Audio Engineering

[int]

I know only a few of us are actually interested in working a sound board, but I'd figure I'd start here instead of an audio forum for once. I've poured over the forums a lot recently, but I think I need a fresh outlook. So here it goes:




I'm having a hard time with drum and guitar tracking for a power-pop/punk/metal/acoustic project. Yea. Lots of variety.

Starting with the kick: I've got a weird phasey sound happening with the kick (D112) about halfway in. When combined with the other mics and instruments the "phasiness" goes away (isn't heard anymore). It sounds ok, but I'm wondering if this is normal? Also, (and this will be a theme) I'm having a hard time getting the right compression on it. Either I get a consistent volume that takes out the meat of the kick, or I find a setting that sounds good but the consistent kick that I need isn't there.

Snare sounds ok, needs more isolation though. I'm using more gain on the quieter stuff, but could use some compression tips for that. The power-pop/punk stuff is consistent and great.

Overheads are killing me. How the hell do these get right? I've got a pair of AT 4041's that sound good - but I can't seem to find the right balance with them. I've tried stereo micing these with no luck (kcan't find a sweet spot with a true stereo balance, so I seperated them and close micing the cymbals - with a little luck. I'm afraid of compression here, so I've avoided it.

I've put up a pair of omnis in various places for the whole kit and found a few sweet spots. Mostly phase problems come around though.

A PZM under the snare sounds awesome. As do the 57s on toms.

With the guitar, I've blended a silver Sennheiser 609 (one of the new ones, on the grill) with my LD condensor (a Shure KSM 32) and found a decent sound with the condenser a few inches off the grill - but it lacks the bottom end meat I'm after (which I can get on the amp itself).

Bass sounds good, as I'm primarily a bass player and can find a good direct tone I like. Vocals take some work in mic choice but I usually get it close. I'm not afraid of compression here either.

Guess I'm just looking for some tips on tracking my least practiced instruments, so any tips or a point in the right direction is appreciated.

[PiccoloNamek]

Also, (and this will be a theme) I'm having a hard time getting the right compression on it. Either I get a consistent volume that takes out the meat of the kick, or I find a setting that sounds good but the consistent kick that I need isn't there.

Don't turn the attack up too high, it will stunt the transient. Setting the amplitude of the threshold to too quiet a level will squish the entire waveform to that level if the ratio is too high. If the output gain is also too high, you will end up with a loud sound with no "oomph". If you set the decay too long, the volume won't come back up quickly enough. What ratio setting are you using? Have you tried overlaying the compressed drum with the original waveform?

And for a personal opinion: try not to use any compression unless you absolutely need it. Go easy on vocals especially, otherwise you end up losing the change in volume causes emotional responses in listeners. You should always try to keep as much dynamic range as possible. Don't clip any peaks, either...

Now, unfortunately, I don't know anything about microphone placement.

[earwax]

For drums, I get pretty good results using a D112 on the kick and two overhead AKG C-1000 condensers set up in a "Y" over the set. Get a step stool and stand over the front of the set and listen while the drummer plays. You'll find the sweet spot. Sometimes I'll add an SM57 to the snare.

For a guitar amp, I usually just stick an SM57 in front of it and look for a sweet spot. (And bless you for getting the sound from the amp rather than software plug-ins. I'm a fan of the old-school. Let the musician's get their sound and then try to capture it.)

I try to use as little compression as possible. Too many recordings have the life squashed out of them these days.

[EL84]

What sort of compressor are you using? Maybe try another compressor for the kick? There's some really wild variation in compressor circuit design out there, and they all sound different. Not that I know much about recording kicks, but when twiddling the knobs isn't working for me, switching devices sometimes does the trick.

Sometimes I only compress the vocal though. For my ears, it always needs some compression, but then I'm recording some crap vocals.

How do you mean the kick starts sounding phasey halfway in? Halfway through the recording time?

I'm also a fan of SM57 on guitar. I like the edge of the cone. For the bottom end meat though, if you could record it, would it work with the rest of the instruments? I have a theory that the guitar tones that blend really well in the final mix don't sound that great by themselves.

[int]

Don't turn the attack up too high, it will stunt the transient. Setting the amplitude of the threshold to too quiet a level will squish the entire waveform to that level if the ratio is too high. If the output gain is also too high, you will end up with a loud sound with no "oomph". If you set the decay too long, the volume won't come back up quickly enough. What ratio setting are you using? Have you tried overlaying the compressed drum with the original waveform?

I've got the attack set pretty fast - I'll try backing it off. I'm using a 4:1 ratio with the threshold around -12, which is only lopping off 3-5db of the kick - which has worked well with other projects. I forgot about adding the uncompressed one in, although I'm trying to take a minimalist approach.


For drums, I get pretty good results using a D112 on the kick and two overhead AKG C-1000 condensers set up in a "Y" over the set. Get a step stool and stand over the front of the set and listen while the drummer plays. You'll find the sweet spot.

That's what I tried - but the hats and crashes were bleeding into both mics, and not giving me enough stereo separation. I'm thinking that's just a coincidence with this drummer. He's consistent, which is good (no compression going into the computer) but he's LOUD.

[int]

What sort of compressor are you using? Maybe try another compressor for the kick? There's some really wild variation in compressor circuit design out there, and they all sound different. Not that I know much about recording kicks, but when twiddling the knobs isn't working for me, switching devices sometimes does the trick.

The one in my mixer (Tascam DM-24) is my go to compressor. But I've tried using Blockfish and my RNC.



Sometimes I only compress the vocal though. For my ears, it always needs some compression, but then I'm recording some crap vocals.

Depends on the song with this group. For the heavier stuff I'm putting some compression on at a 10:1 ratio, fast attack, fast release, -10 threshold. This just lops off the top and gets it's pretty hot to disk.



How do you mean the kick starts sounding phasey halfway in? Halfway through the recording time?
No - I mean with the mic about halfway inside the kick drum. It seems to be the right place for getting the boom and the attack. By itself it sounds strange. When mixed together it's fine, so I'm not that concerned.



I'm also a fan of SM57 on guitar. I like the edge of the cone. For the bottom end meat though, if you could record it, would it work with the rest of the instruments? I have a theory that the guitar tones that blend really well in the final mix don't sound that great by themselves.

I'm noticing that too. I find myself turning up the midrange on the guitar amps and backing off the distortion too. What I'm trying to capture is the bassy palm muted, high distortion stuff. Almost metal-like. There's a big dynamic swing here that I don't seem to be capturing. When that D112 frees up I might try using it. Same with the 57s that are on the drums right now - I'll probably move them over to the guitar amps once the drums are tracked.

Also this project is getting mastered - and I thinking if the whole thing is going to get squashed even more, I should probably use less compression in mixing.

[PiccoloNamek]

I've got the attack set pretty fast - I'll try backing it off.

I thought so. The faster the attack, the more closely the volume change will conform to the shape of the initial transient, which is what you want to keep, at least most of it.


Also this project is getting mastered - and I thinking if the whole thing is going to get squashed even more, I should probably use less compression in mixing.

Don't let them butcher your music! You should make two things very clear to them.

1. There are to be NO clipped peaks!
2. No song should have an RMS value greater than -18 to -16 or so. (Some songs today go as high as -6!)

Your songs should look more like this: http://img.photobucket.com/albums/v135/PiccoloNamek/Gpagoda.png
And NOT like this: http://img.photobucket.com/albums/v135/PiccoloNamek/Oworld.png

If you songs look like the second example, find another company to master your work! And don't let them tell you stuff like "it needs to be made extra loud for radio play" or whatnot. Radio stations have of their own compression and limiting equipment to do that. (Plus, since these devices respond to the input amplitude of the signal, a sound that was pre-loudened will actually come out more quiet in the end!)

I'm saying this because the truth is that nobody wants to listen to music with no dynamic range. All of the punch and impact, and most importantly, emotion is lost, if everything is at a consistent amplitude. It becomes like what noise in the background, and in the end the listener just wants to turn it down so that it will go away. You don't want this, I'm sure.

Always keep this in mind, and everything will be OK. :)

[EL84]

...I find myself turning up the midrange on the guitar amps and backing off the distortion too.

Yeah this is sort of what I'm doing these days too. As a teen I was gain on eleven, scooped mids, but many classic guitar tones that I love are the complete opposite of this approach, e.g. early Van Halen, early Jimmy Page, they don't have much distortion, but sounds really heavy.



What I'm trying to capture is the bassy palm muted, high distortion stuff. Almost metal-like...

I like Dave Mustaine's sound for this, but I don't know quite how to achieve it, I got something like it once with a Hafler preamp. One thing I do which I like is putting a mosquito distorted guitar hard left, and a much less distorted middly guitar hard right, both playing exactly the same thing. At the end, there's all the extra high frequency harmonic distortion stuff, but there's still some definition.





1. There are to be NO clipped peaks!
2. No song should have an RMS value greater than -18 to -16 or so.

Amen to that. Glad others out there still have ears. Isolating sometimes, being aware of modern clipped ultra-compressed recordings.

[int]

I like Dave Mustaine's sound for this, but I don't know quite how to achieve it, I got something like it once with a Hafler preamp. One thing I do which I like is putting a mosquito distorted guitar hard left, and a much less distorted middly guitar hard right, both playing exactly the same thing. At the end, there's all the extra high frequency harmonic distortion stuff, but there's still some definition.


When I was a kid I read an interview of Jerry Cantrell, guitarist for Alice in Chains. What he liked to do is record rhythm guitars twice, and pan one left and one right. Then do leads up the middle (or pan them appropriately if there is more than one track of leads).

This makes the rhythm simple to EQ, compress, and level in mixing. On my board I can set them in a stereo pair and EQ/compress as needed. And since their the same parts, leveling them is easy too. I'll make them a little different by changing one track in the box (with EQ or dynamics/effects processing. One trick on clean rhythms is to put a light chorus on one [even though double tracking the same part is a chorus effect in itself]). I'll then send the stereo pair to an overall delay or reverb, as needed.

For this project I'm upping the EQ on the guitars around 150 hz, 2.5k, then a high shelf at 8k. About 3 db for each.

I think I need to learn how to figure out where to cut, to make those boosted frequencies I'm boosting stand out more (instead of adding EQ).

[int]

Don't let them butcher your music! You should make two things very clear to them.

1. There are to be NO clipped peaks!
2. No song should have an RMS value greater than -18 to -16 or so. (Some songs today go as high as -6!)


1.) I've found clipping is fine AS LONG AS IT'S NOT HEARD. And even if it's heard, as long as it still sounds ok (verrryyy subtle). I don't care much about the radio thing but getting consistency from song-to-song is tough. I'm competing with Blink 182 and the Offspring in some cases here.

2.) This mastering guy is happy with an RMS of -12. Although I think if I go lower I can give him more room to work.

[PiccoloNamek]

I'm competing with Blink 182 and the Offspring in some cases here.

I don't think clipping is fine, ever. To me, it is indicative of sloppiness and carelessness.

Also, what are you competing for? Loudness? If you're trying to get noticed by some hotshot exec or DJ, I can see what you're saying. I don't agree with it, but I understand the situation. But trust me, the listeners (the people who matter) won't want to listen to music that has a consistent amplitude.

It really is a crappy situation. It seems nearly impossible to please both parties.

Also, I would be wary about making things too bright in EQ. A lot of cheap speakers and headphones already have a U-shaped equalization built-in... boosting the high frequencies (especially the upper midrange, from around 2 to 7kHz) will make things sound very harsh.

[earwax]

I have to agree. Clipping is never good. Clipping in the recording is distortion. I think my job, as an engineer, is to capture the sound the musician makes. No more, no less.

[int]

I don't think clipping is fine, ever. To me, it is indicative of sloppiness and carelessness.

Also, what are you competing for? Loudness? If you're trying to get noticed by some hotshot exec or DJ, I can see what you're saying. I don't agree with it, but I understand the situation. But trust me, the listeners (the people who matter) won't want to listen to music that has a consistent amplitude.



I'm competing for volume consistency from song to song. Nothing more. I don't like having to be up there with the major-label crap either.

I agree clipping is sloppy - I can master it myself and do fine. I'm not picking the mastering guy here though - I suggested someone the project and band thinks they need. Their marketing plan is better than mine and are into my suggestion though, so I gotta run with it. He overs a lot - but it's subtle and works with this style of music.

The punk/metal/power-pop aspect of this group dominates - loudness and its associated artifacts are accepted by the listener. Normally I'm with you guys - this is one of those "other" cases.

The acoustic stuff is usually under control.

[int]

I have to agree. Clipping is never good. Clipping in the recording is distortion. I think my job, as an engineer, is to capture the sound the musician makes. No more, no less.

Clipping to tape or disk, bad.

Clipping under careful limiters and ears - sometimes ok.

[PiccoloNamek]

The problem with clipping in the digital domain is that once it happens, what information was there is gone for good, just like when you push a brightness value to 255 in Photoshop. The problem is that the actual total loudness of a CD can never be increased... once all 16 bits are filled up, that's it.

[int]

The problem with clipping in the digital domain is that once it happens, what information was there is gone for good, just like when you push a brightness value to 255 in Photoshop. The problem is that the actual total loudness of a CD can never be increased... once all 16 bits are filled up, that's it.

Right.

The thing is, with modern equipment and dithering some of the problems are nulled - in some styles of music.

There's a technique where you subtly layer a Waves L2 at -3 threshold until the loudness is "proper." Some transients get through near the end, but since so many subtle layers aare catching the back end of them, it seems to be ok.

Going along with your Photoshop analogy - it's sometimes OK to have pure white on the screen. If noone sees or finds it offensive, and the rest of the picture looks good - then why wouldn't you do it?

[PiccoloNamek]

I'm not fully convinced such a technique would work all the time. It reminds me of something I noticed once... there is a song in the FFXII soundtrack called "The Dalmasca Estersand". Unfortunately, the loudness of the track was strongly maximized, which really isn't normal for this particular genre of music. Anyway, at several points in the song, they attempted to "hide" clipping within a strong cymbal hit. The transient is very short and quick, but with my highly resolving system, I saw what they did there anyway. ;) They probably thought they were being clever, but in the end, all they got was a cymbal with static pops mixed in.

Also, please don't think I'm arguing or trying to tell you how to do things. I just find all of this audio discussion to be very stimulating. It is one of the few things I am good at... plus, it's my job, too. (Although I'm really more of a sound editor than an engineer.)


Going along with your Photoshop analogy - it's sometimes OK to have pure white on the screen. If noone sees or finds it offensive, and the rest of the picture looks good - then why wouldn't you do it?

In real life, specular highlights do appear to be pure white... but in real life, very loud sounds don't have harsh static popping mixed in. ;)

[int]

I'm not fully convinced such a technique would work all the time.

That's what I'm saying.


Also, please don't think I'm arguing or trying to tell you how to do things. I just find all of this audio discussion to be very stimulating. It is one of the few things I am good at... plus, it's my job, too. (Although I'm really more of a sound editor than an engineer.)

I know. When I was working at taht music distribution company we would master soundboard audio for digital distribution. When I was training the audio interns I made sure they kept the RMS low and stayed away from heavy limiting.



...but in real life, very loud sounds don't have harsh static popping mixed in. ;)

I've got a Mexican Squier and a solid state Crate amp that beg to differ. :)