ok....long time question

[narcoman]

have a look at the latest info on 1770 - that's where it's all heading.

[gsilbers]

this question should be in the post production forum.

[Jay Rose]

Digital delivery for analog broadcast is pretty well set at -20 dBFS tone and average dialog, -10 dBFS absolute peaks.

Yes, it's low. It throws away a bit and a half of resolution. But Sony decided that -20 dBFS would equal +4 dBu on the outputs of their digiBeta decks, and 10 dB headroom was determined safe, so that's what got written into network contracts.

Digital delivery for digital broadcast, which is virtually all new production, relies on 1770 or Dialnorm. But those are longterm weighted averages of key elements, not the "RMS" you were asking about. Search this forum for more info.

Yes, DVDs are generally mastered louder. And films are nominally -20 dBFS for dialog, which is supposed to reflect 85 dB SPL in a theater seat, but it's really up to the director. And the only limit on headroom for films is Fullscale.

[danijel]

Jay, I've learned from your book, (and I'm thrilled to have your input here at the Gearslutz), so with due respect, I'll point out that these lines I've quoted, are sligtly off

Originally Posted by Jay Rose

.... -20 dBFS tone and average dialog ....
.... Yes, DVDs are generally mastered louder ....
.... And films are nominally -20 dBFS for dialog, which is supposed to reflect 85 dB SPL in a theater seat ....

All the questions asked by the OP are addressed in this post:
Standard mixing levels for movie theater, DVD, broadcast TV, commercials etc

[Jay Rose]

Danijel, I certainly appreciate your comments and link to deeper discussion. That'll probably help the OP more than I did.

But what do you mean by 'slightly off'? My statements in this thread are abbreviated, and may be limited to US practice (the OP is from New Jersey), but they're true for the qualifications I included with them.

At least I think so. Which do you have a specific quarrel with, and what do you consider a more accurate value for those situations?

[danijel]

Ah yes, I was in a bit of a rush myself when typing that last post.... What I meant is, those sentences are imprecise, so they could be interpreted in different ways - eg. when you say '-20dBFS average dialog' you don't say if you mean PPM, VU, RMS, or plain digital peak. I'm pretty sure most novices would interpret that as 'digital peak', so they would end up limiting dialog at -20dBFS....
I might be nitpicky here, but I know how much time it took me to comb through piles of loudness and calibration related posts throughout the internet, decipher them, and actually pick out those that are precisely written, accurate and not subject to interpretation. I'm thankful to guys like Tom Hambleton, Marti Humphrey, Pascal Sijen etc, whose posts are always reliable, and who have corrected me at times, so I've become sensitive in these matters... I'm sorry if I sound like a loudness-policeman ( ), but I've learned so much from these forums (especially Gearslutz), so that I feel like the Amazon forests are dying when the questions are not answered with care

[Jay Rose]

when you say '-20dBFS average dialog' you don't say if you mean PPM, VU, RMS, or plain digital peak. I'm pretty sure most novices would interpret that as 'digital peak', so they would end up limiting dialog at -20dBFS....

Fair enough. I did say "-10 dBFS absolute peak" immediately after, but it could be misinterpreted.

The -20/-10 spec is from just about every US network contract I'd seen, starting when DBeta became common and ending just a year or so ago, and I did specify "digital delivery for analog broadcast". But again, fair enough: a novice skimming the thread could read that wrong.

(FWIW: I mixed a film with Marti in his studio, about seven years ago. Back then - in digital delivery for analog release days - he seemed to be measuring loudness the same way I did: with Dorroughs, looking for where they change color at -20 dBFS. And of course actually mixing to his calibrated ears, using calibrated speakers.)

[digitalson]

Doesn't that sound low -20 dbs ....that's lower than most cds out there....if the peak is that what. Should the rms value or rangee be at? Not much room to work with....also losing so much res from digital.......hmmmmm

[Henchman]

Nope. Rms -20 for dialog is correct. In fact, for film at 85 I think that's a bit hot.

[Jay Rose]

Doesn't that sound low -20 dbs

Yes. And yes, it's lower than almost any audio CD, mp3, podcast, or on-air broadcast...

any other questions?

--

We got the -20 spec in production recording for practical reasons: to allow enough headroom for sudden peaks in a performance, on early equipment where even a couple of milliseconds attempting to record above fullscale would give you horrible clicks.

We got the -20 spec for post because Sony designed their Digibeta decks so that -20 dBFS on the tape would generate +4 dBu on the analog output... the same level that 0 VU would in an analog Beta deck. This was a day when facilities were just starting to convert, and digital decks were often wired to analog consoles or patchbays.

Ditto -10 dBFS absolute top for peaks. This was considered the same practical amount of headroom above "zero level" -- i.e., -20 dBFS -- that you'd get from a good analog deck. (Headroom in analog is a slippery topic because there's no definite point where distortion becomes intolerable, so long as you don't hit the power supply rails... but that's a topic for some other thread.)

I deliberately wrote "top for peaks" rather than "limit". There's nothing in the spec that says you have to use a limiter, and certainly nothing saying you have to slam that limiter and destroy dynamics. The limiter is just a convenient tool to assure safety in an otherwise properly done mix.