22-05-2015, 04:13 PM
Hey guys.
I hope I haven't accidentally implied that I'm as skilled or capable as Brauer... I'm way too humble to be that ignorant. Nor have I tried to imply that my compromised (but continually improving) "studio" isn't limiting me... in fact I'm quite aware of the limitations on more than an academic level... I've had the misfortune of experiencing them firsthand. Like anyone else who takes mixing seriously, I print my mixes and listen to them elsewhere. A friend owns a local studio, and listening to some mixes in the control room is quite sobering and on other mixes is a pleasant surprise. I'm beginning to understand the problem areas my room and indeed my monitors are creating, but of course it's a complex and multifaceted issue and will take time to apply what I'm learning to my own environment.
I'm not a scientist, but I'm very smart... and I've got a much better than laypersons understanding of acoustic science as it relates to the role of the room and how it relates to the benefits and compromises I made when I switched to my current monitors.
Acoustics are a sonofab***h, especially in a small room. Early reflections in the midrange are simple to treat, but bass is never going to be perfect in a domestic setting. Not only do we have to worry about room modes skewing our perception of the bottom third of the spectrum through simple comb filtering, we also have to contend with nonlinear reverberation time across the spectrum... and that's affected by everything from our mix position to the materials used in the construction of not only the room, but the entire structure. And of course when it comes time to apply reverb electronically, well, that can obviously be a bit of a crapshoot. Then we have to factor in the fletcher-munson effect, which comes into play quite severely in a small room where playing music at 80 dB will rattle the walls... midrange bias, much?
My monitors in particular are certainly on the low end of the cost scale, and I understand quality costs, but there are a few key features that make them a sound choice for my dollar in this environment. 5 inch speakers are appropriate for a small room because it forces a roll-off in the reproduction of frequencies around 50 hZ or so, and below 50 hZ is where the absolute worst room issues exist in my current and even most professional mix environments... those waveforms are simply too big to control without the sort of resonant bass traps that cover an entire wall! Unfortunately, unported nearfields, even in much higher price ranges, are becoming increasingly rare... however, many of the small room problems associated with ports can be mitigated by choosing nearfields ported in the front as opposed to the rear, which somewhat mitigates a less than ideal rear wall proximity, which can force rear ports to severely over represent their resonant frequencies. And of course, those resonant frequencies are usually somewhere between 90 hZ and 150 hZ, which is arguably where the kick drum vs. bass guitar battleground tends to take place. You don't want that crap getting to the wall before it gets to your ears or you've lost the fight before you've even begun... we simply can't afford for a low frequency resonance to arrive at our ears late, especially when a bass is growling and the drummer is a double kick speed metal monster. That's gonna sound like mud even after you've attenuated the fundamentals to near non-existence.
Even beyond simple physics, we have nyquist/sample rate shenanigans and bit depth to worry about, though most people worry too much due to the ABUNDANT bulls**t about both all over the internet. Which is why you see the occasional gearslutz post waxing lyrical on the purported benefits of sampling at 192 kHz... of course, when you tell them that most (acoustic) instruments stop producing harmonics at 30 kHZ or well beneath that, and most microphones aren't designed to capture much past 20 kHz (hello Nyquist, are you hearing this?), he'll tell you that you just don't get it. Do you believe in magic, ooooooh, in a young girls heart? Then casually mention that personal computers, especially the ubiquitous studio laptop, by and large aren't fast enough by half to sample audio in real time at 192 kHz, regardless of the quality of the converter, you get banned for life! And we wonder why there's so much distortion in digital music.
And, relevant to the sample rate/bit depth argument, we can rant about mp3 all day. Lots of mastering engineers these days are printing multiple masters of each song, one of which is processed specifically respond more favorably to mp3 conversion. Honestly, given that the public is becoming more and more aware of the issues inherent with mp3 and that lossless streaming is soon to become a reality, I think the mp3 crisis will be over within a few years... so I generally don't worry about "mastering" in that sense, although maybe I should learn a few techniques for the benefit of posting mp3s to the forum?
Anyway... my main point is that I understand the bulk of the issues that I'm faced with (a.k.a. "the fundamentals of audio") and it's sometimes a little depressing how little I can do about them... but I don't let it deter me from experimentation if I think there might be a potential gain. Even if it means I screw the pooch and print a lousy mix, at least I've got a list of processed tracks to look over and debug, to figure out what went wrong and how I might do better next time.
Which leads me to ask, more directly, where we feel as a group the main problem areas are in the mix. I'm giving it a listen right now with fresh ears and agree with Matt about the 4000-6000 range, though I wonder if my monitoring situation is over-representing the air frequencies because they really annoy me. As Dave suggests, however, compromised source material may be the cause of this... it's unclear to me whether the cymbals are sampled or synthesized, but inappropriately sampled cymbals or synthesized cymbals based on heavily modulated white noise tend to create top shelf mix obstacles. Usually, confronted with cymbals so cumbersome on top, I wind up attenuating it, but obviously that dulls the mix as a whole considerably... so aside from replacing them with a new sample, trying to process them otherwise is probably like hanging a fancy chandelier in a haunted mansion?
Other thing that's bugging me is the low mid congestion, which is probably what Don was referring to in noting the apparent 250 hZ excess on the bass guitar. This is a bit of a sticky, since the distorted guitars seem to want way more low middle content in this genre than I'd allow them elsewhere? Which also means a lot of muscular stuff pounding away in the sides channel, and that's no good. Metal is a bizarre mix situation, since the bass guitar and electric guitars are generally downtuned by at least a half octave, if not three quarters, yet low end on double kick is just asking for trouble/intermodulation distortion/god help you if you don't have bass absorbers, so we have to allow more of the mid frequencies to let the kick drum "thump" come through... but the god forsaken guitars and bass are really hungry there, too.
So in a way, the multi-buss strategy makes sense here, to a point, by allowing you to introduce subtle level interactions between the related instruments. But it'll probably never quite work without serious, detailed automation. Automating low cuts on the kick drums during the 16th note rolls makes a lot of sense, and would allow a more bestial thump elsewhere when it's more physically possible. Something probably needs to be done in 200ish range to help the bass and guitars agree as well... could be as simple as frequency dependent sidechaining. I don't really trust multiband compressors for this because the phasing at the crossover frequency creates too many issues where I've tried it, but I know of a technique wherein a polarity inverted copy of an instrument or group is bandpassed surrounding a contentious frequency and gated... then the gate is triggered by another instrument you'd like to occupy said contentious frequency whilst they play together... and the fader of the polarity inverted track is then automated to vary the level of intentional cancellation as necessary. I've used this with some success to help a vocal cut through a wall of guitars without statically dulling them with EQ, and once somewhat successfully to remedy an overabundance of bass spill on EVERYTHING. Usually those kinda ninja moves feel like overprocessing at its worst to me, but maybe in this situation they'd be worth it?
I hope I haven't accidentally implied that I'm as skilled or capable as Brauer... I'm way too humble to be that ignorant. Nor have I tried to imply that my compromised (but continually improving) "studio" isn't limiting me... in fact I'm quite aware of the limitations on more than an academic level... I've had the misfortune of experiencing them firsthand. Like anyone else who takes mixing seriously, I print my mixes and listen to them elsewhere. A friend owns a local studio, and listening to some mixes in the control room is quite sobering and on other mixes is a pleasant surprise. I'm beginning to understand the problem areas my room and indeed my monitors are creating, but of course it's a complex and multifaceted issue and will take time to apply what I'm learning to my own environment.
I'm not a scientist, but I'm very smart... and I've got a much better than laypersons understanding of acoustic science as it relates to the role of the room and how it relates to the benefits and compromises I made when I switched to my current monitors.
Acoustics are a sonofab***h, especially in a small room. Early reflections in the midrange are simple to treat, but bass is never going to be perfect in a domestic setting. Not only do we have to worry about room modes skewing our perception of the bottom third of the spectrum through simple comb filtering, we also have to contend with nonlinear reverberation time across the spectrum... and that's affected by everything from our mix position to the materials used in the construction of not only the room, but the entire structure. And of course when it comes time to apply reverb electronically, well, that can obviously be a bit of a crapshoot. Then we have to factor in the fletcher-munson effect, which comes into play quite severely in a small room where playing music at 80 dB will rattle the walls... midrange bias, much?
My monitors in particular are certainly on the low end of the cost scale, and I understand quality costs, but there are a few key features that make them a sound choice for my dollar in this environment. 5 inch speakers are appropriate for a small room because it forces a roll-off in the reproduction of frequencies around 50 hZ or so, and below 50 hZ is where the absolute worst room issues exist in my current and even most professional mix environments... those waveforms are simply too big to control without the sort of resonant bass traps that cover an entire wall! Unfortunately, unported nearfields, even in much higher price ranges, are becoming increasingly rare... however, many of the small room problems associated with ports can be mitigated by choosing nearfields ported in the front as opposed to the rear, which somewhat mitigates a less than ideal rear wall proximity, which can force rear ports to severely over represent their resonant frequencies. And of course, those resonant frequencies are usually somewhere between 90 hZ and 150 hZ, which is arguably where the kick drum vs. bass guitar battleground tends to take place. You don't want that crap getting to the wall before it gets to your ears or you've lost the fight before you've even begun... we simply can't afford for a low frequency resonance to arrive at our ears late, especially when a bass is growling and the drummer is a double kick speed metal monster. That's gonna sound like mud even after you've attenuated the fundamentals to near non-existence.
Even beyond simple physics, we have nyquist/sample rate shenanigans and bit depth to worry about, though most people worry too much due to the ABUNDANT bulls**t about both all over the internet. Which is why you see the occasional gearslutz post waxing lyrical on the purported benefits of sampling at 192 kHz... of course, when you tell them that most (acoustic) instruments stop producing harmonics at 30 kHZ or well beneath that, and most microphones aren't designed to capture much past 20 kHz (hello Nyquist, are you hearing this?), he'll tell you that you just don't get it. Do you believe in magic, ooooooh, in a young girls heart? Then casually mention that personal computers, especially the ubiquitous studio laptop, by and large aren't fast enough by half to sample audio in real time at 192 kHz, regardless of the quality of the converter, you get banned for life! And we wonder why there's so much distortion in digital music.
And, relevant to the sample rate/bit depth argument, we can rant about mp3 all day. Lots of mastering engineers these days are printing multiple masters of each song, one of which is processed specifically respond more favorably to mp3 conversion. Honestly, given that the public is becoming more and more aware of the issues inherent with mp3 and that lossless streaming is soon to become a reality, I think the mp3 crisis will be over within a few years... so I generally don't worry about "mastering" in that sense, although maybe I should learn a few techniques for the benefit of posting mp3s to the forum?
Anyway... my main point is that I understand the bulk of the issues that I'm faced with (a.k.a. "the fundamentals of audio") and it's sometimes a little depressing how little I can do about them... but I don't let it deter me from experimentation if I think there might be a potential gain. Even if it means I screw the pooch and print a lousy mix, at least I've got a list of processed tracks to look over and debug, to figure out what went wrong and how I might do better next time.
Which leads me to ask, more directly, where we feel as a group the main problem areas are in the mix. I'm giving it a listen right now with fresh ears and agree with Matt about the 4000-6000 range, though I wonder if my monitoring situation is over-representing the air frequencies because they really annoy me. As Dave suggests, however, compromised source material may be the cause of this... it's unclear to me whether the cymbals are sampled or synthesized, but inappropriately sampled cymbals or synthesized cymbals based on heavily modulated white noise tend to create top shelf mix obstacles. Usually, confronted with cymbals so cumbersome on top, I wind up attenuating it, but obviously that dulls the mix as a whole considerably... so aside from replacing them with a new sample, trying to process them otherwise is probably like hanging a fancy chandelier in a haunted mansion?
Other thing that's bugging me is the low mid congestion, which is probably what Don was referring to in noting the apparent 250 hZ excess on the bass guitar. This is a bit of a sticky, since the distorted guitars seem to want way more low middle content in this genre than I'd allow them elsewhere? Which also means a lot of muscular stuff pounding away in the sides channel, and that's no good. Metal is a bizarre mix situation, since the bass guitar and electric guitars are generally downtuned by at least a half octave, if not three quarters, yet low end on double kick is just asking for trouble/intermodulation distortion/god help you if you don't have bass absorbers, so we have to allow more of the mid frequencies to let the kick drum "thump" come through... but the god forsaken guitars and bass are really hungry there, too.
So in a way, the multi-buss strategy makes sense here, to a point, by allowing you to introduce subtle level interactions between the related instruments. But it'll probably never quite work without serious, detailed automation. Automating low cuts on the kick drums during the 16th note rolls makes a lot of sense, and would allow a more bestial thump elsewhere when it's more physically possible. Something probably needs to be done in 200ish range to help the bass and guitars agree as well... could be as simple as frequency dependent sidechaining. I don't really trust multiband compressors for this because the phasing at the crossover frequency creates too many issues where I've tried it, but I know of a technique wherein a polarity inverted copy of an instrument or group is bandpassed surrounding a contentious frequency and gated... then the gate is triggered by another instrument you'd like to occupy said contentious frequency whilst they play together... and the fader of the polarity inverted track is then automated to vary the level of intentional cancellation as necessary. I've used this with some success to help a vocal cut through a wall of guitars without statically dulling them with EQ, and once somewhat successfully to remedy an overabundance of bass spill on EVERYTHING. Usually those kinda ninja moves feel like overprocessing at its worst to me, but maybe in this situation they'd be worth it?
I'm grateful for comments and suggestions. Thank you for listening!