Getting your loudness?

Ice Man

Member
Sep 18, 2006
272
0
16
West Palm Beach, FL
I'm sure it's been asked a few times before, so sorry for beating a dead horse, but I'm curious when during the mixing do you guys go for loudness? When people post screen shots of their mastering chain their L2s/L3s/other master limiters have thresholds between -1.5 and -6.0. In most cases to get a perceived loudness to match that, I'm into the -10.0+ area. So, what 's up, guys? Do you guys limit your individual tracks to get that loudness so your mastering limiter does less work, or what? Any input would help. Thanks!

Regards,
Daniel
 
I don't like limiting. In fact, I fear that if I were to limit my precious dirt guitar tracks they would run away crying about not being appreciated and I'd have to hint at marriage to get them to come back... I'm not fucking going through with it, though, that's just out of the question. They knew what they were in for when they started, end of discussion. Sometimes it's just best to leave things without that limiter on and just see what else you can do in terms of EQ, placement, additional tracking, et cetera.

Anyway, one thing that you may be missing is that each little part of your song needs its own little hole. First thing to try would be switching straight to fucking mono. I'm wagering you've a lot of mud on your hands and most things aren't popping out - this must be fixed, or your mission will surely fail. Tinker with anything you can (but not pan, as of course you don't have that) like EQ (to figure out what ranges are cluttered and what ranges are naked) or mild compression (if something is just popping all over the place in volume and you've no way to make it present apart from a bit of smooshie) so you can give each important thing its own very special happy place in the mix (hands above the bedsheets, though, not that special and happy just yet) and then move back out to stereo to see if you've completely butchered everything and need to scratch and do again. After this, look at the statistics for the whole thing and see what your average and peak volumes are - the trick is to have some difference between them so you have some dynamics (those are hard to get back in with EQ, PM jp22 for details) but have the peak pretty close to 0.0 so you're getting the most out of the volume you already have. If I recall correctly someone posted info from a couple of tracks off of Puritanical Misanthropic Euphoria and it was pretty much riding the top of the level the whole way through - that album makes me feel icky in my tummy wummy so I wouldn't quite go that far, but if that's the kind of 'loudness' you want then study that album.

After you have your total level as high as you can get without clipping, just mildly compress individual things with the highest peaks and see if you can tame them without killing dynamics completely. If you're getting better, try throwing mild compression over the whole mix to see if maybe that's better (of course, try dropping the compression on the highest peaks that you just added because compressing too many things just makes your mix sound like your compressor is jumping up and down on its bed and won't stop no matter how much you yell at it) - it could be that the stuff you've put in there is just too far all over the place to get that same perceived average loudness. Find whatever works best for you out of those two, then bypass the compression you just added and go on to the mastering limiter to see if your best results are better or worse than the stuff you did before. You don't want too many compressors most of the time because they don't always play nicely, and you don't want too much limiting because it's usually better to have that dynamic feel than that extra bit of loudness - which most people will never even use because of hearing damage risk and the inherent amplification capabilities of the vast majority of audio setups - but in the rare case that you do (like if you happen to be Dimmu Borgir, and Satan gleefully exchanges your soul for transients) just ignore screenshots because everyone here should also know that it's a no-no to just use a template without at least tweaking, tinkering, completely throwing out of the window, recovering from lawn and cleaning off, replacing, tinkering some more, and randomly bypassing with eyes closed the said template. You'll also want to make two versions of your mix, one with limiting and one without, and adjust the level of the two so that they sound about equal in volume - you'll of course hear louder as better just out of habit, so you want to get rid of that, and when that extra volume goes away you might just figure out why a lot of people just don't use L2.

Try that, if any of it makes sense, and post clips so we can figure out what you're talking about better.

Jeff
 
this question has been bugging me for a while, might as well ask in this thread..

do you guys ever compress rhythm guitars? i always figure the amp distortion already compresses it a lot, and you want as much dynamics as possible in the rhythm guitars so i usually dont touch them with a compressor, but i see a lot of screenshots and stuff from this forum of people that do
 
In most cases there really isn't any need to compress rhythm guitars. It's pretty much squashing the life out of something that's already gone.

I use it for solos occassionally, when certain overtones or notes on the amp want to pierce your ears, also when certain runs and stuff aren't distinguishable from the slower parts of the solo.

Compression on rhythm guitars though is something I only use when necessary (an oddly dynamic rhythm guitar track... or a sludgy one, where I use the attack of the compressor to bring out the pick and individual notes).
 
I cant touch the breakdown J just gave, but he's dead right. everything has to have it's place and the key is repeated gentle (or not so gentle in some cases) compression that add up to the level you are looking for.

shave a db here a db there always trying to keep your eye on letting the song breathe.

my kick drum may be compressed 3 times before it hits the two buss, individually, then in a drum group then the master. same with just about everything else.

except heavy guitars, I only ever use multiband on them and thts only if I have a "woof" problem Im trying to control.

you also have to keep your room and monitors in mind. if you have a huge spike at 50hz and 10k you cant hear, it will be eating all your headroom and killing any limiter you strap over it.
 
I only compress amps with a multiband in the low end (essential with Fredman micing, important with single and even more so with less than great volume and room conditions), when there's a clean or mild dirt sound, or when there's a high part that needs to overcome thundering walls of chunk, and even then it's only a few decibels (with exception of clean, which depends on number of tracks and a few other things) and mostly EQ that gets done to get my guitar parts noticed.

Jeff
 
There's a very interesting discussion going on in the business right now, about the loudness war. There's tons of material to read, and a good place to start would be the Gearslutz forum, where mastering guru Bob Katz is a regular.

However, I learned nothing from reading just about everything I could find about the loudness war. It just made me more determined to make my tracks as loud as possible. I'm probably stupid and deaf, but I enjoy the sound of a really squashed track, considering it is NOT done with L2/L3 which really kills the attack of the drums. A fine 'ol clipper is a better alternative any day of the week.

To illustrate my point, here's a picture of a pre production demo that I mastered in Wavelab. If you zoom in more you can actually see some wavesforms; I haven't found a way to get rid of them yet. ;)

flat.jpg


If you wanna hear the track, there's info in this thread: http://www.ultimatemetal.com/forum/andy-sneap/277307-ok-so-i-finally-started-new-band.html
 
damn storyteller! that looks insane! butif it works, then it works ofcourse...

i often use subgroupcompression on drums and guitars, not to get more loudness but to make the mix sound more powerfull and punchy. it creates the illusion that the instruments are "fighting" to come through more.
as for mastering i usually use 4 compression stages: fisrst multiband, then a little L2, then a little full band compression to get some punch and then loudness maximizer to give the final boost
 
what do you use like a ummm spectrum analyzer to check were spikes are??

also do you do each track in analyzer and then adjust with eq or maybe miltiband comp?

also anyone wanna give a run down of were they try and fit each instrument in mixes usually? I know it can vary from recording to recording but like some people pull 80hz out of the bass to let kick sit there or whatever.

Any other normal cuts to like let bass sit some were and same with guitar and what do you usually cut them from like what instrument. Maybe just some starting points or something for me to work with in my mixes.....
 
Becoming reliant on spectrum analyzers is one of the worst things you can do for your career as an AE. The moment you start relying on meters, is when you stop trusting your ears and go by what the software is showing you.

The entire thing is about listening. Try to get as far as you can without metering or spectrum analyzers. In mastering, I can understand, but you really shouldn't be doing that at home anyway.
 
+1....the only time I would use a spectrum analysis tool is if I was at my wits end trying to fix a problem at a particular frequency I can't aurally identify (after numerous attempts). If it sounds right, it's right.
 
Thanks for all the replies, guys! I'll scrap the limiter and go for subtle compression through stages. My usual problem is that I get the mix sounding good but lose that coherency and attack and bury certain things like a kick drum or snare because the limiter is chopping off peaks.

JBroll, special thanks to you for taking such time to post all that. Not only is it full of information, but the little jibes of humor made it a good read!

Thanks to all,
Daniel
 
As far as the frequency analyzer, you'll often have something popping out at you (whether you're producing your own stuff or someone else's album has some odd tone that's driving you nuts) and if you can't quite figure out what it is with an EQ notch sweep then the frequency analyzer can come in handy. When you've found that, A/B a compressor and an EQ to see what sounds more natural while killing the bad stuff.

Broken, you may wind up with boxy-sounding shit when you cut around one frequency just to make room for another - supposing your kick sits around 80, its first octave overtone will be at 160 and you may have noticed that if you want more 'thud' without sacrificing beater snap you may get a cardboard kick sound if you just increase 80 without cutting out some of its overtones (I picked this up from that Slipperman thread, for anyone who hasn't read it that's a must read) to compensate. It also may occur that your kick appears boxy in the mix because it has more room around 80 and thus appears louder; this is when I'd try compensating for that by twiddling EQ knobs on both kick and bass around 80 and 160, just to see what fits and doesn't sound like a basketball bouncing.

Storyteller, that looks... interesting. For that little bit of waveform in the beginning, you could try any fade-in plugin that allows above unity gain, and set it so that its gain increase is inversely proportional to your level (that's going to be a pencil thing, you're not getting a function like that - what we're talking about is mathematically the inverse of a derivative of a combination of several orders of sinusoidal functions, more than likely at least a little remnant of the chaotic little pixies called transients, operating noise... no computer will willingly take on a differential equation of that beastliness) but I wouldn't kill that as it'll just sound odd.

Joeymusicguy, I get the same dizzy/queasy feeling out of thumps around 50-90 hertz - read: meat of kick drums, love happy joy joy - so I just have to break and return when working with those. I hate that.

Ice Man, enjoy, I get bored and that comes out. You may have noticed from high gain amp designs that better sounding gear typically has multiple stages for all of that smashy smashy - the way this is done we reduce artifacts and still get the same volume. Have fun.

Jeff
 
IMO one disservice that the internet is responsible for is making people think that compression is bad or to be used sparingly. I can't even count the number of times I've read on gearslutz "I don't compress drums at all."
That's great for that guy but every time I've worked w/ and watched an experienced (rock) mix engineer in action they have been anything but shy about compressing and eqing things.

Again, just my experience, but the result are mixes that change a lot less during mastering b/c the individual tracks have been leveled out more.
I would be curious to hear what the heavy hitters here have to say.
 
Loudness. Hmmm...something I've been thinking a lot about lately.

I happen to be in a bit of a pickle here. I'm mainly just a hobbyist who mixes and
produces my own shit. I've done some commercial work, soundtracks for independent films and such. When I mix my own junk I leave a TON of dynamics in them. That's just my style.

Recently I was contacted by a film editor friend of mine who is putting together a DVD for a UFC-fighting type thing, and he needs some metal-type royalty-free tunes. Well, here I come. I pounded out a 1:30 intro and a quick 3:00 ditty.

I'm running into trouble mixing this stuff. Normally, the song would have a lot of room to breathe, but I feel that I need to slam the shit out of tracks because that's what my editor friend, and the meatheads watching this DVD will expect.

But damn, I am not good at getting my music to ~-10dB RMS without using heavy, heavy limiting. This a completely different style mixing than I'm used to.

I'll limit everything, and I mean everything. Rhythm guitars, bass, drums, drum buss, 2-buss...etc. It sounds like shit, but I can't get the volume up any other way.


Goddamnit.
 
Well, I have that and use it for some things, but does it give you your RMS and peak levels numerically? Or, is it just from reading the graph? I like to use it for EQ purposes and haven't spent much time with it otherwise.
 
*noob alert*

What do you guys use to analyze your average and peak levels?

I think I'd understand this better for myself if I could analyze my own stuff, but I don't know what to use.

Here is a very good tool to analyze your material and it is free.
http://www.voxengo.com/product/SPAN/


I use the nuendo multiband to master my mixes. The soft clip function gives me the loudness I need most of the time.
L1 kills all the transients if you squash it very hard.