I thought there was a calibration specification for when the sound system is set up in theaters so that dialog in the edit suite ends up the same volume in the theater and the same for explosions.

Those home theater calibration DVD's from Avia and whoever talk about some kind of signal they they generate needing to be around 80db in the room.

At least I think that's what I had understood. I was just wondering how that translated to the digital meters in Vegas.

JBJones

A bit long here...
Spot, I ask you a question at the end.

Yes, I think a good tutorial on this could be very useful.

The Avia calibration DVD was mentioned. In their tutorial explanation they mention that the DVD audio spec is a level of -20dB recorded should produce a sound level of 85dB. They also mention that the 85dB level will probably seem way too loud for a home system and a lot of people set the -20 signal to produce about 75dB on a sound level meter.

Using Avia, I set up my home system for about 75dB output when the volume control is set to 0 dB. This seems plenty loud to me. Something like the "Private Ryan" DVD rocks the room pretty good at 0 dB volume, and I can crank the volume knob up +10 dB if I want to get to Hollywood specs, but usually I find that way too loud. Typically all the commercial DVDs I have tried sound theatrical to me with my volume knob set to 0.

On the other hand, how the sound gets encoded from Vegas into AC3 makes a lot of difference and I haven't worked it out yet so that I think I really understand it all.

I made a DVD recently with some big booms in it up near 0 dB on the Vegas audio scale. When I encoded it, those parts were loud, but seemed way compressed vs. the sound levels coming off my commercial DVDs. I used all the default setups when I did the audio encoding.

I think if I change some of the AC3 encoding parameters I can get the loudness I expected, but I haven't taken the time to play with various audio signals and encoding parameters to really understand how to make the audio come out of a DVD as I want it. I understand that the parameters are there so that I can get what I want, but I don't yet really understand when and why to choose what.

Getting a good audio level for dialog, good mix with background sounds and music, but still having all the head room for a big punch where I want it, is finesse that I don't have yet.

Spot,
Will the VASST Surround audio class make this all clear for me?

Rex,
First, I need to say that Avia, if what you've written is correct, is absolutely 180 degrees off base. Saying that setting a digital meter to -20 on output 'should equal______' is ridiculous. That would fail to account for;
>monitor efficiency
>size of room
>amplifier efficiency
>amplifier power
>acoustic factors in the room
>any gain or attenuation in signal path

-20dBFS (Full Scale=digital meter) is the ATSC standard to match 0dB in the analog scale. It bears no relevance to acoustic measurement in a room. The measurement quoted above would only be relevant if they were talking about a specific sized room containing a specific model of amplifier connected to a specific set of speaker monitors.

The Surround VASST gets into some of this, but given that I've never seen anyone spec out a digital output level to be equivalent to a weighted scale dB meter in volume.....that's not on the agenda. Rather, there are calibration techniques shown, but....that's about it.

I think this calibration scale was created for cinema owners out there so that they could set the amplifiers in their rooms to the proper volume without having to play the first.

I'm aware of the ratio of a digital meter to an analog meter. But what I"m not sure about is how the cinema world sets up their screens.

Spot, all of those items you listed as failing to be accounted for would be accounted for in some way by taking the physical measurment of the sound level in the optimum viewing location.

It's physicaly impossible to tune a surround system perfectly for every seat in a theater so obviously they have to choose where they want to be optimum. Size of the room, amplifiyer efficiency, power, room accoustics, would automaticly play into where they set there volume knob on the system when they were measuring the SPL in the room at the designated optimum location.

In other words. If you were 2 feet away from the speakers then obviously the amplifiers would need to be set to a lower volume than if you were 20 feet away.

It doesn't seem likely to me that that when a film is mixed that they set the loudest sound in the movie to -00db on the digital meter and just let everything else be relative to that. If that were the case then a movie with explosions would have dialog set a good bit lower than the maximum 0db on the meter. While a drama without any explosions might have the the dialog set at nearly 0db. That would cause havoc to cinema owners because without playing th movie all the way through, they wouldn't know where to set the volume knob on their amps.

Action movies would have dialog that is too quiet or else dialog in a drama would be deafening.

So, apparently from what RexA said about the Avia DVD's, the spec states a -20dB level should run around 85 in the listening postion.

As far as my responsibility to mix, I think I should follow whatever standard is set and just let the audience adjust their home systems to whatever they like best. If they want to calibrate to 75dB instead of 85dB then that's up to them. But at least all the DVD's they put in their system would be comparable.

JBJones

>Saying that setting a digital meter to -20 on output 'should equal______' is ridiculous.

As jbjones mentioned, I don't think it is ridiculous at all. I did some web surfing and found some great information.

First Dolby has a paper (pdf) on setting up your environment:
5.1-Channel Production Guidelines

Pages 33 and 34 talk about the -20 dBFS pink noise reference and 85 dBC SPL.

Here is a thread in another forum about this subject:
Room Calibration for Film and TV Post

The first message there has some great links to other information. The Blue Sky link seems particularly clear and succinct on the subject. It also has calibration files that can be downloaded.

All of the links do clearly establish that -20 dBFS is equivalent to 0 VU, but they go on to give guidelines for a reference SPL in your monitoring room.

I'm glad this subject came up, or I wouldn't have found my way to all this great information.

K, to make it clear as to what I'm trying to point out...
You CAN'T take a statement like 0dB output from the source equals XXXdB at monitor. Impossible to do.
The comparative is the ATSC standard of -20dB matching 0dBvu. That's the standard.
Now, if you want to say that the optimum listening level is 85dB or 75dB or whatever dB, that's fine. But that doesn't equal anything to the output peak of the finishing system. It simply can't, without setting standards for equipment, and Dolby (nor anyone else) does that.
Reference SPL is quite different from master output levels.
As an example, a playback system monitored at -.03 in Vegas meters, that consists of 5 surround monitors containing 4" speaks and a dome tweet, plus a 12" sub driven by a 40 watt distributive amplifier, in a room that's 12' x 12' will be no where near (apparently) as loud as a system that is made up of 5 8" monitor speaks powered by 200 watts (direct/biamplified) in that same 12'x12' room. Further, the differences in system balance, efficiency, frequency response will cause any number of flutter issues, overburden, and imbalance of the overall system causing one to need to be turned up and one to need to be turned down.
SPL at what distance? Center of room? Back of room? Center speak?
I'm disputing the statement of relativity, not the statement of recommended SPL. To say that XXdBFS equals XXXdB SPL is where the statement is inaccurate. That's my point. IN the case of the Dolby paper, they are measuring on a C weighted scale, which is the normalled scale for this sort of setup. But that's not part of the original post/statement.
None of this even takes into account clarity, accuracy at what level.
For example, a simple one that everyone can relate to is that listening to a low watt, AM radio signal of Van Halen in your car on a single 4" speaker at 85dB will be fairly distorted while listening to the same volume level on a high power, well balanced and efficient system will be clean and punched. We've all heard and experienced this at some point. It's all relevant to equipment, environment, and program material. By the statement that XXXdBFS equals XXXdBSPL, it suggests that two systems of gross difference would have the same SPL regardless of all other factors. Further, it fails to take into account any gain or attenuation stages that are almost certainly taking place from a soundcard of one sort or another.
85dB SPL is the standard for setting up a theatre, no dispute there. And does leave room for up or down. It's a benchmark. -20dBFS is a benchmark for average program as well. But the two aren't necessarily related.
I think we're all more or less saying a similar thing, but maybe somewhere in there, I'm missing why the output level and SPL are tied together.

Spot,
I think you are missing the point. 85dBc (slow) is what you choose to set the system to. You calibrate your room so that 0 dB or -20 dBFS produces 85 dBc (slow) at the listening spot, or 79 dBc for braodcast facilities. The only reason for these setting is to try and produce the same sound in every theatre. The producers want their product to be the same quality in every venue. Lucas created the TAP (Theatre Alignment Program) in 1984. They trained technicians to go into theatres and measure and report the sound and picture quality. This was used for major 70 mm releases to try to improve the overall quality of the presentations.

I have tuned movie theatres for many years, and although the producers and sound mixers may expect that the theatres will play back sound at the calibrated levels that seldom happens. It is just too loud. There is no way that I could mix audio at the calibration level. During one class at Dolby they explained that the 85 dBc was chosen so that noise from HVAC and other sounds would not significantly effect the readings. Reading the Dolby paper seems to contradict that idea. The actual level that you use will indeed vary because of the room, speakers, electronics and the judgement of the listener.

Interesting to hear that 85 dBC is too loud in comercial theaters too. I guess that certainly makes sense. As I said, when I tuned my home theater I found that about 75 dBC from the -20 reference signal with my volume knob set to 0 dB, seemed to work well for me. When I view a comercial DVD and I want "theater" sound levels, I can pretty much set my knob to 0 and be happy.

Using the actual SPL in your room as a reference is just a way to get your mix in line with other mixes. Clearly, you want all your speakers producing equal volume from the same signal at your chosen location. Exactly what that level is, is a bit subjective. Ultimately the goal is to not have sound levels vary significantly from one AC3 mix to another.

So the 85 dBC SPL "standard" seems to be a way to try to get everone on the same page in terms of encoding sound levels. I guess my real goal is just to not be way off relative to other DVD mixes. To do that I need to have some standard relationship of signal dBFS to room dBC SPL in my mixing environment.

I think we all are trying to say the same thing here.

As I said before, we are saying exactly the same thing. It's the XXXdB=XXX SPLgeneralization that is at issue. That's what you calibrate TO, rather than say simply happens when you connect XXX to XXX and monitor what's happening in Vegas at 0dBFS. The bigger concern is where do you calibrate your outputs to, as opposed to calibrating your room to. In other words, you set Vegas to a house standard which is approximate with -20dB average, and then set your monitoring environment to what's comfortable. Unless you have an average listening environment. Most of us don't in terms of our editing rooms. Our control room is 27 x 16. That's pretty big by most smaller studios, yet 85 dB would kill us in that environment. But it's close to the average living room, I guess. So, while it's a recommendation, and a standard in theatre rooms, it also needs to be taken with a dose of rationality.
Further in the equasion is that a heavily compressed/small dynamic mix at -.03 will appear far louder than a very dynamic, marginally compressed mix with peaks at the same level of -.03. (We should for max peaks at -.03)
We use WAVES UltraMaximizer+ on all outputs, in addition to other plugs that help achieve a good dynamic presence yet maintain a master reference point. How you'll encode to AC3 also plays a large role in all this, and unfortunately, there isn't a way to monitor what's being encoded. We've talked about having an AC3 simulator that would simulate how the compression might generally behave, but that's not come to fruition. It's too bad Dolby makes all these great tools available, but doesn't provide a real-time decode that could be plugged into authoring tools at this level.
So in short, no matter how well you set up your room, outputs, etc, it's still only an approximation of what's gonna come out on the encoded side.

Cool.

I think my original question was answered (or confirmed) in all of that.

JBJones

Just read this entire thread and although everyone seemed to be disagreeing you all seemed to be saying the same thing, amazing!
Naturally no meter can define how load anyone will perceive sound. Apart from anything else if you hate the music it's always too loud.
All that this calibtaion process is trying to do is make the apparent loudness the same. If I send SPOT a file of solo piano and our studios are calibrated the same then the SPL should be the same, within reason of course. SPOT might think its a load of crock in which case he'll percieve it as being way too loud or he might think it's the greatest bit of music he's ever heard and he wants it much louder.

I suspect part of the reason for calibrating cinemas is to stop the tendency to wind movies up louder and louder. Must say until I saw a movie in a setup just calibrated by Dolby I didn't realise just how good movie soundtracks could be. My hat goes of to most of the composers, musicians and the engineers, pity is that rarely do the public get to hear the full scope of their work.