The Jazz Guitar Chord Dictionary
  1. #1

    User Info Menu

    Say you record a tune with other musicians, each one into a separate track into a DAW.

    You can now solo your track and see exactly where your notes/chords are by seeing both the wave form and the time lines (from the DAW's metronome) at the same time.

    If you nail every single chord at the instant the metronome suggests, the track would, presumably, sound mechanical, like a chart rendered into music by Sibelius.

    So, what does a more natural sound mean, in terms of what you expect to see on the screen? How far off from the click , in milliseconds, is a more "natural" sound?

    If you're playing swing feel, I'd assume that it will sound less mechanical if you aren't right on the click or the eighths. So, where, in ms, should the chords be?

    If you're editing a track, how many ms off the click would you allow as "close enough" to right on the click? My impression is that you can feel the lack of groove well before you hear a flam.

    I'm guessing this has been studied and I'm also guessing it's tempo dependent.

  2.  

    The Jazz Guitar Chord Dictionary
     
  3. #2

    User Info Menu

    Yikes, that's a rabbit hole. I try not to look and just use my ears, but I succumb to the temptation.

    Should the waveform of a drummer relative to the grid look the same as the bass?, the guitar?, the vocals? Or, should there be some give and take between instruments for a good groove?

  4. #3

    User Info Menu

    The hi-hat(and ride) on main beats should be very accurate. Pretty much all else is all about the feel. Some styles need to sound a bit like machine, but many are laid back a bit.. here and there.
    What makes it tricky, is that most instruments have 2 "starting points" - the initial attack and then the other - when it's tone starts to actually work.
    The lower the frequency, the later it will really kick in. That just means, there is no way of telling that in concrete milliseconds...
    Bass is sometimes suggested to be played very slightly ahead but when starting to modify it somehow later, this might not work the way it was intended by the player.
    I've also had problems with guitar when I played accurately like a robot but it didn't work at all. The moment the pick hits the strings is not when the listener feels the impact of the chord.

    Anyway, whenever going in and start fixing something, you're going to take over the control of the feel.
    It's such a pain to do it later. And the result is never gonna be as good as if it was recorded with skillful musicians who know what they are doing.

    About chords, I would try to maybe see if it's good to match them with snare feel first (this can be laid-back often - depends on how the drummer felt about the groove).
    No milliseconds, just plain "this sucks" & "this doesn't suck" judgement. If this won't work as whole, I've just gone "f@# this, just support the melody".

    Looking at the waveforms really is not the way to do it.
    ---
    Eh, not an expert here. Just had some experience and some advice from people.

  5. #4
    So far, my experience is that I can improve groove by getting the sounds right on the quarters (in guitar and piano comping) to line up with the click. Maybe allowing up to 20ms variation at 100bpm. I can tell it's improved, because I'm more drawn to tapping my foot when I hear it. But, the risk is that it's too homogenized.

    Still, the idea here is to know what great time feel looks like in the wave form. Seems like that would be a good thing to know.

    One thing I've noticed with a small sample size is that the one top pro (drummer) we have on these projects tends to be right on the click. I just checked a few snare drum hits --all less than 10ms from the click. A lot even closer than that. The rest of us vary, but, on average, we're late.
    The more variation, the more ragged the groove.

  6. #5

    User Info Menu

    I have actually monkeyed around a bit with cutting up tracks and shifting things toward or away from the click, but have always done this by feel (not by a number of ms, or any other metric). I experiment by what feels most like a groove, and/or to try to fix groove problems. If I did do this according to a number, I think it would it would depend on the tempo. I mean 10 ms might sound right on the beat at 60 bps, but off at 180. A couple of things I have observed is that bass can sound good a little early, but not constantly, for a whole tune. It's better to vary it a bit and have some sections or choruses right on the beat. Hits generally have to be right on the beat or right on the off beat, but not somewhere in between. But this thread does give me some food for thought, and maybe I'll go back to some recordings and see if anything more concrete jumps out.

    John

  7. #6
    Had a discussion today with a pianist who pointed out different players adjust the exact timing of the notes to create "swing feel" -- and different players do it differently. Not to mention that it's different at every tempo.

    He said that he's heard Herbie Hancock change the approach to swing feel in the middle of a phrase.

    Drummers talk about how differently the greats play a basic ride beat. Different by player, tempo and situation.

    That said, I would like to fully understand, mathematically as well as by feel, even one of these approaches -- just to get some inking about what works and what does not.

  8. #7
    I'm going to try to resurrect this thread, since it's a topic I find fascinating.

    I was fortunate enough to enlist a top Brazilian pro drummer on a project. He submitted his track (using Reaper in Brazil) and then, after some discussion (with him in the role of teacher) recorded another take to demonstrate a different approach.

    His recommended approach for this tune, a samba at 100bpm in 2/4, was in what he called a "jazz" style. His drum track tended to be about 35ms ahead of the click. Now and then, he'd hit something on the click. This is a sophisticated style of playing, not an error. That's about a 64th note early.

    To show us the difference, he was generous enough to record a "metronomic" version in which he was still ahead but only by about 5ms. That's pretty close to the click -- you can't hear it and I'm not certain if you can really feel it.

    But, you can feel the difference (30ms or so) between the two versions. The one that's 35ms ahead feels better and sounds more like authentic samba.

    He recommended that we use our ears, not the waveform, to place the other tracks.

    We still couldn't stop thinking about the waveform.

    If the drums are 35ms ahead, where should bass, keys and guitar be? Bearing in mind, that the answer is likely to be different based on song, style, player, tempo etc. But, what about for this song with this drum track?

    As it happened, the bassist was often 15-20ms behind, but with a lot of variability. Some of his notes were ahead. The drummer suggested that he might have been too far behind, but the bassist, a fine player of this style, didn't think he could do any better. And he liked what he had done. So did I.

    The pianist, a stickler for accuracy, played right on the click or, perhaps, edited in MIDI to get right on the click.

    On guitar, I played the first part of the song a bit behind (unintentionally - probably took me a chorus to switch from recording engineer to performer) and then played the rest more or less on the click, which sounded better. So, I cut up the track and moved the errant notes close to the click.

    Here's a link to the current rough mix of the track. Muito Non, first on the list (no sign-in required for this link, I fixed that).

    Ginga | SoundClick

    There are threads on here regularly dealing with the details of note and scale choices. Not so much about the details of time.
    Last edited by rpjazzguitar; 09-29-2020 at 01:12 AM.