Old 09-17-2020, 05:51 PM   #1
Peterk312
Human being with feelings
 
Peterk312's Avatar
 
Join Date: Feb 2012
Posts: 477
Default Basic Question about Mix vs Master

A basic rule-of-thumb for creating a mix is:

Mix master VU meter is set for -18dbfs = 0 db (VU), and you have gotten your mix to hover around 0 db but the meter of course goes beyond it at times and may occasionally hit +3 db but only briefly. No compression on the master.


Q. If you follow the above parameter, which I think is fairly standard, when you go from your mix to your master, should you generally notice a very large jump in volume?

The reason I ask is because I'm experimenting with Ozone 9 Elements for mastering, and when I set the limiter ceiling at -0.5 dbfs and then check it again with a VU meter on the rendered result, it sounds great but I can't believe how much louder it sounds compared to the original mix. When I look at the waveform for the rendered master result, it doesn't look slammed to the max with peaks chopped off. But it's incredibly louder. It makes me think I'm doing something wrong in the mix. So, I'm asking the above question so anyone with more experience with mastering might provide some insight.
Peterk312 is offline   Reply With Quote
Old 09-18-2020, 12:45 AM   #2
ramses
Human being with feelings
 
Join Date: Jul 2009
Posts: 1,231
Default

Quote:
Originally Posted by Peterk312 View Post
If you follow the above parameter, which I think is fairly standard, when you go from your mix to your master, should you generally notice a very large jump in volume
As always with audio, it depends. In this case it largely depends on what the final destination of your master is. Is it music for film? For Spotify? Youtube? CD-master? Vinyl? For all of the above? And so on. Mastering is done to meet specific ends, and these ends will often relate to specific loudness standards for delivery.

LUFS is still the most common standard I believe, and I believe that -14 LUFS is still considered the standard "allround" level, that is the standard that will grant the best results on the most common destinations.

Also, dB levels are funny, as a +10 dB increase is usually percieved as being twice as loud, which is A LOT louder. It's recommended, when you want to compare the benefits of mastering for example, to verify your interventions at the same percieved loudness as the original mix. YMMV.

Last edited by ramses; 09-18-2020 at 12:51 AM.
ramses is offline   Reply With Quote
Old 09-18-2020, 01:04 AM   #3
ramses
Human being with feelings
 
Join Date: Jul 2009
Posts: 1,231
Default

Quote:
Originally Posted by Peterk312 View Post
It makes me think I'm doing something wrong in the mix. So, I'm asking the above question so anyone with more experience with mastering might provide some insight.
No, I think that having a mix target of around -18dB RMS sounds reasonable. I set my own target at -14dB, in order to know if the mix will (most probably) need additional limiting or not, and I find this target to be the most convenient. If you already know that your mix will probably end up at -14dB (or louder) RMS mastered, I don't really see any point in not having this set as your mix target as well, so that you can deal with overshoots and stuff somewhat already while mixing. As long as you feel that you know what you're doing.

Last edited by ramses; 09-18-2020 at 01:17 AM.
ramses is offline   Reply With Quote
Old 09-18-2020, 01:07 AM   #4
ramses
Human being with feelings
 
Join Date: Jul 2009
Posts: 1,231
Default

Quote:
Originally Posted by Peterk312 View Post
Mix master VU meter is set for -18dbfs = 0 db (VU), and you have gotten your mix to hover around 0 db but the meter of course goes beyond it at times and may occasionally hit +3 db but only briefly.
Oh yeah, different VU time windows will give you different readings. Check the settings for the Reaper master track VU-meter, or get yourself a decent, standardized RMS/VU plugin to rely on. And perhaps a integrated LUFS meter thingy.
ramses is offline   Reply With Quote
Old 09-18-2020, 11:07 AM   #5
Peterk312
Human being with feelings
 
Peterk312's Avatar
 
Join Date: Feb 2012
Posts: 477
Default

I used Youlean Loudness Meter to get an integrated LUFS reading for comparisons between the mix and the mastering stages. Before limiting, the mix was at -28.2 LUFS. After limiting with Ozone Elements, I got -16.2 LUFS. That's about the level you would need for streaming Internet services. It's a huge jump in volume. It's also much closer to what sounds like a "commercial" recording level compared to things like the Ray Charles version of "I Got a Woman," which on the remastered edition comes out to be about -13 LUFS. Of course, who knows where the mix level was before limiting?

But again, my question was is this indicative of the level being to low in the mix, -28.2 LUFS vs -16.2 LUFS, or is this an issue that doesn't really matter and mostly depends on what kind of music you're recording?
Peterk312 is offline   Reply With Quote
Old 09-18-2020, 11:42 AM   #6
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Well yeah your mix was 12 dB quieter (integrates average) than the “mastered” version. If it was actually peaking near 0, then it was probably way too dynamic, and Ozone must have smashed the hell out of it. If it was peaking around -12, then Ozone did nothing but turn it up.

My point being that absolute levels are absolutely meaningless right up until you get to making the final distribution master. All that matters at mix and even pre-master time is the crest factor. It’s sometimes (sort of inaccurately) called Dynamic Range, but whatever. It’s the difference between the loudest peak and the average level. That then tells what the actual absolute loudness can be without push your peaks past zero, which is what actually matters when it’s finally mastered.

If your mix averages -26 and peaks at -12, then when you normalize it you’ll be nuts on -14 average. Same average with peaks at -18 is going to average -8, super loud. -26 peaking at 0 still averages -26, way quiet and probably too dynamic because people will try to turn it up, but then those loud peaks will blow their heads off and possibly break things. OTOH, if it’s averaging -12 and peaking at +6, which will normalize to a relatively quiet -18 average.

Crest factor is all that matters.


(Ok yeah absolute peak CAN matter if there are things in the chain with limits on their output, but in the DAW even that is relative to an extent. You decide how close to or far beyond the limits you want to go based on how it sounds, and can adjust the result to whatever absolute level you want)
ashcat_lt is offline   Reply With Quote
Old 09-22-2020, 07:32 AM   #7
Peterk312
Human being with feelings
 
Peterk312's Avatar
 
Join Date: Feb 2012
Posts: 477
Default

Using dpMeter (by DBProAudio) Mix vs. Mastered version:


RMS integrated:
-32.2 dBFS vs. -19 dBFS


Max True Peak:
-10.9 dB vs. -0.5 dB


Crest:
21.2 dB vs. 18.5 dB


For those who don't know, the Izotope Ozone plug-in makes recommendations to get you started on a mastered version of your mix. I was experimenting with it to see what is it actually doing to the mix.

So, limiting in Ozone Elements only lowered the crest factor by 2.7 dB. RMS is raised almost to the point you would want it (-18 dBFS), although I lowered that myself a bit because even what Ozone Elements calls a "low" intensity sounded too loud. I can't imagine why anyone would go with what they call a "high" intensity level, which sounds slammed to me, but again I guess it's about what kind of material is in the mix that you're mastering.

Looks like setting the limiter ceiling to -0.5 dB allowed for an increase in peak level by more than 10 dB, and that would be obviously louder. What I don't think is audible is the crest factor decreased by just under 3 dB.

But the basic question I had was is it better to have a situation like I have in which the mix appears to be way too low in volume for the final master, which then needs to be turned up quite a bit during the mastering process. Or, is it better to raise everything up in the mix? I'm guessing it doesn't matter as much with digital media, except for maybe which way increases your noise floor more. It's just what I think of as a "healthy" level in a mix based on the -18.0 dBFS = 0 VU rule doesn't transfer to the mastering process as I expected and sounds too weak for a final mastered version, as if I'm doing something wrong in the mix. But I guess not.

Last edited by Peterk312; 09-22-2020 at 07:44 AM.
Peterk312 is offline   Reply With Quote
Old 09-27-2020, 08:53 AM   #8
jrk
Human being with feelings
 
Join Date: Aug 2015
Posts: 2,969
Default

Does it really matter? If you've got your monitoring calibrated to taste, then you mix at at levels that suit you - whatever that comes out at (insert number & units here) provided you're not clipping the master - cool. If you're sending stuff on for mastering - then they'll have preferences for what they want from you (so many bits / so much headroom) and you'll adjust yr mix to that before shipping it.

If you're mastering yrself - then you'll have a target (even if approximate) for LUFS in mind, and if you can hit that without mucking up the sound then all's well.

If you can't get close to your target, *then* you might want to think if there's "something wrong in the mix"

You might not be interested in loudness, and you might be completely happy with the sound, in which case your mastering is reduced to cranking up some gain so that you end up with the (true) peak you decide on. Pretty close to 0, I'd guess.

If you've got reference tunes alongside your mix - and it can be a good thing - bear in mind that these will have been mastered - and may sound a bit much with your mix monitoring levels - so pull down your references a few dB.
__________________
it's meant to sound like that...

Last edited by jrk; 09-27-2020 at 09:08 AM.
jrk is offline   Reply With Quote
Old 09-27-2020, 09:12 AM   #9
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

If you typically monitor/listen at such a high level that -28 LUFS program is in the listening zone range...

Strictly classical music listener right?

You shouldn't have anything suffer in a mix from it. It's when people do the polar opposite that causes big problems. ie. Monitoring at very low levels from listening to volume war program in -8 LUFS territory. Then trying to mix at that monitor level. Red lights going off before you can even hear the track!

So, doing the opposite here. You'd be throwing away 2 bits of digital precision with a mix that peaks at -12db. Using 22 of the 24 bits. Not a big deal and not really losing a damn thing.

I'd just want to be a little cautious with monitor levels like that. One full scare audio event in the system could make for a really bad day! For both electro-mechanical and biological based audio equipment!
serr is offline   Reply With Quote
Old 09-27-2020, 09:14 AM   #10
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Quote:
Originally Posted by jrk View Post
If you're sending stuff on for mastering - then they'll have preferences for what they want from you (so many bits / so much headroom) and you'll adjust yr mix to that before shipping it.
Honestly if they ask for anything other than floating point they’re kinda derpy, and if they say a god damn thing about “headroom” you should find somebody else.

As I said above absolute levels don’t fricking matter until the distribution master, especially if you render to floating point. If your ME doesn’t understand that, then you should not work with them.
ashcat_lt is offline   Reply With Quote
Old 09-27-2020, 09:28 AM   #11
COCPORN
Human being with feelings
 
Join Date: Jan 2008
Location: Oslo
Posts: 315
Default

ashcat_it: What is an exported format that preserves floats or doubles that you can conveniently send to a mastering house? I am not aware of those.
COCPORN is offline   Reply With Quote
Old 09-27-2020, 11:12 AM   #12
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by COCPORN View Post
ashcat_it: What is an exported format that preserves floats or doubles that you can conveniently send to a mastering house? I am not aware of those.
I recently sent an EP worth of music off to a mastering house, and before sending I asked twice what format they wanted. Both times they replied "whatever", so I sent 32-bit FP wav. I got the same format back.

As a side note: I also null tested my material in 32-bit FP vs 64-bit FP and there was no practical difference. It did not null to -inf, but to -98, which is the same thing for me. The 32-bit FP files were of course much smaller.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 09-27-2020, 11:26 AM   #13
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Quote:
Originally Posted by Fabian View Post
I got the same format back.
This kind of surprises me. What were they mastering for? Unless I have a specific request, I usually give my clients 24 bit for uploading and 16 bit in case they want to put it on CD and mp3 for portability and sharing. I just feel like mastering should be the final stage and you shouldn’t have to do anything more to the files you get back. All the dither and stuff should already be handled and all you need to do is “press it”.
ashcat_lt is offline   Reply With Quote
Old 09-27-2020, 01:26 PM   #14
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by ashcat_lt View Post
This kind of surprises me. What were they mastering for? Unless I have a specific request, I usually give my clients 24 bit for uploading and 16 bit in case they want to put it on CD and mp3 for portability and sharing. I just feel like mastering should be the final stage and you shouldn’t have to do anything more to the files you get back. All the dither and stuff should already be handled and all you need to do is “press it”.
This went then to be pressed on vinyl.

But... you may be right, I'll have to check. I got wav files back for sure, but now that I think of it... I have to check to be sure.

EDIT: My bad. You are right, I got 24-bit wavs back. Sorry. But I did send 32-bit FP.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 09-27-2020, 02:17 PM   #15
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Quote:
Originally Posted by Fabian View Post
This went then to be pressed on vinyl.
If you’re satisfied it’s all that really matters. If I had a client with a specific plan for manufacturing, I’d consult the place they’re sending it. For vinyl I might be inclined to send each side as one long 24 bit wave in order to preserve timings and stuff, but again it depends what the pressing house wants/expects.
ashcat_lt is offline   Reply With Quote
Old 09-27-2020, 08:48 PM   #16
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Kalispell
Posts: 14,745
Default

Quote:
Originally Posted by Fabian View Post
This went then to be pressed on vinyl.
May I ask Fabian, why you want to go to vinyl? I assume it's mostly for nostalgia, but are there other reasons. I have friends on VIP that are also doing this, and that seems to be the overall reasoning.

I've also been mastering some of their songs for them as well as mixing a few. Whether for mixing & mastering or just mastering, I've been getting 44.1/24bit files from them, and sending the same back.

This is my mix/master setup using 3-second sign-sweep profiles.

https://stash.reaper.fm/40402/Mixing...tup%20%231.png
Tod is offline   Reply With Quote
Old 09-28-2020, 01:27 AM   #17
jrk
Human being with feelings
 
Join Date: Aug 2015
Posts: 2,969
Default

Quote:
Originally Posted by ashcat_lt View Post
Honestly if they ask for anything other than floating point they’re kinda derpy, and if they say a god damn thing about “headroom” you should find somebody else.
I'm sure you're right. And maybe I'm out of date, but this kind of spec. isn't / wasn't unheard of.
__________________
it's meant to sound like that...
jrk is offline   Reply With Quote
Old 09-28-2020, 01:29 AM   #18
jrk
Human being with feelings
 
Join Date: Aug 2015
Posts: 2,969
Default

Quote:
Originally Posted by serr View Post
I'd just want to be a little cautious with monitor levels like that. One full scare audio event in the system could make for a really bad day! For both electro-mechanical and biological based audio equipment!
Another good reason for calibrating monitors.
__________________
it's meant to sound like that...
jrk is offline   Reply With Quote
Old 09-28-2020, 09:56 AM   #19
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Quote:
Originally Posted by jrk View Post
I'm sure you're right. And maybe I'm out of date, but this kind of spec. isn't / wasn't unheard of.
I mean, my job is to take what I get and do the best I can with it, but if you’re asking me what format to send, I’m telling you 32bit floating point because then I know that the full dynamic range will be there without any worries of truncation distortion or clipping. You don’t have to worry about levels at all. Render whatever you’ve been listening to and send it. If you send me 24 bit fixed, I’ll use it unless there are real audible issues, but with floating point we can be pretty sure there just won’t be.

But “headroom” is a serious pet peeve of mine. Nobody with a clue actually needs the mix file to be 6db down to “leave room for mastering”. Leave room for what? To turn it up 6db? If it’s not clipped, it completely doesn’t matter. Most of the time we’re going to “turn it down” via compression and limiting before pushing it back up anyway. Digital gain adjustments are perfectly clean and quiet and quick and easy. If somebody tells you they need you to keep your mix file peaks under -6db (or anything else, really), they have a fundamental misunderstanding of the process and I would not trust them.

What you maybe should leave is crest factor. Don’t slap a bunch of compression and limiting on the master bus just to make it louder and competitive. That’s (part of) the ME’s job. Now if some of that really is part of the sound you’re going for, it can be okay, but do it for those aesthetic reasons, not just to meet some loudness target. The ME might have some better tools or tricks to eek that last couple dB out of it, and if there’s going to be any EQ or anything happening in the process, it will change all those numbers anyway.
ashcat_lt is offline   Reply With Quote
Old 09-28-2020, 10:04 AM   #20
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Why vinyl...? Yeah, I guess it was because of nostalgia. We started 40 odd years ago with self-produced singles and EPs during late 1970's. Now we are just a bunch old farts trying to be teenagers again

But we also wanted to have something to sell at gigs. Unfortunately, then came Corona...

The mastering house was recommended by the vinyl press factory, so I mixed the four songs, sent them as 32-bit FP to the mastering house, and got back four mastered files. The four files I then had to merge into two files, one for each side of the EP. So there I had to arrange the gap between the songs. Those two files then went to the vinyl presser.

But then I also took the mastered files and tweaked then slightly, and they are on as streaming site close to you (Spotify, Deezer, ...)

EDIT: A note on ashcat's reply above (which he wrote while I was typing), the mastering house we used said exactly the same thing. I shortened it to "whatever" in my post above, but what they really said was more like what ashcat says: "we work with what you give us".
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 09-29-2020, 12:25 AM   #21
jrk
Human being with feelings
 
Join Date: Aug 2015
Posts: 2,969
Default

Quote:
Originally Posted by ashcat_lt View Post
Nobody with a clue actually needs the mix file to be 6db down to “leave room for mastering”. Leave room for what? To turn it up 6db?
wasn't that the "secret"?

Obv, You're right. And I wasn't recommending it. My point was that it doesn't matter so much when mixing.

P.S. re Fabians post above, isn't assembling the (vinyl) sides part of the ME's job? Don't we send sides out as DDPs nowadays? Genuine question.
__________________
it's meant to sound like that...

Last edited by jrk; 09-29-2020 at 12:44 AM.
jrk is offline   Reply With Quote
Old 09-29-2020, 10:30 AM   #22
Fabian
Human being with feelings
 
Fabian's Avatar
 
Join Date: Sep 2008
Location: Sweden
Posts: 7,417
Default

Quote:
Originally Posted by jrk View Post
wasn't that the "secret"?

Obv, You're right. And I wasn't recommending it. My point was that it doesn't matter so much when mixing.

P.S. re Fabians post above, isn't assembling the (vinyl) sides part of the ME's job? Don't we send sides out as DDPs nowadays? Genuine question.
I was not requested to hand them DDPs, only two wav files, one for each side. And the mastering guys did not do that, they sent back the same number of files they got.

For streaming media, I submitted four wav (24/44.1) files to the Record Union guy, who then somehow put it all together and added the cover as icon, and submitted to Spotify etc.

So... no DDPs anywhere in sight for me. Maybe for CDs, I don't know.
__________________
// MVHMF
I never always did the right thing, but all I did wasn't wrong...
Fabian is offline   Reply With Quote
Old 10-05-2020, 07:30 AM   #23
Peterk312
Human being with feelings
 
Peterk312's Avatar
 
Join Date: Feb 2012
Posts: 477
Default

Quote:
Originally Posted by ashcat_lt View Post
But “headroom” is a serious pet peeve of mine. Nobody with a clue actually needs the mix file to be 6db down to “leave room for mastering”. Leave room for what? To turn it up 6db? If it’s not clipped, it completely doesn’t matter. Most of the time we’re going to “turn it down” via compression and limiting before pushing it back up anyway. Digital gain adjustments are perfectly clean and quiet and quick and easy. If somebody tells you they need you to keep your mix file peaks under -6db (or anything else, really), they have a fundamental misunderstanding of the process and I would not trust them.

What you maybe should leave is crest factor. Don’t slap a bunch of compression and limiting on the master bus just to make it louder and competitive. That’s (part of) the ME’s job. Now if some of that really is part of the sound you’re going for, it can be okay, but do it for those aesthetic reasons, not just to meet some loudness target. The ME might have some better tools or tricks to eek that last couple dB out of it, and if there’s going to be any EQ or anything happening in the process, it will change all those numbers anyway.
Thanks for clarifying along with what you said above. This is what I was originally concerned about -- if at the mix stage should you be concerned with how low your volume might be even though you've adhered to keeping the master RMS output hovering around -18 dbFS, or 0 db VU. As I deduced, it doesn't quite matter because it can be turned up at the mastering stage.

I was already aware of not compressing or limiting the whole mix simply for the sake of getting it louder, I just never heard the term "crest factor" or knew what it was. It's confusing in a way because I get the impression a larger number is better (i.e., more dynamic range) but I guess it really depends on what kind of music you're mastering.
Peterk312 is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 03:56 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.