I am wondering how latencies in general and especially with the new Rack 2 VST are handled in a DAW…
The audio module in Rack 2 shows a block size of 256 samples (5.3ms @48kHz) that can not be changed, but the plugin wrapper says “latency: none”, so I began to fear that every instance of Rack would add ~5ms of unreported latency. To test this I put 100 instances of Rack FX in a row, containing only an Audio-2 module patched to itself, but to my surprise I wasn’t noticing any additional latency when I was anticipating >500ms…
So… how does this work? And what should I pay attention to, to always get the lowest latency possible when using the VST in various ways? Can I assume it’s always latency free, with the exception of 1 sample for every cable?
PS: The 256 samples relate to my ASIO driver settings, I put it back to 64 samples now, which I usually try to run everything at… with 64 samples it’s only 1,3ms of (output-)latency, which together with input latency and processing buffer adds up to ~5ms roundtrip latency for the entire DAW, which is really good… if the latency doesn’t increase when using Rack VST I’m very happy with this.
It depends on the VST, some add latency, some don’t. Effects that use oversampling usually add a bit, and some synths do add a bit as well, but usually very little. There are some extreme cases like mastering plugins with additional latencies of 50-90ms (probably mostly because of look-ahead).
But I don’t understand how it really works, I just hope Rack doesn’t add any, and so far it seems it doesn’t.
I sometimes wonder how this works… like: why is it possible to run multiple VSTs in a chain without additional latency, but not possible to do oversampling without additional latency?
My fear was that Rack would add a bit because it might not be optimized for VST performance, but it seems it integrates into the buffer nicely and thus runs latency-free. I am very happy about this, because now running VSTs in the DAW gives me an advantage over running them inside standalone Rack (which always adds a bit of latency per VST) and I should be able to have a powerful low-latency-hybrid-setup.
It also makes me wonder: Would it be possible to run VSTs inside standalone Rack without latency by integrating them into Rack’s buffer?
I guess you must experience latency on the recorders / samplers, but I don’t think you will otherwise experience latency, you may run out of processor and experience hiccups but no latency
edited it is not a reply to @Squinky, is for thread in general , I pressed wrong button