1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

How to treat / avoid overly dynamic instrument tracks

Discussion in 'Recording' started by gunsofbrixton, Mar 8, 2010.

  1. gunsofbrixton

    gunsofbrixton Active Member


    I am still at the beginning of my career as a recording / mixing engineer and this is an issue that I come across quite often: pop / rock songs with a lot of dynamics, e.g. soft verses and loud choruses. As we all know, current rock / pop music is heavily compressed, so that in the end, if you display the wave form of a pop song, it's basically one big "sausage", with the peaks in the "soft" parts at the same level as in the "loud" parts.

    Considering this, I've been telling musicians (most of all drummers) to play with less dynamics, because in the end everything will be evened out anyway. But often they will say something like "We can't do that. Dynamics are really crucial to this song etc."

    How do you approach this issue? I see 3 options:
    a) Tell musicians to play less dynamically.
    b) Use a compressor in a fixed setting throughout the song. This will mean that the loud passages are WAY more compressed than the soft passages.
    c) Adjust the input or threshold of compressors and ride track faders during the song manually or with automation. This will lead to a more even compression and more even levels.

    I would guess that the best thing would be a combination of a) and c).


    Addition: To avoid misunderstandings: Options b) and c) are meant in the mixing stage, not during tracking!
  2. TheArchitect

    TheArchitect Active Member

    You should never tell a musician to play with less dynamics. The musicians should perform the song the way it meeds to be performed. So what if other tunes get smashed with a limiter. That does not mean it should have been or that tune your working on needs to be. Out job is to capture a performance, not fabricate one that is more easily smashed with a limiter for no particular reason
  3. TheJackAttack

    TheJackAttack Distinguished Member

    Just because current music is crushed by compressors does not mean that the initial tracks are crushed. You would ideally record without any limiting or compression with the PEAKS hitting about -12dB. This means that the RMS level is well below -12dB. Once the tune is recorded then you tweak or overdub or add fx or whatever and mix it down to the 2-bus. At that point you can begin to worry about evening the RMS for all songs on an album-pulling down volume if necessary on some songs. Then and only then do you begin to worry about limiting and bringing the RMS up to sh***y non-dynamic radio levels.

    At your stage of the game you're a long way away from worrying about the final product.
  4. TheJackAttack

    TheJackAttack Distinguished Member

    And you don't ride the faders in a recording session. Period. You make sure the levels are low enough to not peak higher than -12dB and leave it alone. Recording is not the same as riding the board in a live PA situation where feedback and the drunken crowd determine your relative levels. Two subspecies (studio/live) of the same animal. "Live" recordings are most often taken from a split to a separate recording setup anyway and mixed later.
  5. albatrocity

    albatrocity Guest

    If you know what sound you're going for, I'd say using compression when tracking is fine, and it saves time in the mixing process if it's done well. If it's not done well, you've ruined your original tracks. I usually go pretty light on compression when tracking if I end up using it. Just make sure you get good levels with enough headroom to work with.

    I completely agree with TheArchitect, though, in saying never tell a musician to play with less dynamics in order to achieve something that's inorganic in the first place.

    If you really want to get the modern dynamic-less sound, you'll probably just have to ride the faders in the mixing process. Using reverbs well can accomplish a sense of space and dynamics even if it's not really that dynamic.
  6. gunsofbrixton

    gunsofbrixton Active Member

    Are you saying I shouldn't worry about dynamics and compression before the mastering stage??? You can't be serious. Drum compression - in fact compression in general - is such a vital part of every pop / rock mix.

    Plus I never said I'd want to ride a fader during recording. You go me completely wrong there. Obviously I had the mixing stage in mind.

    Somehow I am a bit surprised how offended everyone seems to be at the thought of someone trying to reduce dynamics in a rock song. Let's face it - if you want to release a record it needs to have a certain RMS level. And if you stay within reasonable boundaries that doesn't mean it sounds bad.
  7. gunsofbrixton

    gunsofbrixton Active Member

    Let me put my question differently: Assuming you want to mix the drums of a song with a quiet verse and a loud chorus. The snare and kick hits are not nearly as loud in the verse as in the chorus. And you want to even it out a little. How do you proceed? Would you change / automate compressor settings to allow some compression of the softer hits without overcompressing the loud hits, and then simply bring up the quieter parts with the faders?
  8. TheJackAttack

    TheJackAttack Distinguished Member

    I'm saying you shouldn't be compressing the 2-bus. Sure, your guitar and bass players may have some fx going on in their amp heads (unless you are going direct into the interface) and I don't mean those. As far as compressing drums, in a recording studio you don't do much of that in the initial tracking. The recording stage is generally as pristine as is reasonable for the style of music you are recording. In the mixing stage would be where you would do some basic compression/gate/ what have you to tighten the sound up. This still isn't where you remove the dynamics.

    If I misunderstood what you typed then I apologize. At no point did this sound like a mix stage since you had just typed about telling musicians to play without dynamics or feeling. It isn't obvious because we see this type of thing all the time in the forum and in mixes brought in to us to fix if we can.

    Regardless, you don't try to chop the dynamics in the beginning stages. You already stated you are new to this so I outlined the general order of events for you earlier. The minimal dynamics and high RMS levels you want are done in the last stages after you already have all the songs sounding great and the average RMS levels of ALL the songs on a roughly similar level. You're trying to put the cart before the horse.

    Also, you want to mix at the softest level you can hear all the elements of the song. This will save your ears and make your mixes better. You can always turn the volume nod when you get ready to test the tune before you go to the mastering stage. Then turn the volume back down.

    The offense isn't at you in particular. It is at the complete destruction of musicality that passes for acceptable music these days. Between radio and mp3's there is no longer an understanding of what real music sounds like in all it's emotional and dynamic glory. And this current state of affairs is offensive to both highly trained musicians as well as audiophiles. The fact that it is a reality doesn't change the feeling. A pro deals with it and moves on hopefully educating a few people that there is something better out there.
  9. TheJackAttack

    TheJackAttack Distinguished Member

    Ahhh, now we get into a mix question. Different engineers will have different approaches. If I really feel there is too much volume difference for the arc of the song I might bring the verse up. I "grew up" in audio riding a FOH desk and a MON desk so I like to subgroup things. For me then the drum kit is going to it's own submix or aux and that may or may not include the bass to fill out the rhythm section. I can then automate the one aux track to bump up or duck out an instrument group pretty easily. My criteria though in the mix stage has to do with the overall arch of the song and again not with a final RMS goal for the entire album. I''ll deal with that later.

    Sooner or later one of the full time mastering engineers will speak up but what I've posted will get you going for quite a while. You might also consider just making the mixes of the various songs as good as you can and then sending the whole project to a professional ME to finish off. A real ME has software and hardware tools that you will not have and can accomplish your goals quicker and more efficiently.
  10. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Try giving the musicians a good mix, including compression on recorded tracks, to play to. They will naturally play to the dynamics of the monitor mix.

    I actually do compress in tracking, but for many years I avoided it. You can't undo mistakes. You have to have a vision of how the song will be constructed in mixdown and really know how to get a particular result. I mostly use a Drawmer 1960 and an ART Pro VLA to give a little analog feel to tracks on their way into the DAW. If I only had lesser analog compressors I'd just skip it and do it ITB. Even with those compressors I tend toward conservative settings and usually compress more in mixdown.

    Regarding musicians and their desire for dynamics, actual mathematical dynamics and perceived dynamics are not the same thing. Let them play the way they need to play to express themselves musically and mix in a way that works both technically and musically. Much of what sounds like dynamics is really tonal instead of actual volume differences. If you expect the recording to end up on the air it will get crushed anyway, so take that into account. If you want the recording to sound as loud as other tracks on someone's iPod you will have to lose some dynamics one way or another. For better or worse, all the other music people are listening to forms the context in which your recordings will be heard and against which they will be judged.
  11. gunsofbrixton

    gunsofbrixton Active Member

    That's a good point. I think even if I bring up the quieter parts of a drum track a bit, the perceived dynamics will not be lost, because the tonality of the hits will be different.

    How would you go about compressing a drum track with quieter and louder parts during mixing? Would you automate the volume before the signal goes into the compressor, so that the drum hits are compressed in a similar fashion during the whole song?

    Anyway, I will stop telling musicians to play with less dynamics. Bad idea.
  12. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    It depends on the source and the goal. You could try something like a slower compressor followed by more of a limiter. If the quiet and loud parts correlate to song parts (verse, chorus, bridge) I might split them out to entirely different tracks with entirely different gain, eq and dynamics. Parallel compression might help. Try a bunch of stuff and do whatever sounds good.
  13. Guitarfreak

    Guitarfreak Well-Known Member

    When I mix drum tracks I set the compressor while looping a section of the song that I deem to be at an 'average' volume/dynamic level for the song. I set the compressor (on the parts of the kit that need it) to RMS around -3dB, then I play the entire track through and see if anything stands out. If the drummer hits harder it will compress it more, but it won't sound overcompressed, and it he hits it lighter then you have 3dB of space before it leaves the active range of the compressor.

    This is however only a starting point of mine that seems to work most of the time. Try this, then tweak as necessary to get it sounding right.
  14. TheJackAttack

    TheJackAttack Distinguished Member

    GF-you aren't setting the RMS at -3, your setting the peak. The RMS is the average volume without the peaks.
  15. Guitarfreak

    Guitarfreak Well-Known Member

    I meant to say 'average'. I said that I set the compressor TO RMS, to is the key word. To take off an average of 3dB for an average hit. Too much average, so I said RMS. Maybe it's just an 'average' technique...
  16. planet10

    planet10 Active Member

    if the drums are quieter in the verses than the choruses then its obvious that the rest of the instrumentation will be quieter as well then resulting in not giving a $*^t about the dynamics and telling a musician how to play his parts!
    as far as compression in a recording session, i use a touch on bass guitar and a touch on vocals (a touch being no more that -2db of GR) just to catch the loud parts and only the loud parts.
    i hope your not charging for your services as a recording studio.....

Share This Page