what level to set peaks for cinema?
- Thread starter Soldier
- Start date
ryanjf
Well-known member
For films, just don't clip. You have a lot of headroom and more dynamic range, more than television and Internet media. The calibration for the United States is -20dBFS PINK Noise = 85 dB SPL in a certified mix room (this number varies depending on the volume of the room). 85 SPL is LOUD. Therefore, you have 20dB of headroom, and if something hits around 0dBFS, that's a sound that's at 105 dB SPL !
A lot of the mixers I know don't mix to numbers, they mix in a room where they know if something is too loud by ear. I guess that's why you always have to budget for a proper mix stage just to check the mix.
I think films are getting louder in general. It's the loudness war. Super 8 was extremely loud, especially during the train derail sequence. I saw it at 2 different theaters and it was loud in both places. I even asked a couple of my buddies and they said they had to cover their ears.
I'm a big fan of dynamics, where the dialogue can range between -23dBFS and go as low -35ish or higher. If you're mixing for theaters, then I think you have a little more wiggle-room to play with dynamics versus mixing a film for a DVD release. If there's budget, the re-recording mixer will remix the film (with less dynamic range) for home theaters. I know Drag Me To Hell (Blu-Ray) was done this way and a lot of Disney films as well. I can't remember where I read the article, but it does exist.
A lot of the mixers I know don't mix to numbers, they mix in a room where they know if something is too loud by ear. I guess that's why you always have to budget for a proper mix stage just to check the mix.
I think films are getting louder in general. It's the loudness war. Super 8 was extremely loud, especially during the train derail sequence. I saw it at 2 different theaters and it was loud in both places. I even asked a couple of my buddies and they said they had to cover their ears.
I'm a big fan of dynamics, where the dialogue can range between -23dBFS and go as low -35ish or higher. If you're mixing for theaters, then I think you have a little more wiggle-room to play with dynamics versus mixing a film for a DVD release. If there's budget, the re-recording mixer will remix the film (with less dynamic range) for home theaters. I know Drag Me To Hell (Blu-Ray) was done this way and a lot of Disney films as well. I can't remember where I read the article, but it does exist.
Last edited:
ryanjf
Well-known member
Here's a good sticky I'm referring to: CLICK HERE! for GS. You'll find some important stuff there.
prexdealer
Well-known member
ryanjf - I don't totally understand because I am still learning. But I am so grateful and impressed that there are people like you (there are so many great contributors Alex H, Chadfish, Fisk, Alan Chung so many people can't remember all the names but thanks!) contributing on this board. Thank you to all of you.
ryanjf
Well-known member
In the linked forum above, danijel says, "There are no guidelines in terms of average loudness, peak or any other level measurement. You achieve proper levels by properly calibrating your listening environment, so that it resembles the environment of the theater" when referring to Movie Theaters. A couple lines later, he says, "Here, same rules apply as with the theatrical mix, except that the monitoring is different (near-field, no X-curve), the room is smaller, it is calibrated lower, AND there is the dialnorm parameter if your sound is AC3 encoded" about DVD levels.
When you build a mix stage for film, it has to mimic a real theater. To get the stamp of approval from Dolby, they come in and blast and measure the room with pink noise and set the room EQs. These EQs should never be touched again. The place I work for just built a Dolby 7.1 room over the summer and we sealed the room EQs behind some glass (then screwed them in with 6-8 screws) so no one can touch them. These EQs are the X-Curve and is applied to the B-chain of the monitoring signal. Dolby has a good technical white paper on the X-curve on their website and explains A-chain and B-chain pretty well. While they are somewhat a boring and technical read, you can get some great info from it.
Knowing that your mix stage mimics a theater, if something is too loud, then it'll be too loud in the theater so you turn it down. Obviously, you'll probably have to do a test run and see how a known good film sounds in your mix stage. We played The Dark Knight probably everyday just to hear how a DVD mix sounds (I'm not too sure if the DVD release of the Dark Knight is a special DVD mix or a theatrical - I will double check). We also played The Social Network in the same room at the same reference level and the dialogue level matched the Dark Knight's without touching the control room knob.
The room we built in the summer is really small compared to some of the ones you'll find at the Ranch or in LA so we monitor at a lower SPL (82 dBSPL for us) although the bigger rooms will monitor at 85 dB SPL. And yes, I still think 82 dBSPL is too loud for the room we built. I think anything above 80 dBSPL is freakn' loud unless you're in a big room.
The mixer I know was working on a documentary yesterday and from what I saw I'd say his average dialogue level wasn't going past -24dBFS and was probably going as low as -33dBFS. This level sounded just right compared to the rest of the elements in the doc. No LM100 or Dolby Media Meter was used to watch peaks. You can just slap a brick-wall on the DX bus and call it a day if you're worried about clipping. With dialogue mixed at that level, the brick wall won't be touched that much - just catching random peaks that get past the compression.
When you build a mix stage for film, it has to mimic a real theater. To get the stamp of approval from Dolby, they come in and blast and measure the room with pink noise and set the room EQs. These EQs should never be touched again. The place I work for just built a Dolby 7.1 room over the summer and we sealed the room EQs behind some glass (then screwed them in with 6-8 screws) so no one can touch them. These EQs are the X-Curve and is applied to the B-chain of the monitoring signal. Dolby has a good technical white paper on the X-curve on their website and explains A-chain and B-chain pretty well. While they are somewhat a boring and technical read, you can get some great info from it.
Knowing that your mix stage mimics a theater, if something is too loud, then it'll be too loud in the theater so you turn it down. Obviously, you'll probably have to do a test run and see how a known good film sounds in your mix stage. We played The Dark Knight probably everyday just to hear how a DVD mix sounds (I'm not too sure if the DVD release of the Dark Knight is a special DVD mix or a theatrical - I will double check). We also played The Social Network in the same room at the same reference level and the dialogue level matched the Dark Knight's without touching the control room knob.
The room we built in the summer is really small compared to some of the ones you'll find at the Ranch or in LA so we monitor at a lower SPL (82 dBSPL for us) although the bigger rooms will monitor at 85 dB SPL. And yes, I still think 82 dBSPL is too loud for the room we built. I think anything above 80 dBSPL is freakn' loud unless you're in a big room.
The mixer I know was working on a documentary yesterday and from what I saw I'd say his average dialogue level wasn't going past -24dBFS and was probably going as low as -33dBFS. This level sounded just right compared to the rest of the elements in the doc. No LM100 or Dolby Media Meter was used to watch peaks. You can just slap a brick-wall on the DX bus and call it a day if you're worried about clipping. With dialogue mixed at that level, the brick wall won't be touched that much - just catching random peaks that get past the compression.
robotsound
Well-known member
What ryanjf said is exactly right. I calibrated my mixing setup using a WAV file containing pink noise with an RMS average of -20db. I have a little Radio Shack decibel meter, which I hold in the spot where my head will be while I'm mixing.
On my setup, I have an external sound interface and powered studio monitors with volume controls of their own on the back of the speakers. To set the levels, I started by turning the volume knobs on the speakers all the way down, and the master volume on my sound interface ALL the way up. Then I turn on one speaker at a time with the pink noise playback, and slowly bring up the volume of the speaker while measuring it with my decibel meter at the mix position. I want a reading of 83dB (C weighted) for each speaker in isolation. Once the first speaker is set, I turn it off and turn on the other one, and repeat the process.
Now, when I'm doing a mix, I just turn my sound interface all the way up and I'm ready to go. From here on down, the lower you set the volume knob, the less dynamic range your mix will have. The lower you set the volume on your monitors, the more you will turn it up in the mix. I find in my room I turn it down just a little bit if I'm mixing for something like DVD or if it's going to be playing at film festivals where the sound system is less than THX certified.
And finally, although rules were made to be broken, I find that my dialog usually ends up peaking somewhere between -20dB and -12dB on the meters, and everything else in the mix sort of falls into place around that anchor point.
On my setup, I have an external sound interface and powered studio monitors with volume controls of their own on the back of the speakers. To set the levels, I started by turning the volume knobs on the speakers all the way down, and the master volume on my sound interface ALL the way up. Then I turn on one speaker at a time with the pink noise playback, and slowly bring up the volume of the speaker while measuring it with my decibel meter at the mix position. I want a reading of 83dB (C weighted) for each speaker in isolation. Once the first speaker is set, I turn it off and turn on the other one, and repeat the process.
Now, when I'm doing a mix, I just turn my sound interface all the way up and I'm ready to go. From here on down, the lower you set the volume knob, the less dynamic range your mix will have. The lower you set the volume on your monitors, the more you will turn it up in the mix. I find in my room I turn it down just a little bit if I'm mixing for something like DVD or if it's going to be playing at film festivals where the sound system is less than THX certified.
And finally, although rules were made to be broken, I find that my dialog usually ends up peaking somewhere between -20dB and -12dB on the meters, and everything else in the mix sort of falls into place around that anchor point.
ryanjf
Well-known member
rogar-BigGrin(DBG) dialogue peaking in that range sounds right to me though they can go past there for film. For broadcast, it's a completely different story...
Noiz2
Sound Ninja
Ryan is on the money. There is a good sticky on calibrating a room on the Digidesign Users Forum that is very technical but very informative. It also has a link to Dolby Pink which is a good file to have. Some pink noise generators are not properly filtered (the one in ProTools being a sore spot) so it's best to use Dolby Pink if you can get your hands on it.
I'm still setting up the room here on the north coast, but I had my room in SF set up at 80SPL = -18dB Dolby Pink. I tend to mix at a lower level since most of the projects I'm mixing are just me in my small room and my ears get tired at louder levels. I will bounce up a bit sometimes just to check but I found that -80 in that room translated very well to the big screen. For video I will usually slap Massey's limiter on the master and compress and boost a bit.
Things to watch out for in mixing in a small room for a big screen are FX's like reverb. You can put a LOT more on in a small room than you can stand in a big one. I basically would apply to taste and then lower the REV return till I could just notice it and that seemed to come out about right.
I'm not sure what festivals are accepting these days, it seems to change hourly, but you should probably check what they want and can play. It used to be (when most things were submitted on film) that you had to assume everything would get passed through the "matrix" (Dolby stereo surround decoder) so you needed to check your mix "decoded" to make sure things didn't get thrown into the surrounds. I suspect that the opposite is true today and everybody is sending in stereo or possibly 5.1 tracks that never got close to a matrix encoder. IF this film is going out on actual film you do probably need to check your mix through a decoder AND get the mix Dolby SR encoded. The encoding can be done by the place that does the sound optical, but you will have to ask them to do it.
I'm still setting up the room here on the north coast, but I had my room in SF set up at 80SPL = -18dB Dolby Pink. I tend to mix at a lower level since most of the projects I'm mixing are just me in my small room and my ears get tired at louder levels. I will bounce up a bit sometimes just to check but I found that -80 in that room translated very well to the big screen. For video I will usually slap Massey's limiter on the master and compress and boost a bit.
Things to watch out for in mixing in a small room for a big screen are FX's like reverb. You can put a LOT more on in a small room than you can stand in a big one. I basically would apply to taste and then lower the REV return till I could just notice it and that seemed to come out about right.
I'm not sure what festivals are accepting these days, it seems to change hourly, but you should probably check what they want and can play. It used to be (when most things were submitted on film) that you had to assume everything would get passed through the "matrix" (Dolby stereo surround decoder) so you needed to check your mix "decoded" to make sure things didn't get thrown into the surrounds. I suspect that the opposite is true today and everybody is sending in stereo or possibly 5.1 tracks that never got close to a matrix encoder. IF this film is going out on actual film you do probably need to check your mix through a decoder AND get the mix Dolby SR encoded. The encoding can be done by the place that does the sound optical, but you will have to ask them to do it.
ryanjf
Well-known member
I second this tremendously. Setting reverb levels in small rooms can be extremely deceiving.
For Sundance 2012, they can take HDCAM. They "require" LtRt or a stereo mix (or even LoRo) on channels 1/2. According to the FAQ, their theaters have a Dolby Pro Logic decoder. What I've been doing lately is mixing in a 5.1 architecture, but printing the 5.1 and LoRo simultaneously. Much more time efficient especially on longer films. Though LoRo sounds funny..
Noiz2
Sound Ninja
So will the HDCAM auto detect the LtRt or is it going to run the LoRo through the Prologic decoder?
ProLogic is essentially the same as the Dolby Matrix box with out the SR decoding, so you would still have the possibility of your sounds getting sent to weird places.
FOR Those that have no idea what we are talking about...
ProLogic (and the Martix box) take four or more tracks of sound and "matrix" them together in a two track "stereo compatible" format that is know as LtRt (Left total, Right total). They encode it by making changes to the phase relationships and then detecting those changes and routing the audio accordingly. The problem that comes up is IF you have sounds that you have done phase related stuff to it can confuse the decoder. The classic is music that has a lot of phase shifting/ flanging (take just about any R&R from the late 60's) parts of the arrangement will fly randomly into the surround speakers. It's weird on the small screen and extremely distracting on the big screen. The fix is to start monoizing till it goes back to where it should be. But of course you can't tell with out playing it through the decoder and seeing where it ends up.
ProLogic is essentially the same as the Dolby Matrix box with out the SR decoding, so you would still have the possibility of your sounds getting sent to weird places.
FOR Those that have no idea what we are talking about...
ProLogic (and the Martix box) take four or more tracks of sound and "matrix" them together in a two track "stereo compatible" format that is know as LtRt (Left total, Right total). They encode it by making changes to the phase relationships and then detecting those changes and routing the audio accordingly. The problem that comes up is IF you have sounds that you have done phase related stuff to it can confuse the decoder. The classic is music that has a lot of phase shifting/ flanging (take just about any R&R from the late 60's) parts of the arrangement will fly randomly into the surround speakers. It's weird on the small screen and extremely distracting on the big screen. The fix is to start monoizing till it goes back to where it should be. But of course you can't tell with out playing it through the decoder and seeing where it ends up.
ryanjf
Well-known member
I am under the impression that the LtRt will be detected.
People reading this might want to check out Neyrinck's LtRt plug. It's $249 for CPTK users
People reading this might want to check out Neyrinck's LtRt plug. It's $249 for CPTK users