another question about calibration !!!

[vudoo]

I just finish mixing a 1hr45min documentary at my studio. The calibration was -20dbFS = 0 VU = 79 db SPL at mix position, i have a very small room ( 10x11 ).

Anyways, the film was edited on Final cut pro so i sent the video editor the final mix and he ouputed the final master via firewire on a DVCAM. Now, they also need a Betacam SP. This is where i'm a bit confused, since there's about 12 db of headroom on most Beta SP, should the transfer be made at - 12 dbFS = 0 VU so i won't overload the Betacam ???
Does this mean we should make a master DVCAM with a -20 dbFS tone and one with -12 dbFS so that depending wich medium transfered to, we're covered ?? Thanks a lot.

[Richard Fairbanks]

No, a sinetone from Pro Tools at -20dbfs should read on the Beta meter's "0 VU" marks. Adjust the Beta decks record levels to make that happen. Your mix's peak levels should be limited to -10dbfs (for 10db of headroom). Hopefully your existing mix is already in the ballpark and can go straight onto the beta tape. Beta SP's tend to have a soft clipping so if you don't limit the deck will anyway, and I think it is better if you do it. If you've done your mix right those awful beta deck meters will jump around between -10 and +3, generally higher more than lower. Those meter ballistics are really strange, especially on the cheaper decks, and seem to be somewhere between peak meters and RMS meters. I've never been able to figure them out.

[vudoo]

Richard,

Thanks for helping. OK a few more questions.

Your mix's peak levels should be limited to -10dbfs (for 10db of headroom)

Why not use the whole spectrum, i mean when you calibrate it at - 20 dbfs that means you give yourself 20 db of headroom so why have it peak at -10 dbfs max ??? Actually, my mix peak at around -5 to - 3 dbfs with dialog dancing around -22 to -20 dbfs. Not using the last 10 of headroom seems odd to me, is this just for safety reason so that your mix can translate on as many medium as possible ?? thanks

[philper]

Richard,

Thanks for helping. OK a few more questions.

Your mix's peak levels should be limited to -10dbfs (for 10db of headroom)

Why not use the whole spectrum, i mean when you calibrate it at - 20 dbfs that means you give yourself 20 db of headroom so why have it peak at -10 dbfs max ??? Actually, my mix peak at around -5 to - 3 dbfs with dialog dancing around -22 to -20 dbfs. Not using the last 10 of headroom seems odd to me, is this just for safety reason so that your mix can translate on as many medium as possible ?? thanks

Well, you are out of spec for most broadcast networks. If you send this mix to a network it would very likely be returned to you by the QC dept. If you have zeroed @ -20 and have 20 db of headroom you have too much dynamic range for TV and for most theatrical situations. If the loudest parts of your mix are made bearable to listen to, the lower parts will be too low. Using up all the headroom is a music recording thing, especially in the current volume wars in which everything has been squashed into the last few db of available level. Since film/TV audio has to be able to work on anything from a laptop speaker to a big theatre, vhs to 35mm Dolby, we generally limit our dynamic range to what Mr. Fairbanks stated, unless explicitly told we shouldn't.

Philip Perkins CAS

[Richard Fairbanks]

Everything Philip said. Broadcast is a whole new world for you, and in fact the headroom restrictions are less strict than they used to be. It makes no sense unless you understand the history a little, and the analog equipment restrictions that started it all are still in place here and there. Home viewing requires a more narrow dynamic range for a number of reasons, including the way human hearing seems to work and the general ambient noise level of a typical home. For lots of reasons the "target program reference level" has been chosen to be -20dbfs or -18dbfs, depending on what country you are in. USA is -20. Limiting the upper dynamic range has served two purposes, mostly to prevent analog broadcast problems as well as a residual benefit of keeping all program's dynamic range more narrow. (I won't even go into the reasons why this "residual benefit" hasn't been very effective.) It also causes program peaks to generally be around -10dbfs, although this restriction is sometimes lifted a little depending on the network that takes it. Again, this is the way things are, not the way that may seem logical on first glance. There are probably a dozen long threads about this exact subject that have been posted here over the last year or so. Might I suggest you start exercising your search skills?

Your peak levels are not too far off, a little high, but first we should make sure your dialog levels are in the right ballpark. Are you saying -20 to -22 as displayed on an RMS average or VU meter? Or do you mean on your peak meter?

[vudoo]

Thank you guys for the explanation...yes you're right, i come from the music side of mixing so all these restrictions are new to me. This is my second full length documentary mix...the first one did very well, so i guess i was lucky.

The film the i just finished has the dialog dancing around - 4 to + 2 on a VU meter and the loudest peak of the mix will hit at -5 to -3 dbFS on my ProTools. So should i just lower my overall mix or maybe limit the mix a few db so that it won't peak above - 8 dbFS. Thanks again for your help.

[Sonsey]

Thank you guys for the explanation...yes you're right, i come from the music side of mixing so all these restrictions are new to me. This is my second full length documentary mix...the first one did very well, so i guess i was lucky.

The film the i just finished has the dialog dancing around - 24 to - 18 on a VU meter and the loudest peak of the mix will hit at -5 to -3 dbFS on my ProTools. So should i just lower my overall mix or maybe limit the mix a few db so that it won't peak above - 8 dbFS. Thanks again for your help.

Lucky maybe...As I discovered the hard way when I started out, some smaller broadcasters (who shall remain nameless because I still do a lot of work for them) will NOT reject the tape but simply transfer it at a lower volume. For a while I wondered why mix rich, full mixes from my room sounded like s*%t on air. You are lucky to have the wise folks at this forum who can help you out where I just had to take my knocks. On the flip side I won't EVER forget those lessons

I also think you've confused VU and PEAK meters. If your dialog were and -24 to -18 on a VU it would be REALLY low. Do a quick search on VU vs PEAK to learn more. The meters on ProTools are peak meters.

[Richard Fairbanks]

Start by putting a brickwall limiter at -10dbfs and listening. (high freq transients of -8 will never make it through some broadcast chains, or off of betaSP). If it sounds like you're crushing things too much then lower the overall mix (before the limiter of course) 1 or 2 db. I don't know how your average vs. peak levels are at the moment. The idea is to keep average (VU) levels high and peak levels low. Sometimes it can be a battle, sometimes not so much.

[vudoo]

Sonsey,

Oops it was a typo, you read my post as i was correcting it. I meant - 4 to + 2 on a VU for dialog.

Richard,

My average is always dancing around -4 to + 3 on VU and the few peaks i have will hit -3 dbFS ( PT peak meter ) so yeah, i will brick limit it first...because i feel the average program level and dialog is OK.

BTW, what is the ballpark for dialog, is there a standard in the filmm industry ????



Thanks.

[Richard Fairbanks]

Yeah, there are well established standard practices for feature work, and emerging or changing standard practices for broadcast. Check out the room calibration "sticky" post at the top of this forum. It is long and convoluted but contains HUGE and mostly accurate information. There are well defined standards for film theater sound playback, and none for home playback, even though there are electrical standards for the mixes we deliver to broadcasters. Mix levels, mix dynamics, playback levels, speaker eq calibration, metadata, all of these terms and factors are inter-related. Often, big movies dialog is far lower than yours, leaving much room for effects and music to drop your socks. Dialog VU on the order of -12 to -7, maybe -5VU in a loud scene. It is more of what SOUNDS right, based on playback speakers being set to known levels and frequency response(check out that sticky!!!). Their practice is NOT very appropriate for broadcast, not currently anyhow if ever. This all starts to touch on other things you need to learn about, like Dialnorm (visit www.dolby.com). Without going into details, a dialnorm reading for typical broadcast work is around -27 to -24 db. Dialnorm for a typical Hollywood feature is more like -31, which is 4 to 7 db softer. These are generalizations of course.