Preparing for DVD

[Gary Hughes]

The short answer is -12dB for tone and -6dB for peak on your FCP timeline. If outputing to anything other than a DV based device, I turn the levels of that device down till tone is sitting on -20dB. For DV based devices, leave it at -12dB.

And for the master volume control, go to the Audio Mixer under the Tools menu.

Now for the long version:

Let me start by pointing out some pages from the FCP 5 manual that you should read. If you don’t have FCP 5, you can still download the manual from apples site. https://manuals.info.apple.com/en/Final_Cut_Pro_5_User_Manual.pdf (43.6MB)

Page 54 0dB (Analog) Versus 0dBFS (Digital)
Heres an excerpt:

There are several common digital levels used to correspond to 0 dB on an
analog meter:

[Alexander Kallas]

Hi Gary, interesting post,below are a few comments

[garyh357] “What you’re dealing with here is a question of standards, not quality.

dBFS stands for dB full scale. It is the digital equivelant of a dB measurement. It simply means all bits on, or all bits off, or somewhere inbetween. All digital audio devices and media support up to 0dBFS and no more, in it’s final stage (more on this later). Therefore, in truth, headroom is not actually required by any digital media. In other words, it is possible to record audio at 0dB on DVD, miniDV, DV-Cam, iPod, etc…, without clipping (distortion). “

depends on many variables like compression that is used. For example, I can give you 2 bits of audio at 0dB- one heavily compressed and the other not. the compressed one can sound considerably louder- even though it’s at 0dB. In lay terms, we have ‘packed’ the headroom, squashed more signal in. The output will depend on if your player, speakers and ears can tolerate the compression that’s been packed into 0dB. It’s not correct to assume 0dB is all the same without considering all the variables.
As we all know, once digital clips it distorts

[garyh357] “Before I go on, let me say that I don’t advise it in all cases because it isn’t standard. Industry standards say to drop the level of your audio if you are outputing to tape, regardless of format. You should abide by the standards, because that is what keeps us compatible with each other. PLUS, it is what allows a straight dub of a digital tape to an analog tape without worry about overdriving the audio levels (that’s the true reason the standard was invented to begin with). It keeps you backward compatible”

not sure about that. 0dB is the standard, and when it goes to broadcast they all differ. For example the BBC put 0 dB and +6 PPL…. a good example of this is watching cable tv- the US will differ to the rest of the ‘normal’ world in volumes!

[garyh357] “In the case of DVD, it follows the same standards as CD. 0dB is actually the standard for PEAK audio levels on CD or DVD.”[garyh357] “A-Pak, Vegas, and other software that allow or suggest normalization during the digital compression stage, are assuming that you do not have your audio properly set to begin with. This is ok if you aren’t an audio person, or don’t understand all of this. They are “dummy-proofing” the audio settings for the users who do not know better. (Please don’t take that personal!) So why so far from 0dB (like -31dBFS)? My guess is that they are using the “RMS” value of audio, which is completely different from “Peak” values, and they really can’t be discussed in the same argument. It’s like saying “my apple is redder than your orange”, “no, my orange is oranger than your apple”. I will say this about RMS, be cautious when using an RMS scale to set the levels of your audio because it is possible for a negative RMS value to end up with a level that tries to exeed 0dB, therefore causing clipping (distortion). That’s why you like -31 better than -27. (If -27 allowed a +1dB level, then it would clip, but at -31 on the same file, it would only be at -3dB, which is safe from clipping.)”

Just about all audio programs allow you to normalise. There are also 2 audio standards commonly used (+4 and -10) for consumer and pro audio.

[garyh357] “Ok, above I said “All digital audio devices and media support up to 0dBFS and no more…” Let me prove this point. We all know that digital signal is made up of ones and zeros and nothing more. 0dB is simply all ones and no zeros. For a tape to require headroom, it would have to be analog. The tape itself can’t know whether it is holding ones or zeros. It doesn’t say “oops, there’s too many ones here”. All digital tape formats can hold all ones without zeros, it’s just that simple. Digital tape can’t clip or distort. For there to be distortion in audio on digital media, the distortion had to be there before the file was copied to the media. The tape, or other media, is simply a container, like a shelf. It holds the bits, like books. No audio is actually being recorded to the tape, just bits of data, and data can’t clip”

no, digital can distort. if you record onto it, certainly it will distort, if it’s a straight transfer of file then it won’t.

Cheers
Alexander

[Gary Hughes]

[Alexander] depends on many variables like compression that is used. For example, I can give you 2 bits of audio at 0dB- one heavily compressed and the other not. the compressed one can sound considerably louder- even though it’s at 0dB. In lay terms, we have ‘packed’ the headroom, squashed more signal in. The output will depend on if your player, speakers and ears can tolerate the compression that’s been packed into 0dB. It’s not correct to assume 0dB is all the same without considering all the variables.
As we all know, once digital clips it distorts

You may be thinking about this as RMS levels or similar to the way a human ear hears audio. When you talk about signal on a digital device, you MUST talk about PEAK not RMS. As I said before, it is possible for a negative RMS level to exceed 0dBfs. But 0dB peak, will always be 0dB peak, regardless of how much or little the audio is compressed. And as for compression, compression itself is distortion, simply because it changes the waveform. Too much of it certainly can sound like distortion. Again, I’m saying that if the signal is clean to begin with, and it doesn’t go above 0dBfs, it will still be clean on digital tape or a digital device. I’m certainly open to disussing dynamic range via compression and limiting, but you have to separate that from the “overdrive” discussion. They simply don’t have anything to do with each other. When you’re talking about peak level, peak is peak, 0 is 0, regardless of how loud it sounds.

Also, I’m not considering the player, speakers, and ears either, simply because this discussion is not about them, it’s about digital signal. You simply can’t record a clean signal that will distort a digital device because it is too close to 0dB. In order to have distortion on a digital tape, the distortion has to be there before the signal hits the tape (the bits are always undistorted). I’m not saying you can’t overdrive a digital device, they too have analog components, but I am saying that you absolutely can record all the way up to 0dB and it will always be distortion free.

[Alexander] not sure about that. 0dB is the standard, and when it goes to broadcast they all differ. For example the BBC put 0 dB and +6 PPL…. a good example of this is watching cable tv- the US will differ to the rest of the ‘normal’ world in volumes!

I will have to admit that I only work with US broadcast, so I’m just not aware of standards outside the US. I guess I should have mentioned that. However, the real point that I was making is that whatever the standard is, you should abide by it. The above discussion about what digital tape CAN handle was only to state that it for sure can, not that that’s what you should do. YOU SHOULDN’T! Use the standards.

[Alexander] Just about all audio programs allow you to normalise.

True, but to suggest that you must use it at a certain level is the same as saying you must use the “Broadcast Safe” filter that ships with FCP. It just isn’t true. If you’ve done your job and set the levels properly, then you don’t need to use it in the final output stage. That’s not to say that normalizing is always a bad thing. I use it myself, but not at that stage in the game.

[Alexander] There are also 2 audio standards commonly used (+4 and -10) for consumer and pro audio.

Those are not audio “standards”. That refers to the level of the signal that travels over ANALOG cable. It has nothing to do with “digital” and it has nothing to do with the levels that you set on the meters.

[Alexander] no, digital can distort. if you record onto it, certainly it will distort, if it’s a straight transfer of file then it won’t.”

Again, you’re confusing the analog circuitry on a digital deck with a digital signal. It is absolutely impossible to record a digital signal hotter than 0dBfs AND it is absolutely impossible for the “DIGITAL” signal to distort because it is simply bits of data. Certainly, if you overdrive the analog circuitry in a digital deck, you’ll get distortion, but again, it is going to be “ANALOG distortion that is converted to a clean digital signal. When played back, you’ll simply hear a very clean, undistorted and pure, digital recording of analog distortion, although, I’ve never seen a pro or even prosumer deck that will overdrive below 0dBfs, providing you are sending clean signal over the correct cable type and the input levels are set correctly, on the back, to either +4 or -10 respectively. And if you’re using firewire, it is a straight transfer, just like copying a file on a hard drive. Also, since the beginning of the CD player, all consumer level digital devices can handle ouputting signal up to 0dBfs without distortion as well.

Simply put, there are a lot of places in the digital workflow for distortion to be created, but it will never be because of the tape or storage medium itself. For 16 bit audio, 0dB is represented as 1111 1111 1111 1111. Since digital is either a 1 or a 0, anything other than all 1s would equal a signal less than 0dBfs which would also be clean. For distortion to digitally exist, it would have to be 1111 1111 1111 1112 or higher, and we all know that’s not going to happen.

And to bring this back home, the original post was not about a deck, or any type of device with analog circuitry, or even tape, it was about outputting to DVD. Take that audio level right up to 0dBfs for your maximum peak in your program and put it on the DVD, don’t leave headroom because it simply isn’t needed. And if you’re outputting to tape, use the broadcast standard for your country, and all will be good.

Going back to the A-Pak app for just a second, and again, I’ve never used it because I made the switch to FCP, just this year. I started with version 5. If you could set it to normalize to the maximum peak value, you could set it to 0 and know that you would be OK, but if it only looks at RMS, I wouldn’t use it at all. I’d use STPs normalize, like in my workflow that I posted earlier.

I hope this helps,
Gary