Would Appreciate An Official Word on How Dual Buffering is Handled

[musicman691]

Gentleman:
Please don't feed the troll anymore. Just let him roll up his own bad karma.

[darbyclash34]

Yes please, let's get back to the topic at hand, and not get to far astray, as that's just going to make it less likely for this question to finally get an official, legitimate explanation from the horses mouth.

Originally Posted by Darryl Ramm

As a user you only get to control the size of the input buffer. What Avid calls the "hardware buffer". Same place as it has always been in the playback engine settings, except now the other buffer (the playback buffer) is of fixed (large) size you can't control.

You set the hardware buffer lower if you need lower recording latency.

You set the hardware buffer higher if you need more stability/to run high processing loads.

If you can't find a hardware buffer size that works for you then you need to print tracks/simply the session, change what your are trying to do e.g. use outboard monitoring hardware, upgrade/change your computer hardware etc. Do whatever/all is needed that you can.

So this is again exactly why this question needs answering so bad. So in the above quoted text, tell me again exactly why does increasing the INPUT buffer setting, which is the buffer allegedly only used on tracks receiving live INPUT, decrease the CPU utilization and can alleviate some CPU overload issues when mixing with nothing record enabled? Am I doing something wrong that's causing all my tracks to use the input buffer instead of the playback buffer? This occurs on every session, and occurs on both my main spec'd out Hackintosh rig running 10.9.5, and an i7 MacBook Air with 8gbs of RAM. If I'm mixing, no recording whatsoever, shouldn't everything be using the playback buffer? So altering the input buffer should have no effect, or at least a negligible one, on my CPU resources.

I have the following two theories as to what may be going on (in both scenarios all tracks are not record or input enabled, and the session consists of audio tracks, aux inputs, and master faders):

The input coming into a submix or aux return, etc is inherently not recorded playback. It's receiving signal from recorded tracks, but what if I mute a track suddenly, or alter the signal it's getting? If there was any type of background rendering of this internal bus, this change would have a noticeable delay as it either rerenders the file, or it would have to resume processing while doing a background bounce of that aux inputs signal. Regardless, I'd say most examples of aux input usage I can think of sound like they meet the criteria for the input buffer, which the manual lists as " Aux input with live input" which is tempting to say means an aux with a live instrument feeding it, but in one of the above referenced posts from one of the Audio Engine developers, they mention thinking about it as a disk playback path and a live input path. Since there are no files to playback for an aux input, doesn't that mean that they must use the buffer set in the playback engine?

Or, does your Punch mode dictate how this operates? I just got this thought while typing this out, but if you have any type of punch record enabled are all your audio tracks always using the input buffer? I'm going to do some experimenting with this and report back.

But I do think it's sad how little we ever hear from the higher ups anymore. The mods are great, but they're not Avid employees, while they are valuable resources and can be of great help, they're not the people who built the thing. Those types of people rarely if ever chime in anymore. Where are the Audio Engine developers who were involved in the previous threads? I would love to hear from them.

Thanks,

Shawn

[Darryl Ramm]

The questions I answered was not explain what cause the issues/confusion the thread was originally asking about. I was not sure they were even a serious question or more just being rhetorical. But this thread was way off base with lots of unnecessary incorrect information flying around. So those answers are just what can you practically do. And of course you can also check your version of Pro Tools for problems or look for problem plugins and usual debugging for things might be causing CPU errors.

But yes I think it would be great for Avid to explain what is being asked here, aux behaviors etc. And while most Pro Tools documentation is of a high quality this is something where maybe a little better clarity (as this thread has shown) and a little more detail in the Reference Guide would have helped.

[JFreak]

I wouldn't mind if Avid paid us, but no, at least I haven't seen a single dollar coming from them.

[Niilo Nuori]

Originally Posted by Darryl Ramm

But yes I think it would be great for Avid to explain what is being asked here, aux behaviors etc. And while most Pro Tools documentation is of a high quality this is something where maybe a little better clarity (as this thread has shown) and a little more detail in the Reference Guide would have helped.

Yes and many of us with technical backgrounds would even prefer a detailed technical answer (with proper flow charts )

And let people ask questions, even stupid ones (ignore them if you have to) and stop blaming somebody for not being skilled enough with PT to know everything. Peace!

[Amack]

How large is that "large playback buffer"?

Originally Posted by Darryl Ramm

As a user you only get to control the size of the input buffer. What Avid calls the "hardware buffer". Same place as it has always been in the playback engine settings, except now the other buffer (the playback buffer) is of fixed (large) size you can't control.

You set the hardware buffer lower if you need lower recording latency.

You set the hardware buffer higher if you need more stability/to run high processing loads.

If you can't find a hardware buffer size that works for you then you need to print tracks/simply the session, change what your are trying to do e.g. use outboard monitoring hardware, upgrade/change your computer hardware etc. Do whatever/all is needed that you can.

[Darryl Ramm]

This has been covered multiple times on other threads on DUC, including those you have been pointed at. Can you just leave other users threads alone and let them get answers to what they want.

[Amack]

First I've heard that one!?!? Can you please point to a post and/or documentation that supports your assertion that "(the playback buffer) is of fixed (large) size you can't control."

Thanks!

Originally Posted by Darryl Ramm

This has been covered multiple times on other threads on DUC, including those you have been pointed at. Can you just leave other users threads alone and let them get answers to what they want.

[Shan]

http://duc.avid.com/showpost.php?p=2...&postcount=157

http://duc.avid.com/showpost.php?p=2105862&postcount=9

Shane

[DC-Choppah]

From the reference guide:

When Delay Compensation is enabled, Pro Tools allocates 16,383 samples at 44.1/48 kHz, 32,767 samples at 88.2/96 kHz, or 65,534 samples at 176.4/192 kHz of Delay Compensation for each mixer channel. For sessions with plug-ins on mixer channels that result in more the 4,000 samples of delay at 48 kHz, select this setting.

To maintain phase coherent time alignment, Delay compensation should always be enabled during playback and mixing. Delay Compensation should also be used in most recording situations.



It is my understanding that when ADC is activated, we now have a big buffer for playback so PT can keep up with all the plugins in real time as you record. The music is just delayed enough so that it has time to do what it needs on the playback side. Meanwhile the record buffer is short depending on your setting for buffer size and if low latency monitoring is on or off.

AUX tracks do not automatically go into low latency mode, while record-enabled audio tracks do. So that may be the difference between aux tracks and audio tracks. It confused me for a while until I wrapped my brain around it.