people raved about hydratone to be the god of eq and to be 'magical'. let's say you compare it to something like voxengo's gliss. with gliss(without dynamic mode), when you take away 15db at 1khz, doesn't matter if small or big q, you'll get your 15db...let's say you're a customer and you want to compare both...and you cut with hydratone 12db and you cut with gliss 12db...the hydratone one will sound more neutral, because in reality, depending on the mode you've used(in this case the ff 110 mode, which people found to be super neutral and good for cutting), you really didn't take away 12db, in reality there is much less gain involved. that's the reason why it sounds more neutral. a lot of people were fooled by that(me too). this shows you can't only rely on ears.
this goes even further...but that's something i can't prove, only come to conclusions through different people.
there are analog-fans, who claim that with their analog-eqs they can push 15db and it still sounds neutral(compared to a plug-in). chances are big that those eqs don't really deliver 15db more.
the industry is full of bullshit and marketing. people should know why something sounds different.

i have a theory why people might like it and why you like it...it's similar with hydratone, other convolution eqs(with hardware irs) and urs(when the filters are engaged). and it's the same with A0: rather high harmonic distortion(compared to 'cleaner' eqs like kjaerhus golden, sonalksis and gliss). you should download christians vst plugin analyser and try better see for yourself.

with A0, when sat is turned off and there is no gain on the filters, harmonic distortion is rather low. but when you now add gain or take gain away, there is an increase in harmonic distortion. sonalksis, kjaerhus and gliss don't have that(as much).

maybe christian can explain the technical side of things. i just know that with a lot of eqs that hardware-people seem to like, there is high harmonic distortion going on.

maybe you could try it with the new neve from the uad-1 and say what happens, when filters show gain.

what is that? is that a new eq for fl studio?

It's Edison i think...

Anyone tried that Waves one that "clones" the eq of real hardware eq's and then recreates it in software (using convolution presumably?)

I thought that looked like a more flexible aproach to what Hydratone does (though couldn't you do this with any IR processor?).

old train 63

I know what "mistery eq" is, and I like both bmanic and christian (yes, I think christian is one of the most promising developers here)

JUST AN OPINION IMHO IMHO IMHO
christian, you pay too much attention on fft magnitudo, phase plots, irs length. You haven't considered distortion in these examples, for example. This "mistery eq" sounds good, but IRS plot are awful. Why? Are you sure this eq sounds EXACTLY as other eq, or do you think the issue could be resolved only in its irs? Have you listened demos/compared them on HIGH-END monitors?
If you consider only plots, sonalksis plots are FAR BETTER (IMHO)
About IR lenght... ok a short IR can't reproduce EXACTLY what you are asking to your equalizer... but in a real studio/mixing mastering environment, who cares it? I have hardware gear with no labels on it, but I can use it for mastering. I don't use leds, I don't care if I'm gain-ing 2db or 0.5 or 5. I stop only when my ears are satisfied. Any comments to this?

I only evaluate the magnitude and phase response, assuming, that there was no distortion (or only minimal distortion here). Of course those ripples might also come from distortion and I have mentioned this in my other post.
By no word I said, that sonalksis is better because the plot looks better. In fact I have only listened to the demo using AKG K270 studio headphones (my roommate was sleeping) and just to hear how the tune sounds like.

The problem is not, that they are truncated, it's the fact, _how_ they are truncated.
If you reduce the size of an IRs you only loose frequency ressolution in the first place. The resolution can be calculated very simply this way: samplerate/IR-Size (as far as I remember). In the case of a 256 IR you'll get a resolution of approx. 172 Hz (@44100Hz). This is enough for high frequencies, but poor for low frequencies. On the other hand that doesn't mean, that you can't get a smooth low end. There still may be a detailed structure, but it's not excactly the same as the original.
But now to the far more important thing: If you truncate your IRs (like done in several products) you will introduce artefacts due to the truncation. Like with normal audio a fade needs to be applied. If not, you get aliasing. A hard cut _always_ introduces frequencies which aren't there (a single pulse equals all frequencies). So a special amount of ripple depends on the length of the IR and not on the EQ sampled (beside the fact, that hard cuts are unusual). Again I didn't say a word about the sound quality. This might sound nice, but you always have the "coloration" (ripple) based on the truncation (and based on the EQ). If you do it by windowing (e.g. fade-in/out) you loose some more resolution, but you remain the coloration of the original EQ.
I totally agree here. If it comes to _using_ the EQ, you should at least trust your ears.
Anyway, the ear has a certain quantisation as well. I know, some people think that they have golden ears and that they can hear everything, but there is always a limit. In case of gain (broadband) at least everyone will notice 1dB. A difference of 0.1dB is already very hard to hear and it's impossible to hear a change of 0.01dB. So I guess no-one will hear a difference between an EQ Setting of Gain 5dB Freq. 300 Hz and 5.1dB Freq. 305 Hz. That's just a guess, but in fact that doesn't matter, because it's even harder to hear if you don't have a comparison. So at least the "end-user" of the tracks will never hear the difference.

What I wanted to say is: be careful and don't _believe_ everything. Make your own experience and try to understand how things work. You don't need to be able to code these things, but you should know the advantages/disadvantages so that you can _use_ the tool called EQ.

Christian

P.S.: If you know, how things sound/work you can directly _apply_ the change. If you spend a lot of time _searching_. If you have two EQs with different labels (EQA:2dB=5dB:EQB), than it's impossible to experience the difference, because you are comparing apples with oranges. Only if both labels are correct you can focus on time structure (transient behaviour) or harmonic distortion, etc.

luckily thats not the thing that fooled me in comparison to other eqs, i always compensated that by boosting harder or less on the other eq for fairer comparison. but i did follow the freq readings so the inaccurate centre freq reading in the lows fooled me, and in the highs the low q fooled me.

but now i understand the good qualities i was hearing, and can set them up fairer they are the same, except most hydra modes boost more below the centre freq which i really dislike (you get muddy sloppy sound), only 'avln' is ok here. anyway im sure i could emulate that with 2 bands on any other eq.

heres a quick comparison of a simple (but stupidly big) 4.9db kick drum boost. hydra amk, Q, waves ren, gliss, posihfobit. hydra sounds different in the subsonics, Q and ren insanely close, but i find all the last 4 equal to my ears.

http://www.badongo.com/file/953107

will do better comparison later.

edit: link fixed.
Last edited by martian on Mon Jun 26, 2006 9:42 am; edited 1 time in total

I always thought analog filters had more phase distortion than digital because they were less linear in the time domain?

Digital is in theory perfect in the time domain, since it's all math, right?

In fact the analog filters phase variations depends on the filter amplitude variations.

In the digital domain, it is possible to design a filter where phase is not dependant on the amplitude.
It also depends on the the technology you use for the digital filter.

Fabrice
Eliosound

"there is no eq god"

damn guy, it exists. In 90s, when everyone was using analog eqs, the "god eq" was the digital one. Because S/N was better. Nowadays, when everyone has at least a good digital eq in his pc (cracked or free, it doesn't matter) the 'god eq' is the analog hi-end one. Infact big ears mixers have good gear, and they use professional toys that boyz don't have. Really big ears are GOOD in mixing, but this slight coloration (from analog ones, a sort of foot-print) is what people understand as 'state of art'. It's like neumann. It isn't the best microphone, but you have listened it in million best-sellers hits. SO it is the 'standard'. If you don't use it on a voice you are fooled. When all boyz buy a Neumann, big studios start with old Telefunken, old AKG, and this new 'colour' is the new 'state of art', because it is different form the medium production.
Today, what is the best eq?. In the age where digital eqs used in long chains destroy the transparency of the audio source, the best eq is the eq wich better preserves your audio. Only beginners need eq with leds, writes, good graphics. Professionists are able to use the tool in the best way, even the wavelab eq, because they use it in the right amount. If they need 12dB eq, they don't use wavelab eq. They listen, and understand quickly something is wrong. SO they change the tool, the amount, the slope, the frequency, and they stop when something they are looking for is emphatized.

Today we are not looking for S/N. We are looking for the right colour, for the right sound, and we dislike the eq that is mudding. Why a digital eq sometimes is not enought? It isn't a problem of internal variables. Every commercial eq now is at least 64 bit floating, and you can't have more than 128 bit fixed point, due to processors. What is incredible, all this power is far from the quality of a vintage unit. If you put a vintage eq on your music several times, at the end only the S/N is worse. If you put a digital eq in your music several times, very often transients are destroyed. If your digital eq preserves 99%, put it 10 times and you got 90% or less. And a long plug chain is a very frequent situation. When I started with digital, in 1996 I liked the digital plugs. In 2000 I was mixing again on analog, because something was wrong. I was not ready for the new troubles, simply I dind't expected them. Initially the big S/N was the best thing. But soon I discovered my music had a new weaky point: it was less spatial, all was glued, plugs were not working in the way I was looking for.

About IRS. A linear system could be represented by its IR. But analog gear is linear? I should say no. It can't. This situation recalls me the digital vs analog race. I remember the age where dealers said: "our sampler is 16 bit, it can reproduce your moog, sample it". We descovered we needed 24 bit, after 32, after we discovered analogs preserve better low frequencies, play on a real moog (not the good Arturia emulation), really the floor quakes, even at low levels. Simply before we were 50%, now we are 80%, but we are far.

I see there are 2 kind of people.

Develpers say "magic doesn't exist". They are right. I'm an engeneer, I know hardware eqs are filters, and they are causal, you can't forecast the future. Plug-in developers believe that it is all a matter of market advertisement.

Musician instead say "magic exist". Magic is the right eq in the right place with the right amount. In general hardware eq are often magic, because they preserve better your transients, they don't have calculation errors, they don't need oversampling, and they have the right colour, the colour the professionist is asking to a tool. And even if magic doesn't exist, the client NEED to SEE hardware leds, because probably he already owns a software good eq. I'm a musician too, I own hardware gear and I know hardware is slightly better again. Better for me is because a tool is not diffused, there are only few owners, so it is the secret weapon for a 10% improvement in the right direction.

excuse me for the long answer.
Last edited by Zaphod (giancarlo) on Mon Jun 26, 2006 11:15 pm; edited 1 time in total

" If you have two EQs with different labels (EQA:2dB=5dB:EQB), than it's impossible to experience the difference, because you are comparing apples with oranges. Only if both labels are correct you can focus on time structure (transient behaviour) or harmonic distortion, etc."


1)
When I put an eq in my chain, I put ALWAYS a look to a spectrum analizer if I'm checking. But I check AFTER I've listened. The problem of "12 dB not correctly reported" doesn't exist using tools. Everyone has a spectrum analyzer. Anyway, I remember, I check after. Before the sound, after the curiosity of checking I'm doing correctly.

2) If you sample hardware, you'll discover it isn't precise neither accurate. But it plays awesome. So it's not only for IR truncation the matter. Sometimes it's the original hardware, it isn't precise/accurate enought. But what's the matter? We have BEAUTIFUL records done using not-accurate tools. GML are accurate? I say no. But they sound 10x better than any digital plug. So we can't compare hardware with software because labels are wrong?


3) transients are destroyed with little corrections sometimes. You don't need to go deeply 12 dB on both to understand one is better. Put 1 dB in wavelab at the frequency you prefer, it is enought for mudding everything (but wavelab eq is not the evil. Just sonalksis works better, and infact it costs more)
Last edited by Zaphod (giancarlo) on Tue Jun 27, 2006 1:22 am; edited 1 time in total

In the court of law.