• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Can anyone MEASURE the multichannels in a loud movie?

Head_Unit

Major Contributor
Forum Donor
Joined
Aug 27, 2018
Messages
1,341
Likes
688
The idea keeps coming up that surround channels need "less power" than the front LCR, because
(1) The average power is less --> I say this is irrelevant because it is actually the PEAK levels that clip the amps
(2) All the channels never peak at the same time. Many say no, some including me say they surely do sometimes. Never any EVIDENCE except for one movie mixer who once posted yes they surely do.

Is anyone set up to be able to MEASURE this for scenes like when the semi is crashing everything along a street in Terminator 2 and similar?!? It would be wonderful to have actual DATA posted.
 
OP
H

Head_Unit

Major Contributor
Forum Donor
Joined
Aug 27, 2018
Messages
1,341
Likes
688
I don't have anything ripped here, nor any disc player connected to my Macs. OK full discosure I do have an old iMac with a disc drive. I just don't want to spend the time downloading and setting up and figuring out, too many other things going on with work and house painting. I assume/hope others are already set up to do this. I'll look at data-bass :)
 

3ll3d00d

Active Member
Joined
Aug 31, 2019
Messages
207
Likes
173
an example to illustrate, the gas station attack scene in terminator salvation. Average signal level across the scene. It's how things commonly look (C louder than LR which are louder than surrounds).

1661761531853.png
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,053
an example to illustrate, the gas station attack scene in terminator salvation. Average signal level across the scene. It's how things commonly look (C louder than LR which are louder than surrounds).

View attachment 227366
That looks to be about what people have described. Center louder by a bit, and surround abouts 10 db lower than fronts.
 

KMO

Addicted to Fun and Learning
Joined
Aug 9, 2021
Messages
629
Likes
901
In terms of channel calibration, surround channels are 3dB down in the cinema environment.

So if that was preserved in the transfer to home presentation, the maximum possible surround level would 3dB lower than mains, as the surround tracks would have been lowered 3dB in the transfer. Same any noise burst that was "equally in all channels" in the cinema mix.

You see something like that in the graph above.

(To be more precise, home calibration puts all the channels at equal level, usually aiming at reference being 85dB = -18dBFS, peak 103dB. In a cinema transfer you wouldn't see more than 100dB in surrounds, due to them being calibrated as 82dB = -18dBFS in the original track.)
 

3ll3d00d

Active Member
Joined
Aug 31, 2019
Messages
207
Likes
173
one more example which covers the entire soundtrack of No Time to Die, same pattern though bigger gap between channels in this one (and this different holds if you look at peak levels rather than average btw so it's not just that it's on average lower, it's consistently lower).

1661768923059.png
 

Cbdb2

Major Contributor
Joined
Sep 8, 2019
Messages
1,530
Likes
1,485
Location
Vancouver
Try the movie U-571. Theres depth charge explos that reach 0dbfs (max level) on all 5.1 channels. We watched it in the mix theatre and had to check levels with the meters because is was so loud.
 

3ll3d00d

Active Member
Joined
Aug 31, 2019
Messages
207
Likes
173
the 1st depth charge scene in that film still has C louder than the others

1661803832253.png


looking at the waveform, there's still visually less going on in the surrounds

L

1661803923595.png


SL

1661803940437.png


the point at which SL hits almost full scale is literally a single sample, once you zoom in you can see it's quite possibly literally a single sample that spikes (so probably some mostly irrelevant v high frequency content). Look at the 1st bit that goes to -1

1661804111741.png


then zoom in even more

1661804207553.png


Ultimately, if designing/sizing a system, then IMV it's fair to assume the surrounds can be less capable than the mains in terms of raw output
 

Cbdb2

Major Contributor
Joined
Sep 8, 2019
Messages
1,530
Likes
1,485
Location
Vancouver
the 1st depth charge scene in that film still has C louder than the others

View attachment 227480

looking at the waveform, there's still visually less going on in the surrounds

L

View attachment 227481

SL

View attachment 227482

the point at which SL hits almost full scale is literally a single sample, once you zoom in you can see it's quite possibly literally a single sample that spikes (so probably some mostly irrelevant v high frequency content). Look at the 1st bit that goes to -1

View attachment 227483

then zoom in even more

View attachment 227484

Ultimately, if designing/sizing a system, then IMV it's fair to assume the surrounds can be less capable than the mains in terms of raw output
For HT systems probably, don't know if that's true for theatres with rows of surrounds. Thats what there mixed for and where there mixed. And some directors want it ass loud as possible and like your in the middle of it. All I can say is I saw the meters. ( https://www.dorrough.com/vision.html )
 

3ll3d00d

Active Member
Joined
Aug 31, 2019
Messages
207
Likes
173
All I can say is I saw the meters
The link is a load of marketing talk so it's hard to tell exactly what it is measuring or where in the signal chain it sits. The raw content on the disc is quite unambiguous though the marketing talk sounds like it is doing something with the same intent as my app does.

For reference, the previous graphs referring to peak or average levels are computed by https://github.com/3ll3d00d/beqdesigner/blob/master/src/main/python/model/signal.py#L930 so it's basically, for average, it uses https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.welch.html (hence is an estimate of power spectral density) and, for peak, it is essentially the max per bin of all the slices in a spectrogram.

In any case, that section of the track is an example of one where the surrounds are relatively loud. I am sure there are other such scenes out there so if sizing a system for the absolute worst case then having big/capable surrounds is the way to go if it's practical to do so (space/budget allows).
 
Last edited:

KMO

Addicted to Fun and Learning
Joined
Aug 9, 2021
Messages
629
Likes
901
All I can say is I saw the meters. ( https://www.dorrough.com/vision.html )
But as I noted above, the scale for the meters in a cinema mix, which I think you were suggesting you were monitoring, is not equal. 0dBFS on a surround channel is 3dB SPL quieter than 0dBFS on the fronts.

When transferred to an equal-calibration-level home mix, which is what 3ll3d00d is dissecting, you would expect a reduced surround waveform amplitude on lots of things.

I can quite imagine quite often the original cinema mixer would just do "pop this in all channels, equal meter level" for a general quick and easy thing to do for an "all round" effect, and because of that calibration difference, it will be 3dB SPL quieter in surrounds, and that will actually be fine - the "equal meter level" is slightly front-focussed.

The graphs in that post are roughly in that ballpark - that central spike at 1:35 being 0.6 rather than 0.8 - that's 2.5dB lower.

So, cinema:

-18dBFS = 85dB SPL (fronts)
-18dBFS = 82dB SPL (surrounds)

Home:

-18dBFS = 85dB SPL (all channels)
-21dBFS = 82dB SPL (all channels)

Something "-18dBFS all channels" in the original mix will be 85dB/82dB SPL. That then becomes "-18dBFS/-21dBFS" in the home mix to preserve the 85dB/82dB SPL intent.

(Actually, I say "intent" - that 3dB difference may well not have been a conscious intent, but rather just something arising "automatically" from the calibration. But still, it's what the original mixer heard and was happy with, and what cinema audiences heard.)
 
Last edited:
Top Bottom