Stephen - when using multiband compression to shape a bass track, is there a good "rule of thumb" to how to set the cross over frequencies? For example, if I wanted to the break the bass track frequency band into 3 frequencies, which 3 frequency bands should it be?
As with most questions of this nature, there are no set rules, as it all depends on the song, the sound of the bass, the way the bass was played, as well as the sound you are going after. It might be over used, but the saying applies here: Use your ears!
That's the short and simple answer, anyway.
The longer answer is that I can't recall a time where I have ever used a multi-band compressor on a bass track. I have used some bass enhancers that generate harmonics to help the bass cut through more, and they usually have a cutoff frequency control that I sweep around until I get the desired effect. However, as I got my start in the analog world where we used single-band compressors like the LA-2A for bass, I pretty much always use simple single band compression on bass when I feel it needs some compression.
About the only situation I can think of where I might want to use a multi-band compressor on bass would be for a really busy slap bass type sound where there are going back and forth between low notes and then hard slapped high notes, and the higher notes are cutting through way too much. In that case, I may try to put a really fast multi-band compressor on there and try to isolate those high slap notes to a band or two and really clamp down on them.
The biggest problem I usually have with bass is getting it to the point where it cuts through a busy mix without it getting all boomy or muddy, and also making sure it can be heard on speakers (or earbuds) that don't have any deep bass response. I have discussed this many times before, and it usually involves rolling off lots of the deep frequencies from the bass and bringing out more mid-range frequencies either with EQ or by adding additional harmonics through some sort of distortion.
There are lots of psycho-acoustic effects you can take advantage of, if you are aware of them. One of them is that even if you totally filter out the fundamental frequency of an instrument, especially something like bass, as long as the harmonics are still there, our brains will fill in the missing information, and we will "hear" the fundamental in our minds and still know what note is being played.
This psycho-acoustic affect is very helpful for getting a present and full bass sound in a very dense/busy mix without making the whole mix too muddy. You can use a low shelving EQ to roll off a lot of the very deep bass frequencies from the bass, during busy parts of the song (you probably want to leave more deep bass in when the bass is much more exposed in the mix), and crank up the mids, or add some distortion, and then the bass can cut through the mix as much as you need it to, while still having the feeling of being very full and deep.
I do a LOT of automation with bass these days, especially with mixes that change back and forth from very sparse to very busy/dense. I will usually have the clean bass track, on which I may automate an EQ to pull out more bass during busy parts of the songs, and then I'll have a bass amp track, blended in to taste, and then some kind of very mid-range heavy harmonic track that is usually created by some sort of distortion effect, such as the Sound Toys Decapitator. I adjust the blend of these as needed to get the bass where I want for each section of the song.
Bass is a hard thing to get right, and you really have to listen to it in context with the rest of the tracks, or else any changes you make while listening to it solo usually won't work at all when all the other tracks are brought back in. I'm constantly going back and reworking the bass, with minor adjustments (or, sometimes major) as I get further along in the mix and get all the other tracks sitting where I want them.
Sorry... not much of a direct answer to your question, but I guess I'm not the person to ask since I rarely (maybe never?) use a multi-band compressor on bass. In fact, I rarely use it on any tracks, except for de-essing vocals (and sometimes de-essing drums for overly loud/splashy cymbals). If I did use a multi-band compressor, I still wouldn't have any fixed "rule of thumb" frequencies I would use. If I was splitting it into 3 bands, I would try to isolate the lower fundamental frequencies in one band, which would depend on the key of the song and the sound of the bass, and then the second band would be the mids and the most important harmonics, while the third band would be more string noise and amp hiss/noise type stuff. In that situation, it would be a lot of sliding the cross-over points around to find the best compromise, and then I'd probably use a slower attack & release on the low band, and quicker attack/release on the mid band, and then adjust compression amount and make-up gain on each band as needed to make it work in each section of the song and for the sound and style of playing of the bass player.
Follow up via email:
Stephen - I don't know if you answered the question or not exactly. I have read this article before and it was very informative. However, my question is one of being able to "even out the bass" to give the track a more balanced bass response. As you mentioned/answered in a previous question, the lower frequencies tend to excite the compressor more because the frequencies spend more time above the threshold. However, when I place a compressor on the bass guitar track it also greatly affects the higher bass frequencies. My question was when using a multiband compressor, are there suggested frequencies (I have a 3 band compressor you might say with ability to adjust frequency bands) with which to set the frequency bands to help even out the frequencies and bass response to get a more even sound? Is what I'm asking make sense?
Based on this question/reply, as well as some other questions you have asked in the past, I'm wondering if perhaps you are too worried about "flat" or "even" frequency responses, and trying to mix too much with your eyes and analyzers, rather than just listening.
I never approach mixing that way, and almost never look at an analyzer. I trust my monitors and my ears, and my only goal is to make things sound "good", which is more about emotion and feeling than anything technical. Sure, I work to achieve a good balance in the mix, but not in technical terms of having any kind of smooth frequency response, or certain frequency curve. It's all about can I hear everything clearly (that I want to be heard), and does the overall sound serve the emotional message of the song. In some cases, you may not want things to sound balanced... you may want something to really jump out and grab the listener. Or you may intentionally want some sections to be smaller, or even dull sounding, so that you can make other sections really jump out even more.
There are no right or wrong ways of mixing, and it's the artist who always has the final say of how he/she wants to hear things when I'm mixing. However, I will say that if you are relying on analyzers and response curves to mix, then you are probably never going to be happy with the results.
Again, I can't really help you as I have never come across a situation where I needed to use a multi-band compressor on the bass. I guess I have never had to work with a bass sound that was so horrible that I needed to split it up that way to try to fix it. Now, some of the single-band compressors I work with do have filters on their sidechains, or other kinds of "tilt" or bias type controls that allow you to fine tune how much they react to the lower frequencies (lower frequencies have more energy, and usually make a compressor react more), but that's still different than a multi-band compressor since the compression is still acting on the entire signal at once, you are just modify which frequencies it responds to more (kind of like a de-esser is tuned to react only to the sibilant frequencies).
Of course, then you have people like Bruce Swedien, who don't believe in using any compression at all for tracking or mixing, and just balances everything out with fader moves and automation.
I like to use compression, but I don't rely on it to do all the work for me. I still do a lot of volume automation on my tracks, including bass, to make it work in the mix and balance things out.
Bass is really not an instrument where you want it to have a balanced frequency response. By it's nature, it is going to have very strong deep low frequencies (which I usually end up thinning out), and then some low mid, mids, and only a tiny bit of hi-mids (string noise, plucks, etc.) and almost nothing in the high frequency range that's of any use. The two areas I key in on are the lows to make sure the bass still sounds full, but not muddy, and the mids to make sure you can hear it in a busy mix or on small speakers. I don't care what the relative levels are as far as on any kind of meter/graph between the lows and the mids, I only care on how it sounds with everything else in the mix. For mixes with lots of really heavy distorted guitars, I'm going to make the bass have a LOT of mids if I want it to cut through and have people be able to pick out the notes it is playing... if I then soloed the bass after I got it working in that situation, it would sound really thin and wimpy on its own, as it wouldn't have much deep low end at all.
So, again, I don't go into this trying to think I need to balance all the frequencies of the bass. It's not about that at all. I never try to even out all the frequencies. It's all about how it works in the mix.
Maybe if there was something wrong with the way the bass was played or recorded so that certain frequencies were WAY louder or WAY softer than everything else, then maybe I would get a multiband compressor so I could isolate that frequency range and fix it. Otherwise, I'm not thinking the same way you are when I'm working with bass, or pretty much any other track... it's never about trying to make all the frequencies even.
But, if you really want some numbers to work with, I would break down the key frequency ranges as follows, keeping in mind that they will shift by a certain amount depending on the type of bass, key of the song, how it was recorded, etc. None of these numbers are exact, or hard and fast cutoffs:
So, if you were going to split it into 3 ranges for multiband compression, I would probably do the low band for everything below 200Hz, the mid band for 200Hz to 1000Hz, and the high band set to 2000 to 6000hz with a good roll-off above that (if possible). Those would probably be the 3 critical ranges you could use to really shape the sound of the bass.