How do you judge an EQ?

[martian]

I prefer a true ABX tester and I always do at least 30 tries. If I want 'real' results I do it 100 times (no, I'm not kidding!).

i was trying not to boast, but id usually do 200 at least.

[Aleksey Vaneev]

the interesting thing is that there are clear differences between the 2 that upsample to get the symetrical high freq bell (gliss and pohifopit) and the other 2 (waves ren and sx Q). as said before the 2 that upsample sound harsh and cold! more distortion from the oversampling? christian?

waves and q sound equal to me.

I have not checked the files (server maintenance), but since oversampling in GlissEQ uses polyphase filter, this introduces some phase coloration. I personally always perceived it as better sounding. In many cases I have deliberately enabled high quality, because it made a difference with high-hats, for example.

Generally I dislike such descriptions as harsh/smooth, because in reality every single user hears a different thing. If your bell is high and thin, ANY equalizer is going to sound 'harsh', so I fail to see a general reasoning in marking some plug-in as sounding 'harsh' or 'smooth'. Not to say that almost every single parametric EQ has an almost identical phase response for the given EQ shape. Except those using polyphase oversampling, because that adds some additional phase coloration.

And, of course, this is not about bell curve symmetry alone.

[martian]

thats aleksey for the input!

yes my terminology is really only relative in that instance, and i see its ultimately meaningless. esp since the eq curve was not dialed to enhance anything, but to sound 'harsh', and only for the sake of comparison.

any eq that sounded brighter/harsher/whatever to whomever in comparison would be used differently and compensated in real world application.

apples oranges again. anyway at least from comparison i understand a new difference now.

[bmanic]

I prefer a true ABX tester and I always do at least 30 tries. If I want 'real' results I do it 100 times (no, I'm not kidding!).

i was trying not to boast, but id usually do 200 at least.

The more the merrier but I would never have patience to do 200! Doing 100 is already a "full time job" for me which is why I rarely do it. There is no way that I could concentrate long enough to get a truly valid result with 200.

Cheers!
bManic

[defjamm]

the interesting thing is that there are clear differences between the 2 that upsample to get the symetrical high freq bell (gliss and pohifopit) and the other 2 (waves ren and sx Q). as said before the 2 that upsample sound harsh and cold! more distortion from the oversampling? christian?

waves and q sound equal to me.

I have not checked the files (server maintenance), but since oversampling in GlissEQ uses polyphase filter, this introduces some phase coloration. I personally always perceived it as better sounding. In many cases I have deliberately enabled high quality, because it made a difference with high-hats, for example.

Generally I dislike such descriptions as harsh/smooth, because in reality every single user hears a different thing. If your bell is high and thin, ANY equalizer is going to sound 'harsh', so I fail to see a general reasoning in marking some plug-in as sounding 'harsh' or 'smooth'. Not to say that almost every single parametric EQ has an almost identical phase response for the given EQ shape. Except those using polyphase oversampling, because that adds some additional phase coloration.

And, of course, this is not about bell curve symmetry alone.

hi aleksey, i have 3 questions. won't you implement an even higher oversampling-mode(just like in elephant) in your eqs or do you think the benefit wouldn't be that big? i mean just as an option.
why is there no oversampling in curve-eq? and question 3, when is marquis going to have an oversampling-mode?

[defjamm]

The ABX software that I use can be found here:

http://www.kikeg.arrakis.es/winabx/

Cheers!
bManic

for some reason i used to have a problem running this, may try again now i have new system.

it's rock solid here, thank you bmanic.
you guys are crazy, 100 and 200 blind-tests

[Barbarossa]

Very interesting informations in this thread: http://www.kvraudio.com/forum/viewtopic ... 85#1956785

Seems that usual digital filters don't even match the analog frequency respond cuves!

[defjamm]

the interesting thing is that there are clear differences between the 2 that upsample to get the symetrical high freq bell (gliss and pohifopit) and the other 2 (waves ren and sx Q). as said before the 2 that upsample sound harsh and cold! more distortion from the oversampling? christian?

waves and q sound equal to me.

I have not checked the files (server maintenance), but since oversampling in GlissEQ uses polyphase filter, this introduces some phase coloration. I personally always perceived it as better sounding. In many cases I have deliberately enabled high quality, because it made a difference with high-hats, for example.

Generally I dislike such descriptions as harsh/smooth, because in reality every single user hears a different thing. If your bell is high and thin, ANY equalizer is going to sound 'harsh', so I fail to see a general reasoning in marking some plug-in as sounding 'harsh' or 'smooth'. Not to say that almost every single parametric EQ has an almost identical phase response for the given EQ shape. Except those using polyphase oversampling, because that adds some additional phase coloration.

And, of course, this is not about bell curve symmetry alone.

hi aleksey, i have 3 questions. won't you implement an even higher oversampling-mode(just like in elephant) in your eqs or do you think the benefit wouldn't be that big? i mean just as an option.
why is there no oversampling in curve-eq? and question 3, when is marquis going to have an oversampling-mode?

[justjazz]

...

[wazun]

Here is another little a/b comparison. This time I tried to make it more apples vs apples by closely approximating Wavelab Q eq with the mystery eq. I used whitenoise and voxengo span + my ears to get as close as I could on the music clip (without spending many hours on it). It got pretty close and I think in this clip it's rather obvious that Wavelab Q kills the transients in a bad way (this could be due to aliasing, I don't know).

Wavelab Q

"Mystery" eq

Cheers!
bManic

No matter how often I try, I can only hear a very small difference here, and I can't hear any difference with the transients
Does anyone know a good way to train the ears to hear those "small" differences?

[Christian Budde]

Here is another little a/b comparison. This time I tried to make it more apples vs apples by closely approximating Wavelab Q eq with the mystery eq. I used whitenoise and voxengo span + my ears to get as close as I could on the music clip (without spending many hours on it). It got pretty close and I think in this clip it's rather obvious that Wavelab Q kills the transients in a bad way (this could be due to aliasing, I don't know).

Wavelab Q
"Mystery" eq

No matter how often I try, I can only hear a very small difference here, and I can't hear any difference with the transients
Does anyone know a good way to train the ears to hear those "small" differences?

The sample you quoted was even more easy in comparison to the one I posted. Anyway, my example could be exposed by listening to the bass. I unfortunately forgot to match those frequencies. At 20Hz there was a difference of about 10dB which was noticable for example. The transient behaviour was very subtle.

Kind regards,

Christian

[Zaphod (giancarlo)]

Here is another little a/b comparison. This time I tried to make it more apples vs apples by closely approximating Wavelab Q eq with the mystery eq. I used whitenoise and voxengo span + my ears to get as close as I could on the music clip (without spending many hours on it). It got pretty close and I think in this clip it's rather obvious that Wavelab Q kills the transients in a bad way (this could be due to aliasing, I don't know).

Wavelab Q

"Mystery" eq



Cheers!
bManic

No matter how often I try, I can only hear a very small difference here, and I can't hear any difference with the transients
Does anyone know a good way to train the ears to hear those "small" differences?

It depends on monitors too. Anyway, remember: the eq is NOT the final step in production. After eq you compress, fx... after there is the mastering step: again compress, eq... A small difference here could be a big difference later.

Don't expect a single plug or hardw unit makes a miracle. But good tools with a good knowledge make it.


Now I have a question for you: if you buy a top-end eq (a manley for example) and you match with wavelab eq (you correct frequencies and phase), where is the difference? only few people, as correctly reported by christian can hear the difference. And my question is: why people buy manley? why people buy avalons? If you think a little eq plug is the same of a 5000$ unit (or more) do you think it is only a marketing miracle?

ehi people, no-one here is speaking about the 'time' to get a result? How much time I need to fix a track using plugs and using hardware? Why if I turn a knob in avalons I get a result "immediatly" and with medium level plug I need to correct and correct again, and put other things over and over?
I am the only one here? Or you haven't ever tried hardware before?

my little 2c

[Barbarossa]

It depends on monitors too.

I don't agree. I can hear the differences on my Acer-laptop "speakers", too although i don't have a very good ear.
I think it's more about learning to listen, or better: learning to listen to certain aspects.
I remember when i started to make music: i couldn't even hear a difference between uncompressed and compressed signals.
I think the longer you work with music production, the better you learn to analyze the signal, by listening to certain aspects of the signal.
As a good starting point i wouldn't recommend EQs but compressors. Using one compressor means to learn his sound, get a feeling for it. With that comes a better listening capability of transients and dynamics. If the compressor is well understood, then a copressor comparison is already possible.
The next step could be stereo imaging/reverbs. EQs last.

[Zaphod (giancarlo)]

agreed....
My reference to monitors was for ... beginners.
I think good monitors could help a bit.. And agreed again for compressors. Eq troubles are very subtle. They anyway contribute to the final result, this is the point.

[Christian Budde]

To demonstrate how destroyed transients may also sound, I did pure harm to those poor little transients. Here's the result:

• http://www.savioursofsoul.de/Christian/ ... sients.wav
  http://www.savioursofsoul.de/Christian/ ... high-H.wav

Listen to the bass, one is blurry, while the other is tight. The magnitude remains untouched. If you hear resonant phantom frequencies, then that is happening in your head. What I did was only to delay some frequencies to destroy their transient behaviour. The transient structure is still intact above 1.2kHz, so the highs still sound tight.
The group delay applied to the poor transiens can be seen here: