Title: A Step-by-step Recording Walkthrough
Description: pics + mp3's - 56k, beware
Description: pics + mp3's - 56k, beware
Drew - April 7, 2006 03:40 AM (GMT)
Ok, part one... Basically, this is just raw tracking, or "getting everything you need recorded." I just did a quick level set, but that's been it so far - the real fun begins when we get into the mixing, and when we do it'll be cool to hear the difference between this mp3 and the final product after we've done everything to it. I'd keep going, but hopnestly it's 11 and I've got work in the morning, and don't feel like working on this until three in the morning. So...
While working on this (i wrote and recorded a simple little song idea from the ground up for you guys),I took detailed notes, took screenprints, and took pictures. Again, this is only complete through the tracking, but I'll continue on (probably mostly with screenprints, now that mic placement is basically behind us) as I work. Here's the notes I took, with their appropriate graphics:
5:45
Ok, I always start by programming a drum track, even if it's just something simple that I'm going to replace later. Drums are a lot more fun to track against than a click. So, to begin, I use the "fill every two steps" function in Fruity Loops to give myself a simple click, and then start jamming along to that with my riff, trying to "hear" a drum track, and dialing in the tempo. Key here is learning how to think like a drummer - it may be worth while to sit down and work out a few drum beats you already know if this is hard. "When the Levee Breaks" by Led Zep and "I Am One" by the Pumpkins are two I remember trying.
MP3: click track with kick and snare
I'll start with the kick and snare. I generally program two measures' worth, one with a slight variation, to give the drums a little more of a sense of direction. Then I'll delete my click and start adding hihats. Once I get something I can live with, I'll export the wave to the folder I'm saving the project in, then I'll export a wave of each kit component - kick, snare, hi-hats, toms if any - seperately. You can select or unselect indivutual tracks by clicking on the green light next to the knobs in fruity loops. I also saved an individual hit of two different snares, figuring I'll want to add some snare crashing when I hit the chorus (sorry, my roots are in Nirvana \m/)
MP3: more-or-less finished beat
That set, I'll close out of Fruity Loops (hopefully I won't need it again) and open Sonar to start tracking. For now, I'm just going to use the drum loop I created of everything in the kit together, to keep disc load down while tracking. So, I'll import that into Sonar and stretch it out for a bit longer than I expect to need it. Also, I'll set the project tempo to 65, the tempo I programed the beat at (though I suspect it's really 130, and I programmed it at half-time to get away with some accents FL would not have otherwise allowed). 169 measures should do it nicely.
I guess I need an arrangement before I can start tracking, so I'll run off to this place that does great burritos up the street while I think.
6:55
Back. I've got a few ideas (and I'm not looking to do anything uber-complex anyway) so I'm going to wing this. First, as a bookmark for myself, I'm going to drop a crash in everywhere there's supposed to be a changeup, as sort of a placekeeper for myself. So, track #2 becomes my crash cymbol.
7:10
Ok, 94-odd measures later, I have a simple arrangement. No key changes or anything, nothing absurd, just a verse-chorus-verse sort of thing with a solo thrown in the middle. "It doesn't have to be perfect, it's just for demonstration." Time to tune up and break out the AKG's for my acoustic.
7:57
After much farting around, and a false start when I had my levels too high and saw I was peaking out at -1.6 (still not clipping, but way higher than ideal - -6db or so is a good target to shoot for when tracking, as an intermediary between risking clipping and creating too much noise when you have to boost your signal later on in the mix), I get a "good enough for my purposes" acoustic guitar track in one take. I've got my guitar (Martin MC16-GTE, strung up with Elixer phospher bronze 12's) mic'd up in the classic "X-array" manner, when you stick two small diaphram condensors next to each other at about a 90 degree angle, pointing approximately at the 14th fret. This gives youa good stereo spread while almost completely elimitating the problem of phasing. Mics in question are a set of AKG C1000s's. Here, I'm actually pointing an inch or two above the 15th, but that's where the mics fell initially, and I liked the warmer, deeper tone I was getting there more than the brigher one at the "accepted" position. And, as they say, when in doubt trust your ears. This might come back to bite me when I start mixing properly, but whatever.
X-array on an acoustic:
Audio clip - this is what a stereo-mic'd acoustic sounds like. I took a clip out, muted one channel, muted the next, and then played them both together. It's subtle, but you can hear timbral differences between the two that combine to give you a richer, fuller, more expansive sound. There are other micing methods where this is more pronounced, but like I said the plus of the X array school of thought is with the mics that close together, phase cancellation is a non-issue, and in a busy mix, the slightly smaller sound this one gives you is actually a blessing.
MP3: Acoustic stereo guitars - solo'd
Normally I'd put a bass down next, but it's 8:01 as I'm writing this, and out of respect to my neighbors I'll get the electric rhythm tracks down next. Time to mic up - again, I'll be using a C1000s. Conventional wisdom is you don't use small diaphram condensors on guitar amps, but so far I've gotten good results on rhythm tracks with one, so screw it. I'll be using a dynamic for leads, anyway, and the differentiation in the mix will be nice.
Mic position - slightly off axis, slightly back, near the center of the speaker. I just jiggled it until I liked what I heard.
Slightly off-axis small diagram condensor, for rhythm parts:
8:23
Just listening to the last low B5 chord (hey, I'm a seven stringer, I had to end on one) feed back through the monitors on my final take. I tossed the first take for the first section of rhythm guitar one, but aside from that everything else here is first take. There are four parts here, comprosing two total tracks of rhythm guitar. As there's a lot of empty space between the distorted sections, I figured it made more sense to cut them individually, rather than standing there waiting for a minute or so during the solo section, and then going back and muting all the amp hiss and string noise and whatnot later on.
Guitar one, currently on the left, is Channel 3 Modern on my Nomad, set with the gain a touch over half up, presence at 2 o'clock, treble maybe 1:30, mids 9:30, and bass about 2 as well. Guitar two, on the right, has a bit more gain, the same presence settings, but treble at 11, mids at 2, and bass at 9. Neither sound phenominal alone (though the first isn't too bad), but together you get this massive rhythm guitar sound. Cool. Time for some bass. I'm a hack bassist, this could take a while.
Rhythm track 1 settings:
Rhythm Track 2 settings:
MP3: And a sound clip - double-tracked electrics with dissimilar tones
8:58
Listening back to my first complete bass take - second if you count the one where a minute or so in, again, I glanced at the meters and realized I was clipping. Turns out my compression ratio on my J-Station I'm recording through here was 4:1 - fine for a gutiar, way low for a bass. Upped it to 8:1, and the "slap" transients that were clipping out when everything else wass coming through at like -9 went away. So far, listening to this take, while it's nothing I'd release or anything, it's clean enough that I just may leave it. I even snuck an occasional fill in without sounding like too much of a dumbass.
MP3: Bass tone clip - solo'd
I set levels for a rough mix still now EQ or effects, but I brought the drums up +6.0, both sides of the acoustic stereo track back -4.5, both electrics back -6.5, and the new bass at -5.0. It still neads loads of work, but it at least sounds like it might be a song now.
Now, let's see if I can squeeze a clean guitar or two in there.
9:13
Ok, I've got a couple potential interlocking parts I'm going to try. Time to track.
9:48
Man, clean tones are a bitch. Eventually I ended up going with a very Edge-meets-Billy-Corgan riff with a lightly crunchy clean tone. Cheesy as hell, but whatever, it'll be fun to mix. Normally this is where I'd stop and then just jam over this for a while to work out a melody, but screw it, why stop here - I'm just going to do a take or two over the top and improvise something. Mic swith to my Audix i5, then time to screw with the mic placement and dial up a tone.
Clean settings:
MP3: Clean riff sound clip
10:03
So far the first take isn't hideous, although listening back to it, I'm going to have a FUN time getting this tone to poke through the chorus guitars - I should have spent more time on placement to get alittle mroe presence. Oh well, it'll make for a challanging example.
MP3: Lead guitar solo'd - clip
Something's missing though - I'm thinking some Ebow on the outro... Now, the "correct" way to use an Ebow is on the neck pickup with the volume down on the guitar, heldo over the pickups, but only dirty communists do everything the "correct" way. Throw that sucker on the bridge pickup in harmonic mode and hold it over the fretboard, and weird things happen. That's what I'm hearing. So....
10:18
Ok, one track turned into three improvised tracks of random harmonies, just to give this one a much needed touch of weirdness and exoticism in it's closing moments, between the sudden inclusion of the rumbling low B5 and the weird sustaining legato stuff. Listening to the lead, you can kinda tell I had this in the back of my mind at least from my improvised solo, probably earlier (ok, quite a bit earlier. I'm sorry, this is the stuff I dig about recording.)
Listening back, I'd say the rough tracking is done. I'm going to go post this up now, along with a few audio clips of various parts of the mix, and if I have time start mixing properlty - all I've done so far is record tracks and set levels. Getting the right performances down aside, that's the easy part. Making it sound like a CD is where things get tough.
MP3: Unmixed new song - a quick level set and some hard left/hard right panning, but no more
11:00
Ok, uploading files now... Man, this takes forever.
...and, 11:38 EST and I'm done typing this out. Time to get some sleep before work tomorrow, lol, and either tomorrow after work if I have time, or over the weekend I'll mix this into something that sounds like a song, and finish this.
While working on this (i wrote and recorded a simple little song idea from the ground up for you guys),I took detailed notes, took screenprints, and took pictures. Again, this is only complete through the tracking, but I'll continue on (probably mostly with screenprints, now that mic placement is basically behind us) as I work. Here's the notes I took, with their appropriate graphics:
5:45
Ok, I always start by programming a drum track, even if it's just something simple that I'm going to replace later. Drums are a lot more fun to track against than a click. So, to begin, I use the "fill every two steps" function in Fruity Loops to give myself a simple click, and then start jamming along to that with my riff, trying to "hear" a drum track, and dialing in the tempo. Key here is learning how to think like a drummer - it may be worth while to sit down and work out a few drum beats you already know if this is hard. "When the Levee Breaks" by Led Zep and "I Am One" by the Pumpkins are two I remember trying.
MP3: click track with kick and snare
I'll start with the kick and snare. I generally program two measures' worth, one with a slight variation, to give the drums a little more of a sense of direction. Then I'll delete my click and start adding hihats. Once I get something I can live with, I'll export the wave to the folder I'm saving the project in, then I'll export a wave of each kit component - kick, snare, hi-hats, toms if any - seperately. You can select or unselect indivutual tracks by clicking on the green light next to the knobs in fruity loops. I also saved an individual hit of two different snares, figuring I'll want to add some snare crashing when I hit the chorus (sorry, my roots are in Nirvana \m/)
MP3: more-or-less finished beat
That set, I'll close out of Fruity Loops (hopefully I won't need it again) and open Sonar to start tracking. For now, I'm just going to use the drum loop I created of everything in the kit together, to keep disc load down while tracking. So, I'll import that into Sonar and stretch it out for a bit longer than I expect to need it. Also, I'll set the project tempo to 65, the tempo I programed the beat at (though I suspect it's really 130, and I programmed it at half-time to get away with some accents FL would not have otherwise allowed). 169 measures should do it nicely.
I guess I need an arrangement before I can start tracking, so I'll run off to this place that does great burritos up the street while I think.
6:55
Back. I've got a few ideas (and I'm not looking to do anything uber-complex anyway) so I'm going to wing this. First, as a bookmark for myself, I'm going to drop a crash in everywhere there's supposed to be a changeup, as sort of a placekeeper for myself. So, track #2 becomes my crash cymbol.
7:10
Ok, 94-odd measures later, I have a simple arrangement. No key changes or anything, nothing absurd, just a verse-chorus-verse sort of thing with a solo thrown in the middle. "It doesn't have to be perfect, it's just for demonstration." Time to tune up and break out the AKG's for my acoustic.
7:57
After much farting around, and a false start when I had my levels too high and saw I was peaking out at -1.6 (still not clipping, but way higher than ideal - -6db or so is a good target to shoot for when tracking, as an intermediary between risking clipping and creating too much noise when you have to boost your signal later on in the mix), I get a "good enough for my purposes" acoustic guitar track in one take. I've got my guitar (Martin MC16-GTE, strung up with Elixer phospher bronze 12's) mic'd up in the classic "X-array" manner, when you stick two small diaphram condensors next to each other at about a 90 degree angle, pointing approximately at the 14th fret. This gives youa good stereo spread while almost completely elimitating the problem of phasing. Mics in question are a set of AKG C1000s's. Here, I'm actually pointing an inch or two above the 15th, but that's where the mics fell initially, and I liked the warmer, deeper tone I was getting there more than the brigher one at the "accepted" position. And, as they say, when in doubt trust your ears. This might come back to bite me when I start mixing properly, but whatever.
X-array on an acoustic:
Audio clip - this is what a stereo-mic'd acoustic sounds like. I took a clip out, muted one channel, muted the next, and then played them both together. It's subtle, but you can hear timbral differences between the two that combine to give you a richer, fuller, more expansive sound. There are other micing methods where this is more pronounced, but like I said the plus of the X array school of thought is with the mics that close together, phase cancellation is a non-issue, and in a busy mix, the slightly smaller sound this one gives you is actually a blessing.
MP3: Acoustic stereo guitars - solo'd
Normally I'd put a bass down next, but it's 8:01 as I'm writing this, and out of respect to my neighbors I'll get the electric rhythm tracks down next. Time to mic up - again, I'll be using a C1000s. Conventional wisdom is you don't use small diaphram condensors on guitar amps, but so far I've gotten good results on rhythm tracks with one, so screw it. I'll be using a dynamic for leads, anyway, and the differentiation in the mix will be nice.
Mic position - slightly off axis, slightly back, near the center of the speaker. I just jiggled it until I liked what I heard.
Slightly off-axis small diagram condensor, for rhythm parts:
8:23
Just listening to the last low B5 chord (hey, I'm a seven stringer, I had to end on one) feed back through the monitors on my final take. I tossed the first take for the first section of rhythm guitar one, but aside from that everything else here is first take. There are four parts here, comprosing two total tracks of rhythm guitar. As there's a lot of empty space between the distorted sections, I figured it made more sense to cut them individually, rather than standing there waiting for a minute or so during the solo section, and then going back and muting all the amp hiss and string noise and whatnot later on.
Guitar one, currently on the left, is Channel 3 Modern on my Nomad, set with the gain a touch over half up, presence at 2 o'clock, treble maybe 1:30, mids 9:30, and bass about 2 as well. Guitar two, on the right, has a bit more gain, the same presence settings, but treble at 11, mids at 2, and bass at 9. Neither sound phenominal alone (though the first isn't too bad), but together you get this massive rhythm guitar sound. Cool. Time for some bass. I'm a hack bassist, this could take a while.
Rhythm track 1 settings:
Rhythm Track 2 settings:
MP3: And a sound clip - double-tracked electrics with dissimilar tones
8:58
Listening back to my first complete bass take - second if you count the one where a minute or so in, again, I glanced at the meters and realized I was clipping. Turns out my compression ratio on my J-Station I'm recording through here was 4:1 - fine for a gutiar, way low for a bass. Upped it to 8:1, and the "slap" transients that were clipping out when everything else wass coming through at like -9 went away. So far, listening to this take, while it's nothing I'd release or anything, it's clean enough that I just may leave it. I even snuck an occasional fill in without sounding like too much of a dumbass.
MP3: Bass tone clip - solo'd
I set levels for a rough mix still now EQ or effects, but I brought the drums up +6.0, both sides of the acoustic stereo track back -4.5, both electrics back -6.5, and the new bass at -5.0. It still neads loads of work, but it at least sounds like it might be a song now.
Now, let's see if I can squeeze a clean guitar or two in there.
9:13
Ok, I've got a couple potential interlocking parts I'm going to try. Time to track.
9:48
Man, clean tones are a bitch. Eventually I ended up going with a very Edge-meets-Billy-Corgan riff with a lightly crunchy clean tone. Cheesy as hell, but whatever, it'll be fun to mix. Normally this is where I'd stop and then just jam over this for a while to work out a melody, but screw it, why stop here - I'm just going to do a take or two over the top and improvise something. Mic swith to my Audix i5, then time to screw with the mic placement and dial up a tone.
Clean settings:
MP3: Clean riff sound clip
10:03
So far the first take isn't hideous, although listening back to it, I'm going to have a FUN time getting this tone to poke through the chorus guitars - I should have spent more time on placement to get alittle mroe presence. Oh well, it'll make for a challanging example.
MP3: Lead guitar solo'd - clip
Something's missing though - I'm thinking some Ebow on the outro... Now, the "correct" way to use an Ebow is on the neck pickup with the volume down on the guitar, heldo over the pickups, but only dirty communists do everything the "correct" way. Throw that sucker on the bridge pickup in harmonic mode and hold it over the fretboard, and weird things happen. That's what I'm hearing. So....
10:18
Ok, one track turned into three improvised tracks of random harmonies, just to give this one a much needed touch of weirdness and exoticism in it's closing moments, between the sudden inclusion of the rumbling low B5 and the weird sustaining legato stuff. Listening to the lead, you can kinda tell I had this in the back of my mind at least from my improvised solo, probably earlier (ok, quite a bit earlier. I'm sorry, this is the stuff I dig about recording.)
Listening back, I'd say the rough tracking is done. I'm going to go post this up now, along with a few audio clips of various parts of the mix, and if I have time start mixing properlty - all I've done so far is record tracks and set levels. Getting the right performances down aside, that's the easy part. Making it sound like a CD is where things get tough.
MP3: Unmixed new song - a quick level set and some hard left/hard right panning, but no more
11:00
Ok, uploading files now... Man, this takes forever.
...and, 11:38 EST and I'm done typing this out. Time to get some sleep before work tomorrow, lol, and either tomorrow after work if I have time, or over the weekend I'll mix this into something that sounds like a song, and finish this.
Drew - April 7, 2006 03:16 PM (GMT)
Some more thoughts, in a moment of downtime at work:
distorted rhythm guitars
Really, it's kind of unintuitive, but a lot of the time lowering the gain will actually give you a "bigger" guitar sound. I mean, granted my Nomad's got a ton of gain on tap and the Blaze bridge pickup in use here has a good amount of output to it, but even the highest gain rhythm track here has the gain somewhere between 6 and 7 tops, and the brighter, more scooped one next to it actually has a bit less - maybe a touch over halfway up. Resist the impulse to crank the gain when you're looking for a big rhythm sound. Likewise, while one side of the mix is pretty scooped, the other is not - midrange is your friend when you're talking about guitars, and you need at least a moderate amount of it in there. I would never double two tracks of "scooped" rhythm, simply because without some mids your guitar is just going to get lost behind the bass and drums.
Mic Placement
I should have stopped and done a couple mp3's of different mic placements and the sort of tones you get from them. Tonight or tomorrow I'll make a point of doing this, and maybe compare the i5 against the C1000s too, just you you can have a reference. Really, the slightest nudge can make a huge difference. It takes a while to get this absolutely perfect, but it's well worth the time, and the more you do it, the better you'll get and the easier it'll be to dial in a good tone in the future.
Also, remember that your mic is "listening" to your amp from speaker level - you tend to dial in a tone that sounds good to you at ear level in the room, but when you're close-micing, you're actually capturing something completely different. Listen to your amp from right in front of it, tweak your EQ and gain as needed, and try to get THAT tone happening. You CAN do multi-mic setups where you combine a close-mic with a room mic for a bit of depth, but that's tough to do in a bedroom studio partly because of phase-cancellation issues (a problem anywhere), and partly because the room mic, since you're in the room too, also picks up the "acoustic" ping-y sound of your pick on the strings, which I don't like.
Any questions or comments before I move on and start mixing, tonight or tomorrow?
distorted rhythm guitars
Really, it's kind of unintuitive, but a lot of the time lowering the gain will actually give you a "bigger" guitar sound. I mean, granted my Nomad's got a ton of gain on tap and the Blaze bridge pickup in use here has a good amount of output to it, but even the highest gain rhythm track here has the gain somewhere between 6 and 7 tops, and the brighter, more scooped one next to it actually has a bit less - maybe a touch over halfway up. Resist the impulse to crank the gain when you're looking for a big rhythm sound. Likewise, while one side of the mix is pretty scooped, the other is not - midrange is your friend when you're talking about guitars, and you need at least a moderate amount of it in there. I would never double two tracks of "scooped" rhythm, simply because without some mids your guitar is just going to get lost behind the bass and drums.
Mic Placement
I should have stopped and done a couple mp3's of different mic placements and the sort of tones you get from them. Tonight or tomorrow I'll make a point of doing this, and maybe compare the i5 against the C1000s too, just you you can have a reference. Really, the slightest nudge can make a huge difference. It takes a while to get this absolutely perfect, but it's well worth the time, and the more you do it, the better you'll get and the easier it'll be to dial in a good tone in the future.
Also, remember that your mic is "listening" to your amp from speaker level - you tend to dial in a tone that sounds good to you at ear level in the room, but when you're close-micing, you're actually capturing something completely different. Listen to your amp from right in front of it, tweak your EQ and gain as needed, and try to get THAT tone happening. You CAN do multi-mic setups where you combine a close-mic with a room mic for a bit of depth, but that's tough to do in a bedroom studio partly because of phase-cancellation issues (a problem anywhere), and partly because the room mic, since you're in the room too, also picks up the "acoustic" ping-y sound of your pick on the strings, which I don't like.
Any questions or comments before I move on and start mixing, tonight or tomorrow?
bryan k - April 7, 2006 03:54 PM (GMT)
this is very cool.........keep it coming.
motionblur - April 7, 2006 06:34 PM (GMT)
Agreed. Very cool for beginners.
Drew - April 8, 2006 06:09 AM (GMT)
Day two....
7:52
Ok, I just finished cooking, eating, and cleaning up after dinner on a rainy, dreary Friday night. I don't really have the heart to go out, so instead I mixed myself a stiff gin and tonic, sat down in front of my computer, and pulled up the project from last night. Now the fun begins... First, though, an upgrade I've been meaning to make to my workstation for ever - what WOULD Devin Townsend do?
Listening back to the playback here... Promising, but we've got a lot to work on. The drums sound kind of low and weak (dry too, but that's a different story), and the lead sounds muffled by the instruments around it. I'm kinda digging that clean part though, although it's going to have to stay back in the mix, and I really like my acoustic and distorted tone on this.
So, first things' first, we're going to replace the original stereo drum mix with our individual component tracks we created yesterday. That done (and set up to the +6db that the original beat was at) and the original beat gone, before I do anything else I'm going to trim out all the high hats before the guitars come in and all but the last few kick and snare hits, to give us a little drum fill of sorts at the beginning. The slight gap between the kick loop at the beginning and the main loop is where I deleted a single kick, to give it a slight pause before the drums dropped back end - I like how it sounded more that way.
MP3: Simple drum fill by deleting beats
Next, in my "Master" section at the bottom of Sonar, I've created a drum bus. A "bus" in mixing terms is a number of outputs grouped together through a single set of controls, so to speak, and it makes it a lot easier to work with them as an entire unit than in peices. If your multitrack program doesn't offer bussing, no sweat, you can do everything I'm about to do by adjusting volumes and applying effects indivdually, but it's convenient to work with. So, now I'm going to change the output of all the drum tracks from "Master" to "drum bus," and then send the drum bus output into the master.
That done, I'm going back to liven up the crash cymbol hits a bit - alternating between the left and right, adding additional ones, ect. The left seems a bit quieter than the right, and is coming up a few dB lower,so I leave the right at +1.0 and put the left at +5.0.
8:40
Some copying, pasting, and dragging later (plus a phone call, I'm normally a little faster ;)), I've got a marginally more interesting drum track. Still nothing anyone would confuse for a decent real drummer, but a little easier to work with. Time to start EQ'ing, effecting, and doing all that other fun stuff.
First, though, before we really delve in, one thing that helps a lot when working on a mix is having a "reference mix" on hand. A reference mix is basically just another CD you really like the sound of - as you're working on a song, and getting fixated on the minutia, you tend to lose sight of the big picture. You'll be adding a reverb and listening to it, and tweaking it, andf listening some more, and your ears willg et used to the sound and demand a bit more of it. This is why you have a reference mix on hand - it gives you a fresh perspective so every once in a while you can step back and say "holy shit, there's WAY too much reverb on that snare drum!"without having to step away from it for a while. So, since I consider Devin Townsend about the greatest producer I've ever heard, I'm going to grab my copy of "Terria," and in particular listen to a track called "Nobody Here" that also has a lot of acoustic, clean, and distorted electric guitars in it, at a similarly slow tempo. Sure it's not an instrumental, and on a Satriani forum I should probably use a track of Joe's, but the only acoustic-driven track that's coming to mind is "Starry Night," and I hate the mix on SBM. Besides, WWDTD?
We'll start by seeing where we stand in comparison to Devin's mix. First and foremost, his is way clearer - that's what we're shooting for. Drums are pretty quiet, but in general they're mixed back on Terria - I might go a little higher here. The lead guitar, again, is the biggest issue, as it sounds pretty muffled but is already mixed fairly high. Aside from that, the levels are basically ok - we're at a point where we can start drilling down and working on effects and EQ curves.
Let's start with the drums. We've got a nice "live" sounding drum tone right now, but the problem with that is it takes up rather a lot of room in the mix. We can leave it as is and sacrifice elsewhere, but I'm a guitarist, not a drummer, and I'd rather have nice lush guitar tones and a tight drum sound than vice versa. So, let's start with the kick.
First, I've put a 6:1 compressor on the kick with a threshold of -20db (eyeballing the waveform, most of the attack seems to be below -12dB even though the track's peaking at about -6, so this isn't as low as it might seem at a glance) with a 35ms attack and a fairly fast (149ms) release. The idea here is I'm trying to clamp down on the kick after the initial attack, to preserve a lot of the sharp transient, while supressing some of the rumble that follows. This'll leave me some room to boost the bass. Then, I used the internal track EQ in Sonar to tighten it up a bit more, with a -5db low shelf filter centered around 170hz, with a Q (sort of a "width" over which the adjustment takes effect) of .8 of an octave. This gives me a fairly tight kick sound. I may do more later, but let's move on to the snare for now.
For the snare, I eventually decided on an -8db low shelf filter at 560hz with a Q of 1.2, and a 5.5:1 compression at -17.5with an attack of 8 and a slightly longer release of 310. Listening to this, though, the snare's now a little too quiet, now that I've cut out some of it's low end so it's brighter and sharper, so eventually I'm going to boost it a bot. It's already up the maximum of 6db on the track, so I could either boost it in an audio editor or just wait until I'm done with EQ'ing the drums, and either cut the rest of the drums and boost the entire mix in the bus, or cut all the rest of the tracks (or both). For now, let's tighten up the hi-hats and crashes.
I applied a -6dB low shelf at 2.7k with a Q of 1.0, and a light 3:1 compression at -25dB with an attack of 10 and a release of 250 - really, this isn't much of an audible change, but that's good - we still have a fairly "live" cymbol sound, but now it takes up less space in the mix.
Next, I did what I'm pretty sure is an old LA studio trick, where you sent a FX out from your drum submix into a compressor, compress the hell out of the signal, and mix it a bit behind oyur unaffected drum mix. This gives the drums a lot more perceived body without robbing them of impact. For those of you not using busses, make a submix of your drums, mix it down to a new track, and open it and compress it there for the same effect. For now, I've got a 16:1 compression at -15, with the FX send level at -4, and the FX bus's output set -7dB below that of the drum bus. Now the drums actually sound a bit louder than they did, so I've kicked the drum bus back to -2.5db. This in turn means the compressor isn't being driven as hard, so I increased the send to -2.5dB. This means there's actually a bit more of the compressed drums in the mix, but I like what I'm hearing, so I'll live with that.
MP3: Drums - EQ and compression (compare to the "beat" mp3 from the previous day)
10:20
So, I've now got a basic drum sound I'mprepared to go forward. However, it's still totally dry, and while there's some natural room ambience on the samples I'm using, it's going to sound a bit more like a drum sound you're used to after we add some reverb. So, I'm creating a new FX bus, "drum verb" that we're going to route the individual components of the kit to. We could just stick a second FX send in the drum bus, but this way we can adjust the amount of reverb on each peice of the kit, which is a plus. After adding an FX send from each drum track to the new 'verb track, I then call up a reverb plugin and sort of guestimate send levels based on the amount of 'verb I want on each component (lots on the crashes and hi-hats, quite a bit on the snare, less on the kick), then pull up a fairly warm, active reverb and set the mix to 100 and then the FX bus level to taste - I end up at about -8.7, though I won't commit to this until quite a bit later in the mix, and I suspect it might be slightly too much. I check against "Nobody's Here," and my drums are definitely significantly wetter than Devin's, though his are admittedly very dry on this album, and I'm working with a sparser mix. I kick the verb back to 10.5 anyway, and it seems like a good compromise.
10:40
Ok, we'll accept the drums for now and move on to another instrument. Logically, since the drums are more-or-less mixed, it makes sense to take a look at the bass guitar. Bass is always a battle for me as I never quite know what to do with it. As such, usually I cut kind of a lot out. So, I'll start by soloing the bass track and the drums, and then opening the track EQ and sweeping a narrow notch cut back and forth in case anything strikes my ear. 280-ish seems to work well with the kick, and when I un-mute everything it seems to fit in pretty well with the rest of the instruments as they are now, so I Cut that down pretty steeply in a narrow band. Then, since harmonics of freqency ranges tend to repeat at mathematic intervals, I try scooping a narrow band similarly at twice and four times that frequency. Again, so far so good. Finally, I curb in the highs a bit, since you really can't hear them within a mix on a bass, and roll off the super lows, below 44hz. While playing with Q's for that low shelf filter, I notice that if you set it past a certain point it actually boosts the frequencies slightly above it a little, and it seems to help the bass carve out harmonic space of it's own, so I run with that.
MP3: Bass guitar - EQ'd (compare to the "bass solo'd" mp3 from the previous day)
11:05
Let's move on to the acoustic guitars - these are one of the more full-frequency instruments in the mix, but on one hand we've got tentative territory staked out for a good chunk of other instruments right now, and on the other that just means they're one of the instruments where we can cut a lot out while still leaving a lot. First, I set up another bus for the acosutics, and route them into it. Listening to them solo'd, they're really pretty "live" sounding as they are, so I have a feeling I'm going to leave them dry. A piezo acoustic will almost certianly need reverb to sound "natural," but when you're using a mic, a lot of the time an added reverb just won't be needed. Again, no promises, and in a half hour I might decide to add one, but for now I like what I'm hearing.
A few minutes of experimentation, and rolling off the bass by about five and a half decibels below just under 400 helps clear up some of that boomy low end I was worried about while tracking this, and a fairly narrow mid scoop at 1375 seems to help things "open up" somehow. A little bit of a high end rollof seems to be be working, too, just under -9 at around 17k.
11:25
The electrics are pretty straightforward. Again, give 'em their own bus, and apply an EQ to them while they're there. I doa gradual rolloff below 129, a high pass filter set at just about 70, a very slight mid scoop at 1163 (this will help the lead cut, though I may move it up or down a bit once I start working on the lead), and roll off most of the high end above 7.3khz. Through my speakers, this is subtle, but through my phones, there's a huge decrease in "hiss" in the distortion when I mute the track, and it frees up a lot of room for the cymbols.
MP3: Electric guitars, EQ'd (compare to the "electric solo'd" mp3 from the previous day)
11:50
A few minutes to screw with the clean guitar. Basically, I destroyed the thing. First, I ran it through the track EQ, chopping 6db off everything below 8.3khz, then running a low-pass filter at about 11.8khz. Next, I sent it through a light chorus - I'm not a big chorus fan, but the one thing it does well is rob a tone of all "body" within a mix. This is why I DON'T like it, but there's a time and a place for that, namely when you've got a guitar part you really want to sit back and not draw attention to itself. That's what we're doing here. Then, after that, I ran the track through a tempo delay, with half notes in one channel and quarter notes in the next, with the feedback pretty high so they really trail off. This on top of the chorus gives us a very spacious, open clean part that, because we've basically cut all of it out of the mix and then run it throug a chorus anyway, doesn't take up very much space. Cool.
If you listen to just the intro, it's finally beginning to sound like a real song. However, the lead still sounds like crap. So, let's work on that one now. It's better than it was, since we've cut a lot out of the backing already, but it still sounds muffled.
MP3: Intro, now that it's mostly mixed
12:15
After trying all kinds of crazy notching and filtering schemes, I'm kind of leaning towards a simple high pass filter at 320hz with a Q of 1. It seems to help tighten things up in the distorted section, while still letting the tone sound warm elsewhere. It's still not perfect, but it's not cutting too badly, so I'll take that for now and move on to ambient effects on the lead.
I tracked the part enirely dry, with no delay or reverb, figuring I'd add it in the mix. As I'm a delay junky, let's start there. I'm sending this one to a seperate bus so I can filter it further.
12:25
A bit of experimentation later, and I've got a stereo multi-tap delay up and running, using the delay times from the tempo delay I already have open, with a low pass at 1.04khz applied to the delay at the bus, and a high pass at about 7k on the delay plugin itself. Again, I'm using 100% delay and just adjusting the send amount to taste. Again, I'll want to revisit this eventually, but for now I'll leave it and move on because I just remembered I had those ebow tracks at the end, and tose are positively begging for some high-feedback delay and creative panning.
12:49
While tring to figure out what to do with the Ebow tracks after I panned them around randomly, I remembered I could access Fruity Loops' plugins via VST, and that there was some pretty warped stuff in there. So, I randomly grabbed a "free filter" which I'm not really sure what it does, exactly (some sort of EQ), but the default settings gave the tracks a nice lo-fi vibe. I fed that into the fruity loops delay, and that into Sonar's normal chorus, then finally into Fruity Loops' "Blood Overdrive" to give it a little more grit. The upshot is, after a fairly pretty, straightforward song, you get 15 seconds of just bizareness right at the end. Cool.
Time for some final muting and cleaning.
1:00
I cleaned up the loose string noise before the electrics came in, a little bit of string noise before the bass came in (you can here both on the solo clips), and deleted everyhting in the acoustic tracks before I actually started playing, so there's dead silence until (and during, everywhere else) that drum fill that opens this one off. I also added volume envelopes to the acoustic guitars, to drop them from -4.5 to -7.5 dB as the electrics came in, to keep the mix more even at that point.
This is at the stage where I'd call it a rough mix - now I'd let it sit for another day or so and then come back to it with a fresh set of ears and see if anything sounds weird a day or so down the road, but the basic picture is there.
MP3: More-or-less final Rough Mix
Comments? Questions? There's rather a lot of information here, I'm sure there's something I could explain better.
7:52
Ok, I just finished cooking, eating, and cleaning up after dinner on a rainy, dreary Friday night. I don't really have the heart to go out, so instead I mixed myself a stiff gin and tonic, sat down in front of my computer, and pulled up the project from last night. Now the fun begins... First, though, an upgrade I've been meaning to make to my workstation for ever - what WOULD Devin Townsend do?
Listening back to the playback here... Promising, but we've got a lot to work on. The drums sound kind of low and weak (dry too, but that's a different story), and the lead sounds muffled by the instruments around it. I'm kinda digging that clean part though, although it's going to have to stay back in the mix, and I really like my acoustic and distorted tone on this.
So, first things' first, we're going to replace the original stereo drum mix with our individual component tracks we created yesterday. That done (and set up to the +6db that the original beat was at) and the original beat gone, before I do anything else I'm going to trim out all the high hats before the guitars come in and all but the last few kick and snare hits, to give us a little drum fill of sorts at the beginning. The slight gap between the kick loop at the beginning and the main loop is where I deleted a single kick, to give it a slight pause before the drums dropped back end - I like how it sounded more that way.
MP3: Simple drum fill by deleting beats
Next, in my "Master" section at the bottom of Sonar, I've created a drum bus. A "bus" in mixing terms is a number of outputs grouped together through a single set of controls, so to speak, and it makes it a lot easier to work with them as an entire unit than in peices. If your multitrack program doesn't offer bussing, no sweat, you can do everything I'm about to do by adjusting volumes and applying effects indivdually, but it's convenient to work with. So, now I'm going to change the output of all the drum tracks from "Master" to "drum bus," and then send the drum bus output into the master.
That done, I'm going back to liven up the crash cymbol hits a bit - alternating between the left and right, adding additional ones, ect. The left seems a bit quieter than the right, and is coming up a few dB lower,so I leave the right at +1.0 and put the left at +5.0.
8:40
Some copying, pasting, and dragging later (plus a phone call, I'm normally a little faster ;)), I've got a marginally more interesting drum track. Still nothing anyone would confuse for a decent real drummer, but a little easier to work with. Time to start EQ'ing, effecting, and doing all that other fun stuff.
First, though, before we really delve in, one thing that helps a lot when working on a mix is having a "reference mix" on hand. A reference mix is basically just another CD you really like the sound of - as you're working on a song, and getting fixated on the minutia, you tend to lose sight of the big picture. You'll be adding a reverb and listening to it, and tweaking it, andf listening some more, and your ears willg et used to the sound and demand a bit more of it. This is why you have a reference mix on hand - it gives you a fresh perspective so every once in a while you can step back and say "holy shit, there's WAY too much reverb on that snare drum!"without having to step away from it for a while. So, since I consider Devin Townsend about the greatest producer I've ever heard, I'm going to grab my copy of "Terria," and in particular listen to a track called "Nobody Here" that also has a lot of acoustic, clean, and distorted electric guitars in it, at a similarly slow tempo. Sure it's not an instrumental, and on a Satriani forum I should probably use a track of Joe's, but the only acoustic-driven track that's coming to mind is "Starry Night," and I hate the mix on SBM. Besides, WWDTD?
We'll start by seeing where we stand in comparison to Devin's mix. First and foremost, his is way clearer - that's what we're shooting for. Drums are pretty quiet, but in general they're mixed back on Terria - I might go a little higher here. The lead guitar, again, is the biggest issue, as it sounds pretty muffled but is already mixed fairly high. Aside from that, the levels are basically ok - we're at a point where we can start drilling down and working on effects and EQ curves.
Let's start with the drums. We've got a nice "live" sounding drum tone right now, but the problem with that is it takes up rather a lot of room in the mix. We can leave it as is and sacrifice elsewhere, but I'm a guitarist, not a drummer, and I'd rather have nice lush guitar tones and a tight drum sound than vice versa. So, let's start with the kick.
First, I've put a 6:1 compressor on the kick with a threshold of -20db (eyeballing the waveform, most of the attack seems to be below -12dB even though the track's peaking at about -6, so this isn't as low as it might seem at a glance) with a 35ms attack and a fairly fast (149ms) release. The idea here is I'm trying to clamp down on the kick after the initial attack, to preserve a lot of the sharp transient, while supressing some of the rumble that follows. This'll leave me some room to boost the bass. Then, I used the internal track EQ in Sonar to tighten it up a bit more, with a -5db low shelf filter centered around 170hz, with a Q (sort of a "width" over which the adjustment takes effect) of .8 of an octave. This gives me a fairly tight kick sound. I may do more later, but let's move on to the snare for now.
For the snare, I eventually decided on an -8db low shelf filter at 560hz with a Q of 1.2, and a 5.5:1 compression at -17.5with an attack of 8 and a slightly longer release of 310. Listening to this, though, the snare's now a little too quiet, now that I've cut out some of it's low end so it's brighter and sharper, so eventually I'm going to boost it a bot. It's already up the maximum of 6db on the track, so I could either boost it in an audio editor or just wait until I'm done with EQ'ing the drums, and either cut the rest of the drums and boost the entire mix in the bus, or cut all the rest of the tracks (or both). For now, let's tighten up the hi-hats and crashes.
I applied a -6dB low shelf at 2.7k with a Q of 1.0, and a light 3:1 compression at -25dB with an attack of 10 and a release of 250 - really, this isn't much of an audible change, but that's good - we still have a fairly "live" cymbol sound, but now it takes up less space in the mix.
Next, I did what I'm pretty sure is an old LA studio trick, where you sent a FX out from your drum submix into a compressor, compress the hell out of the signal, and mix it a bit behind oyur unaffected drum mix. This gives the drums a lot more perceived body without robbing them of impact. For those of you not using busses, make a submix of your drums, mix it down to a new track, and open it and compress it there for the same effect. For now, I've got a 16:1 compression at -15, with the FX send level at -4, and the FX bus's output set -7dB below that of the drum bus. Now the drums actually sound a bit louder than they did, so I've kicked the drum bus back to -2.5db. This in turn means the compressor isn't being driven as hard, so I increased the send to -2.5dB. This means there's actually a bit more of the compressed drums in the mix, but I like what I'm hearing, so I'll live with that.
MP3: Drums - EQ and compression (compare to the "beat" mp3 from the previous day)
10:20
So, I've now got a basic drum sound I'mprepared to go forward. However, it's still totally dry, and while there's some natural room ambience on the samples I'm using, it's going to sound a bit more like a drum sound you're used to after we add some reverb. So, I'm creating a new FX bus, "drum verb" that we're going to route the individual components of the kit to. We could just stick a second FX send in the drum bus, but this way we can adjust the amount of reverb on each peice of the kit, which is a plus. After adding an FX send from each drum track to the new 'verb track, I then call up a reverb plugin and sort of guestimate send levels based on the amount of 'verb I want on each component (lots on the crashes and hi-hats, quite a bit on the snare, less on the kick), then pull up a fairly warm, active reverb and set the mix to 100 and then the FX bus level to taste - I end up at about -8.7, though I won't commit to this until quite a bit later in the mix, and I suspect it might be slightly too much. I check against "Nobody's Here," and my drums are definitely significantly wetter than Devin's, though his are admittedly very dry on this album, and I'm working with a sparser mix. I kick the verb back to 10.5 anyway, and it seems like a good compromise.
10:40
Ok, we'll accept the drums for now and move on to another instrument. Logically, since the drums are more-or-less mixed, it makes sense to take a look at the bass guitar. Bass is always a battle for me as I never quite know what to do with it. As such, usually I cut kind of a lot out. So, I'll start by soloing the bass track and the drums, and then opening the track EQ and sweeping a narrow notch cut back and forth in case anything strikes my ear. 280-ish seems to work well with the kick, and when I un-mute everything it seems to fit in pretty well with the rest of the instruments as they are now, so I Cut that down pretty steeply in a narrow band. Then, since harmonics of freqency ranges tend to repeat at mathematic intervals, I try scooping a narrow band similarly at twice and four times that frequency. Again, so far so good. Finally, I curb in the highs a bit, since you really can't hear them within a mix on a bass, and roll off the super lows, below 44hz. While playing with Q's for that low shelf filter, I notice that if you set it past a certain point it actually boosts the frequencies slightly above it a little, and it seems to help the bass carve out harmonic space of it's own, so I run with that.
MP3: Bass guitar - EQ'd (compare to the "bass solo'd" mp3 from the previous day)
11:05
Let's move on to the acoustic guitars - these are one of the more full-frequency instruments in the mix, but on one hand we've got tentative territory staked out for a good chunk of other instruments right now, and on the other that just means they're one of the instruments where we can cut a lot out while still leaving a lot. First, I set up another bus for the acosutics, and route them into it. Listening to them solo'd, they're really pretty "live" sounding as they are, so I have a feeling I'm going to leave them dry. A piezo acoustic will almost certianly need reverb to sound "natural," but when you're using a mic, a lot of the time an added reverb just won't be needed. Again, no promises, and in a half hour I might decide to add one, but for now I like what I'm hearing.
A few minutes of experimentation, and rolling off the bass by about five and a half decibels below just under 400 helps clear up some of that boomy low end I was worried about while tracking this, and a fairly narrow mid scoop at 1375 seems to help things "open up" somehow. A little bit of a high end rollof seems to be be working, too, just under -9 at around 17k.
11:25
The electrics are pretty straightforward. Again, give 'em their own bus, and apply an EQ to them while they're there. I doa gradual rolloff below 129, a high pass filter set at just about 70, a very slight mid scoop at 1163 (this will help the lead cut, though I may move it up or down a bit once I start working on the lead), and roll off most of the high end above 7.3khz. Through my speakers, this is subtle, but through my phones, there's a huge decrease in "hiss" in the distortion when I mute the track, and it frees up a lot of room for the cymbols.
MP3: Electric guitars, EQ'd (compare to the "electric solo'd" mp3 from the previous day)
11:50
A few minutes to screw with the clean guitar. Basically, I destroyed the thing. First, I ran it through the track EQ, chopping 6db off everything below 8.3khz, then running a low-pass filter at about 11.8khz. Next, I sent it through a light chorus - I'm not a big chorus fan, but the one thing it does well is rob a tone of all "body" within a mix. This is why I DON'T like it, but there's a time and a place for that, namely when you've got a guitar part you really want to sit back and not draw attention to itself. That's what we're doing here. Then, after that, I ran the track through a tempo delay, with half notes in one channel and quarter notes in the next, with the feedback pretty high so they really trail off. This on top of the chorus gives us a very spacious, open clean part that, because we've basically cut all of it out of the mix and then run it throug a chorus anyway, doesn't take up very much space. Cool.
If you listen to just the intro, it's finally beginning to sound like a real song. However, the lead still sounds like crap. So, let's work on that one now. It's better than it was, since we've cut a lot out of the backing already, but it still sounds muffled.
MP3: Intro, now that it's mostly mixed
12:15
After trying all kinds of crazy notching and filtering schemes, I'm kind of leaning towards a simple high pass filter at 320hz with a Q of 1. It seems to help tighten things up in the distorted section, while still letting the tone sound warm elsewhere. It's still not perfect, but it's not cutting too badly, so I'll take that for now and move on to ambient effects on the lead.
I tracked the part enirely dry, with no delay or reverb, figuring I'd add it in the mix. As I'm a delay junky, let's start there. I'm sending this one to a seperate bus so I can filter it further.
12:25
A bit of experimentation later, and I've got a stereo multi-tap delay up and running, using the delay times from the tempo delay I already have open, with a low pass at 1.04khz applied to the delay at the bus, and a high pass at about 7k on the delay plugin itself. Again, I'm using 100% delay and just adjusting the send amount to taste. Again, I'll want to revisit this eventually, but for now I'll leave it and move on because I just remembered I had those ebow tracks at the end, and tose are positively begging for some high-feedback delay and creative panning.
12:49
While tring to figure out what to do with the Ebow tracks after I panned them around randomly, I remembered I could access Fruity Loops' plugins via VST, and that there was some pretty warped stuff in there. So, I randomly grabbed a "free filter" which I'm not really sure what it does, exactly (some sort of EQ), but the default settings gave the tracks a nice lo-fi vibe. I fed that into the fruity loops delay, and that into Sonar's normal chorus, then finally into Fruity Loops' "Blood Overdrive" to give it a little more grit. The upshot is, after a fairly pretty, straightforward song, you get 15 seconds of just bizareness right at the end. Cool.
Time for some final muting and cleaning.
1:00
I cleaned up the loose string noise before the electrics came in, a little bit of string noise before the bass came in (you can here both on the solo clips), and deleted everyhting in the acoustic tracks before I actually started playing, so there's dead silence until (and during, everywhere else) that drum fill that opens this one off. I also added volume envelopes to the acoustic guitars, to drop them from -4.5 to -7.5 dB as the electrics came in, to keep the mix more even at that point.
This is at the stage where I'd call it a rough mix - now I'd let it sit for another day or so and then come back to it with a fresh set of ears and see if anything sounds weird a day or so down the road, but the basic picture is there.
MP3: More-or-less final Rough Mix
Comments? Questions? There's rather a lot of information here, I'm sure there's something I could explain better.
Rocklegend - April 8, 2006 10:43 AM (GMT)
Excellent article....Well done
How did ou get the Hi-Hats to sound quite real...in the feeling of the playing...it's rather like a human playing them.I notived that in Fruity Loops that you had 3 Hi-hat tracks...how come....
Could you do an article on Mastering....
How did ou get the Hi-Hats to sound quite real...in the feeling of the playing...it's rather like a human playing them.I notived that in Fruity Loops that you had 3 Hi-hat tracks...how come....
Could you do an article on Mastering....
Drew - April 8, 2006 03:56 PM (GMT)
I really don't know that much about mastering - previously I'd just normalize the wave file to 0db, then run it through Sonic Foundry's "Wave Hammer" plugin, which is a compressor/volume maximizer. I'd add a light compression, say 2:1 at like -3db, to take the edge off some of the peaks, and then use the mazimizer side to boost the volume by about 5-6dB - this would get me up to CD level without really having any perceptable impact on the dynamics of the wave. I'm fuzzy as to how it works, but, well, it did, so...
The thing with plugins like this is it's VERY easy to overdo it. Sure, you can boost 8-9dB before you really hear much distortion and clipping, but the dynamics of your track begin to sound squashed before you hear clipping, and that's the aural equivalent of getting hit across the eardrum with a sledgehammer over and over again - you get what's called "ear fatigue" when your dynamics are always hard on, and there's no room for the music to breathe. Your ear gets sick of it after a while - if you compare a CD that was mastered REALLY hot to one that was more moderate, the really hot one is generally harder to listen to for extended periods of time.
Of course, if you send a mix off to a mastering house, volume maximization is only one of the mny things they do to it. For home recording to share some mp3's with friends on the net, this is enough, but if you're releasing a professional CD, it's worth spending the money to have a good professional masterer give your music that extra something.
For the record, this "rough mix" mp3 is unmastered.
The hi-hats... I used, um... looks like 5 hats made the finished track, but I had a lot more open. I've got three types of hats, each with three samples loaded - "closed" hi-hats, "open" hi-hats, and "muted" hi-hats, where they're half closed. Then, each of the hats is at three velocities,a hard hit, a medium hit, and a soft hit.
Now the problem with fruity loops is that while it's really a pretty cool sequencer, if you use the same sample several times in a row, well, it SOUNDS like the same sample being repeated, and not like a real cymbol being hit a couple times. Varying it up like this gives you 1.) slightly different attacks, as if the drums are being hit by a stick in slightly different manners, much like a real drummer, 2.) slightly different volumes, so you can have some hi-hat hits with stronger accents than others. This combines to give you a more life-like drum performance. IF you'll notice, I'm also doing this with the kick and the drums, and this is part of the reason I use two crash cymbols (although, they're different sounding cymbols and panned to two different spots, so it's really more that there's at least two crashes on your average drum kit).
I'm using samples from the Drumkit from Hell here, but Fruyity Loops' "Real Drumkit" is actually quite good with a little bit of reverb on it, and while there aren't as many options in the default template as I've loaded here, if you dig around in the "packs" folder and fid the "realistic" kit folder, there's actually seven or eight hi-hats, three kicks, seven or eight snares... So even with Fruity Loops stock with no added samples, you can do a lot.
By the way, if you will be saving templates with a lot of samples in them... see that black box in the lower left hand corner that says "all?" That's your filter, and it's a great time-saver - if you want to create a grouping of all the hihats in a template, select the big green light on the right of your sample, click and drag down to get all of the peices you want to group together, then go to "Channels," and "Group Selected." enter a filter name, and then it'll greate a filter where you can just look at that component of your kit. It makes it easier once you've got the basic pattern down, if you want to just tweak your hi-hats without any distractions or scrolling or anything.
The thing with plugins like this is it's VERY easy to overdo it. Sure, you can boost 8-9dB before you really hear much distortion and clipping, but the dynamics of your track begin to sound squashed before you hear clipping, and that's the aural equivalent of getting hit across the eardrum with a sledgehammer over and over again - you get what's called "ear fatigue" when your dynamics are always hard on, and there's no room for the music to breathe. Your ear gets sick of it after a while - if you compare a CD that was mastered REALLY hot to one that was more moderate, the really hot one is generally harder to listen to for extended periods of time.
Of course, if you send a mix off to a mastering house, volume maximization is only one of the mny things they do to it. For home recording to share some mp3's with friends on the net, this is enough, but if you're releasing a professional CD, it's worth spending the money to have a good professional masterer give your music that extra something.
For the record, this "rough mix" mp3 is unmastered.
The hi-hats... I used, um... looks like 5 hats made the finished track, but I had a lot more open. I've got three types of hats, each with three samples loaded - "closed" hi-hats, "open" hi-hats, and "muted" hi-hats, where they're half closed. Then, each of the hats is at three velocities,a hard hit, a medium hit, and a soft hit.
Now the problem with fruity loops is that while it's really a pretty cool sequencer, if you use the same sample several times in a row, well, it SOUNDS like the same sample being repeated, and not like a real cymbol being hit a couple times. Varying it up like this gives you 1.) slightly different attacks, as if the drums are being hit by a stick in slightly different manners, much like a real drummer, 2.) slightly different volumes, so you can have some hi-hat hits with stronger accents than others. This combines to give you a more life-like drum performance. IF you'll notice, I'm also doing this with the kick and the drums, and this is part of the reason I use two crash cymbols (although, they're different sounding cymbols and panned to two different spots, so it's really more that there's at least two crashes on your average drum kit).
I'm using samples from the Drumkit from Hell here, but Fruyity Loops' "Real Drumkit" is actually quite good with a little bit of reverb on it, and while there aren't as many options in the default template as I've loaded here, if you dig around in the "packs" folder and fid the "realistic" kit folder, there's actually seven or eight hi-hats, three kicks, seven or eight snares... So even with Fruity Loops stock with no added samples, you can do a lot.
By the way, if you will be saving templates with a lot of samples in them... see that black box in the lower left hand corner that says "all?" That's your filter, and it's a great time-saver - if you want to create a grouping of all the hihats in a template, select the big green light on the right of your sample, click and drag down to get all of the peices you want to group together, then go to "Channels," and "Group Selected." enter a filter name, and then it'll greate a filter where you can just look at that component of your kit. It makes it easier once you've got the basic pattern down, if you want to just tweak your hi-hats without any distractions or scrolling or anything.
Drew - April 8, 2006 06:16 PM (GMT)
Update - I wanted to run this past Eric before I posted this for bvious reasons, but I just got the go ahead, so here's a 51-second clip from "Nobody's Here," the tail end of the second verse through the second chorus. This is what I was comparing my mix against while working, whenevr I wanted to put things in perspective.
Listening back this morning, I can still here some things I think could be better, so there might bea 3rd installment coming.
Anyway
MP3: "Nobody's Here" clip (DevinTownsend) used as reference for the mix above.
Now, go buy this album, Devin's work is absolute genius.
Listening back this morning, I can still here some things I think could be better, so there might bea 3rd installment coming.
Anyway
MP3: "Nobody's Here" clip (DevinTownsend) used as reference for the mix above.
Now, go buy this album, Devin's work is absolute genius.
Criss - April 9, 2006 12:58 AM (GMT)
This needs to be stickied !!
Thanks a bunch Drew :D you deserve a medal...or 5 ^_^
Thanks a bunch Drew :D you deserve a medal...or 5 ^_^
Mapu - April 9, 2006 05:02 PM (GMT)
yea dude, really a hard work to explain and give every details on recordin, man, congrats, you are awesome.
Besides, what a beatiful piece of music you wrote, i love it
Besides, what a beatiful piece of music you wrote, i love it
Axeshredder - April 10, 2006 02:15 AM (GMT)
Nice Job Drew...Topic Pinned :rock2:
Drew - April 10, 2006 11:02 AM (GMT)
Thanks guys! I think there's going to be a part 3 to this, as I'm still not totally happy with the mix... Anyway, I just hope some of this is helpful. :)
Oh, and I should really resize some of those graphics, hey? :lol: I forgot my screen definition was set to be more-or-less massive when I did this. :lol:
Oh, and I should really resize some of those graphics, hey? :lol: I forgot my screen definition was set to be more-or-less massive when I did this. :lol:
Eric - April 10, 2006 01:01 PM (GMT)
Thanks A LOT for posting this kind of valuable info on the forum... Great tutorial!
RAIN - April 10, 2006 03:11 PM (GMT)
O.K. Drew, thanks so much. I do have a question about using effects sends that you can help me with I hope.
Let's say I want delay on a lead track (or several of them) and I use a send to an effect bus. Using Sonitus Delay on the bus, would I run the MIX level in Sonitus Delay to 100% and then use a post send on the track to run the amount of delay?
A - When I increase the send to the bus I get more volume, so I imagine it is correct to alter the track main volume to compensate for this?
B - What if my lead track is panned 50% to the right. How do I get the delay to match that panning?
Using the effects sends is a whole new world to me.
Let's say I want delay on a lead track (or several of them) and I use a send to an effect bus. Using Sonitus Delay on the bus, would I run the MIX level in Sonitus Delay to 100% and then use a post send on the track to run the amount of delay?
A - When I increase the send to the bus I get more volume, so I imagine it is correct to alter the track main volume to compensate for this?
B - What if my lead track is panned 50% to the right. How do I get the delay to match that panning?
Using the effects sends is a whole new world to me.
Drew - April 10, 2006 05:55 PM (GMT)
Rain,
When using a FX send, you're probably best off setting the FX mix to 100% wet, so the bus is playing just the wet signal with no dry signal. Then, increase the send until the delay is as loud as you want it to be in the mix - this way you can adjust the delay volume without changing or affecting the volume of the original signal. I'm afraid I didn't screenprint the delay settings for the lead gitar in the mix, where I was running it through a bus - the delay on the clean rhythm guitar is set with a mix of 34% because that was being applied directly to the track, and not a bus, so that one's not a great example.
For panning, there'sa couple ways to go about it - you can either set the delay's stereo position in the plugin - in your instance, you'd be doing a 100% wet delay panned 50% to the right (though, this delay plugin I was using doesn't support that, you just get a left volume and right volume. The one I used on the lead guitar lets you specify the volume and pan for each of the repeats). Or, you can adjust the pan of the bus itself. If you look at the last screenprint there, you can see that there's a send and a return pan control on the "Lead Delay" bus. I didn't go this route because I was using a "ping pong" stereo delay, with one repeat hard left and a second repeat hard right, so thus I needed the effect itself to be centered so I could have repeats on both sides.
Hope this helps!
When using a FX send, you're probably best off setting the FX mix to 100% wet, so the bus is playing just the wet signal with no dry signal. Then, increase the send until the delay is as loud as you want it to be in the mix - this way you can adjust the delay volume without changing or affecting the volume of the original signal. I'm afraid I didn't screenprint the delay settings for the lead gitar in the mix, where I was running it through a bus - the delay on the clean rhythm guitar is set with a mix of 34% because that was being applied directly to the track, and not a bus, so that one's not a great example.
For panning, there'sa couple ways to go about it - you can either set the delay's stereo position in the plugin - in your instance, you'd be doing a 100% wet delay panned 50% to the right (though, this delay plugin I was using doesn't support that, you just get a left volume and right volume. The one I used on the lead guitar lets you specify the volume and pan for each of the repeats). Or, you can adjust the pan of the bus itself. If you look at the last screenprint there, you can see that there's a send and a return pan control on the "Lead Delay" bus. I didn't go this route because I was using a "ping pong" stereo delay, with one repeat hard left and a second repeat hard right, so thus I needed the effect itself to be centered so I could have repeats on both sides.
Hope this helps!
RAIN - April 10, 2006 06:45 PM (GMT)
Again thank you. This is an awesome help for me.
How about a little modification to what I meant on my lead delay issue.
Let's say I have a solo section that has two guitars "battling it out" (so one plays, and when it ends, the other one comes in, and they switch off for a while). I want to have all the leads patched to the same delay bus, but I want the "duelling" guitars to come from different pans, say one 25% to the right, and the other 25% to the left.
How do I accomplish that panning in the same bus? If I just use the panning control on the bus itself I get everything going in the same panning direction.
Also, does that mean that you are still always getting a dry signal from the original track? If so, how do you use effects that don't have a wet/dry mix?
Is there a way to make your complete track volume go to the bus?
How about a little modification to what I meant on my lead delay issue.
Let's say I have a solo section that has two guitars "battling it out" (so one plays, and when it ends, the other one comes in, and they switch off for a while). I want to have all the leads patched to the same delay bus, but I want the "duelling" guitars to come from different pans, say one 25% to the right, and the other 25% to the left.
How do I accomplish that panning in the same bus? If I just use the panning control on the bus itself I get everything going in the same panning direction.
Also, does that mean that you are still always getting a dry signal from the original track? If so, how do you use effects that don't have a wet/dry mix?
Is there a way to make your complete track volume go to the bus?
Drew - April 10, 2006 11:28 PM (GMT)
| QUOTE (RAIN @ Apr 10 2006, 06:45 PM) |
| 1.) Let's say I have a solo section that has two guitars "battling it out" (so one plays, and when it ends, the other one comes in, and they switch off for a while). I want to have all the leads patched to the same delay bus, but I want the "duelling" guitars to come from different pans, say one 25% to the right, and the other 25% to the left. How do I accomplish that panning in the same bus? If I just use the panning control on the bus itself I get everything going in the same panning direction. 2.) Also, does that mean that you are still always getting a dry signal from the original track? If so, how do you use effects that don't have a wet/dry mix? 3.) Is there a way to make your complete track volume go to the bus? |
Ok, I put numbers in there, so I can address these point-by-point.
I'll start with three because it's the easiest - 3. ) yes. It's your "Master" bus. Everything runs through it - tracks that don't get sent through a bus default to outputting through the master, and if you send a track into a bus, you generally send the bus to the master.
1.) This one's tricky. Unless you can find a stereo delay that preserves the stereo orientation of the original source, I'm going to hazard a guess that what you're trying to do isn't possible with a single FX send. You'd probably have to either create a FX send for each track, or just apply the FX directly to the track and use the mix control to adjust the amount of delay.
More importantly, though are you sure you WANT the delay to be in the same space in the stereo spectrum as the original? I mean, "never say never" and all that, but I can't think of a singler instance whenI've ever added a delay to the same exact spot as te original track - generally I run my leads down the center and then use a stereo delay with repeats off on the sides - maybe hard left/hard light, maybe a bunch of repeats gradually going further and further out, whatever. It sounds clearer if you don't have the exact same frequencies hitting the same space in a mix, so maybe this is a hypothetical situation what wouldn't really work that well in a mix anyway.
Also, note the difference between a bus and a FX send - a bus is a way to group a bunch of tracks together, while an FX send is a way to run a whole bunch of tracks through the same FX plugin, while varying the amount of each track that goes into it. It's sort of like sticking an independant volume control independant of track level on everything going into a bus, if it helps... They're subtly different things.
2.) Related to that last point - in a bus, no. You route the entire output into the bus, and the output from the bus into the master. In a FX send, yes. You use the track level (or, possibly a bus - look at the drums in the above mix. They're all sent to a drum bus, but each track also has a FX send that goes into the drum reverb FX send. Compare this to the lead, where the track goes straight into the master, but there's a FX send to a stereo delay). That's kind of the point of FX sends - they allow you to set the level of an effect independant of the level of the track, and to change one without changing another.
EDIT - maybe I'm misreading your first question. Do youy want the delay to track the stereo orientation of the guitar parts in a FX send, or do you just want to group all of your lead tracks for convenience into the same bus, but have different tracks in different stereo positions? If it's the later, then this is no problem - unless you specifically assign it to be otherwise, the busses are stereo. So, set your panning in the track itself, match up all the levels to where you want them to be relative to each other, and then re-route them all from the master o your "lead guitar" bus. Then you can either use this bus to quickly adjust the level of all of the tracks relative to the rest of the mix, or use it to apply the same effect to all the tracks that you wouldn't typically have a "mix" control on, like compression.
RAIN - April 11, 2006 01:08 PM (GMT)
For the record, you are being a tremendous help. Very much appreciated.
Pardon my ignorance, but how do I create an FX send? (unless I'm not understanding something, which is very likely). When I go into the bus pane and right click, I can only select either "Insert Stereo Bus" or "Instert Surround Bus". So, what I've been doing is inserting a stereo bus, adding an effect like delay to that new bus I created, and insert sends on the lead tracks to that bus.
Pardon my ignorance, but how do I create an FX send? (unless I'm not understanding something, which is very likely). When I go into the bus pane and right click, I can only select either "Insert Stereo Bus" or "Instert Surround Bus". So, what I've been doing is inserting a stereo bus, adding an effect like delay to that new bus I created, and insert sends on the lead tracks to that bus.
Drew - April 11, 2006 04:34 PM (GMT)
| QUOTE (RAIN @ Apr 11 2006, 01:08 PM) |
| For the record, you are being a tremendous help. Very much appreciated. Pardon my ignorance, but how do I create an FX send? (unless I'm not understanding something, which is very likely). When I go into the bus pane and right click, I can only select either "Insert Stereo Bus" or "Instert Surround Bus". So, what I've been doing is inserting a stereo bus, adding an effect like delay to that new bus I created, and insert sends on the lead tracks to that bus. |
That's exactly the right way to do it. Adding a send from one track into the bus, from the prospective of Sonar, sort of "reclassifies" it as a send, and not a bus.
Try this - after you create a FX send this way, go to the "output" control of another track, and pull up the list of available busses. Your FX send bus will not be in them. Do this before adding a send into that bus on a track, and it WILL be. Cool, eh? ;)
RAIN - April 11, 2006 05:34 PM (GMT)
Never noticed that. Very cool man, and thank you again. :) :clap2:
sridhar11_2 - April 13, 2006 01:16 AM (GMT)
Good work with this thread an. AMAZING. :D Dont have time to read the whole thing from start to finish right now but i'll do it eventually.
Just one question which might sound dumb or stupid but here it goes - How do you properly loop a drum sample? I ve used the groove clip looping and extended it till where i needed but there's always a gap between every consecutive loop of the drum sample. I have no idea about recording or drum machines or anything in general so any help is appreciated. Thanks
Just one question which might sound dumb or stupid but here it goes - How do you properly loop a drum sample? I ve used the groove clip looping and extended it till where i needed but there's always a gap between every consecutive loop of the drum sample. I have no idea about recording or drum machines or anything in general so any help is appreciated. Thanks
Nathan H. - April 13, 2006 08:32 PM (GMT)
Great article! I've dabbled with recording, but it's great to see a walk through of how to really set it all up.
-Nathan
-Nathan
Drew - April 17, 2006 07:45 PM (GMT)
| QUOTE (sridhar11_2 @ Apr 13 2006, 01:16 AM) |
| Good work with this thread an. AMAZING. :D Dont have time to read the whole thing from start to finish right now but i'll do it eventually. Just one question which might sound dumb or stupid but here it goes - How do you properly loop a drum sample? I ve used the groove clip looping and extended it till where i needed but there's always a gap between every consecutive loop of the drum sample. I have no idea about recording or drum machines or anything in general so any help is appreciated. Thanks |
Hmm. What program are you using, Sonar? I've never really played with the groove loop clipping tool, but it seems more like a way to specify downbeats than a proper way to loop.
Also, are you using pre-made drum loops or creating your own?
I do my own beats in Fruity Loops (though, I'm looking forward to working with a real drummer on the album I'm writing for :D) and then importing them into a new track in Sonar. Then, I set the Sonar time signature to the time signature it's reading the beat as (clip properties, I believe, will tell you this, but I don't have the program open in front of me, I'm at work), and simply drag the end of the beat out into the track - it'll automatically "loop" it as far as you drag it.
If you're doing this and still hearing a slight hesitation from repeat to repeat, then the problem is with the loop itself - you may have to delete a millisecond or two of silence from either the very beginning or the very end of the beat. If it's the beginning, it's easy - open the loop in your audio editor, zoom in to the first drum hit, and delete everything before where the waveform begins to swell into the initial attack. If it's after, you have a bit of experimentation on your hands with an audio editor that allows looped playback just to find how much you need to trim out before it loops smoothly.
sridhar11_2 - April 18, 2006 01:34 AM (GMT)
| QUOTE (Drew @ Apr 17 2006, 07:45 PM) | ||
Hmm. What program are you using, Sonar? I've never really played with the groove loop clipping tool, but it seems more like a way to specify downbeats than a proper way to loop. Also, are you using pre-made drum loops or creating your own? I do my own beats in Fruity Loops (though, I'm looking forward to working with a real drummer on the album I'm writing for :D) and then importing them into a new track in Sonar. Then, I set the Sonar time signature to the time signature it's reading the beat as (clip properties, I believe, will tell you this, but I don't have the program open in front of me, I'm at work), and simply drag the end of the beat out into the track - it'll automatically "loop" it as far as you drag it. If you're doing this and still hearing a slight hesitation from repeat to repeat, then the problem is with the loop itself - you may have to delete a millisecond or two of silence from either the very beginning or the very end of the beat. If it's the beginning, it's easy - open the loop in your audio editor, zoom in to the first drum hit, and delete everything before where the waveform begins to swell into the initial attack. If it's after, you have a bit of experimentation on your hands with an audio editor that allows looped playback just to find how much you need to trim out before it loops smoothly. |
Thanks. I dont know how to use a drum sequencer so all i needed was justa stand in drum beat so i could instruct the drummer how to play the beat. I figured out how to cut out the problem area from the samples so i'm set for a whle now. I do use Sonar. Thanks for the thread.
desertdweller - April 24, 2006 09:30 AM (GMT)
Drew, I said it when you PM'd me this thread and I'll say it again. You're a god among mortals on this stuff. Your acoustic work is absolutely mind-blowing. You remember our chats about 6 months or so ago about V-miking the acoustics. Well, you're partially responsible for the acoustic sounds on the album I'm working on now because of it, so thanks man.
BTW, what's this 590 posts BS? Shouldn't you be up around 5000 or so? :P
BTW, what's this 590 posts BS? Shouldn't you be up around 5000 or so? :P
Drew - April 24, 2006 04:33 PM (GMT)
:lol: Vince, you ass.
(DD's a fellow mod over at sevenstring.org and an all around stellar guitarist, and modesty aside, is probably better behind a mixing console than I am. And when I say "probably," I mean "and I'll be on the phone with the guy constantly while mixing my own album :lol:)
Seriously dude, I think I mentioned this in a prior discussion, but you have to try micing with an X array slightly ABOVE the "correct" spot. It was total dumb luck on my part when I did it, but it sounds so much more like I think my acoustic actually does sound, and an X-array on the 14th is pretty damned natural to begin with. I think it just adds the slightest trace of the room sound or something into the mix.
(Oh, and Vince is also a god amongst mortals at drum programming - check out his solo album over at www.vincelupone.com or his myspace page, which I can't remember the address off the top of my head but I'm sure is linked off his page.)
(DD's a fellow mod over at sevenstring.org and an all around stellar guitarist, and modesty aside, is probably better behind a mixing console than I am. And when I say "probably," I mean "and I'll be on the phone with the guy constantly while mixing my own album :lol:)
Seriously dude, I think I mentioned this in a prior discussion, but you have to try micing with an X array slightly ABOVE the "correct" spot. It was total dumb luck on my part when I did it, but it sounds so much more like I think my acoustic actually does sound, and an X-array on the 14th is pretty damned natural to begin with. I think it just adds the slightest trace of the room sound or something into the mix.
(Oh, and Vince is also a god amongst mortals at drum programming - check out his solo album over at www.vincelupone.com or his myspace page, which I can't remember the address off the top of my head but I'm sure is linked off his page.)
Mindcrime - April 25, 2006 03:24 PM (GMT)
Hey, Great write up!!
Is there any way you could post us your template file for fruity loops using all the various velocitys and stuff?
Thanks!!
-chris
Is there any way you could post us your template file for fruity loops using all the various velocitys and stuff?
Thanks!!
-chris
Drew - April 25, 2006 04:19 PM (GMT)
I could, but as I'm using samples I imported, unless you had them (and had them saved in the same directory) it wouldn't do you much good...
I didn't really vary the velocity however - I just used different samples that were hit with different amounts of strength, etc. So, it's sort of a roundabout way to the same effect, I guess...
I didn't really vary the velocity however - I just used different samples that were hit with different amounts of strength, etc. So, it's sort of a roundabout way to the same effect, I guess...