Reference level

[Vaughan Odendaa]

Hi guys,

Attended a Cedia Home Cinema Designer course yesterday and something that is bugging me pertains to reference level requirements. It was said that each main channel has a maximum peak output of 105 dB's at the listening position and the LFE channel requires 115 dB's.

I have no problem with that. But I assumed the subwoofer requirement of 115 dBs maximum peak was due to speakers set to 'large'. So the LFE track is handled by the sub (or subs), and each main channel handles it's own LF information. But that 115 dB figure was specific to the LFE track and the LFE track only. But if you using bass management (which is a big part of the recommended practices in the course), those requirements fall out the window.

What the Cedia instructor told me was that the requirements would remain the same. I explained to him that the LFE track + directed main channel bass when summed and low passed should result in much steeper requirements (look, I'm perfectly willing to accept that I'm wrong, but in this case I'm confident I'm right). The levels jump to 120/121 dB's for the sub (again, maximum peak, assuming it is calibrated flat).

The instructor went on a tangent concerning 80 Hz crossovers and didn't seem to understand what I was talking about. He then said that low bass didn't exist in the main channels (as a counter to my argument as naturally the sub will be more stressed by handling all bass curtailed from their channels). I asked him if he was sure and he didn't know.

But the guy is an expert on everything else. Which took me by surprise since the course is supposed to teach us how to maximize performance for home theater by using objective standards used in the industry. So I am correct ? Is there something else I have missed out ? Or is the Cedia instructor correct ?

If I'm right and I'm designing a theater to achieve maximum dynamic impact then his philosophy would result in 5-6 dB's of compression (on the sub side) and/or possible damage, assuming the sub or subs were capable of 113-115 dB's.

Advice would be appreciated.

 

[Vaughan Odendaa]

Does anyone know ?

 

[Matt34]

Does anyone know ?

Well, it is like 2 AM in the states, give it a few hours.

 

[Vaughan Odendaa]

Sorry, yeah I know our times are a bit messed up. I'll wait for your responses.

 

[InTheIndustry]

Does anyone know ?

The instructor is correct.

I need more of an explanation of why you think bass management would change the 115db ref lvl to fully understand what you're saying on that front.

As far as large vs. small there's NO good reasoning that someone with a sub(s) should ever run their speakers at large. There are several different reasons for this, one being that Dolby Digital tracks are mixed so that the main 5-7 speakers receive only 80hz & up with the intent/assumption that there is a seperate powered subwoofer present.

Perhaps the instructor struggled with your concept because it's based on false assumptions/logic?

Another note on why bass management could possibly not do 115db on LFE is the capability of the subwoofer(s) being used vs. room size? Just a thought?

 

[Vaughan Odendaa]

The reason why I believe the figure of 115 dB changes is (as I said earlier) because you are summing bass from each main channel in addition to the LFE channel. The total sum is far greater than the LFE on it's own.

You are telling me that a 115 dB's + redirected bass from each main channel will still end up 115 dB's at the sub out. I'm saying, no it won't. I'm saying the requirements are THAT much higher because the subwoofer has to handle that much more. That doesn't strike me as faulty logic. It sounds pretty darn logical to me.

My Cedia instructor didn't understand THAT. FYI, I'm not arguing with the concept of bass management, I believe in it's effectiveness. I simply disagree with the claim that the sub+redirected bass will require 115 dB's maximum peak output at the listening position (calibrating flat to 75 dB's).

If he is somehow correct then none of this makes any sense.

 

[AcuDefTechGuy]

As far as large vs. small there's NO good reasoning that someone with a sub(s) should ever run their speakers at large. There are several different reasons for this, one being that Dolby Digital tracks are mixed so that the main 5-7 speakers receive only 80hz & up with the intent/assumption that there is a seperate powered subwoofer present.

You mean unless the front towers have large built-in powered subs like the RBH T2/P or the Definitive Technology BP7000SC, right?

 

[Vaughan Odendaa]

I'm not overly concerned about the bass management debate. That's for another thread. I'm only concerned about the max dynamic output from the LFE+redirected bass. My lecturer says it's 115 dB's. Is that correct ? If so, please elaborate and explain why. I think it's closer to 120/121 dB's.

I mean, if you sum more than one signal and those signals are in-phase, the signal won't remain constant. It can only increase. So you add 5 ch's of redirected bass to the LFE for the sub to churn out (remember, these are all dynamic peaks, whether the recording engineer mixed this content at those levels is not my concern, I'm strictly talking the absolute dynamic peaks that is possible in the chain)

So either I don't know what I'm talking about or I'm more clued up on this than the Cedia teacher. Thoughts ?

 

[InTheIndustry]

You mean unless the front towers have large built-in powered subs like the RBH T2/P or the Definitive Technology BP7000SC, right?

2 notes on that....

1. While "technically" it is a "tower", it's really just a monitor/bookshelf speaker sitting on top of a powered subwoofer. Think about it.... The sub is what's getting the LFE chanel through the RCA cable while the actual main channel speaker still gets its normal wiring. This is totally different than what is refered to as a trulyfull range passive speaker that can cleanly play down to 20hz. The speakers you're referencing should still be set to small.

2. Bass is placement sensitive. PLACEMENT SENSITIVE. Having your .1 low freq channels litterally nailed down to your main L/R is NOT ideal unless where your L/R happens to be is exactly where your .1 LFE will perform its best. This rarely works out in practical use. Corner loading a sub in the left & right is usually a good place, but conversely, the same is not true for the main L/R.

Do my two points make sense to you?

 

[InTheIndustry]

I'm not overly concerned about the bass management debate. That's for another thread. I'm only concerned about the max dynamic output from the LFE+redirected bass. My lecturer says it's 115 dB's. Is that correct ? If so, please elaborate and explain why. I think it's closer to 120/121 dB's.

I mean, if you sum more than one signal and those signals are in-phase, the signal won't remain constant. It can only increase. So you add 5 ch's of redirected bass to the LFE for the sub to churn out (remember, these are all dynamic peaks, whether the recording engineer mixed this content at those levels is not my concern, I'm strictly talking the absolute dynamic peaks that is possible in the chain)

So either I don't know what I'm talking about or I'm more clued up on this than the Cedia teacher. Thoughts ?

I'm sorry, but I'm failing to follow your line of thinking.

Reference is reference. Sure, you can turn things up louder and get more DB, but that is not reference standard.

Lets say you throw 5,6,7, even..... 100 LFE channel subwoofers into a room, you still calibrate the system to 115db for that channel (LFE). Once the 115 DB reference standard can be met adding more subs is for EQ/management.... NOT output. If you direct signals under 80hz to your main speakers you are asking them to perform the LFE portion of the sound track and, therefore, asking them to essentially be subs.

Besides all of that, your concept of "redirected bass" is a false idea acording to Dolby Labs. In a soundtrack bass is its own channel that is meant to be managed through placement which is why the idea of putting LFE signals to the main speaker channels in a room is a bad idea. It would be a nightmare, if not impossible to place correctly unless each custom one-off piece was specifically designed to work in a specific placement in a custom one-off room.

For the fun of debate & illustration... lets move past that....
If your passive speakers could perform 20hz signals at 115db (which only a few can. & by few I mean a handfull).... you wouldn't have a need for the .1 speaker. You would simply set your speakers to large & let the main system would be doing all of that work. But, as I mentioned above and in response to ADTG, that would be a bad idea.

If you're only concerned about max output and not management or what is correct then I don't know what else to tell you, friend. Like I said above, you can always make things louder. But that doesn't make it better. Reference is reference. To go over that would be considered "incorrect". But, at the end of the day, it's all about what sounds best to the listener, right?

I hope this all makes sense. I'm not trying to be smart towards you, just as direct and encompasing as I can.

 

[Vaughan Odendaa]

I don't think you understand the crux of my argument at all. You combine 5 channels of deep bass together with the LFE channel and you simply CANNOT have 115 dB's. Mathematically, that is impossible.

It makes no sense. You can't tell me that summing multiple signals (in phase) plus a maximum peak level of a 115 dB from the LFE will equal 115 dBs. That is what the lecturer essentially led me to believe, that it wouldn't make any difference either way and that is what I have issue with.

As far as my concept of bass management, it's just fine. If it is in error then you haven't demonstrated it yet because you haven't pointed out one factual error I've made. I have said, clearly, that each main channel has it's OWN bass information recorded, it is full range. 5 full range channels plus an LFE channel for super low, and loud bass. There is plenty of bass recorded in these main channels. I don't care about LFE in the main channels as that is not relevant to my argument. Actually, there is no LFE in the main channels. The only scenario where that can happen is if you tell your receiver that you have no subwoofer. Then it will reroute LFE bass to the main channels. In any case, those are two different things.

Like I said, I am not debating the merits of bass management. I could care less whether main speakers should be set to 'large'. That is not what my thread is about. I am a big proponent of bass management. I use it on all my systems. I am simply saying that summing redirected bass from each main channel together with bass in the LFE does NOT equal 115 dB's. And so....this REQUIRES a better solution. A solution to achieve that. Since we are discussing the highest dynamic range possible in a home theater system, you can't seriously tell me that utilizing bass management still requires 115 dB's from a sub given the information I've brought to the table.

The idea that you need 115 dB's at reference when using bass management, THAT is what I am discussing here so I don't understand why you can't follow that. Does it not make sense to you that essentially what you are doing is adding considerably more workload to the sub ? It has to handle all the bass recorded in the LFE and now it has to cover all the bass in the main channels as well. That means MORE work. Multiple signals summed means MORE level.

So, again I ask, what is the situation here ? If I am wrong, show me the math. Explain the theory as to why I am wrong. I don't care about output and output only. I care about objective facts. If I buy a subwoofer that can handle 115 dB's then that is great, especially if I am setting speakers to 'large', since the sub will handle ALL the information, as per reference level spec for the LFE. But what about when speakers set to 'small' ? What then ? If, at reference, you are 5-6 dB's over that spec, then at best, you've just added 5-6 dB's of compression, so your dynamic output is flattened, and at worst you have bottomed the subwoofer out completely.

Does that make sense to you ? Because I don't know how else to explain it. I think I have been clear enough.

Lets say you throw 5,6,7, even..... 100 LFE channel subwoofers into a room, you still calibrate the system to 115db for that channel (LFE). Once the 115 DB reference standard can be met adding more subs is for EQ/management.... NOT output. If you direct signals under 80hz to your main speakers you are asking them to perform the LFE portion of the sound track and, therefore, asking them to essentially be subs.

You are still not understanding what I am talking about. Yes, you can add 100 subs, calibrate them together to achieve 75 dB's and you should never exceed the maximum dynamic output of the LFE track. I don't care about that because I'm not discussing that. I am saying, you take the OTHER main channels (left, center, right, surrounds etc) which has ample bass content mixed in, you ADD that together with the LFE.....then what ? Does the maximum output peak still remain 115 dB's ? If that is true, please demonstrate the math behind it and please explain the theory behind it.

Bass under 80 Hz in the main channels is not considered LFE either. It is main channel bass. LFE bass is LFE bass. Unless you set your subwoofer to 'off', the main channels won't be voicing any LFE content whatsoever.

If your passive speakers could perform 20hz signals at 115db (which only a few can. & by few I mean a handfull).... you wouldn't have a need for the .1 speaker. You would simply set your speakers to large & let the main system would be doing all of that work. But, as I mentioned above and in response to ADTG, that would be a bad idea.

Forgive me for sounding a little agitated but my topic has still not been addressed in any way, shape or form. I'm not debating the merits of bass management. I don't care about 'large' vs 'small' so let's not discuss that, okay ? My understanding of bass management has not been shown to be in error. I just need to have someone explain the actual situation to me. Explain the math to me...assuming my argument is at fault. That's all I ask.

 

[Vaughan Odendaa]

Another note on why bass management could possibly not do 115db on LFE is the capability of the subwoofer(s) being used vs. room size? Just a thought?

If your room is flat, and you calibrate flat and if the mixing engineer maximizes the dynamic peaks in the LFE track then the maximum dynamic peaks you can expect from your movies at your listening position is a 115 dB's. Again, assuming you are handling the LFE on it's own. BUT WHAT ABOUT THE OTHER CHANNELS ? The sub still has to handle that. Does it not ? Should we just forget about all the bass that you've just redirected and assume it's evaporated into the ether ?

The total output required at reference from the sub should be what ? Please enlighten me.

 

[WaynePflughaupt]

I think I get what you’re saying, Vaughan. You’re saying that adding all the other 5 to 7 channel low-freq signals to the LFE is going to boost its level significantly, right?

Technically you’re correct. Shoot, assuming that each additional channel will increase the signal at least a couple dB you’re looking at something like a 10 dB increase, from 115 to 125 dB.

But my thinking is that it ultimately doesn’t matter. Why? Because that would mean your system would be grossly out of calibration. Don’t take my word for it - go increase the level of your sub 8-10 dB and try living with that for a while!

So - even if all the re-directed low freq to the LFE channel means a significantly-increased signal to the sub, ultimately that’s governed by your standard calibration, which typically has the sub running about 10 dB hotter than the mains, not 20+ dB hotter. In other words, you’re going to reduce the sub level to compensate, which means you end up right back where you started. Unless you just like everything bass-heavy, that is...

Regards,
Wayne A. Pflughaupt

 

[Vaughan Odendaa]

I think I get what you’re saying, Vaughan. You’re saying that adding all the other 5 to 7 channel low-freq signals to the LFE is going to boost its level significantly, right?

Yes, I'm saying that if you sum all the main channel bass with the LFE (according to bass management lore) that dynamically the subwoofer will be more taxed than it would ordinarily be if it just voiced LFE content on it's own.

So instead of handling LFE (115 dB maximum peaks) if set to 75 dB's, you have main channel bass to contend with as well. We know that this main channel bass exists -- there is overwhelming evidence for it. So we have to add that variable to the equation. If we calibrate flat, that information still exists -- it ain't going nowhere -- its simply been shifted.

So what you are suggesting is that it doesn't matter if the total sum is much higher than 115 dB's because ultimately you'll just calibrate and lower the level ? Isn't that the equivalent of messing up the mix ? If you reduce the subwoofer trim then at reference level you would have affected the level of LFE+redirected bass so dynamically the system has been crippled.

To me, it seems like there can be no free lunch and something has to give otherwise why bother creating 5 full range channels to begin with ? Why not dump all LFE content into the sub in addition to all main channel bass in one foul swoop if there was no added strain for the sub ? But I'm not the bass management guru here, I'm just trying to understand this as best I can...

 

[Pyrrho]

You appear to be assuming that all of the channels will have their bass in phase with the bass in low frequency effects channel. If the bass is not in phase with the low frequency effects channel, it can actually reduce the amount of output required of the subwoofer if one adds in the other channels' bass.

To keep matters simple, imagine a 2 channel stereo recording, done in a minimalist style. Thus, two microphones are used, one for the right channel and one for the left channel. There is zero reason to suppose that the bass from one channel will be in phase with the other channel.

Also, the reference level requirements are simply standards for what the equipment must be able to do to be certified or whatever the standards are for. Thus, there is no need for the standard to be different based upon what one does regarding bass management.

You might want to get someone from Cedia to respond, in order to get the story on what the standard is all about. Perhaps contacting them directly about the standard might deal with the issue in the best way possible. Perhaps the standard has already taken into consideration the issue that you are concerned about; after all, the standard for the low frequency effects channel is already 10 dB higher than the other channels. After all, if you are reproducing a natural sound, if the peak volume you require is 105dB for the main channels, why would one want a bass speaker capable of 10dB greater?

 

[Vaughan Odendaa]

You appear to be assuming that all of the channels will have their bass in phase with the bass in low frequency effects channel.

If that is the case then we can assume that all main channel bass is not in phase either. There is no reason not to assume that the bass would be in phase with the LFE channel. Otherwise what is the point of having 5 full range channels with deep bass recorded in them (with 105 dB peaks from each) if some or most of it will be out of phase. Makes no sense.

Also, the reference level requirements are simply standards for what the equipment must be able to do to be certified or whatever the standards are for. Thus, there is no need for the standard to be different based upon what one does regarding bass management.

So where does that extra information go to then ? All that redirected bass, if it's out of phase, then the entire concept of bass management and reference level would seem to be invalid. You can either have all main speakers handle the information recorded or you let the sub handle all that information as well as it's own LFE information. You are saying that using bass management shouldn't affect what the sub is essentially 'seeing'. I can't imagine how this couldn't be further from the truth.

You might want to get someone from Cedia to respond, in order to get the story on what the standard is all about. Perhaps contacting them directly about the standard might deal with the issue in the best way possible.

I honestly don't think the Cedia guys would know the answer to that question. I did ask the question but I got a vague response. The answer was 'no, it won't change anything', and then 80 Hz was brought up, so the vagueness of it all just let me in the dark.

 

[Vaughan Odendaa]

http://www.hometheaterhifi.com/volum...pril-2000.html

A quote for theater system with no BM:

"If you are doing the math right now, you may have already concluded that in a text-book Dolby Digital theater, one heck of a subwoofer system is called for! Let me explain.

In a movie theater on the island of Utopia, with ideal hardware, any single screen channel should be capable of a clean 105 dB peak with it's own respectable bass. The LFE channel should be capable of a 115 dB peak. Drive all channels to the max and the system should be able to slam you with approximately 120 dB of bass information. Thank you Dolby."

 

[moves]

The instructor is correct.

I need more of an explanation of why you think bass management would change the 115db ref lvl to fully understand what you're saying on that front.

As far as large vs. small there's NO good reasoning that someone with a sub(s) should ever run their speakers at large. There are several different reasons for this, one being that Dolby Digital tracks are mixed so that the main 5-7 speakers receive only 80hz & up with the intent/assumption that there is a seperate powered subwoofer present.

Just wondering about the reasoning behind small and large settings... i just checked the setup on my avr and when it said large for my fronts.... under SUB a 'yes' sign popped up. Under small.... it says to use this setting only with THX certified speakers... I am using totem Forests with a pioneer SC 27 So I am wondering how come you are recommending a 'small' setting. I also have a sub.

 

[MDS]

I think I understand the question as posed but I also think there are a few misconceptions about digital samples and how they relate to 'reference level'.

- Dolby reference level is simply a standard for playback in a movie theater. It is 105 dB for each channel and 115 dB for the LFE channel and we try to replicate that in our home theaters.

- The AVR/pre-pro does the 10 dB boost for the LFE channel.

- Reference level is achieved when the digital sample PEAKS, ie. '0 dB' (all ones).

- No sample can exceed 0 dB and the LFE channel is no different.

- Digital samples don't have any SPL until they are converted to analog and amplified, thus 'summing' a time slice of different frequencies and sample values from the Small channels cannot increase the level beyond 0 dB - it is the max.

- When you calibrate the AVR you are setting it up so that the voltage it delivers in response to a PEAK sample is sufficient to drive the output SPL to reference level.

- All of the calibration related to reference level is moot if the average level of the soundtrack is greater than the -30 dBFS test tone you used to do the calibration. If the average level is greater, the output SPL will be greater too but -30 dB is more or less a de facto standard for movies.

 

[skers_54]

Vaughan,

If you're considering just pure signals, then you are correct that the sum will be greater than 115. From what I gather, you want to combine one 115 dB tone with 5 105 dB tones (all at say, 40 Hz). Adding those together will yield an spl of ~117 dB. You can use this calculator

http://www.doctorproaudio.com/doctor/calculadores_en.htm

Keep in mind that spl is log-based so the mains are an order of magnitude down. So you would be correct from an academic sense. But the difference is much smaller than you were imagining.

In the real world, what I think ITI and your instructor are saying (someone correct me if otherwise) is that movie signals are constructed such that this type of combination does not happen. IE, there wouldn't be the same 40 Hz tone on both the LFE and main channels. There's always exceptions (I'm sure DVD-A and SACD don't follow this rule). So from CEDIA's standpoint, 115 dB at the listening position is all you'll have to go given the practical implications of the mixing process.