Audio level for dialogue and music

Hi,

I am piecing together clips and trying to maintain the same audio level. Some clips are very soft and others peak a lot. My understanding is that dialogue should average 0 db in the FCP audio meter and music underneath should be 6 db below the dialogue. Is this correct?

> My understanding is that dialogue should average 0 db in the FCP audio meter and music
> underneath should be 6 db below the dialogue. Is this correct?

If your reference tone is -12dB, your levels should peak around -8dB. 0dB would be well into the red. How loud the music is, however, is really dependent on how loud your dialogue is, what the frequency content of the dialogue and music is, and the mood you're trying to create. Also, if all your music, sound effects and dialogue from scene to scene are the exact same level, you'll probably have a pretty flat overall sound mix. The numbers help, but a lot of it is down to your ear.


www.derekmok.com

Yeah, I am 'playing it by ear' regarding the music. There is a REALLY big difference in the loudness of some clips and others, that is why I am trying to understand what the average should be, because I don't trust my ear, not the volume on the speakers. So if I understand right, the tone during BARS and TONE is set to -12 db, which should be the average of the dialogue, and the sound should not hit -8 at all if it can be helped, right? This is totally new to me.

This hasn't come up in a long time.

[www.kozco.com]

It's a stiff download (I gotta fix that one of these days) but it shows broadcast VU meter (Betacam SP) properly adjusted and a music track in proper relationship. Stick with it because it gets louder and more interesting near the end. After you render to your DV timeline, your Final Cut bouncing ball meters should track, and that's really what you're watching here.

The VU meter is interesting because it's only correct in a very narrow range of loudnesses unlike the Final Cut meters which can be miles off and you might never know.

Note at the beginning of each video, the Final Cut -12 is adjusted--at the tape machine--to the broadcast -20 standard before the recording is made. You should stick with the -12 because that's the DV standard.

Even that will only get you in the ball park.

You do have to use your ears and use the meters as rough guidance. The Final Cut meters are peak reading and they don't even catch all the peaks. This is very nearly useless for show balancing, but it does tell you when you're in trouble.

Never allow the sound to hit zero, ever. Zero is the point at which there is unrecoverable sound damage, crunchy, ticky sound when you didn't record either crunchy or ticky.

You will find that human voices vary widely on the meter depending on the microphone and the actor. Listen to them.

Half or double loudness is roughly 18 to 20 on those meters. Go up 18 and the show will seem to have doubled in volume.

And yes, as a very rough, fuzzy guideline, most show peaks should be in the -6 to -4 range.

After you finish your show, listen to it in a dark room. Ignore the meters. Way most of the complaints we get for sound are from people listining in their office and notice an evil change in volume *to their ears*. This may be the first time you ever actually listened to a television show. It will wreck your regular TV watching habits.

Koz

Koz,
Thanks..very good infg that answers questions I also had.

Me thinkith that this would be a good post to put in the FAQ for audio.

Very very helpful info. So I take -12 db to be the average, establish -6 db as peak, make adjustments accordingly, and when the program is ready to be transferred to Digibeta, I bring the tone down to -20 on the tape machine and everything follows suit.

I was told that the sound may be have to be 'dial normed' to -27 db, meaning dialogue level is brought to -27 db via a special machine. Is the process still the same, i.e. leave all at -12 db and worry about pull down when ready to make Digibeta tape?

I'm not quite sure that's exactly what Koz was referring to. If you master to -12 bars and tone, then your show should average at -12. If you send out -20 bars and tone(which can be adjusted), then your program should be set to average at -20 inside FCP. If Koz is changing the recording levels at the deck, this is a different thing. Just changing FCP's bars and tone does not change the audio levels of your program at all. The bars and tone and your audio mix should always match- no matter what your output.

Andrew Balis
Lead Final Cut Pro Instructor
www.moviola.com

Hi, Andrew:

What I am trying to understand is, if I set the tone to -12 and all the audio is set to average -12 (peaks at -6) as everyone has recommended, if I want to submit the program to a station where they require -20, do I need to go in and manually set the tone to -20 and do the same with all the audio, before I transfer to Digibeta?

From what Koz said, I thought that the engineer can simply drop the audio output level to -20 while outputing to Digibeta, meaning I don't have to manually go in and adjust the audio to the entire sequence.

if you want to submit your work to a station that requires -20db line up level then you need to make a dub of your programme, first adjusting the source deck's play back level, so that the tone registers -20db on the record deck, or by first adjusting the record decks rec level, so that the tone from the source deck register at -20db on the record deck

> if I set the tone to -12 and all the audio is set to average -12 (peaks at -6) as everyone has
> recommended, if I want to submit the program to a station where they require -20, do I
> need to go in and manually set the tone to -20 and do the same with all the audio, before I
> transfer to Digibeta?

No, you don't. The important thing is that your show is mixed properly in relation to the tone, and that the standard for your levels matches the output format.

So, if you're using -12dB reference tone, you'll be mixing your sound levels in relation to that. For example, you have a huge explosion that you're mixing to -6dB there. When you're using that timeline to output to DigiBeta, which wants a -20dB tone, then while playing back the -12dB tone to the deck, you adjust the input level of the deck so that the -12dB tone is being recorded at -20dB, a difference of -8dB. That means your whole show is being recorded 8dB lower and your explosion is now at -14dB.


www.derekmok.com

Thank you for the clarification. That was what I surmised by previous posts, but I wanted to make sure I understood correctly.

And like it said on the explaination, if you encounter a deck that will not adjust its input levels (or you can't reach the deck), you can pull everything down 8 dB on the timeline and output that. Same difference except it's a lot more work than simply pull out the knobs and adjust. If you do it on the timeline, you now have a non standard timeline and you have to remember to put it back. And has been pointed out, you have to pull down the whole timeline tone and all.

The DV timeline is always produced the same way.

There was one variation that was popular for a while, and that was to put the tone at -12 and put the show peaks there, too. If you got a station or customer used to you doing that, it worked out because they simply ignored your tone. The show sound was at broadcast levels, not DV levels. The problem was the first time you hit a "foreign" or new client that rediced your tone to -20 and your whole show would go into the toilet.

And, as I'm finding out by making that video, the explosion could go as high as -4 or -3 or so and still pass muster. -6 may be slightly too conservative for special effects peaks.

Koz