|
|
Post by popmann on Nov 11, 2013 18:50:24 GMT -6
This isn't what I'd call self explanatory...in fact, whne I switched to Cubase and asked at that forum, I got a HUGE response of incorrect procedures. I was actually just asking what the quickest way was to export a quicky mix to redbook...not what needs to be done--but, the responses showed me most people who eval mixes on end user systems aren't doing it correctly.
In my eternal fight against loudness idiocy, I thing this is an important procedure to do...prior to taking any kind of "mastering" steps. Every step after #1 is simply for the quick eval file and can be chucked when done if you want.
1-export the mix to stereo WAV in native project sample rate (keep THIS file, as this is what you would provide to a mastering engineer)
2- Perform any sample rate conversion to 44.1 into a new project if applicable
3- PEAK normalize that WAV to -.04dbfs
4- insert dither post fader on it's channel (slots 7&8 in Cubase) set to 16bit
5- export that wav (from the channel output) to a new 16/44.1 file (be it redbook or mp3 or AAC or whatever you need for eval)
This came up on another small forum....again...and it occurred to me I've never posted this here. All this does is make sure you get a full scale redbook (or mp3). The order is important. Without doing this, you're not going to hit consumer converters or analog stages properly--and your mix will sound tiny. This (obviously) does nothing to increase RMS loudness that gets addressed in mastering, BUT it does give you a clear picture of what your mix sounds like.
If you think through the alternative--say we just burn the step 1 file to CD of export a quick mp3 from the 2buss of the project...if you're properly gain staged, you will be outputting a file in the -12 or -10dbfs PEAK range...which means the RMS will likely be -24dbfs or greater. Convert that to 16bit, and you have a seriously low resolution AND small sounding mix.
If you're wondering how this fights the loudness war--look, I'm not going to preach to you the evils of DR5 mastering...well, I will-but not right now. That's a CHOICE...that you should be making based on a full scale optimized copy of the mix. Not based on what amounts to a 12bit tiny quiet copy that would be a straight conversion of the 2 buss.
It takes minutes. If you knew how many rough mixes I get that have one of two issues, you'd see why I publicly post this....they're either tiny sounding exports...OR...they actually mixed the song clipping the master fader in order to get full scale...neither of those sound good--even if you DO want to crush it into oblivion for an end result.
|
|
|
|
Post by cowboycoalminer on Mar 1, 2014 9:00:18 GMT -6
My head is spinning. I must have missed this post the first go around. Any chance of you doing a simple screen shot video that gives a parallel visual of these steps and instructions? That would help me a lot.
Thanks Pop
|
|
|
|
Post by henge on Mar 1, 2014 9:20:24 GMT -6
Great info pop! A vid would clarify for all. Does step 2 leave you with 24 bit/44.1?
|
|
|
|
Post by popmann on Mar 1, 2014 12:05:50 GMT -6
It's the step that brings you to 44.1. You don't change the bit depth. So whatever you're working at. I tend to flip Cubase to 32bit for mixing...but, honestly, that more a safety net than anything--isn't universally supported, and actually adds another step not covered re:what you deliver to mastering.
Screenshots....the problem with that kind of thing is--helpful for you, being on Cubase...not helpful on Reaper, Sonar, DP, Logic...etc...I did this for years with a hardware recorder. It the concept and order you need to get in order to apply things to WHATEVER digital mixer you work with (that bounces without leaving digital). Maybe a simpler more conceptual list:
1-render your mix to a stereo file
2-Convert sample rate.
3-peak normalize to -.04dbfs
4-add dither noise post fader
5-export file at 16/44 wav for CD....or directly to AAC/MP3
Does that make it easier? The order is fairly important in order to ensure there never being an issue. None of this should alter the sound of the mix...unless you hear the dither being added (note, never use Waves IDR on Ultra)...it simply converts it to a downlevel full scale presentation.
a modern CD is ALSO crushed up with dynamics processing...but, you DO want to have a mix--even if you desire to test Soundcheck's limits to turn you down....better to hear your MIX while mixing...get it balanced and sounding great...the more balanced it is, the easier it is to get it crazy loud. Know what you've actually got. Then, you are making an educated call if you ALSO want to limit it or multiband loud mojo plug it.
|
|
|
|
Post by Johnkenn on Mar 1, 2014 13:29:28 GMT -6
I was told there would be no math...
I need to read this a couple of times.
|
|
|
|
Post by lolo on Mar 1, 2014 18:36:08 GMT -6
Regarding dither fellas? And the end of your mixbus chain, do you insert a dither plug before sending to mastering. Say for instance you mixing at 48/24 exporting to 24 bit file to send to mastering. Do you dither and set the dither to 24bit? This was something i never did, and then a while ago i read a couple of posts by Bob Olhsson that suggested this is pretty important |
|
|
|
Post by popmann on Mar 1, 2014 18:56:34 GMT -6
Yes. And do not normalize. No reason to....let them normalize it for whatever their DAC and workflow suggests. Since I enable 32bit float for bounces...I use the same Apogee plug as I do to spit out the 16bit file.
I don't honestly think it's that super important. It's more safety...it's to ENSURE the least significant bits go away. But, in a day of frankly more hiss in my DAW (courtesy of Slate and such)...than I've ever had, it really becomes less of an issue, IMO. I've never tested with/without...some things I simply take on theoretical face value.
I will say...Waves IDR in "Ultra" mode--PLEASE never use that. Holy jeez...I spent like and hour going through a mix once going "where is the nasty scratchy high end?"...when I found it? Some Waves plug I'd set to "ultra" for--some reason...I just heard it as some nasty and harsh--like I'd boosted something nasty...
|
|
|
|
Post by lolo on Mar 1, 2014 19:03:48 GMT -6
Yeah I normally just put the apogee dither plug on slot 8 in Cubase pretty much from the get go when i start mixing
|
|
|
|
Post by Mark Kano on Mar 3, 2014 21:55:40 GMT -6
Crap. This is why I need to spend more time on audio forums. I just finished a track I'm planning on releasing and this was my mix/master process. Session was recorded at 24bit 44.1 btw.
Bounced stereo mix through master fader at 24/44.1, no buss processing of any sort except a trim tool.
Imported to new session and applied Slate VTM, Scheps 73, VCC, Ozone 5 eq and VBC all on the master fader.
Bounced again at 24/44.1 and imported back into the same session
Used Waves Center and then Slate FGX for volume boost. Bounced again but this time I converted to 16bit during the bounce and selected Dither via the Slate FGX.
I used to slap as much as possible on the master fader during the first mix but have learned to take it in stages. I have not however given much thought as to what stages are best for conversion and dither. Did I screw this up?
|
|
|
|
Post by popmann on Mar 3, 2014 22:52:37 GMT -6
Not if it sounds good. And, I intentionally didn't address "self mastering"...as that's a little different concern--and not something you need to even look at until you hear your mix full scale on consumer systems, IMO. I would wonder why you need so much processing....but, you know, it's art...when I see mastering chains like that, I immediately have two thoughts "who mixed THAT"....and "who decided to use 44.1?" ....if the same guy, you should fire him for sure.  winky...joking...but, seriously...that's a hell of a lot of vaseline to be smearing on a lens of the whole mix...you might make a list of what you think all that "mastering" stuff is bringing to the sound. And then bypass it and make a list of what it's taking away. See if you can't isolate things and incorporate them into the mix. For example--VTM....great for heavy Egtrs...maybe a bass or vocal if it has some tracking "issues"--but, that can be put on a guitar buss--and likely cranked even more...while leaving the other aspects untouched. Maybe the limiter is really just trimming wayward drum transients that you didn't catch during the mix? Drum buss. Fix that without making the vocal weird or sucking the low end from the bass. Also...why TWO bounces for "mastering"? That's the only technically iffy thing there...the amount of DSP is simply...subjective, I guess. That said...what I posted above is not replacement for mastering...it's simply the method for a quick full scale eval. For you....for a client...it's the way you make a CD or mp3 of your MIX for a consumer system. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 4, 2014 4:14:01 GMT -6
Great topic.
I can't get it, that this is not a general option in all DAWs, in fact i did not see it in any DAWs. But it is not complicated, it is just logical.
Export to wav with project SR and bitdepth, if exporting for a mastering house, procedure ENDS HERE -> SRC -> Normalize, if you are not already in the ballpark with your peaks-> Dither, if final format bitdepth is different from project bitdepth -> Export to final format.
Astonishingly, many people do not really get the order of the steps or what they do, and even more surprising, you don't get these instructions in the DAW manuals.
Maybe, because it would be striking, that there is no "one knob does it all"-option in their software for this pretty common procedure...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Mar 4, 2014 5:04:11 GMT -6
I guess the one thing, that breaks this up into at least 2 steps, is the normalization, you have to make an offline render to accomplish that, so the software gets the highest peak value. But i do not get, why the export routines of DAWs don't offer this as an optional service in any DAW i have seen yet. It would take the technical side of things away from the user, providing him a perfect full export without thinking about details and having to reimport, doing a seperate project etc...
|
|
|
|
Post by Mark Kano on Mar 4, 2014 13:48:37 GMT -6
I've been tracking at 16 bit 44.1 for a long time just out of the fear that any conversion would introduce unwanted noise and also to preserve cpu resources. I have also fired myself from engineering many times:) I just happen to work for cheaper than anyone I know, and I suck less and less at it each time... or so I think:) Sometimes the additional steps in mastering are because certain plugins will turn back into pumpkins and I would like to be able to go back. In this case I wanted my original mix with no buss goo, the next mix with buss goo, and finally the volume boost.
I need to read up on what dithering is. If I'm selecting "convert during bounce" in PT I'm not currently sure what the dithering is doing in addition to this.. I do notice that when I try to get my levels up to compete with modern tracks, it's so distorted that I have to back off. I know some of this is in my mixing but just making sure I'm not also hurting myself with my process. I really appreciate everyone sharing their knowledge here!
|
|
|
|
Post by Johnkenn on Jun 27, 2014 18:16:15 GMT -6
So - just really reading through this...Since I think I'm switching to the AOM Limiter, I'm looking to use a new dithering plug. Thinking I might use the Ozone 5 M-bit 2...One thing I've never done is normalize. I always thought of normalizing as being bad...is that not the case? Are you just basically cutting off any info past .04dbfs? Is it necessary to do Sample Rate Conversion and bit rate conversion in two separate moves? I've always bounced from say - 24/48 to 16/44.1 in the bounce in PT's...
|
|
|
|
Post by wiz on Jun 27, 2014 18:30:52 GMT -6
John, Normalizing means bringing the highest peak of the song to a set point. It doesnt cut anything off, that would be limiting.
Eg, you have a song that you play through from start to finish, you look at your DAW meters (peak responding, post fader) and they say -12dB.
This means that during the song, the loudest instantaneous peak that was reached hit -12dB.
If you normalise that file, to the above mentioned .04dBFS then it "turns up" the file so that the biggest peak is now .04dBFS , not -12.
The only thing I do differently, to what popmann posted above, is I set things to -1dBFs, just an old habit from mp3 conversion, I leave some headroom so it doesnt clip.
cheers
Wiz
|
|
|
|
Post by donr on Jun 27, 2014 18:33:26 GMT -6
Doing this is quite simple in DP. No matter what your sample rate or bit depth, you can offline bounce to .mp3. Put MOTU
Masterworks Limiter (with dither on) or your limiter of choice at the end of the master fader plugins, lower the limit threshold until the mix just hits your ceiling (usually -.1 dB for me) peaking, and let DP do the math.
I too have never thought about SRC as a separate step, when you don't have to making the mp3 in DP.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 28, 2014 7:13:08 GMT -6
Glad to read this again. I needed the understanding of normalizing as I too thought it destructive in some way.
|
|
|
|
Post by Johnkenn on Jun 28, 2014 9:08:59 GMT -6
So I can't find any kind of normalization plug that measures in full scale (dbfs). Well, I guess I only have what's in Pro Tools. How much are you yanking it down?
|
|
|
|
Post by mulmany on Jun 28, 2014 11:45:52 GMT -6
So I can't find any kind of normalization plug that measures in full scale (dbfs). Well, I guess I only have what's in Pro Tools. How much are you yanking it down? John, normalize is an audio suite plug process. It's not pulling down its pulling the mix up. Because you should be mixing at a set monitor level and mixing to what is comfortable. This causes you to mix with the ears natural loudness curve (fletcher - Munson), you then apply the normalization to the finished mix... taking advantage of the whole 16bit depth. Basically you are trying to mimic what the mastering house is going to do to prep for 44.1/16 pressing. |
|
winky...joking...but, seriously...that's a hell of a lot of vaseline to be smearing on a lens of the whole mix...you might make a list of what you think all that "mastering" stuff is bringing to the sound. And then bypass it and make a list of what it's taking away. See if you can't isolate things and incorporate them into the mix. For example--VTM....great for heavy Egtrs...maybe a bass or vocal if it has some tracking "issues"--but, that can be put on a guitar buss--and likely cranked even more...while leaving the other aspects untouched. Maybe the limiter is really just trimming wayward drum transients that you didn't catch during the mix? Drum buss. Fix that without making the vocal weird or sucking the low end from the bass. 
