You’re basically just loading part of your resources, cores in this case, and the load you see in VST Performance is that of your busiest thread (while a few of them just sit watching a couple do all the work ). The current engine uses physical cores for time-critical processing and virtual cores for ASIO Guard / non-time-critical processing.Īs sometimes an image says more than a thousand words, in the attachments you can see: 20 seconds (2 leftmost squares) with no tracks selected, 20 seconds with all tracks recording (no monitoring enabled) and 20 seconds with all of the tracks having monitoring enabled. Tracks being recorded can use ASIO Guard, if monitoring is disabled. Kind of an edge-case and indeed a simplified example, but helps understand the mechanism.ĪSIO Guard is designed to use virtual threads: if you disable it, you effectively use the physical cores only. This in turn means that if you have a small project with just a few, say ‘X’, VST Instrument tracks with FX processing on them, you’ll be using ‘X’ amount of cores (‘X’ being 1/4 of the available cores) - if you now duplicate those tracks, the additional instances will be allocated to different threads, previously unused, the result being little to no increase in the VST Performance. VST Performance measures the CPU load within the constraints of the ASIO driver, a CPU meter does not.Īlso, if you have a single core which is working hard, the VST Performance meter will show its value, not an average of all the cores (due to the fact that if one maxes out a single core, the application will stutter, regardless of how low the load is on the other cores). The posters above are correct in saying the VST Performance is not the same as the system’s CPU usage. Something is clearly wrong because that discrepancy between the ASIO meter and the CPU meter is just too big. Is that it? And if that is, that doesn’t seem like it would affect my overall ASIO performance, but challenge it more.Īnyway, off to do some reading. It seems to only impact performance on the track I’m recording on, right? Like, I can set my project buffer to be something quite high, but with ASIO Guard enabled, I can still record a MIDI part and not experience much latency. No matter how much I’ve studied exactly what that is, I never fully get it.
CUBASE 8 VS CUBASE 9 DRIVER
So ASIO drivers… On the Mac, I think most of this stuff is plug and play, so it’s worth double-checking the manufacturer’s website to see if there’s other driver I can install.Īlso, ASIO guard. It would more likely be the hard drive than the RAM and it’s not the hard drive.
So I can’t even fathom what could possibly be bottlenecking and causing an 80% ASIO performance spike… Except something to do with ASIO.Ħ4 GB of RAM, so it’s not the RAM amount, is it the RAM speed? Unlikely. So that is a really light load, which IS reflected in my CPU meters not moving much at all and that hard disk meter within Cubase showing zero. All I have is IK Multimedia’s Modo bass, 2 audio files, and the UVI workstation sampler playing one patch. That bottleneck… hmm… The hard disk meter isn’t reading anything in that pic, I’m running an NVMe drive and my Samples disk and they both stream you know over 300 MB/s or something much higher.
Without having read the articles, and I’m going to read them here in a bit, I just want to give a quick comment…
0 kommentar(er)
