How did you arrive at these conclusions ? Explain the math to me in detail please.
1) All input signals contain same LF content (presumed faulty) ? Who cares, if the signal has equal amplitude and is in phase, then the assumption stands. There are mixes out there with coherent signals. Mixes out there with incoherent signals. And a mixture of both. Doesn't mean one can't have equal amplitude signals.
2) It is not impossible to have equal amplitude signals in all channels + LFE. Not impossible but perhaps improbable. Doesn't mean it can't or won't happen.
3) You are telling me that there WILL be a difference in peak output in a room between 5 individual full range channels + LFE vs redirected bass and LFE ? How in the world did you arrive at that conclusion... ?
1) All input signals contain same LF content (presumed faulty) ? Who cares, if the signal has equal amplitude and is in phase, then the assumption stands. There are mixes out there with coherent signals. Mixes out there with incoherent signals. And a mixture of both. Doesn't mean one can't have equal amplitude signals.
2) It is not impossible to have equal amplitude signals in all channels + LFE. Not impossible but perhaps improbable. Doesn't mean it can't or won't happen.
3) You are telling me that there WILL be a difference in peak output in a room between 5 individual full range channels + LFE vs redirected bass and LFE ? How in the world did you arrive at that conclusion... ?
Last edited:
I thought the dolby spec for movies was to filter every chanel at 80hz (or something) and send it to the LFE track? That was the whole point of an LFE track. Just because a track hits 0 DBFS dosnt mean theres any LF in it. And the "loudness" is in the upper mids. I believe U-571 hits 0DBFS on all channels quite often. Some think its the "loudest movie ever".
Back to the question. First I talk of movies only (because they have standards that dolby has setup). If your mains arn't crossed over at too high a freq (say 120 hz or so) and since most of the subs power goes into sub 60 hz anyway, that 5db of headroom is enough and the instructor was right.
This is getting ridiculous.
Let's look at the picture again.
Now, if you had a disk with a copy & paste 40Hz tone on all six channels encoded at full strength on each channel, that would be 0dBFS or -3dBRMS per channel.
Without BM that would be 105dBSPL Peak/102dBSPL RMS for the speakers, and 115dBSPL Peak/112dBSPL RMS-LFE at the seats, if calibrated to "reference" levels. For a total of 123.23dBSPL Peak/ 120.23dB RMS in room at the seats, assuming a non-modal hypothetical perfect room. (In a real room it could be more or less, but that is muddying the waters.)
Now let's flip on BM. Looking back at the chart you can see that the LFE channel is digitally reduced -5dB before anything else is done to it. It is boosted +15dB after the DSP in the analogue section. +5dB to make up for the initial reduction, and +10dB more to give it more reach over the other channels.
The satellite channels are reduced -15dB, and then summed together before being added to the LFE signal to keep the final amplitude equal to the upper satellite signal.
So, for the satellites:
0dBFS - 15dB = -15dBs
-15dB from five channels combined + -5dB from the LFE = +3.23dBFS
As you can't go higher than 0 digitally, you have just clipped the input in the DSP. Your subwoofer will now make an awful noise, and have loads of power dumped through it for no good reason. Turn the Master Volume down -3.5dB, and you are good again.
There is nothing made up here. You can see the Dolby paper, and I have the disk. It clips every time in AVRs from several different brands.
If you made a new test signal with random frequency and phase on each channel, and normalized them to 0dBFS, you get something a little different.
Without BM on, you still get 105dBSPL Peak/ 102dBSPL RMS from each satellite, and 115dBSPL Peak/112dBSPL RMS from the LFE, but when you play them at the same time you only get 116.75dBSPL Peak/ 113.75dBSPL RMS in room, assuming a non-modal hypothetical perfect room[.i]. This is because you are using incoherent/uncorrelated sounds now.
Flipping on the BM you get:
-15dB from five channels combined + -5dB from the LFE = -3.25dBFS
This will pass through the DSP just fine, and the final in room SPL Peak/RMS will be the same as the non-BMed trial.
You must also add in 0dBFS+ signal situations caused bey the digital to analogue conversion method, which can add another dB or 3 to the output after the digital section. 😉
This is getting about as bad as all the people talking about a brickwall filter on the LFE channel years ago. Heck, there are many who still believe in it. 🙄
Let's look at the picture again.
Now, if you had a disk with a copy & paste 40Hz tone on all six channels encoded at full strength on each channel, that would be 0dBFS or -3dBRMS per channel.
Without BM that would be 105dBSPL Peak/102dBSPL RMS for the speakers, and 115dBSPL Peak/112dBSPL RMS-LFE at the seats, if calibrated to "reference" levels. For a total of 123.23dBSPL Peak/ 120.23dB RMS in room at the seats, assuming a non-modal hypothetical perfect room. (In a real room it could be more or less, but that is muddying the waters.)
Now let's flip on BM. Looking back at the chart you can see that the LFE channel is digitally reduced -5dB before anything else is done to it. It is boosted +15dB after the DSP in the analogue section. +5dB to make up for the initial reduction, and +10dB more to give it more reach over the other channels.
The satellite channels are reduced -15dB, and then summed together before being added to the LFE signal to keep the final amplitude equal to the upper satellite signal.
So, for the satellites:
0dBFS - 15dB = -15dBs
-15dB from five channels combined + -5dB from the LFE = +3.23dBFS
As you can't go higher than 0 digitally, you have just clipped the input in the DSP. Your subwoofer will now make an awful noise, and have loads of power dumped through it for no good reason. Turn the Master Volume down -3.5dB, and you are good again.
There is nothing made up here. You can see the Dolby paper, and I have the disk. It clips every time in AVRs from several different brands.
If you made a new test signal with random frequency and phase on each channel, and normalized them to 0dBFS, you get something a little different.
Without BM on, you still get 105dBSPL Peak/ 102dBSPL RMS from each satellite, and 115dBSPL Peak/112dBSPL RMS from the LFE, but when you play them at the same time you only get 116.75dBSPL Peak/ 113.75dBSPL RMS in room, assuming a non-modal hypothetical perfect room[.i]. This is because you are using incoherent/uncorrelated sounds now.
Flipping on the BM you get:
-15dB from five channels combined + -5dB from the LFE = -3.25dBFS
This will pass through the DSP just fine, and the final in room SPL Peak/RMS will be the same as the non-BMed trial.
You must also add in 0dBFS+ signal situations caused bey the digital to analogue conversion method, which can add another dB or 3 to the output after the digital section. 😉
This is getting about as bad as all the people talking about a brickwall filter on the LFE channel years ago. Heck, there are many who still believe in it. 🙄
Last edited:
I've received PMs, so let me answer them here.
You need to keep this part of my post in mind,
"For a total of 123.23dBSPL Peak/ 120.23dB RMS in room at the seats, assuming a non-modal hypothetical perfect room. (In a real room it could be more or less, but that is muddying the waters.)"
This was a hint to use your noodle. 😉
Acoustical and electrical summing are a little different. 1+1 may always be 6 in electric land, while in ascousticville 1+1 could be anywhere between 3 & 6 before boundary issues even get added in. The distances between the speakers, and the frequency(ies) in question are the major factors here. We are talking bass, remember?
Two separate mono sources are not always limited to a 3dB gain. 😉
When you throw room issues into the mix it is also easy to see how some people could measure mid 120s levels, while others never do, without falling on the old "user error" crutch. Of course, it isn't always just a crutch. 🙂
You guys did notice the bone stock incoherent in room totals, right? These would be similar to a "real" large room, with non-coupled/non-collocated speakers.
You need to keep this part of my post in mind,
"For a total of 123.23dBSPL Peak/ 120.23dB RMS in room at the seats, assuming a non-modal hypothetical perfect room. (In a real room it could be more or less, but that is muddying the waters.)"
This was a hint to use your noodle. 😉
Acoustical and electrical summing are a little different. 1+1 may always be 6 in electric land, while in ascousticville 1+1 could be anywhere between 3 & 6 before boundary issues even get added in. The distances between the speakers, and the frequency(ies) in question are the major factors here. We are talking bass, remember?
Two separate mono sources are not always limited to a 3dB gain. 😉
When you throw room issues into the mix it is also easy to see how some people could measure mid 120s levels, while others never do, without falling on the old "user error" crutch. Of course, it isn't always just a crutch. 🙂
You guys did notice the bone stock incoherent in room totals, right? These would be similar to a "real" large room, with non-coupled/non-collocated speakers.
Last edited:
Why don't you come with me to AVS forum (you remember that forum right, used to post there, yes ?) and we can discuss this with other like minded individuals ? 😉
Cedia training - Page 2 - AVS Forum
Cedia training - Page 2 - AVS Forum
- Status
- Not open for further replies.