Page 4 of 5

Re: MegaLD Dumping Project

Posted: Sat Feb 13, 2016 6:21 am
by LAPProject
Knock knock! Anybody home?

I'm writing/checking in on this initiative tonight because I come bearing some bad news - while trying to find a buyer for my LaserActive hardware, my Sega PAC has unfortunately ceased working. Since I had a full capacitor swap on it the day I got it, I dread to find out what the problem is. Because of my acute awareness of just how fragile the LaserActive hardware is, I just wanted to express my renewed interest in getting these discs dumped and backed up. From what I remember, my LaserActive library (3D Museum, Hyperion, Pyramid Patrol, Rocket Coaster, Triad Stone) is all fairly common, and so Nemesis already has all of those discs. However, if he sent them back, etc. I am more than willing to lend them out.

Re: MegaLD Dumping Project

Posted: Sat Feb 13, 2016 7:13 pm
by Sik
It could still be some dumb thing that broke up. I mean, it's not the LD drive which is prone to mechanical wear :P

Re: MegaLD Dumping Project

Posted: Mon Feb 22, 2016 4:17 pm
by LAPProject
I'm gonna have the PAC looked at.

Regardless, looks like Nemesis is MIA :( He's the last hope, unless he is willing to distribute his custom hardware and software to another pair of capable hands.

Re: MegaLD Dumping Project

Posted: Tue Feb 23, 2016 8:34 am
by Mask of Destiny
So I was thinking about this today and I wonder if a dumping system could be rigged up with one of Pioneer's "industrial" LaserDisc players that have an EFM output port like the LD-V4200, LD-V4300D or LD-V8000. These players have an RS-232 port that can be used to control the player and the EFM port gives access to the raw data of the digital track. There was an accessory called the DA-V1000 that allowed these to be used as an LD-ROM drive. That device seems rather difficult to find now, but the players that worked with it pop up for relatively low prices on ebay. Apparently, EFM stands for eight-to-fourteen modulation, which is the same encoding used for CD-ROM data. Raw EFM is a bit of a pain to work with as there's no separate clock line, but modern micro-controllers are fast enough you could probably use one for decoding if an appropriate decoder IC can't be found.

Re: MegaLD Dumping Project

Posted: Wed Feb 24, 2016 2:38 am
by Mask of Destiny
This project is quite promising too forum thread, github project. Essentially, what this person is doing is capturing the raw FM output produced from the pits and lands on the LD with a high frequency ADC and then doing all the decoding in software. Current output is arguably better than capturing off composite on the player's he/she has been experimenting with, but not as good as the very best player. In time, it should be possible to extract the digital audio/data track this way.

Re: MegaLD Dumping Project

Posted: Wed Feb 24, 2016 6:10 am
by LAPProject
That latter link is VERY promising, since that would allow us to at least preserve the contents of the discs, even if we don't have what it takes to decode them right now. The data section of the disc should be relatively easy to identify because it's all based on CD.

The issue with that would be getting that person the discs... Nemesis has the big collection at the moment.

Re: MegaLD Dumping Project

Posted: Wed Feb 24, 2016 9:10 pm
by Mask of Destiny
LAPProject wrote:That latter link is VERY promising, since that would allow us to at least preserve the contents of the discs, even if we don't have what it takes to decode them right now. The data section of the disc should be relatively easy to identify because it's all based on CD.
Well it's EFM, which is the encoding CD's use, but I don't think it exists physically on the disc in the same format as a CD. From my understanding, a standard NTSC laserdisc effectively has 4 signals with different carriers. There's the analog composite video, analog audio left, analog audio right and a digital signal which normally contains PCM audio. My understanding is that on an LD-ROM disc, the data track replaces the PCM audio rather than being in a completely separate part of the disc. So it won't necessarily be all that distinct. Given a high enough quality capture of the RF signal it should be possible to decode the digital track, but it's not guaranteed that the existing captures by happycube are good enough for that. I would guess that they are since the composite video signal is the highest bandwidth portion of the combined signal and presumably most sensitive to noise, but we can't say for sure until he/she has been able to decode the digital track.

Regardless, I don't think it's necessarily the best option in the short term. Getting the digital data out of the EFM port of one of the players I listed should be a lot easier and we can capture audio and video with a traditional capture card. I think it's interesting though because it potentially gets us as close as possible to a perfect copy of the physical disc.
LAPProject wrote:The issue with that would be getting that person the discs... Nemesis has the big collection at the moment.
I'm not proposing sending discs to happycube. His capture technique does not require anything particularly exotic, just a laserdisc player with an RF test point (it appears most Pioneer players do) and something to capture it with (his setup uses an older capture card that allows raw ADC access). This (along with the EFM port method) theoretically opens up dumping to people who can't justify the expense of acquiring a working LaserActive and appropriate PAC.

I've briefly talked with the founder of the MADE (a local video game museum) and they would be willing to host a LaserActive preservation effort. They're a 501(c)(3) so for those in the US, any donations would be tax deductible. Also, loaning discs to an organization seems a bit less dodgy than an individual. Assuming I can convince my wife, I'd be willing to acquire an appropriate player and one of the less expensive LaserActive titles. Once I got things working I would donate those to the MADE and preservation work could begin.

I imagine Nemesis is just busy with real life at the moment, and will resume work/reappear once he has time. Given that, I think the most logical approach would be for this effort (assuming it gets off the ground) to focus on titles he doesn't currently have and/or is unwilling to dump.

Re: MegaLD Dumping Project

Posted: Tue Mar 08, 2016 8:12 pm
by Mask of Destiny
Minor update: I'm less convinced the industrial, EFM port equipped players, are worth it. The schematics from the 4300D make it appear to just have the RF output passed through a low pass filter and an amplifier. It does not appear to be properly digital at the point the signal leaves the device. However, it seems likely that LD-ROM data could be extracted via S/PDIF on a player with such a port.

My current thinking is to pursue raw RF capture as the primary approach with the S/PDIF option as a backup unless I can scrounge together an S/PDIF recording solution with hardware I have on hand. I'm planning to pick up an Embedded Artists Labtool for raw RF capture. I don't think the default firmware can do true continuous capture, but the ADC in the LPC4370 it's based on is capable of up to 80 megasamples/second and the LPC4370 is a USB 2.0 High Speed device so the hardware itself is definitely up to the task. The firmware is open source making modification for true continuous capture more feasible.

Re: MegaLD Dumping Project

Posted: Thu Jun 16, 2016 1:10 pm
by Nemesis
Hey MoD, i'm very interested in the angle you're looking into, but I've got no spare time to pursue it myself right now. I'm going to work on completing the digital content rip and analog video/audio recording via the LaserActive itself and get that up online, but if you can come up with a complete process for ripping the raw RF, I'd like to work together to make that happen for the disks I've been loaned, one way or the other. At this stage I've only really got one last technical hurdle to overcome before I'm ready to make "proper" rips of all the disks I have, so I expect that to start soon, if all goes to plan. It'd be great if after doing that ripping process, we could move onto the raw RF capture process, so it all gets done back to back.

All that said, I've wanted a high resolution, continuous analog capture solution for a long time. It's actually pretty incredible I think that in the world today, such a device hasn't been created long ago, because it could directly benefit so many different fields. I've got a few scopes with pretty nice resolution and capture rate, but all of them have what really amount to tiny buffers, only useful for setting triggers and inspecting around the trigger point, not for doing continuous capture.

That said, there is one interesting product that's only recently hit the market, that comes very close to meeting requirements here. Check these guys out:
http://www.saleae.com
You may be familiar with them, they released a logic analyzer which did continuous capture a few years back. That was little more than a Cyprus CY7C68013A wired straight to some probes, but it did the job, and it got some significant attention. It was also seriously overpriced (around $150 IIRC). Like a lot of other people, I got myself a $15 clone on ebay, and borrowed their software to do the capture, and it worked a treat. That's digital signals only mind you, but they went on from that to design and build more powerful versions of the same thing, continuous capture data loggers, with analog support this time. Check out their range, they only came on the market last year. The Logic Pro 8 claims to be able to do a 12-bit capture at 5MHz, with -10V to +10V. That's not quite in the 6MHz range you mentioned before as the target, but it's quite close. The next generation of that device would probably far exceed the requirements. The price is scary though, just a little under $500 USD.

Again though, I find it pretty incredible that this problem still seems so be so far out of reach. Why I can't go on ebay right now and get a high-end ADC sitting on a minimal board that plugs straight into a PCIe 16x slot, with a driver that dumps out a raw data file to my SSD at the fastest possible bit depth and data rate is beyond me. Hell, don't even bother writing a real program to work with it, don't even give me a user interface, just dump out the raw samples to a wav file, that's all you need. Maybe I should go build a card like that and sell it, I'd probably make a killing.

Re: MegaLD Dumping Project

Posted: Thu Jun 16, 2016 1:24 pm
by Nemesis
Oh, I should really have posted my apology first. Sorry, as you might have noticed I've been pretty well absent from SpritesMind (and pretty much everywhere else) for the past year or so. The last year has been difficult and crazy for me in real life. Things are just settling down now, and I'm finding I actually have the occasional bit of free time here and there. This project is number one priority for me, and will be until it's completed, so I'm spending all the free time I now have actively working on it.

My goal is to have all the disks I've been sent ripped and uploaded before the year is out. I'm hoping to have the first good, complete rip up well before then, just one last technical hurdle to overcome to get a complete process in place. Hoping to have that knocked over in the next couple of weeks, just waiting for some new gear to arrive from China. My biggest problem right now is that I can get the complete digital data accurately ripped off the system, but only in one uninterrupted pass, meaning I need to be able to stream it off at the disk playback rate. Parallel and USB based solutions for the controller ports aren't up to the task, so I'm rigging up a microcontroller to expose a magic address I can just continuously write 16-bit data blocks to in a single move instruction on the 68000, which I can then stream to my computer via USB. I need a few nothing parts that I didn't have on hand, so they're in the mail. When they arrive, I should be able to get something working with what I have on hand. I already have a tested working analog capture process, and I've deciphered all the critical hardware registers to make the player do whatever I want, so I can capture in the right video and audio modes. Everything's ready to go, and the digital capture process actually works, for all raw 2352 sector bytes plus 96 bytes of subcode data, I just need to be able to push the bytes off the system fast enough at this point.

Re: MegaLD Dumping Project

Posted: Thu Jun 16, 2016 7:41 pm
by Mask of Destiny
Nemesis wrote:Hey MoD, i'm very interested in the angle you're looking into, but I've got no spare time to pursue it myself right now. I'm going to work on completing the digital content rip and analog video/audio recording via the LaserActive itself and get that up online, but if you can come up with a complete process for ripping the raw RF, I'd like to work together to make that happen for the disks I've been loaned, one way or the other. At this stage I've only really got one last technical hurdle to overcome before I'm ready to make "proper" rips of all the disks I have, so I expect that to start soon, if all goes to plan. It'd be great if after doing that ripping process, we could move onto the raw RF capture process, so it all gets done back to back.
Cool. I got distracted by real life and other projects, so I haven't moved on this. I'll try to pick it back up this weekend.

All that said, I've wanted a high resolution, continuous analog capture solution for a long time. It's actually pretty incredible I think that in the world today, such a device hasn't been created long ago, because it could directly benefit so many different fields. I've got a few scopes with pretty nice resolution and capture rate, but all of them have what really amount to tiny buffers, only useful for setting triggers and inspecting around the trigger point, not for doing continuous capture.
Nemesis wrote:The Logic Pro 8 claims to be able to do a 12-bit capture at 5MHz, with -10V to +10V. That's not quite in the 6MHz range you mentioned before as the target, but it's quite close. The next generation of that device would probably far exceed the requirements. The price is scary though, just a little under $500 USD.
6MHz bandwidth is sufficient for raw capture of a normal composite signal, but not one FM modulated at 7.1Mhz. I believe roughly 10MHz of analog bandwidth is required. This would seam to disqualify the Embedded Artist's LabTool as well, but according to the manual it's capable of 12MHz analog bandwidth at certain voltage ranges. While the firmware it ships with can't do continuous capture, the exact same micocontroller has been used by the software defined radio community to do continuous capture. Whether it can actually sustain a sufficient sample rate for LD RF capture is a little less certain, but seems doable. happycube's RF captures were 8-bit at 28 MS/s and supposedly the real world transfer limit for USB 2.0 HS is about 40 MB/s so it doesn't seem completely out of the question. It's ability to do simultaneous analog and digital captures is also quite attractive for general usage outside of this project.
Nemesis wrote:Again though, I find it pretty incredible that this problem still seems so be so far out of reach. Why I can't go on ebay right now and get a high-end ADC sitting on a minimal board that plugs straight into a PCIe 16x slot, with a driver that dumps out a raw data file to my SSD at the fastest possible bit depth and data rate is beyond me. Hell, don't even bother writing a real program to work with it, don't even give me a user interface, just dump out the raw samples to a wav file, that's all you need. Maybe I should go build a card like that and sell it, I'd probably make a killing.
It's a fairly niche market. You've got diagnostic/reverse engineering users, but the market for those is mostly focused on the high end and for whatever reason PC based capture doesn't seem to be very popular outside of relatively low-end products. The SDR community has some growth potential though. I think you'd probably get more traction with a USB 3 product rather than a PCIe one these days.

Re: MegaLD Dumping Project

Posted: Sun Aug 28, 2016 4:41 am
by Mask of Destiny
Minor update. Finally, managed to acquire an appropriate player for this: a CLD-V2800. This isn't a great player generally, but it has RS-232 support which will be handy for controlling playback while "ripping" a disc. It's also conveniently the same model as one of the players used by happycube so I know it has a usable test point already. There's no digital audio output, but it does have a digital audio decoder so I can theoretically tap the signal internally if I need to fall back on something else for capturing the digital data track.

Re: MegaLD Dumping Project

Posted: Wed Nov 23, 2016 12:45 pm
by bakudandan
-Bi Ryojon Collection (Pretty Illusion - Minayo Watanabe) (Japan) [NEC]
-Bi Ryojon Collection II (Pretty Illusion - Yuko Sakaki) (Japan) [NEC]

These games are not pornographic, they have a model in suggestive poses, but nothing worse than you would find in the Victoria's Secret catalog.

Re: MegaLD Dumping Project

Posted: Mon Jan 02, 2017 5:34 pm
by MNc99
Any updates to spare, gents?
The suspense is killing me.

Re: MegaLD Dumping Project

Posted: Tue Jan 03, 2017 2:16 am
by Mask of Destiny
So I acquired my Labtool a while back, fiddled around with some code for composite decode (simpler place to start than a laserdisc RF signal) and started hacking on the firmware a bit. First, I dramatically sped up the process of doing a normal capture (stock code was really inefficient). Unfortunately, even with this speed up there's too much of a delay to use the existing process for a continuous capture application (not an unexpected outcome). I then did some work on a streaming solution, including writing some optimized ARM assembly for packing a stream of 12-bit samples from the DAC to RAM for transfer. None of that streaming code is tested though and I don't remember the exact state I left it in.

I'm currently a bit busy with a different project, but hopefully I'll be able to get back to this in a few weeks. Sorry for dragging things out.