Mixdown levels for mastering

[BassBob]

Hi, I'm at the mixing stage of my project. When mixing down tracks what is the
proper range the master faders meter should be at? In terms of levels where
should the master faders absolutely not be? Thanks

[MiltonRob]

Keep them around -6 db, try not to go over that point. It leaves room for the Mastering Engineer room to do the magic they do

[Mark Bliss]

-6 dB (peak) is a common recommendation, but it does vary per the preference of some mastering engineers/services.
Technically, I generally wouldn't have a problem taking a shot with anything that isn't pushed to close to peaking at 0 with no limiter/master bus processing.
Bottom line- Depends

[jlouvar]

Hi, I'm at the mixing stage of my project. When mixing down tracks what is the
proper range the master faders meter should be at? In terms of levels where
should the master faders absolutely not be? Thanks

Hi and welcome to the Mixcraft forum.

I asked this same question a while back, and there doesn't seems to be an absolute answer. So with that in mind, what I do is listen for and correct any and all tracks clipping (gain staging), and just keep the master bus/track from pegging the red. In other words, if it sounds good, it is good.

[Ian Craig]

My answer is that it depends what you are doing. If you are putting music on the internet directly, you could try using a wave editor program to check the levels before you upload the thing and if you do that then you could happily set the master fader in Mixcraft to -6db. I did read here a year or two ago that the master fader should be set to 0dB and the track levels should be kept low to avoid distortion (it was useful advice and since then I have used that as a standard). I don't know about sending things off to be mastered by someone else, sounds expensive to me (I've learned to do it myself). Also, if you are encoding to mp3 the resultant mixdown file should not be above -1.5dB to allow room for the encoder to work cleanly. If you mixdown to wav you don't need to worry about that you just need to make sure that it does not go above 0dB.
It would be possible to mixdown to wav using -6dB then import the resulting file into an empty Mixcraft project and normalise it in Mixcraft so that it then stayed below 0dB and then mixdown again. That would keep the overall levels under control so it wouldn't be too loud or too quiet when you go to use it.

[Mark Bliss]

I agree Ian, and perhaps I misread the intent of the OP as well.
Level while mixing is different than level for mixdown, and I may have misunderstood. I thought the intent might be preparing for mastering. In re-reading I think I may have been wrong.
While mixing, I target a maximum of -6 peak, but am not super particular about it. ("Give or take some") and my master fader is always left at 0 (unity)
Methods vary.

[dpaterson]

I've a whole thread on this very same topic if some can remember.

I use Lurssen, TRackS-5, or iZotope for mastering (basically do three different masters and decide which sounds the best). In all cases though: it would appear that they all "expect" a -6dB input in order to comfortably do their thing. Put another way: the higher the input level to these things the less they are able to do. Based on my (now extensive) research, testing, and experience: this is what works. You can always push the master using the mastering software. But it doesn't work the other way around i.e. you cannot feed the mastering software an input level of just below 0dB and then a) expect it to produce a decent master or b) lower the output level of the master because it's now too hot (this just results in unbearable compression).

[BassBob]

Thanks to everyone for taking the time to answer my question. To be specific,
I'm mixing down in preparation for mastering. Again, thanks everyone I really
appreciate the help.

[jlouvar]

Thanks to everyone for taking the time to answer my question. To be specific,
I'm mixing down in preparation for mastering. Again, thanks everyone I really
appreciate the help.

In that case, keeping the signal close to -6dB with the master fader set at (0) unity gain is probably your best bet.

[Mark Bliss]

Again, I would ask the person doing any "mastering" work for you about their preferences.
If that person were me, I would suggest the following:
1: Render a copy of your mix for reference, as you mixed/mastered it. Peaking at no higher than about -1 would be good. Just no clipping.
2:Turn off all master bus processing and set your master level (or a gain plug in if necessary) to obtain a mix down level peaking at about -6
3:Share some other references to obtain an idea of what kind of sound you are going for.

Hopefully that's helpful.

[dpaterson]

Also, if you are encoding to mp3 the resultant mixdown file should not be above -1.5dB to allow room for the encoder to work cleanly.

This is interesting. Is this based on personal preference and experience or is there a technical reason for this??? Not questioning you at all but having never seen such a statement or recommendation made (and I mix down to .MP3 a lot of late to share test stuff on SoundCloud): just wondering if there's a technical reason and why.

Notwithstanding your answer I currently work like this:

• Mix out from Mixcraft to .WAV file ensuring -6dB peak/-18dB RMS (-12LU RMS).
• Open .WAV file in Sound Forge Pro 12 and remove any DC Offset and save to .WAV file.
• Master .WAV file in Lurssen, TRackS-5, or iZotope Ozone 8 Advanced pushing the mix but ensuring that peaks never exceed -0.1dB (and no clipping obviously).
• Pull up mastered .WAV file in Sound Forge Pro 12, check peaks, normalize to -0.1dB if necessary, add fade ins and outs, and save to highest available quality .MP3 file.

Are you saying in that in the last step I should be normalizing to -1.5dB and then save as .MP3???

[Ian Craig]

Yes, it is common knowledge that encoders need headroom and I think it was expressed by iZotope in the RX6 manual and various other places that it should be -1.5. Obviously, this is just followed practise, I haven't been analysing for clipping, personally I go for -1.7 just to be sure. It could be bollocks, I don't know. Right, I'm watching a film (a rare thing for me)

[Mark Bliss]

The common recommendation on mp3 (from reliable sources) has been to allow for some headroom due to inter-sample peak errors in the conversion codec.
How much advised varies. The most common recommendation I have seen refer to older codecs as compared to more modern codecs.
Older recommendations recommended 1.5 or more, but it is commonly stated that more modern mp3 codecs, (which include the LAME codec built into Mixcraft) are "safe" at -1dBfs.
But we are splitting hairs, and allowing the loudness war nonsense to cloud our judgement to believe this is a big deal. Just leave some headroom either way IMO.
https://www.musictech.net/tutorials/10m ... ple-peaks/

And Dale- reminder. For Sound cloud submissions, mp3 isn't necessary, and in fact, if you desire- uploading a wav file allows users to download a quality file, (while streaming is of course still an mp3)

[dpaterson]

Yeh. It’s just that right now I’m often uploading sort of test tracks for that project over at Seymour Duncan so .MP3 uploads are quicker is all. Because it’s such a long song a .WAV file is +170MB and our upload speeds are way slower than our download speeds so it takes time (and takes even more time if you sit and watch an upload!!! LOL!!!). Quality not really an issue at this stage. Just never knew this about .MP3 files is all.

One interesting factoid about SoundCloud is that if you enable the “Direct Download” permission for a file then the original uncompromised or unoptimized file is what is available for download i.e. without any SoundCloud optimization. The streamed file is of course optimized.