Reference Levels ( -20=0 vs the Dynamic Range)

[Diegux]

Hi everyone
I want to ask the forum a question I can't figure out by myself regarding reference levels.
I work in a post production place. All our rooms are calibrated @ -14....that seems to work fine when we do radio. We have a D-Beta machine and an Avid system,which are calibrated @-20.We usually turn down 6dB on everything that is taken digitally from the rooms (@-14) in order to layback at the D-Beta.That seems to work fine cause everybody is happy with it. But it still doesn't make much sense to me
For example, yesterday I had to layback to D-BETA a DAT tape with a mix from a famous latin singer (that was obviously peaking almost @ -0 dB FS , and in order to respect the -20 reference level of the video machines, I had to turn down the mix.... +20 (analog) would cause a distortion....but turning down the level I lost dynamic range...the peaks were far from 0 dB FS
I can't seem to find an answer to this....why would you calibrate in such a way that makes you loose Dynamic Range...I understand that more headroom is needed in the digital world...but this is not making total sense to me...am I missing something or is the -20 thinking "in order to have more headroom you loose dynamic range"?
Whoa!That was long...thanks a lot for your interest and answers
Diego

[Diegux]

By loosing this "low level" you are in fact loosing dynamic range.Actually you are loosing more than a bit (6dB) resolution of your digital system dynamic range when you peak so far from 0dB FS.
-20dB FS=0 VU is still not making sense to me....does anybody else have any thoughts on this?
Thanks again
Diego

[Dylan]

It seems to me the question is what the hell are the reference levels in the post production industry...Having worked in radio and tv audio for many years now - I've seem 'em all. Musicians give me the hottest mixes they can, Digibeta mixes need to be compatable with analog decks, and radio (often distributed nationally over the internet) will get squashed to death 2 or 3 times before it airs. All that said...generally I'm referenced at -18db....

[Mark Haliday]

You are not "losing" dynamic range.
Dynamic range is the difference between soft and loud passages. All you are doing is scaling down levels. If you do -20 db and then crank up your monitors +20db you should be hearing exactly the same thing ... well ... almost
You *are* inn fact losing definition in the low levels (this can affect reverbs for example), but I would dare say that it is almost irrelevant to broadcast where standards are not as "high fi" as on records

[MGagnon]

Diegux , you should find kind of answer at this address: http://www.digido.com/Audiolevels.html [img]images/icons/cool.gif[/img]

[Extreme Mixing]

I think that once all mixes are all digital. the zero VU standard for mixing should go away . In my opinion, you should respect your mic pres while recording, but fill those meters in the mix.

Steve

[Diegux]

Ok Guys,thanks for keeping this post alive,cause I'm far from satisfied. The digido article is great,thanks for the link.
My question though,has not been answered.
Steve,I would tend to think like you, ignoring the analog VU reference and trying to maximize your system. But what happens (at least in my everyday job) is that all d-beta masters are maxed at (the most) -6.
What I still don't get is...what's the point in calibrating @-20 in order to have an extra headroom in the digital world because of media\system properties if you are not going to use it (when you max at -6,you are actually mantaining the same headroom,14 DB, that is used in analog)
any other opinion?
Thanks again for your time.
Diego

[Ted Maul]

If you're laying back to D-Beta (a digital format for audio, too?) from a DAT, you're going digital to digital, therefore no need for attenuation, right?

[Diegux]

If you're laying back to D-Beta (a digital format for audio, too?) from a DAT, you're going digital to digital, therefore no need for attenuation, right?

Well, if you submit a d-beta with a mix maximied to 0dB FS,you'll be generating +20 Analog peaks (0VU = +4)after D\A...that's more signal that lots boards\equipment can handle
My question is not about D-Beta format,my question is about -20=0 reference level.

[Park Seward]

When inventing the digital video recorder, care had to be taken to use these new VTR in an existing analog environment. The video and audio in/outs had to match the existing environment. Also, the headroom had to be at least as good as the existing video recorders. The way to calibrate that is to go to the max, undistorted output you want from a device and back down from there. If you want a +24 dbm max output, and your zero reference is +4 dbm, you them end up with a -20 dBFS.

The headroom could have been set anywhere. But since a VTR can be used in unrehearsed situations and since digital at 20 bits has over 100 db S/N ratio, giving you generous headroom was a plus, not a minus.

Due to the early 16 bit systems having trouble at lower levels caused the audio industry to try and max out the levels as much as possible. Since the new digital VTR far exceeded the audio performance of older analog VTRs, there was no desire to get any more S/N ratio.

Different operating levels are a pain but they are not going away. Different machines and different organizations will set their operating levels where they want them.