"mastering" question

[thekeepsake]

Hi,

This has probably been asked (though I did a search for it and didn't come up with the exact question, so)...

The bands I record want their mixes loud, and I've learned all too well about destructive over-limiting in attempts to please clients. So my question is simple:

How are some of you home-recording guys doing rough mastering to your mixes? I'm doing this in Wavelab, so a plugin list or signal path would be great.

What I've been doing is some EQ with Waves LinEQ, multiband compression with Waves C4 and then limiting with L2. I want to find a less instrusive method.

Obviously the least intrusive method is to tell the band to take it to a mastering engineer, but they're on budgets (mostly I do demo recordings, very few actual releases at this point) and "the customer is always right"*

*Not always.

Any advice would be much appreciated. I'm looking for transparency over loudness.

[ericblam]

Often times, i'm in a similar situation. The band has no money for mastering, so i have to do it.

This is from the www.waves.com:

Mastering loud mixes without the squashed flat tops - Linear Phase Multiband and L2

Here we'll use Linear Multiband and L2 to create a radio-friendly mix that doesn't look like a close-cropped, GI-style wave shape.
Place LinMB and L2 in the master chain - LinMB first, followed by L2. Load the "Adaptive Multi Electro Mastering" preset. Under the "Master" heading, you'll see control boxes with arrows. These controls globally affect all of the frequency bands. Grab the THRSH (Threshold) control and drag downwards. This lowers the thresholds across all frequency bands. When the animated DynamicLine begins to dance like a snake, you are beginning to apply compression. Stop when the DynamicLine is moving through about half of the total Range represented by the blue rectangle. Make note of any frequency bands that are really showing a lot of action. This indicates a lot of activity in the band, something which may be normal for the mix or which could indicate a frequency balance problem. Also notice if the lowest frequencies are very active. Either of these conditions may be normal, or they might suggest the mix balance could use some additional work.

Assuming all looks and sounds well, you can proceed to the L2. Grab the left Threshold slider, and drag it downward until you begin to see very slight activity on the ATTEN (attenuation) meter. Leave the slider at that position. Adjust the Out Ceiling slider to a ceiling of -0.1 dB to extra-guard overshoots (there are still playback systems around that display signal clipping when the signal is at 0.0dB). At this point, you should be hearing a very full mix, where most of the peak information is still intact.

In general, I've noticed that at alot of the mastering sessions i've gone to, the mastering engineer will use small amounts of compression at several different points in the signal chain. Then do the final limiting. Try doing the same thing in waveburner. Use one or 2 compressors doing 1 or 2 db of gain reduction. Then strap on the L2.

eric@blam

[joel hamilton]

What eric said. Stack up a few with drastically differing attack and release curves, so you dont wind up with a "pumpy" sound at all, because every device is reacting differently. That is a pretty key thing.

[thekeepsake]

Thanks a bunch, guys. So you wouldn't get in and do anything real serious with EQ? Sometimes I'd get frustrated from muddying up the mix and boost certain freq. ranges around 4k or 6.5k a heck of a lot.

I guess a lot of the problem is you become committed to how the premastered mix sounds and are trying to get it to sound "as good" meaning "the same but more loudness." I understand completely why it isn't preferable that an engineer masters a project they've mixed.

[joel hamilton]

Using the waves gold or platinum bundle:

Use a PAZ analyzer to confirm your room modes (meaning, if you have a big spike at 650 and a vally at 3 K, you were probably correcting for monitoring or your room.

Make small EQ moves based on that info. I have found this to be helpful when I have to "master" something that I mixed in the same room.

I do for sure use C4, to the point where it is shaping the sound. Getting the high end from multiband compression seems to be nicer in the digital world than getting it from eq.

I will also De-ess across the two track if I have to, to keep the cymbals in check, and esses I guess sometimes. That is the only way to get some high end compression that approaches tape like results IMO, is to use a de-esser to lick the top off of a big transient, just like tape.

[thekeepsake]

Using the waves gold or platinum bundle:


I will also De-ess across the two track if I have to, to keep the cymbals in check, and esses I guess sometimes. That is the only way to get some high end compression that approaches tape like results IMO, is to use a de-esser to lick the top off of a big transient, just like tape.

One last question then: when you de-ess for this purpose, is there a particular range of frequency you look to grab or is this kind of a what-sounds-best deal?

[mjau]

I don't know if this is possible...but maybe there is someone else in your area in a similar situation (i.e., having to engineer and finalize a project), and if so, maybe you two can sit in on each other's finalization. I've got friends who send me demo stuff to finalize, and the only reason they do it is because I'm a new set of ears. None of us have perfect (or even good) rooms or monitors, but it does help to have different opinions.
Just a thought...

[cgarges]

So you wouldn't get in and do anything real serious with EQ? Sometimes I'd get frustrated from muddying up the mix and boost certain freq. ranges around 4k or 6.5k a heck of a lot.

How about instead of boosting a bunch of stuff, cutting frequencies in the problematic areas? I find my self doing WAY more of this than boosting. Sometimes, when things sound really great and big and full when soloed, all that hugeness adds up to cause muddiness in the mix. Start thinning things out. How much and where is program-dependent, but the more you do it, the faster you'll get a grip on where you need to take stuff out. Does it sound great in your mix environment, but not translate? Then send at least ONE project to a decent mastering engineer and then ask where the problem stuff is occurring. (If you want a recommendation for a good, inexpensive mastering engineer, PM me.) Then you can apply that knowledge to your future mixing and mastering experiences. Not that that information is going to be a cure-all, but at least it'll give you an idea of where to start.

Also, try to attend a mastering session at some point. It will probably be very enlightening. Most of the mastering guys I've ever worked with are doing EQ tweaks of like a dB and a half! Not much, but when it's in the right place, MAN does it make a difference!

Chris Garges
Charlotte, NC

[thekeepsake]

thanks everyone for the advice. turns out one of the things i'm working on is going to be mastered at west westside, so i'm going to try and stare at alan while he works.

i've heard of a lot of people that master each other's work and i'm also going to pursue that route, but with the tips on this thread so far i'm feeling a lot better about doing some of it myself.

[wwsm]

I'm fairly new to this "posting" thing, but I figured this is a good place to add a few thoughts. Remember that mastering is not a procedure... It's a way of listening and working. It has a lot to do with perspective. It's fairly easy to "flatten out" the EQ response of a mix. Any good multi-band compressor can do that. Same goes for volume. Everyone has a L2. But finding and embellishing the "feel" of a record can be way more important and more effective. Lots of gear (digital especially) can solve problems, but do nothing for the feel of the mix.

That all being said, I feel there really is no "one" way to suggest a signal chain. Certain things may be obvious as stated earlier. Limiter is best last... EQ-ing into a compressor tends to help smooth out moments where the EQ could get "peaky" (also helps prevent clipping). Multiple compression can be effective, but "less is more". One compressor that does it right can blow away two that are just trying. Same with EQ (as cgarges said). A little in the right spot always sounds better. Add things in slowly and really listen for what they're doing (or not doing) to your mix. If you're doing a lot in any one place, normally there's a better way to do it. Try a different approach.

A couple of suggestions I could add to some of the already good advice from above... First. Don't try to master right after you mix. Get away from it for a while, and maybe even try to switch to a different set of speakers. Second. Don't play it too loud (until you're done). Volume plays havoc on how you hear and use compression (especially if your trying to get maximum volume... but that may be another thread entirely). Listen back at several volume levels. Third. Find one or two of the most important instruments on the CD and try to keep a relative volume for these instruments throughout the whole sequence. And forth. I agree with mjau... get opinions from other engineers on what you're doing. Sharing comments on each others mixes and mastering is a great way to learn.

Remember this too... CD's that sound loud (and great) usually have a lot going on before they get to mastering. Trying to make the CD loud at the last step rarely works well. If you're working in digital, leave room to allow your plug-ins to work. If you have to bring your digital volume down to prevent your EQ from clipping, you're just adding another stage of DSP. And chances are the more DSP you use, the more your final master will suffer... Less is more.

Alan Douches

[Disasteradio]

I was a total L2 masochist for the last year and a half until I read up about compressing mixes and I realised that the more work you put in compressing individual or groups of tracks, the more you can get out of compressing your mix. Now I'm only using maybe -1 > -2dB thresholds and my mixes sound way loud, but still have space in there. I guess what I'm getting at is the earlier you put the work in getting stuff balanced and loud, the more you can get out of stuff at the mastering stage.

ahem, how obvious.

aside from limiting, I barely touch EQ at mastering stage, my tracks are pretty much samey frequency-wise (writing 90% with just DSP + sampling in Buzz) - I don't trust EQ with destructiveness.. maybe when I can get 96K mixes out of the thing then I'll start tweaking my EQs. blah.

ooh ooh! + Sonic Foundry CD architect is a must-have for checking the "steps" between hotness / brightness between tracks on a CD if you're sequencing the tracks up.

[jajjguy]

Alan, thank you for your very informative post, and welcome to TapeOp. I hope you keep posting, as your knowledge and experience will be much appreciated around here.

[EasyGo]

Yeah, thanks for the great post Alan! Welcome indeed.

I really enjoyed reading the post, especially the bits about making subtle adjustments and leaving headroom to allow the plug-ins to work.

Cheers

[thekeepsake]

I third that thanks, Alan. I have a new perspective on it now. I should add that for the mix in question, I was being rushed by the band and couldn't leave any time in between mixing and mastering. In the future, I realize that I should insist on waiting as if the sound really matters, we should all be a little more patient.

Thanks to everyone who responded for your helpful insight. I feel much more confident now.

[stillafool]

I don't do mastering, but I have a friend who worked at a computer music store who did a litttle. He played me a mix he did that sounded excellent, and I asked how he did it. He said he used, along with the waves ren comp, waves ren eq, and the L1, some program for the mac that took an eq image of one file and applied it to another (I can't remember the name of the program). He basically just took an eq image of a commercial recording and applied it to his own. He said the program really made his mix come alive. I asked him if a program like that was out for the pc, and he said no (that was a few years ago).