What genius can help decipher delivery specs?

[djh108]

Hi,


I primarily do music production, but occasionally get gigs that include post mixing. I'm mixing a cable TV show this week and I'm having trouble understanding the delivery specs, hoping those wiser than I can enlighten me. Here they are the specs, verbatim:


" 2.1.8 Mixing
Episodes that are delivered with poorly mixed audio will be rejected.

• § Average audio levels should be recorded at -12dB, including whispered voices.
• § Transient audio, like gunshots, are the only exception: they can peak up to -8dB.
• § Absolutely no audio should rise above -8dB, ever. Any audio above -8dB will likely be distorted by
  the compression added by our carriers.
• § Audio must be mixed on two channels in stereo. Stereo audio must be fully mono compatible. The
  audio channels must be in proper phase.
• § Program audio must reflect 1 kHz reference tone level set to -12dB. Audio levels must be consistent
  throughout the program with norms averaging -12dB
  NOTE: Full Mono compatibility means that when the left and right stereo channels are actively combined to mono, there is no discernible change in audio level or fidelity
  Summary of the 3 key audio levels you should pay attention to:
  • § -12 dB: reference 1kHz tone, normal audio level – this is where most audio will be (including whispering).
  • § -8 dB: maximum peaks; no audio should ever go above this level.
  • § -20 dB: minimum audio level. No audio should fall below this level, as it will be inaudible on a
    viewer’s television.

Sooo...I guess I'm to assume their values, such as their overall avrage of -12db is an RMS value (or LUFS)? Isn't that crazy hot for broadcast? And then they say nothing can peak above -8db...seems to me that ratio is going to compress the living daylights out of the material to have an average of -12 and no peaks above -8. What am I missing here??



Thanks folks!

[its2loud]

Time to call or email someone at the network and get it straightened out.

That's what I would do. Most delivery specs are wrong or outdated. That has been my experience.

[minister]

These look like standard PEAK values, NOT RMS or LKFS. These look old skool. But, pretty simple.

[philper]

And you are going to be making a mighty compressed mix, mon ami.

[papalou]

These specs are crazy.
Lack of precision ("average" ? RMS, LUFS, LKFS ???), and a compressed signal that'll give headache in seconds.

[tom_lowe]

Anything below -20dB won't be audible, well assuming they mean -20 dBFS that is insane, I mean who wants backgrounds or quiet parts that high.

Awful spec.

[TrentWilliams]

Originally Posted by tom_lowe

Awful spec.

+1

[tom_lowe]

Originally Posted by djh108

Hi,


I primarily do music production, but occasionally get gigs that include post mixing. I'm mixing a cable TV show this week and I'm having trouble understanding the delivery specs, hoping those wiser than I can enlighten me. Here they are the specs, verbatim:


" 2.1.8 Mixing
Episodes that are delivered with poorly mixed audio will be rejected.

• § Average audio levels should be recorded at -12dB, including whispered voices.
• § Transient audio, like gunshots, are the only exception: they can peak up to -8dB.
• § Absolutely no audio should rise above -8dB, ever. Any audio above -8dB will likely be distorted by
  the compression added by our carriers.
• § Audio must be mixed on two channels in stereo. Stereo audio must be fully mono compatible. The
  audio channels must be in proper phase.
• § Program audio must reflect 1 kHz reference tone level set to -12dB. Audio levels must be consistent
  throughout the program with norms averaging -12dB
  NOTE: Full Mono compatibility means that when the left and right stereo channels are actively combined to mono, there is no discernible change in audio level or fidelity
  Summary of the 3 key audio levels you should pay attention to:
  • § -12 dB: reference 1kHz tone, normal audio level – this is where most audio will be (including whispering).
  • § -8 dB: maximum peaks; no audio should ever go above this level.
  • § -20 dB: minimum audio level. No audio should fall below this level, as it will be inaudible on a
    viewer’s television.

Sooo...I guess I'm to assume their values, such as their overall avrage of -12db is an RMS value (or LUFS)? Isn't that crazy hot for broadcast? And then they say nothing can peak above -8db...seems to me that ratio is going to compress the living daylights out of the material to have an average of -12 and no peaks above -8. What am I missing here??



Thanks folks!

you don't need a genius to decipher them, you need a blithering moron to make sense of it as they were clearly written by one.

[djh108]

Thanks so much for jumping in, guys. My email notifications for this thread aren't coming for some reason, so I didn't realize these responses were here until just now.

Originally Posted by its2loud

Time to call or email someone at the network and get it straightened out.

That's what I would do. Most delivery specs are wrong or outdated. That has been my experience.

I actually tried that yesterday, wasn't a tremendous amount of help as we were clearly speaking different languages, but my initial guess was somewhat affirmed by the confusing conversation I had, which is basically this...

Originally Posted by minister

These look like standard PEAK values, NOT RMS or LKFS. These look old skool. But, pretty simple.

So my game plan is to just do a balanced mix and use those peak values as my ceiling. The network contact said its rare to have a mix rejected, and I got the sense his motivation for writing things that way simply to avoid getting mixes from novices that are jacked and limited out to wazoo (which, ironically, is exactly what those specs would create if taken literally).


That said, sounds like you've seen similar specs from yesteryear, Minister. Wondering if there's any advice you'd offer mixing to a situation like that...does indeed sound like its an old skool setup with older gear involved...Sounds like the only loudness control they do is hard limiting the ceiling, but nothing to normalize mixes that average too soft..would you approach it the way I described above if you were in my shoes? Or maybe lean it a little on the compressed side? - thanks so much...

[its2loud]

going back a few years to -10 peak specs for most networks, the usual method was to throw a brickwall limiter on all DIA, MX, FX, and FULL MIX Busses.

Regarding the " § -20 dB: minimum audio level. No audio should fall below this level, as it will be inaudible on a
viewer’s television. " spec. I would take this to mean that no spoken dialog should fall below this number. There's no way you're gonna drive up music or BG's to that level without it sounding ridiculous within the content of the show.

The things that really puzzles me is the -12dB reference 1K tone spec. I think they just mean they want all the dialog sitting around -12. Which is about where it would sit if you were mixing to around -24LKFS.

Again, too much is open to interpretation and guessing. These specs are horrible. Good luck