Start a new topic

P.S Loudness Variations between ATMOS and 2.0

Evening all. 

I have a question regarding the loudness metering in PS. Is anyone getting 1.5 - 2.0 db differences in level when looking between the ATMOS and Stereo Integrated Dialogue measurements?


For example. Feeding a full Atmos mix to the renderer software shows the atmos mix has a dial average of -28.0 whereas the stereo is at -26.2 


However, when you look at the BS-1770-3 LKFS Reading.... The atmos mix is at -25.1 and the stereo is at -24.9. (I would expect around 0.2 LKFS of difference between the two mixes)  Essentially meaning the 5.1 and 2.0 mixes are very close in average loudness. 


This 2.0 different must be something to do with the fold-down process in the renderer I presume? 


I have tried meausring my re render printed master in Pro Tools and Re-Render exports from the DAMF and the same dial average loudness difference issue applies...


Im sure Im missing something obvious! Anyone shed any light! 


Cheers

Andy


Hi Andy, 


Differences in the speech integrated measurement  between Atmos and Stereo can occur and are program dependent.   The speech measurement uses the Dolby Dialog Intelligence algorithm to detect speech and only measures the portion of the content where speech is detected.    1770-3 is an integrated measurement of the entire file and generally meters louder as is the case here.   


When performing loudness measurement of the Atmos mix (in reality a 5.1 re-render with a limiter implementation) the existing speech is often in the clear in the center channel and/or doesn't have as many other sounds mixed with it and the metering is performed on all channels.    With a stereo re-render the speech is in the left and right along with everything else so depending on what else is going on in the mix during the speech detected portions of the content, the stereo can meter a bit louder.   In this case 1.8LKFS louder vs .2LKFS louder for the integrated measurement.    This isn't entirely unexpected and again is program dependent.   


If there is a lot of content in the rears and overheard during portions of speech you might be able to get this measurement more aligned if you tweak the Trim controls manually but this will affect the entire stereo mix which may or may not be desirable.


Given how close the integrated measurement of the Atmos and Stereo are I don't know if that makes sense to pursue.


If you are delivering a stereo re-render separately and it is outside the delivery specification you are working with you might need to attenuate the file in your DAW.


Hope this helps.


Best,

Adam


1 person likes this
Adam, that is a perfect reply. Thank you very much for your prompt response, greatly appreciated. I'll worry a little less about it for now i think! Best
Login to post a comment