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

Mixing Help?
MUFF WIGGLER Forum Index -> Production Techniques  
Author Mixing Help?
dykesh
I know this is a broad question but I'm getting to the point that I must really suck at recording or mixing stuff. I've got a really great home studio with a 9U eurorack of goodies. My main question is: how much is everyone EQing or compressing your orginal synth tracks? It just seems like everything I record is very flat or one dimensional no matter how much delays and reverbs I use and I just can't get happy with how the overall tone of frequencies sits in the mix compared to other tracks. My tracks just never sound as open or alive as say Depeche Mode or NIN, etc. I know I cannot get all the way there but any help and opinions would be really appreciated...
modularland
The secret to mixing is to think of your music in three dimensions:

1) frequency space
2) volume space
3) panning space

the reason your mixes are 'flat' is because you've made a 'flat' mix- rather than one being 3D.


for example- a drum kick has some very low frequencies- but it has others as well- if you 'isolate' the drum kick, with EQ, to a certain frequency range, then it will 'live' in that frequency dimension

now your synth comes in- if the synth has too many dominant frequencies that overlap the kick drum, you get a 'flat' sounding mix. use EQ to reduce the amount of energy the synth puts into frequencies in the drum kick's range, and the synth will 'pop' out of the mix since it will then live in its own frequency space.

the same principle applies to volume and panning... compression is used to reduce the variety of volumes appearing in the mix, but in doing so 'flattens' the overall sound, which is why pop music is often fatiguing on the ears- there is no variety of volume...

make sense?
dogoftears
XYZ mixing as modularland mentions is one of the keys.
i have been trying to achieve your described goal for some time now-- to be able to produce full sounding 3-dimensional mixes out of my home studio. it is possible, and yes it is difficult. trial and error goes a long way.

the most important thing IMO is to separate frequency space... don't have two 1k leads on top of each other etc... or if they must be in the same section, do a call and response. i do lots of little tiny hi-q super shallow band pass cuts when EQing stuff. it is important not to cut too deep and to *never* add with digital EQ. compress lots of stuff, but dont over compress any thing unless you specifically want that kind of compressed sound. if you find yourself having to over compress something to turn it up loud enough in the mix, you've done something else wrong. retrace your steps. every mix is like a little puzzle-box... figure out all the little things to do and you have success, a lot of time you don't even notice it coming and then you "solve" the mix.

you may want to also meter the average loudness of your mix, called RMS. there are many different ways to meter RMS and ME's like Babaluma could chime in here and talk about that but if you run Voxengo Elephant in "pure" metering mode, it should give you a good reflection of your RMS. many professional recordings have an RMS of -10 or great. actually i believe there was a ricky martin single w/ an RMS of -3, which is just absurd really. If you look at even -10 RMS on your level meter, it will look like a level-sandwich, and yet this is industry standard. your RMS *will* make a difference in the way your song sounds when you take it in the car etc...

as far as getting that "Z" axis of stereo-space going... you can never have enough! the more stereo breathing room and modulation you give your sounds, the richer it tends to make the mix. i've noticed this especially on large sound systems-- songs where i really put the time into stereo spread, panning, and sample delay type fx have an incredible 3-dimensional quality. so once you've got your EQing and compression down, definitely start opening up all those cool stereo processing plug ins you have. 9U of Euro is great but if it's all mono that's still quite some post-production to get it to sound like Depeche Mode, hehe. I often use my DSI Evolver if only because it spits out perfect stereo sounds as a matter of preset.

hope this helps i'm sick of typing now smile
mojopin
now that i make music with the modular i rarely need compression or eq. those things are taken care of within the patch and performance of the part. I do need to run high pass filters on a most everything that shouldn't have lowend since the mx-4s passes dc. I also use that mixer to make virtually every patch in stereo so I don't have a problem with getting massive mixes. So as mentioned before, place sounds around the soundfield and maybe pan. Use delays and reverbs and mix summed to mono. Get that sounding good and then take it off and you'll be blown away. thumbs up
skweeegor
dogoftears wrote:
it is important not to cut too deep and to *never* add with digital EQ


I've read this many, many times. Why specifically is this a general rule? I know logically it makes sense to cut frequencies in order to accentuate a specific range rather than boosting, but why not make use of both if needed? Is there something in particular about digital eq's that makes this a bad idea? What made me re-think this was seeing a video on Depeche Mode's live setup about a year ago. They had all the tracks running through Ableton and its packaged effects, and I distinctly remember seeing the Ableton EQ-8 plug on many tracks with some pretty crazy boosting on certain frequencies - +6db and more! Does the rule change for live use or something? I doubt that DM's people don't know what they're doing.
Curtischip
skweeegor wrote:
dogoftears wrote:
it is important not to cut too deep and to *never* add with digital EQ


I've read this many, many times. Why specifically is this a general rule? I know logically it makes sense to cut frequencies in order to accentuate a specific range rather than boosting, but why not make use of both if needed? Is there something in particular about digital eq's that makes this a bad idea? What made me re-think this was seeing a video on Depeche Mode's live setup about a year ago. They had all the tracks running through Ableton and its packaged effects, and I distinctly remember seeing the Ableton EQ-8 plug on many tracks with some pretty crazy boosting on certain frequencies - +6db and more! Does the rule change for live use or something? .


DM's Abelton rig is not what you think its for, you've been misinformed if you think they run "The Show" on Ableton... its for something far more creative and very much improvised than one would think and is not at all the mix of the gig, far from it, Ableton is about 5/% of the sound you hear.
The EQ your referring to is extreme for a reason, due to placement and layering, take the live snr for example 2 mics plus triggers also layered with a Radar track, triggers going to Ableton, which is then feed back down the multi so at foh you will have at least 4 to 5 channels for a snr.
regarding Live EQ, its very much a different approach to the studio and you tend not to have the the 3d dimension space as discussed above, also you must be cautious with hard panning as well, and as a rule you never boost only cut, and most of the shaping is done pre desk.. e.g crossovers and P.A processors, then graphics and parametric's, desk eq is to chop and shape and fine tune, Studio mixes can have great depth and colour like fine intricate paintings live tends to be more of a Van gogh bright bold and in your face.

listening to a CD or a recording is like watching a porn movie, where as live music is has like having Sex.. sweaty sticky and smelly hihi

Quote:
I doubt that DM's people don't know what they're doing

not this shit again
Sheez you youngsters so assuming, DM's ppl know exactly what they are doing they don't have that high rep and standard of quality for nothing, After all They are the only Electronic stadium act!... for good reason!
skweeegor
Thanks for the explanation.

Curtischip wrote:

Sheez you youngsters so assuming, DM's ppl know exactly what they are doing they don't have that high rep and standard of quality for nothing, After all They are the only Electronic stadium act!... for good reason!


However, I don't know where this is coming from. I asked the question because I was wondering if my assumptions were wrong. I also think you misread what I said since you're arguing against a point I did not make:

Quote:
I doubt that DM's people don't know what they're doing


Translation: DM's people know what they're doing.

Sheez, you old farts are so assuming. hihi
Curtischip
skweeegor wrote:


Sheez, you old farts are so assuming. hihi

hihi hihi thumbs up

not arguing just informing.. just come across as cranky at times you kids get off my lawn love
skweeegor
Totally cool; I figured it was just a misunderstanding, and you answered my question, so hey! thumbs up
dykesh
Thanks for the info. I've been working at this for years and really thought it would be a little easier with mostly synths and drum samples and not dealing with live guitars and bass, etc; but I'm just not getting what I want out of it. I really have not been carving out space for the synth tracks, so I'll work on that. I'll try and get some kind of example up tomorrow if anybody's interested. Keep the ideas flowing...

I think many people still do not like the sound of a digital EQ when boosted. I have some really nice analog hardware and I think it sounds better than digital. I think it's more personal opinion than anything else...
dogoftears
well, there are some exceptions to the digital EQ rule, but generally, digital EQ's tend to destroy sound in a very digital way as soon as you start to boost. there is an art to cutting that should make it so you never have to boost-- if you want your 1800hz louder, then cut yr 1k and yr 2k.... then turn the volume of the channel up!! tada. mission accomplished.

this is most important when you are using high-quality source material, like a modular synth for example. your sound is only as audiophile as it's lowest common denominator, if you start boosting with a digital EQ, your sound is now only as audiophile grade as that EQ plug in. there are some exceptions to this, for example i've found that the UAD EQ's are great for boosting when done correctly. but if you have a mix full of tracks with boosts in them its going to be over-colored and bleeding into itself like crazy.

generally when using digital tools to "make it sound good" you want to use them as transparently as possible. for example mojopin mentions that he hi-passes all his sounds but the bass. i do this too-- every single sound in the mix is hi-passed at 200 hz. when i first started making music with this formula, that 200 hz HP would just timbrally obliterate so much of the character of my sounds. but these days, now that i know how to design sounds for very specific frequency ranges, the HP is barely audible. as often as possible i try to use an analog HP to roll off the low end. this is tough though as most analog hi-passes are not surgical/exact enough.

i highly recommend the Air Windows plugs in for transparent channel-perfecting... a lot of them are free... try running "Channel" at 50% on all your channels and buses. AFAIK it slews a lot of digital stepping in small amplitude changes, a barely audible effect but one that adds up on a whole mix.
modularland
I would also recommend listening, and I mean really listening, to some older recording of jazz combos with or without vocals.

When I say "listening" I mean playing on good quality speakers, turning off the computer, leaning / sitting back, and thinking about the music you are hearing.

Listen to where the instruments live in the mix, where the frequencies are, which pops out, which is in the back, etc...

When you can visualize what you are hearing on an old recording of a jazz combo, you can then apply that to making your own recordings.

The reason I pick jazz is because it has drums, bass, guitar, maybe a horn or two, and maybe a voice- all instruments that live in their own fre/vol/pan dimension very distinctly.

Synths suffer from being too powerful- they can create frequencies in all ranges with great definition, so its hard to make 3D mixes often because we forget to emulate what our ears hear in nature: different things with different frequencies in different points of space.

The reason for an 'older' jazz recording is that, in those days, making a correct mix wasn't an option- the only people allowed in studios were ones who knew how to play instruments and ones who knew how to engineer- so the results were better than what we all make, since there are no filters keeping us out of our studios :-)
dykesh
I guess I'm not really getting the right frequencies out of the modules so I'll work on that tonight for sure.
dykesh
Is anyone routinely running filters in series to get a LP, and then a HP to tame the low bass frequencies?
Babaluma
dykesh wrote:
I know this is a broad question but I'm getting to the point that I must really suck at recording or mixing stuff. I've got a really great home studio with a 9U eurorack of goodies. My main question is: how much is everyone EQing or compressing your orginal synth tracks? It just seems like everything I record is very flat or one dimensional no matter how much delays and reverbs I use and I just can't get happy with how the overall tone of frequencies sits in the mix compared to other tracks. My tracks just never sound as open or alive as say Depeche Mode or NIN, etc. I know I cannot get all the way there but any help and opinions would be really appreciated...


people will tell you this rule and that rule etc. but there are no rules in audio. always use your ears as the final arbitrator. if it sounds good, it is good.

imnsho, the best thing you can do to make your mixes better is PRACTICE. keep at it. it takes years and years of practice to get good at, just like anything else. the reason your mixes sound flat and lifeless compared to a lot of commercial releases, is that the engineers who recorded, mixed and mastered those albums probably have a great deal more experience than you do.

next on the list (and these are just all personal/subjective statements, so please take them with a pinch of salt) are these (and in no particular order):

learn about GAIN STAGING. i cannot over-emphasise this enough. from the source to the mic to the pre to the compressor to the eq to the speakers, or even within your digital plugin chain, each item will have an optimum point for both input and output gain. if you can optimise each stage, your audio will have far more life than without this optimisation. experiment, experiment, experiment.

the beauty of 24 bit: we now have more dynamic range than the human ear can perceive. USE IT! no need to crank anything into the red any more (except perhaps the last mastering digital limiting stage IF you want it LOUD). most audio interfaces are set to work at their best sound quality with analogue outboard gear at the analogue level of 0dB VU, which is normally set somewhere around -18dBfs digital level. if you stick to recording and mixing with -18dBfs RMS average in mind, then your track will be at exactly the right level to interface perfectly with your mastering engineer's expensive class a all discrete transformer balanced analogue outboard gear. wink

nice gear is not necessary, but it can help greatly if you have sorted the above two things. a nice pre amp/d.i, compressor, eq etc. can do wonders in the right hands. in my experience i have never found any plugin that sounds as good as my analogue gear, and i'm constantly trying new stuff to see, as are many in the mastering community. we are all still pretty much agreed, especially where compression is concerned, that an equivalently great sounding digital compressor is a far way off. you have to remember that analogue electronics has about 100 years more R&D than digital, which is still pretty much in its infancy. and when i say good gear, i mean good gear, not the latest piece of chinese prosumer crap. good gear is expensive. better to save a long time and get good gear, than keep "upgrading" your crappy chain with slightly less crappier items. i am definitely NOT saying amazing music and sounds can not be created with cheaper gear, just that if you want the best/top tier sound and build quality, it costs a lot of money.

whilst the argument about loudness continues, try not to get too caught up in it. most of the best sounding recordings from the 50's through to the early 90's have an RMS of somewhere around -20dBfs. in my personal opinion, listeners of the future are going to look back on the mid 90's and beyond with complete and abject horror.

and to repeat for emphasis: USE YOUR EARS! i've lost count of the number of times i've read stupid posts by people talking about how they are looking at waveforms/VUs/FFTs/visualisations/gain reduction graphs etc., and are trying to eq out what look like uneven bumps etc., EVEN WHEN THEY CANT HEAR them! it is just ridiculous. the visual tools are there to supplement what your ears are telling you, not to replace them. when i'm mastering, i'm shutting my eyes, listening to the music, and twiddling the knobs. that's all that matters.

hope that was all of some help.

sorry for the long rant. we are still a bit worried for our lives here, but hanging in there!
MrBiggs
Anyone point to any good sources of info about things like gain staging and such? Lately I've been finding myself fighting my equipment almost as much as actually making sounds, and I'd like to wrap my had around some of these basic concepts.
I took to re-reading my Ultralite manual cover to cover last week, which helped with the routing and channel issues that I have been having. But I've not really quite figured out some other things like volume levels. I spent three hours playing and recording my synth the other night and I couldn't even get the levels high enough to make a dent on the Cuemix Oscilloscope. It was nice and loud through my headphones plugged into the Ultralite, however. So somewhere between the final stage of the modular (VCA and Malekko's Output module) to the Ultralite's input trim to the mix channel level to the input and mix level of into the computer (sometimes Ableton Live, other times Wave Editor which has no level adjustment) to the headphone output my staging is off. Secondly, and this may be more of an Ultralite thing, I don't seem to know if what I'm hearing is the sound as it enters and leaves the Ultralite, or if it's the result of having run through the DAW. That is, I'll add an effect in Live but not hear it and I usually can't figure out why. I'm sure it has to do with monitoring the channels, but since MOTU Cuemix doesn't affect anything in or out of the computer, it just loses me.

Any good websites or books?
Babaluma
amazing thread here, with loads of great posts by joel hamilton:

http://messageboard.tapeop.com/viewtopic.php?t=54487

and dug these up from SOS:

http://www.soundonsound.com/sos/apr98/articles/gainstructure.html

http://www.soundonsound.com/sos/oct00/articles/soundcard.htm

http://www.soundonsound.com/sos/1995_articles/oct95/gainstructure.html

http://www.soundonsound.com/sos/oct08/articles/qa1008_5.htm

a good google search on "gain staging" should help too.

and make sure your soundcard is set to work best with your other gear, -10 or +4.
MrBiggs
Babaluma wrote:
amazing thread here, with loads of great posts by joel hamilton:

http://messageboard.tapeop.com/viewtopic.php?t=54487

and dug these up from SOS:

http://www.soundonsound.com/sos/apr98/articles/gainstructure.html

http://www.soundonsound.com/sos/oct00/articles/soundcard.htm

http://www.soundonsound.com/sos/1995_articles/oct95/gainstructure.html

http://www.soundonsound.com/sos/oct08/articles/qa1008_5.htm

a good google search on "gain staging" should help too.

and make sure your soundcard is set to work best with your other gear, -10 or +4.


Good good. Thanks Babaluma. I googled it or something like it and after hitting three links realized that who the hell knows who's writing this -- and two of the three were forums. The SOS stuff is just what I was looking for. I shoulda thinked of that.

What does your last statement mean? -10 or +4 what?

EDIT: With the Ultralite I've read to keep the input trims down, but then I've also read that the mix channel faders are just for monitoring. If these are both true, then I have no idea where the preamp levels are set. And i know I'm likely missing some obvious something...
Babaluma
MrBiggs wrote:
What does your last statement mean? -10 or +4 what?


on some interfaces, you can set the input and output gains between "pro" level +4dBu vs. "consumer" level -10dBv. but you have to be careful that it's not just dropping the level in software (and thereby loosing headroom) in the -10 mode (like on echo products, i always keep everything set at +4).

http://en.wikipedia.org/wiki/Line_level

i seem to remember that the actual level difference is about 11dB, but someone with a more technical background might be able to jump in here. the confusion arises because one is reverenced to "v" and the other to "u".
Babaluma
ok, for good gain staging, start with the source, and work your way through the chain slowly.

eg, you have a loud source you want to mic, the mic is quite sensitive and you have a quiet room, so backing off the distance of the mic from the source may be a good idea so that you don't overload the mic itself.

the mic is plugged into a mic preamp. the mic has a high output, so you engage the mic pre's -20dB pad. there should be some way to monitor the output level of the mic pre. you want this hovering around "line level" if you intend to record it digitally. if you don't know where line level is on the output of your mic pre, then you need to do some more research.

ok, so far all signals and gain stages are analogue. but you want to record them digitally, right?

so the next stage is the physical input on your soundcard/interface. my echo audiofire 12 has physical level meters on the front panel, as well as "virtual" ones in the software mixer for both input and output gains. forget the output gains for the time being. not important here. things can get a little confusing now.

every digital interface will be preset so that an analogue line level hitting its inputs will be referenced digitally to the old analogue standard of 0dB VU. on some hi end converters you can set this reference level yourself with an internal trim pot. i emailed echo, and they told me that their converters are referenced such that 0dB VU is referenced to -18dBfs. that means that most of your average signal level coming in to the converters should hover around -18dBfs, and that you have 18dB of digital headroom before clipping. there should be a way of seeing this on your software meters. if you are recording at 24bit, having this much headroom is great, and noise will not be a problem, you still have oodles of dynamic range. if you are still recording/mixing at 16bit then lord help you. it's always a toss up between headroom and noise, but with 24 bit you are eliminating most of the noise problem, and you should therefore be concentrating on keeping as much headroom as possible, not cranking the gain into the red and clipping the converters.

the other problem with pumping higher than 0dB VU levels into your soundcard, is that the analogue stages within your soundcard, before it hits the converter, including the opamps and filters etc, especially on "prosumer" gear like most of ours, does NOT handle "hotter than line level" signals at all gracefully. it might be fine to clip your lavry/burl/cranesong/forssell converter at 0dBfs, because the analogue stages before the conversion have been meticulously designed with +30dB of analogue headroom, but this is most certainly NOT the case with lower end gear. i can hear it on my echo converters easily. push a mix into the converters too far past 0dBVU/-18dBfs average RMS and you can hear the mid range start to sound "pinched" and grainy. back it off a little and it sweetens up.

so, you have your analogue signal, it's not clipping anywhere so far (in the analogue realm or digitally), you are recording at 24bit with an average RMS of around -18dBfs (or wherever your interface is calibrated), and you have a way to meter these signals at various stages. great.

that's the input stage done with. the next is output, how you route your signals, mix them together, and then send them to your monitor chain. the same basic rules apply. there should be a way to meter the output of your final mix, and then route this to your headphones or monitors, which will also probably have their own gain stages that need optimising. i've already ranted long enough so am gonna stop here, much to your benefit. wink

it's all about being extremely logical and methodical, following the signal through from start to end, and checking it at every stage along the way. rather like programming a modular synth patch. it stops being a "hassle" and becomes second nature when you realise that the the quality of your recordings is dramatically improving.

i can highly recommend the sonoris meter vst plugin if you want an extremely precise way to measure digital gain stages. as it's a plugin, you can insert multiple instances of it wherever you like to see what the gain is at a particular input or output stage. great for checking that none of your software plugin chain is clipping digitally anywhere.

http://www.sonorissoftware.com/catalog/meter-p-34.html

MrBiggs
Thanks for that educational post. Too bad everyone's asleep (and I should be) or else I'd go upstairs and flip all my switches to 'on' and work this out.

My mic preamps are the first two inputs of my interface, which also handles the DA/AD, so that simplifies things. I've also learned in the last week that these first two inputs are mic/instrument level, while the other analog inputs are line level. When I made this connection, I immediately switched my modular's output from hitting input 1 (mic) to input 3 (line). This made a huge difference of course. So while it's no longer clipping like crazy, now it's quite low. You make a great analogy with the signal flow of the modular -- just working out the stages as the signal leaves the modular, hits the guitar pedals, and into the MOTU will help. I also think I've figured out to ignore the MOTU's mix stage when I'm using the DAW, as the software will route all the inputs in and out to the MOTU's outputs -- the mix stage is for when it's being used as a mixer. If I have this right, that will make a huge difference and explains a lot -- it would also explain why I get so much feedback when using a mic into Ableton.

I often can't believe I've had this thing for three years and I'm still learning this stuff.
mojopin
Babaluma wrote:
MrBiggs wrote:
What does your last statement mean? -10 or +4 what?


on some interfaces, you can set the input and output gains between "pro" level +4dBu vs. "consumer" level -10dBv. but you have to be careful that it's not just dropping the level in software (and thereby loosing headroom) in the -10 mode (like on echo products, i always keep everything set at +4).

http://en.wikipedia.org/wiki/Line_level

i seem to remember that the actual level difference is about 11dB, but someone with a more technical background might be able to jump in here. the confusion arises because one is reverenced to "v" and the other to "u".


i have an old apogee rosetta that i use. i think i calibrated it to around -18dbfs to match the max output of my modular. but i set the input impedance to -10db as it is 15k ohm vs 9k ohm on the +4db. seeing as modulars have higher output impedance, i figure it doesn't hurt to raise the input and thus widen the ratio as much as possible. i am not sure if it really matters though.
Babaluma
mojopin wrote:
i am not sure if it really matters though.


well, that's where you've got to ask yourself, "does it sound better"?
MUFF WIGGLER Forum Index -> Production Techniques  
Page 1 of 1
Powered by phpBB © phpBB Group