Plugin Delay Compensation

[Shan]

3. The answer will either open a can of worms....


Neil.

PS Respect to you Shane for spending the time you have investigating this issue.

Hmmm...like maybe increasing this fixed buffer from 2000 to 4000 is quite simple to do?


Shane

[Shan]

Hey Shan,

just a stab in the dark here, but could the mystery 2000 samples of PDC be the key to "low latency monitoring" mode, ie: PDC buffer gets bypassed for faster response at the expense of being able to use plugins?

Quite possible. It would be nice to know how this works instead of us just guessing and deducing from doing tests. But...we could probably figure it out or get pretty close on our own.

Shane

[yoursickstudio]

I still believe in these times and in this competetive world the focus and goal should be put at :

FULL plugin delay compensation.

let me repeat : FULL

[yoursickstudio]

latency is a much bigger problem for mixing
just my opinion

when mixing i don't want to get hassled with delaying tracks etc...
during recording i don't mind removing some temp plug ins to achieve no latency

[DoubleDigit]

So far what I hear that sounds reasonable (and some of which I have verified through testing) is that (we think) there is PDC in LE, but only up to 2000 samples. The delay shown is what goes over and above the possible compensation and is in fact a "net" delay. Low latency monitoring is possibly the result of disabling this PDC (thereby keeping us from running plugs). This all sounds possible to me, but REALLY where is Digi on this??? I really wish someone would clear this up once and for all, especially if Digi is trying so hard to get us all mixing in the box.
This is important stuff, especially between tracks with bleed from other tracks (drums etc) as phase is concerned.

For now, I plan on doing the following: For any tracks which contain information common to other tracks (snare and overhead, or parallel compression auxes, for example) I will be aware of phase problems due to plugin delays and compensate by nudging, inserting and bypassing identical plugs, or printing the effect with the Audiosuite equivalent.
For tracks without common information (guitar track and a vocal overdub track, for example) this delay is not such a big issue- no more than any other very slight delay would be. In fact, unless the delay is greater than 660 samples (which would equate to 15ms in a 44.1 session) it is likely that we wouldn't even percieve it (between two different sounds). It would follow that these delays (again between two different sources, ie. tracks with no sounds common to both) would not begin to cause percievable timing issues (like noticeable misalignment between kick and bass, for example) until the delay were over around 2000 samples ( approaching 50ms in a 44.1K session).

Does all this stand to reason with you guys? Again...Digi...please chime in and shed a little light.

Ted