dcumpian

pharohoknaughty
I like to hear things the way they will be, and not have to concentrate too much on technical details like changing to low latency plugs or bypassing the effects before recording. So this proposal would let you leave any FX plug in Sonar and it would not affect the VST synth.
 
For example , I use JJP Vocal which has a latency of 473, and I like to leave it in place at all times.

This is the source of all your latency issues while tracking. You just can't do that and expect mo latency. Even moving your instruments to another computer won't change anything because the latency is in your FX chain.
 
Record your parts first, completely naked except for any FX in the instrument itself that you may want to print. Once you've assembled all your tracks you can start the mixing process and start adding FX. I usually take it a sterp further and render all of my tracks as audio just to simplify the project. If I discover that I really need to add another instrument, it is no big deal to temporarily add a VST, enable the global FX bypass, record the new part, render, dump the VST then disable global FX bypass.
 
Regards,
Dan
 

Thanks for the advise and comments.
 
Moving the synth to another computer is akin to using a dedicated synth like a Korg Triton. Of course there will be digital latency, but unrelated to the DAW. The DAW has compensation for audio inputs and its own latency. So if you hit record, and there are lots of high latency plugs and high buffer settings, Sonar will line things up, making the new tracks sync to the prior tracks.
 
I suspect a Triton has latency on the order of 2 or 3 milliseconds.  The latency of a synth on a dedicated computer is going to be related to the A/D/A processing time, plus a little bit for overhead. And of course the terrible latency of MIDI transmission between machines.
 
So my point is that the latency on the DAW could be set very high and comfortable, say even one second if needed. But the synth (or guitar effects processor) would be independent. Not every audio devise on the computer has to have the same latency, but they can be synced.
 
I have proposed this idea as long as 15 years ago but it never gains traction. So I am the odd man out.
 
Your advise to record without effects and render the tracks so forth is good advise given today's state of the art. But I am pretty sure it could change if someone would bifurcate the two systems. The state of the art can move forward.
 
Thanks again for your comment.