Thoughts on frequency dependent compression

Discussion in 'Miscellaneous [BG]' started by Rockin John, Feb 24, 2005.

  1. Hello folks.

    This post is by way of a sort-of academic thought rather than something I wish to rush out and do. However, I do understand it's a common enough concept.

    The senario:

    If I wish to compress my bass' signal I feed it into a compressor. That compressor will / should compress all the components of the note (fundamental + however many harmonics). So, giving (say) low B a damned good pull, the compressor kicks in at some pre-determined point and the whole note becomes compressed.

    That implies that all the components of the note are compressed in relation to the largest amplitude (usually the fundamental, I guess). But that seems a bit pointless because, surely, the lower amplitude components of the note get compressed, too?

    It makes me wonder whether: a) simply applying some bass cut would achieve the same end; b) apply frequency selective compression so the the big thumpy bass nots get squashed but the higher pitch harmonics get left alone.

    Or am I talking silly (as usual)? :eek:

    Views welcome.

  2. paintandsk8

    paintandsk8 Pushin' my soul through the wire...

    May 12, 2003
    West Lafayette, IN
    I believe you are talking silly. I ran some tests in sound forge.

    I mixed a 50 Hz 1000Hz and 10000Hz sine wave @ equal levels. I then made a spectrum analasys graph. Then I applied heavy single band compression. Then I made a new spectrum analysis graph.

    The result: All frequencies stayed at equal levels in relation to each other.

  3. paintandsk8

    paintandsk8 Pushin' my soul through the wire...

    May 12, 2003
    West Lafayette, IN

    Now that I have read your post a few more times, I'm not sure whether I actually answered your question, or for that matter what your actually asking.

    Would applying bass cut achieve the same end as single band compression? - no

    Would applying bass cut achieve the same end as applying multi-band compression? - Depending on the settings used you could get a very similar result.

    Maybe if we knew what your end goal of this theoretical compression was, then we would know what type of effect to recommend. So what do you want your bass to sound like in the end?
  4. Petebass


    Dec 22, 2002
    QLD Australia
    John your whole theory revolves around the fundamental being the largest amplitude harmonic. On a bass guitar, it often isn't. The second harmonic is often quit a bit bigger.

    Have a look at page 2 of THIS THREAD . They were talking about amps rather than compression but in post 36, Mark Reccord recorded 2 of his basses through a spectrum analiser, then posted a screen-shot. Both basses showed a 2nd Harmonic that was much larger than the fundamental.
  5. "Hello" and thanks to you both.

    This is not about trying to create a particular bass sound, although there are slight real-world implications for me. The tread was spawned by something I'd read just the other day (and I've been looking everywhere for the papers since, but can't find them). BTW the frequency response thread is excellent: I've saved it on my computer.

    As I remember, the article talked about the "Muffling" effect of compressors. From memory, it dealt with the idea of larger amplitude signals (Eg. Bass Guitar, bass drum) triggering a compressor. The result was that all signals in the programme content - including the high frequencies were reduced. That gave a muffled effect.

    I believe the article was concerned with compressing audio for recording on disk (or whatever). Again from memory, the (studio) compressor had an EQ placed somewhere in the line. This device somehow reduced low frequency, high amplitude peaks such that the compressor was not "fooled" by them. It was me that extended this idea to compression of a bass guitar alone, and wondered whether the same reasoning applied.


  6. jondog


    Mar 14, 2002
    NYC metro area
    You can buy multiband compressors, I used to use and Akai hexacomp (6 bands, each w/ level and comp level controls). I ended up thinking about it like a fancy equalizer that had the added benefit of controlling peaks.
  7. Passinwind

    Passinwind I Know Nothing Supporting Member

    What you're referring to is known as sidechain EQ. Basically, you patch the EQ in just before the detector section of the compressor. You'll need a compressor with a sidechain loop, which is fairly common but far from universally so. I use it on bass occasionally, and it can do some cool things, but carrying an extra parametric EQ in my rack for it seems like overkill for the most part. Do a google on it, you'll find plenty of reading material online. The Rane website is a good place to start if you want more insight.
  8. I think, Passinwind, you describe exactly what I was trying to say.

    Since the original post I have picked up a compressor (made by Behr....... :bag: no, won't say that word in full in case World War 3 begins on TB) that has a sidechain in / out. I also have one of their parametric EQs.

    I will check you references when time permits but, I wonder, can you suggets a way of connecting the gear together so I can at least experiment?


  9. Finger Blister

    Finger Blister

    Jul 8, 2003
    Valid concern.

    My compressor has a Low Pass Filter.

    The Largest Amplitude Signals don't trigger compression.
    The result is that the high frequencies aren't muffled by over-compression
    triggered by Lows.

    I've found that a Sonic Maximizer "re-conditions" the compressed
    signal and gives it back the 'life' that compression can take away.