Mixing vs. Mastering

I’m curious how this group would distinguish between mixing and mastering. LearJeff–I think I read one time (on your website?) that you thought the line was greyed between the two (or could be) – I think I agree.

Obviously, a missed note is a tracking issue and should be fixed there. A rolloff in the EQ is a mixing issue, along with some compression, effects, etc. But what are the things that I need to be saying “this goes to the mastering section”?

So, let me phrase the question this way: At what point do you actually dump the song from N-track (let’s call this the mixdown stage) and begin working in another tool (Wavelab, SoundForge, Ozone, etc.) or send it to the Mastering house? Next, what are the requirements that you look for when saying “that can’t be fixed anywhere else than in Mastering and I think I’m done”? While I realize that much of this is subjective (I know, I know, use my ears :D ), I also have learned that there comes a time where the added effort tends to start screwing UP the mix instead of making it better.

Thanks much and looking forward to the responses.

Mixing is the process of adjusting the relative levels and EQ of the individual tracks to create a balanced mix.

Mastering is the process of adjusting the level and EQ of the finished mix to create a pleasing overall sound, and one that is relative to other tracks to be included on the same CD.

Mastering is also to prepare the track for insertion onto a CD - 16bit/44.1KHz, and to edit the intro/outro of the track.


At our place, we consider anything that you might do to the stereo mix part of “mastering”. We (meaning me and most of the freelance engineers that work in our space) pretty much do nothing to the stereo mix once it’s been made. This includes compression or limiting, sweetening EQ, etc. but generally does not include running the mix to (or through) one of the 2 track decks. So even we’re kind of loose on the terms…

There are a few real good mastering houses here in Chicago, and we prefer to let those folks do their thing on clean mixes. Sometimes, even running the mix across a tape machine can impart a sound the ME doesn’t like (this happened with a recent record…the computer files won over the mix off the Studer). Mind you, this is when sending stuff to an ME rather than doing it yourself. Real mastering facilities have equipment and listening environments that blow away what most people have access to, and when you’re paying money for a fresh set of ears that are focused on a “mix”, supplying such mix with extra compression or EQ is generally a bad idea, since that can compromise the final result. That’s why the Studer tracks got sent back…

If you’re doing the mastering yourself, I say get the mix done as best as you can with nothing on the master buss. Once you get it sounding as good as you can on a variety of different palyback systems, then you can consider it “mastering” time. If the mix suffers when you start limiting, adding EQ or whatever to the master buss, then either the mix wasn’t done in the first place, or you’re being too heavy handed with the 2-buss effects. See, that’s the thing about real mastering people and their gear. The difference between the raw files and the mastered ones can be subtle, and it takes well made stuff and experienced ears to hear the positive changes from the over done ones (or more importantly, the point where “good” crosses over to “too much”). Making radical changes in the sonics of your mix when trying to master it is usually a good sign that you have more work to do on the mix…

[quote=John,Aug. 04 2005,16:57][/quote]

Thanks for your reply. Sorry that I was unclear in my question.

I understand the purpose of each fairly well, I think. I’m looking for instruction. As I stated in my original post, I’m interested in when you might finish with N-track and move into another application. Let me give you an example, lest I be misunderstood.

Say that I’ve got 5 perfectly recorded tracks in one song – no mistakes. I’ve EQ’d, minimally compressed and “effected” each track and I’m happy with it. Kinda. I take 15 minutes and mix it down in N-track (with or without mastering effects/compression if I’m only going to do it again in Ozone??? Oh, and by the way, Ozone makes a great VST and DX plugin, so I don’t really have to leave N-track to hear the changes or mix it down–maybe that’s part of the confusion here…), port it over to the mastering software (I could do it as two stereo tracks which is alot easier on my Dell) to see if that makes enough sonic difference for me to be happy with it. I decide that I am. I burn it to a CD, and take it over to your place, we have a couple of beers (my treat, your stereo) and we both agree that it’s too whatever (muddy, bright–you pick).

I go back home with all the thoughts you’ve articulated, and I’ve dutifully jotted down on paper. I think I’ve got it this time…

However, I’ve now buried the cymbals…or kick…or whatever. You pick. Now, I’m down two CD’s and a 12-pack, and am starting to wear out my welcome at your pad, because your girlfriend hates Country music. Or punk. You pick.

So…I think to myself…“HEY! Why don’t I find a way that says “this is as good as I can do on the MIX”. Once I’m happy with that, then I’m going to force myself to figure out how to do a better job at MASTERING.” Then I think to myself…“HEY!!! I know a GREAT forum where I could ask this question, and get a decent response!”.

Your turn. :D

I’m sorry to say that I have nothing valuable to add to this thread, other than to compliment billthecat on a very clever response!! :D

In all seriousness, I have similar questions… thanks for asking first! :p

I think clava’s bang on the money. Once you leave the individual tracks behind and start messing with the stereo bus you’re entering the mastering zone. Generally speaking…

Mastering is also making each individual song fit in with its’ sorrounding songs.

An occasionally wise man once told me : Mix for content, master for sound.
Oh well…

One thing I found helpful was to bring in a song from a record that I really like and analyze the eq and levels. I found a great native meter that follows Bob Katz’s level system and a frequency analyzer that has improved my mixes and masters. Here is a shot of them.


Why don’t I find a way that says "this is as good as I can do on the MIX

See, this IS the crux of the problem, and I don’t have any magic bullets other than keep a lot of notes, burn a lot of CD’s, and buy yer most understanding friends a few beers…

Jeromee has a good idea with the analysis software approach, especially if you don’t really know what you should be listening for. More experienced mixers can listen to their reference material against their mix job and get an idea where any deficiencies may lie without looking at something like a spectrograph, but that kind of tool can sure be useful.
I find it best to focus on one or two things that I think hamper a mix and get them to my satisfaction first…“Hmm…too heavy handed with the 'verb on the percussion, and that bass is muddy…” Fix that stuff and check again. Either the glaring problems are covering smaller nagging ones (which can cause you to wear yourself out almost needlessly) or the glaring problems are the only real bad spots and once you get 'em the mix is done. The other thing I found was that my mixes all had the same sonic footprint that I didn’t like (i.e. too muddy in the bass, or too crispy or too rolled off on the high end, or whatever). Here, analysis tools can help you determine what that footprint may be, and how well you’re doing at changing it. Once you’ve done this enough times you get a much better idea what to listen for right off the bat and mixes happen faster. It takes practice and listening, and sometimes it can be a drag.

Also it never hurts to save your “new gear” money and spend it instead on a little time with a pro ME. Pick one song that is representative of the sonic footprint of your mixes and have the ME work it over. Best if you can be there when it happens. The good ones are willing to help you find out where you can better your efforts, and it needn’t cost more than a $200 or so. Like paying for private instrument lessons when you get right down to it…

Cava’s dead right, but Bill’s question remains – when do we stop adjusting the total sound by fiddling with tracks and start fiddling with the stereo master?

The answer is, unfortunately, pretty boring: it’s a matter of judgement. The more you do it, the better you get at recognizing the point of when to quit fiddling with tracks and focus on the master.

One clue is just like you said: when track fiddling stops improving the mix.

More practically, for most of us home-recording folks, we move on to the next stage when we’re sick and tired of the previous stage, even though we can hear lots of mistakes and potentially fixable things. When we’ve reached the point of diminishing returns on our efforts.

There are some general guidelines, though. When mixing, don’t pay too much attention to the overall loudness and dynamics as far as listenability is concerned: focus on the dynamics of the SONG and the PARTs, but not on whether “it’s too loud here but too quiet there, and I need to pump up the overall levels a lot anyway”. Focus more on the compositional aspects of dynamics rather than the technical ones.

Another part of mastering that’s definitely not mixing is when you make adjustments to songs so that they fit well together in a collection. But that’s obvious, I guess.

To muddy the waters more, though, for many pro mastering jobs, the mastering house doesn’t just want a stereo mix. They want “mix-minus plus seps”, if I remember the terms correctly. For example, a mix without vocals or bass, and separate bass and vocals submixes. But the reason isn’t so that they can remix the song for you, it’s so they have more precise control over dynamics. Obviously, it’s nice to be able to EQ and compress the bass separately – which is why we use a multiband compressor. Separating the bass out gives more flexibility. Having the vocals separate means they can compress the band’s attack without hearing the vocals duck at the same instant (a very annoying artifact, and sometimes difficult to twiddle a multiband compressor to avoid when doing heavy compression).

But that’s pretty advanced, and very time consuming, and I bet lots of pro work is done without going that far. Way more time consuming than I can manage!

I might not be contributing much to this thread but i can definitely tell you that you really need to put the mix away for a while to get a fresh perspective.

I was tracking and mixing a few songs myself last December for 2 weeks and I was nitpicking at EVERY thing…“oh this pitch is off” “too muddy” “too bright” etc etc.

Needless to say…I was discouraged.
I gave up and put it away for 4 months until a producer i am working with wanted a rough demo of the songs i am recording in a professional studio and he thought they sounded really good as well as 4 session musicians i had to give copies to.

My point is…i think the moment when you are dragging a project too long for sonic is the time when u need to stop twiddling the knobs and come back ot much later and judge for yourself. There is NO way you can acquire objectiveness when you are recording/mixing AND mastering your work yourself

Leveling various tracks or volume matching on the CD so you don’t have weird level jumps from tune to tune is a part of the mastering job.

I’m with others who have said that any processing on the multitrack is mixing and any done on the two bus is mastering.

So, then. Those of you who have VST type mastering software – Do you “master” on the fly while within N-track, thus making Ozone (or whatever) a hyped up Mixdown, or do you force yourself to wait until you’ve committed it to a stereo mix? :;):

really appreciate all the responses.

Of course, Einstein disproved the validity of concepts like; synchronicity, causality, and consequently, before and after.

So, why not try mastering first, then using “Ali’s Patent Track Disassembler” do the mixing afterwards?

As someone on here used to say; “Use your beers!” :cool:


Bill, I generally separate it into two stages, for a number of reasons. One is the obvious “head space” thing. Another is that I usually master a collection of tunes together for a single release (CD, EP, etc), so I import 32-bit master mixes from each song into a single n-Track song file for the CD. This allows me to:

a) put stuff on the master channel that I want the same on all cuts, like ambience master reverb, helping give unity to the collection
b) quickly change from one song to another to compare
c) drag files in different orders to hear what they’d sound like one following the other

and things like that. Very handy. When I’m done, I drag all the files to the left and mixdown each one independently (using Solo button).

There’s another reason why I advise against putting plugins like compression on the master channel when mixing. It makes mixer controls operate in a very counterintuitive way. For example, raising the fader on one track can cause all the other instruments to get quiter, without affecting the volume of the first track much. So, unless you really understand what compression is and does, I advise against using it on the master channel during mixdown.

Another reason I avoid master channel FX in mixdown is that I forget to turn 'em off when I make a submix for some purpose, or render a MIDI track to audio. With the new Freeze feature (which I don’t have, running V3 at this time), that might now be a non-issue.


As always, a very thorough response. Thanks much for the insights. Anyone else?

If I need to do “much” while mastering I usually go back and remix. By “much” I mean any major overall EQ fixes, or de-essing, or anything that can actually change the mix very much.

This is where I has a different philosophy that many other folks. I don’t mind keeping effects normally reserved for mastering enabled while mixing. That said I keep a very keep eye on them while mixing. If there’s a mastering limiter in the output effects insert it very easy to forget about it. When that happens there’s a chance of turning something up that can’t be heard, but all turning it up does is bang against the limiter harder. Eventually that track can be heard at the expense of the limiter pushing down the other tracks overall. That’s NOT a good thing unless that the effect you are going for. Trent Rezner (NIN) does that kind of thing regularly, but it’s something that is generally a very bad idea. The same hold true for any effects in the output insert. You can really screw yourself doing this and I do it regularly.

Of course, the argument can be made that using what are considered traditional mastering tools this way is not really mastering, and I’d agree with that. With that thought in mind, whether agreed upon or not, mastering is what’s done to make a set of individual mixes be grouped together as a single body of work - to take a bunch of recordings done over time that may not be sonically coherent and to get them ready for an album by making them all have similar/appropriate volumes and EQ – you want them to sound like they came from the same session.

So, mastering a single song is one step. When all is done, it’s time to remaster.

We love Phoo because he’s the exception that proves every rule! :wink:

We all know that rules are made to be broken. Beginners can follow them and stay safe (but with no guarantee of success). Experts should know the rules, the reasons for the rules, and the reasons why they’re breaking them. What Phoo’s doing fits more in the latter category here.

My way of thinking
- I don’t use any effects / automation / equalization at the master channel during mixing.
- The mixing will finish only when I say: This song won’t need mastering.
- Mixdown at the final midia quality (If CD´s: 44.1 KHz/16 bits).
- New Project with the mixed track.
- MultiBand Compressor / Eq / Reverb
- Work with Master fader to have 0 dB peaks.
- Sometimes (ok, usually) back to the mixing phase

Me break the rules? No, never. I make the rules. Of course, they are mine and yours…ummm…anyone else’s WILL be different. :)

Actually, I use the stuff in the output as tools to let me know when I shouldn’t be using them. If I need to EQ in there to get the overall tone sounding right then I look at the EQ curve and I go back to the mix and change things until the EQ doesn’t need to be used. Likewise, I love Har-Bal, but I’ve gotten away from using it except to look at the EQ curve. I rarely change a mixdown in it anymore – If I need to I go back to the mix.

The same holds true for limiting in there. If the mastering limiter is nuking any more than a few db then the mix is too hot overall going into it. It’s much better to knock off 1 db at a time in a few passes than it is to knock off more in one pass. I do like having it there for the sound. I keep endorphine in there for a similar reason, but it’s only very slightly touching the mix.

After lots of mixing and mastering I found that I have end up doing the same things to almost every song. That’s were I came up with this combination. To be sure no one gets the wrong idea, I do not start with any of these enabled except the mastering limiter. That is set to no gain with the hard limiting threshold a -.5 db. It stays there until the very end. It’s only when I get a mix where I like it do I slowly add these in.

The only other effect that’s always there is MDA Image. That is rarely used when doing a legitimate mixdown. It’s in there so when it’s enabled the stereo is flipped - left and right become reversed. That so the mix can be A-B-ed for center balance and spread. The sound should not shift from side to side when flipping, but it should stay centered. I have a hard time with centering because i have a bit more of of a hearing loss in one ear than in the other. Something truly centered and perfectly even on the meters might sound off centered. Flipping left and right exposes truly off center stuff.

I also A-B having these enabled and disabled. The mix should pop out with the stuff on but not change except to get a little louder-fuller-clearer and only a little better at that. As I said earlier, if there’s too much done in there it’s time to go back the the unmastered mix.

And after all that…yes I really screw myself sometimes, really bad. I really depend on the fresh ears of other listeners - mostly teryeah, Sean and fytrius. One thing that is the hardest to learn is not the mixing itself. It’s the ability to listen to and hear others unbiased criticism in a way that can be used constructively. A LOT of ego goes into the whole recording process. The last thing anyone wants to hear is that something they’ve worked on for hours/days/weeks/months sounds like crap to someone that is hearing it for the first time. In most cases if someone says “that sucks” it probably does. Finding friends with ears you trust is hard, but must be done. And those extra ears need to get past the “that suck” comments and need to learn how to translate “that sucks” into “that suck because…” and they need to be able to explain it in a way that is truly helpful. THAT’S the best way to get a good mix.

By the way, posting a new song in this forum for critique is one way to get these kinds of relationships started, but it’s not a god end-all. That will happen overtime when you get close to a few individuals that hear a lot of your stuff over time. First time listeners can hear what needs to be done, but they probably won’t tell it like it is since they don’t know you, or they will tell it like it is but that won’t go over well because you don’t know them. Also, a lot of insight can be picked up over time of hearing a lot of different stuff from someone. The real problems in the mixes will be come more apparent when it becomes obvious that every mix has the same little “things”.

Oh well…that was an interesting tangent…time to eat something. :)