Why are my mixes -3db on another computer?

[Nick96]

I am mixing songs in Cubase 3.1 SX at my studio. I mix so tracks peak at -3db, then I use something like Waves L2 to bring it upto 0db. I take these mixes home and import them into Wavelab and they peak at -3db. Is this because of a setting I am missing? I am not to experienced with this stuff so if anyone could tell me why I very much appreciate it. Thanks people :-)

[WidgetNinja]

Kinda coincidental that it seems to be back to the unprocessed peak level, right?
You sure you are including the plug in the mixdown?

If you were changing bit depths implicitly, say importing a 16 bit file as 24 bit, the difference should be greater than 3db.

I know on some of my gear, like my Fireface (or was it in SX?), it has a setting for calibrating to 0- +3db. It could just be you have a meter setting inconsistency.


What format file are you mixing to?
What Format is the project set up for?
Do you have a dithering plug in the chain?

And most importantly, is L2 making it to your mix? Try smashing it to hell to verify.

[tony666]

The L2 raises your average levels but no peak levels. The sound appears to be louder (rms levels go up indeed) but your peak level is still at -3dBFS.

[DontLetMeDrown]

Originally Posted by tony666

The L2 raises your average levels but no peak levels. The sound appears to be louder (rms levels go up indeed) but your peak level is still at -3dBFS.

WTF? This is not correct.

The meter on the right-- "Cut Ceiling" adjusts the peak level. If I am wrong, please explain.

[tony666]

Originally Posted by DontLetMeDrown

WTF? This is not correct.

The meter on the right-- "Cut Ceiling" adjusts the peak level. If I am wrong, please explain.

The L2 inserted on a track with the fader at -3dB will use this as ceiling.

When he inserts the l2 on the master fader he's probably bouncing the individual track's outputs , instead of the master ouput.

[jdunn]

I too was going to say that the L2 doesn't work like that, but if the fader is at -3dB, who knows?

What you might try is bouncing the mix to disk first, then take the resulting stereo file and process it with the L2. Set the output ceiling to -.3dB. The master fader should be at zero, no?

If you're mixing with the master fader down 3dB, try putting it at zero and lowering the individual tracks so they don't overload the master buss. This is proper gain staging, which is important in a DAW.

[maxy]

jdunn - can you explain why having the master at zero and the individual track down is proper gain staging? This is something I've been wanting to understand.

Thanks
Maxy

[junior]

not sure, but could it have something to do with your pan law settings in cubase? they're set to -3dB by default (page 209 in the manual).

http://www.harmony-central.com/artic.../panning_laws/

[jdunn]

Originally Posted by maxy

jdunn - can you explain why having the master at zero and the individual track down is proper gain staging? This is something I've been wanting to understand.

Thanks
Maxy

I'll try, but I might get it wrong. Let's say that you have to lower your master fader 3dB or it will clip. This is most likely an indication that the individual channels are hitting the master fader too hot.

Try just lowering each channel by an equal amount, say 2dB at first. Then put your master fader at zero and see what happens. Basically, nothing should be clipping, starting with the channels, then the busses, then the master fader.

So, you're lowering the channels until you can set the master fader at zero without clipping. Then you won't have to digitally attenuate the whole mix at the master fader. I think it's better to trim the gain at each channel as opposed to doing at on the main bus. Less potential loss of detail.

I'm not a digital math expert, this is just how I do it. Someone more knowledgeable than I may say that it doesn't matter either way, but I think I'm technically correct on this.

[Geert van den Berg]

I am not a Cubase user, so I don't know how this is in Cubase. But I am very interested in the bussing and routing in Cubase and Nuendo.

In PT's the inserts on the master fader are post-fader. Thus if you lower the master fader and you have an L1 or L2 inserted with the ceiling set to 0 dbFs it will go to 0. The only thing that happens is that you drive the limiter less. If you lower the treshold -3 and then set the master fader to -3 you probably end up with almost the same thing, the same level as without the L1.

If you would do the same thing on an audio channel where the inserts are pre-fader, lower the fader lowers the ceiling as well. If you lower the treshold, the limiter will start working. Lowering the channel fader won't drive the limiter less but basically you are lowering the ceiling. so if you lower the ceiling by -3 and you pull down the channel fader with -3, you have a ceiling of -6.

I have no idea if the same goes for Cubase, but it could be the cause for Nick's problem.

[wildstar]

INSERTS 1-6 ON THE MASTER FADER A PRE. 7 & 8 ARE POST. MAKE SURE YOUR INSERTING THE L2 ON INSERTS 7 OR 8 ON THE MASTER FADER.

[Nick96]

Ok guys I will break it down some more because I really want to get to the bottom of this. By the way, I track and mix at 44.1khz and 24 bit into a RME multiface.

I mix in cubase, don't know about the panning law although I knew it was in the settings. I will try it but have never had the problem before. I did things a little differently this time round. This is how I mixed the tracks.

1. Mix with individual tracks peaking at -3db so the master bus gets a very occasional peak of upto 0.5 of a db giving an absolute peak of -2.5db . Generally speaking I aim to not go over -3.

2. When I am happy with the mix I will place a Really nice compressor as a hardware insert in the master section. This will fatten things up a little, providing mabey another 1 or 2 db in level. By now I am peaking at about -1 db.

3. I mixdown the file to a Wav file at 44.1 24 bit.

4. I import the tracks into Wavelab. Apply an L2 dithering to 16 bit, with the ceiling at -0.2db and work the threshhold ( I think thats what it's called, can't see comp right now) to about 6db if that makes sense. Basically to a point where the loudness of the track can compare with something commercial the band likes.

5. Usually I will save these files after they have been rendered with L2 in wavlab as 44.1 16 bit in wavelab. When I save as I get the option as to what format I save it as.

6. Thats it. I got the tracks home and ripped them with nero into wave files. (wavelab won't rip off a cd with my comp at home) This is because I didn't think the tracks were loud enough. I wanted to see what the level was like in Wavelab. I don't remember this happening before. Is this the way you guys would recomend?

Any suggestions would be greatly appreciated. Thanks for all the responses. I can feel the love awwwwwwwwwww :-)

[Nick96]

Any feedback would be greatly appreciated :-) I was away for a few days and only got to read the post yesterday. Thanks

[tony666]

1. Skip Nero and import the raw 16 bit wav file into Wavelab to watch levels. (maybe Nero is doing something/limiting)

2. I stil don't understand what files you render with the L2 in Wavelab? Individial bounced tracks (those at -3dBfS) or "master" tracks (those at -1 dBFS) of a whole song? What's the level of these tracks in wavelab before L2?

Good luck

[sleepwalker]

By this logic, your master fader is completely pointless. Why have it if you have to leave it at 0? If the digital mix bus is floating point, then these peak values actually mean a lot less than it looks.

As far as I'm concerned, provided your gain staging isn't ridiculous, just pay attention to the meter on your master. This is(probably) going directly to your DA converter. Just make sure your peaks aren't too high. -3-8dB sounds good to me.

I personally mix with the limiter engaged, so the above advice wouldn't apply.

Originally Posted by jdunn

I'll try, but I might get it wrong. Let's say that you have to lower your master fader 3dB or it will clip. This is most likely an indication that the individual channels are hitting the master fader too hot.

Try just lowering each channel by an equal amount, say 2dB at first. Then put your master fader at zero and see what happens. Basically, nothing should be clipping, starting with the channels, then the busses, then the master fader.

So, you're lowering the channels until you can set the master fader at zero without clipping. Then you won't have to digitally attenuate the whole mix at the master fader. I think it's better to trim the gain at each channel as opposed to doing at on the main bus. Less potential loss of detail.

I'm not a digital math expert, this is just how I do it. Someone more knowledgeable than I may say that it doesn't matter either way, but I think I'm technically correct on this.

[jdunn]

If you're mixing entirely ITB, then the master fader is not pointless. I'm mixing through 16 busses into an analog summing bus, so the master fader is set at zero.

So you don't watch overs on the individual DAW channels, only the master?

[Nick96]

Thanks for all of your responses. I would like to focus on one part of my problem if I may. If I am mixing in the box, 44.1/24 bit. I insert a stereo comp on the master bus and render the mix real time in Cubase. At what point should I dither to 16 Bit? Is it just a matter of setting the Waves L2 after the stereo comp to dither, and then choosing 16 bit 44.1 in the save properties in cubase export funtion?

What is the panning law and how do people set it in Cubase and Nuendo? Thanks again for all your help :-)

[Cojo]

Originally Posted by Nick96

Thanks for all of your responses. I would like to focus on one part of my problem if I may. If I am mixing in the box, 44.1/24 bit. I insert a stereo comp on the master bus and render the mix real time in Cubase. At what point should I dither to 16 Bit? Is it just a matter of setting the Waves L2 after the stereo comp to dither, and then choosing 16 bit 44.1 in the save properties in cubase export funtion?

What is the panning law and how do people set it in Cubase and Nuendo? Thanks again for all your help :-)

Hi there!

The panning law is how much the level is decresed when you have your pan pot centred. This apply to mono tracks. If you use the balance of a stereo track you will not get a decrese in level when pot is centerd. Why, you may ask? Because for you to get the illusion of having the same volume regardless of the pan setting.

About the wavelab/computer thing: have you tried to reload it into wavelab on the same computer to see what result you got there?

/Cojo

[DAWgEAR]

Have you seen this?

http://forum.cubase.net/phpbb2/viewt...highlight=dbfs