Head-Fi.org › Forums › Equipment Forums › Sound Science › Does Equalizing actually distort music?
New Posts  All Forums:Forum Nav:

Does Equalizing actually distort music? - Page 3

post #31 of 53
Quote:
Originally Posted by bigshot View Post

Yes, maybe you like ketchup. But it's covering up what the chef created for you. 

 

Agreed 100%, but as you said some folks like ketchup.

We can view it as a waste of what was originally offered but at the end of the day it's the satisfaction of the consumer that matters. A good case in point is a family member of mine that loves music almost as much as I do, but they don't invest in anything that would be remotely considered high end for listening. I did a blind test many years ago trying to prove that I wasn't nuts for preferring certain cables over others (at the time Copper vs Silver vs. Carbon Fiber). They could hear a clear difference just as I could but they preferred the cheap out-of-box vanilla copper cables. I could argue how the silver had much better high end and clarity (though too much for me), or the carbon fiber was tighter and more neutral (imho) but the thing is for the aspects I knew they liked in most music types that didn't matter. Yes the cheaper cables subtley filtered the audio, to me lowered the quality, but it was in a way that was closer to what they appreciated. At the end of the day they enjoyed listening to the tracks as much as I did, just for different reasons. 

Now we can argue the quality of EQ, Linear Phase and so on, but that's just the means to an end. If you use EQ to achieve a sound you prefer over non then don't let that little voice of doubt destroy your listening experience. Yes it's muddying it sonically, no it's not what the source intended, but it's yours now and your ears are king.

 

BTW I'm kinda playing devil's advocate here. I don't process audio on my main listening rig at home, all colorisation is from the physical components in the mix and not logical meddling.

But on my PMP I do make use of EQ and BBE to make the sound more enjoyable in non-pristene listening environments.

post #32 of 53
You don't need high end equipment or fancy cables to get accurate sound. Your family member probably is better at discerning accurate sound reproduction than you are. If you like imbalanced sound that's fine for you though.

For me, accuracy is important. I listen to a lot of acoustic music... classical and jazz. I don't want an upright bass to sound like an electric bass, and I don't want the strings in an orchestra to sound all mushed together like a synthesizer. I want an orchestra to sound like it's sitting right there in front of me. I want a piano to sound like it's in the room. My goal is naturalism.

Not easy to accomplish, but not as expensive as chasing random coloration with different combinations of imbalanced high end equipment.

.
Edited by bigshot - 8/29/13 at 1:35pm
post #33 of 53

Okay, bowing out now since this has degraded from an interesting debate to some baseless/weak assumptions. It's easy to believe someone who doesn't agree with you 100% hasn't a clue, it must make your life very easy, so best of luck in your bubble.

post #34 of 53
You're always welcome back to visit Sound Science.
post #35 of 53

Depending on where/how you listen, the acoustic space you're in can also unbalance the music--every location outside an anechoic chamber has its resonances. I've never gotten as deeply serious about this as it's possible to do with the right equipment, but as just a simple example, you can use an EQ judiciously to partically correct for a boomy room resonance that happens to focus right on your easy chair. Of course, if you're determined never to move that chair then you should start looking at room treatments, but audio equipment can also play a role, moving you closer to the mastering engineer's intent, instead of further away.

http://www.soundandvisionmag.com/blog/2012/10/03/rethinking-room-correction

post #36 of 53

Hi everybody. Ill try to answer the question about distortion of equalizers.

 

Equalizers make use of digital filters to attentuate or amplify certain frequencies. There are two kinds of filters that are normally used. They both have pros and cons with respect to distortion.

 

One type is FIR(Finite Impulse Response) filters. They can be designed so that the have linear phase (aka. constant group delay), which is the ones commonly used in audio.

This means that all frequencies will be delayed an equal amount in time going through the filter. So if you hear a bass guitarre that plays a 200 hz with some 1k ovetones, then the 1k part will mix right back into the 200 hz where it is supposed to and not be delayed any more than the 200hz going through the filter. This will (arguably) make the music sound like the 200 hz and the 1k tone come from the same place, as it should.

On the other hand FIR filters tends to have what is known as pre-ringing. For example a drum impact should be heard as a sharp transient followed by some ringing from the drum. The sense of impact is very important for the perception of the drum. When a drum impact is run through a FIR filter, the filter will cause ringing BEFORE the impact of the drum, which sounds very unnatural and it deminishes the sense of impact. This effect becomes worse the more drastic the filter is designed, and the lower the cutoff frequency (assuming a low pass filter in this case).

 

The other type of filter is what is called minimum phase filters, which are of the IIR(Infinite Impulse Response) type.

These filters do not have the pre-ringing issue but they lack the properties of linear phase. So now the 1k overtone will not mix into the 200hz note exactly the way it came in to the filter.

This can make imaging suffer. The more drastic the filter, the more the phase response is distorted.

 

Something that is valid for both types is this: If you make extreme equalizations with sharp drops in amplitude response, you will distort the sound more.

So try to make your equalization is gentle as possible while still adressing whatever you are trying to equalize. For example a very sharp  high order notch filter will reak havoc with the phase response. Also oversampling the digital signal before equalization will help the filters distort far less (this is why you use 192kHz in mastering). However, your equalizer may not support this.

 

Now you may think that if you equalize your music, then it is tainted. But please recognize that all your music have probably been through a lot of these filters already. However, those filters were designed by recording engineers who knows what they are doing. So thats really the key: Know what you are doing.

 

I hope this helps.

 

BR

Casper

 

P.S. For you nerds out there: I intentionally left some technical stuff out and some i simply forgot about. This was from the top of my head.

post #37 of 53

In the real world listening to music in the home, most good quality digital equalizers are completely transparent for all intents and purposes. Distortion is entirely theoretical for human ears because it's well below the thresholds of audibility.

post #38 of 53
Quote:
Originally Posted by bigshot View Post
 

In the real world listening to music in the home, most good quality digital equalizers are completely transparent for all intents and purposes. Distortion is entirely theoretical for human ears because it's well below the thresholds of audibility.

That is simply not true. However, for some purposes it is arguably true. Of you are just making some gentle equalizing and are boosting the base in general with a gentle filter which does not have to steep rolloff, then I am inclined to agree - especially if you're using a minimum phase filter.

 

Butif you are trying to cancel out resonances in your headphones with some notch filters, then it is easy to create plainly audible effects.

Or if you are using a drastic linear phase filter in the bass region, and don't have oversampling, and you are running 44.1 kHz sampling rate, you will have very audible effects.

 

See this link.

 

BR

post #39 of 53

Who does massive EQ spikes? Answer: No one. Any tool can be grossly misused to make it work badly. But that isn't what the tool is for. It's absurd to say that good digital equalizers distort sound, because in the real world, they never get anywhere near to doing that.

 

If headphones require massive notch filters, it's the headphones that suck, not the equalizer. Equalization *corrects* sound, it doesn't distort it. Every transducer can benefit from equalization to one degree or another.


Edited by bigshot - 6/1/14 at 11:13pm
post #40 of 53
Quote:
Originally Posted by bigshot View Post
 

Who does massive EQ spikes? Answer: No one. Any tool can be grossly misused to make it work badly. But that isn't what the tool is for. It's absurd to say that good digital equalizers distort sound, because in the real world, they never get anywhere near to doing that.

 

If headphones require massive notch filters, it's the headphones that suck, not the equalizer. Equalization *corrects* sound, it doesn't distort it. Every transducer can benefit from equalization to one degree or another.

I find it a quite bold assertion that "no one" is notching out peaks in their headphones. It is perfectly reasonable to do. You just have to not over do it which is my point entirely.

You are right; the headphones may not be the best and most are'nt. But it is probably cheaper to use a little EQ in place of bying a new perfect headphone .

 

In digital filtering there is generally no such thing as a free lunch, for the reasons explained in my first post. However, if you're careful you will have a pretty cheap lunch.

That is all I am saying.

 

I suspect you did'nt have a look at the link i sent. It is very informative and makes the pros and cons in EQ very clear and plainly audible.

 

BR


Edited by chlyhne - 6/2/14 at 12:01am
post #41 of 53
Quote:

Originally Posted by chlyhne View Post

 

Or if you are using a drastic linear phase filter in the bass region, and don't have oversampling, and you are running 44.1 kHz sampling rate, you will have very audible effects.

 

Oversampling or high sample rate are not useful when equalizing bass. In fact, they can make the response less accurate with short FIR filters, or with IIR filters that are implemented using single precision floats. Increasing the sample rate is mostly useful for high frequency equalization, especially with IIR filters, which are hard to make accurate (in terms of simulating an analog filter) as the Nyquist frequency is approached.

 

Quote:

Originally Posted by chlyhne View Post

 

One type is FIR(Finite Impulse Response) filters. They can be designed so that the have linear phase (aka. constant group delay), which is the ones commonly used in audio.

This means that all frequencies will be delayed an equal amount in time going through the filter. So if you hear a bass guitarre that plays a 200 hz with some 1k ovetones, then the 1k part will mix right back into the 200 hz where it is supposed to and not be delayed any more than the 200hz going through the filter. This will (arguably) make the music sound like the 200 hz and the 1k tone come from the same place, as it should.

On the other hand FIR filters tends to have what is known as pre-ringing. For example a drum impact should be heard as a sharp transient followed by some ringing from the drum. The sense of impact is very important for the perception of the drum. When a drum impact is run through a FIR filter, the filter will cause ringing BEFORE the impact of the drum, which sounds very unnatural and it deminishes the sense of impact. This effect becomes worse the more drastic the filter is designed, and the lower the cutoff frequency (assuming a low pass filter in this case).

 

The other type of filter is what is called minimum phase filters, which are of the IIR(Infinite Impulse Response) type.

These filters do not have the pre-ringing issue but they lack the properties of linear phase. So now the 1k overtone will not mix into the 200hz note exactly the way it came in to the filter.

This can make imaging suffer. The more drastic the filter, the more the phase response is distorted.

 

FIR filters can be minimum phase, too, they do not have to be linear phase. Also, the phase shifts from minimum phase filtering can actually be beneficial in some cases, when the target is an overall flat response, and the system (headphones etc.) to be equalized is minimum phase or close to minimum phase. In that case, the minimum phase EQ - at least partly - cancels out the existing phase errors, because the inverse of a minimum phase filter is also minimum phase.

post #42 of 53

stv014,

 

I agree completely. Thanks for clarifying that.

 

BR

post #43 of 53

If someone is doing a tight narrow band notch filter to correct for a 40dB spike in the core frequencies, they have a lot more to worry about than a little phase error! You're talking about extreme situations that just don't come up in practice. In mixing, maybe you might run across a situation like that, but if you're mixing, you're working with pro grade equipment that can handle it. For just listening to music, equalization is about correcting for overall imbalances in the response curve or fine tuning room acoustics that room treatment can't get. Doing stuff like that, you never get remotely close to audible distortion. All you get is improved sound. For the purposes of people EQing their home stereos, it just isn't an issue.

post #44 of 53
Quote:
Originally Posted by chlyhne View Post

Hi everybody. Ill try to answer the question about distortion of equalizers.

It's useful to define "distortion" first. Some people consider any change to a signal to be a form of distortion, but generally, for audio, distortion implies nonlinearity. So in that sense equalizers don't distort beyond small amounts of residual distortion in the analog electronic portion of a hardware device.
Quote:
One type is FIR(Finite Impulse Response) filters. They can be designed so that the have linear phase (aka. constant group delay), which is the ones commonly used in audio.

The most common type of equalizer by far is minimum phase. This includes all analog hardware EQs, and most digital plug-ins. Linear phase EQs are more recent, and personally I'd never use one because of the pre-ringing.
Quote:
The other type of filter is what is called minimum phase filters ... This can make imaging suffer. The more drastic the filter, the more the phase response is distorted.

Phase shift per se is not audible, unless it's a huge amount - tens of thousands of degrees at midrange frequencies.

--Ethan
post #45 of 53

Severity depends on the software and the hardware used.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Does Equalizing actually distort music?