Settle a bet…

[Kevin]

HUH?
“Digital Audio can’t distort”…. try it, you won’t like it.

MR. Old School is correct.
0DB analog
-12db digital

[Gary Adcock]

[Jeff18876] “He says we should peak our audio at -12db in FCP and then at 0db on Beta. he says that if we go above -12db in FCP then the signal sounds distorted. Which I think is mostly due to the junk speakers he has in his suite. “

Sorry no meat on that bone.. the old guy does know what he is talking about.

Part of the problem is the difference with analog audio (in SP) vs Digital audio (in FCP). Remember that analog audio could (and did) go above 0db whereas Digital Audio (in FCP) clips hard above 0db.

So in the BetaSP analog world audio can go above 0db, going above 0db would cause distortion but not clip until much higher. In the Digital Realm 0db is a HARD ceiling. I actually have my clients set their audio to -18db instead of -12db.

Gary Adcock
Studio37
HD and Film Consultation
Chicago, IL USA

[J. Nedy]

Actually, I think your friend is right.

Funny thing is, I had the same situation not long ago. Anyways, an anologue meter is more of an averaging meter and a digital meter is a peak meter, meaning the anologue meters takes into account 20db of headroom, while the digital meters don’t need this.

If your hitting 0db on your PVW-2800 meters that means you are in a good sound range and have some headroom if you need it. 0db’s in FCP means your distorting your audio, and you DO have digital distortion becuase after all, if not, then why have the red at all?

Once my co-worker explained this to e, I gave it a try, and low an behold, he was right, now I edit everything so it bounces around -12db, lay it to tape at 0db, and it sounds wonderfull, no distortion, and the sound floor is way far below my range that it’s not much of an issue.

[David Rowan]

On the DVC Pro D-650’s and D-750’s that I use the audio meter has a little arrow marking where audio should peak (like for tone) that marker is at -20db. I have been told that digital audio clips and sistorts much worse than analog

DWR

[David Roth weiss]

Jeff[Jeff18876] “This way we will get the most dynamic range and fullness of sound out of our program.”

Jeff,

In addition to what the others have said, this concept doesn’t translate to digital audio in the same way it does with analog audio. In the analog world, volume or “amplitude” affects variables such as dynamic range and signal to noise ratio. In the digital world however, increasing amplitude simply raises the amplitude. Of course this is a relatively simplistic explanation, and we could get into a very deep discussion of these things, but thats a whole kettle of fish best left for another discussion.

BTW, in addition to creating very audible distortion, severe clipping of digital audio files can actually corrupt the file itself.

DRW

[Bryce Whiteside]

I thinking you both should read this article Broadcast Engineering Managing Audio Levels.

[Jeff18876] – I keep trying to tell him that our audio in FCP (or any DIGITAL application for that matter) should peak at just below 0db, the very top of the meters in FCP, just before they peak and lite up the red-peak-dot.
The mean level should hover around -12 with excursions above -12 which on an analog BetaSP machine would be 0 VU with excursions above 0 VU.
According to p. 628 of Apple Pro Training Series : Advanced Editing and Finishing Techniques in Final Cut Pro HD (2nd Edition) by by DigitalFilm Tree and Michael Wohl:
Dialogue: Common Average and Peak Levels in Final Cut Pro
Delivery…..|….Average….|….Peak
Format……|….Volume…..|….Volume
Broadcast TV____-12 dB__________-6 dB
VHS tape_______-12 dB__________-3 dB
Theatrical_______-20 dB__________-3 dB
Dolby Digital

He says that if we go above -12db in FCP then the signal sounds distorted.
Well actually when they hit 0 dB they get clipped but if they stay between -12 dB and 0 dB they are fine but that includes all transients in the audio even small ones that are loud but extremely narrow in time.

Short answer is peaking should occur between -3 dB and -6 dB on the audio meters adjacent to the sequence timeline. 1 Khz tone reference should at -12 dB in FCP and 0 VU on the analog BetaSP with peaking occuring in program audio above these values.

I think no one won this bet because the actual answer was in the the middle of both of you arguments.

HTH with my 2¢
Bryce Whiteside

Don’t worry Mr. B. I have a cunning plan…

PowerBook 1.67 Ghz ATI 9700 128 MB 2 GB
Final Cut Pro HD
DVD Studio Pro 3
Motion

[Shane Ross]

I do what Bruce does. I have my audio hover around -12, and peak at about -6.

Shane Ross
Alokut Productions
http://www.lfhd.net

[Jeffrey]

Well, I stand corrected. This will put a grin on my co-workers face.

But let me ask this:
I came to my now de-bunked theory by a problem we had when we did our audio the way you all stated here. We would calibrate our FCP’s tone at -12 in FCP and 0 on Beta, but we found that when we do this, as i have just tested on a spot i just finished, that when your tone in FCP is at -12, and you set your Beta to 0db by that tone, and you set your program audio to bounce at -12 in FCP then it bounces at like -12 on the Beta deck, wayyy below the 0db mark.

So even though the tone went through at -12db (FCP) and 0db (Beta), and the audio is bouncing at -12db in FCP, the program audio is wayyy too low on the Beta deck!!! Am I missing something here?

BTW, we route the signal from the MAC thru Firewire into a DSR-45, then out of the DSR-45 via XLR to a Mackie 1402-VLZ then, via XLR, to a UVW-1800. (The meters on the Mackie read 0db with tone and wayyy below 0db during program audio)

Jeff

[Chris Tomberlin]

[Jeffrey] “So even though the tone went through at -12db (FCP) and 0db (Beta), and the audio is bouncing at -12db in FCP, the program audio is wayyy too low on the Beta deck!!! Am I missing something here?”

The real issue here is the difference between the analog and digital worlds. We still have one foot in each. Contrary to an earlier post, things like signal to noise ratio and headroom still apply in the digital world but in a different way. The “noise” in digital audio is dithering noise, for instance. I know this may be unconventional, but I master to a digital tape format from FCP with reference tone at -20 and program audio peaking anywhere from -6 to -3. When I make a BetaSP or other analog dub from that digital master, I adjust tone on the analog tape to be 0db with program peaking around +3. I do this by using the playback volume knobs on the digital deck – when they are popped in (unity), the -20 tone is 0db on the beta deck, but after the tone plays I pop them out to an adjusted lower level that puts the program on the beta deck peaking around +3. This method gives me the best signal on both the digital and analog tape formats.

Chris Tomberlin
Editor/Compositor/Owner
OutPost Pictures

[Tony]

Jeff,

First I think you need to beg forgiveness from your elder for being to quick to rush to judgement about his technical advice. Digital unlike analog does not have anywhere near the same headroom.

From what you describe somewhere in the signal flow chain (most likely out of the dvcam deck to the Mackie mixer) you are no longer at unity reference which results in analog levels being too low on the betacam deck.

I suggest the following plan to make your life simple which is to go back to your senior citizen editor and beg him to instruct you on how to properly establish unity reference throughout the entire signal flow.

Once a unity reference level is established whether it be digital or analog what was recorded will be played backed at the same levels as the original.

One last item ask him to explain the differences between VU meters and peak meters. The latter will reveal alot of the mystery you are experiencing.

Good luck,

Tony Salgado