> @casm: many thanks for the links, that was very helpful.
> I still have some questions regarding the VBI data. I am not a expert, but i assume
> it is something like the copyprotection from a VHS, just complete different data and
> no copyprotection.
To be perfectly honest, I don't know enough about how Macrovision and other related VHS copy protection schemes work in order to be able to draw a comparison, and really don't want to give you an answer based on a quick skim of a couple of articles. The same applies to any similar schemes used for laserdiscs.
As relates to the VBI on laserdiscs, however: the VBI data needs to be preserved for a number of reasons, but one of the main ones relates to how laserdiscs store and access data. More on that below:
> I dont know, but wouldnt it be enough, to capture the stuff with
> underscan enabled from a broadcast monitor, so that you have the "whole" frame?
> Capturing with a good card like a Blackmagic in lossless format?
OK... Think about how a laserdisc stores video for a moment. Each frame is stored individually, like on film - which is totally unlike how, say, VHS stores its video, which is essentially a stream.
So, going back to what Aaron said here for a moment (and I realise you may already understand this, but I'm going to work through it for the benefit of anyone reading this post):
"At first, the VBI part of the scan just contained a couple of special signals for synchronization, but over time, people started adding data to them."
We now know that there's more than just basic start & end of frame info in there. Combining that with this statement a bit further down the page:
"The thing I discovered about laserdiscs, though, is that they encoded several very important bits of metadata on some of the VBI lines. Specifically, line 12 has what is known as a "white flag", line 16 contains some control bits, and lines 17-18 contain information about the current frame number and chapter. Even more importantly, all of this metadata is crucial to the way laserdiscs are controlled and operated in video games."
Putting it all together: the VBI contains, in essence, data attached to each individual frame used for control and playback of the frame in question. The original players referenced this data in order to be able to do things like skip to or pause on a specific frame.
Now, coming back to your question re: capturing the VBI data using something like a Black Magic card: sure, it's possible. But the issue then becomes what to do with the data stored in the VBI when the LD frames are converted to a digitally-stored streaming file. You also need to be able to play back those video files without the VBI data visible, which brings its own set of headaches.
DAPHNE worked around this issue through its use of framefiles, and possibly the best overview of how those work is here. However, framefiles allow DAPHNE to ignore VBI data altogether (see Matt Ownby's posts in this thread for more info on that). This is fine for playing the games, but from the standpoint of historical preservation it's less than desirable.
(Side note for anyone who has read this far: in one of the posts in the above-referenced thread, Matt Ownby said the following: "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." I happen to agree with him on both counts, and would prefer if that didn't happen here.)
This leaves us with other options to consider that might satisfy the requirements of storing the video digitally from both a presentation and preservation standpoint.
A format such as Motion JPEG may be an option, since it does store frames individually and plays them back one after the other rather than as a stream of data. VBI data could be preserved as part of each frame, and, potentially, referenced by an emulated (or simulated) laserdisc player.
For streaming formats, it may also be possible to read two video streams in sync: one containing the VBI data, and one containing the video. Again, the emulated or simulated player could use the VBI data as a marker by which it would access video in the streaming format - but that would come back to the issues of not being able to land on an exact frame division within the stream, and not storing the data in a format as close to original as possible.
All of the above gets trickier when you get into emulation of corner cases such as Williams' Star Rider - the VBI was used extensively in that game for in-game data (look here for example video from Matt Ownby that shows this), and it also had hardware that did panning of the video frame as it was displayed. But things like that are the exception rather than the rule.
Personally, I feel that the best approach is a Motion JPEG or similar codec capable of storing the entire video frame with VBI, followed by playback through an emulated or simulated LD player. But I'm a preservationist at heart, and can see where other considerations may not make that a universal fix.
In any case, we need to be preserving these discs at the highest possible quality, now, and while we still can. I do realise that efforts in this direction have not been lacking, but if we are going to settle on a 'universal' method of doing this now would really be the time since a lot of re-ripping will likely be involved and the time to be able to do that successfully is only decreasing.