ten band graphic eq

[xstatic]

Actually, sidechaining an EQ to a comp does not really act like a multiband compressor. It's actually a really cool trick. Compressors all have their own individual frequency curves in which they respond to before compressing a signal. When you have the threshold set, that is a basic setting. The frequency curve that the comp responds to is based around that threshold, but is not always the same for all frequencies. At 80 hz the comp might trigger sooner than say at 2 khz. Maybe think of the Fletcher Munson curve as a loose not necessarily related graphic image. Sidechaining an EQ allows you to retune that curve as it were. If you sidechanin a graphic EQ and boost 2khz, the comp will respond to 2khz a little sooner than it would have prior. Conversely, by cutting alot of low frequencies you can greatly change a compressors reaction to low frequencies so it is more likely to start compressing when higher frequencies hit the set threshold. When sidechaining an EQ to a compressor the EQ boosts and cuts do not affect the audio path itself, so the signal is not EQ'ed, but just affects the compressors algorithm so to speak (at least pertaining to threshold and frequency). Kind of like Key filtering gates. This would allow you to "tune" your compressors reaction to various frequencies at various thresholds which can greatly affect how a compressor sounds and acts. It's one down and dirty way to really make something like an Alesis 3630 sound much better than it really is.

Some (or I guess even all) of this may or may not be correct, but this was how it was all explained to me years ago. I have used this trick on occasion and it definately changes how a compressor acts. It can be very hard to get used to, but sometimes it really does the trick well

 

[Ford Van]

a 10 band can be great but I would use it after the mix, not while you're mixing. they tend to have crappy headroom and you wouldn't want to screw up a mix because of that thing. get the mix good and apply it after. they really can work wonders if you use it right.

Crappy headroom? Dude.....just shut up please.

Fucking headroom.................I hate how people throw that word around and generally have NO idea what it means in audio circles!

 

[SouthSIDE Glen]

Actually, sidechaining an EQ to a comp does not really act like a multiband compressor.

Yeah, I meant that as a *very, very* rough analogy, only in that they both (in a general way) were a way of getting the compressor to react differently at different frequencies. Obviously the details of how they do that (and why one would want them to do that) are entirely different.

An interesting idea, though. I may have to play with it when I get an opportunity and hear what kind of effect(s) it has. Though I have to admit that I don't do a whole lot of sidechaining in general - that's not a judgement right or wrong on sidechaining, just the way the hand usualy plays for me - I'm intrigued by the idea you describe.

G.

 

[danny.guitar]

Crappy headroom? Dude.....just shut up please.

Fucking headroom.................I hate how people throw that word around and generally have NO idea what it means in audio circles!

Yea I was kind of wondering what he meant by headroom when applying the term to an EQ?

But then again I don't really know what everyone is talking about here so I figured I wouldn't say anything.

 

[FALKEN]

Crappy headroom? Dude.....just shut up please.

Fucking headroom.................I hate how people throw that word around and generally have NO idea what it means in audio circles!

WHOA........what makes you think I don't know what it means?

I have a 15 band graphic and the thing clips at anything past 0db on the meter. Sometimes with everything set to unity the thing will clip with a hot track. The output has to be turned down from the get-go, and any eq'ing will definitely require that the output be turned down, or the thing clips like mad. ugly, nasty clipping. Even my massive does this on occasion, and it technically has headroom up do +28 db, which is pretty high. many of the more inexpensive boxes might only have headroom up to say +14 db...not very high at all for a typical mix. any sort of eq curve can have more drastic results than you intended or might even be hearing. its important to listen with focus that you are not adding any unwanted distortion through the output circuitry. Even cutting can make the resulting waveform louder than the original. When mixing there is so much going on and so much to pay attention to that it might be easy for a novice to miss something like this. I hate it when an otherwise perfect mix gets completely ruined due to a simple oversight that you don't notice until the next day. my suggestion was simply to leave the thing off until the mix was done. sheesh......

 

[SouthSIDE Glen]

The thing to remember is the wide effective bandwidth of graphic filters (indeed of all frequency-specific filters of respectable Q), including the 50% or so of energy affected outside the bandwidth as defined by the Q value. A boost of, say, 10dB at the center frequency results in a total energy boost of far more than 10dB cumulative across the affected bandwidth. (I've never been very good with calculus, but for those who know their way around integrals, the calcualtion of the actual value should be relatively simple.)

If a filter circuit has a theoretical design for 24dB of potential gain, that only means that it can theoretically cleanly boost a pure sine (or square or sawtooth or...) at that frequency by 24dB. But that does not include considerations for the output stage, general signal chain gain staging, and, as demonstrated above, a complex waveform with plenty of energy distributed across the filter's operational spectrm.

Even cutting can make the resulting waveform louder than the original.

Huh????? Unless you're talking about a mis-calibrated or mis-set output stage, the filtering itself should not cause this to happen *at all*.

And to try and bring this back towards the center of the discussion, my only original point was that any EQ that does not allow one to surgically address problematic frequencies is - by definition - not the ideal tool for corrective pre compression EQ, regardless of the quality of the EQ itself.

One can have the absolute best, rubber-gripped, torque driven, carbon steel-tipped Phillips-head screwdriver, but that doesn't make it anywhere near the best tool to use to drive a Torx-head screw.

I'm not sure I understand why such a fundamental concept is so controversial sounding?

G.

 

[danny.guitar]

WHOA........what makes you think I don't know what it means?

I have a 15 band graphic and the thing clips at anything past 0db on the meter. Sometimes with everything set to unity the thing will clip with a hot track. The output has to be turned down from the get-go, and any eq'ing will definitely require that the output be turned down, or the thing clips like mad. ugly, nasty clipping. Even my massive does this on occasion, and it technically has headroom up do +28 db, which is pretty high. many of the more inexpensive boxes might only have headroom up to say +14 db...not very high at all for a typical mix. any sort of eq curve can have more drastic results than you intended or might even be hearing. its important to listen with focus that you are not adding any unwanted distortion through the output circuitry. Even cutting can make the resulting waveform louder than the original. When mixing there is so much going on and so much to pay attention to that it might be easy for a novice to miss something like this. I hate it when an otherwise perfect mix gets completely ruined due to a simple oversight that you don't notice until the next day. my suggestion was simply to leave the thing off until the mix was done. sheesh......

I'd consider getting a new EQ then...

There's no reason a cut should cause something to clip unless the master volume on the Equalizer is set real high, and why would it be?

And if minor adjustments (+3db?) are causing it to clip then I'd say you're mixing too hot, or your tracks were recorded too hot.

The headroom comes from the tracks you're mixing (there should be enough room for you to make EQ adjustments, etc.). That's what (I thought) the term headroom means...

 

[FALKEN]

look. my master recorder has headroom up to +24 db. If I am using all of that headroom, then the moment I send the mix into a piece of gear that only has +14 db, I am probably going to get some distortion. Its a fact.

Eq cuts can cause a signal to become louder. do some research.

http://www.tcelectronic.com/media/nielsen_lund_2003_overload.pdf

 

[Ford Van]

WHOA........what makes you think I don't know what it means?

I have a 15 band graphic and the thing clips at anything past 0db on the meter. Sometimes with everything set to unity the thing will clip with a hot track. The output has to be turned down from the get-go, and any eq'ing will definitely require that the output be turned down, or the thing clips like mad. ugly, nasty clipping. Even my massive does this on occasion, and it technically has headroom up do +28 db, which is pretty high. many of the more inexpensive boxes might only have headroom up to say +14 db...not very high at all for a typical mix. any sort of eq curve can have more drastic results than you intended or might even be hearing. its important to listen with focus that you are not adding any unwanted distortion through the output circuitry. Even cutting can make the resulting waveform louder than the original. When mixing there is so much going on and so much to pay attention to that it might be easy for a novice to miss something like this. I hate it when an otherwise perfect mix gets completely ruined due to a simple oversight that you don't notice until the next day. my suggestion was simply to leave the thing off until the mix was done. sheesh......

Well, if you would quit running a +4 output to it's -10 input, you probably would notice a significant increase in your alledged "headroom"!

 

[Ford Van]

do some research.

http://www.tcelectronic.com/media/nielsen_lund_2003_overload.pdf

Maybe YOU sir should read a little closer!

 

[SouthSIDE Glen]

look. my master recorder has headroom up to +24 db. If I am using all of that headroom

Well, question #1 is why on God's green Earth would you even want to be even coming close to +24 with your signal? Even counting 0.5ms transients and even counting wanted preamp saturation, etc. etc. etc., that's running your signal awfully damn hot.

And that's also what turning down the output gain is for: your not supposed to be sending a +24 signal, or even a +14 signal, out to the next box. Try a little gain stage control and you'll still be able to saturate all you want without having to worry about signal chain "headroom".

Eq cuts can cause a signal to become louder. do some research.

I just read that (quite excellent) paper your referenced, and I believe you are erroneously interpreting what it says. It says (boldface added editorially)

"Removing part of the spectral content of a broad band signal is very likely to cause a peak level increase, despite the reduced energy in the resulting signal."

It goes on to give a LF square wave and the use of filtering to counter DC offset as an example, along with charts of the effect.

An increase in peak levels combined with a reduction in overall energy does not make something louder. Take a look at Figure 3, which is their example of this effect. The resulting skewed dotted line is not going to sound louder than the original solid square wave, even though it's initial peaks are of greater amplitude.

But OK, I'll take your point assuming that you didn't actually mean "louder", you just meant higher theoretical peak amplitudes. OK, that's an interesting technical issue, and I'll grant that possibility for the sake of discussion.

But it's an issue that you shouldn't need to worry about if you kept the +14 and +24 nonsense inside the boxes and didn't send them out of the output stage to the next input that hot. You might just as well be mixing +4 and -10 if you're going to let your gain staging get that unwieldly.

G.

 

[FALKEN]

Well, question #1 is why on God's green Earth would you even want to be even coming close to +24 with your signal? Even counting 0.5ms transients and even counting wanted preamp saturation, etc. etc. etc., that's running your signal awfully damn hot.

Glen why is everything you say completely wrong?

if I am using 24 db of headroom on my recorder it doesn't mean my signal is "awfully damn hot". It means the exact OPPOSITE. It means that my RMS is "awfully damn low" because the peaks are +24 db above the average, which is still at 0. Maybe I am mixing classical or something. maybe I just really like dynamic range. who the hell cares.

And that's also what turning down the output gain is for: your not supposed to be sending a +24 signal, or even a +14 signal, out to the next box. Try a little gain stage control and you'll still be able to saturate all you want without having to worry about signal chain "headroom".

Thanks for the lesson on how to use a VOLUME KNOB, mr. KING OF THE OBVIOUS. I think you should get a HR.com AWARD for explaining this to me. You are a god.

I just read that (quite excellent) paper your referenced, and I believe you are erroneously interpreting what it says. It says (boldface added editorially)

"Removing part of the spectral content of a broad band signal is very likely to cause a peak level increase, despite the reduced energy in the resulting signal."

It goes on to give a LF square wave and the use of filtering to counter DC offset as an example, along with charts of the effect.

An increase in peak levels combined with a reduction in overall energy does not make something louder. Take a look at Figure 3, which is their example of this effect. The resulting skewed dotted line is not going to sound louder than the original solid square wave, even though it's initial peaks are of greater amplitude.

But OK, I'll take your point assuming that you didn't actually mean "louder", you just meant higher theoretical peak amplitudes. OK, that's an interesting technical issue, and I'll grant that possibility for the sake of discussion.

OF COURSE WHAT I MEAN IS HIGHER PEAK AMPLITUDES. WE'RE TALKING ABOUT HEADROOM RIGHT?? WHAT ELSE WOULD I MEAN?? AN INCREASE IN CHOLESTEROL? AND WHY IS THIS HIGHER PEAK AMPLITUDE "theoretical" ????? why would you use a word like that to dis an entire paper? one word and who are you??

But it's an issue that you shouldn't need to worry about if you kept the +14 and +24 nonsense inside the boxes and didn't send them out of the output stage to the next input that hot. You might just as well be mixing +4 and -10 if you're going to let your gain staging get that unwieldly.

G.

have you EVER MIXED ANYTHING???? things change and you make adjustments. you make adjustments on top of adjustments on top of adjustments. Sure you can set your boxes to play nice. But as soon as you make an adjustment you might be in distortion city and not noticing!! because you're intensely focused on the vocal or bass line or something you didn't notice that now the snare is clipping a bit. I would never have expected anyone to argue the point that those crappy graphic eq's can be nothing but distortion boxes when used incorrectly. I think I'm done here.

 

[FALKEN]

Maybe YOU sir should read a little closer!

please explain ?

you really need to stop dissing people without explanation!!

 

[SouthSIDE Glen]

if I am using 24 db of headroom on my recorder it doesn't mean my signal is "awfully damn hot".

It does if the peaks are at +24.

Maybe I am mixing classical or something. maybe I just really like dynamic range. who the hell cares.

I've heard your music, remember? I even complimeted you on your work. What you do is about as far from classical music as Tolouse Latrec is from being an MBA basketball player. So don't lay that boogie woogie on me.

And you know damn well that I am a crusader for dynamic range on this board. But you don't need (or want) to push your gear to the limit to get all the range that you need. What the hell do you need 120 or 130 or 140 dB of dynamic range for? That's WELL beyond the range of what the human ear can deal with, WELL beyond what's required to capture even the subtlies on a classical composition, and that's what you got when you are riding +24 on your meters.

Thanks for the lesson on how to use a VOLUME KNOB, mr. KING OF THE OBVIOUS. I think you should get a HR.com AWARD for explaining this to me.

Apparently you needed the lesson, after complaining that there's a problem because one device in the chain has less "headroom" than another.

OF COURSE WHAT I MEAN IS HIGHER PEAK AMPLITUDES. WE'RE TALKING ABOUT HEADROOM RIGHT?? WHAT ELSE WOULD I MEAN??

Geez, settle down, Castro. You said "louder". That is entirely different from "peak amplitude". Geez, are you going to give me another award for having to re-explain yet another fundamental to you?

And I then cut you the slack that it was probably just an honest semantic mis-statement on your part that you said "louder" and that you were indeed referring to amplitude. For that you jump down my throat?

WHY IS THIS HIGHER PEAK AMPLITUDE "theoretical" ????? why would you use a word like that to dis an entire paper?

Or for crying out loud, read what I said, not what you want to believe I said. I said it was an excellent paper. The "theoretical' was referring to the fact that the demonstration they give is an a very limited, labratory-like setup (for good reason; I'm not dissing that, before you try and burn my ass with your flamethrower again, Fidel.) It was showing the effect of high-frequency cuts on a low-frequency square wave. Low frequency cuts on a high-frequency complex waveform won't necessarily have the same result. So while the theroy they demonstrate is viable, it is not applicable in a blanket-type fashon to all EQ cuts on all program material in general.

I would never have expected anyone to argue the point that those crappy graphic eq's can be nothing but distortion boxes when used incorrectly.

The only argument there must be in your own head, because I never said anything along those lines.

All I said was that low-resolution graphic EQs are not the best tool in the box to use for corrective pre-comp EQ because they cannot target the specific problem frequencies that need correction. Not because they're "crappy", not because of anything to do with "headroom" or anything else for that matter. Only because BY DESIGN they are not meant for that kind of work, any more than a hammer is meant to be used to drive in a screw.

Then you go off on me like I'm attacking your $4000 Manley - which isn't even a graphic EQ, so I don't know where THAT probelm came from - as if I were commenting on your dick size. And somehow that turns into a silly fucking argument about headroom and white papers on esoterinc types of distortion, none of which anybody on this board needs to give a rat's ass about until LOOOONG after they've resolved a million other far more important and more fundamental problems they have. Like learning about how to not abuse levels.

Happy fucking Thanksgiving, Castro boy. Thank you for making it less enjoyable for all of us.

I think I'm done here.

We can only pray.

G.

 

[FALKEN]

dude. thanksgiving is over. if some dude on an internet post "ruined your thanksgiving".....whatever, its not my problem. Nobody was talking about dick size. I was making a point which still has not been directly addressed which you have now agreed with but still argumentatively which I don't understand in the slightest. if anything I think you are worried about dick size because any time in any thread someone says something contrary to what you say you take it personally. like you might be afraid to learn something because it would imply you didn't already know something.

and you constantly assume that because some technique doesn't work for you then it definitevely does not work. it becomes a fact, not an opinion. and its never anything from outside the box or an angle that might not have been considered. All I did was through out a separate thought and you take it that I am arguing against yours.

And by the way, it would not be uncommon to be running one of my mixes at 0 on the tape meters and be using all of my available headroom. My music is extremely dynamic, thank you very much. If you don't think that running an equalizer (ANY equalizer) over the mix bus during mixdown can limit that, then that's your problem.

 

[SouthSIDE Glen]

All I did was through out a separate thought and you take it that I am arguing against yours.

If that is indeed what happened, I apologize for misunderstanding you. But cut me some slack, Fidel (since I don't know your real name); here's the way things went when we first crossed swords (as it appears on my screen, anwyway):

Post #9: I thought it sounded (at the time) like maybe Adiel was talking about the use of corrective pre-comp EQ, and I said that a 10-band graphic was not the right tool for that particular job, followed by the *disclaimer* that "there's more than one way to skin a cat."

Post #10: You reply (at least it looked like a reply based upon position and context): "my massive passive probably doesn't have more than 10 points on it. but I haven't counted." Now to me, that appeared as though you were equating your MMP with a 10-band graphic. You can - I hope - understand how I would find that both confusing and erroneous.

We went back ad forth for a while hashing out a few things fairly reasonably, we both made mistakes and corrected them, including, in post #15 yet more disclaimers on my part in the form of "Again, YMMV, IMHO, QUIDPROQUO, ETC. " The "again" being because I had used a couple of IMHOs in the previous intervening posts. (Am I really always saying I'm right and everybody else is wrong and that my opinions are godlike? It sure doesn't look that way here - and I can guarantee you I sure don't think that way. How many more disclaimers must I use???)

Then comes post #19, in which you state "But I still stand by my statement that you can do a lot with a 10 band. its not the gear its how you use it and what your goals are." My reply, in post #20, was "And what if your goals are to correct problem frequencies in the mix before comping?", which was the thrust of my point all along, BTW. You never answered.

Next thing we know we're talking about headroom, which is an ENTIRELY different, way, way, off-thread topic. And where twice, in posts 25 and 28, you referred to EQ cuts making the signals "louder". Now we both now know you meant to say that in some cases it could increase the peak amplitudes of some frequencies, and I never disagreed with that. But based upon your *twice* using the term "louder" - which means something altogether different - I disagreed with you.

Now if you want to go ahead and lay blame on me for being a cocky asshole or something like that based upon the above account, that's your right to do so. But you'll have to excuse others for not being able to quite see it that way.

And by the way, it would not be uncommon to be running one of my mixes at 0 on the tape meters and be using all of my available headroom. My music is extremely dynamic, thank you very much.

And I don't doubt that for a minute. I also don't doubt that you could very well eliminate many of the distortion and headroom problems you describe without comprimising the quality of your productions by cranking it back a bit. There's no need to be running your analog RMS at a full 0dBVU...*especially* if your peaks are that dynamic.

If you don't think that running an equalizer (ANY equalizer) over the mix bus during mixdown can limit that, then that's your problem.

What I said was that if you kept the massive use of above zero headroom "inside the boxes" (post 31), that you wouldn't have the problem. In other words, play the gain stage game between the boxes, and inside the boxes use however much you need.

Look, just because a box has a design limit of 14dB of "headroom" doesn't mean it doesn't have enough dynamic range between the noise floor and the ceiling to handle the full signal. Even with distortions caused by gains or cuts or monkeys flying out of my butt, proper gain staging on the I/O and proper positioning of the signal within the useable range of the box should - except in the very most extreme cases - give your material plenty of room to breathe.

G.

 

[xstatic]

What i don't understand is how you could just say that a 10 band EQ has crappy headroom. That is a pretty broad statement that is far from true. YOUR 10 band EQ may, but how can you generalize that statement to include all 10 band EQ's? I happen to know that the API 10 band EQ's have plenty'o'headroom and seriously kick some ass. On kick drum there is nothing like it. The bottom line is that you don't really know what the headroom for the specific 10 band EQ in question is. Do you even know which 10 band EQ is being discussed here?

In the end, if it isn't capable of +22 or so, than it is just a bad design, or designed for something completely different. If you are coming close to that +22 than there is an excellent chance that you are adding a ton of gain somewhere. This however would not be a problem with the EQ though. If you hit the headroom limit of the EQ, than it is a mix problem.

 

[adiel]

ten band eq

This makes me wonder if the EQ wasn't just sidechained to the comp to help tune the compression itself and not for actual EQ'ing.

Yes!!! thank you all for leading me in the right dirrection, i found that this proccess is call pumping. a groove pump to the kick & bass freqs by sidechain.