Mastering plugins - when do you use them?

[knightime]

Hi guys,

I've been reading, with great interest, a few posts debating the use of compressors and limiters on the Mastering channel of a project.

Now whether to use them or not seems to be quite a hot topic for debate, but I was just wondering how you all feel about the use of Mastering plugins, more particulary when you use them.

(Now don't shoot and abuse me!) I currently have two (yes, two!) Mastering plugins on my Master channel, T-Racks on the first insert and WaveArts Mastering plugin on the second. This has the effect of making the project I'm working on as loud as commercial tracks. I adjust individual channels as needed, cutting/boosting EQ as needed, and everything seems to work fine.

I was just wondering what you guys do in terms of using these types of plugins on your Master channel during your project - do you use them, do you not? Is it preferable to render to audio then mix/master later on?

I feel it's easier to mix/master as a project is going on - what do you guys think?

Cheers!

[billybobakimbo]

Personally I prefer to kinda get the mix right as I progress through the track (sound design, programming,arrangement etc), but this method only evolved for me as I moved away from mixing on a 'desk' to an 'in the box' system (within the DAW).
As to when to master and when not.. that's a biggie.. Here's a couple of threads at the new DV Forum which may be of interest to you..Mastering for Release..
and..
Multi Band -Friend or Foe

[Delphine]

I guess it's down to the individual, I prefer not to have anything on the master and have dynamics processing at the source or across busses but some people I know ( esp those who write dance stuff ) Like to strap at least a limiter over the master and switch it in and out during mixing to get an idea of how their track sounds when it's got commercial RMS levels.

As far as 'mastering' goes, if you are doing it yourself, I guess that's the same thing. If I do that I have always done it in Wavelab because of the burning to CD, start and end time and mainly because of the CPU issues. I couldn't run a whole track with plugins used for mixing, then add even more on the master to try and do a self mastering job.

But the main thing with me, was to burn the 'unmastered' mix, then check it on other systems as when I burned a 'mastered' version and found issues, I found myself going back and tweaking the mastering plugins, when really it should have been sorted at the source sounds in the mix.

[knightime]

Interesting comments guys.

I don't plan to have my tracks 'professionally' mastered as I feel I have the tools and (20 odd!) years of knowledge to be able to do it 'right' (although I'm still learning new techniques daily!).

There seems to be a 'Pandora's Box' issue when it comes to mastering and 'loudness', and I know many an article has been written discussing this issue. One of the reason I started applying the mastering plugins on my master output channel was due to the reference material I had chosen as a base. Without the plugins I didn't feel I could accurately compare my tracks to the references because they had been mastered so 'loudly', and so to get a reasonable comparisson I felt the need to apply the plugins during the composition stage. If I hadn't applied them and just used the volume knob on my audio inerface I don't think my monitors (or ears!) would be intact right now!

Modern music these days is loud, and I don't think there's any turning back from that. Unless someone with a whole loada clout, someone well and truely established in the idsutry, turns round and says, 'Let's stop this fasion for loudness and go back to quality', I think we're all gonna follow the trend to try and get the loudess mix we can get.

I think, Delphine, that you hit on something that may be significant. One of the reasons you said for not putting the plugins on your master channel was to do with CPU issues. Are you saying that if your PC had more power that you might actually find youself using mastering plugins at the composition stage too? I'm wondering, as our computers become more and more powerfull, whether these types of plugins become 'the norm' in the search for that 'loud' mix? Hmmm...

[Danny1969]

I do the same as you Knightime, compressor and / or limiter on the master bus so I can hear what will happen to the drums when it gets mastered, or what offending vocal sib is magnified by the compression.

I think the main thing is to do two bounces. One with the master bus processing and once without, just in case you do need it mastered by someone else. I always lower the overall gain of the one without as well, no more than -8 dB peak, I figure the more headroom I leave him the better.

In terms of mastering yourself I would sooner send it out if it was an important project. The trouble with digital is the more processing you do with it the worse it sounds. Some of the mastering houses that are affordable have some great analog EQ's and compressors, and they can get the level pretty hot without it sounding all smashed and crunchy.

[Delphine]

Originally Posted by knightime

I think, Delphine, that you hit on something that may be significant. One of the reasons you said for not putting the plugins on your master channel was to do with CPU issues. Are you saying that if your PC had more power that you might actually find youself using mastering plugins at the composition stage too? I'm wondering, as our computers become more and more powerfull, whether these types of plugins become 'the norm' in the search for that 'loud' mix? Hmmm...

Well I wasn't in that boat until recently. I now have a dual core system which is perfectly capable of running that but I have yet to try. Certainly, it would be quicker that's for sure but I have a great liking for Wavelab and its numerous tools, and generally prefer working with a visual of the stereo file. No reason other than preference I guess and being used to working with something I have got to know. I also used Reason for ages and obviously there's no VST support.

There's no link whatsoever though between loudness and doing a self mastering job in your daw app or an editor like wavelab. How your mix your tracks and the sounds used will have a lot of impact though. Getting rid of extreme low end by HP filtering, evening out dynamic ranges within parts and generally having a good solid mix is going to give you a stereo file that should be pretty easy to get commercial RMS levels if that's your thing.

If you find you need to use a commercial track for 'loudness reference', you might want to look at using an audio editor anyways. You can grab a section of it and perform a global analysis which will give you a reading of the RMS levels - which are responsible for percieved volume. You will find a lot of tracks in the same genre which have been released recently, have roughly the same levels and you can compare your own mastered stuff against those. Then again, you might have plugins which give RMS levels, I don't know.

[Ten Flew South]

It really is all down to the mixing. I've been focusing on songwriting for ages as that was a priority to learn (obviously ), but now I'm examining the technical side more, and the way you mix is going to determine the end level.
A very quick and good example to show RMS vs. Peak.
If you have 2 guitar parts recorded, and playing simultaneously, at some points where the signals combine, you will get a doubling of the peak level, but you will not perceive the sound as being twice as loud.
Losing that much headroom unnecessarily is going to be the biggest obstacle in getting a loud 'master'.
The only way to avoid it is by being obsessive about frequency seperation, both at the arrangement/instrumentation stage, and at the mixing stage. If you can achieve a great balance between frequencies, then you will encounter less headroom problems later on, it sounds condescening maybe to someone with your experience but it's damn important!

Be completely systematic. Kill off everything below 30hz. Hi-pass every track, every group, and again on the master channel, make sure there is no possibility for low frequency artifacts to be introduced anywhere.
Use good quality filters with steep curves to do it.

Make full use of the 30hz to 120hz range for bass, that's like 2½ octaves so should be plenty. If you've got a deep kick then make sure you carve out room for it, keep deep bass notes short and gate the kick too.
120hz - 350hz is the biggest problem area, if you frequency analyse your individual recorded tracks , you'll see that virtually every instrument has presence in this range, they call it the 'misery range' for this reason
You've got to use panning and eq cuts and more hi-passing to make sure you don't overcrowd this range, otherwise again you'll get disproportionate peaking compared to your RMS levels.
Pan reverb away from the source sounds, use stereo widening to give yourself extra room in the stereo field where you need it, and one other thing I've been blissfully ignoring all this time - Mix automation. Automate, automate. Panning, levels, effect sends, the works, give yourself loads more room in the mix to play with.
I've only just started to observe these basic rules, and they help, A LOT!

I read that when you mix, it should take about 4 hours to get a demo quality mix, and yes you should start from scratch. It should take about another 1½ to 2 days to get a polished mix, including all the automation.
If your mix is rounded enough, then you could put an exciter and a decent limiter on the master and crank it up a little bit, then have it go through a good quality loudness maximiser as well, but personally I would leave it at about -3dB peak and send it off to a mastering engineer when I have a track I think deserves the treatment, I don't have an acoustically sound enough room to do it myself anyway.

Also, keep everything at 32bit, dither once only, on the final master bounce. This has more of an impact on the mix detail you can achieve than anything to do with levels though, as even at 32bit an over on the master outs is still an over, even if it's impossible to get overs on track channels.

[knightime]

See, now that's why I love this forum more than any other - you guys have such invaluable knowledge I think I'm gonna write a track, just for you!

You know yesterday I said I had '20 years' experience? I take that back completely - I might have 20 years in the business, but when it comes to mastering - completely different ball game!!!

I think I've been approaching the whole 'mixing/mastering' from completely the wrong angle. It's beginning to dawn on me (especially after Delphine and Ten's comments), that it would be a very good idea to treat ech stage separately (compose, mix, master), although I wholeheartedly agree with Danny regarding the use of compression/limiting on the master bus for comparison.

I think I've always had an issue regarding the output level I set on my audio interface. Because my commercial references are so loud I've always found it a pain in the a$$ to keep riding the volume knob on the audio interface to accomodate my 'quiet' compositions in comparion to my 'loud' commercial references. For that resaon I started applying the mastering plugins across my master output bus. Like I said though, I think I've been trying to do too much in one go, so I think I'm gonna change my approach.

Just out of interest, do you guys experience something similar, or do you leave 'comparisons' till way later on in the mixing process, therefore negating the need to keep 'riding' your audio interfaces volume knob?

On a slightly (but maybe connected?) topic, I have a quick question regarding audio input levels and normalising. Although I have set up my audio interfaces analogue inputs to their highest setting, I'm finding that the signal comming into my Motu 828MKII borders on the inaudible; you can barely see the waveform. This is from both the 8 analogue inputs and also the mic/line direct front inputs. I'm finding myself normalising every piece of audio comming in, and on some clips this is resulting in an increase in background noise level too. Something tells me this isn't normal - surely I should be hearing more than a faint whisper when recording? I'm thinking that this may have something to do with my 'commercial comparison' issue and the audio interface/main monitor level settings (hope you're still with me!). At the moment I have the output settings on my monitors to 50%. I usually ride the volume on the Motu anywhere from -72db to -26db (anything higher than -20db and my ears start to bleed). How do you guys balance your monitor/audio interface volumes? Do you think I should start of on 0db on the audio interface then adjust my monitors? Have I been working backwards all along...?

[Delphine]

Originally Posted by knightime

Just out of interest, do you guys experience something similar, or do you leave 'comparisons' till way later on in the mixing process, therefore negating the need to keep 'riding' your audio interfaces volume knob?

I guess that's down to knowing what to expect at the end of the mix. Nothing wrong with getting an idea of how everything sounds 'pumped up' but you might be one of those people who get spurred on when the music is more 'in-yer-face', I think it's all placebo and whilst I like things to actually sound good as I am writing, you can get too bogged down on those kind of details too early on and get sidetracked into mixing before you even have a track laid out. It's hard to resist but the actual procedure is less of an immediate desire when you have gone through a track writing and mixing process a fair few times and have come to know what gives you the results you require at the end of the whole thing.

One thing I would be cautious of, is using commercially mastered material as a reference so early on before a proper mixdown has even begun. You don't say in detail where you are in this track, whether you have mixed it already or are simply applying some sort of mastering suite over an unmixed track as you write for 'psychological reasons'. I'd have to go with the easy route and say turn the monitors up if things don't feel loud enough. One thing that can actually ruin things later on, is having all your levels too high thinking you need to be close to 0db on everything for maximum volume output. It's actually quite the opposite.

A digital summing of a bunch of audio files peaking at 0db could arguably sound rather flat and thin, if you are recording software there's absolutely no reason whatsoever to be maxing them out. All you would be doing later on, is making gain changes anyway. Let the parts breathe, you have so much more mixing options available when you have lots of headroom to play with. Don't think that 'quiet' parts compromises your desire for a 'loud mix' at the end of the line, they absolutely don't. Pretty much all my mix faders on tracks are set at unity as the individual tracks only hit like yellow or even green, there's just no purpose in getting near the red on anything inside a digital system. Prior to rendering, you can adjust your master fader to get a reading of about -4db or thereabouts, I find that amount sufficient to leave the necessary headroom which I need when I do self mastering jobbies and obviously you are safe in the knowledge that all parts are 'breathing' and not maxed out. The result is a much better mix which has depth and clarity.

Ultimately, the dynamic range of the summed parts is going to have an impact on how 'loud' the track is after you do your mastering on it and this begins before the summing of the parts with the dynamic range of the individual parts within the composition. Unless you are making stuff which is extremely in your face, a compromise needs to be met between the natural dynamic range and the amount of compression you should apply. This obviously various on the parts in question and for the most part, unless compression is used an effect, you only need little amounts to ensure dynamics are kept in check and aren't compromising the actual sound. Also, it's sometimes better to use volume envelopes if you need to iron out the occasional peaking transient. If the meat of a signal was averaging quite low but you get an odd shooting peak, you'd think of adding a compressor to catch it, but a simple volume envelope in the velocity editor is probably better as the compression required may affect the actual sound at that particular point. Anyway, that's neither here nor there just something that's worth mentioning.

But as TFS and myself were saying, apart from compression, make use of HP filters for mixing purposes as well. When I first used them and understood their importance, I wondered how the hell I worked without them in time gone by. HP filter everything, even your bass and kick. How much would depend on what you are writing I guess but you can rest assured it will go a huge step in you getting the loudness you require later on. Inaudible low frequencies eat headroom, that's a fact.

You can HP filter a whole mix at 30hz with a 12db per octave slope and whilst you hear no difference, you can see the gain on the meter drop by some margin sometimes. You just don't need that sort of rumble. Now in a mix with parts that haven't been HP filtered, each part on its own might not sound like it has any issues when you solo them, esp non low frequency sounds but it's the summing of them all and the cumulative effect where it comes into play. A multiplication of small amounts of non-needed rumble all adds up and can cause havoc esp when trying to get your mix loud and clear.

So HP your parts, compress them within sensible limits to even out your dynamics and check your efx returns too for anything untoward. If you adopt this approach and are using good source sounds and samples and are keeping your track levels well away from the red, there's absolutely no reason whatsoever for your track not being able to get as 'loud' as the commercial tracks you are referencing with. With a good base mix hitting at around -4db, you can slap on around 6db of input gain on a 'transparent' limiter these days without it even sounding remotely squashed. With a lo and hi cut, you'd be getting more and getting even lower RMS levels too if that's your bag.

Not sure about your recording issues, that doesn't sound right at all though. Have you checked your cables and the entire signal paths?

[knightime]

Hey Delphine, thanks for that reply. Between you, TFS and Danny I've received some of the soundest advice that's ever been offered.

What you're saying, of course, is so true - I have gotten completely sidetracked of late trying to get instruments sounding 'good' from the outset, and the trade off has been that 90% of my most recent compostions haven't even really got off the ground as I've been trying too hard to get the 'perfect' sound from the outset. Compose, mix, master, HP/filter as often as I can, let the tracks breath. I'll take everything on board, starting immediately.

Audio issues I'll be investigating tonight - as far as I'm aware there are no cabling issues as I'm getting the same result from the direct inputs as well as the analogue - so I'm gonna do some experimenting with my monitors/audio interface/DAW software to see where the problem lies.

Once again thank you for some well thought out and sound advice - it has been most appreciated!