Questions about sound pressure and sampling rate...

[Caravaggio]

G'day, 1st off. Do you ever have problems with too much pressure in your songs? Sometimes the backing oscillators or orchestral sections I have start to feel like they're "pushing" in on my ears if they play for too long. It sounds great, but it literally feels heavy, if that makes sense. I've been trying various things like cutting the end frequencies, but that of course makes it sound different, and lowering gain or volume levels of course makes it too quiet (They're pretty quiet already, which is the weird thing). Any ideas on what functions or options I can use/pull back on while causing the least damage?

The 2nd thing I wanted to ask has probably been discussed, but terms like 44 and khz are too short for the search engine (I got some things for 44100, but not much.). Does anyone mix in higher hertz than 44,100?

I ask because working mainly with vsti instruments I've been using a lot of fx, distortions and every other thing under the sun. It sounds okay but I inevitably get unwanted sounds when too many effects (sometimes one or two are enough) are stacked and after reading that higher sample rates may help alleviate that problem I wondered if it wouldn't be worth it to squeeze as much quality as I can out of a sound before the sound starts to tear in various ways. That makes sense but it seems to be under debate. Other people claim that the conversion from something like 88 or 96 to the final 44k release does more damage to the sound than just using 44k in the first place.

And if any of you have changed that figure, did it cause trouble with existing projects?

Any help appreciated.

[Sole Signal]

At first it may seem that cutting the low frequencies (I assume that's what you mean by "end") will make your track sound worse or less powerful, but to the outside listener it will often actually improve the listenability and sound better. For example, if you've got a low pad cluttering up space with your dedicated bass line and/or the higher frequencies of your kick, you'll inevitably end up with a bunch of mush. This may be the "pressure" that you're referring to. Carve out a spot for your bass in the pad using EQ and ditto for the kick. Once you've done this, you can then often raise the volume levels to match your previous perceived loudness, with the added bonus that you will now notice improved clarity in the low end.

I would suspect that EQ problems are causing your other issue with "stacked" instruments. I know it seems laborious, but taking the time to separate your synths from each other with EQ keeps things sounding much cleaner and will certainly improve the listener fatigue that you seem to be describing. In short, switching to an (arguably) better sample rate probably won't do anything to help you out. 44k should be all you need.

Also, you may find this handy.

[Moseph]

I'm not sure if it's the same thing you're describing or not, but I sometimes feel like the midrange frequencies of things I'm listening to are, for lack of a better description, too close to me in relation to the other frequencies. It happens even on professionally produced music, and it's caused by the headphones I'm using -- they're kind of cheapo. What kind of listening setup do you have?

I've never seen any actual evidence that mixing at above 44.1 khz improves anything. Everything I've heard about it has been anecdotal. Personally, I don't see any reason to do it unless you plan on making a version of the music with a higher sampling rate in the future.

[Fray]

According to the tests outlined in "Mastering Audio: The Art and Science" (really good book btw!), the only improvement you ever really get from going above 44.1kHz is that it's easier for the A/D converter to prevent aliasing without coloring the sound. Basically if you have any input with a frequency more than half of your sampling rate, you'll get this raunchy-sounding aliasing effect. Even very good human ears can't hear higher than 20kHz, so with 44.1kHz you should be fine, right? Well, the trick is designing an analog filter that cuts out enough of the frequencies over 22.1kHz without changing the signal beneath 20kHz in the audible range. Going up to 96kHz gives the filter more room to prevent aliasing without messing up audible frequencies. Even at 44.1kHz though, the coloration is going to be pretty subtle -- most people cannot even hear it, and it's definitely not what's causing your listener fatigue.

Like Audix said, chances are your mix has too much in the midrange (anywhere from say 500Hz to 6kHz). Here's what you can do:

1. Fire up your EQ on the whole mix, or just the track giving you problems. Leave it flat to start out with.

2. Set the Q on one band to 1.0, and set its gain to something fairly high like 12dB. You want to be able to hear the change from the EQ very clearly, but if your mix starts to distort you may need to turn it down or just lower the mix volume temporarily.

3. Play your track while slowly changing the frequency of the band. You want to find the frequency where the "heaviness" you described is the strongest.

4. See if you can narrow it down a bit by increasing the Q. This way you'll change the tone less.

5. Now drop that band down to -6dB or so an see if that fixes the problem.

Also, you may just be using too many effects. When you overdo it on those, you can end up with all kinds of problems, listener fatigue included. See if you can cut out the effects that aren't really necessary to getting the sound you want.

Admittedly I find these problems pretty difficult to deal with, so if someone comes along with better advice, feel free to ignore mine

[Moseph]

According to the tests outlined in "Mastering Audio: The Art and Science" (really good book btw!), the only improvement you ever really get from going above 44.1kHz is that it's easier for the A/D converter to prevent aliasing without coloring the sound. Basically if you have any input with a frequency more than half of your sampling rate, you'll get this raunchy-sounding aliasing effect. Even very good human ears can't hear higher than 20kHz, so with 44.1kHz you should be fine, right? Well, the trick is designing an analog filter that cuts out enough of the frequencies over 22.1kHz without changing the signal beneath 20kHz in the audible range. Going up to 96kHz gives the filter more room to prevent aliasing without messing up audible frequencies. Even at 44.1kHz though, the coloration is going to be pretty subtle -- most people cannot even hear it, and it's definitely not what's causing your listener fatigue.

Good stuff; this hadn't occurred to me.

[Fray]

Good stuff; this hadn't occurred to me.

Thanks, I love to get all nerdy about this stuff

[Caravaggio]

Thanks for the replies. I actually wasn't thinking about the mid-range, but I'll test it now. I'll try things like a more visual equalizer and experiment with parts dropped out to identify tracks or combinations of tracks to narrow it down.

I've got some 15$ cheap sony headponhes (which I think are actually better than some 30 or 40 dollar ones I've had) and some mid-range bose desktop speakers. The speakers might be a little bass heavy, but it's not as bad with commercial music, so I never really thought about that.

Upon thinking about it more I also worry that I might have a habit of riding the stereo separation a little too hard in my effort to make things more organic. As usual, I'll just have to keep trying different things.

Cotton candy.

[SnappleMan]

Yes, cutting frequencies is what you need to do, but also take into account all the effects that you're using. Reverbs can very quickly drown a track in a heavy, muddy frequency stew. Always check the EQ levels of the reverb, use the high pass and low pass filters to only keep the reverb frequency you want.

Also, you said that you add distortion, this completely fills up the spectrum. Distorting a B3 organ and then playing a full chord will literally eat up your entire bandwidth. Your goal in mixing a song should be to find and isolate the sweet spots of all your tracks with as little frequency crossover as possible, of course, this doesn't mean brick-wall everything till it's separate, just make sure you familiarize yourself with the sounds you're working with and don't overlook all the frequency boosts that the effects you're using bring to the mix.

[Zephyr]

A little panning and stereo can go a long way in clearing this up just a bit, but EQ will probably be more important.

[Fray]

I've got some 15$ cheap sony headponhes (which I think are actually better than some 30 or 40 dollar ones I've had) and some mid-range bose desktop speakers. The speakers might be a little bass heavy, but it's not as bad with commercial music, so I never really thought about that.

You're on the right track comparing your stuff to pro mixes -- it's not as good as getting real monitors, but you can at least partially get around the limitations of your setup by learning how a good, portable mix sounds on your particular system. If commercial music (a lot of which is actually pretty fatiguing anyway) doesn't hurt your ears on your speakers, neither should your own mix.

So umm... let's hear your mix already