> Again big thanks, casm.
No worries; just glad it's useful. And part of the reason for doing this is so that I have a reference as well - it's been a decade since I had to deal with a lot of this, so it's helping me to bolster my memory.
> I hope i can read everything in the evening after job-time. I
> agree with anikom15, MJPEG is not good enough, as it is not a lossless codec.
True, and I do agree on that point (which is why I mentioned using MJPEG or similar). To my mind, the basic idea behind MJPEG is essentially what we're looking for - each individual frame stored as a frame, not a stream. What I'm not sure of is if there's a lossless equivalent that can store the data in the same basic way.
Per-frame compression may be acceptable depending on how much of a performance hit that would cause in playback (and I admit that that might be unrealistic in a gaming environment), but lossless codecs are the key. See also:
> After reading your first post, i also understand now, that it would be silly to
> convert the games to codecs that would generate smaller sizes, as they cant be
> accurate enough, to jump to exact frame numbers, but this shouldnt be a problem
> nowadays with a average PC setup and a big enough HD.
I understand where you're going with this. My gut feeling is that whatever format is used should be capable of randomly (and directly) accessing frames, the same as an LD player would have, without landing on a frame transition. In theory, this should never happen in a properly-designed codec since one frame doesn't necessarily stream into another, interlacing possibilities aside.
However, the more I think about it, the more I think that what I'm describing is a codec specifically tailored to playback of media sourced from laserdiscs. It might also be able to be used for teleciné-style film transfers, but to the best of my knowledge nothing like this currently exists in the format being described.
> Back to anikom15 suggestions, i maybe can help with point 4, which sounds the easiest
> to do. Ok, it may be not the "best and most accurate" method, but the result would be
> the same IMHO and only different seen by someone who goes into very technical details
> (sorry if i am totally wrong with that). MAME already has some compromise for systems
> like atari2600 i.e., why not for laserdiscs? Can only repeat anikom15 here:
> "If good enough is not good enough, and you want an absolute perfect preservation of
> an analog medium, then of course there is no practical way of digitizing the signal.
> At that point the problem will be unsolvable."
True, but that can be said of any analogue to digital conversion. There is no good way in a digital domain to store the infinite possibilities within a finite analogue range. About the best you can do is decide on what constitutes acceptable loss and move on.
> From what i try to understand, the laserdisc-players seems to be the main problem,
Not really. We know how they work, and there's not very much stopping them from being emulated (other than someone having the motivation to do so). The thing is, though, because we understand how the video is stored and structured on the discs we can pretty much bypass the players altogether for digital playback. This is a good thing, because given how laserdisc players are dying off and there are very, very few people able to repair them, we will hit a point fairly soon where there just aren't any useful ones left.
> so i need to ask, what kind of laserdisc-player is needed for those games?
It depends on the game. Many, many models were used, and it's largely game-dependent. See http://www.d-l-p.com/games/ for examples; you'll need to pull up each individual game and check its Tech Center entry for the info.
> i assume that
> a consumer model (those that played just movies at that time) will not work. what
> kind of models would work? some exact model-numbers would help me.
Some consumer models *may* work. In general, though, industrial players are preferred as they generally had better playback, output, and build quality.