How do you judge an EQ?

[bmanic]

Ok, here are the eq sweep examples. One is Wavelab/Nuendo Q eq and the other is the same "mystery" eq that has been used before. I verified that both cut about the same amount (checked with white noise and SPAN in wavelab, over most of the frequencies during the sweep). They both sweep from about 30hz to 10kHz. The shape of these eq's are NOT the same but they cover about the same area.

Wavelab Q equalizer example

"Mystery" equalizer example

Any comments are well come. Once again, this is not a scientific test in any way and is not trying to be but I think it shows quite clearly the differences. Pay special attention to the low end and the critical 2-5kHz area at the end of the sweep. Hint: listen to the hihats.

Cheers!
bManic

[Eiosis]

One of the items on my list is to do a worldwide doubleblind tests, [...] With untrained subjects I could verify the literature and prove that no one was able to hear a difference.

This idea is pretty cool, but I found myself that performing A/B comparisons for the same EQ settings (which gives the same amout of boost with the same Q, about the same shape, etc..) are not very relevant. In fact, most of the time we could not hear the difference.

After two month I will make an evaluation tool available, so that everyone can see, how he performed...

Nevertheless, this idea is rather interesting.

Personally I think this discussion is better off in the DSP Developers forum instead of here in the Effects section. I think Poonna (the original starter of this thread) has lost interest after page 2.

I think that the reason why this thread may seem a little off is that the tings are not so simple. As I personaly experienced, it is rather difficult to perform convincing and eloquent A/B comparisons on EQs. Many other people with better listening abilities, better monitoring and listening environnement or simply just another point of view may find that it is not the case, and I respect this. Osbiously if you make an A/B test between an high end passive equalizer and a poorly designed VST equalizer, the difference will be much more noticeable.

People are talking a lot about good/bad EQ plugins, yet I don't seem to get the gist of the discussion. Analog-modeled or linear-phase EQs, coloured or transparent, I think I get the concepts of what these are. The thing is, I don't have any idea how these actually sound, and I don't know how to judge them.

My personal experience told me that one of the best way to judge an EQ is to use it in a real situation. Before investigating about EQ issues, I used EQs without knowing what is the technology behind them. After using it many times for mixing purposes, I found some EQs good and others bad. So I sought to understand which were the reasons of these differences.

I noted that the ones I founded good has certain technical characteristics, and those which were not appropriate to me had other characteristics.
My conclusion was that if I can not judge an EQ by performing A/B comparisons, and that I can not use an EQ in a real situation during a sufficiently long time to evaluate it, then I have to rely on technical arguments. That's one of the reasons, I think, this topic derived a little.

I think it would be great if some kind souls here could provide audio examples of these things, providing a raw clip, several EQ'ed clips using different kinds of EQs, perhaps with description of what to look (or listen) for in each clip, why one is superior to another, etc.

Also never forget that EQ is a matter of taste, and that technical arguments are not the only way to judge an EQ. The best way is to use it. With experience, you will be able to know more and more quickly if an EQ is good or not.

Fabrice,
Eliosound

[bmanic]

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

[Eiosis]

Here is another little a/b comparison. [...] It got pretty close and I think in this clip it's rather obvious that Wavelab Q kills the transients in a bad way

Bloody interesting, bmanic, thanks !
So I have to change my mind about A/B comparisons.

We can clearly hear that the wavelab EQ does not manage correctly some of the transients.

Fabrice,
Eliosound

[Kingston]

because of these subtle yet audible differences, the psychological effect is that the other seems easier to use. One won't get carried away trying to avoid the grit, and can concentrate on getting the sound right in the first place.

This was a massive problem for me in the beginning days of my mixing. EQ was/is my most important mixing tool ([edit] after faders... doh!), and initially I was wondering why my mixes sounded like shit.

Not all of it was because of my mediocre skills back then.

Although wavelab Q sounds pretty f**king decent in comparison to what digital EQs we had back in 97.

[mammath]

Off Topic: Kingston, when can we expect your offering?

Sorry about that guys

[Barbarossa]

bmanic, thank you for this example. I think this one nails it down quite good: no matter what someone likes more, but no matter how hard someone tries, he would never get the same "feeling" out of the two EQs.
Now i would be interested in technical explanations: If we assume, that both are phase-minimum EQs and the filters are similarily adjusted, what kind of effect is this, that smoothens/smears the transients?

[Kingston]

If we assume, that both are phase-minimum EQs and the filters are similarily adjusted, what kind of effect is this, that smoothens/smears the transients?

The assumpition is wrong. The mystery EQ (I know what it is but can't reveal), is a design similar to manley massive passive, far from minimum phase. It's the kind of design where the phase error was in full knowledge *left in* and even emphasised. This way we can (even algorithmically) preserve transients as well as technically possible. The trade off - well audible in the mystery example - is an odd smootheness of the EQ sound, in form of emphasised phase warp.

Yes filter phase errors can sound *good*. It would've actually worked better in this comparison if the voxengo HarmoniEQ phaseP mode would've been used. It works on a similar concept, and as Aleksey hinted some time a go, could be improved further.

[Kingston]

OT answer.

Off Topic: Kingston, when can we expect your offering?

Sorry about that guys

aarrgh. shouldn't have said a word before the release was already actualising. Now I suffer the fate of an early announcement, no matter how unofficial it was. Sorry about that fellas. It's ready when it's done.

well actually, it's been done quite some while, but there are certain things out of my hands that need be done. hopefully this summer.

f*ck! I said it again.

[Jonny Quest]

The mystery EQ (I know what it is but can't reveal), is a design similar to manley massive passive

Wasn't the manley massive passive based on some of the old Pultecs?

[Kingston]

The mystery EQ (I know what it is but can't reveal), is a design similar to manley massive passive

Wasn't the manley massive passive based on some of the old Pultecs?

yes it is. It's a modern take on a passive equaliser with a tube gain makeup stage. Plenty of ways to do passive equalising as well. Mostly the same rules - curve shapes, number of discrete filter stages (orders) and phase behaviour - apply to digital design as well.

[Jonny Quest]

The mystery EQ (I know what it is but can't reveal), is a design similar to manley massive passive

Wasn't the manley massive passive based on some of the old Pultecs?

yes it is. It's a modern take on a passive equaliser with a tube gain makeup stage. Plenty of ways to do passive equalising as well. Mostly the same rules - curve shapes, number of discrete filter stages (orders) and phase behaviour - apply to digital design as well.

Ah, perhaps a clue to the mystery eq then?

Whether related or not, wonder which of the Pultec software sims (BombFactory, UAD, URS) comes closest to the original.

[defjamm]

The mystery EQ (I know what it is but can't reveal), is a design similar to manley massive passive

Wasn't the manley massive passive based on some of the old Pultecs?

yes it is. It's a modern take on a passive equaliser with a tube gain makeup stage. Plenty of ways to do passive equalising as well. Mostly the same rules - curve shapes, number of discrete filter stages (orders) and phase behaviour - apply to digital design as well.

Ah, perhaps a clue to the mystery eq then?

Whether related or not, wonder which of the Pultec software sims (BombFactory, UAD, URS) comes closest to the original.

don't know but i doubt it's urs. urs is more about marketing and nice guis. they use average algos. the uad-1 one has some highend roll-off at 44k due to their oversampling...if i would have the original i wouldn't trade it for one of them

[Kingston]

Ah, perhaps a clue to the mystery eq then?

No, not really. As was mentioned, there are plenty of ways to do a passive EQ. There's plenty of them around, featured in consoles etc, as they are cheap and easy to implement in volume manufacturing. There's less buffering and only one gain make up stage to deal with. The filters are just caps and resistors and you can rest assured all cheapo manufacturers have used them through the ages paired with shit opamps in mixers and channelstrips alike.

What's great with pultec and massive passives (and other great ones) is that they are over-engineered bruteforce beasts that spare no costs in getting the I/O and sound perfect.

In fact, pultecs (and fairchilds etc) are by todays standards *far* overspecified I/O-wise, originally designed to deal with absolutely ridiculous amounts of gain. We don't need that today, but we still like to hear that distortion coming from those brute-force designs.

Manley did away with the distortion and specified the I/O by todays standards, still keeping the sound similar... as did plenty of other manufacturers.

[xander]

Wow

Freakin great info mates!

Keep it comin

Cheers,
Alex