Analog hardware latency

[Peters86]

Hello guys,

I guess this question has been asked a thousand times before but I can't seem to find a definitive answer!

I use PT 12 Native on OSX with an Apogee Ensemble Firewire for recording and mixing. I have a Warm Audio WA2A and EQP-WA I use for mixing, I don't use plugins for this. Now, when mixing, I send the track signal through output 5 --> compressor --> EQ --> input 5 by using the i/o plugin on the track. I then Bounce the track in realtime to a new track and click "hide and make inactive" on the existing unprocessed track.

My question: do I have to compensate for latency when doing this? I use a relatively low buffer setting (64 samples) and I never hear any phase issues or noticeable delays of the signal when doing this (for instance when sending out and mixing seperate drum tracks). But something tells me a latency/delay is induced by routing the signal like this and that I should compensate for it.

Thank you guys very much.

[YYR123]

What does it look like when compared side by side to the original ?

Before you hide and make inactive look at the wave forms.....

I bet the delayed samples are minuscule but you should check it.

[Drew Mazurek]

Regular PT (unlike HD with Avid's hardware) has no idea about the delays induced by 3rd party converters.

It's easy to test and plug in the delay amount in the I/O setup.

[Carl Kolchak]

There will be inherent latency in the Apogee, as it takes a small amount of time to convert the digital audio to analogue on the way out, and to convert the analogue signal to digital on the way back in (usually a different amount of time for each process).

You can measure the total time by drawing a 1 sample spike on an audio track, and recording it to another track by sending the signal out of one set of hardware outputs, and back in to another set of inputs - physically connecting them with patch leads.

With Pro Tools in slip mode, and the timeline set to samples, select from the start of the 1 sample spike on the source track, to the start of the 1 sample spike on the record track.

The counter will display the total number of samples selected, which is the total latency for combined input & output.

Now repeat the experiment with your external hardware units in the chain.

You can then input the figure in to the nudge value, and whenever you record a hardware insert, select the recording, and nudge it back.

You can also look at the I/O setup page, and once you have zeroed the value in, put it in the hardware compensation field, to have PT do it automatically.


Sent from my iPhone using Tapatalk

[john1192]

Originally Posted by Carl Kolchak

There will be inherent latency in the Apogee, as it takes a small amount of time to convert the digital audio to analogue on the way out, and to convert the analogue signal to digital on the way back in (usually a different amount of time for each process).

You can measure the total time by drawing a 1 sample spike on an audio track, and recording it to another track by sending the signal out of one set of hardware outputs, and back in to another set of inputs - physically connecting them with patch leads.

With Pro Tools in slip mode, and the timeline set to samples, select from the start of the 1 sample spike on the source track, to the start of the 1 sample spike on the record track.

The counter will display the total number of samples selected, which is the total latency for combined input & output.

Now repeat the experiment with your external hardware units in the chain.

You can then input the figure in to the nudge value, and whenever you record a hardware insert, select the recording, and nudge it back.

You can also look at the I/O setup page, and once you have zeroed the value in, put it in the hardware compensation field, to have PT do it automatically.


Sent from my iPhone using Tapatalk

just to add to carls great description ..i use Grid Mode in Samples View set to 1 sample .. it will not go any lower than this anyway ... easier to count the samples .. and set as carl says to sample in the main timeline ..

i go through this on every new version of pro tools .. seems like things change each time, so do not be afraid to check each time .. set up a session ready to go and test .. !!! and i do it for each sample rate ... 44,48,88,96, etc ..

cheers john

[Peters86]

Thanks a lot guys. That was the answer I was looking for. Perfect!!

[chrismeraz]

Of course, it's perfectly correct that you need to figure out the latency for a round trip through the converters.

I would add that you don't need to bother trying to figure out the latency for a hardware unit like an EQ or a compressor. You need to compensate only for the converters, so once you figure out that latency, you can use the same value for all your other outboard hardware.

[Carl Kolchak]

Indeed, provided it's an entirely analogue device - if it's a multi-effects unit, or digital delay etc, there's going to be an ADC / DAC round trip in the device, in addition to the Apogee (or any) converter.


Sent from my iPhone using Tapatalk

[ChuckS]

Originally Posted by Carl Kolchak

Indeed, provided it's an entirely analogue device - if it's a multi-effects unit, or digital delay etc, there's going to be an ADC / DAC round trip in the device, in addition to the Apogee (or any) converter.

This delay is not compensated for in analog recording chains. Chalk up any delay in the effect unit itself as the device's 'character'. When I calculate converter delay I use a loopback rather than having a device inline.

[Carl Kolchak]

True, but there are all sorts of things we couldn't do in the analogue days, that we now can in the digital age, so it makes sense to take advantage of improvements.


Sent from my iPhone using Tapatalk