EQ Dummy, please help me understand

[progmr]

Can anyone please give confirm if my understanding of a 4-band parametric EQ is correct.

From what I have read it appears that it permits one to select a frequency (in this case up to 4) and either boost or decrease the gain on it and those close to it (depending on the envelope spacing?).

Now my question is:

I am trying to EQ a 12 string guitar who's middle B is extremely overpowering. If my assumption is correct my 4 bands should be between 82 (low E) and about 1045 (high C) and I can decrease the gain of a band at 246hz with a tight envelope to remove the offending note. Is this correct? Going forward in my understanding is it logical to assume that I need only set the EQ freqencies for range of the instrument I am trying to modify? What happens to those frequencies (assuming my 12 string situation above) that are not covered by envelopes set, are they assumed to be zero? My machine goes up over 2500 HZ which is out of frequency range for the guitar...so can I assume that I would never use this high of a setting?

Thank you in advance.

 

[Son of Mixerman]

The human ear roughly hears from 20hz to 20khz. When recording a instrument, you must realize that every instrument generates harmonics that extend beyond the primary frequencies of the instrument. To cut these harmonic would significantly reduce the natual sound of he instrument. You could cut the "B" with a narrow Q to reduce the offending note, but its more musical to have a Q that about 1.4 wide so that you don't get phasing artifacts from the EQ, too narrow can make things worse, cut the 246hz at increments of -1.5 db while sweeping the Q to hear the most musical setting. You might want to cut off everything below 80Hz so you don't hear body motions while playing. You need to study Harmonics a little so you understand the total sonic picture.

SoMm

 

[Ethan Winer]

Prog,

> I am trying to EQ a 12 string guitar who's middle B is extremely overpowering. <

Son of MM is correct that EQ cannot usually lower a note that's too loud, because most notes are comprised of a fundamental pitch and harmonics. But in your case it's worth a try because it may well be just the fundamental, or just the second harmonic, that's stricking out too loud. Sometimes compression can help this too when EQ fails.

If you'd like to read a detailed primer about equalizers, see "The art of equalization" about 1/4 way down the list on my Articles page:

www.ethanwiner.com/articles.html

I do need to correct one thing SOMM said: The notion that EQ adds "phasing artifacts" is a myth. Yes, phase shift is fundamental to how EQ works, but it's totally benign. As proof, play the finest sounding CD you have ever heard. Sounds great, right? You can be sure that CD was mixed and mastered using standard EQ having phase shift.

In fact, when the goal is to lower the level of a single frequency with EQ, a very narrow band is the only way to do that properly.

--Ethan

 

[Son of Mixerman]

Re: Re: EQ Dummy, please help me understand

I do need to correct one thing SOMM said: The notion that EQ adds "phasing artifacts" is a myth. Yes, phase shift is fundamental to how EQ works, but it's totally benign. As proof, play the finest sounding CD you have ever heard. Sounds great, right? You can be sure that CD was mixed and mastered using standard EQ having phase shift.

In fact, when the goal is to lower the level of a single frequency with EQ, a very narrow band is the only way to do that properly.

--Ethan

Mmnnnnn?

Myth? Maybe.
Not exactly....maybe we can call this a partial correction or more of a clarification. The EQ's used for mastering is quite a different animal than most eq's you would find in a home studio or most project studio's. If you have a good eq as in GML, Neve, SSL your more than likey not going to hear any artifacts...depends on the AE. If you are using a low end digital or analog eq you are more likely to hear the artifacts. The artifacts are more evident obviously if they are in the 2k to 4k and your boosting that region 6db with a Q of .4...but if your in the 150Hz to 400Hz range lowering 3 db with a Q of .4, the artifacts may not be as obvious, its all depends on your ears and the quality of the eq and the degree in which boost or reduction and in which frequency range.
Not all dynamic processors are created equal. Therefore ones mans myth may be another man nightmare. Ill trade you! Your myth for my nightmare


SoMm

 

[chessrock]

I would think a multiband compressor would get you where you want to go a lot better than EQ.

 

[Outlaws]

Get a diferent guitar.

 

[13th_Omen]

I think they should come up with a guitar that has 6 outputs, one fro each string, so you could track each string. Seems like it would be fun!

 

[Outlaws]

I think they should come up with a guitar that has 6 outputs, one fro each string, so you could track each string. Seems like it would be fun!

They do. Its called a Variax...I think.......

 

[Ethan Winer]

Re: Re: Re: EQ Dummy, please help me understand

SOMM,

> The EQ's used for mastering is quite a different animal <

So far as phase shift is concerned, how are they different?

> If you have a good eq as in GML, Neve, SSL your more than likey not going to hear any artifacts <

What kind of artifacts? What do these artifacts sound like?

> Not all dynamic processors are created equal. <

Of course I was addressing EQ, not dynamics.

I don't mean to single you out, but it's become common for people to point the finger at "phase shift" as the cause of whatever shortcomings they have with their productions. In my experience phase shift alone is totally benign and has no "sound."

--Ethan

 

[Middleman]

I would take Chessrocks approach to the problem.

A multiband to focus in on the errant note or a regular compressor just to even the sound out across the strings.

A parametric could be used but, if the note is that bad, another guitar would be my choice.

 

[Son of Mixerman]

Re: Re: Re: Re: EQ Dummy, please help me understand

What do you mean "no meaning to single me out"?
heh heh... You accidentally single yourself out Ethan as soon as you say "in my experience"
Experience is rarely global so generalisations are risky, personal experiences shared are always enlightening.

SOMM,

> The EQ's used for mastering is quite a different animal <

So far as phase shift is concerned, how are they different?

Filters and crossovers between the bands. Phase shift is a relationship correct. In analog component who have to choose components that segregate signals into bands or banks, the accuracy of splitting is not perfect so the overlap can vary from band to band. If your are boosting a band some of those frequencies in the overlap can shift. Analog is more forgiving as far as Ive experienced, maybe its the slew rate, but cheaper digital eq's the phenomina is more evident.

What kind of eq do you normally use and have you actually tried to look for artifacts? Ive only recently discovered this artifact issue with a particular software eq.

> If you have a good eq as in GML, Neve, SSL your more than likey not going to hear any artifacts <

What kind of artifacts? What do these artifacts sound like?

Ive heard it described as pre-echo by Dave Collins and Bob Katz, I usually hear a slighty phasy nastiness in the upper frequences. It may be that Im more phase sensitive which makes my experience different, I know Bob Katz is jitter sensitive

> Not all dynamic processors are created equal. <

Of course I was addressing EQ, not dynamics.

Sure..traditional description of eq's and compressors, expanders are separated. But the end result of a VCA limiting a specified (wideband) output is unusually similar to specified (narrowband) eq cut. Both result in a reduction in output. Specific verses general, even though the compressor can change the waveform differently, that is a choice of design.
Maybe its because I use to deal with radar stuff that I make less of distinction.

I don't mean to single you out, but it's become common for people to point the finger at "phase shift" as the cause of whatever shortcomings they have with their productions. In my experience phase shift alone is totally benign and has no "sound."

--Ethan

Phase shift has no sound? Flangers, chorus' and delays all operate using varying levels of phase shift and obviously have a sound. Phase cancellation might not have a sound if your processing sine waves.
Maybe I haven't been around homerecordists as much as you, but I rarely hear of phase shift as being an excuse for shortcomings...If someone says copied track and now it sounds thin in my DAW, I usualy suspect phase problems, but thats easily fixed by a 35ms nudge right. Usually the only place I hear people having phase problems it in cymbals mic'd with O/H's.

Im not sure it can be benign on its own when the ear is highly sensitive to phase, it can cause vomiting if it is too extreme. Your brain uses phase for finding where up is and where left is. Radar systems use phase for tracking and targeting. But when masked by 48 tracks of other pieces of audio, yeah..artifacts for the most parts are pretty benign.


SoMm

 

[Ethan Winer]

Re: Re: Re: Re: Re: EQ Dummy, please help me understand

Son,

> In analog component who have to choose components that segregate signals into bands or banks, the accuracy of splitting is not perfect <

I've been talking only about EQ, not active crossovers which are a whole 'nother category.

> What kind of eq do you normally use and have you actually tried to look for artifacts? <

I've used many kinds of EQs over the past 40 years. Analog, digital, plug-ins, cheap ones, good ones - even some I designed and built myself.

> Ive heard it described as pre-echo <

I have never heard a pre-echo when using any EQ. Have you?

> I usually hear a slighty phasy nastiness <

We're getting closer...

> Phase shift has no sound? Flangers, chorus' and delays all operate using varying levels of phase shift and obviously have a sound. <

Bingo! This is the crux of everyone's misunderstanding. Phase shifter and flanger effects use phase shift internally in the same way an EQ does, but what they really do is alter the frequency response. That is what you hear when you run a track through a flanger. It's the comb filtering that drastically changes the frequency response, not the phase shift per se.

I have heard people claim they can hear phase shift in an EQ because when they boost the treble they hear a "phasey" sound. So they wrongly assume that sound is the damaging phase shift everyone talks about. In truth, what they are hearing is comb filtering that was already present, but subdued. When the treble is boosted, the comb filtering becomes more apparent. The EQ did not add the phasey sound, it merely made it more apparent.

> I rarely hear of phase shift as being an excuse for shortcomings <

I hear this all the time from newbies, and also from working professionals who should know better. The "problems" caused by phase shift have been repeated so many times by magazine writers and audio salespeople that it's now commonly accepted, even though there's not a shred of truth to it.

> copied track and now it sounds thin in my DAW, I usualy suspect phase problems <

Yes, but again it's not the phase shift you are hearing. Rather, it's the change in frequency response when the normal and shifted versions are combined. This is a huge distinction!

--Ethan

 

[Ethan Winer]

EQ and phase - the final word

Folks,

I clarified and consolidated my last few replies, and added them to an existing mini-article about phase shift in equalizers. Here's the link:

www.ethanwiner.com/EQPhase.html

--Ethan

 

[Blue Bear Sound]

Not sure I agree with all of it... but here's another interesting article:

http://www.rane.com/note115.html

 

[Son of Mixerman]

Re: Re: Re: Re: Re: Re: EQ Dummy, please help me understand

I've been talking only about EQ, not active crossovers which are a whole 'nother category.

The role of the crossover is to divide the audio frequency spectrum into two or more bands, and distribute
each band to the appropriate driver (e.g. low frequencies to the woofer and high frequencies to the tweeter).

You said you designed eq's but yet your not understanding the use of crossovers in eq's. Is it semantics on definitions? To divide the audio into the separate bands you need some sort of passband filter, which is just like a crossover. Depending on how you handle getting the different bands into the right section of the processor. There are IIR and FIR filters in a Linear Phase Eq that try to reduce the myth.

> Ive heard it described as pre-echo <

I have never heard a pre-echo when using any EQ. Have you?

Yes... and so have other people. Bob Katz, Dave Collins....

> I usually hear a slighty phasy nastiness <

Its a hard thing to describe, thats as close as I can get. You know it when you hear it...its the nasty part that gets you, its distortion I think in the overlaps.

> Phase shift has no sound? Flangers, chorus' and delays all operate using varying levels of phase shift and obviously have a sound. <
Bingo! This is the crux of everyone's misunderstanding. Phase shifter and flanger effects use phase shift internally in the same way an EQ does, but what they really do is alter the frequency response. That is what you hear when you run a track through a flanger. It's the comb filtering that drastically changes the frequency response, not the phase shift per se.

Can you tell me how a comb filter works without a phase shift? Did you know that comb filtering is an artifact? Often used on purpose for FX manufacturers. There isn't a frequency responce change, maybe a blurring and masking problem.

I have heard people claim they can hear phase shift in an EQ because when they boost the treble they hear a "phasey" sound. So they wrongly assume that sound is the damaging phase shift everyone talks about. In truth, what they are hearing is comb filtering that was already present, but subdued. When the treble is boosted, the comb filtering becomes more apparent. The EQ did not add the phasey sound, it merely made it more apparent.

Chorus units often are comb filters.
EQ's do shift phase.

See farther down for people of authority to refer this to.

> I rarely hear of phase shift as being an excuse for shortcomings <

I hear this all the time from newbies, and also from working professionals who should know better. The "problems" caused by phase shift have been repeated so many times by magazine writers and audio salespeople that it's now commonly accepted, even though there's not a shred of truth to it.

Since you've heard from the newbies.

You can contact Bob Katz at his site and discuss the incorrectness of Chapter 8 section IVof his book "Mastering Audio" as well as talk to Daniel Weiss about perpetuating a myth. You Also might want to contact George Stimson and have him make corrections to his book "Introduction to Airborne Radars" I assure you these people are not salespeople and are authorities on the subject. They are the sources to the info that helped explain the anomalies I "thought" I heard.

> copied track and now it sounds thin in my DAW, I usualy suspect phase problems <

Yes, but again it's not the phase shift you are hearing. Rather, it's the change in frequency response when the normal and shifted versions are combined. This is a huge distinction!

--Ethan

Phase is the degree to which the cycles of a wave or signal coincide with those of a reference signal or wave of the same frequency. Its a time based problem and not a pitch problem. It may fool some people as being frequency changes when actually its a reduction of the power. Amplitude reduction. The only way for the signal to change frequency is if a local oscillator has been added. If thats true, it still requires a phase shift to sum the diff of the local oscillator. Like an arpeggiating flanger or pitch transposer. You will hear artifacts on those as well.

A copied track that is exactly located peak to peak so that they are in phase will be louder, or nulled if out of phase 180 degrees. Anything in between works its way to chorusing and delay, or "phasing".

I think I've covered everything?

I refer all further questions to Bob Katz.


SoMm

 

[Ethan Winer]

Blue Bear,

> Not sure I agree with all of it... but here's another interesting article <

Geez, what's not to agree with?

I happen to be friends with John Roberts, the engineer credited with insipiring that article. John and I shared a house many years ago, and both wrote for R-e/p magazine, and we were very much in agreement about the importance of myth debunking. Anyway...

I agree with that article, though it addresses mainly EQ for room correction, not tracking / mixing / mastering uses. For example, much of it discusses problems associated with combining multiple narrow bands. This is not a factor with traditional console and parameteric equalizers, unless you're manipulating nearby frequencies at the same time.

Often people believe things that defy basic scientific fact. That's okay - everyone is entitled to believe whatever they want. But it's always good to keep in mind Scotty's famous phrase from Star Trek, "Captain, I canna change the laws of physics."

--Ethan

 

[Blue Bear Sound]

It's not the content so much that I question -- more the implied "sanctioning" of using an EQ to "flatten" a room in the first place...!

Being a signal chain minimalist, I can't get over putting an EQ in the monitoring chain!

 

[Ethan Winer]

Son,

> You said you designed eq's but yet your not understanding <

No need to challenge what I know or don't know. Let's stick to the issues and the rest will sort itself out, okay?

> To divide the audio into the separate bands ... <

Maybe for a 1/3 octave graphic EQ, but those are not used much in recording. Also, the graphic EQ designs I've used don't work that way at all. Rather, they include multiple parallel resonant circuits within the feedback loop of a single op-amp.

> Bob Katz, Dave Collins ... Daniel Weiss ... George Stimson <

That's the classic "Argument from authority." As in "I can't explain it but these experts agree with me." If you'd like to invite any of those fellows here to discuss this, I'd very much welcome the opportunity!

> Can you tell me how a comb filter works without a phase shift? <

Who said it does? Time delay is the basis for flanger effects, and phase shift is the basic mechanism for a phaser effect. Comb filtering has many causes including acoustic cancellation in the air. That's one of the points I made in my revised EQ-Phase article linked above.

> Did you know that comb filtering is an artifact? <

An artifact of what?

> There isn't a frequency responce change, maybe a blurring and masking problem. <

I honestly have no idea what you are saying!

> Phase is ... a time based problem and not a pitch problem. It may fool some people as being frequency changes when actually its a reduction of the power ... The only way for the signal to change frequency is <

Huh? EQ does not shift pitch. It changes the frequency response. Again, what are you trying to say here, and how does it relate to the audibility of phase shift in equalizers?

--Ethan

 

[chessrock]

I have no idea what the heck you guys are arguing about . . . but I'm thankful there are people like Ethan and SOMM on this earth to worry about this stuff. God bless you guys.

 

[Ethan Winer]

Bear,

> more the implied "sanctioning" of using an EQ to "flatten" a room in the first place...! <

Oh yes, I absolutely agree with that!

> Being a signal chain minimalist, I can't get over putting an EQ in the monitoring chain! <

More important, EQ is absolutely the worst way to address room problems, in small home-sized rooms anyway. Forgetting the addition of another device, room EQ usually makes things worse. And clearly it cannot solve the main issues such as severe low frequency nulls from acoustic interference, and overall mud caused by excessive LF reverb.

--Ethan