A
Altruist
New member
I mean I have used one and really I have no idea what I am looking for. I know what I am looking at but what are you looking for a straight line?
lpdeluxe
The Precision Bass Guy
I use mine to identify problem areas and then to see if my treatment really works. A flat line?? In my dreams.
C
chessrock
Banned
You're basically looking for general areas that would be candidates for possible boosts and cuts.
A no-brainer would be kick-drum or bass guitar. If you're staring at a mix for example, and you'd like to hear a little more oomph out of the kick, then watch for a few minutes and figure out where the kick mostly resides.
This will give you a quick little shortcut as to where you should start sweeping.
For vocals, you can use it to figure out where the singer's sibilant range is, making it quicker/easier to figure out settings on your de-esser or multiband comp.
A no-brainer would be kick-drum or bass guitar. If you're staring at a mix for example, and you'd like to hear a little more oomph out of the kick, then watch for a few minutes and figure out where the kick mostly resides.
This will give you a quick little shortcut as to where you should start sweeping.
For vocals, you can use it to figure out where the singer's sibilant range is, making it quicker/easier to figure out settings on your de-esser or multiband comp.
A
Altruist
New member
ahhhh okay that makes sense. I was staring at mine yesterday and I was like what the hell is this good for. So it just helps you pin-point problems easier. Thanks
Geoff
Geoff
K
kylen
New member
Altruist - What analyzer are you lookin at ?
I use a DEQ2496 1/6 octave Behringer on the outside - just before the monitors (in parallel actually - not in series). In Sonar or Cool Edit Pro apps I use Voxengo GlissEQ2, CurveEQ, Ozone3. In that order - not all at the same time, hehe.
Once I set the peak & average decay times, resolution (1/3 octave, 1/6 octave, or full) and in some cases set slopes (3dB/octave, etc), then the readings make sense after a while and you can begin to relate shapes to sounds that you hear thru your monitors. It does not replace monitoring - just another tool to assist.
GlissEQ2 is special and has overlays (different from Ozone type overlays) that can be imported in real time from any other instance of GlissEQ2 (in the same app). In other words if you have Gliss on the kick drum and want to make sure the guitar dosen't sit on top of the kick you can super-impose each spectrum on top of the other (they're different colors so you can tell the difference) and freq carve whichever instrument you want out of the way visually. Of course you should listen to it too thru adequate monitoring. There is no other spectrum analyzer that does this...
I use a DEQ2496 1/6 octave Behringer on the outside - just before the monitors (in parallel actually - not in series). In Sonar or Cool Edit Pro apps I use Voxengo GlissEQ2, CurveEQ, Ozone3. In that order - not all at the same time, hehe.
Once I set the peak & average decay times, resolution (1/3 octave, 1/6 octave, or full) and in some cases set slopes (3dB/octave, etc), then the readings make sense after a while and you can begin to relate shapes to sounds that you hear thru your monitors. It does not replace monitoring - just another tool to assist.
GlissEQ2 is special and has overlays (different from Ozone type overlays) that can be imported in real time from any other instance of GlissEQ2 (in the same app). In other words if you have Gliss on the kick drum and want to make sure the guitar dosen't sit on top of the kick you can super-impose each spectrum on top of the other (they're different colors so you can tell the difference) and freq carve whichever instrument you want out of the way visually. Of course you should listen to it too thru adequate monitoring. There is no other spectrum analyzer that does this...
K
kylen
New member
This just in today...free VST spectrum analyzer, basically a combination of GlissEQ2 and CurveEQ RTAs...SPAN:
http://voxengo.com/freevst/
http://voxengo.com/freevst/
Middleman
Professional Amateur
It's a great learning tool for seeing what is happening in your tracks. If you use soundfonts, which are pre designed, you can use an analyzer and a parametric to shape them into the right EQ zone.
You can also use them to view the shape of a song similar to what you want yours to sound like, and then apply a similar curve to your mix.
I tend to use it more for the low end to get the kick and bass to work together. Also to analyze tracks for excessive or out of desired EQ range as a clean up step prior to finalizing and printing.
Overtime you tend to learn what is happening and how to fix it by ear, then the analyzer becomes less important. However, it always get used at some point along the way when there are problems.
You can also use them to view the shape of a song similar to what you want yours to sound like, and then apply a similar curve to your mix.
I tend to use it more for the low end to get the kick and bass to work together. Also to analyze tracks for excessive or out of desired EQ range as a clean up step prior to finalizing and printing.
Overtime you tend to learn what is happening and how to fix it by ear, then the analyzer becomes less important. However, it always get used at some point along the way when there are problems.
B
Bulls Hit
Well-known member
kylen said:
Wow cool just what I was looking for, thanks!
S
shackrock
New member
it's cool to run a spectrum anylizer while listening to a commercial cd of what you want your mixes to sound like at this particular point.
for example, i had some rappers in who sounded just like 50 cent...errr that is they wanted to, ha!
When i was mixing, and doing a bit of "mastering" (final eq over it all) I tried to make it look like a certain 50 cent sound during the vrs/chrs.
ozone is tite because you can save a picture of the spectrum at any particular point, and overlay it while you play with the eq (I like to use "average" for this not "peak" mode, etc. and try to get closer to matching what it looks like during mixing expecially, and a bit on that final eq like i said).
that is no way probably how it SHOULD be done, but the results rule, so i cant argue.
for example, i had some rappers in who sounded just like 50 cent...errr that is they wanted to, ha!
When i was mixing, and doing a bit of "mastering" (final eq over it all) I tried to make it look like a certain 50 cent sound during the vrs/chrs.
ozone is tite because you can save a picture of the spectrum at any particular point, and overlay it while you play with the eq (I like to use "average" for this not "peak" mode, etc. and try to get closer to matching what it looks like during mixing expecially, and a bit on that final eq like i said).
that is no way probably how it SHOULD be done, but the results rule, so i cant argue.
R
Robert D
New member
Altruist said:
Actually, a straight flat line would sound terrible, way too bright. Since low frequencies require more energy to achieve the same perceived volume as high frequencies, a proper spectral spread is going to have a generally downward slope to it from left to right, with no big peaks or valleys.
RD
R
Robert D
New member
kylen said:
Hey Kylen - we have a couple of things in common, both in the Bay Area, and both voxengo fans.
Oh, the coolest spectrum analyzer is in Wavelab, a rotatable 3d spectrasonic display that gives an incredible visualization of your sound. I don't use Wavelab much since becoming a Samplitude user, but I keep it around just for that.
Cheers, RD
C
chessrock
Banned
Robert D said:
Actually, a straight line would sound like a bunch of pink noise.
Which would be great for checking your monitors or mixdown area. I mean a tech head might really have some use for it, but I don't think the casual listener would care for it much.
Guys, watching a reference song and "trying to match" your own to it is likely going to come out sounding like dog crap. I'm sorry to be the one to break it to ya. Use it as a general tool for mixing and trouble-shooting.
R
Robert D
New member
It certainly is not a quick fix for unexperienced ears. It's easier to interprate what your ears are telling you than what a spectrum analyzer is telling you. And if you've ever tried one of the spectrum matching programs, you probably found out that trying to force the spectral curve of your favorite commercial release to one of your mixes doesn't yield the results you expected in most cases. It is helpfull though, especially for the visual mathmatician type, to correlate spectrum analysis with what your ears are telling you, and can help pinpoint a troubling build up of energy in a certain frequency band. Just don't expect it to suddenly make everything clear because you can see whats happening. It's a little like handing a layperson an eeg strip or an ultrasound and them trying to make a diagnosis. You really have to know what you're looking at, and even then it's of limited value.
RD
RD
K
kylen
New member
Hi RobertD ! Nice Sunny California Day today 
I'm moving to Tennessee pretty soon - once I sell the house...
Yep Voxengo is way cool !
Sounds like the Wavelab thing does waterfall too - interesting !
Somebody mentioned - chessrock ? - don't depend too much on spectrums, yes that's true - it's your ears and 'feel' you get in open space that gives you that final deisions. I did go waaaayyy over board while learning it (still do probably) so I could get all that spectrum matching stuff out of my system.
I still use CurveEQ to do a 15-20% spectrum match from time to time though...or use a spectrum match kind of like a preset and change it to how it sounds best. Spectrum matching kind of needs a whole new thread itself. The best I can say is overdo it and see what it does then back off and use it like any other tool. That's how I learned compressors - by really killing stuff, hehe.
I'm moving to Tennessee pretty soon - once I sell the house...
Yep Voxengo is way cool !
Sounds like the Wavelab thing does waterfall too - interesting !
Somebody mentioned - chessrock ? - don't depend too much on spectrums, yes that's true - it's your ears and 'feel' you get in open space that gives you that final deisions. I did go waaaayyy over board while learning it (still do probably) so I could get all that spectrum matching stuff out of my system.
I still use CurveEQ to do a 15-20% spectrum match from time to time though...or use a spectrum match kind of like a preset and change it to how it sounds best. Spectrum matching kind of needs a whole new thread itself. The best I can say is overdo it and see what it does then back off and use it like any other tool. That's how I learned compressors - by really killing stuff, hehe.
C
chessrock
Banned
The problem with all the "matching" shit is that the original you're trying to match up with is . . . well, different. There's going to be different instruments and different arrangements, more than likely, and stuff is just going to sit in different places, naturally.
Example: You might look at a particular song and notice: "Damn, there's a lot more high end stuff going on past 10 khz that mine doesn't have." So you pull out your EQ and do a hi-shelf boost on yours . . . and it subsequently sounds like shit because maybe you don't have anything to bring out. Maybe the original just looks that way because they have a bunch of tamborines, shakers, and bells or bright hi-hats accounting for all that 10 khz stuff.
So if your arrangement doesn't have any bells or tamborines going on like the reference track, you're just going to be bringing up a bunch of screetchy, crappy noise from a guitar amp, sibilance from your vocals or unneccessary ringy crap from your cheap Sabien crash cymbal you found at a garage sale.
Similarly, you could look at a song that's got some sort of bass synth on it that's thumpin' away down to 40 or 50 hz. You say to yourself: "Hey, mine doesn't have any of that stuff, I think I'll just put a wompin' low-shelf and get me some more thump action."
Bad move. If you don't have a bass synth or similar deal as the original song you're trying to match, then you're probably just going to be bringing out a bunch of rumbly crap; resonance from your kick drum shell . . . a truck passing by outside . . . etc. etc.
You have to look at it like this: The spectrum analyzer is giving you a readout of how/where a particular artist's instruments and arrangements on a particular song are residing. And it's the instruments and arrangements that cause it to look the way it looks -- not some brilliant engineer's EQ curve.
Example: You might look at a particular song and notice: "Damn, there's a lot more high end stuff going on past 10 khz that mine doesn't have." So you pull out your EQ and do a hi-shelf boost on yours . . . and it subsequently sounds like shit because maybe you don't have anything to bring out.
Maybe the original just looks that way because they have a bunch of tamborines, shakers, and bells or bright hi-hats accounting for all that 10 khz stuff. So if your arrangement doesn't have any bells or tamborines going on like the reference track, you're just going to be bringing up a bunch of screetchy, crappy noise from a guitar amp, sibilance from your vocals or unneccessary ringy crap from your cheap Sabien crash cymbal you found at a garage sale.
Similarly, you could look at a song that's got some sort of bass synth on it that's thumpin' away down to 40 or 50 hz. You say to yourself: "Hey, mine doesn't have any of that stuff, I think I'll just put a wompin' low-shelf and get me some more thump action."
Bad move. If you don't have a bass synth or similar deal as the original song you're trying to match, then you're probably just going to be bringing out a bunch of rumbly crap; resonance from your kick drum shell . . . a truck passing by outside . . . etc. etc.
You have to look at it like this: The spectrum analyzer is giving you a readout of how/where a particular artist's instruments and arrangements on a particular song are residing. And it's the instruments and arrangements that cause it to look the way it looks -- not some brilliant engineer's EQ curve.
K
kylen
New member
OK well I don't mean to say that spectrum watching and matching is useless - just tricky. I'm coming at it from a remastering point of view so there's another thing.
Content, arrangement, key are all pieces of the puzzle both for your audio needing the EQ/Dynamics balance and some commercial reference you may be comparing to.
I'll just throw these ideas out to chew on - or up , whatever the case may be
In a song whose sound field seperation & depth I admire, also where the tempo and instrumentation are similiar to my piece I will take a look at a verse and compare it to my verse on a spectrum analyzer, then I'll do a chopruse, etc. I'll look for stuff like overall freq range, overall freq balance and slopes, and decay times of various bands and instruments within those bands. Kind of making a mental note of what a good mix and my mix both look like on an RTA. I'm listening too of course - this is most critical, hehe.
I'm finding I can make some pretty good spectral adjustments by using stereo balancing and dynamics processors to bring up the bass and lo-mids if necessary, but what I seem to be watching for in the spectrum analyzer is where I have to push transients down - like the guitar at 160Hz that's too loud for a few seconds as it hits some resonant thing somehow, or maybe a 4KHz spike on the vocal cause they forgot to work the mic. That kind of stuff I can usually get with a combination of Soniformer2 (32-band MB) and possibly using a spectrum match like CurveEQ. So the combination of the dynamics processor pushing down along with a touch of subtractive EQ is how I try to repair stuff - that's when I first ask Mr. Bob Ludwig to borrow the slope of the track I like...
CurveEQ can then compare the reference with my piece and calculate a 12 point EQ (a pretty broad slope). Then I adjust it so it's pushing down where I want a couple of dB...I can also allow it to adjust the general slope up a few dB per octave if I want it brighter, etc.
A pro would simply do this by ear - some of you guys here qualify for that. I'm sure I'll be doing that too as time goes on also. Of course if I was mixing I would just do a remix but since I'm doing a remaster I operate on the 2-buss in stereo and on the mid-side components. Sometimes just adjusting the slope of the sides on a muddy mix can help a lot too. Since I'm in headphones now (moving - speakers are packed and shipped to Tennessee) I would be doing zero audio without spectrums & matching - don't get the wrong idea about phones though - that's another thread too .
Anyway - tricky is what we're saying for spectrum & curve matching. I'm calling it a good home learning tool - I would rather have sat in one of your guys studios and watched & listened but I gots what I gots !
Happy Spectruming everyone !
Content, arrangement, key are all pieces of the puzzle both for your audio needing the EQ/Dynamics balance and some commercial reference you may be comparing to.
I'll just throw these ideas out to chew on - or up , whatever the case may be
In a song whose sound field seperation & depth I admire, also where the tempo and instrumentation are similiar to my piece I will take a look at a verse and compare it to my verse on a spectrum analyzer, then I'll do a chopruse, etc. I'll look for stuff like overall freq range, overall freq balance and slopes, and decay times of various bands and instruments within those bands. Kind of making a mental note of what a good mix and my mix both look like on an RTA. I'm listening too of course - this is most critical, hehe.
I'm finding I can make some pretty good spectral adjustments by using stereo balancing and dynamics processors to bring up the bass and lo-mids if necessary, but what I seem to be watching for in the spectrum analyzer is where I have to push transients down - like the guitar at 160Hz that's too loud for a few seconds as it hits some resonant thing somehow, or maybe a 4KHz spike on the vocal cause they forgot to work the mic. That kind of stuff I can usually get with a combination of Soniformer2 (32-band MB) and possibly using a spectrum match like CurveEQ. So the combination of the dynamics processor pushing down along with a touch of subtractive EQ is how I try to repair stuff - that's when I first ask Mr. Bob Ludwig to borrow the slope of the track I like...
CurveEQ can then compare the reference with my piece and calculate a 12 point EQ (a pretty broad slope). Then I adjust it so it's pushing down where I want a couple of dB...I can also allow it to adjust the general slope up a few dB per octave if I want it brighter, etc.
A pro would simply do this by ear - some of you guys here qualify for that. I'm sure I'll be doing that too as time goes on also. Of course if I was mixing I would just do a remix but since I'm doing a remaster I operate on the 2-buss in stereo and on the mid-side components. Sometimes just adjusting the slope of the sides on a muddy mix can help a lot too. Since I'm in headphones now (moving - speakers are packed and shipped to Tennessee) I would be doing zero audio without spectrums & matching - don't get the wrong idea about phones though - that's another thread too
. Anyway - tricky is what we're saying for spectrum & curve matching. I'm calling it a good home learning tool - I would rather have sat in one of your guys studios and watched & listened but I gots what I gots !
Happy Spectruming everyone !