Mastering DJ Set Ableton?

[Misfire]

Hey all, wanna say happy holidays firstly.

I am a trance DJ and I arrange sets in ableton, my latest project is my 2009 Trance Year Mix, similar to Armin van Buuren's yearly year mixes. Therefore, lots of transitions, overlays, and mashups are expected.

I am not a master engineer or anything but i've mastered my previous sets with results that I like and are enjoyable to listen to. However, with this project I want to really make it sound as clean as I can.

I would hire a professional, but right now I do not have the money, or I wouldn't be asking for help in a mastering forum . So lets get to it.

My current master rack consists of SSL E Channel (Waves Mercury), SSL G, PSP MixPressor, MixSaturator, PSP Xenon.

I have a few questions on this actually. I want to achieve similar sound to Armin's recently released 2009 year mix. Overall, the mix sounds very clean, low end isn't muddy, basslines come through, the highs are bright, the click of the kick stands out the most, and the snares are wide smack in your face. Transitions are also extremely clean and leveled. There also seems to be more 'space' between the instruments.

I take one of the songs (original) he plays in his mix and compare to his year mix, armin's version sounds a lot better. I realize I probably won't achieve his sound because he used a "professional", but I really just want to get pointed in the right direction.

I will attach clips to so you guys can hear the differences yourselves but first question.

1. I have always done mastering on the first pass, not exporting out as .wav to work with first. I know this is not the best way, so I want to export the set as a wav first so I can work with it. Would I lower the volume of all the individual tracks so the peak doesn't exceed -3 or -6db, reduce master volume, or brickwall it with limiter?

2. What does a DJ set typically need in terms of mastering? I mean the tracks themselves are individually mastered already. I already use a compressor to glue the transitions, mashups, and overlays together, and I know how to use a limiter. Does it need EQing? I know in Armin's mix, it definitely sounds like he boosted some frequencies.

3. Here are the samples, the track is called Cressida - Onyric (Stoneface & Terminal Remix). I arranged and overlay-ed the same way Armin did in his year mix so you can compare. There is a big difference, his highs are crisp and punch through (esp that snare). How can I achieve this effect? or make it closer to sounding like this?

Onyric Original: http://chenzen.org/downloads/onyric%20original.mp3 [no plugins, only peak limiter to brickwall]
Onyric Armin: http://chenzen.org/downloads/armin%20sample.mp3

I appreciate all the feedback and help I can get on this. Hopefully people can share their thoughts on this.

[miscreant]

As an old-time dance producer who masters for a living these days, I've done a few DJ mixes, and more than anything I go for a consistent level across the entire mix. The fact that all of the tunes are already mastered is irrelevant; they all sound different, and the challenge is getting them to sound like they work together. This often means automating EQ and compressor settings throughout the entire mix. I also frequently hand-draw volume adjustments to acheive a more balanced feel from front to back.

DJ mixes are actually one of the more difficult things to master as there are so many disparate elements going into it. Many electronic producers aren't working with the best knowledge of mixing, so it can be hard to wrangle everything to a middle ground.

That said, I don't know that mastering is what put the Onyric track on Armin's mix over the top for you. It sounds to me like it might be a differen mixdown altogether, or there may be an additional loop layered in over the original track. When I was doing mixes frequently myself, I would often drop in extra percussion loops to beef up tunes that needed a little more oomph.

Frankly I like the sound of the original better. The one on Armin's mix sounds too hyped and treble-heavy for me. But, different strokes.

[Jesse Graffam]

Mastered quite a few mixes myself. They almost always come out best when each track is at least fully EQed by myself *before* the mix is mixed/rendered/whatever... Then after it's mixed, final adjustments and final levels n such.

The rest of the mixes I've had to "morph" chains between tracks (by laying down a bit before & after each track and sample-accurate crossfade), which can be a real challenge to keep from having one or the other go all "wonky sounding" during the morph. A pain in the butt, but pretty much the only way to really get optimal results if you're forced to master already mixed mix, and the EQs in the mix aren't so hot.

Armin probably has a lot of the tracks he plays remastered anyways, if they need it... and chances are he's recording his mixes in the presence of a few experienced recording engineers. Every little bit all adds up.

[formlabs]

Remember, when mastering DJ mixes -- each individual track has likely already been processed when mastered for original release. Volume automation and perhaps a TINY bit of limiting to account for any EQing you did, is usually sufficient. No need for more compression in most cases.

Cheers!
Brian

[miscreant]

As with anything, it goes on a case-by-case basis, though I find a dB or two of gentle compression through the STC-8 tends to set the whole mix in a cohesive space.

[Misfire]

thanks a lot for the responses all, they helped a lot!

So when I am EQing, I should be doing it before hand? I lowered the volume of one track so it doesn't exceed -6DB or -3db, and for transitions I would need to lower both of the tracks so the total doesn't exceed -6DB or -3db, this is with EQ applied right?

Afterward, I would export this and do the compression and limiting as needed correct?

[miscreant]

You really can't do this by the numbers. You have to go by your ears. Just because two tracks show -6db doesn't mean their apparent volume is identical.

[Misfire]

Originally Posted by miscreant

You really can't do this by the numbers. You have to go by your ears. Just because two tracks show -6db doesn't mean their apparent volume is identical.

Alright got it, just one thing I'm confused is am I doing the EQ and lowering the volume on the tracks at the same time first, then rendering it out and applying compression and limiting after?

Thanks a lot for the help

[Misfire]

Just to clarify my question,

I know that EQ will raise the volume of the tracks, so I am wondering if its EQ first and making sure volumes hit a certain level. Because I know that when you export it out with -6db or 3db you have the headroom to add things, but if you already EQ it beforehand, do you need to EQ it again?

Can I just do everything in one go?

EQ and watch for volumes hitting the master, then on the master put the mastering plugins (limiter, comp)

also would it be wise to EQ individual tracks or slap EQ on the master?

[Jesse Graffam]

If you're properly EQing each track, then you shouldn't have to apply any EQ afterwards except to compensate for gear (and you shouldn't ever have to do that)... but this assumes that your room is completely flat. Otherwise just do your best before hand with the EQ work, to make it sound like reputably great sounding tracks (on many systems) sound on your setup.

Btw, you can also use an EQ from the approach where you're just turning down the stuff you don't like or sounds too loud. Only compensating for large obvious "valleys" with wide musical EQs (think Q=0.5). And be careful not to notch too thing, or drive too fast. Take your time to figure out how the change you just made effects the song in the long term. That's something you don't have the luxury of doing while you're mixing... plus mixing pre-EQed tracks leaves you with less to think about and can end up in a better mix - performance wise.

[807Recordings]

As I have done more than my fair share of DJ sets, and even helped many in Ableton I can try and give a few pointers, especially on the muddy section.

Make sure that you do not go into the plus dB region on any track. Ableton has its own compressors that come into work when this happens and when you combine this with the output it really can muddy things up.

Second as this is a multi-tracked set (if I understood correctly) I would go over the transition areas and slightly tweak the levels. I find automation is much better than using a compressor. In fact I never use a compressor on tracks already produced as it just leads to a boring, over fatiguing mix that ppl will listen to once and never again.

You said you want it clean so be sparing on eq also. As a Dj who mixes on 3 decks most the time I try and stay away from the gimmics of too much eq. If the tracks fit together you should not need to do it too often. Remember less is more in many cases.

Sometimes depending on your audience a bit of parallel compression can help on the quite parts.


Feel free to ask me anything else if you think I can help.

[Hardtoe]

Originally Posted by 807Recordings

Make sure that you do not go into the plus dB region on any track. Ableton has its own compressors that come into work when this happens and when you combine this with the output it really can muddy things up.

What? Ableton has no "automatic" compressors on the channels? You might overload the mixbuss, but I cant imagine what you are referring too??

[criticaldistance]

Originally Posted by 807Recordings

As a Dj who mixes on 3 decks most the time I try and stay away from the gimmics of too much eq. If the tracks fit together you should not need to do it too often.

not to totally derail the thread, but many many years ago when i first started dj'ing, one big name american trance dj mixed an entire 3 hour vinyl set, without once touching the eq on the mixer... purely using levels only... left the crossfader in the middle the entire time. good track selection goes without saying...

[LFF]

I have mastered quite a few DJ mixes and they always pose a challenge. Getting the levels and EQ right on each song is always fun but never easy. Biggest tip I would give is to go easy on the EQ and compression and watch the fades in terms of drastic eq changes. Make sure you never go into the red.

For what it's worth - every DJ who has had something mastered by me has only returned for more mastering.

[807Recordings]

Originally Posted by Hardtoe

What? Ableton has no "automatic" compressors on the channels? You might overload the mixbuss, but I cant imagine what you are referring too??

Right, Try sticking your sine wave in and pushing it over 0dB and then render that output. I am sure you will find it not exactly the same. There is a compression on the bus if over loaded. Many thread talk about this and more so on the master bus.

Before you rant off about what you do not know pay a little more attention.

[Hardtoe]

I'd hardly call my reply a rant, but in any event I was just curious, not meaning to insult you?

Anyway, I think this is just a digital overload (ie clipping), not any kind of compression.

No offense intended, just interested in what you were saying.

[807Recordings]

Originally Posted by Hardtoe

I'd hardly call my reply a rant, but in any event I was just curious, not meaning to insult you?

Anyway, I think this is just a digital overload (ie clipping), not any kind of compression.

No offense intended, just interested in what you were saying.

Don't worry I don't take offense but rather making a point. I have done a lot of recording multi-tracking and tried various DAW and solutions. Even mutli tracked to two inch tape many a times.

Best solution I found for me was my Prism and Samplitude mutli-track. I used Ableton of course many a times and the one area I repeatedly noticed was when the busses or channels got hot. Of course the audio warping also made a difference and I personally don't use it. You can always bring levels up after you dump down the mix to a stereo file but if you have clipped it then that is hard to fix, if not impossible.

Other side note if possible I found was to make sure you are recording in a 24bit resolution and exporting the same.

[aof21]

Originally Posted by uncajesse

Mastered quite a few mixes myself. They almost always come out best when each track is at least fully EQed by myself *before* the mix is mixed/rendered/whatever... Then after it's mixed, final adjustments and final levels n such.

The rest of the mixes I've had to "morph" chains between tracks (by laying down a bit before & after each track and sample-accurate crossfade), which can be a real challenge to keep from having one or the other go all "wonky sounding" during the morph. A pain in the butt, but pretty much the only way to really get optimal results if you're forced to master already mixed mix, and the EQs in the mix aren't so hot.

I would second both of these points. If you have control over the individual elements in Ableton, try to get your mix as smooth as possible before applying anything to the master bus (or before exporting the stereo file, however you end up doing it). You can't rely on a compressor or limiter to smooth out the levels between songs because then you will end up with some songs possibly too low and others being flattened by the comp/limiting and it won't sound so good.

Originally Posted by SafeandSound

A touch of fader automation on the quieter tracks (up) and the beat matches(down)
should do this trick if you are feeling naughty take a dB off the top with your favourite limiter.
and if it's destined as a compressed format stop it at -0.3dBFS(ceiling) to avoid overshoots when it's an MP3.

and I would also +1 all of that advice.


Set your levels so that you can volume automate up the quieter tracks. If you run into a situation where you have one or two tracks that are just really weak compared to the rest of the mix, then you have two options:
a) just don't use those tracks in your mix! or b) you may have to completely "re-master" them before you put them into the mix to see if you can get them to match better with the other tracks in the mix.

don't go too crazy trying to "improve" the sound of the entire mix with some special mastering chain. Unless you are really experienced and have really great gear to work with, there's probably not a lot you can do to "improve" the sound of the already mastered stuff. Ideally, you are aiming for just balancing and leveling. If you get too overzealous trying to put some magic chain across all the tracks to "give it your own sound" it can be easy to overprocess you don't realize it until later after you've heard your mix on a lot of different systems and realize you actually squeezed some of the mojo out of it (speaking from experience here unfortunately!

[Misfire]

Originally Posted by aof21

I would second both of these points. If you have control over the individual elements in Ableton, try to get your mix as smooth as possible before applying anything to the master bus (or before exporting the stereo file, however you end up doing it). You can't rely on a compressor or limiter to smooth out the levels between songs because then you will end up with some songs possibly too low and others being flattened by the comp/limiting and it won't sound so good.



and I would also +1 all of that advice.


Set your levels so that you can volume automate up the quieter tracks. If you run into a situation where you have one or two tracks that are just really weak compared to the rest of the mix, then you have two options:
a) just don't use those tracks in your mix! or b) you may have to completely "re-master" them before you put them into the mix to see if you can get them to match better with the other tracks in the mix.

don't go too crazy trying to "improve" the sound of the entire mix with some special mastering chain. Unless you are really experienced and have really great gear to work with, there's probably not a lot you can do to "improve" the sound of the already mastered stuff. Ideally, you are aiming for just balancing and leveling. If you get too overzealous trying to put some magic chain across all the tracks to "give it your own sound" it can be easy to overprocess you don't realize it until later after you've heard your mix on a lot of different systems and realize you actually squeezed some of the mojo out of it (speaking from experience here unfortunately!

thanks a lot for this post and the others that contributed. I definitely did find leveling up all the tracks and transitions equals a much better mix! That was the biggest part for me, because I used to mix at full volume and just slap a compressor on it with -4 to -6db reduction, sounded good at the time, but now with minimal compression it just sounds even better. Volume automation truly is better than compression.

But whats even better I found after you render out a nice level'd out file, run it through a nebula neve emulation and omg, I am getting the dynamic range of Armin's mix. I have pretty much achieve what I wanted from the start, larger dynamics and perfectly level'd mix.

His mix still have those really hyped highs that feel good to me at least, but I decide to back off from that and apply a little bit of color instead, it sounds great already.

EDIT: I actually learned a lot from posting this thread, thanks a lot guys. There really isn't much going on in making a nice DJ mastered set. Its all about having that volume automation going on, dont need to EQ that much. But man the most drastic difference is that neve 5088, it opens it up so much I love it. I'll be sure to post the set here for you guys to check out.

[aof21]

[filin]

Originally Posted by 807Recordings

Right, Try sticking your sine wave in and pushing it over 0dB and then render that output. I am sure you will find it not exactly the same. There is a compression on the bus if over loaded. Many thread talk about this and more so on the master bus.

Before you rant off about what you do not know pay a little more attention.

Hi

You're spreading lies / talking crap. The channels do nothing of the like that you're speaking of, the master clips if rendering above 0db (as all do).

Bye

[807Recordings]

Originally Posted by filin

Hi

You're spreading lies / talking crap. The channels do nothing of the like that you're speaking of, the master clips if rendering above 0db (as all do).

Bye

And what is clipping then if your such a genius?
I suppose you rather just post rants instead of trying what was suggested.

It is not a lie but rather something we noticed. When pushing a channel too hard you will alter the signal. I am not sure on the latest version but it was a factor on early versions.

[Misfire]

Thanks guys, you helped made this!

Here it is, released to the podcast.

You can get it here

Episode 58 (Year Mix 2009) | Danny Chen's Trance Fixation Podcast

direct DL: http://media.blubrry.com/trancefixat...r_Mix_2009.mp3

[filin]

Originally Posted by 807Recordings

And what is clipping then if your such a genius?
I suppose you rather just post rants instead of trying what was suggested.

It is not a lie but rather something we noticed. When pushing a channel too hard you will alter the signal. I am not sure on the latest version but it was a factor on early versions.

I really don't see why you need me to define clipping for you, I would hope that you already know the ins-and-outs of digital clipping, particularly as you're here in the Mastering Forum.

I will however other you a definition of rant; " A rant is a speech or text that does not present a well-researched and calm argument; rather, it is typically an attack on an idea, a person or an institution, and very often lacks proven claims. "

Now, please go and open up Ableton and try what you yourself suggested. I already have (i did so before my first post in this thread).

Then apologise.

[Hardtoe]

Just for the record I didn't rant either...

[807Recordings]

Originally Posted by filin

I really don't see why you need me to define clipping for you, I would hope that you already know the ins-and-outs of digital clipping, particularly as you're here in the Mastering Forum.

I will however other you a definition of rant; " A rant is a speech or text that does not present a well-researched and calm argument; rather, it is typically an attack on an idea, a person or an institution, and very often lacks proven claims. "

Now, please go and open up Ableton and try what you yourself suggested. I already have (i did so before my first post in this thread).

Then apologise.

I am well aware of what digital clipping is but I am not in the mood to get into this.
Anyways as I do not use Ableton anymore just due to preference I found a test someone did on this.

mnml.nl :: View topic - The Ableton headroom test


Seems indeed on the 32bit system which is now a 64bit system for internal summing (correct me if I am wrong) you have about 35dB of head-room. My tests showed different results but I can not go back to this computer at the moment because it is stuck in Canada. If it was an error somewhere in the path then so be it.

Good practice either way is to keep your levels lower. When adding in EQ or similar pluggins they may not be able to hit the full 32bit or 64bit headroom.

[filin]

Originally Posted by 807Recordings

Anyways as I do not use Ableton anymore just due to preference I found a test someone did on this.

mnml.nl :: View topic - The Ableton headroom test


Seems indeed on the 32bit system which is now a 64bit system for internal summing (correct me if I am wrong) you have about 35dB of head-room. My tests showed different results but I can not go back to this computer at the moment because it is stuck in Canada. If it was an error somewhere in the path then so be it.

Actually, if you read that link more carefully you'll see that headroom is closer to +65db.

This is besides the point; you were incorrect about the channel compression issue, and you were out of place to accuse others of ranting when, in actual fact, it has been you handing out unsubstantiated claims.

But anyway, this is boring now.