Frequency/Spectrum Display?

[masteringhouse]

A spectrum analyzer can be useful to help find problematic frequencies until your ears are trained better. However, don't fall into the trap of trying to match one song to another (similar to programs like Har-Bal). There's far too many variables for this to work with pro results.

In addition to spectrum meters there are also phase meters that are actually another nice thing to have if they are included with the package.

Once your critical listening skills become sharpened there will be little or no need to rely on these.

offtopic:

I'm glad to see that Glen is on the case. Also the usual crew (Benny, Farview, Ryan, Chessrock, John, etc.) Their input to this forum has been extremely valuable. I'm finding less of a need to help monitor audio BS since much of my input would be redundant, though I still like to drop in on occasion. This forum has really grown-up from when I first started posting and has become a great source for budding engineers.

Keep up the great work dudes!

 

[chessrock]

I'm a lot friendlier to the idea of spectral analyzers than most on this board.

I honestly couldn't do without them. Sometimes my ears get a little tired, or they just lose all perspective and they need a little visual help to confirm what they might be hearing or not hearing.

Let's say I'm dialing in sounds for a heavy guitar track, and there's some spike up there in the high mids that's just shredding my ear drums. If I track a sample of it and have a look at it on the graph, I can instantly tell where the offending range is. Usually somewhere around 2-3 khz, and if it's higher than that, then we probably have some problems that need to be addressed before we proceed, etc.

Also, when I de-ess vocals, I generally use a MB-compressor ... and since everyone's sibilant range is a little different, I can generally pull up the graph and have a look at where that particular vocal's sibilant range is at with pin-point accuracy, and then dial that in to the MBcomp.

When sound-checking, I sometimes pull it up and have a look for where the peaks are on the kick drum, and compare it with any peaks that might appear on the bass guitar track. If they're both spiking pretty heavily at 80 hz, for example ... then I might want to have a listen for potential conflicts.

From there, if it sounds fine, then I don't worry about it. But if I hear something that seems to lend credence to what I saw on the spectral graph, then I might chime in and mention something. Hey Mr. Bass player ... where do you envision your instrument's roll in the mix? Heard more than felt? Featured or complimentary? etc. And just work on adjusting the tone from there if you think there might be some conflicts with what the kick is doing.

When mixing, I might have a peak at it to figure out some possible danger zones. Like if I want to add some high end to the mix, where do I start? Well, looking at the graph, I know there are a lot of cymbal crash transients popping up all over the place around 4-5 khz ... so I might want to proceed with caution and try a high shelf a little farther up from that range, so as to avoid danger on the cymbal crashes. That kinda' thing.

.

 

[SonicAlbert]

The only thing I think is a little wrong in this thread is the concept of "use your ears not an RTA", which is correct but implies that you use either one or the other. How about using an RTA *and* your ears? Why diss a tool that can be so helpful? There's no reason why a person can't use *both* their ears and an analyzer.

It's another tool in the arsenal. Of course you don't want to "mix with your eyes" based solely on what you are seeing on an analyzer. But at the same time, an RTA can be a great help in locating problem frequencies quickly and maybe calling attention to something that might otherwise be overlooked.

I own the Elemental Audio Inspector XL, a great program, and I also use the Behringer Ultra-Curve 2496 (one of the best units they make). The 2496 has a bunch of different metering and I keep that unit on the master buss when mixing, switching between the Peak/RMS meters and RTA analyzer. It's really nice to have that instant reference, I wouldn't want to be without it.

 

[djdarwin]

I'm a lot friendlier to the idea of spectral analyzers than most on this board.

I honestly couldn't do without them. Sometimes my ears get a little tired, or they just lose all perspective and they need a little visual help to confirm what they might be hearing or not hearing.

Let's say I'm dialing in sounds for a heavy guitar track, and there's some spike up there in the high mids that's just shredding my ear drums. If I track a sample of it and have a look at it on the graph, I can instantly tell where the offending range is. Usually somewhere around 2-3 khz, and if it's higher than that, then we probably have some problems that need to be addressed before we proceed, etc.

Also, when I de-ess vocals, I generally use a MB-compressor ... and since everyone's sibilant range is a little different, I can generally pull up the graph and have a look at where that particular vocal's sibilant range is at with pin-point accuracy, and then dial that in to the MBcomp.

When sound-checking, I sometimes pull it up and have a look for where the peaks are on the kick drum, and compare it with any peaks that might appear on the bass guitar track. If they're both spiking pretty heavily at 80 hz, for example ... then I might want to have a listen for potential conflicts.

From there, if it sounds fine, then I don't worry about it. But if I hear something that seems to lend credence to what I saw on the spectral graph, then I might chime in and mention something. Hey Mr. Bass player ... where do you envision your instrument's roll in the mix? Heard more than felt? Featured or complimentary? etc. And just work on adjusting the tone from there if you think there might be some conflicts with what the kick is doing.

When mixing, I might have a peak at it to figure out some possible danger zones. Like if I want to add some high end to the mix, where do I start? Well, looking at the graph, I know there are a lot of cymbal crash transients popping up all over the place around 4-5 khz ... so I might want to proceed with caution and try a high shelf a little farther up from that range, so as to avoid danger on the cymbal crashes. That kinda' thing.

.

ChessRock:

Nice contribution. That was exactly what i was looking for. That is exactly how i would use my RTA.

I can hear that the bass and the kick are masking eachother and dont sound "just right." I can cut and boost where ever i want and it will make a difference but im not practiced enough to know exactly freq to cut and boost to make it sound perfect.

Thats where the RTA comes in. I can see where the Kick peaks and where the bass peaks and cut and boost accordingly. Maybe this wont cause my mix to "sound the best" but it gives me a great starting point to work from.

I wasnt planning on boosting or cutting while just staring at my RTA but more or less using it as a point of refference to work off of.

Ears are much more important than the eyes, but why not use the tools available to help your ears out a little right?

Hell... if chessrock does then i better right?

Just some clarification. Glen doesnt support this, but is this wrong?:

Using the RTA on the mix to locate "blank" or "missing" spots in your "Wall of sound." Just like chess said, you can identify where there is room for more high end or whatever. Got room for cymbols in your "wall of sound" why not add them right? I want my RTA to show me a full mix, a full wall of sound before i compress/master.

Is this wrong?

Keep in mind this is hip hop and if you dont got a loud wall of sound you got nothing. Period.

 

[mshilarious]

In addition to spectrum meters there are also phase meters that are actually another nice thing to have if they are included with the package.

Oooh I do like my phase meter. Even though I mostly just look at the 1-dimensional version, Wavelab has a really cool XY graph that is like one of those Media Player visualizations. It helps add ambiance since I don't have a lava lamp

I don't mind the spectrum meter either, it comes in pretty rainbow colors

Working on microphones and such, I live and die by the FFT. Couldn't function without it. But that has little to do with mixing.

 

[kylen]

Using the RTA on the mix to locate "blank" or "missing" spots in your "Wall of sound." Just like chess said, you can identify where there is room for more high end or whatever. Got room for cymbols in your "wall of sound" why not add them right? I want my RTA to show me a full mix, a full wall of sound before i compress/master.

Now I'm getting confused...easy to do!
Are you basing your arrangement on what the RTA tells you? I don't think I fully understand that concept. If the song arangement needs cymbols put it in, then mix it so stuff fits. If the song sounds too full - then rearrange and possibly remove something. That kind of stuff you have to hear IMO. Did I misunderstand...or am incompentent or both

 

[darkcubus]

these are all very true. i do agree with using both if you know how and when to use them. i have just always been under the impression it's best for me to know what's going on without one to truly know what i'm doing with one. but if you have em both and know how and when to use em, use em!

 

[bennychico11]

Using the RTA on the mix to locate "blank" or "missing" spots in your "Wall of sound." Just like chess said, you can identify where there is room for more high end or whatever. Got room for cymbols in your "wall of sound" why not add them right? I want my RTA to show me a full mix, a full wall of sound before i compress/master.

but what's a good frequency look like? What's a good high end look like? Hmm, at 5kHz it says I have -50dB...is that good? There's a roll off around 15-20kHz...should I boost there? It means I have room for high end right? Or is that just because of the sample rate it was recorded in?

What does a good spectrum look like? Should all the frequencies be flat all throughout? No, the human ear doesn't hear that way.
Just because the spectrum reads one thing doesn't mean you need to add highs or mids or lows because it appears to be lacking in that area. Each room, each instrument, each ear, and each mix is different. There's no way I'm ever going to EQ more mids in a guitar just to get the spectrum analyzer to fill in the mids a bit....now the guitar sounds like crap.

Do not change the way your instruments sound individually to satisfy your visual needs.


---edit---
by the way, it's spelled 'cymbals' guys.
sorry that was buggin' me

 

[djdarwin]

but what's a good frequency look like? What's a good high end look like?
What does a good spectrum look like? Should all the frequencies be flat all throughout? No, the human ear doesn't hear that way.
Just because the spectrum reads one thing doesn't mean you need to add highs or mids or lows because it appears to be lacking in that area. Each room, each instrument, each ear, and each mix is different. There's no way I'm ever going to EQ more mids in a guitar just to get the spectrum analyzer to fill in the mids a bit....now the guitar sounds like crap.

Do not change the way your instruments sound individually to satisfy your visual needs.

Thanks for the advice benny.

I wasnt planning on boosting an allready existing instrument in my mix but rather adding a new instrument. If the RTA shows im lacking in the mids, is it wrong to add a new guitar or new piano if it fits?

Sometimes you can sit there listening and say "i need more mids... lemme throw in another guitar" and it will be the right decision. Sometimes but not always (at least in my case). Sometimes you add a new guitar and it sounds muddy and is masking your other mids.

If i used an RTA i would know wether or not there was room for the guitar in the first place correct? It would save me hours of recording and tinkering to just know.

Like i said, i wouldnt use the RTA to boost freq on existing instruments cause that will more than likely destroy the nice "natural" sound of the instrument. I would however use my RTA to determine if i could throw in little things like cymBALs/shakers/whatever.

I would never just start boosting till my meters were flat across the board.

Is this wrong or a bad approach?

 

[bennychico11]

I wasnt planning on boosting an allready existing instrument in my mix but rather adding a new instrument. If the RTA shows im lacking in the mids, is it wrong to add a new guitar or new piano if it fits?

well, with the approach I sorta agree....but that is compositionally speaking.
Many composers add more harmonies to help thicken up the chords. Happens all the time. Composers will even had other instruments to give a different sound. Maybe he's using 5 trumpets already, but what would it sound like if he added a flugelhorn? That would change the way the brass sound.

Definitely with that approach I see where you're going. However, base it on the ears and not visually. If you added a guitar, does it thicken it up a bit and make it sound better...or is it just clashing with the other rhythms/harmonies. A composer might add a maj7th or 9th in a chord to add a different feel to it, but if he starts adding more and more harmonies to the chord it may start to actually subtract from the song as a whole and stuff may clash.

If you add another instrument, do it for compositional reasons....again, not because the meters tell you that you have room to.

again, just my opinion.

 

[kylen]

If you add another instrument, do it for compositional reasons....again, not because the meters tell you that you have room to.

again, just my opinion.

2nd that opinion!

 

[djdarwin]

well, with the approach I sorta agree....but that is compositionally speaking.
Many composers add more harmonies to help thicken up the chords. Happens all the time. Composers will even had other instruments to give a different sound. Maybe he's using 5 trumpets already, but what would it sound like if he added a flugelhorn? That would change the way the brass sound.

Definitely with that approach I see where you're going. However, base it on the ears and not visually. If you added a guitar, does it thicken it up a bit and make it sound better...or is it just clashing with the other rhythms/harmonies. A composer might add a maj7th or 9th in a chord to add a different feel to it, but if he starts adding more and more harmonies to the chord it may start to actually subtract from the song as a whole and stuff may clash.

If you add another instrument, do it for compositional reasons....again, not because the meters tell you that you have room to.

Great.

Thanks again Benny. That was my plan from the beginning, but i dont think i worded it correctly in my original post.

Using the RTA as a reference to help my ears. It will can show me where i "might possibly if necessary" add something new.

Always keeping the quality of the mix in mind as I work.

Anyone else agree or disagree with this approach? Tips/tricks to using the RTA?

Thanks everyone...

 

[SouthSIDE Glen]

Using the RTA on the mix to locate "blank" or "missing" spots in your "Wall of sound." Just like chess said, you can identify where there is room for more high end or whatever.

Regardless of the genre of music, a "hole" in the response curve doesn't necessarily mean something is missing; that could simply mean that because of the instruments and arrangements used that there are *natural* spots in the curve that are not filled. That is neither a bad thing nor an indication of there being actual holes in the wall. It's *normal* for there to be an unevenness in the response curve.

And, OTOH, a perfectly "even" response across the spectrum has another name: it's called "white noise". Perfect spectral response, but not exactly what I'd want to listen to .

It's perfectly possible - hell, it's been done a million times - to have a great hip hop composition that's in your face, that gets the blood flowing, that makes the old fart in the car three lanes over from you get a queesy stomache, with nothing more than a beat track, a synth sequence, lead vocals and choral backup vocals. There's going to be more ups and downs in that response curve than there are in a San Francisco car chase, and that's how it's supposed to be. Filling in those holes via signal processing will do nothing but muddy the mix.

As far as the idea of using the RTA to determine the composition or arrangement of the song, I'm not even sure where to begin on that, that just seems so wrong to me 9 ways to Sunday. Let me just grab a few at random:

First of all, if you need a meter to tell you that you have a "hole" in the upper mids - or any other general section of the spectrum - then your ears just are not ready to mix. And if you can't train them any better than that, you need to find another hobby or line of work. Honestly and seriously. I'm not saying that to be mean any more than I was being mean if I said if you can't run 100 yards in under 10 seconds, you have no business being a cornerback in American football or if you're 5'2" and can only vertical jump 5" you probably should not take the position of power center on a basketball team.

Second, you should compose a song to convey feeling and emotion, not to please an oscilliscope. I don't care what the genre of music is, there are millions of examples where the producer or arranger purposely emphasizes a particular frequency range or leaves another one out in order to convery specific feeling. The same can be said for selection of instruments and their arrangement within a song. In such cases the response curve will certainly not be "even" in the least. 100% frequency spread 100% of the time will get boring and monotinous real fast.

As to those who ask "what's wrong with using ears and an RTA both?", I would reply, well, if it works for you, that's fine, and I won't tell you to change. But to those not yet set in their ways, I reply, "what's the point?" It's unnecessary, it takes longer, and it is more dependant on outside sources than simply training you ears like 99% of pro engineers have.

There is no reason to have to depend upon an RTA to do your mixes right, and there is no advantage to doing it that way either. As good, and often better, results can be had with a good set of ears alone as can be had with an average set of ears and an RTA. So why bother with the excess baggage? It's just an unnecessary complication, not a way to make it easier - which after all is what the ulterior motive to this thread is, isn't it? Aren't we really looking upon using an RTA to make the job easier? Well, if so, we are seriously barking up the wrong tree.

And, as I said earlier, not only do those with well-trained ears and the disclipine of concetrated and critical listening simply have no need for an RTA, use of an RTA before the ears are ready can hamper or at least slow down the ear training process. When one is not dependant upon their ears, one tends to not use them as much or as critically. An RTA may make it easier on the ear, but that also means that the ear is not exercised as hard and, just like a muscle, tends to atrophy when not regularly exercised.

Finally @ Tom: Hey you, if my being on the case means you are around here less, than take me off the case, I'm doing this board more harm than good. Don't make me come out there and force you to eat a cheesesteak made with mozzerella cheese instead of cheese whiz just because you disappeared on me .

G.

 

[kylen]

Regardless of the genre of music, a "hole" in the response curve doesn't necessarily mean something is missing; that could simply mean that because of the instruments and arrangements used that there are *natural* spots in the curve that are not filled. That is neither a bad thing nor an indication of there being actual holes in the wall. It's *normal* for there to be an unevenness in the response curve.

It's perfectly possible - hell, it's been done a million times - to have a great hip hop composition that's in your face, that gets the blood flowing, that makes the old fart in the car three lanes over from you get a queesy stomache, with nothing more than a beat track, a synth sequence, lead vocals and choral backup vocals. There's going to be more ups and downs in that response curve than there are in a San Francisco car chase, and that's how it's supposed to be. Filling in those holes via signal processing will do nothing but muddy the mix.

As far as the idea of using the RTA to determine the composition or arrangement of the song, I'm not even sure where to begin on that, that just seems so wrong to me 9 ways to Sunday.
G.

Excellent

 

[djdarwin]

Damn...

I get all excited then you gotta come through and ruin it all glen.

Anyways thanks for the response. Very informative indeed.

I understand your point (to the best of my ability) and completely agree.

I was only planning on using the RTA to look at problem spots, not use it to actually construct my mix.

Just like chess said in his earlier post. Find the peaks of your bass drum and find the peaks of your bass guitar and EQ where there is problems. NOT pick your bass and pick your kick based on the RTA. NOT add 3 more kicks cause my meters aint topped out on the RTA in the lows.

Just use it as a tool in my box... like benny and chess seem to cautiously give me the OK to do...

Nothings more important than the ears. Bottom line. Not like you can look at an RTA and be like "damn thats a great song!!" Its silly. Its not what im trying to accomplish or wish to put into practice.

But f*ck all that...

Glen/Chess/Benny/(people who know more than me)...

How can i train my ears and practice? Every day i get of work and Mix Mix Mix Mix MIx mix... But sometimes i cant tell if the mix i made last week is better than the mix i made yesterday.

Any way to practice or train my ears other than what i am currently doing: Mixing ALL THE TIME?

Anyway to get more benifit out of the time i spend mixing?

 

[Hi_Flyer]

take some time AWAY from mixing. take a break for a few weeks, and then come back with a fresh perspective.

instead of trying to figure this stuff out on your own via the internet, try to sit in and watch an experienced engineer. you'll learn more, and you'll learn faster.

 

[bennychico11]

Any way to practice or train my ears other than what i am currently doing: Mixing ALL THE TIME?

listen, all the time.
Not to your stuff, to other people's.

Buy CDs and listen to them not only because you like (or don't like) the band, but to hear what the engineer did. Try and pick out what kind of effects they might have been using....listen to how a guitar sits in the mix. Listen to how much reverb you notice on the vocals, and listen to how subtle the reverb might be. Is the bass and kick drum clashing with eachother? Compositionally, think about the rhythms.

You may actually start listening so intently to stuff that you start to get annoyed. You'll start to be the only one in your group of friends who shouts out "OMG! Did you hear that effect on the guitar and how it panned from left to right?!"
It'll be a good thing.

 

[djdarwin]

You may actually start listening so intently to stuff that you start to get annoyed. You'll start to be the only one in your group of friends who shouts out "OMG! Did you hear that effect on the guitar and how it panned from left to right?!"
It'll be a good thing.

I started doing that about 6 months ago. Its a great thing. Friends hate it, but still a great thing.

 

[chessrock]

Using the RTA on the mix to locate "blank" or "missing" spots in your "Wall of sound." Just like chess said, you can identify where there is room for more high end or whatever. Got room for cymbols in your "wall of sound" why not add them right?

In a way. Sort of. Think of it like this; if you feel like your mix is lacking in the high end ... and you look at the spectral analyzer and you notice that it does sort of curtail and fall off towards the right end of the spectrum graph ... then you just used your eyes to help verify what you heard with your ears.

Now there's a few ways you can go about this. If you've already got tamborine tracks, shakers, and plenty of cymbals in your mix ... then chances are your mix is just lacking some clarity, and you might try a high shelf, or perhaps a wide boost somewhere in the higher registers.

On the other hand, suppose your mix doesn't really have any cymbal ... doesn't have any shaker or tambo or anything like that. Well, then that should give you a clue that you might want to start adding some of those elements in to your mix in order to fill it out more in the higher registers. But here's the kick : If you need a spectral analyzer to tell you that, then you've got some problems in your composition/arrangement ability. That should just be something that comes intuitively to you as an arranger or composer.

I want my RTA to show me a full mix, a full wall of sound before i compress/master. Is this wrong? Keep in mind this is hip hop and if you dont got a loud wall of sound you got nothing. Period.

If your name isn't Phil Specter, then this is probably flawed thinking. But maybe it isn't. Only you can answer that.

But consider this: If you see a billboard or a magazine advertisement where every inch is covered with text: "Buy Now!" "Lower your interest!" "SAVE!" "LIMITED TIME ONLY!" ... does this actually make you want to look at it? Similarly, if you read a post from someone on this board (not to name names) who doesn't ever use a paragraph break and fills every available inch with text ... don't your eyes get kinda' tired trying to read it.?

On the other hand, let's say you've got another advertisment with nothing but a photo of Heidi Klum, half-naked on a bear-skin rug ... with the name of the product at the top of the page, and a quick, one-sentence tagline. "Got Heidi?" Or something stupid like that. In graphic design, they call that "white space," and visually, the use of white space is very important in that it prevents the viewer from tuning out due to sensory overload.

Music isn't much different. If you try to jam too many different elements together, then you run the danger of making everything sound like a busy, cluttered mess, and the ear doesn't know what to focus on. The mix runs the danger of losing power, impact and amplitude, and like Southside says, you're getting close to White Noise territory.

 

[SouthSIDE Glen]

Any way to practice or train my ears other than what i am currently doing: Mixing ALL THE TIME?

As you can see in my signature space, I'm actually working on a notebook on critical listening, but it won't be ready until this fall. In the meantime a few quick ideas in short:

- first of all, try that technique given to you in the other thread using an equalizer to mess around around with a full mix and *CAREFULLY LISTEN* to just what each center frequency band on the EQ sounds like when that single band is boosted and cut. I used capital letters there because really listening and paying attention to the actual character of the sound of each band is the key. Do that for a half hour a day for a couple of weeks. Then get a friend to do it for you randomly, making you guess which band she's* moving and in which direction and, eventually, by how much. Before you know it, you'll have a great idea of just what each general band's character sounds like which will be a huge start.

- listen to a few quality commercial mixes (one at a time) with the following bullet list in you hand: instrument location, soundstage panning layout, reverb use and character, frequence content and spectral balance (using the skills you got above), song structure, instrument arrangement, and lyrical content. Put the song on and listen the first time concentrating on just instrument location for each individual instrument (including vocals) and how it is used throughout the song. Then play the song over and concentrate on the overall soundstage layout (all the instruments together) and concentrate on how they relate to eachother and integrate (or don't integrate) together. The do it over again concentrating only onthe next item on your list, and so on. Then the next time try it again, but listen to more than one list characteristic at once. As you go along, you're memorizing the list itself. Before that long, you'll be finding yourself auotmatically making these analysies to any song your hear more or less at the same time and more or less as second nature.

There are lots of other tricks and exercises that involve stuff like critically listening to live performances in person, critically comparing speakers and monitors in retail showrooms, etc. but those above should get you started for now. The rest I'll be typing out in detail over this summer.

*I say "she" because you can kill two birds with one stone. Get your gal to help you by having her play back tracks from Barry White, Gato Barbieri, or Etta James. As you impress her with your aural sensitivity, you'll be getting her in the mood for some oral sensitivity

G.