by Xavier » Sat Apr 28, 2012 6:53 am

Please let me know if theres anything I can do to help!

Misty
Interested
Posts: 15
Joined: Mon Nov 14, 2011 4:08 am

Post by Misty » Sat May 12, 2012 1:21 am

Can't believe I missed the last few messages - for some reason it unsubscribed me to replies!

Anyway, really glad to read about your progress!

re: games, I've got a decent-size collection I'm glad to either loan, or dump on my own hardware as appropriate.

I think I posted my game collection earlier, but here's what I've got - not as big as BlueBMW's.

MegaLD:

Blue Chicago Blues (JP)
Hyperion (US)
Pyramid Patrol (US)
Space Berzerker (US)

LD-ROM2:
Manhattan Requiem (JP)
Quiz Econosaurus (JP)

MNc99
Interested
Posts: 20
Joined: Thu May 17, 2012 11:07 pm

Post by MNc99 » Thu May 17, 2012 11:39 pm

Hello;
I'm MNc99 and I first posted about Mega-LD emulation here.

I just wanted to post and say I'm so glad there are people with the knowledge finally giving this project the time it deserves.
You guys are great.

Cheers!
MNc99
Image

LAPProject
Interested
Posts: 37
Joined: Sat May 19, 2012 9:13 pm
Contact:

Post by LAPProject » Sat May 19, 2012 9:19 pm

Hello,

I am the head of the LaserActive Preservation Project (http://laseractive.wordpress.com). I was alerted to this page by a previous link in the thread.

Needless to say, I am thrilled by this turn of events. Dumping the data tracks of Mega LD discs is a huge step towards preserving this unique platform, and brings us one step closer to emulating the LaserActive.

During my time maintaining and updating the LAP Project site and Youtube channel, I have come to know several collectors who own many Mega LD and LD-ROM2 titles.

Due to the rarity of these discs, I propose that we all work together to get as many games dumped and recorded as possible. Whether or not that should be under a new forum or other kind of discussion community is anyone's guess.

I would like to hear your feedback on such an idea. Once we make a decision, I'll begin emailing as many LaserActive owners as I can informing them about this project. If we work together, we may indeed be able to dump the entire North American Mega LD library.

Nemesis
Very interested
Posts: 791
Joined: Wed Nov 07, 2007 1:09 am
Location: Sydney, Australia

Post by Nemesis » Tue May 22, 2012 1:03 am

I just wanted to post in this thread to say I'm still alive and I'm still going to carry this project forward. I haven't done more work on the video ripping process as of yet, but I'll be getting back to it in the next couple of weeks. I'm splitting my time between half a dozen different projects right now, along with work and family committments, so I'm just doing a bit at a time as I'm able to. Most of the work has already been done in terms of working out the ripping process, so it's really a matter of a bit more experimentation and testing to make sure everything is being done properly. I don't want to make a mistake which will result in having to rip all these disks again.

TascoDLX
Very interested
Posts: 262
Joined: Tue Feb 06, 2007 8:18 pm

Post by TascoDLX » Tue May 22, 2012 8:18 am

Excellent! In that case, I will add that I now have a fully working PAC -- Thanks to BlueBMW -- so I should be able to do minor testing should the need arise. Otherwise, I await the continued dumping efforts of Nemesis. ;)

MNc99
Interested
Posts: 20
Joined: Thu May 17, 2012 11:07 pm

Post by MNc99 » Tue May 22, 2012 11:44 pm

That's great news, Nemesis.

I know there's been a lot of talk around here about what formats we should dump to so I think I should re-hash my comments from the initial thread at Eidolon's inn for the sake of convenience.

Initially, the first dumps made should be as close to 1:1 as possible, and damn the disc space requirements. The video should be kept pristine and virgin so we have something to work from. From there we should make more readily-available compressed versions. This way we can keep both the archivists and the more practically-minded happy.

All we need is somewhere reliable to store the data...
Image

LocalH
Very interested
Posts: 152
Joined: Tue Dec 19, 2006 5:04 pm

Post by LocalH » Wed May 23, 2012 4:42 pm

There is no such thing as "pristine and virgin" composite video. I'll say this again (with one caveat at the bottom), lossless compression of analog video is a complete waste of space when you can use high-bitrate I-frame-only MPEG-2 (say, 25Mbps or higher) and retain all the quality. Seriously, that's what the professionals use all around the video industry, and they handle higher-than-broadcast quality video on a daily basis. Composite chroma resolution is nowhere near 4:4:4, and that's the only real reason to choose lossless over MPEG-2.

"More readily available versions" would just be lower bitrate MPEG-2 with predictive frames, or possible MPEG-4 ASP or AVC, and may perhaps have a bit of processing done to it to kill artifacts like crosstalk between the luma and chroma (which manifests as the commonly known "chroma crawl").

By all means, the initial capture should be lossless if one has the drive space, but only so that it can be encoded with a good software encoder like HcEnc or Quenc, or even Cinema Craft if one has money to spend on it. But by no means should the lossless need to be kept or distributed once the base MPEG-2 video is encoded at a sufficient bitrate (minimum 25Mbps with no P or B frames, as mentioned earlier, and if you want to play it safe then 50Mbps is absolutely enough to be virtually lossless, the error would be so slight that it's just not worth the extra long-term storage).

Notions of "1:1" ripping are only applicable to digital data. It is impossible with current technology to get an exact 100% representation of the analog data, because even at 13.5MHz sampling rate (which is industry standard), you're quantizing a continuous stream of varying voltages, which are not limited to pixels, into a grid of pixels. Automatically, just from the act of digitization, you already have miniscule loss from the original, even if you write the RGB24 values for each pixel directly to disk with no modification.

Huge
Very interested
Posts: 197
Joined: Sat Dec 13, 2008 11:50 pm

Post by Huge » Wed May 23, 2012 5:20 pm

LocalH wrote:It is impossible with current technology to get an exact 100% representation of the analog data
We are dealing with Ultra Hacker Maestro Extraordinaire Nemesis here, so I'd wary of using the word "impossible" in any context.

Chilly Willy
Very interested
Posts: 2984
Joined: Fri Aug 17, 2007 9:33 pm

Post by Chilly Willy » Wed May 23, 2012 5:23 pm

What you see for analog rips of LD movies are people rip the LD four or more times, then average all the frames together. Since it's analog instead of digital, that actually does give you a better image.

LocalH
Very interested
Posts: 152
Joined: Tue Dec 19, 2006 5:04 pm

Post by LocalH » Wed May 23, 2012 7:44 pm

I say impossible because one would basically have to build their own sampling hardware from scratch, using much higher than 13.5MHz sampling rate, then write your own software to deal with it, only to downsample the end result to 4:3 720x480 anyway (which I recommend over square pixel 640x480, as you can correct the aspect ratio on the fly and you retain that little bit of extra detail).

That's ok, because you don't need an exact representation of the analog waveform. 13.5MHz is good enough, as it can more than cover the detail found in virtually any composite signal, given all the inherent flaws in the format and the limitations in the hardware used to record the video. Any mass-market hardware (even professional-level) will only be sampling the signal at 13.5MHz, and you then have interoperability with a large amount of hardware and software.

Willy: Yes, and that's a very good idea that I would recommend Nemesis do. I think he'd already mentioned that he was considering doing so. It's still not 100% accurate to the original analog signal, since the very act of digitizing quantizes the waveform into discrete pixels that simply don't exist in the analog realm. Noise that is inherent to the signal is essentially non-deterministic, and that's why that works. It still won't remove chroma crawl, as chroma crawl is pretty much deterministic (no matter how many times one samples a signal, the crosstalk will be so close to the same that it would take hundreds of captures to even have a chance at getting rid of it). That's ok, as there are hardware and software comb filters that can do a really good job of cleaning up that type of noise. It might also be acceptable to run a very slight spatial-only denoiser on the merged rip just to clean it up that little bit more (although one would have to do some testing with varied parts of the video and make sure absolutely no detail is lost). I would suggest using no temporal denoising, however, as it blends between frames and can cause extra loss of detail and "smudginess" where an object will move but a part of the detail inside of it appears to smear because the temporal algorithm blended it together.

One thing I was thinking of - it was stated earlier in the thread that some discs contain two interleaved video streams. The only realistic way that I see this to be the case is if the two streams are stored as alternating fields. Either that, or the laser head skips every other frame, which I would think would cause a bit extra wear (and may even cause the unit itself to generate some noise as the laser continuously reseeks). If it is stored as alternating fields, then the best way to store it in an MPEG-2 stream would be to separate the fields into individual 720x240 frames, de-interleave them, and once again correct for aspect ratio on playback. This would have the effect of making the file higher quality at the same bitrate, since the image only has half the number of pixels as a normal video frame. The separation and deinterleaving is easily done with Avisynth, which is yet another reason that the "impossible" statement I made is relevant, as it would be much harder to work with a data stream that is incompatible with every single video application on the planet.

MNc99
Interested
Posts: 20
Joined: Thu May 17, 2012 11:07 pm

Post by MNc99 » Wed May 23, 2012 10:48 pm

I concede my point, since you're clearly far more informed on this matter than I am.

I know this project is in safe hands, at least :D
Image

Post Reply