First off, I want to say that donating the ROMs/PROMs to MAME was done with the complete approval and knowledge of one of the lead Daphne developers (Mark B.). I would have liked to see Daphne get it done first, but Mark felt he would be too busy with life to get to it quickly (very understandable, as this is a hobby to most of us). I still would like to see it in Daphne (I did make sure that they (Daphne) got the ROM dumps first). Also, FYI, the ROM dump I may have sent people in the past was proved to be lacking in a few PROMs. I was able to dump them again and get a good complete dump.

That said, MAME is making amazing progress. Check out the MAME work-in-progress. It should be released in a few weeks.

http://philwip.mameworld.info/

I've been in contact with the MAME developer all along the way, and he is in full support of any help he can provide to the Daphne team. I have passed his email along to Warren and Mark.

Joe (joemagiera at ameritech dot net)
joemagiera@ameritech.net

This is very exciting news.

Yes! Cube Quest is a game I never thought I'd see emulated. Very cool!

Thanks joemagiera. I don't think I've ever seen Cube Quest before!
Aaron has written some cool stuff "About Laserdiscs" on his blog recently too: http://aarongiles.com/

And now as of MAME 0.127, it's fully supported:

http://mamedev.org/releases/whatsnew_0127.txt

Great work guys, I can't honestly say I've ever seen this game before but I can assure you that I'll be giving it a damn good flogging over the next couple of weeks.

The only thing that concerns me about mame's laserdisc support is the rumored huge file sizes. I know the first rule of laserdisc emulation is "thou shalt not complain about large files" but still.... I would be all about supporting a higher-quality storage format if the arcade games didn't output to crappy ntsc resolutions, making anything greater than dvd quality a moot point.

From what I heard the average laserdisc game will be 5-10 gigs, which doesn't sound like a lot until to consider that a full laserdisc set could easily fill a 100 gig drive! That's some serious storage investment!

Regardless, getting a new game emulated is never a bad thing, so I'll refrain from complaining too much right now.

Anyone know how large the image weighs in at?

I think it's going to be more like 10-15 gigs per laserdisc, comparable to the size you'd get by compressing 30 minutes of NTSC video using huffyuv and no audio compression.

One reason they went with this format is to make it difficult for people to "collect them all." Also, I think this format is perfect for archival use and preservation. However, I do agree with you that it is overkill for playing the games.

By "image" I assume you mean both the roms AND laserdisc video together?

The roms appear to be only 63.2 kb.
The chd (which I haven't found) is rumored to be 12 GB! OUCH! That's going to require a serious commitment for the collector. Of course, in Korea where you can download at 100 MB/s, it's doable, but here in the states, where we invented the internet, but are stuck at around 3 MB/s, it's something else entirely.

10-15 isn't horrible.... it's still too big, but it isn't horrible.

I personally think laserdiscs fall into a totally different category than the average emulated arcade game and they are an exception to the "romz kiddiez" deal you usually see with people hoarding mame roms. For one thing, unlike arcade pcbs, in this day and age it's actually more practical to run an emulated version on the pc in a cabient rather than the actual hardware. The laserdisc players themselves are a pain in the butt and the disc are huge, making it impractical to switch them out. Not to mention the fact that stuff like the laser ace (correct name?) is no longer available and you might need several players to run a laserdisc collection. You've also got to consider that most of the laserdisc games (obviously not dl,sa and cliff hanger) don't really belong to anyone anymore. When the bottom fell out of the ld game market, most of the companies dropped their stuff faster than you can blink. Most of them don't even have info on the games, much less original prototypes, ect.

I'm certainly not endorsing piracy, but in this case, it might make more sense to make the images more managable to get them out there to the masses for preservation. Afterall, the more people have a copy, the less likely the footage is to be lost years from now.

Well after waiting two days for the chd to download I tried it. It's cubequest alright, and it's emulated alright... too bad it's completely unplayable.

I have a athlon xp 3800+ with a gig of ram, which I consider to be a fairly decent pc. Mame can barely run cubequest at 50% on my system. Looking at the source code, at first glance anyway, aaron's video rendering system seems to be to blame, which is what I was afraid of. Mind you the cubequest game itself uses a fairly decent chunk of processing power, but still not nearly enough to account for any major speed decrease.

Mame might be able to run the simplier games at full speed or at least a playable speed on the average system, but for most of us, we'll need to stick to daphne for the foreseeable future.

CubeQuest uses 3 CPU at 5MHz, with a complex logic to be reevaluated each instruction, for the other lasergame the situation is different, and only a good HD is required.

Now this is a game I have wanted to see since I was a kid, but 10gig ???

I think I'll wait till they encode the video correctly

Which is still nothing major that any decent pc should be able to handle or at least be close to handling. Don't believe me? Try running cubequest without the chd... you should get at least a 25% performance increase. What slows it to a crawl is the way the laserdisc artwork was implemented. Mind you it's a very accurate way to read the ld info (pretty sure mame actually reads the blank lines in the video to get the frame info) but it's very slow. What should have been done, imho was to use a utility to parse the blanker info ahead of time into a simple text file, similar to daphne's framefiles... that way processing power isn't wasted converting video into data. But mame is all about accurate emulation, even if the accuracy gets in the way of practicality. Of course the fact that the video frames themselves are virtually uncompressed couldn't help either. That's a lot of data to be swapping out on-the-fly.

MAME reads the VBI to emulate the behavior of the LDs player, trying to be the most faithful possible with the original arcade, the system used by the daphne it is an adaptation that can create many problems of fidelity.
An example is DL2 that for succeed to stopping the scenes to the correct frame has been split the video, thing that won't happen in MAME.

I think that the VBI data is stored as "metadata" inside the CHD file (ie I think it actually is pre-processed somewhat), however I was thinking on my commute to work this morning that almost all of the VBI is predictable, so it could be compressed down to almost nothing. For example, instead of storing VBI for every single field, you could just store it for the first field, and then store the algorithm that says "this VBI repeats itself in a predictable way for the rest of the disc." It's kind of moot though since the VBI data itself doesn't take up much space, but it is an optimization I will definitely keep in mind if I tackle the VBI issue at some point (more on this in one of my next responses).

The framefile system definitely has its weaknesses to be sure. The biggest problem with framefiles is that some people seem to have such a hard time understanding how they work. However, I feel that the framefile system has worked quite well when the framefiles are created correctly. In fact, since the framefile system has worked for all the games that Daphne emulates, this should prove to you that the VBI data is _not_ necessary for playing the game correctly. It is more "nice to have" for preservation, and I do applaud Aaron Giles for going to the trouble of storing the VBI data for laserdiscs.

I may devise a replacement for the framefile system in the future in order to facilitate some of the tricks used by the more advanced laserdisc games.

As for Dragon's Lair 2, I think Daphne emulates it fantastically, and I also think the video quality is superb. I don't know how you can complain about the way we split up the files when the game plays correctly.

Remember, Daphne is focused on PRESENTATION. MAME is focused on DOCUMENTATION.
Both are worthy goals, so let's not get into a war about which one is better. Let's just agree that they are different.

Bravo! Excellent points!

Like many others, I appreciate both DAPHNE and MAME for what they do. I appreciate the attitude of preservation and I appreciate the attitude of documentation.

Matt_O: BTW: THANK you for sticking with the preservation perspective.

Brad O.

I have to agree w/Matt. To be honest I like accuracy to the arcade as much as possible when I play the game. And when I say that I mean I want it to look and have the same feel as when I played them on the arcade cabinets. I don't really care what goes on behind the scenes to accomplish that. Well it does interest me and I find it fascinating, but I'm not going to care how it was achieved in respects to when I'm actually playing the game. I give major kuddos to MAME and to the Daphne developers. I think the Daphne devs came up with a very clever way to make it possible for us to play our favorite laserdisc games. I find it annoying when a great game comes out for MAME but it just runs way to slow for anyone to play it because computers can't run fast enough (MK4 for example). I'd rather just have a version that is playable and still keep the arcade accuracy as far as playability, graphics and sound. But don't get me wrong, I understand where the MAME devs are coming from and I respect them for all they've accomplished.

I only have a 250gig drive on my main gaming rig (well, I started out with that, but also have a 500gig for back-up of it and my everyday computer - instead of going external, as this is really easier for me), and I have no problem with doing the whole set of games.

Bonus for me is, I don't want 'em all. I don't have all those available in Daphne. (I remember when those were the largest file sets I had on my computer...actually, besides Windows install, they've been the largest until CQ.) It might follow I'm not a pokerom, either.

As well, many people these days have Core duos, and in two years neither storage nor performace for things like this will even matter for almost anyone (well, anyone who's not in the poor house and can only afford a PIII give-away, which should be minimal, if not extinct. Sheesh).

Lastly, bandwidth is ever increasing.

All in all, the only things I can see to say are:

Gosh, it's cool Mame has done this.

- and -

Gee, it would be nice if the video were made available to the Daphne team in the meantime.

I believe the video came from the Daphne team in the first place. They certainly had it a couple of years ago anyway.

The video and all upcoming laserdisc captures in MAME have nothing to do with existing captures. MAME is re-capturing every laserdisc to get the best quality possible and, more importantly, include the data outside of the visible frame that is in the VBlank area of the laserdisc and was left out in all Daphne caps. This data is important for accurate emulation since it contains per frame information that the laserdisc player uses.

I disagree; if your statement were true, then Daphne would not be able to accurately emulate any of its games, and I think many people who own original machines agree that it does emulate these games accurately. In fact, the VBI data is not important at all for accurate emulation of most laserdisc games. It _is_ important for documenting how frames were encoded on laserdiscs.

Matt, nobody discounts the enourmous contribution Daphne has given to the community over the last years. You are absolutely right about Daphne being a great project to give us sometimes better-than-the-arcade emulation and simulation of games that were once thought to be gone forever. And, quite frankly, a project that probably gave Mame at least a little push to finally cave in and add laserdisc emulation.

However.... as an early 80s arcade fan I always wanted to see Cube Quest, Star Rider, Firefox, and the other more elaborate and intricate laserdisc games emulated to finally experience them - and the pedantic Mame approach certainly makes these more advanced usages of laserdiscs finally possible and gives us these games. Dragon's Lair can nowadays be played in glorious Hi-Def on any PS3, but Cube Quest, and soon Star Rider and Firefox, those are very different and more complex beasts.
Secondly, I do agree with the Mame team in their endless quest for absolute perfection in the emulation, and the Mame approach to laserdiscs at least does get closer to what the idea behind a CAV laserdisc was in the first place - single frame after single frame of video with additional data in the VBlank. Daphne is afterall very much built around Mpeg and with that alone is in the video portion of the emulation the antithesis to laserdisc. You can certainly get Mpeg to behave the way these more complex games need to, but Daphne is not there yet.
Finally, if at one point somebody makes the final, crazy jump and devises a way to capture the actual raw analog data stream from the laserdisc and turns it into a digital stream, most likely this will happen within the Mame project, or at least Mame will provide the framework to easily replace the current data with a closer-to-the-source approximation. That is the final goal every laserdisc fan should strive for.

Having said that - if Mame does lead to a healthy competition and gets Firefox, Cube Quest, and Star Rider into Daphne, all the better!

I agree with you here. Having the laserdisc separated into fields, with each field having VBI data is probably necessary (or at least makes the most practical sense) for emulating Firefox, Cube Quest, Road Runner, and Freedom Fighter (I don't know about Star Rider).

I just don't like people saying "VBI data is necessary for accurate emulation" without qualifying that statement with "for certain advanced laserdisc games" because I have been in the ROM code for most of the games Daphne emulates, and most of the ROMs do not use the VBI data at all (Cliff Hanger and MACH 3 are two exceptions). I've been in the ROM code for Firefox and it _does_ use the VBI data.



I think it's more like healthy cooperation. The reason Daphne doesn't emulate the more advanced games is not so much based on its usage of mpeg, but more on the lack of talented developers who can figure out how these advanced games work. I don't think I would've ever figured out how Cube Quest's hardware worked, which is why I am so grateful for people like Phil Bennett.

If the Daphne developers are interested, I do have the original source code for Firefox. This might help it get emulated one day.

Are you talking about the code for the arcade ROMs? That could be very helpful.

Yes, the original source code from Atari for Firefox that the arcade ROMS were created from.

Wow, where'd you get that? I'd love a copy! You can email me at http://www.ownby.org/emailme (and I will try the address you used to register on these forums too)

I can't believe that I have just seen Cube Quest

Thank you, thank you, thank you