suggested workflow for overdubbing in a session with latency inducing plugins

[zwolf]

My questions pertains to my native rig, running 2020 on a 2012 mac mini with a UA Apollo x8.

It's acknowledged and documented by many that there's a long standing bug with the preference to "allow sends to persist with LLM" wherein this makes a doubled signal coupled with direct monitoring software that's impossible to work with. I've dealt with this by not having this option checked and trying to run lean sessions without many plugins while tracking. But I've come to the point where I have sessions that I've done some premixing work that need further overdubs by the bands that I'm working with. So in this scenario, I have both LLM and delay compensation active. But when I record enable a track with a high latency plugin on it, even though the insert is bypassed in LLM, the track is out of time. This doesn't happen with "allow sends to persist with LLB", but then you have the doubling issue.

I'm wanting to settle on the best workflow for dealing with this. The only options I can come up with are below. I'm hoping someone has a better idea!

- Turning off DC for overdubs. Pros: track you're monitoring plays in time. Cons: you have to remember to turn it back on. What you hear prior to the punch has no processing.

- printing the track you want to overdub on with all processing and DC, then monitoring that prior to the punch point. Pros: you hear what you expect to hear. Cons: makes it necessary to have cue sends be post fader and also muted prior to the punch. It's a PITA.

Anyone have any better insight into this?

[JingleDjango]

I used to freeze tracks with delays longer than 32 samples. Now I find it safest and most reliable to print stems, record overdubs in a new session with no plugins active and import the new tracks into the original session later.

[JingleDjango]

I should mention that I'm not running HDX (anymore) and all of my plugins are either running native or off of UAD2. Sometimes I will just engage LLM and monitor direct through the Apollo instead.

[PatriotsBiker]

Mine is somewhat dated and evolved over time. I found it easiest and least confusing for me to use this less elegant solution.

1. Bounce my session's output track WITHOUT the instrument being overdubbed.
2. Print a stem of the track destined to be over dubbed. Include, for reference, any aux track FX used to create that instruments final sound within reasonable limits.
3. Create a new session
4. Import session level data such as markers, bpm, etc, etc.
5. Import both of the bounced tracks from above.
6. Import the raw track to be overdubbed, complete with plugs used.(deactivate the plugs for now)
7. Then i do my over-dubs. No, they won't have all of the in the sauce sound they normally would, but that doesn't matter here.
8. Bounce the new dubbed track.
9. I re-open the original session and import the new over-dubbed track. (Not the entire mix, just the audio from the instrument WITH the overdubs.
10. Mute the original and add a new track instead of overwriting.

I do this for bigger sessions when foresight has failed me, resulting in me adding guitar or keyboard parts after I've hit the mixing phase.

I know it's a few steps, but the time and aggravation it ends up saving is well worth it. Especially with subsequent additions and modifications due to the bulk of what is in place already. Just bounce the latest tracks needed as described above.

I know I can de-activate as much as needed to accomplish the same thing within the same session. Doing it this way allows me to now have to worry about destroying or changing anything I've already done. Less of a management nightmare.

This bit of overkill sort of evolved starting before freeze and commit functionality hit us. Perhaps I need to learn more modern techniques.

ps - The big "gotchya" is when I use drum software such as Superior Drummer and Strike 2 and others before that. Not only will you have to re-adjust any phase issues you've previously fixeds if it was the drums that need to be modified or replaced.

[zwolf]

Originally Posted by PatriotsBiker

Mine is somewhat dated and evolved over time. I found it easiest and least confusing for me to use this less elegant solution.

1. Bounce my session's output track WITHOUT the instrument being overdubbed.
2. Print a stem of the track destined to be over dubbed. Include, for reference, any aux track FX used to create that instruments final sound within reasonable limits.
3. Create a new session
4. Import session level data such as markers, bpm, etc, etc.
5. Import both of the bounced tracks from above.
6. Import the raw track to be overdubbed, complete with plugs used.(deactivate the plugs for now)
7. Then i do my over-dubs. No, they won't have all of the in the sauce sound they normally would, but that doesn't matter here.
8. Bounce the new dubbed track.
9. I re-open the original session and import the new over-dubbed track. (Not the entire mix, just the audio from the instrument WITH the overdubs.
10. Mute the original and add a new track instead of overwriting.

Thanks for sharing your workflow. Can you clarify what you mean in step 2? Are you talking about the overdub-to-be track in question? I'm not sure what you mean by printing a stem of it if you haven't yet tracked the part. Surely there's something I'm misunderstanding....

[Sardi]

Originally Posted by zwolf

Thanks for sharing your workflow. Can you clarify what you mean in step 2? Are you talking about the overdub-to-be track in question? I'm not sure what you mean by printing a stem of it if you haven't yet tracked the part. Surely there's something I'm misunderstanding....

I assume he means a track of the instrument already recorded and then you would drop in on that track to fix mistakes etc.

I’m assuming you just want to record a whole new instrument over the top of an existing tracked session. If so, you can skip step 2.


Sent from my iPhone using Tapatalk

[zwolf]

Originally Posted by Sardi

I assume he means a track of the instrument already recorded and then you would drop in on that track to fix mistakes etc.

I’m assuming you just want to record a whole new instrument over the top of an existing tracked session. If so, you can skip step 2.


Sent from my iPhone using Tapatalk

Ahh, of course - thanks!

[PatriotsBiker]

That was it. Sorry about the floating second usage part in there. My bad. I didn't realize the question was as specific as it was. But yes, I do new tracks and spot fixing that way when processing gets in the way.

[zwolf]

I've settled on a workflow and edited my tracking template accordingly. My main concern in creating this post was addressing the scenario wherein I have a band in for tracking and need to facilitate punch in fixes and overdubs while also beginning to perform some pre-mix processing so the playback is inspiring. Up until now, I deactivated plugins on record enabled tracks to deal with the cue latency, but that's a PITA and also makes for different sounding/uninspiring pre-roll.

For situations with a more fleshed out production in need of a later stage overdub or fix, I'll utilize a process similar to the one outlined above by PatriotsBiker.

I'll just run it by you all here and let me know if you think there's a better way. I'm self taught and apparently more confused by latency than the next person, so there's that...

Essential requirement: no/low latency monitoring in artist cue mixes of their instruments and PT playback

1- LLM and delay compensation engaged in PT
2- Cue mixes of individual instruments set up in UA Console software
3- post fader sends on recorded tracks in PT sent to a Cue channel in Console
4- parallel compression buses created for instruments that benefit from doing so. I don't need to worry about the latency of the plugin chains on these auxes as they're not record enabled, do I understand that correctly? (and up to this point, so long as I don't have a plugin instantiated on a record enabled track, latency is not an issue for any overdub or punch, correct?)

Once a punch or overdub is needed on track/s with plugins instantiated:

5- A print track that's fed from the stereo bus is created
6- record necessary pre-roll of mix. Determine punch spot and mute this recording following it.
7- solo this print track
8- play the preroll and punch in. Band hears the mix as it's been as well as their direct monitored Console mixes of their instruments, plays along and takes over from punch spot.
9- mute and unsolo the print track and un-record enable punch tracks for band to hear what they did.

That's what I've come up with in hopes of achieving a workflow where I don't have to worry about latency. I want to be able to use my high latency UAD plugins as it's just more fun and inspiring to mix as you go and keeps the band engaged. And I need a guaranteed process for avoiding the dreaded "wait, stop....everything sounds weird and doubled" comment that sends me searching for the rouge offender and kills the vibe and stresses me out. If the "allow sends to persist during LLM" function worked as advertised and inserts would be active, delay compensated, and not doubled when record enabled, this wouldn't be necessary.

But I'm not gonna wait for a fix from AVID.

[zwolf]

One last bump and I’ll let it lie. Just looking for another set of eyes on this.


Sent from my iPhone using Tapatalk