MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Creating separation using filters (only)? Advice needed
MUFF WIGGLER Forum Index -> Production Techniques  
Author Creating separation using filters (only)? Advice needed
Paranormal Patroler
I've found myself in the following situation and I would appreciate some advice on how to best tackle it. I'll go into quite some details below but feel free to ask me for more information if you think it's necessary in order for you to give a more educated suggestion. I appreciate your time and answers, especially if you go into the trouble of explaining your reasoning.


The synth and its options for mixing, panning, EQing
I am using a Roland XV synth for live purposes which can unfortunately only send the same stereo image to all its outputs (RCA, Stereo jacks, headphones) with no way of assigning different instruments to different routings, apart from the typical L/R panning. The synth offers three effects per Patch (MultiFX, Reverb, Chorus) and the same per Performance*. My initial intent was to use the onboard MultiFX as an EQ to work with each instrument on the mix, but after extensive internet searching, since the manual was being super cryptic, I realized that it's impossible to have the effects running separately on each Patch while in a Performance Mode. This means that during a performance the instruments can only be processed through a common MultiFX, a common Reverb and/or a common Chorus.

The on-board MultiFX options that I can use for mixing purposes are the following:

    - Stereo EQ (Low 200Hz/400Hz, MID1 200Hz - 8KHz, MID2 200Hz - 8 KHz , High 2KHz/4KHz/8KHz)
    - Multi Band EQ (250Hz, 500Hz, 1KHz, 12,5KHz, 2KHz, 3,15KHz, 4KHz, 8KHz, Q 0,5/1/2/4/8)
    - Compressor
    - Limiter



Instrumentation and filtering options
The sounds are of my own design and some of the ROMpler's available structures (re: each patch's audio path) allow for a VCF before the final VCA. It's not always the best structure to use for every given sound, but I'll see if I can work with such a structure under all circumstances if push comes to shove. Here are my available filter options: LPF, BPF, HPF, PKG (this one is a like a Peak Filter which cannot act as a Notch, but only as Resonance) and two LPFs which can work with the on-board Filter Envelope to reduce the levels without changing the CutOff frequency (???). The sounds I've designed and use are:

    - Upright Base
    - Saxophone
    - Trombone
    - Trumpet
    - Drums (Kick, Snare, OHH, CHH and maybe Ride/Cymbal)
    - El.Piano


There's also an overdriven guitar sound in there but that's not so important as I tend to introduce it separately. I will probably need some separation between some of the instruments as the whole live M.O. contains a fair amount of improvisation and there's no way to use the arrangement to predetermine where each instrument will lie. I tend to fly all over the place with my sequencing and that's a given, it won't change.


External Hardware and the idea of parallel processing
I also have external hardware that I intend to use live. One idea I'm working on is to use all three available stereo outputs, even though they offer the same mix, and to process each one in parallel using

    -two mono fully parametric EQs (LowCut, 3 bands, HiCut)
    - a Three Band stereo filter which doubles up as a three band compressor
    - a stereo compressor (which has no on-board dry/wet) but I want to use it for glue and warmth
    - a convolution reverb which I can either program for glue, reverb or as a crazy processor depending on the IR


I then think of combining the processed mixes together on a three stereo-channel tube mixer (it's actually 2 x 3 mono). The whole thing could obviously could turn things to shit (is there such a thing as too much warmth?) but it's something I'm enjoying exploring as an option for now. In any case I think it's absolutely essential to get a fine mix going from the synth itself before using any external gear. Any other stuff I add (whether serially or in parallel) should be considered as icing on the cake and fodder for tweaking live so a good starting point is super important. I don't want to do mixing live; or at least not in a corrective way, more in a playful/creative manner.


The big questions
Is it possible to use the available filters to create separation between the instruments and if so what would be your guiding tips as a starting point for each instrument, given the type of instruments I use? Keep in mind that I also have the final MultFX as a master bus option where I can use it to apply some EQ to the whole mix. How would you use that? I also have the hardware items which I can use in parallel or serially, but always on instances of the same original mix. Everything is a like a mix bus effect in a way. I'm going for a jazz-funk style so the kick will be more prevalent compared to your typical jazz recording.


A final word about panning and live processing
As far as panning goes I've been studying Rudy Van Gelder's mixing and so far I'm contemplating some of felixer's suggestions for LCR panning: so I was thinking Base and Drums should be in the center, with the three brass instruments panned hard left and right. I'd love to but there's no way I can pull of the opposite reverb trick that RVG usually does, as there's just one Reverb with no way of panning it, but I do use a Strymon Deco right before the house's console so I create a strong stereo image using that. I haven't tried hard panning the brass yet but I do move them a bit and then ask that the house hard pans my LR channels, so the Deco works its wonders. I'm open to suggestions in any aspect of the synth mix, so feel free to give me advice there as well. I'm also open in terms of how I could repurpose my external hardware, especially serially, as there's a big discussion on whether compression should be before EQ or not. I guess that's another can of worms entirely but it's on the table if you feel it's important.

I like my setup and I like how open it can be and for some reason I enjoy that I'm only opting for stuff that could be considered a mixer's tools instead of delays and whatnot. Subtle processing and live remixing is what I have in mind. The typical keyword nomenclature applies in what I have in mind: glue, warmth, stereo imaging, unity and contrast, groove and freedom to work with my crazy sequencing without having to constantly think of frequency bands; I obviously listen to what I play and tend to try keeping things separate but an initial good mix is important, right?


*A Performance consists of 16 Parts and each Part contains one Patch. So in essence you can use up to 16 Patches simultaneously, where each Patch is in essence an instrument. So, think of a Performance like being able to play up to 16 different instruments.
felixer
the easiest way out of your dilemma (not having seperate outputs for the different sounds) is to get some extra midimodules. then you have the advantage of choosing the best box for the job. and as you already have a mixer they would slot right in. for live maybe keep the setup as compact as it is, but for studio it doesn't matter how much junk you have to move hihi edit: there are many small/lightweight and cheap midimodules around. and they only have to make 1 good sound to be useful to you Mr. Green
the valve mixer might be good: i used to put my K2000 thru a pair of ARTpre's. and my tech always said how much of a difference that made. the eq in the roland is not going to be much use: the low band is to high and the high band is too low to correct the middy character pretty much all roland gear has ...
another option would be to get some knobby box with midi and assign level/pan parameters to that. then you have some hands-on control. but implementing that could be a nightmare and may not be possible ...
the sounds you are using seem good: distorted guitar and brass usually mix well ... the bass goes below that and the piano and drums can cut thru if you have the right sound for them: thick layers of brass are not the easiest thing in the world to mix but if you don't put them too loud you can make it work ... put the compressor after the mixer and you'll be able to use it as glue. forget about seperating sounds with eq: it never works and you are left with 'amputated' sounds.
prob a midicontroller is the simpelest solution. hope the roland's midi implementation allows that!
good luck!
Paranormal Patroler
Hi felixer,

thanks for chiming in! Let me reply to each point you raise:


felixer wrote:
the easiest way out of your dilemma (not having seperate outputs for the different sounds) is to get some extra midimodules. then you have the advantage of choosing the best box for the job. and as you already have a mixer they would slot right in. for live maybe keep the setup as compact as it is, but for studio it doesn't matter how much junk you have to move hihi edit: there are many small/lightweight and cheap midimodules around. and they only have to make 1 good sound to be useful to you Mr. Green


My main concern and question is if and how I should take advantage of each instrument's final filter option to create a better mix. If at all possible. That's why I mention the different filter options available (LPF, HPF, BPF, PKG). In the studio I'll use more MIDI devices, if not for easier mixing, but for the added value that I get when using physical modeling for some of my sounds instead of ROMplers. I enjoy the VL70m's playability but I don't want to add too many beans to the broth when I'm playing live. I like how the Roland ROMpler can be coaxed to produce all the sounds itself; it's such a let down that it doesn't have better routing options, it would be helpful.


Quote:
the valve mixer might be good: i used to put my K2000 thru a pair of ARTpre's. and my tech always said how much of a difference that made.


I bought the Erica Fusion mixer for its added tube warmth and because I got intrigued by the way it compresses the incoming sound when a bass source is used. But I'm rather worried I might be overdoing it in that area, between the tube, the colouring of the external compression and the tape overdrive at the end, I think there might be too many compressors and too many warmers.


Quote:
the eq in the roland is not going to be much use: the low band is to high and the high band is too low to correct the middy character pretty much all roland gear has ...


What would you suggest in terms of external EQing to correct that? I have two excellent and super playable mono EQs which can pretty much do anything and I'm dying to use 'em. Did I mention how playable they are? They invite you to mess with the knobs.


Quote:
another option would be to get some knobby box with midi and assign level/pan parameters to that. then you have some hands-on control. but implementing that could be a nightmare and may not be possible ...


I'm playing live with the velocity parameter, it changes a lot, so levels are not stable at all. As for the panning, I think I'm going to go for your LCR suggestion as described above. Wish I could do the reverb trick that I'm hearing in most Rudy Van Gelder recordings: he sends each brass instrument on one side and its reverb on the other side. Makes things really sit nicely in the mix. I think I'll add some on-board reverb to the El. Piano to push it backwards and keep both that and the Upright Bass in the middle. I have no idea how to pan the Drums.

It's the Drum vs Upright Bass that worries me most. Any suggestions on filtering the two to create a decent mix?
Again, the sound is jazzy but I'm keeping the Kick quite prevalent as I tend to groove stuff; more funk than jazz in that sense and so are my rhythms. In most jazz recordings I'm listening to the Kick is almost non existent, or at least very secondary.

Quote:
the sounds you are using seem good: distorted guitar and brass usually mix well ... the bass goes below that and the piano and drums can cut thru if you have the right sound for them: thick layers of brass are not the easiest thing in the world to mix but if you don't put them too loud you can make it work


It's three separate instruments, but the playing style is very complex and counterpointy. That's one more reason why I think panning between the three brass instruments in a zig-zag manner (re: left, right, left) is a good idea. They always follow each other in that manner so I'm hoping it will add to the stereo image instead of making it worse. As always, this will have to be tested, but you know that better than I do.


Quote:
... put the compressor after the mixer and you'll be able to use it as glue.


Yeah, about that. I get compression from the following sources: the hardware filter can act as compressor with side chain on specific bands, the compressor itself (FMR RNLA) can be coaxed with an extra EQ to work with sidechain, the mixer itself adds a but of compression if pushed, and the Deco adds some tape compression. It'll be a pain in the ass to keep things nice, warm, but not overdo it. Or maybe not.

That's why I'm considering parallel compression on one pair of the outputs. I've also read that reverb can act as a nice glue if added in moderation. Apart from the RNLA every other processor I have has a wet/dry option. I was thinking of using the three pairs of outputs sending the same mix to the compressor, the reverb, the filter (band compressor?) and mix those. I could grab the output of the mixer, pass it through the EQs for some tweaking and then go through the Deco for the added tape push.

seriously, i just don't get it

There's a ton of way to go about it, I know, I know. I'll have to test to see what works for me. But some advice by people who have tried things won't hurt, right?

Thank you for your answers, I really appreciate having you on this thread. You've given me much food for thought in the past.
felixer
the two mono eq's are usefull. but maybe look for a stereo one? the old neumann rackunits are nice ... if only because you can adjust things with one hand without looking ... i often adjust eq without looking at the settings.
parallel processing is good. it means you have the different sounds on seperate faders. and you can clearly hear what is going on. so there is less chance to overdo it: you can alwas push up the clean version.
as for the drums vs bass: try to pan the kick and the bass slightly apart. that keeps 'm out of each other's hair. then again if they are tight you want to keep 'm together and look at them as 1 instrument ...
drums in general should not be too stereo: it sounds unnatural to have the toms 20 feet apart ... but of course it creates a big hole in the middle where the other stuff can live. the other way around, i think, gives more possibilities: keep the drums fairly mono and fold the other sounds around it. the kick being so prominent is a result of the 80ies/disco/techno. on lots of older recordings (and live work) it isn't so important. now i like bass (as i think it is the more interesting instrument), so on my mixes the bass is often louder then the kick. but bass is undervalued: esp around where i live/work the bass is thought of as some muddy drone underneath: the players are not used to hearing what they are doing and as a result are very limited. they never heard of ron carter or chris squire meh
the 'reverb on the other side' trick is nice. you can do that with delays/harmonizers too. it keeps things clear and at the same time easier to blend. in general i use delays more then reverb as i don't like the wash that often comes with it. and with delays you can make nice stereo effects.
ps maybe post some music you are doing. now i feel a bit in the dark and can give only general advice ...
Paranormal Patroler
felixer wrote:
you can do that with delays/harmonizers too. it keeps things clear and at the same time easier to blend. in general i use delays more then reverb as i don't like the wash that often comes with it. and with delays you can make nice stereo effects.
ps maybe post some music you are doing. now i feel a bit in the dark and can give only general advice ...


I read somewhere that a lot of people don't know that you can use delays to create depth instead of reverbs. Now you mention it as well. Any insight on that? Are we talking Haas delay or do you have something else in mind?
felixer
no, not haas delays. not too crazy about that as it does weird things when summed to mono. simply two different times (try 331&537ms, works well on almost everything except drums) one for left, one for right. adjust the feedback so that the endresult is equally long: more feedback on the shorter time. you don't want to sync them to tempo as the idea is to make 'm bounce against any rhythm. hence the two non-related times. if you want to go at it scientifically you prob want prime numbers: 331 is one, 537 is not, 541 would be one). on my old&trusted spx90 (max delaytime=500) i go for 300 against 500 and set the feedback so that it just doesn't come together again at 1500 ms). this is just one set i found works well with vocals and any kind of solo's. maybe 97 against 137 for chords/ensemble work. i go for a loose/random type of feel and add just a bit. experiment! most delays have some multitap mode ...
btw those two times i gave at the start of my post is a good example of 'setting without watching'. they come from my modded copycat tapeecho. originally the tapeheads are spaced equally. i found that a bit boring, so got out the screwdriver and started moving 'm: the last one all the way just before the erase and write heads (for frippertronics/eno-type stuff), the first one very close for rockabilly-type slap echo's and the other two 'somewhere in between'. this is what i came up with. then later i got a digital delay and copied the settings by ear ... reverb is basically a lot of those non-related echo's ... the ideal machine for this might maybe be an ursa major spacestation: a multitap delay dressed as a primitive reverb. another box i really enjoyed was the a/da stereo tapped delay. both now vintage and quite expensive: don't bother and get a nice modern 24bit box instead. often getting two boxes is the easier way to do it. behringer makes this em600 'echo machine' that is really nice. and cheap Mr. Green
Paranormal Patroler
I have my eyes set on a couple of boxes that could pull the multitap delay off. I also bought a Roland Demora which is quite pristine and powerful, I'll see if I can program it accordingly.

I need to prepare some videos of my live sets and upload them on Youtube. I had a gig last week and the kick & upright base sounded like mud together which pushed me into drone territory for a bit to sort it out. No EQ on my setup yet to fix it on the fly but I'm getting there. I liked your idea of considering the bass and kick as one instrument, so I tried having the main sequence of the base being derived from the kick pattern (with some extra hits on different dynamics to keep things interesting). It worked quite well, it wasn't as boring I imagined! But I understand you meant that from a mixing perspective which I obviously need to work on more.
felixer
i find i always need eq on the kick. ussually to cut at the mud frequencies (around 300-400 Hz). and i like my bass pretty forward so i usually boost at 1,5-4 kHz. most simple deskeq's can do that ...
but that is with a band and acoustic drums. with electronic sounds anything goes ... put the principle is still the same ...
MUFF WIGGLER Forum Index -> Production Techniques  
Page 1 of 1
Powered by phpBB © phpBB Group