Do Logic's Busses add latency?

Like say if you have a pre-master bus and then stems beneath that, and then a separate bus for each track and then the actual standard track. So essentially something is going through 3 buses in series.

Does stacking buses in series this way add latency?

The reason I ask is that I'm creating an autoload that will have some of this stuff prebuilt...and so I'm wondering is it too elaborate of a set-workspace...

Thanks for any info!

Well I hate to say it but yes they do !! Not all the time but lately I've been experiencing annoying latency issues on track which were routed to bus.

Also I now have latency on a midi ultrabeat track when opening a project on a mac book pro (9.1.3 and even 9.0.2 before) when there was no problem on the imac (9.1.1) when I made it.

any idea what's going on ?

Well I hate to say it but yes they do !! Not all the time but lately I've been experiencing annoying latency issues on track which were routed to bus.

You were probably using latency generating plug-ins on the Bus or Aux receiving that bus. The plug-ins are creating the latency, not the bus. If you'd put the plug-ins on the track itself and not on the bus/aux, you'd get the exact same latency.

Well I hate to say it but yes they do !! Not all the time but lately I've been experiencing annoying latency issues on track which were routed to bus.

 

You were probably using latency generating plug-ins on the Bus or Aux receiving that bus. The plug-ins are creating the latency, not the bus. If you'd put the plug-ins on the track itself and not on the bus/aux, you'd get the exact same latency.

how do you get around this?

i like to send my kick and snare for example to a bus,, then compress that buss (i.e. multipressor, and layer it in) but the latency will not alow this.

any ideas?

many thanks

Set your Plug-in Latency Compensation to All.

Logic Pro > Preferences > Audio >General

Set your Plug-in Latency Compensation to All.

 

Logic Pro > Preferences > Audio >General

ok thanks

i had that but i was adding plug ins on the fly so thats where i messed up..looks like you need to start stop to get it reset..getting to used to ableton live i guess.

many thanks

on another similar note

i am running an eventide 8h000fw with adat through my sound card.

when i set up a buss with an I/O plugin, there is a latency on the way back in.

my delays are way off.

i havnt used logic in a bit but in live i dont have this issue.

any ideas would be welcome

i am using 256 buffer

many thanks

i was adding plug ins on the fly so thats where i messed up..looks like you need to start stop to get it reset..

Yep, that's correct. When adding plugins it's best to first stop playback - same when routing or rerouting audio ((re-)assigning sends, buses, auxes, inputs, outputs etc) Logic is not designed for live use, but Live is - that is probably where it's got its name...

bump

You should create a new topic for a different question.

See the Forum Guidelines, item #4.

http://www.logicprohelp.com/viewtopic.php?t=29410

No, busses do not add latency.

Actually, you're wrong! I'm afraid they do: even when empty!!! Even just two in series is noticeable...

It's very slight though...but I do notice it playing a synth...in fact I now switch back to stereo output when tracking...

No, busses do not add latency.

 

Actually, you're wrong! I'm afraid they do: even when empty!!! Even just two in series is noticeable...

 

It's very slight though...but I do notice it playing a synth...in fact I now switch back to stereo output when tracking...

Sorry but that's just incorrect information.

No, busses do not add latency.

 

Actually, you're wrong! I'm afraid they do: even when empty!!! Even just two in series is noticeable...

I'm afraid you may have something else creating the latency in your project. A bus does not add latency. A bus is used to route signal from one channel strip to another. It's like a virtual cable if you want. A bus does not add anymore latency than, say, a guitar cable.

When you say "even when empty" I'm assuming you're referring to the Aux that receives the signal from the bus and not the bus itself? And Aux is just a channel strip, which is a device allowing you to route audio signal, change its volume and pan, and insert plug-ins to process the signal. A channel strip (whether it's an audio, software instrument, aux or output) does not create latency. The plug-ins inserted on the channel strips, on the other hand, can create latency.

All,

This thread has interested me so I did some minor testing. I can understand that parallel bussing creates no latency as signals are being paralleled i.e one signal being split into many (for whatever reason). In series I'm assuming that sends are being daisy chained so I set this up as follows:

- Channel #1 with UB playing a kick.

- Set the output of Channel #1 to Bus 1

- Set the output of Bus 1 Channel Strip to Bus 2

- Set the output of Bus 2 Channel Step to Bus 3....and so on.

- I did this until I had Bus 10 going to Stereo Out.

- I put a Channel EQ and Meter on Channel #1 and Bus 10

The difference in Channel EQ and Level Metering could suggest that there is latency in "extreme" series bussing:

I repeated the set-up however; this time I set all outputs Channel 1 --> Bus 10 off and created sends. Channel #1 --> Bus 1, Bus 1 --> Bus 2 etc.....and saw the same thing.

I'm not sure you'd see a noticeable difference if series bussing was being applied to 2-3 channels however; I can see in extreme cases that latency could be introduced, There are also some caveats such as the resolution of metering, channel EQ resolution etc. I'm also led to believe that Channel EQ's themselves can create latency.

Interesting results methinks.....

Andy

Bounce the source and bussed signals out and null test them.

Logic's metering displays are too inaccurate to draw any meaningful conclusions in this manner.

Bounce the source and bussed signals out and null test them.

I just tested it:

TEST #1:

An EVP88 goes through a series of 25 chained busses (bus 1 > Aux 1 > Bus 2 > Aux 2 > .... > Bus 25 > Aux 25 > Stereo Out). No noticeable latency when playing live.

TEST #2:

Two audio tracks play the same audio file.

Track 1 goes straight to the Stereo Out.

Track 2 goes through the 25 busses in series then to the Stereo Out.

Reverse the phase of Track 1, you get perfect silence. The peak meter on the Stereo Output doesn't display any signal:

Conclusion: busses do not introduce latency.

I've attached the project below, so you can try it yourself:

• drag any audio file to track 1,

• option-drag the region to track 2 to make a copy,

• press play and you'll hear perfect silence.

Bus-test.logic.zip

Bounce the source and bussed signals out and null test them.

 

I just tested it:

Nice proof! Nice proof indeed... 25, to be sure to be sure! Very good.

I'm an idiot...but yes, there are lots of things in my project that are causing latency; it's hard to track down what's what when there's so much going on...i/os, plugs, synths, ADAT sends; 4 stems split into 8 submixes down into 64 channels (sometimes themselves sub-bussed)...there was probably something else in the chain...I'm a gainstage freak.

Busted.zip

Bounce the source and bussed signals out and null test them.

 

I just tested it:

 

Nice proof! Nice proof indeed... 25, to be sure to be sure! Very good.

 

I'm an idiot...but yes, there are lots of things in my project that are causing latency; it's hard to track down what's what when there's so much going on...i/os, plugs, synths, ADAT sends; 4 stems split into 8 submixes down into 64 channels (sometimes themselves sub-bussed)...there was probably something else in the chain...I'm a gainstage freak.

latency sucks

Bounce the source and bussed signals out and null test them.

 

I just tested it:

Cheers David! 8)

Perhaps we should add a new (audio-forum) internet meme:

NTSFU

(Null Test or Shut the Heck Up)

Bounce the source and bussed signals out and null test them.

 

Logic's metering displays are too inaccurate to draw any meaningful conclusions in this manner.

Yep - just tested it too, perfect silence..

David is right - Busses do not add noticeable latency.

beej is right - Logic's metering is too inaccurate to draw any meaningful conclusions.

Good stuff....a very interesting thread.....

Andy

Conclusion: busses do not introduce latency.

Here is a question (not an argument):

Using something like EZ drummer on an instrument track and output to Stereo 1-2.

In addition, use a Send ( at unity) that will go to an Aux strip and also out Stereo 1-2.

A Gain insert on the Aux strip with phase invert will null out the signal.

Here is the question: Why does the Bus Insert turn Orange when LLM is activated (low latency mode limit setting doesn't matter)?

Here is the question: Why does the Bus Insert turn Orange when LLM is activated (low latency mode limit setting doesn't matter)?

It's being disabled to so that your low latency monitoring signal won't enter other latency inducing parts of your mixer.

Let's start with two tracks, drums and moog bass.

The drums are already recorded but you need to record more bass.

You happen to have a bunch of latency inducing plug-ins on your Stereo Output.

LLM won't disable those plug-ins for for the drum track but will ensure you low latency on the bass track by routing the signal from the original Channel Strip to the stereo output, BUT after the last latency inducing plug-in in the chain!

So if you had a send from your bass too it would go through the latency inducing plug-ins on your output, causing the two signals to get out of phase with each-other.

Hope that makes sense.

Nice proof! Nice proof indeed... 25, to be sure to be sure! Very good.

I was going to go to 64 but I got tired and figured 25 was enough for our purposes.

there are lots of things in my project that are causing latency; it's hard to track down what's what when there's so much going on...i/os, plugs, synths, ADAT sends; 4 stems split into 8 submixes down into 64 channels (sometimes themselves sub-bussed)...there was probably something else in the chain...I'm a gainstage freak.

You'd have to take some of things out of the chain one by one until you can eliminate the latency, figuring out what part of the signal chain is the culprit.

• Low Latency Mode will help you figure out what plug-ins create latency.

• Synths are sound sources, they cannot create latency.

• i/o's, sends, those are routing tools. Like busses, they do not create latency.

• stem split into submixes into 64 channels, sub-bussed etc.. are all normal mixer routings. They do not create latency.

• however, the use of i/o's and ADAT indicates that you are going out of Logic, through buffers that may be introducing latency. I would look there first.

I had a look at your attached file which is my file that you have modified... not sure what do with that.

Here is the question: Why does the Bus Insert turn Orange when LLM is activated (low latency mode limit setting doesn't matter)?

 

It's being disabled to so that your low latency monitoring signal won't enter other latency inducing parts of your mixer.

 

Let's start with two tracks, drums and moog bass.

The drums are already recorded but you need to record more bass.

You happen to have a bunch of latency inducing plug-ins on your Stereo Output.

 

LLM won't disable those plug-ins for for the drum track but will ensure you low latency on the bass track by routing the signal from the original Channel Strip to the stereo output, BUT after the last latency inducing plug-in in the chain!

 

So if you had a send from your bass too it would go through the latency inducing plug-ins on your output, causing the two signals to get out of phase with each-other.

 

Hope that makes sense.

It would if it were a different question.

In this case, it is simply EZ Drummer and a Send insert. The send is immediately bypassed when LLM is activated.

It would if it were a different question.

Well, I don't know how to put it in any other way than that it is a safety measure for more complex routing.

It has nothing to do with EZ drummer, any software instrument would do.

You can always use Low Latency Safe mode if you want it to pass audio.

I never noticed this before. I use PDC and LLM regularly.

As for the Software Instrument tracks, nothing will happen until after play has been activated. After that point, the send(s) will always be muted during LLM (unless LL safe mode is used). This will include the Sends used on the Aux tracks too. If the Software Instrument is removed, the Send insert is unaffected.

Sends that are present on AUX channel strips cannot be made Low Latency Safe.

"Each digital process—plug-in processing, changing the volume or pan level, and so on—adds an amount of latency (a small delay). Each of these processing latency values is added to each other." ~ Logic Manual

Strange.

I never noticed this before. I use PDC and LLM regularly.

 

As for the Software Instrument tracks, nothing will happen until after play has been activated. After that point, the send(s) will always be muted during LLM (unless LL safe mode is used).

Yes! The signal path gets "pinged" when using the play command.

Select a second channel strip (make one shiv if you don't have one ), press play and then select your first channel strip.

You will now have to press play again for it to register LLM.

This will include the Sends used on the Aux tracks too. If the Software Instrument is removed, the Send insert is unaffected.

Sends that are present on AUX channel strips cannot be made Low Latency Safe.

I think that the orange coloring of Sends on an AUX is a small graphical bug.

I can easily hear that those sends are passing audio and I think that's why the do not offer the LLS function.

"Each digital process—plug-in processing, changing the volume or pan level, and so on—adds an amount of latency (a small delay). Each of these processing latency values is added to each other." ~ Logic Manual

 

Strange.

But some of that process will occur within the buffer range so you won't experience more delay than the buffer time.

I never noticed this before. I use PDC and LLM regularly.

 

As for the Software Instrument tracks, nothing will happen until after play has been activated. After that point, the send(s) will always be muted during LLM (unless LL safe mode is used).

 

Yes! The signal path gets "pinged" when using the play command.

Select a second channel strip (make one shiv if you don't have one ), press play and then select your first channel strip.

You will now have to press play again for it to register LLM.

 

This will include the Sends used on the Aux tracks too. If the Software Instrument is removed, the Send insert is unaffected.

Sends that are present on AUX channel strips cannot be made Low Latency Safe.

 

I think that the orange coloring of Sends on an AUX is a small graphical bug.

I can easily hear that those sends are passing audio and I think that's why the do not offer the LLS function.

 

"Each digital process—plug-in processing, changing the volume or pan level, and so on—adds an amount of latency (a small delay). Each of these processing latency values is added to each other." ~ Logic Manual

 

Strange.

 

But some of that process will occur within the buffer range so you won't experience more delay than the buffer time.

So for this specific discussion about the send insert, it is only the selected track that turns orange.

I agree that it is a bug.

So the null test is void if a parameter on one strip is different than the other. It seems to be better suited for phasing issues and little to do with latency. However, I do agree that it is the use of plug ins (and buffer settings) that create the latency - not the channel strip itself.

So for this specific discussion about the send insert, it is only the selected track that turns orange.

I agree that it is a bug.

I've never said that?

I think that the orange coloring of Sends on an AUX is a small graphical bug.

I can easily hear that those sends are passing audio and I think that's why the do not offer the LLS function.