The guy who produced this song for himself did some sort of widening panning tricks with most of the guitar tracks, every single one of them except the bass is detuned at some point by something really close to 17 cents. If you don't pan the guitars exactly the way he did you'll end up with these gross resonances that seem to never quite resolve and are fighting each other in different ways throughout the track, (as the track progresses, things that may have been detuned are back at 0 and things that were at zero are detuned.)
The solution is to find either exactly the panning positions that the original artist used or you can automate pitch shift plugins to correct the difference, (since every track is specifically pitch shifted by a certain ammount not randomly,) or you can apply some degree of auto-tune to every guitar track to either pull things enough in the right direction that it sounds pleasing in the panning positions of your choice or all the way. I didnt go that route so I don't know how it would turn out but 80% of the pitch shifting seems straightforward.
Also, there are major timing issues centered around the drummer and the guitar player(s), who are basically rushing their downbeats. The swing factor is pretty consistent throughout, except if you decide to cut up the drums to marry them with the bass more smoothly, (there are some explicitly "oops" moments,) then you'll find the swing notes leading up to those downbeats will end up sounding wrong as a result. So, if I you don't want to move virtually every 16th note transient in the session onto a 16th note triplet grid, (to taste,) then don't even start doing timing edits. The better the drummer sounds the more the timing mistakes of the bass player will stick out, then the bass player will make the rhythm guitar parts sound bad, then the rhythm guitar will make the singer sound bad.... You have to go full circle on it once you start, especially with the guitar solo, because once the drummer is quantized to spell out the 16th note triplets which the guitarist is riffing on, it becomes super obvious when the guitar player falls off, but sounds great if they hit it together almost like a pre-ordained unision figure. Some degree of quantizing to grid and some degree of quantizing to each other is probably good, if everything is robotic its not the best, but as long as they rush or pull back together, (ie quantized to each other,) then it retains a human feel.
If you start bringing the guitars in tune just solo each track and listen to the bass, the bass itself is never out of tune.. (maybe for a split second on one note, but chaulk that up to musician error not some sort of edit or effect.)
Anyways, removing the widening effects, (feels almost like reverse engineering,) and fixing the timing mistakes is an excersise in itself, (I've put 70+ hours into my version already, I had to undo a bunch of work after I decided to go backwards and bring things into tune,) but it will leave you with something a little sweeter. I haven't decided if I'm going to auto-tune the vocal to some degree or perhaps leave it as the lone widened/L+R pitch shifted element, but I think the song might sound right with some degree of pitch correction in the vocal.
I mixed a snare and bass drum sample behind the already recorded drums to give them more "crack" and "thwack" respectively, loved the bass sound, basically just high passed and cut some select mids out of all of the different guitars, paralleled the drum overs and squashed them, (trying to hide drum edits as much as possible by making the drums a little trashy, but behind the replaced kick and snare sounds,) made a few reverbs for different guitars, loved the bass guitar sound, squashed it anyways, sent super high-pass vocal signal to a reverb, squashed that a few times reverbed it, split that left and right and offset them by some ammount of ms less than 30, then side-chained a compressor onto that with the input of the vocal itself, so like, an "exciter" and then continually mixed it down another .2 to .3 dbs for like hours, before realizing that the out of tune guitars were making the vocal sound worse than it was, and also that the tuning issues were screwing with my sense of eq and level balance. So I went almost back to the beginning on a few things, (the vocals mostly I was trying to counterbalance the tuning changes with volume automation until I realized what was going on, bleh.)
So basically I wanted to warn everyone who does this track that it has significant tuning issues and timing issues that should be addressed before you start sending auxes and copy-pasting trim plug-ins. Its really rewarding if you get the rhythm section to marry each other though, I will post my results when I get there.
I'm posting this here because this is the highest thread in the section and I want anyone who's touching the tracks for the first time to have a chance to see this before they work themselves into a corner. Be warned!
The solution is to find either exactly the panning positions that the original artist used or you can automate pitch shift plugins to correct the difference, (since every track is specifically pitch shifted by a certain ammount not randomly,) or you can apply some degree of auto-tune to every guitar track to either pull things enough in the right direction that it sounds pleasing in the panning positions of your choice or all the way. I didnt go that route so I don't know how it would turn out but 80% of the pitch shifting seems straightforward.
Also, there are major timing issues centered around the drummer and the guitar player(s), who are basically rushing their downbeats. The swing factor is pretty consistent throughout, except if you decide to cut up the drums to marry them with the bass more smoothly, (there are some explicitly "oops" moments,) then you'll find the swing notes leading up to those downbeats will end up sounding wrong as a result. So, if I you don't want to move virtually every 16th note transient in the session onto a 16th note triplet grid, (to taste,) then don't even start doing timing edits. The better the drummer sounds the more the timing mistakes of the bass player will stick out, then the bass player will make the rhythm guitar parts sound bad, then the rhythm guitar will make the singer sound bad.... You have to go full circle on it once you start, especially with the guitar solo, because once the drummer is quantized to spell out the 16th note triplets which the guitarist is riffing on, it becomes super obvious when the guitar player falls off, but sounds great if they hit it together almost like a pre-ordained unision figure. Some degree of quantizing to grid and some degree of quantizing to each other is probably good, if everything is robotic its not the best, but as long as they rush or pull back together, (ie quantized to each other,) then it retains a human feel.
If you start bringing the guitars in tune just solo each track and listen to the bass, the bass itself is never out of tune.. (maybe for a split second on one note, but chaulk that up to musician error not some sort of edit or effect.)
Anyways, removing the widening effects, (feels almost like reverse engineering,) and fixing the timing mistakes is an excersise in itself, (I've put 70+ hours into my version already, I had to undo a bunch of work after I decided to go backwards and bring things into tune,) but it will leave you with something a little sweeter. I haven't decided if I'm going to auto-tune the vocal to some degree or perhaps leave it as the lone widened/L+R pitch shifted element, but I think the song might sound right with some degree of pitch correction in the vocal.
I mixed a snare and bass drum sample behind the already recorded drums to give them more "crack" and "thwack" respectively, loved the bass sound, basically just high passed and cut some select mids out of all of the different guitars, paralleled the drum overs and squashed them, (trying to hide drum edits as much as possible by making the drums a little trashy, but behind the replaced kick and snare sounds,) made a few reverbs for different guitars, loved the bass guitar sound, squashed it anyways, sent super high-pass vocal signal to a reverb, squashed that a few times reverbed it, split that left and right and offset them by some ammount of ms less than 30, then side-chained a compressor onto that with the input of the vocal itself, so like, an "exciter" and then continually mixed it down another .2 to .3 dbs for like hours, before realizing that the out of tune guitars were making the vocal sound worse than it was, and also that the tuning issues were screwing with my sense of eq and level balance. So I went almost back to the beginning on a few things, (the vocals mostly I was trying to counterbalance the tuning changes with volume automation until I realized what was going on, bleh.)
So basically I wanted to warn everyone who does this track that it has significant tuning issues and timing issues that should be addressed before you start sending auxes and copy-pasting trim plug-ins. Its really rewarding if you get the rhythm section to marry each other though, I will post my results when I get there.
I'm posting this here because this is the highest thread in the section and I want anyone who's touching the tracks for the first time to have a chance to see this before they work themselves into a corner. Be warned!