November 12, 2016 at 12:58 am #65005hansParticipant
Since DPS only supports stereo we have to downmix multi-channel content, e.g. movies.
As far as I understand 5.1 material becomes therefore about 8dB quieter and 7.1 stuff even 10dB.
So I was wondering what input level DPS is calibrated to? I guess handling native stereo material the same as multi-channel one doesn’t work that well?
Then there’s the problem with voices (center channel) and surrounds being to quiet when downmixed. Some media players give you the possibility to change mixing levels yourself. Usually center and surround channels are at -3dB.
Are there any better recommendations?November 16, 2016 at 11:16 pm #65010xaqmusicKeymaster
Here is a short answer made way too long 🙂
I imagine the final level of a downmix from 5.1 or 7.1 can vary depending on the downmix algorithm. The normal method for summing signals (when no dynamic range control is used) is SUM = ((A*0.5)+(B*0.5)) so each signal is attenuated by 6dB before summing just in case both signals are the same and they result in a potentialy clipping +6db. But in 5.1 or 7.1, not all channels should be treated equally.
Summing the center channel with FL/FR usually requires some attenuation as you mentioned because you are essentially copying a duplicate of the center signal to both left and right channels. This has the effect of an increase in level for the center channel (relative to L/R) because this mono signal will sum a bit in the atmosphere on its way to your ears. However, the perceived loudness will vary depending on frequency content.
Summing the front with rear/side channels may be handled differently by a good algorithm because ProLogic encoding or HRTF filtering may be applied to give some spacial sense to the surround signals. There is less reason for total signal attenuation in this case.
So when you say there is a 8-10dB drop in level after downmix, what are we comparing? SPL in the room? Peak level of each 5.1/7/1 channel to peak level of the downmix? You can see there is some ambiguity here.
Fortunately DPS can handle all of that! In Movie mode, most profiles have a -20 or lower threshold for the Bongiovi AGC. This means it will be sensitive to a very wide range of content. The output gain is also set accordingly so there will be no net loss of volume. You may also calibrate the input level (thus increasing sensitivity) by increasing the volume for the Digital Power Station driver and lowering the volume of your computers output device. DPS will respond to the perceived loudness of the material so you can get away with boosting the center channel if desired. Experiment with this and you will see/hear.
Also, I recommend using the movie’s LtRt (stereo) mix instead of downmixing if possible. This LtRt mix has been released from the studio to ensure good dialog levels on most stereo systems.
- You must be logged in to reply to this topic.