Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Augustine Leudar]

I read here that with digital the channels do not need to get near the red but -12db suffices because of the headroom. That being the case in my multitracker when i come to render the final file is it better to render it at -12 db ???? or is it just the tracks that need to be -12db but the master outs need as usual to be as close to the red as possible without clipping ?

[.nathan.kosakowski.]

technically digital doesn't have _any_ headroom.  anything over 0 at the top and its crap.  to compensate engineers pick a value, usually between -20dbFS to -12dBFS.  tape has a _loose_ headroom of 14dB, add a few more for errors just to be sure...  -12dbFS is on the risky side, i'd try aiming for averaging -18 or so.

is the recorder 16bit or 24bit?  when mixing down to a 16bit format (DAT or whatever) one would do a few practice run throughs of the song, trying to get the peak as close to the top without overs.  since playing the mix again and again is entirely repeatable, this works.  when tracking someone live, you obviously down know whats coming next, so aim for you average level to be safe, leaving headroom just in case.

if mixing down to a 24bit format, just make sure it averages somewhere above -20dB and don't worry about it.  there is plenty of resolution so you don't need to nit pick.

again, 16bit is another thing, aim for getting it a few dB under 0 by running through it a few times.

one more thing, be concious that there are several different ways to use the term "-12dB".  on analog 0 was supposed to be you average, so there are values like +6dB and such.  on digital there is nothing over 0 so analog's 0VU = -18dBFS or whatever.

if you are mixing entirely on the computer, mix to a 24bit file and just let it be, in mastering (whoever that may be) the final peak adjustments can be made.

[Augustine Leudar]

Thanks great reply.  rMs ? What i mainly wondering is if upon rendering  the track to -12 db (or thereabouts)should then be normalised to 0 db as otherwise it would be too quiet wouldnt it ?
(yes to 16 bit has to go on a cd)

[bobkatz]

Augustine Leudar wrote on Sat, 07 January 2006 17:01

I read here that with digital the channels do not need to get near the red but -12db suffices because of the headroom. That being the case in my multitracker when i come to render the final file is it better to render it at -12 db ???? or is it just the tracks that need to be -12db but the master outs need as usual to be as close to the red as possible without clipping ?

Render for what purpose? For further processing, or as a final "master" to be sent to a plant for duplication?

If for further processing, which may include analog processing or SRC or coding (e.g. MP3) I recommend no higher than -3 dBFS as an arbitrary point below full scale, sufficiently high to satisfy the "SNR addicts" and sufficiently low to stay away from any potential requantization overloads. But if you're talking "headroom", the only "headroom" you should leave would be from the true value of the output level of a digital to analog converter. Since the meter you're using is a true digital peak meter, you have a lot more room to go up than -12 dBFS.

Regardless, for a 24 bit recording, you will not detect any audible "snr" improvement recording any hotter than -12 dBFS, because the noise floor is significantly lower than that anyway.

BK

[Augustine Leudar]

I will be sending it off for mastering later however this is a copy  someone wnated to hear on cd first, so i thought id burn them a copy off. Im assuming I want to get as close to 0 db as possible ? Perhaps I should use a limiter, its a film soundtrack type soundscape of the jungle and a compressor changes the frequencies too much.

[bobkatz]

Augustine Leudar wrote on Sun, 08 January 2006 14:30

I will be sending it off for mastering later however this is a copy  someone wnated to hear on cd first, so i thought id burn them a copy off. Im assuming I want to get as close to 0 db as possible ? Perhaps I should use a limiter, its a film soundtrack type soundscape of the jungle and a compressor changes the frequencies too much.

Why not tell the client to turn his own volume control up...  and leave the darn thing alone, if anything you do changes the sound...

BK

[Augustine Leudar]

bobkatz wrote on Sun, 08 January 2006 22:53

Why not tell the client to turn his own volume control up...  and leave the darn thing alone, if anything you do changes the sound... BK

Thats what I did in the end. So the snr isnt such an issue with digital, but most cds are still recorded as loud as possible.

[bobkatz]

Augustine Leudar wrote on Tue, 10 January 2006 21:51

bobkatz wrote on Sun, 08 January 2006 22:53 Why not tell the client to turn his own volume control up...  and leave the darn thing alone, if anything you do changes the sound... BK Thats what I did in the end. So the snr isnt such an issue with digital, but most cds are still recorded as loud as possible.

I guess it's the definition of what "as loud as possible" means that creates the great divide. A client doing a country music album just sent me a Brooks and Dunn cut for an example. It's unlistenable. It sounds WORSE than my car radio.

BK

[Patrik T]

Augustine. Since you wrote that your music piece had some soundtrack-qualities to it, I just wanted to mention this:

One of my favourite albums is an audiophile one. It was recorded with one stereo microphone (some vintage AKG tube-story), through a cumstombuild mixer for preamping, and straight to tape. There is one singer playing guitar and two additional guitarplayers by his sides. The performance is taking place in a room with ace acoustic properties. Everything goes through the same microphone at the same time.

As you see, the setup is extremely basic.

One of the tracks on that CD, perhaps my personal favourite has got the following digital thinkable figures:

- An average RMS level of about -27 dB throughout the whole track.
- Highest peak at -6 dBFS.

The recording sound ALIVE. Things JUMP out from the speakers. It doesn't even use the 16:th bit. This recording is released like this.

I have found that some other recordings which I love for their sonic qualities have about the same properties in their digital format on CD. They are AWAY FROM zero dB full scale.

It's all a matter for the listener to stretch out for the volume-knob and adjust is until the listeners ear is pleased with the soundpressure. Actually, most people with ears will tend to do exactly so. And the really funny thing is - they will just judge the music with what they hear.

They won't give a s*** about how many bits that where used either in tracking or in the final mix. The only time they might give a s*** about the issue is when the recording is constantly using ALL the top bits ALL of the time. But in that case, sadly, most won't understand why things sound weird.

[Augustine Leudar]

Patrik T wrote on Thu, 12 January 2006 14:44

Augustine. Since you wrote that your music piece had some soundtrack-qualities to it, I just wanted to mention this: One of my favourite albums is an audiophile one. It was recorded with one stereo microphone (some vintage AKG tube-story), through a cumstombuild mixer for preamping, and straight to tape. There is one singer playing guitar and two additional guitarplayers by his sides. The performance is taking place in a room with ace acoustic properties. Everything goes through the same microphone at the same time. As you see, the setup is extremely basic. One of the tracks on that CD, perhaps my personal favourite has got the following digital thinkable figures: - An average RMS level of about -27 dB throughout the whole track. - Highest peak at -6 dBFS. The recording sound ALIVE. Things JUMP out from the speakers. It doesn't even use the 16:th bit. This recording is released like this. I have found that some other recordings which I love for their sonic qualities have about the same properties in their digital format on CD. They are AWAY FROM zero dB full scale. It's all a matter for the listener to stretch out for the volume-knob and adjust is until the listeners ear is pleased with the soundpressure. Actually, most people with ears will tend to do exactly so. And the really funny thing is - they will just judge the music with what they hear. They won't give a s*** about how many bits that where used either in tracking or in the final mix. The only time they might give a s*** about the issue is when the recording is constantly using ALL the top bits ALL of the time. But in that case, sadly, most won't understand why things sound weird.

veeedddy interesting meester bond...
but excuse my ignorance, what program allows you to calculate the rms db level for a track and how ?

[Patrik T]

Augustine Leudar wrote on Mon, 16 January 2006 16:33

veeedddy interesting meester bond... but excuse my ignorance, what program allows you to calculate the rms db level for a track and how ?

There are programs that can provide you with a basic and average RMS level. Hm, I think Wavelab has got such a thing. And Cool Edit or Adobe or whatever it is called nowadays.

They can read the track and spit out some AVERAGE figures. The RMS will increase when the music is louder. The RMS will decrease when the music becomes softer. If the program material is quite static, the average RMS can indicate where things are at.

If you are mixing in the box I can recommend to use a K-20 meter (if there is one avaliable) or to switch your mix/output bus to read RMS-level. In the later case you will be safe and good and nice and everything else if you keep the level around or below -20 dBFS.

I always keep my master fader at zero - never touching that one -  and work my track faders butts off if I get too close to clipping on the output. This in conjuction with a -20 dBFS creates mixes that won't get overfilled. The more you approach full scale, either in separate tracks or in the master bus, the more congested it will be.