PDC is killing my love for LOGIC

[skijumptoes]

Just a rough/quick test based on that chart, i rendered out a click track and aligned the peak to the downbeat then played it out my monitors through mic, and results are:-

 

 

Software ON, PDC ALL - as we expected/know will throw the recording back due to the Bus 'bug' (As some call it).

 

All other modes result in the same recording result. I sent my recording delay to 0, normally i would have it offset slightly. But the deviance is important here, there doesn't seem to be any other than on the first combination.

 

I think the answer for the OP is to run Software Monitor:On, PDC: Audio/Instruments. However, more precise details down to ms and samples maybe is something that you are interested in?

[Dewdman42]

To really do these tests properly we should use a separate LPX project for each test because if you setup a bunch of tracks/channels, even though some are muted, their PDC could be effecting the other unmated tracks you are testing. Its also worth keeping in mind that the source track with the recorded click will be effected by PDC too during playback.....it would be better to have an external sound source, but I realize that is maybe impossible to do.

 

That being said, i just tried to do similar test.....

 

I read in Edgar Rothermich's books that its possible to use hardware monitoring (software monitoring turned off) and still get the software reverb into the monitor mix, but with a few ms of delay. But he doesn't explain how to set that up in the book and I have tried everything I could think of and could not make it happen. Turning off software monitoring really does seem to completely block the incoming audio from being sent over a send to an aux, or to the outputs at all, its kind of strange. I don't know if that's a bug or if Edgar Rothermich stated things wrong in his book. I have sent an email to ask about it.

 

One thing that is interesting, with software monitoring on.. if I put a latent plugin directly on the audio track we're recording to, then we will hear the delay while recording, but LPX seems to be be smart enough to register the audio to the region correctly in that case, not late like we heard it from the latent plugin while recording it.

 

But when bussing or sending to an AUX channel with a latent plugin, we hear the delay while recording and its not registered to the track on time, its registered late... unless you change PDC to AUDIO/inst only. Then it seems to do that right thing, which doesn't make complete sense to me the way PDC is supposed to work, but anyway...there it is.. it works, though with PDC in that mode, its possible other existing tracks and AUX configurations in the project would be out of sync while you're working that way, so there is that.

 

I do have to say that is maybe an oversight or design flaw...if it works for latent plugins on the audio channel, it ought to work for AUX channels too. That is an inconsistency. I'm calling it a bug now. It is at a very minimum, very confusing for users to understand what is going on and could be improved by guiding users to the right solution. But I see an inconsistency there in how the record offset is calculated when AUX channels with latent plugins are used vs when audio channel plugins with latency are used.

 

Regarding using hardware monitoring in combination with an AUX reverb...should theoretically be possible according to Edgar Rothermich, but I can't figure out how to do it either.

 

Personally I think the best approach is to avoid complicated monitoring situations. Use hardware monitoring whenever possible. Use outboard FX if you have to while recording tracks.. Use temporary plugins in software monitoring mode, that don't impose any extra latency. Then plop the UAD in there during mix down or directly on tracks that are already recorded. But don't try to use latent plugins while tracking. It just opens up a can of worms.

[Dewdman42]

and by the way, through this I also discovered that the recording delay slider maxes out at +/- 5000 samples. so if you have some really big PDC hog on an AUX channel, you can't use the recording delay slider to compensate for it either, because it probably doesn't go high enough.

[robinloops]

Now use the audio file editor to zoom way in and select from one spot on the track you record from to the same spot on the track you record and look for how many samples you’ve selected. This is your offset. Repeat to get exact.

 

As far as this test. You can run it inside the project in question as well. Then just delete the tracks afterwards

[robinloops]

If you mean to use hardware monitoring with a logic reverb it’s pretty easy. You need to have software monitoring on for it to work.

 

Set a send for the audio track being recorded/monitored. Since you’re using hardware monitoring for the dry signal, make it a pre fader send. Put the fader to zero do you don’t hear the dry track (which would cause phasing issues because of being slightly offset from the live tracking because of latency). Now send to the effect you want to add and make sure to use 100%wet and 0%dry in that plugin.

 

However I find this to be a convoluted monitoring situation, which I too try avoid, so I isually use an external reverb unit for live tracking. To do this I use a hardware mixer with sends/returns.

[Dewdman42]

Now use the audio file editor to zoom way in and select from one spot on the track you record from to the same spot on the track you record and look for how many samples you’ve selected. This is your offset. Repeat to get exact.

 

As far as this test. You can run it inside the project in question as well. Then just delete the tracks afterwards

 

if the latent plugins have more than 100ms latency then you can't use the slider this way. Plus, to be honest its not really a suitable solution because you'd have to constantly be adjusting that slider as you move around from track to track and different things.

[Dewdman42]

If you mean to use hardware monitoring with a logic reverb it’s pretty easy. You need to have software monitoring on for it to work.

 

Set a send for the audio track being recorded/monitored. Since you’re using hardware monitoring for the dry signal, make it a pre fader send. Put the fader to zero do you don’t hear the dry track (which would cause phasing issues because of being slightly offset from the live tracking because of latency). Now send to the effect you want to add and make sure to use 100%wet and 0%dry in that plugin.

 

 

I've been trying to do that, but with software monitoring off, the pre-fader send is not being sent.

 

However I find this to be a convoluted monitoring situation, which I too try avoid, so I isually use an external reverb unit for live tracking. To do this I use a hardware mixer with sends/returns.

 

no argument there.

[robinloops]

For it to work you have to enable hardware monitoring (so the signal can be played back through Logic). Then you have to manually disable hardware motoring (so to speak) by lowering fader levels. You can also enable independent input monitoring level meters (in the mixer) so you don't have to change the actual track level when monitoring live (and using internal reverb). Also only works with tracks armed for recording.

 

Another even more complicated solution is to use an in/out utility on a track with the a reverb (or any other effect) that can work as a wet signal only (to avoid phase issues with the dry signal) in conjunction with a send/return on an external mixer. But then were getting even more convoluted... And a hardware reverb unit and mixer can be found for very little money as they don't have to be very high end since they are only for tracking and nothing actually goes through them for recording... But are being sold for pennies because they aren't up to par with todays standards...

[Dewdman42]

For it to work you have to enable hardware monitoring (so the signal can be played back through Logic).

 

how do you do this? I"m assuming you mean turn off software monitoring mode. And yes I already did that and yes I have hardware monitoring enabled in my audio device. That is not the problem though, the problem is that the pre-fader send is not being sent by Logic. Try it yourself! Maybe it used to work and doesn't now? as soon as your turn off software monitoring, the send is cut off from being sent to the bus. The meters also show this to be true.

 

I think maybe you were meaning to leave software monitoring on and combine that with hardware monitoring...and that does work but also results in record offset problems as mentioned earlier in the thread.

[Dewdman42]

alright robreed2, I found a way to combine hardware monitoring with software monitoring to get the best of both worlds...

 

Its a bit involved and I agree it should be easier...but...can be done like this...

 

1. Setup your audio tracks to record, does not matter whether you use the Input monitoring on each one
    
    
   
2. Turn off software monitoring
    
    
   
3. Go into the environment, create an input object for each mic'd inputs you will be recording.
    
   
   
    
   
4. in the inspector of each one, make sure the input channel is set to each of the actual input channels from your audio device that your using for your drum mics or whatever.
    
   
    
    
   
5. Go to the main mixer. The input channels you created in the environment will now be shown there also. Set the output of each of the drum sounds to a new bus. When you do that an AUX channel will be created. Put a reverb on it. Make sure all of these input channels are outputting to that bus (and thus to that aux channel).
    
   

 

That's it.

 

You should have zero latency hardware monitoring happening on the dry sound, the input sources will be duplicated through these other input channels and sent to the software reverb for you.

 

it turns out that when you create input objects directly this way, they ignore the software monitoring button and always get the sound regardless of that setting, so you can have software monitoring turned off to get the correct record timing we talked about and still hear the bussed reverb like this...but just remember there will be some ms of delay to whatever you have on that bus if there is latency there.

 

And another nice thing, that reverb used this way will have ZERO IMPACT on the record offset back over the audio track you're recording to, they are totally separate.

 

Note also that once you have recorded these tracks, then you will want to have the recorded audio tracks use the normal send to go to the reverb

 

 

Also, the input channels you created are like live channels that are always listening to those inputs. So you will want to mute them when you're not wanting to record from them and probably just remove them from the project once you're done recording tracks..they are only needed during the tracking phase in order to get the software reverb into your monitor mix while software monitoring is turned off for timing reasons.

[JakobP]

Actually there is a much simpler way to achieve the reverb when software monitoring is turned off (or on for that matter), simply create an aux in the mixer, insert the reverb, and set the input of the aux directly to the interface input...

[Dewdman42]

Except that way you can only feed one input at a time to the bussed reverb, or two in stereo I suppose

[JakobP]

Yes you're right, it's just a way to have a "monitoring reverb" without having it in the signal chain when recording. Avoiding the reverb in the signal chain maybe would bypass the PDC problems discussed here?

[drakepeterdrake]

Dewdman, we worked on the weird external instrument vs internal midi latency thing last year. I'm fascinated by your suggestion above using the environment. I don't have monitoring problems when recording, as I use a console and monitor the input directly. However, due to LPX's way of handling PDC in auxes and busses, mixing is a pain. As I said in an earlier post, with PDC set to ALL, LPX calculates the worst-case latency of any track in your project, then delays auxes by that much. The result of that is that is I am sending signal on a send from the console to a software reverb in LPX, I get horrendous pre-delay - sometimes 200ms. I found that for my purposes, the easiest thing to so when mixing is to put LPX PDC not in ALL but in "audio and software instrument" mode, then calculate the latency introduced by plugins that are in busses, like the background vocals. I then introduce a negative offset for those tracks in the inspector, essentially doing manual PDC. This leaves my reverbs, which are in auxes, unaffected by other tracks' PDC.

 

However, I'm wondering if I could use the method you described above, or something similar, to take my auxes with reverbs "out of the equation" as far as LPX PDC is concerned. They would be treated as live tracks, and even with PDC set to ALL, should be unaffected by LPX's PDC. They would receive signal from the console on inputs and send the 100% wet signal back to the console, with the only latency being that generated by the reverbs themselves. (UAD reverbs typically have about 25 ms at 44.1k) Do you think this is possible, and if so, could you show me? I have never ventured into the environment. Thanks very much.

Peter

[Dewdman42]

It’s true if you are only going to overdub two tracks at a time, it’s not necessary to create the input channels.

 

But in the other hand if you wanted to overdub more and more tracks sharing one fx bus, then you’d need to setup the input channels so that a bus feeds the aux channel and your pre-recorded tracks feed the bus as well as the live inputs.

 

The “problem” encountered by the op seems to be only relevant when using a shared fx bus. Turning off software monitoring cures the record offset “bug” but then this is a way to keep using a shared fx bus with both pre recorded and live input feeding it, and live input could be 8 or ten mics at once or whatever. Other wise when software monitoring is turned off, to work around the offset “bug”, the live input is cut off from all busses and sends.

[Dewdman42]

Peter will respond to your post when I get to Computer, on iPhone now

[robinloops]

For it to work you have to enable hardware monitoring (so the signal can be played back through Logic).

 

how do you do this? I"m assuming you mean turn off software monitoring mode. And yes I already did that and yes I have hardware monitoring enabled in my audio device. That is not the problem though, the problem is that the pre-fader send is not being sent by Logic. Try it yourself! Maybe it used to work and doesn't now? as soon as your turn off software monitoring, the send is cut off from being sent to the bus. The meters also show this to be true.

 

I think maybe you were meaning to leave software monitoring on and combine that with hardware monitoring...and that does work but also results in record offset problems as mentioned earlier in the thread.

You have to have software monitoring ON or nothing will sound from logic, not even the wet signal from the effect. You just lower the track level so you don’t hear the dry signal and use a pre-fader send. Doesn’t require environment or any trickery. Simply use software monitoring but o to monitor the wet signal through logic.

 

You can even just set a the reverb on the audio track being monitored ( with software monitoring on) and put it to wet only.

[Dewdman42]

You have to have software monitoring ON or nothing will sound from logic, not even the wet signal from the effect. You just lower the track level so you don’t hear the dry signal and use a pre-fader send. Doesn’t require environment or any trickery. Simply use software monitoring but o to monitor the wet signal through logic.

 

You can even just set a the reverb on the audio track being monitored ( with software monitoring on) and put it to wet only.

 

Hey Robinloops, read the whole thread. There is a bug when using software monitoring and a bus that causes incorrect record offset. That is what the OP is trying to work around now.

[Dewdman42]

Dewdman, we worked on the weird external instrument vs internal midi latency thing last year. I'm fascinated by your suggestion above using the environment. I don't have monitoring problems when recording, as I use a console and monitor the input directly. However, due to LPX's way of handling PDC in auxes and busses, mixing is a pain. As I said in an earlier post, with PDC set to ALL, LPX calculates the worst-case latency of any track in your project, then delays auxes by that much.

 

Alright so first let me just clarify one thing which may be important to keep in mind while you plan your strategy. The way LPX handles PDC, according to Edgar Rothermich in his books; is that when there are latent plugins on audio and instrument channels, then those track regions are fed early to the mixer so that they will sound in sync with the clock. When you are recording your parts to those tracks, however, they can't be fed early..so they will just sound late...unless you use low latency mode to disable the latent plugins.

 

For AUX and OUTPUT channels its handled a bit differently. If there are latent plugins on AUX channels, its considered like a live feed, so what LPX does is to delay everything else, so that everything lines up, unless of course PDC is set to audio and inst only, in which case that behavior doesn't happen. And if you use low latency mode, then LPX will determine if AUX plugins need to be disabled also I have found.

 

That is how Edgar describes LPX PDC. Its not clear to me whether that is the whole story or not, but for now let's just go with that.

 

The main point being...LPX does not cause AUX channels to be delayed even more. AUX's are delayed by their own plugins.

 

Perhaps you were referring to the record offset when AUX's with latency are in the signal path of something being recorded and yes, there seems to be a bug related to that. The work around for now I think is to turn off software monitoring mode, which causes the record offset to be much much tighter, regardless of audible delays due to latency on the bus sends. But when software monitoring is off, the bus feeds from audio tracks don't work anymore.... so...

 

However, I'm wondering if I could use the method you described above, or something similar, to take my auxes with reverbs "out of the equation" as far as LPX PDC is concerned.

 

Yes I absolutely think so. note however, when we were looking into this months ago, we were looking at external midi issues and at this point I do not know the ramifications of that. My studio is currently being relocated and I don't have external midi gear at the moment to test with. Its possible this may lead to some solutions in that area too.

 

They would be treated as live tracks, and even with PDC set to ALL, should be unaffected by LPX's PDC.

 

Yes. Well more correctly stated... the aux channels were never effected by PDC. Rather the opposite, they effect other channels. And in particular if you have a bussed AUX channel in the signal chain on an audio or inst track you are recording, then the record offset is not being calculated correctly and the events end up recorded into the region late. However, by splitting the incoming audio, one fork goes to the audio track you're recording to and its not bussed or sent to any AUX; the other fork goes directly through the input channel to an AUX with latent FX. That disassociates the AUX from the audio track you're recording to, its an independent channel in the mixer. That may or may not effect PDC in the mixer per Edgar Rothmich's description of how PDC works, but the record offset of the audio track you're recording to doesn't experience that Aux bug, it rather registers incoming audio exactly where its supposed to.

 

And also software monitoring can be turned off if you are using hardware monitoring, and you'll still hear those independent input channels through the bussed FX...because the software monitoring button seems to only effect record enabled audio tracks or audio tracks with the I button engaged for input monitoring.

 

They would receive signal from the console on inputs and send the 100% wet signal back to the console, with the only latency being that generated by the reverbs themselves.

 

Yes exactly and meanwhile the dry sound would be recorded to your audio track exactly at the correct time it was meant to be registered to the region.

 

Do you think this is possible, and if so, could you show me? I have never ventured into the environment. Thanks very much.

Peter

 

So look at my previous post. This is dead simple environment work. Just open the environment window and by default it will be showing the mixer objects page. Go to the new menu, and create a new channel strip input object. when you select that object, to the left you will see an inspector and the "channel" property will refer to which audio device input channel that input object will be referencing. So make as many of those input objects as you will need to record at once. One stereo or multiple mono's, whatever you need.

 

when you go back out to the Logic mixer you will see them in the mixer. (if you delete them from the mixer, they will disappear from the environment too). Unfortunately there is no way I know of to create them in the mixer without using the environment to do it. Once they are there then they are receiving audio directly from the audio device input channels and you can configure them in the mixer to bus or send the output anywhere you want. See my previous post with one example using two input channels feeding one FX bus.

 

its worth a try Peter, and I'd be curious how this may effect external midi and inst recording since you will be eliminating AUX FX busses while recording tracks, it just might eliminate that problem for now. Probably won't cure the external midi metronome issue though.

[robinloops]

I thought it had evolved to finding a direct monitoring solution to avoid offset but wanted to use some reverb for tracking purposes. In those cases you can leave software monitoring on, and only 'monitor' the wet signal of the effects you want to use (to mix with the direct dry signal). It then would record the dry signal without any reverb. Then after recording and starting to mix, you either change the send back to 'post fader' or use a different sends for 'mixing' reverb' and tracking reverb.

 

In that scenario, it is no different than software monitoring except you only 'monitor' the wet signal through logic but still monitor the dry signal live/direct. While this will cause latency with the reverb trails, making it sound slightly different when tracking live but are still monitoring the dry signal in real time (no latency) and the wet signal will only have latency.

 

It's a best of both worlds scenario, direct monitoring but with the ability to add Logic effects. The only limitation is that it only works with record armed tracks and only for as many inputs as you have for software monitoring. If one needs to use an external mixer to sum more inputs than they have on their interface, then the same trick (wet software monitoring with dry direct monitoring) an external mixer needs to be used and an in/out utility in conjunction with the external mixer's aux send/return.

[Dewdman42]

as mentioned by others, there are a few variations on the above theme...so here are the variations as I can think of them, in case people want to try different ways of accomplishing the basic basic overall goal, which is to split the incoming audio so that the audio track you're recording to doesn't have to bus or send anything to an FX Bus, you keep it isolated so that the record offset bug mentioned earlier will not mess up the recording into the regions.

 

So options are:

 

1. use the method described above, create input channel objects which split off the audio device input away from the recording audio track to monitor FX through them to a bus.
    
   
    
    
   
2. You can create AUX channels instead of INPUT channels, and AUX channels can take their input from any input, so you can duplicate the input that way too. And you don't need to go into the environment to create an aux channel, you can do it from the mixer's submenu. In this case you'd need one AUX channel for each input channel (or each pair of stereo input channels), and then if you want to use a bus FX you'd need one more AUX channel to host the FX and send them all to that one.
    
   But you could, for example, set it up like this:
    
   
    
   One thing I prefer about using input channels instead of this AUX channel approach, is that they will be sorted to the right end of the mixer next to all the output channels. Plus you can use the INPUT blue button on the top right of the mixer to hide them. If you create AUX channels then they will be intermixed with your other AUX channels in the middle of the mixer and unable too hide them without hiding all your other AUX channels. Eh..could be other pros and cons. But it will work either way I think. I just think I prefer using Input channels because it designates them for that special purpose and I will not lose track of the fact they are doing that. but if you don't want to open the environment at all, you can do it this way.
    
    
   
3. You can use the I/O plugin to get the input from a single (or stereo set) input channel on your audio device. I messed around with this a bit and decided it was more hassle, but might be useful in some simple situations. It basically can only take one mono or stereo input at a time, so would be a way to handle perhaps situation where you're only overdubbing one track at a time, without having to create extra channels and such.
    
   

 

That's actually all I can think of right now... will update this post later if we think of something else applicable.

[Dewdman42]

I thought it had evolved to finding a direct monitoring solution to avoid offset but wanted to use some reverb for tracking purposes. In those cases you can leave software monitoring on, and only 'monitor' the wet signal of the effects you want to use (to mix with the direct dry signal). It then would record the dry signal without any reverb. Then after recording and starting to mix, you either change the send back to 'post fader' or use a different sends for 'mixing' reverb' and tracking reverb.

 

If you have an FX bus and the audio track you are recording is sending or bussing to that AUX channel, then the record offset will be incorrect when you record audio to that audio track when software monitoring is on. That is the bug we are trying to work around. You can only do that when software monitoring is on because when software monitoring is off, the sends and busses from the record enabled audio track are cut off.

 

In order to work around that record offset bug you can turn off software monitoring, which also turns off the ability to send or bus the audio track to the Aux channel.

 

With this above trick, you can use the FX independent of the audio track you're recording to. You still might want software monitoring on if your audio device does not support hardware monitoring but the record offset bug would persist because of software monitoring mode...so... bottom line, if you're using a bussed FX with software monitoring on, you're going to end up with incorrect record offset. Find a way to turn off software monitoring and setup this alternate monitoring channels.. Theoretically if you don't have actual hardware monitoring available in your audio device, you could still setup alternate input channels in the LPX mixer to monitor them..and they would essentially act just like software monitoring, but with it turned off to avoid the bug.

 

In that scenario, it is no different than software monitoring except you only 'monitor' the wet signal through logic but still monitor the dry signal live/direct. While this will cause latency with the reverb trails, making it sound slightly different when tracking live but are still monitoring the dry signal in real time (no latency) and the wet signal will only have latency.

 

No because the software monitoring bug will be present if you're using a latent FX bus.

 

It's a best of both worlds scenario, direct monitoring but with the ability to add Logic effects. The only limitation is that it only works with record armed tracks and only for as many inputs as you have for software monitoring. If one needs to use an external mixer to sum more inputs than they have on their interface, then the same trick (wet software monitoring with dry direct monitoring) an external mixer needs to be used and an in/out utility in conjunction with the external mixer's aux send/return.

 

If you are using record armed tracks to monitor through software monitoring and you plan to use an FX bus, then no that will not work, it will result in the record offset bug mentioned in this thread.

Edited February 8, 2019 by Dewdman42

[Dewdman42]

I just did a quick test and even with the above tricks to separate the FX bus from the audio tracks being recorded to, if software monitoring is ON, then the recorded audio track is registered to the region late. So the bug does appear to be related to software monitoring in general.

 

I think the fact that the AUX channel has latency causes the other channels to sound later and LPX calculates the record offset based on that. Bleh..

 

However if you absolutely must use software monitoring for some reason, then change PDC to audioinst only and then I guess it will be recorded correctly, though any other existing tracks in your project may not be playing properly in sync while you're in that mode.

[drakepeterdrake]

Hi, thanks for replying. I'll experiment with it. However, I am most definitely referring not to record offset, but to PDC-induced inaccuracies in playback. Again, I am not using software monitoring at any time. What I did, as an experiment, in a new session, was to feed a click from a synth into my console. Then I used an aux send on the console to send signal to a UAD reverb in an aux channel in LPX, with the output of the UAD assigned through LPX into a return on my console There was negligible predelay audible in the reverb, less than 30 ms, with PDC set to ALL The console has a worst-case 2ms delay, UAD plugins have about 25ms delay. When the UAD reverb was removed or bypassed, the round-trip time delay was very low, less than 10ms.

 

I set up an audio channel, loaded in enough plugins to need about 200 ms of PDC. Still, no audible predelay on the reverb. Then I made bus, or sum, or some background tracks, and loaded that with 200 ms worth of plugins. Immediately there was a 200 ms predelay audible on the reverb. When I then set PDC to "audio and software instrument tracks", and put as many plugins as I wanted into any audio and/or bus channels, there was no added predelay audible in the reverb. That's why I now, when mixing, manually fix the latency in any audio sum tracks and leave PDC in "audio and software", not "all."

 

It would just be nice to not have to do that, so I will experiment with your scenario. Edgar's book make sense, but LPX should have a way of determining which busses need PDC, and which ones, such as those carrying my reverbs, do not. Admittedly, most people don't use consoles anymore, but I still like knobs, and UAD and others like Liquid Sonics have some wonderful software reverbs. I have hardware reverbs which of course have no latency issues when used with the console, but getting a hardware EMT 140 just doesn't make a lot of sense these days...

 

P

[Dewdman42]

So here is a setup that more or lesses emulates "software monitoring" but without actually having to have the software monitoring button ON. This will avoid the aux software monitoring bug while still keeping PDC set to all.

 

 

Notice that in this case the monitoring happens entirely through the AUX1 channel, which has its own send to the other AUX2 channel with the bussed FX.

 

So you could have other channels feeding that bussed effect, for example.

 

 

Notice in the above example also how the audio tracks are also sending to the FX bus. Note that during recording with software monitoring oFF, that send won't happen. But then when you playback the tracks, they will already be setup to send to the same FX bus that live inputs are going to.

 

But I think I still like using INPUT channels for this monitoring purpose, mainly because they will be shown to the far right on the mixer in ALL mode, next to the output channels. And they will be hidden from the mixer Tracks view...

 

Edited February 8, 2019 by Dewdman42

[Dewdman42]

 

I set up an audio channel, loaded in enough plugins to need about 200 ms of PDC. Still, no audible predelay on the reverb. Then I made bus, or sum, or some background tracks, and loaded that with 200 ms worth of plugins. Immediately there was a 200 ms predelay audible on the reverb. When I then set PDC to "audio and software instrument tracks", and put as many plugins as I wanted into any audio and/or bus channels, there was no added predelay audible in the reverb. That's why I now, when mixing, manually fix the latency in any audio sum tracks and leave PDC in "audio and software", not "all."

 

As I said... when you have an AUX channel with latency, then all other channels are delayed to match it. So when you added that bus with 200ms of latency, all other channels were delayed to match it, including audio, inst, other aux channels, etc. everything. Unless you switch PDC mode to audioinst only, then that behavior doesn't happen, but then of course your project will not be playing at all in sync. You really want PDC set to ALL whenever possible because you want all of your preexisting tracks to play in sync, the problem comes in when you're trying to overdub more tracks and software monitoring is a problem. So personally I think if you have direct monitoring as an option, then use that with the above stuff. Leave PDC on all. Well you might need to turn PDC to audioinst only for recording software instruments...where they would otherwise be delayed by the PDC ALL and hard to record your tracks that way... Or else use low latency mode which will disable your bus FX. Or use temporary bus FX that aren't high latency.

 

It would just be nice to not have to do that, so I will experiment with your scenario. Edgar's book make sense, but LPX should have a way of determining which busses need PDC, and which ones, such as those carrying my reverbs, do not. Admittedly, most people don't use consoles anymore, but I still like knobs, and UAD and others like Liquid Sonics have some wonderful software reverbs. I have hardware reverbs which of course have no latency issues when used with the console, but getting a hardware EMT 140 just doesn't make a lot of sense these days...

 

P

 

yea I am using an X32 for my audio device and it has some built in reverb and other FX which I plan to use when I need to while recording tracks. I am simply going to avoid like the plague using any bussed FX with latency while recording tracks. Even with the above trick...while it will work and the record offset will be correct, there could be some latency on the bussed FX if those plugins have a lot of latency, which means they will not be in sync with the tracks themselves...they will be late..which might be ok enough for a reverb and if the latency is not too big, but when you start talking about 100 or 200 ms, its a problem.

Edited February 8, 2019 by Dewdman42

[robreed2]

Bloody hell.>! Lots to take in..... BUT a massive thanks..

I think we have it.... If I can keep PDC set to ALL, to build mix with UAD on buses, and switch off Software Monitoring, to enable me to overdub a guitar or vocal, and it to be replayed in time.. then use the routing system shown to send the over dubbed vocal or guitar to a internal reverb or delay when track laying them ( I can live with a slight delay on these, just to give a bit of feel)

So I think Im happy >> I just need to try it all out tomorrow..

[Dewdman42]

Eventually i would like to find out how exactly Cubase and/or ProTools handle all of these situations in order to make feature requests to Apple. When you say it "just works", I can't help but thinking, it should just work in LPX too. Look at all the fiddling we are having to do. But..I do think you can get it working to get the job done and LPX has many other benefits. I think LPX has more dials and knobs to adjust related to Latency, which basically means it can be done, but if you don't even know how to do that, then look how confusing it is. Just sayin'.

[drakepeterdrake]

Dewdman, I think I have made it work for my purposes. In the environment I created an input, named it "emt feed," assigned input 21 to it, had it output to bus 10. I assigned the send on my console to that input. Created an aux, bus 10, placed the UAD 140 into it. As shown in the screenshot, there is about 200 ms of latency in the sum marked BV. This has no impact on the predelay of the UAD in Aux 5 while it is being fed signal from my console. All this is while PDC is set to "ALL." So, for my reverbs, I have essentially crested live feeds for them that are not subject to PDC, which is wonderful. There are a couple of oddities: while the EMT 140 plugin is fully functional, for some reason the input meter on the plugin itself shows no input, even though I am cranking into it. This is a minor issue. Also, no level shows on the meter of the "emt feed," also not an issue. Output does show on the main console from the UAD 140, which is sent back to the console. Software monitoring is off.

 

My questions are these: In your earlier post you said the input I created, here called "emt feed" would show up on the main console, but it doesn't. Not a big problem. Also, if I set up these live feeds for the 6 or so reverbs I use in a typical mix, can I use this setup in other songs or do i have to create live inputs for each song.

 

Thanks again.

P

[Dewdman42]

a couple things

 

PDC should always effect all channels in the mixer, even these. If you put some other AUX channel with 400ms of latency, then this one will be delayed along with everything else in the mixer. But what will be done right is the record offset over on the audio tracks you're recording to will work properly.

 

To see the channel strips, the Input channels only show up on the mixer "ALL" view unless you add those channels as Tracks, then they will show up on the "Tracks" view of the mixer. I like keeping them on ALL cause I want my tracks view to be more consolidated, the input channels are kind of set it and forget it to a point. Then even in the ALL view you can hide them, in the top right corner of the mixer you see the types of channel strips to show or hide, there are a separate set of filters for each of the three main views Single, Tracks, All.

 

I also created a VCA, which is attached to all of my input channel faders and that shows up on my Tracks view, so that I can mute all my input channels quickly or change their overall level quickly, including from the Tracks View.