Avid Pro Audio Community

Avid Pro Audio Community

How to Join & Post  •  Community Terms of Use  •  Help Us Help You

Knowledge Base Search  •  Community Search  •  Learn & Support


Avid Home Page

Go Back   Avid Pro Audio Community > Pro Tools Post Production > Post - Surround - Video

Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 02-18-2021, 11:56 AM
Nick Camaron Nick Camaron is offline
Member
 
Join Date: Feb 2009
Location: Somewhere in the vicinity of Alpha Centauri, MW
Posts: 155
Default Dolby Atmos Top Front Content: Object vs. Bed=huge difference in Mixdown.

Hey dear Atmos buffs,

while setting up a 7.1.4 Atmos Studio with production suite I discovered a mixdown oddity which apparently can’t be affected with neither setting of the “trim and downmix controls”.

Here we go:

A signal panned to (+/-)60,60, 60, 100 = (TopFront R or L) is mixed down vastly different depending on being “object” or “7.1.2 bed”.
In essence, when I toggle this exact panning between Object/Bus in Protools and the monitoring in the Renderer is in “Stereo downmix”, the object mode results in about double volume than Bus mode!
This happens only with the top fronts. All other speakers mix down as expected.

When I dig deeper into it I see the first difference appearing in the 7.1.4/7.1.2 to 7.1 downmix/monitoring where Objects in this position are fed partially into R,C,Ls,Rs whereas Bed content in this position is exclusively downmixed to Rs. The Object pattern goes down to the 5.1 and then adding up in overly loud reproduction in the Stereo.
It gets even weirder! When I do the same experiment with a 7.1 Bed (instead of 7.1.4 Bed) the contrary happens. The object is mixed down lower than the bed! Less difference but still considerable. What the heck?
As I said, no setting in the downmix controls affects this being so.

To compare I set up my own straight downmix matrix for quick reference with the awesome RME total mix matrix and these DIY downmixes behave exactly as I would expect. No overblown objects in the stereo.

What's going on here?
Top Front Objects mix down into Stereo a lot louder than 7.1.2 beds?
Top Front Objects mix down into Stereo lower than 7.1 beds?
So different that it isn’t the same mix at all.

I´m sorry for this complicated matter but can anyone confirm this?
Do I have my wires crossed or is this what the Renderer does?


Thanks so much for help

Nick

Mojave, 6 CoreTrashcan
PT 2020.12 Ultimate native
Dolby PS Renderer 3.5, Dolby Audio Bridge In, Aggregate Out (Fireface,DAB)


Edit: I just could exactly reproduce this on a different computer

Last edited by Nick Camaron; 02-18-2021 at 03:32 PM. Reason: Sys info added
Reply With Quote
  #2  
Old 02-18-2021, 02:36 PM
gives's Avatar
gives gives is offline
Member
 
Join Date: Feb 2001
Location: Portland, OR
Posts: 1,350
Default Re: Dolby Atmos Top Front Content: Object vs. Bed=huge difference in Mixdown.

Quote:
Originally Posted by Nick Camaron View Post
Hey dear Atmos buffs,

while setting up a 7.1.4 Atmos Studio with production suite I discovered a mixdown oddity which apparently can’t be affected with neither setting of the “trim and downmix controls”.

Here we go:

A signal panned to (+/-)60,60, 60, 100 = (TopFront R or L) is mixed down vastly different depending on being “object” or “7.1.2 bed”.
In essence, when I toggle this exact panning between Object/Bus in Protools and the monitoring in the Renderer is in “Stereo downmix”, the object mode results in about double volume than Bus mode!
This happens only with the top fronts. All other speakers mix down as expected.

When I dig deeper into it I see the first difference appearing in the 7.1.4/7.1.2 to 7.1 downmix/monitoring where Objects in this position are fed partially into R,C,Ls,Rs whereas Bed content in this position is exclusively downmixed to Rs. The Object pattern goes down to the 5.1 and then adding up in overly loud reproduction in the Stereo.
It gets even weirder! When I do the same experiment with a 7.1 Bed (instead of 7.1.4 Bed) the contrary happens. The object is mixed down lower than the bed! Less difference but still considerable. What the heck?
As I said, no setting in the downmix controls affects this being so.

To compare I set up my own straight downmix matrix for quick reference with the awesome RME total mix matrix and these DIY downmixes behave exactly as I would expect. No overblown objects in the stereo.

What's going on here?
Top Front Objects mix down into Stereo a lot louder than 7.1.2 beds?
Top Front Objects mix down into Stereo lower than 7.1 beds?
So different that it isn’t the same mix at all.

I´m sorry for this complicated matter but can anyone confirm this?
Do I have my wires crossed or is this what the Renderer does?


Thanks so much for help

Nick
Without know anything about your system info etc. I would just contact Jeff Komar Avid Product Specialist . He trolls here quite often.

G
__________________
Gregory Ives, Composer/Sound Designer/Musician
http://www.gregives.com
https://pro-labs.imdb.com/name/nm1170521/?ref_=sch_int

Protools 2021.7 Ultimate HD Native Thunderbolt Avid MTRX. DAD MOM,Mac Pro 3.5GHz 6-Core Intel Xeon E5 64GB of Ram
OSX Catalina 10.15.7 Digital Performer 9.5.2, Logic Pro X 10.6.3
Reply With Quote
  #3  
Old 02-18-2021, 03:38 PM
Nick Camaron Nick Camaron is offline
Member
 
Join Date: Feb 2009
Location: Somewhere in the vicinity of Alpha Centauri, MW
Posts: 155
Default Re: Dolby Atmos Top Front Content: Object vs. Bed=huge difference in Mixdown.

Quote:
Originally Posted by gives View Post
Without know anything about your system info etc. I would just contact Jeff Komar Avid Product Specialist . He trolls here quite often.

G
Ok thanks. How to contact Jeff though? Hello Jeff, you hear me?
Reply With Quote
  #4  
Old 02-18-2021, 07:32 PM
Mixchump Mixchump is offline
Member
 
Join Date: May 2002
Location: Toronto, ON
Posts: 376
Default Re: Dolby Atmos Top Front Content: Object vs. Bed=huge difference in Mixdown.

OK, without knowing the specifics of what you're doing in your home-brew downmixer, I'm guessing that the discrepancy lies in the values in the front height channels possibly summing with signal already feeding the 7.1.2 bed?

I would love if somebody could explain why Avid and Dolby settled on a max bus size of 7.1.2, and not 7.1.4? I've taken to avoiding using the .2 in the bed, and creating a Quad bus that feeds 4 object paths for the .4

If that's what you're doing as well, then I may be facing the same downmix dilemma you're discovered.

Have you measured with a tone to calculate the exact level differences?

Is it possible that the pan-law for your height objects (if you're also using a Quad bus?) is set to one of the optional shapes?

Is it possible that you have some binaural processing going on that might be causing this?
Reply With Quote
  #5  
Old 02-19-2021, 04:29 AM
Nick Camaron Nick Camaron is offline
Member
 
Join Date: Feb 2009
Location: Somewhere in the vicinity of Alpha Centauri, MW
Posts: 155
Default Re: Dolby Atmos Top Front Content: Object vs. Bed=huge difference in Mixdown.

Quote:
Originally Posted by Mixchump View Post
OK, without knowing the specifics of what you're doing in your home-brew downmixer,
My home brew downmixer simply reflects the measured relative output values of each Renderer Output in the specific Monitoring mode (Made in LoRo Mode. It is basically a static rip off and seems to resemble quite well). But this is just a side show and not part of the problem.

Quote:
Originally Posted by Mixchump View Post
I'm guessing that the discrepancy lies in the values in the front height channels possibly summing with signal already feeding the 7.1.2 bed?
There are no double feeds as far as I can see, and no summing, because I having a solo signal.

Quote:
Originally Posted by Mixchump View Post
I would love if somebody could explain why Avid and Dolby settled on a max bus size of 7.1.2, and not 7.1.4? I've taken to avoiding using the .2 in the bed, and creating a Quad bus that feeds 4 object paths for the .4

If that's what you're doing as well, then I may be facing the same downmix dilemma you're discovered.

Have you measured with a tone to calculate the exact level differences?
Indeed. Further experiments this morning seem to blame the 7.1.2 bed concept and how ProTools (and DAR?) are handling the mixdown & routing.

7.1.2:
Rtf Objects switched to 7.1.2 bed are summed -3dB to Rtf+Rtr. These Rtf+Rtr are summed -4.5dB(?) to RS only in the 7.1.
This seems to result in a 6,3 dB dip in the Stereo Mixdown(compared to object mode)

7.1:
Rtf Objects switched to 7.1 bed are more or less adequately distributed to C,R,Ls,Rs resulting later in a 3dB boost in the Stereo Mixdown. (compared to object mode)

(Values read out on the short term Loudness measurement in DAR)

So these are my findings. Please don't nail me down on these since I´m just investigating and I might be confused.
But when I am right there is a 9,3dB!! difference for Rtf Objects toggled to Bus, depending on 7.1 or 7.1.2.

My conclusion then is to steer clear of 7.1.2 beds totally.
7.1 beds are not perfect either but seem to be close enough to get halfway corresponding mixdowns.

Quote:
Originally Posted by Mixchump View Post
Is it possible that the pan-law for your height objects (if you're also using a Quad bus?) is set to one of the optional shapes?
You mean height mode "free form or wedge" etc. in the panner?
It appears that in wedge mode the discrepance is alleviated due to automatic avoidance of extreme Tops position. When I get that right. But thats just a workaround I´m afraid.

Quote:
Originally Posted by Mixchump View Post
Is it possible that you have some binaural processing going on that might be causing this?
I´m not sure. Not that I know of.


Phew this is a difficult one and it is hard to phrase this correctly.
I need a coffee break now

cheers

Last edited by Nick Camaron; 02-19-2021 at 04:41 AM.
Reply With Quote
  #6  
Old 07-22-2021, 04:50 AM
Protone Protone is offline
Member
 
Join Date: Aug 2008
Posts: 51
Default Re: Dolby Atmos Top Front Content: Object vs. Bed=huge difference in Mixdown.

Quote:
Originally Posted by Mixchump View Post
...I've taken to avoiding using the .2 in the bed, and creating a Quad bus that feeds 4 object paths for the .4
Unrelated and probably a bit stupid question: how do you route quad bus to objects? I have only managed to route mono and stereo busses to objects.
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is On

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Atmos Object Automation Phasing problems kava Pro Tools 2020 2 12-01-2020 10:48 AM
Dolby Atmos Ceiling Speakers - Rear vs Front Ported Speakers antonis Post - Surround - Video 6 06-27-2019 09:01 AM
Dolby Atmos with S5 MC-Pro critictalk Post - Surround - Video 4 01-01-2018 08:21 AM
Dolby Atmos IrelandM Avid Pro Mixing General Discussion 2 07-17-2017 01:00 AM
Dolby Atmos peter5992 Post - Surround - Video 20 10-29-2012 01:51 PM


All times are GMT -7. The time now is 10:53 AM.


Powered by: vBulletin, Copyright ©2000 - 2008, Jelsoft Enterprises Limited. Forum Hosted By: URLJet.com