My testing also wasn't quite ideal and I had a hard time finding the right RetroArch settings to cut down on latency. It's probably not that great in this regard, despite having a Game Mode. Indeed, it's an older rather cheap-ish Samsung Smart TV I couldn't find any latency numbers on. I saw your tests, and the one thing that's not accounted for is the TV's own input lag I believe? Do you have the specific model? Weird, although I guess there isn't all that much interest in the scene in finding out which of the "known to be not very accurate" emulators is the least inaccurate. Interesting, this seems to contradict everything else I read about QuickNES. Our tests are probably not too far off if you count those things! So that's how I got to around 2 - 2,5 frames. And then my TV is very low lag - only 14ms in gaming mode, while older Samsung TVs usually had way over 30ms. By using frame delay, which is not listed on your testings, I was able to shave another 10ms ( almost a frame). Nestopia is just too demanding for running Runahead, but the tradeoff is not too bad.ĮDIT: I saw your tests, and the one thing that's not accounted for is the TV's own input lag I believe? Do you have the specific model? Because I believe you could definitely achieve better results. And all that while being faster! That's probably because FCEU is very hacked together, while QuickNES is more optimized. 67.1% (QuickNES) vs 45.6% (FCEUMM) according to tests. It's surprisingly fairly solid, and more accurate than FCEUMM. I also disregarded QuickNES at first because of accuracy concerns - but first I found it it has vastly superior audio to FCEUMM. and compare to default settings you'll be absolutely blown away by how much precise the jumps can be, specially compared to default settings which would give you upwards of 4 - 5 frames of lag. I'm also using a CRT shader, which is ZFast, and everything runs fullspeed. I had some prejudice because I thought it wasn't very accurate, but it's actually more accurate than FCEUMM and only 15% or so less accurate than Nestopia). Retroarch KMD, QuickNES core (This core is fantastic. Considering my TV has 14ms of input lag, there's only 1 frame coming out of the mini, virtually identical to real hardware! The real thing would also give me 2 frames when plugged into my TV.īellow is my configuration if anyone is interested: a Real NES has 1 frame of input lag on a CRT, and the Mini NES through my UHDTV has 2 frames. With the bellow settings, I was able to achieve latency indistinguishable from real hardware. I own real hardware for almost everything but the NES, and thought the Mini would be a good addition to my collection, as I was confident I could get the input lag to be very low using Retroarch, but I never thought I'd get this to be as low as it did! I'm very picky about input lag and have been running some tests on the NES Classic.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |