Question about master bus levels

[paul_g]

I know this topic is well worn and quite old but I had a question based upon observations.

I started in PT 8.0 back in the day. When I tracked (recorded) my sessions, I always set input levels around -12db (using the older Pro Tools meters). Almost everything I do is based upon knowing down the chain there will be a summing up until the master fader. I've recorded countless sessions, large sessions, full bands, (13 drum mics / upright bass / guitars / b3's etc) and my levels were all -12 or less. I consider anything -6db a peak I don't want.

When I compress, I do so conservatively. I am old school and have never participated in the loudness wars of the early 2000's. My master tracks on these sessions at peak were -.02 but never bumped over 0db.

Since moving off of PT 8, 9, and 10, I record with the same conservative approach. Watching my input levels extremely carefully, never going above -15 to -12 db. No single track ever goes higher than -6db after EQ and Compression.

Yet, since moving onto the 64bit platform of Pro Tools 20xx, master track on my sessions are hitting +2db, sometimes +3db, yet no single track is peaking above -6db

I can open up old sessions from 13 years ago that are much "louder" and those are rock steady below 0db on the master fader.

Is there something that has changed in Pro Tools that increases the signal down the chain to the master bus? I feel like I could record at -20db and still end up with issues. I'm so baffled by this.

Keep in mind, my old setup was a 003 / Persons DigiMax FS, and now my rig as of 2020 is entirely new. (see sig block for details).

To counter this, I started using VCA's and pulling large sections of the mix down. This helps somewhat, and I read that as long as no single track is peaking, I'm ok to just bring the master down, but I doubt the advice as if it's red when it's hitting the master, turning something that is red down, IMO just means, red at lower volume.

I really think I'm missing something here and I'm at an utter loss. I realize this is an impossible question to answer, I guess the question is, has there been structural differences between PT 10 and PT 20xx enough to sum signals at higher levels at the master bus?

[Darryl Ramm]

First thing is simply how many tracks are you summing together? besides a little phase cancelling all these signals eventually add up.

And what exact metering are you looking at? The metering change a few years ago to default to sample peak makes the same levels look hotter than before...but may be a better meter to use.

And it's also easy to slip up and confuse pre- and post-fader metering levels.

[paul_g]

Originally Posted by Darryl Ramm

First thing is simply how many tracks are you summing together? besides a little phase cancelling all these signals eventually add up.

And what exact metering are you looking at? The metering change a few years ago to default to sample peak makes the same levels look hotter than before...but may be a better meter to use.

And it's also easy to slip up and confuse pre- and post-fader metering levels.

Hi Darryl,

The current project I'm working on has 44 tracks, and that doesn't include the aux sends with effects on them. But again, mostly everything is recorded between -19 and -12 db.


Metering I use Pro Tools Classic and Linear on the Master fader. Like I said, I'm not slamming the master all the time, just hitting peaks maybe 3 times in a 4 minute song.

I have an idea where some of my peaks are coming from now that I've had time to really analyze it. I'll need more time to figure it out, but I have a sneaking suspicion it's the virtual drums. That's really the big thing that changed. Going from a couple of great session drummers, awesome mic's and really getting good levels, to assembled VI drums where different fills are at different levels. I'll have to do more research.

[Sardi]

Originally Posted by paul_g

I read that as long as no single track is peaking, I'm ok to just bring the master down, but I doubt the advice as if it's red when it's hitting the master, turning something that is red down, IMO just means, red at lower volume.

Don’t doubt it. It’s correct.

I advise you do some reading on floating point architecture to understand how this works in audio.


Sent from my iPhone using Tapatalk

[Stephen Bond]

I am old school too. Nothing over, channel, buss etc.

Floating point is good behind the scenes with the new bit depths. You can go over and just pull Aux / Master down and not be clipping. Just reset the metering and double check after attenuation.

Stephen

[Darryl Ramm]

It’s just math… like Sardi suggested it’s maybe a good idea to understand the basics to know how this all work

With 64-bit float mixing and 32-bit floating point files there is no internal clipping at 0dBFS and colors don’t really now mean anything on internal busses.

Make sure you are using 32-bit floating point files for maximum flexibility.

You still have to watch levels going into plugins.

Try deliberately run a mix hot and convince yourself you can just pull down a fader OK and this all works. Yes watching stuff punch into the red can be discerning at first.

[paul_g]

Originally Posted by Darryl Ramm

It’s just math… like Sardi suggested it’s maybe a good idea to understand the basics to know how this all work

With 64-bit float mixing and 32-bit floating point files there is no internal clipping at 0dBFS and colors don’t really now mean anything on internal busses.

Make sure you are using 32-bit floating point files for maximum flexibility.

You still have to watch levels going into plugins.

Try deliberately run a mix hot and convince yourself you can just pull down a fader OK and this all works. Yes watching stuff punch into the red can be discerning at first.

So what you are saying is, I should go to the menu bar, select Setup > Session > And change my 24bit session to 32Bit Float?

Can I make this change after everything is tracked and my mix is complete?

Yeah, I will have to spend some time learning this....

Cheers

[Darryl Ramm]

Originally Posted by paul_g

So what you are saying is, I should go to the menu bar, select Setup > Session > And change my 24bit session to 32Bit Float?

Can I make this change after everything is tracked and my mix is complete?

You can make that change. But it won't make any difference to the existing mix that setting only affects new files that are created. Existing files will stay as they are and if they are clipped at 0dBFS they'll stay clipped. If you do set the session to use 32-bit float files and then re-record stuff in the session then that new/updated content will be 32-bit float... and that won't clip at 0dBFS like fixed point files will.

In contrast if you do a Save Session Copy In and choose to convert files then you can convert every file to 32-bit float. But again that won't help you if the files were already clipping... and it can cause confusion because now you have all 32-bit float files but some might be clipped at 0dBFS because they were fixed point before, and some might not be clipped at 0dBFS. If you do that keep the original session around if you need to look at what was there before. And it can just be more obvious to change session new file format like you mentioned, than to do a Save Session Copy In.

You can play around with all this, looking at the Workspace window to see the file bit-depths/formats.

But there is really just no reason not to work at 32-bit float from the start, it's a slight increase in disk size, and SSDs are low-cost and fast.

Yeah, I will have to spend some time learning this....

Yep, and it will be worthwhile.

[paul_g]

Originally Posted by Darryl Ramm

You can make that change. But it won't make any difference to the existing mix that setting only affects new files that are created. Existing files will stay as they are and if they are clipped at 0dBFS they'll stay clipped. If you do set the session to use 32-bit float files and then re-record stuff in the session then that new/updated content will be 32-bit float... and that won't clip at 0dBFS like fixed point files will.

In contrast if you do a Save Session Copy In and choose to convert files then you can convert every file to 32-bit float. But again that won't help you if the files were already clipping... and it can cause confusion because now you have all 32-bit float files but some might be clipped at 0dBFS because they were fixed point before, and some might not be clipped at 0dBFS. If you do that keep the original session around if you need to look at what was there before. And it can just be more obvious to change session new file format like you mentioned, than to do a Save Session Copy In.

You can play around with all this, looking at the Workspace window to see the file bit-depths/formats.

But there is really just no reason not to work at 32-bit float from the start, it's a slight increase in disk size, and SSDs are low-cost and fast.



Yep, and it will be worthwhile.

Thank you for all this. Yes, I'm fortunate because none of the audio tracks are clipping, even after eq and compression they're still way below -6db, in pre-fader mode, and even less in post fader mode. What I'm dealing with is exactly as you first stated, and that is everything is accumulating up. to the master fader.

I will try a save copy in. Right now I have nothing clipping and nothing is beyond zero db at the master, but even then that is still too hot for my old school liking.

Thanks again Darryl!

[Rich Breen]

Originally Posted by paul_g

...Is there something that has changed in Pro Tools that increases the signal down the chain to the master bus?...

All the info about 32 vs 24 is fine, but to answer the original question, the math in the bus summing blocks is improved from older versions but wouldn't change overall gain in any way - 0 is still 0. Do you perhaps have HEAT enabled on your tracks now and didn't before? That would definitely account for more gain.