How do you judge an EQ?

[oursvince]

[...] and the fact you can enter the values manually, what is very rare in free eqs...

Double click a bobble to enter the value manually in posihfopit.

Regards,

Christian

I didn't know that Christian, This was the only thing I thought posihfopit missed and you just learned me it was in, go back to posihfopit thanks! And congrats and cheers for your work! Vince

[Sascha Franck]

Double click a bobbEle to enter the value manually in posihfopit.

What's Boris Becker got to do with that?

[bmanic]

Hihi, the next Electri-Q version 1.5 (which is going to be released tomorrow) will have a maximum phase mode. With that the transient smearing happens *before* the transient. That's a funny thing. In theory, a narrow peak filter may turn a booom into a whoop.

Actually, this has been done before. A guy at our university coded an EQ that can do this.

It was never released to the public though.

Cheers!
bManic

[donkey tugger]

Generally, the more money one spends on 'eq' ,be it in making music or listening to it, the more 'level 42' one becomes.

[martian]

my advice to OP;

before listening/comparing and being fooled understand that some eqs have asymmetric high end bells. all 'regular' min phase digital eqs give the exact freq and gain you specify. whereas convolution and analog eqs have roughly labelled parameters. ie. 1db, 110hz boost on hydratone may actually be 2db, 135hz on the impluse.

listen/compare and realise you can exactly match the sound of most digital min phase eqs. try a lin phase eq and note the difference, esp on transients and using high q values, finally try some more interesting dynamic/harmonic eqs.

be sure to get it all out your system and get back to honing your skills.

[Debutante]

Well, it's not easy to describe how a good EQ sounds like, but i will try to explain it with an example, you can do on your own: choose a real fat sawtooth trancelead, set the usual hipass and then choose a medium Q bell band and try to reduce the "noise" in the 6-10kHz area. If you are sucessful, then the synth will sound a bit muddy, but much more balanced. This is the first check, how well the EQ can handle this task, without destroying the whole sound. Now you need to compensate a bit for the lost noise-treble with a hishelf. Set it to 12-15kHz and listen, if the EQ brings out a silky shimmer or if it just produces quite harsh treble, that never seem perfect. Either they are too much, or they are to less. You'll be never satisfied. With a good EQ fitting to you, you'll become satisfied.

Same procedure for EQ-testing works quite well on 909 open hihats, too. Hipass + loshelving to get rid of the rumble and create the correct size of the hihat's body and then reduce the peaks in the treble with high Q settings until you have a flat frequency response.

A good EQ will allow to quite easily reach perfect results (a good source material is needed ofcourse), while a not so good EQ, will correct one thing and always produce a problem somewhere else.

How this can be technically explained, since filters are not that magic, i have no clue.

That's a very interesting test. Thank you for this one. I think you should also try WWAYM NWEQ for this.

[3*s]

I'll post some audio examples soon for you guys.

Please do! Some comparisons of posifopit, electri-q, and maybe the voxengo and kjaerhus EQs with whatever high-end hardware you have at your disposal would make my day.

[Christian Budde]

I'll post some audio examples soon for you guys.

Please do! Some comparisons of posifopit, electri-q, and maybe the voxengo and kjaerhus EQs with whatever high-end hardware you have at your disposal would make my day.

If anyone has IRs of famous hardware EQs, please let me know. I'm currently looking for some hardware to emulate with the charm of Electri-Q (WYSIWYG-Editing)

Christian

[jPod]

If anyone has IRs of famous hardware EQs, please let me know. I'm currently looking for some hardware to emulate with the charm of Electri-Q

Would some of the IRs from the preamps section at Noisevault be useful?

[FaX]

This may sound stupid as I prefer to actually get on with production/composing as opposed to being hung up on tech specs and details.
For me it's quite simple -

Clarity and phase of signal post eq.
Lack of frequency smeering.
Good definition and frequency control of the bands in question.
A nice clean well defined sheen and polish to the end results.

Not much help I know but one of the reasons why I'm not a big fan of Live's built in Eq's unless using them as fx processing tools.

I will often use HP/LP/BP/BR/AP/Notch filters instead of eq's where possible also within a synth patch itself if needs be and not rely on eq-ing everything in the mix.

I'm not a big one for heaps of compression either if I can avoid it.

[bmanic]

Well here are some quick test files. What I've tested here is basically the EQ's ability to boost low end. It usually leads to some problems like "mud" or sort of lengthening of low transients. I only had a quick go at it with Wavelab's "Q" plugin and Sonalksis EQ. I also created some examples from a "mystery" eq (high end stuff) that I can unfortunately not tell more about at this point in time (I'm not hyping it! no! ).

What I did was a -14dB dip with PLParEQ3 lowshelf at 80Hz using the highest quality it allowed (no linear phase here). I then used the tested EQs to "repair" the damage. WARNING!: Subtle is the name of the game (though not in my opinion! ..but certain people think I should always issue this warning.. ).


Original audio clip taken from the limiter shootout thread.

Audio clip with only the PLParEQ3 dipp applied. Use this for your own experiments.


Wavelab Q eq. It has some of the typical plugin eq problems in the low end. The low end starts sounding muddy somehow because of "sluggish" transient. I used this plugin as the reference while tuning PLParEQ3 to get as "transparent" result as possible. This is simply to show that an inferior plugin eq can create problems that are not related to the shape of the eq curve only but rather because of destruction of the transients. This eq plugin I had to boost by far the most to get to the desired results.

Sonalksis SV-517 eq plugin. It might not be the smoothest eq in the high end but it's one of the better when it comes to treating the lows. About 2dB less boost was required than with wavelabs Q.

Mystery eq number one. Only 5dB of boost was used on this clip, however, I'm pretty sure the readout on this eq is a bit flawed. IMHO it sounds quite a lot better than any of the above EQs as it keeps the punch of the kick drum as it should be. It has a slight overall colouration of the entire track and a small loss in the extreme top which dulls the audio clip a bit.

Mystery eq number two. About 4dB of boost in the lowshelf of this one. I also boosted the highs a bit (about 0.5dB) to show what magic this thing can do at even minimal settings.

This test is in no way scientific at all. I'm trying to learn what actually makes a good eq good. First I thought it was mainly the shape of the eq curve combined with zero aliasing but I'm starting to suspect that there is much more to it than this. Phase of course plays a role and this probably more than anything messes with the transients.

Cheers!
bManic

[defjamm]

my advice to OP;

before listening/comparing and being fooled understand that some eqs have asymmetric high end bells. all 'regular' min phase digital eqs give the exact freq and gain you specify. whereas convolution and analog eqs have roughly labelled parameters. ie. 1db, 110hz boost on hydratone may actually be 2db, 135hz on the impluse.

listen/compare and realise you can exactly match the sound of most digital min phase eqs. try a lin phase eq and note the difference, esp on transients and using high q values, finally try some more interesting dynamic/harmonic eqs.

be sure to get it all out your system and get back to honing your skills.

yes, i absolutely agree. i was so fooled by hydratone but at least i'm not the only one
every one should try christians free plugin analyser, see how you get fooled by hydratone if you own it.
and then go and buy electri-q, it's awesome!

[dover666]

by the size of it's knobs (knoobs??) erm, steepness of the rolloff, esp imp in low ends and extreme highs
partly cloudy, chance of showers.

[Christian Budde]

Well here are some quick test files. What I've tested here is basically the EQ's ability to boost low end. It usually leads to some problems like "mud" or sort of lengthening of low transients. I only had a quick go at it with Wavelab's "Q" plugin and Sonalksis EQ. I also created some examples from a "mystery" eq (high end stuff) that I can unfortunately not tell more about at this point in time (I'm not hyping it! no! ).

What I did was a -14dB dip with PLParEQ3 lowshelf at 80Hz using the highest quality it allowed (no linear phase here). I then used the tested EQs to "repair" the damage. WARNING!: Subtle is the name of the game (though not in my opinion! ..but certain people think I should always issue this warning.. ).

Here are the deconvolved filters used. They have been windowed to 2048 samples, which turns out to be perfectly fine.

Original audio clip taken from the limiter shootout thread.

Audio clip with only the PLParEQ3 dipp applied. Use this for your own experiments.

PLParEQ3 Dip (Magnitude)
PLParEQ3 (Phase)

Wavelab Q eq. It has some of the typical plugin eq problems in the low end. The low end starts sounding muddy somehow because of "sluggish" transient. I used this plugin as the reference while tuning PLParEQ3 to get as "transparent" result as possible. This is simply to show that an inferior plugin eq can create problems that are not related to the shape of the eq curve only but rather because of destruction of the transients. This eq plugin I had to boost by far the most to get to the desired results.

Q
Q Phase

Sonalksis SV-517 eq plugin. It might not be the smoothest eq in the high end but it's one of the better when it comes to treating the lows. About 2dB less boost was required than with wavelabs Q.

Sonalksis
Sonalksis Phase

Mystery eq number one. Only 5dB of boost was used on this clip, however, I'm pretty sure the readout on this eq is a bit flawed. IMHO it sounds quite a lot better than any of the above EQs as it keeps the punch of the kick drum as it should be. It has a slight overall colouration of the entire track and a small loss in the extreme top which dulls the audio clip a bit.

Mystery
Mystery Phase

Mystery eq number two. About 4dB of boost in the lowshelf of this one. I also boosted the highs a bit (about 0.5dB) to show what magic this thing can do at even minimal settings.

Mystery 2
Mystery 2 Phase
By the phase and magnitude, I can tell that this EQ isn't a biquad, like the first two implementations.

Have fun,

Christian

P.S.: Since I now figured out about the "Mystery" and "Mystery 2" I will soon built them into my "Electri-Q", wuhahaha....

[poonna]

Wow, lots of stuff to put into my head! Thanks a lot guys.

Poonna