My mix walkthru

ahjteam

ahjteam

Anssi Tenhunen
Jan 3, 2008
9,386
3
38
Virrat/Helsinki, Finland
www.keikkapaikka.info
Spiritbeast said:
This sounds really good BTW! Is the only compression you have just on the overall drum bus, not the individual parts?
Click to expand...

I used it on other stuff too. I used auto release and had the limiter "off" on every instance of Vcomp, but here is what I did...


And before anyone starts to bash about piracy... Eventho I did use Waves plugins, I actually have paid for my licenses. You should too.

DRUMS

Kick:
Slate Trigger (after an hour of struggling with the original kick, I just 100% replaced it with AC kick with 7ms on the attack fader to tame the initial transient so that it wouldn't sound so heavy metal)
Datube (just for some character, ~45% drive)

Snare:
Rvox (2-4dB attenuation)
Slate Trigger (NRG rooms only, blended 30%)
Rbass (@122hz; the snare didn't have anything below 250hz, so I added some bottom with this)
IK T-racks Clipper (1-3dB attenuation)
EQ (100hz highpass, +5dB @*200hz, one surgical resonance cut at 434hz, +5dB hishelf @3khz )

Toms:
EQ (65hz highpass, 4 deep cuts at 200-750hz region to carve out mud)
Vcomp (6:1 ratio, 3-6dB attenuation)

OH and HH tracks:
1khz highpass

Room (really low in the mix)
5khz lowpass
Vcomp (6:1 ratio, 6-12dB attenuation)

Drum Group:
Vcomp (4:1 ratio, about 6-12dB attenuation)

The snare, toms and room were also sent to a plate reverb from Waves Rverb

BASS

Bass DI track:
Podfarm (Gate - Rock Classic aka "Ampeg SVT Classic": low drive, no low mids or high mids, only bass and treble, with 30% blended DI for taste - Compressor)
C4 (6dB attenuation on the below 120hz range, 3dB attenuation on other bands)
L1 (1-6dB limiting on all hits)

The other bass track that comes at the end of the song:
Cubase distortion (just for different character)

Bass group:
Vcomp (4:1 ratio, 3dB attenuation)
EQ (noticed that the low end was too much so I added a -5dB low shelf @180hz, +3dB @*115hz and -5dB at 290hz)

GUITARS

Individual tracks, except 6 tracks that didn't have anything:
Renessaince Axx (1-2dB attenuation on most tracks, 6dB on the two hi-gain rhythm tracks)

Guitar group:
Vcomp (2:1 ratio, 2-4dB attenuation)
EQ (lowshelf -6dB @ 250hz)

VOCALS

Individual tracks:
Nothing or Renessaince de-esser (5.5khz highband split, attenuating 10dB)

On one track there is the "telephone eq": 1khz highpass, 5khz lowpass and +3dB @ 4.5khz)

Vocal tracks had different amounts of sends (but only the three effects that I used on the whole mix; delay, reverb and flanger/chorus)

Vocal bus:
Vcomp (4:1 ratio + de-esser, 1-6dB attenuation depending on the part)
Renessaince De-esser (same 5.5khz highband split, attenuating 7dB)

When the whole vocal chain is soloed, the vocals don't have almost anything below 200-250hz

SENDS

amounts depending on the track, flanger is only on backing tracks, not on leads

Delay (vocals, guitar solo)
H-Delay: 1/2 delay with 60% feedback
Compressor: (3:1 ratio, attenuating 8dB, 1ms attack and 500ms release) The lead vocals are sent to here as sidechain so that the delay would duck when the lead vocals are singing so that the delay would not drown the vocals

Reverb (Snare, Toms, Room, Vocals, Delay, 1 guitar track)
Rverb: 2.2 second Plate1

Chorus (6 background vocaltracks)
MetaFlanger ("Chorus light" preset)

FX group bus:
EQ (100hz highpass, +4dB @ 450hz, -5dB highshelf @ 7khz)

MASTER BUS

C4 (just for general balance taming, none of the bands cuts more than 4dB on any band)
Vcomp (1.5:1 ratio, ~3dB attenuation)
PuigTEC EQP1A (really weird looking eq but sounds really musical, just did what I thought sounded good; boosted lows +3 @60hz, boosted highs +8 @12khz)
Linear Phase EQ (hishelf +3dB @5khz)
L2 (attenuating 0-3dB on most hits, 6dB on like maybe 3-4 hits in the whole song)

I mixed into the C4 + Vcomp + L2, but added the Pultech and high end EQ boost later


Thats it. It's a bit of a "radio mix" (guitars pretty low in the mix and vocals really loud), but if excluding the snare, bass and master bus, the inserts are literally almost empty. Total of 43 plugins on 64 tracks and not a single automation. But the material was really well recorded too tho.



edit:


jimwilbourne said:
yeah dude.
this is a gorgeous song/mix.
Click to expand...


Thanks :)

Give the song a go yourself, Pedro did a good work on the recording. You can get the files here. http://www.ultimatemetal.com/forum/rate-my-mix-tone-threads/652838-new-year-gift-song-mixing.html
 
OneDaySky said:
Wow thats quite detailed Anssi.
Thx for sharing man
Click to expand...

np. I just wanted to point out that the "sound good at the source" combined with "not a lot of plugins" approach usually works better than the "fix it in the mix with all the insert slots filled on all the tracks" approach for a lot of stuff. The song could've benefited from a bit of automation, especially with the delay, but I mainly did it for practice. And with EQ's I didn't use more than +-6dB with 1-4 bands except with the toms I used 5 (1 highpass, 2 surgical and 2 wide)

And if excluding the EQ's, C4 and L2 that are fairly transparent sounding, all of the plugins I used have quite a bit of character to them (as in they distort, saturate or do something to the frequency spectrum somehow), so I didn't use them "just to make things louder", because I have like about 15dB of headroom with the mix when I bypass L2. I try to mix with unity gain input and output on the plugins when possible, because when you bypass the plugin that is the only good way to tell if the result was worse than the original instead of just louder, which always sounds "better". Especially with the Rvox distortion gets really ugly sounding if you drive the compression too hard, it somehow brings up the nastyness that it causes, but when used with taste, it sounds pretty cool. I think it actually works in very similar way as the BBE sonic maximizer.
 
I really appreciate you typing all that up for us. Like I said, the mix is excellent. Refreshing to hear some mixes completely different then metal, especially ones pulled off so well. Suprising how little you did the OH's themselves - and the fact they just got their love from the group bus. The use of RVox over a normal compressor is definitely interesting.
When I record, I'm certainly going to try a grouping approach. Would you advise grouping the lead guitars with the rhythm ones? Also, even though Joey drum samples are "pre-mixed" (Though I'm finding this definitely not to be true, the cymbals seem a little loud), am I still going to need to EQ things as much?
 
Spiritbeast said:
When I record, I'm certainly going to try a grouping approach. Would you advise grouping the lead guitars with the rhythm ones? Also, even though Joey drum samples are "pre-mixed" (Though I'm finding this definitely not to be true, the cymbals seem a little loud), am I still going to need to EQ things as much?
Click to expand...

I usually try to get the basic balance ready with just faders and static panning. Set panning, close all faders and start raising them one by one starting from: kick, bass, snare, oh, lead vocals, rhythm guitars and after that what ever is left. After that: Listen, set levels, repeat. That usually is 50% of the rough mix. Then if there is something that can't be fixed with just level, process only when needed. Because of my FOH mixing background, I usually tend to use as little tracks as possible; If something sounds totally awful (for example if you have a hihat track that sounds TOTALLY disgusting), you can actually try to leave stuff muted if it's not something really vital like the bass guitar or vocals. I have done a few of mixes where I didn't use the overhead or snare track at all. And I even once bounced out a mix where I noticed I accidentally had left the toms muted, but the band didn't notice a thing so I just couldn't give a fuck after that :D

And a funny story about using "as little as possible" was when I was still living with my room mate (he is a DJ/producer who owns a electronic music label called Finrg, their biggest artist is Proteus) and he wanted me to overdub some guitars for a song. I asked for a mix with just drums, bass, vocals and original guitars as reference, and we both laughed that the stripped down version sounded better than the full mix with all that synth stuff. He only had spent like over 400 hours on the song, it had like over 170 tracks :D The track is here btw (he didn't use the guitars we tracked because they were slightly out of tune tho so I understood he used the earlier guitars):


edit: I noticed on his youtube channel that he still is doing the "fucking loads of tracks" thing:


But no need to be stupid and anal about the "no processing" rule; if your guitars are only DI-tracks, it's understandable that you need to ampsim or reamp the signal, but in addition to that I try to have "nothing" on it except the basic HP/LP filtering at first, and the LP isn't always even necessary. But they should be printed out already before the mixing stage anyway to save CPU. Making decisions early and committing to a sound is usually not a bad thing.

And on the grouping: I had just 5 groups (drums, bass, guitars, vocals and fx) in this mix, but nothing is set on stone. If it works better for your workflow that you have kick and snare separate from the rest of the kit and the leads separate from the rhythm tracks; go for it. There is no rules, just make it sound good :)

Spiritbeast said:
I really appreciate you typing all that up for us. Like I said, the mix is excellent. Refreshing to hear some mixes completely different then metal, especially ones pulled off so well. Suprising how little you did the OH's themselves - and the fact they just got their love from the group bus. The use of RVox over a normal compressor is definitely interesting.
Click to expand...

Well, I like stuff that has a character to them. I am not 100% sure how the Rvox works, but I have a feeling that Rvox does something to the presence region which brings out the vocals, or what ever you want to bring forth, nicely. The basic compressors are bit meh, but they get the job of taming dynamics done if you have nothing better at hand.

But there are times when you need a super transparent processing. The L1 is awesome because it literally is that tasteless and odourless thing. When you put it on bass to tame the wild dynamics of the "not so audible 0-40hz" that takes 70% of your headroom if you haven't highpassed the track.

The "normal" compressors are cool too if they have some sort of character to them. For example I love the Electro Harmonix Black Finger compressor pedal, because it has a really nice character to it. I also I found out about the "Gclip on snare/kick -trick" by myself in ~2006 at live gigs by accident using the cheap Behringer T1952 hardware "tube" compressor. Don't so much like the compressor in it, but love it because of it's limiter. Sounds fucking brutal and awful, but works miracles on snare and works on kick too, it's similar to why people use Gclip.

I have to say I love the dbx 166XL (the 266XL is very similar, but I hate it), it's so "idiot safe" dynamics unit (two knob [threshold, release] gate, one knob [threshold] limiter and a really basic 5 knob [threshold, ratio, attack, release, makeup gain] compressor) and it sounds good. Love it as a whole, it also is really versatile compared to the Behringer thing, which kinda is a one trick pony.

From the software stuff the Waves CLA compressors and Waves Renaissance comp from the normal compressors is cool too. I briefly tried the CLA stuff at friends place, I think it will be my next software purchase. The Rcomp is not the best compressor out there, but I've used it like since when I started so I've become accustomed to the sound and metering and I find it really CPU friendly and good sounding. And it's like ridiculously cheap at the moment ($38 currently at Waves' front page on sale AND if you register a new account, you get the Musicians Bundle 1 for $49).
 
Last edited by a moderator:
Wow, I've gotta say, thanks a ton dude. I've been lurking these forums hours on end (I'm a little obsessed) and bookmarking everything along the way that has a helpful tip. I'll have to be honest... it's loaded with shit you say. I just can't wait to start recording so I can have something to finally show for all of my research.

Would you say for most my compression, I should use compression with "character" - like all of the things you were suggesting, as opposed just using flat out ReaComp? Well actually - is it fine to use that for ALL my compression?

I think my main issue is where will I draw the line between "chill out on the processing" and "MORE PLUGINZ = MORE SOUNDZ". A toolbox of an EQ, compressions, and a limiter (and also basic effects such as Reverb and Delay) is what I'd like to use primarily, but on the other hand you hear people saying, "Oh I inserted some saturator on all my tracks and it sounds so much nicer!" Honestly, I'd rather not go down this road, but if it sounds better, it sounds better. I guess trying these things will have to be my first goal, but I've tricked myself into thinking that when I put a plugin on something in sounds better.... when it really doesn't. That's also a skill I'll have to develop.
 
Spiritbeast said:
I think my main issue is where will I draw the line between "chill out on the processing" and "MORE PLUGINZ = MORE SOUNDZ". A toolbox of an EQ, compressions, and a limiter (and also basic effects such as Reverb and Delay) is what I'd like to use primarily, but on the other hand you hear people saying, "Oh I inserted some saturator on all my tracks and it sounds so much nicer!" Honestly, I'd rather not go down this road, but if it sounds better, it sounds better. I guess trying these things will have to be my first goal, but I've tricked myself into thinking that when I put a plugin on something in sounds better.... when it really doesn't. That's also a skill I'll have to develop.
Click to expand...

No rules really what to use, but start out raw and use bypass a lot and just try to match the input and output loudness. The peak volume doesn't really matter as it can be louder or quieter depending on the plugin, but the RMS loudness is what matter. So when you bypass the plugin, the signal should sound as loud as with the plugin on. if it sounds better with the plugin on bypass, just take it off. That way you can only tell if it sounds better instead of just louder. This is one of the reasons why the amount of plugins in my mixes have gone down like 80%. Usually the old school approach of Source to EQ/Comp or Comp/EQ and FX sends is enough on 80% of the well recorded tracks, but you don't even need to use those if they don't need any processing.

So: Start with the basic raw sound on everything without plugins or effect sends and set the volume and then if there is an element that is bothering you, fix only that and listen again.
 
You know, I honestly never paid that much attention to the output level. If the output was louder, I'd try to push down the fader of the track, not the plugin, a little to make it match. Little tips like this help so much.

I think I'm pretty prepared for the mixing challenge ahead of me, except for one thing: I need to develop a better ear for the use of the paragraphic EQ.
Are there any tutorials that give hints? Here's my problem: I'll start with a narrow boost all the way up. The tradition is to sweep it across the spectrum, trying to find "unpleasant" sounds. I thought this sounded simple enough, until I realized: Pretty much everything sounded funny with such a high spike. So perhaps I should make the boost less small, but that might hurt my ability to find the odd spots.

I'm not trying to eviscerate my raw tracks Ermz style (his guides are lovely, but since I'm using modeled guitar tones, programmed drums, such an approach will probably not be as needed.) - but there are things such as the fabled 3-4k spike on the guitar that will need to go.
 
Spiritbeast said:
You know, I honestly never paid that much attention to the output level. If the output was louder, I'd try to push down the fader of the track, not the plugin, a little to make it match. Little tips like this help so much.

I think I'm pretty prepared for the mixing challenge ahead of me, except for one thing: I need to develop a better ear for the use of the paragraphic EQ.
Are there any tutorials that give hints? Here's my problem: I'll start with a narrow boost all the way up. The tradition is to sweep it across the spectrum, trying to find "unpleasant" sounds. I thought this sounded simple enough, until I realized: Pretty much everything sounded funny with such a high spike. So perhaps I should make the boost less small, but that might hurt my ability to find the odd spots.

I'm not trying to eviscerate my raw tracks Ermz style (his guides are lovely, but since I'm using modeled guitar tones, programmed drums, such an approach will probably not be as needed.) - but there are things such as the fabled 3-4k spike on the guitar that will need to go.
Click to expand...

I personally like to use these two as rule of thumbs with digital parametric EQ's:
- rather cut than boost and don't boost more than +6dB. This is what I was taught and I don't know what this is based on, but I have noticed that after you boost something like a lot, it usually sounds really phasey and bad in general, kinda like if you boost on the 500hz area +18dB it sounds like you are in a toilet or something. This +6dB rule doesn't apply to analog EQ's as much, because if the analog EQ is good, it doesn't sound like ass even if you boost 20dB, which you sometimes have to do when getting the click with the kick for example, but still the general rule is that if you have to boost something +20dB, there is something wrong with the source sound and should be fixed at the source
- boost wide, cut narrow and use minimal amount of bands; If you notice you have 5 cuts in the same region, try using just one wider band instead that does similar curve

On the "boost EQ first, sweep the frequency and cut later": It's a VERY good practice, truely useful live when doing sound checks, as you can find the feedback frequencies that way. BUT in studio it's more of a doubled edged sword as it can lead you to cut too much. I would rather say try to listen at the sound first, then you can find the general area where the problem is (like 0-100, 100-200, 200-500, 500-1k, 1k-6k, 6-20k etc), try to fix it at source if it's a major problem (mic placement, change mic, change instrument if possible etc... Replacing the lead singer might be a bit awkward tho sometimes :(), but if it's a small problem (like a an annoying resonance or mud on a snare), you can try to fix it with the EQ.

First use a medium-wide Q and boost ~9dB and sweep the ~900hz area for the metallic "ping" resonance or the ~400hz on the mud, you can usually hear one frequency that is louder or more awful sounding than everything else, so that is most likely the frequency area you are looking for. Put the boost back to zero, then listen to it back at the zero gain a few seconds to "reset your ears", then cut it "a bit too much" at medium-wide Q, then narrow the Q until the problem is gone without affecting the other stuff around. If it comes louder, it means you didn't find the exact frequency, so sweep the frequency a bit more, but now you have like ~20-50Hz area to look for so it's way easier. But if you have something like a guitar that is too bright sounding, use a wide scoop that affects the whole 1-5khz area where the presence is, but don't cut as deep as when going surgical, only like -1dB to -4 dB instead of -12dB to -20dB.

But once again, general rule of thumbs without even hearing the sound source, just do what sounds the best :)
 
I love this band and song and loved mixing this. It practically mixed it self from the source.

But man, you've got some wierd stuff going in with your vocal correction, it's tuned but there is a noticable amount of the wrong notes you've picked.

"and the RAIN came down"

"YOU dove right in"

for example.

:)
 
Barnacle said:
I love this band and song and loved mixing this. It practically mixed it self from the source.

But man, you've got some wierd stuff going in with your vocal correction, it's tuned but there is a noticable amount of the wrong notes you've picked.

"and the RAIN came down"

"YOU dove right in"

for example.

:)
Click to expand...

I don't own autotune (yet), so I used the less than stellar pitch correction thing on Cubase, and I took out the vibrato out intentionally from a few spots because it was bothering me. The central note is the same as it was originally. Also on the choruses there is a flanger/chorus, that may sound like badly tuned vocals :)
 
You must log in or register to reply here.

Similar threads

ahjteam
MERRY XMAS: Mixing time (groovemetal multitracks inside)
Replies
11
Views
2K
Practice Room
Uncle Junior
U
P
Trying to get a Swedish chainsaw HM-2 guitar sound with VST plugins
Replies
2
Views
1K
Practice Room
DIYistkrieg
DIYistkrieg
ahjteam
Guitar Tone Competition 3: Summary
Replies
26
Views
5K
Practice Room
MarkG
M
gotet10
New Progressive/Djent-ish mix I've been working on. Link to multi's inside
Replies
9
Views
716
Off Topic
Jonny Crrash
J
Morgan C
Song from my band's upcoming EP.. please help finalise the mix
Replies
0
Views
263
Backline
Morgan C
M
Top Bottom
Back