> That single frame does a great job of shutting any "why not just use a capture card" > arguments right up in terms of resolution, but to really put it into perspective how > far the ld-decode project has come in the past year, here's a video of the Domesday > National Disc intro decoded by ld-decode as of about a year ago: > https://drive.google.com/file/d/1DQZ0bwqcRSaAQRfIuqhDg6xc7C1ttka_/view > > Meanwhile, here's that same intro decoded with the latest WIP version of ld-decode > off of Github: https://drive.google.com/open?id=1YAPvAjiguQ7zUZ2vQZL7Ix4HCLsmPBNE > > Of immediate note are: > - Drop-out detection and correction is massively improved (or rather, it actually > exists in the first place) > - Audio decoding is now implemented > - Color gamut is wildly improved > - Image stability is wildly improved > > There are, nonetheless, still areas where it can be improved. Small dropouts of only > a few pixels aren't yet detected, but there are some theories that it should be > possible to detect them based on a rapid loss of frequency lock, rather than the > current method which is to detect a sudden spike above or below the expected signal > levels. > > Anyway, suffice it to say that things are rolling right along. The main issue is > going to be that the mainstay of the decoding process, ld-decode, is written in > Python, and there's little interest on the Domesday86 team's part to port it to a > compiled language. It's likely that someone will eventually have to port it between > languages, and potentially make it GPU-accelerated, and that person is most likely > going to have to be me.
First of all, its really, really nice to see how far and well this capture method has come. The difference is clearly visible. I already told you that yesterday and i meant it that way.
What makes me sad is that you still "somehow" need it, to point at the "why not just use a capture card" argument, as i was this guy.
But you dont tell, at what state the whole LD-development was, at that time. Or that the previous (MAME) capturing methods and especially the tools for it, where not practical / handy for anyone who really tried to preserve anything LD-related (and that to a point, where it was nearly impossible without tremendous afford). You dont mention, that i just pointed at this and had a solution, what needs to be done to make this (old) capturing method properly and for everyone available. You dont mention that the argument come from a born (out) of necessity-thing, because of the still ongoing process of LD-rot. You dont mention, that even people from the MAME-dev team totally gave up on any chance to preserve something LD-related. You dont mention what came out of this discussion, which is the fact, that there is such a thing like a Domesday86 project . A project, which maybe would not be known of any of us here. You dont mention, that because of that fact, i said (and also did) i will either spend the money (that was planned as a budget for the capturing card and other affords needed) for upcoming LD-development or spend it for the DU to buy other cool things needed for MAME.
But you can mention the "shutting any "why not just use a capture card" arguments right up" or "there's little interest on the Domesday86 team's part to port it to a compiled language."
IMHO thats just not needed and doesnt do anything good, for any further and advanced process. At least nobody can say i am a bad loser, i still wish you a merry xmas and a happy new year.