Let's all sit in a circle and talk about mastering

elektro80 · Posted: Thu Jan 29, 2004 11:31 pm Post subject:

Ahhh.. ok.. I mostly use plugins.. personally I am aiming for stereo concert sound and you all know which synths and gear I love. One thing I do not aim for is that chrunced concert sound.. where details disappear.

There is no point in driving up against that 0dB digital limit where no real saturation happens.. you just get clipping. Keep a nice headroom and use some eq before you push the signal into a limiter.. and then a compressor. One trick is to use the combination of limiter and compressor to create that "loud" saturated top peak where the PA loudspeakers almost starts to burn.. but you can adjust the attack/threshold/slew in both units so you get different effects. try this.. on some loud synth solos.. and you can also add some distortion from plugs like supercamelphat and others. But do listen to the real thing first.. a synth thru a vintage PA. I love bass synth voices thry those old Ampeg bass amps.. NICE! And you can also add that other trick I mentioned earlier in this thread.. by running a 100% reverb feed into an enevelope which complenets the main signal. You can even fade out the reverb earlier than than the main signal so you do not get that lone reverb tail.. and you can run the reverb feed into some slight distortion and then into a stepfilter array. ( This is really a stepsequencer which has does not control CV pitch.. but filter values in a VCF. Howard´s analog sequencer design is perfect for this. For every step the filter has a new position. Throw in a limiter like the Waves L2 in order to cram the signal dips and the signal peaks together .. and if the signal is "steppy" .. it has pauses.. not completely legato.. you can also add a compressor after this and tweak the attack/slew/release on both units and go hunting for some sweet spots in the signal. Very often you can find really hot stuff in this combination. If you wanna go really wild.. use a stereo path. Send the synth signal first into a stereo reverb... which spreads the signal out.. then into pair of VCFs which are stepped by the sequencer ( you can use the same filter CVs for both VCFs, the stereo reverb will give you enough "stereo space" in order to make the bass shake .. in stereo.. Actually.. a stereo plate reverb will do fine too ) .. then into stereo limiters/compressors. The bass synth sequence at the start of my song "A Scanner Darkly" is done this way. I have been doing this since way way back.. and you can do this live too. this adds far more life to a part than phasing.. Phasing.. hmm.. It is great fun to use an old style sequence wher you have right and let channel echoes ( Roland Space Echo .. or modified Tandberg reel to reel machines ) .. keep the clean signal panned to mid.. then send the sidechhain to the phaser and then into the echo unit.. with the delays set to pan out left and right.. I prefer using reel to reel decks. Some very soft tube distortion on top of this and you have a killer sound. You can also setup many chains of this type for each instrument .. for more ambient effects... like what I did in "utost Savagery" you then use volume pedals for playing the various feeds in the mix. You can easily create patches which are VERY playable.
All this can be done digitally. bu the trick is adding "analog" signal behaviour... some slight tubish or solid state distortion here and there.. and of course compression. Loudspeakers does compression. You can aim for that type of compression instead of the usual "pop music" type of compression. Instead of just loud you create a different feel which is much more subtle and far more effective. if you cannot quite understand how to do this, you can try miking the various parts off a PA. When I think of it.. the big TDl transmission line loudpeakers.. old stereophile thingies.. those works fine too. Some of the boomy bass parts in "Blinded By Memory" comes from runing the signal thru some TDLs.

Btw.. forgot.. the stepfiltered reverb thingie.. do the signal mix at 40 - 50 % clean/reverb and adjust the reverb time .. you will know when you have it right!

Cyxeris · Joined: Oct 30, 2003 Posts: 1125 Location: Louisville, KY

See, arent these threads delicious? Wink

It just occured to me that I literally do not use any distortion whatsoever. Perhaps I'll start experimenting with that later on. 3rd album perhaps. The second is shaping up to not bennefit from that sort of sound. It's looking highly symphonic and thematic. One of these days I'll be in the mood for more experimental stuff.

I look at digital audio like a movie screen. There are rigidly defined parameters wherein the content itself can reside. You can have any content you like, but it has to fall within the limits of the screen. Exceed these limits and you have no picture there, just a truncated version of the original image. Analog would be more like the equipment used to film the movie. You can play with different lenses and filters and lighting and whatever, and come up with some killer imagery. And you cant really compare, say, a widescreen plasma display and a 6-figure movie camera. A similar distinction, I believe, exists between digital and analog recording.

Digital is the format, analog is the technique. I suppose. Or something like that.

Cyx

mosc · Posted: Fri Jan 30, 2004 8:45 am Post subject:

Wow, lots of great information here in this thread.

I realize, after reading this, that I don't do mastering at all. I mix. I look at tape and CDs as just media to play my mixes. Of course, I just compose electronic music. I might view things differently if I was a recording engineer or producer.

I want my music to sound good on great speakers, not necessarilly on home budget systems or in cars. It seems to me if you are targeting budget stereo, then you have to make it sound less good on great systems, either by adding or taking away something you otherwise wouldn't.

Someone mentioned radio stations running the music through all kinds of processing chains. Actually, that makes sense. If you figure the music is going to be played in a car, process for the car.

Nobody mentioned mastering for mp3. MP3 format makes a mess of music. I surprised how we tolerate it. I know, it's the only way most of us have of getting our music heard. I've seen, on other lists, people rant on about how they wouldn't use anything less than 24 bit at 96 KHz for their recordings, and then they release MP3s.

So, I ask. What is the purpose of mastering? Why is it a separate process?

paul e. · Posted: Fri Jan 30, 2004 1:17 pm Post subject:

Cyxeris · Joined: Oct 30, 2003 Posts: 1125 Location: Louisville, KY

Cyxeris · Joined: Oct 30, 2003 Posts: 1125 Location: Louisville, KY

One way to think of mastering is to think of a circle. Let's say that the optimal final product is a 10 CM diameter circle. Your mixdown may translate to a 9 cm x 7.5 cm oval. Mastering would be the process of adjusting your frequency content and levels in order to fill the circle. Normalizing would be like upping the oval to 10 CM x 8.2 CM. It's the act of efficiently utilizing your available headroom and balancing out your frequency content, and also making these aspects of all tracks on the album as uniform as possible. That is why the tracks on a CD may all sound very similar holistically, while all CDs sound different from one another.

Of course, the best option is always to get the mix to as close to the circle as possible, but ultimately it's far more efficient and productive to focus on getting your mix itself well balanced and clean, and then make your overall corrections in the mastering stage. Otherwise you'll spend alot of valuable snooze-time tweaking all of your parts up to the perimeter of the circle. If you can get your mix balanced well and everything set well in the mix, then you can make your sweeping changes to that good mix in one go, rather than in 30, over and over and over until everything is both balanced well in the mix AND up to mastered levels. Just tackle one problem at a time and do so intelligently.

Even within the scope of creative endevors, never overlook the relevance and applicability of raw unadulterated laziness. It can go a loooong way.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Cyx

mosc · Posted: Sat Jan 31, 2004 7:49 am Post subject:

Good discussion. I've learned a lot.

So, I guess I do mastering, but I integrate it into the mix process. Since I don't use any softsynths, I have ample prcessing power.

Cyxeris · Joined: Oct 30, 2003 Posts: 1125 Location: Louisville, KY

I try to do as much "mastering" in the mix as possible, but there is always room at the end for some minor overall tweaking and shaping. The less, the better.

I have a vision of a scenereo where I have a song both mixed and roughly mastered before I commit a sequence to audio. What a lovely idea.

Cyx

nothing happens · Posted: Wed Feb 04, 2004 8:32 am Post subject:

This is all way more in depth than I know anything about mastering, etc. But even I can tell a lot of modern pop/rock stuff sounds like jankola from being over-compressed/limited to try to make it as loud as possible. I think you really lose the attack on drum hits that way, the beat gets mushy, as well as wussifying all the dynamic contrasts.

I beg to differ about rock music not having much dynamics, though. Perhaps you need to look for some better rock music. Er, I mean, not all rock music has dynamics, but quite a lot of it does. But then, rock music isn't the most popular thing on a website full of electronicko-heads.

My overly simplistic, highly uninformed idea of mastering is this: after I mix from my 8-track tape or whatever else into the computer in Cool Edit, I do a little EQing if there sems to be a serious need for any, normalize, limit out any big peaks that don't seem to belong there and are going to get in the way, normalize again, give it some relatively nonintrusive compression (usually either 2:1 > -10 or a preset called "Classic SoftKnee" depending on the characteristics of the song in question), normalize again, and that's about it.

I basically shoot for the little meters to spend most of the time flickering around/topping out in the -3-and-up area (what that is for RMS I don't know, how would I find that?). A little more for particularly loud parts/songs, a bit less (say in the -6 neighborhood) for softer parts, maybe even less than that for really soft parts. This fairly bonehead process of mine gets that done and sounds pretty good unless I really screw something up. Volume-wise it competes well with (and is even sometimes a little louder than) most CDs I listen to, so if I put it in my car CD-changer on random, my own stuff doesn't sound out of place. I usually listen to lots of indie label and older post-punk stuff, and they seem to be pretty tasteful about mastering, so I figure if my stuff can hang with them I'm at least getting a decent level. But I'm probably doing it all wrong in the first place...

I just like to keep things simple.

mosc · Posted: Wed Feb 04, 2004 11:02 am Post subject:

Cyxeris · Joined: Oct 30, 2003 Posts: 1125 Location: Louisville, KY

You're going to get some degree of artifacting with such a technique, but if you're working in the 24 (if not 32) bit domain, as well as perhaps some dithering, then these are negligable. One thing to make note of, though, is that each time you normalize, you reduce the amount of headroom available to the next stage in the process.

Cyx

DES · Posted: Wed Feb 04, 2004 2:25 pm Post subject:

Ah to be able to get the true sound of analog tape in the digital world...whoever figures THAT out will be lifted to Saint-hood! Laughing

Unfortunately because of the physics involved with tape - signal saturation and you have pieces of metal becoming magnetized and each piece of metal is magnetized differently - and then the phase differences between these pieces as they go across the tape-head - it is very difficult to emulate in the digital world. What a lot of engineers do is once they have the track(s) basically mastered, they run it out to a 2 track machine and then pick up the signal off the tape machine and bring it back into the digital domain. It works wonders and you only have to buy the tape for a 2 track machine as opposed to 24 track tape which is very expensive. The end results are not the same as doing it all on tape - but really does add a certain warmth and cohesivness that you would not have being totally in the digital domain....

Dave

DES · Posted: Wed Feb 04, 2004 2:32 pm Post subject:

Cyxeris - I don't know which Waves package you have - but try the renaissence EQ's and compressors - I have found them to be quite useful. Another recommendation would be to try EQ's and compressors from different companies on your tracks. Plug-ins do have a certain sound to them which is due to their algorithms...different manufacturers products use different algorithims and as such sound, well, different..this can be used to spice up the mix.

Dave

Cyxeris · Joined: Oct 30, 2003 Posts: 1125 Location: Louisville, KY

mosc · Posted: Wed Feb 04, 2004 4:17 pm Post subject:

elektro80 · Posted: Wed Feb 04, 2004 4:28 pm Post subject:

I agree with Howard. Fact is, if we dig into the tape tech we will find a lot of very entertaining issues which kinda raises the question: how the hell did anyone manage to make great recordings using tape?

OK, it is possible to use tape for signal processing, and that might be the best way to go about it. I have used tape A LOT, and not portastudious, but the real ( hehe.. reel.. get it? ) shit. Fantastic machines, but digital is far better... and digital will become even better. Many of the most significant "tape-effects" can be simulated digitally one way or another.

nothing happens · Posted: Wed Feb 04, 2004 5:06 pm Post subject:

As a matter of fact, I do do all that stuff in 32-bit. Thanks for asking. Then convert to 16 when done, to burn to CD, convert to mp3, whatever.

One reason vocals in rock bands require so much compression to be heard is because of the predominance of big loud distorted rhtyhm guitar hogging up all the midrange. The modern rock music aesthetic is practically built to make the singer and the guitar have to compete for inample space in the mix. And then we wonder why we head down to the local rock club and we can't understand any of the lyrics. (Then again maybe if we could we'd realize they often aren't worth hearing.) I'm just as big an offender as anyone (my guitar amp has a 15" speaker), but lately I've been trying to be more cognizant. Rock musicians seem to have forgotten that old adage about the "space between the notes."

mosc · Posted: Wed Feb 04, 2004 5:15 pm Post subject:

elektro80 · Posted: Wed Feb 04, 2004 5:21 pm Post subject:

Tape for professional audio recording is in essence dead. There will still be some use for tape, but that is for pretty fringy tasks. Tape machines are to expensive, too messy, too huge, consisitent sound quality is so and so, tape itself is too expensive and volatile and then.. the business is now about nonlinear editing. Tape is dead. And I still have a few old reel to reel machines I cherish.. I keep them just because they look good.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

elektro80 · Posted: Wed Feb 04, 2004 5:34 pm Post subject:

I did mention that tape can be used for processing. Of course it can.. it works well too. But .. some of the great myths about tape sounding warm.. the great saturation etc.. are not quite true. Of course, a good machine and a very good tape can be used in order to do this.. to a certain extent... but alot of this is pure mythology created by people who do not understand the issues with the technology.

Better tools would be various breeds of line tube amps, tube compressors, certian breeds of hybrid or solid state devices... like the brand new SSL XLogic Superanalogue Channel.

elektro80 · Posted: Wed Feb 04, 2004 5:44 pm Post subject:

I am actually waiting for a new circuit design to show up soon. It would be a digital or analog device which has a limiter, then EQ, then a compressor and last another limter. The very special thing here is that the EQ varies its settings depending on volume, and it must also have a feedback option for using compression factor as a CV for the filters.

I can elaborate on this.. anyway.. if the various processing sections have sensible designs.. this unit can be used for some very interesting simulations and processing. I guess digital attempts as plugins will show up first, then TC or someone will make this as a hybrid diftal/analog device.

DES · Posted: Wed Feb 04, 2004 7:39 pm Post subject:

Michael Chocholak · Posted: Wed Feb 04, 2004 8:56 pm Post subject:

mosc · Posted: Wed Feb 04, 2004 9:50 pm Post subject: