Avis pour mixage = UAD2 ou plugins FLUX

http://www.rhythminmind.net/presetblog/2009/03/digital-eq-fact-myth/
https://fr.audiofanzine.com/homestudio/forums/t.327714,post.5381101,p.7,les-legendes-urbaines-en-audio.html

You can't use a noise signal when doing a null test and get accurate information about the effects of phase shift. If you do a null test using phase shifted sine waves vs. phase shifted noise, you will get drastically different results.

Noise signals are useful for frequency response testing, but for phase shifting, they are essentially useless.

Thanks to Brian Brock on the second page of the thread, we have a link to the original signaltonoize.com blog posting about this issue. It seems that the argument presented over there has been mis-characterized by nearly everybody, including myself since I'd only read other's opinions about it. I haven't read the gearslutz thread because just the thought of reading that forum makes me lose the will to live, but the relevant bits of the signaltonoize posting are this:

All Digital Parametric EQ’s are the same.

Notice the qualification, digital parametric.

I know your sitting there saying but hey my Waves, URS, Oxy, UAD, Vox, blah blah sound completely different. This is why. With emulation plugins you have EQ+X. X can = Saturation, bandwidth limits, softclips, & so on. Now take away X what do you have? The EQ algorithm.

So, in other words, all digital parametric EQs are the same, except that they're not. The EQ algorithms are the same, but the EQs sound different because of other enhancements. Fair enough, but usually you can't take away those enhancements, so they still sound different. This is important, because the author was not denying your ears, nor calling you all fools for shelling out $$ for more EQs. However, the author's highly qualified point has been lost under the debate. Further, some have further clouded the issue by calling this the

"all digital EQ's are the same unless they're broken" theory

Nowhere does the original author state or imply this. Instead, he recognizes that some parametric EQs have EQ algorithms plus other enhancements, as well as this:

Lets look at Digital EQ types.

1. Parametric/Series – “PEQ”.
(These are the most common track EQ’s, Most channelstrip EQ’s, Most Plugin EQ’s, Native DAW eq’s, it’s never ending.)
2. Parallel – More uncommon in the digital world. Algorithmix has one.
3. Linear phase – Mastering EQ

These 3 types will be different. But PEQ’s are the most commonly used/emulated.
You can take any respectable PEQ and create all of the curve/phase characteristics of any other digital PEQ. It is the “X” that is the variable.

So to sum up, the author's point is not terribly controversial after all. Digital Parametric EQs don't re-invent the wheel, they all work on frequencies in the same way. (This is such a twee point that it seems obvious, but it's really the gist of the whole thing.) However some EQs are enhanced by other things, and further, there are other types of digital EQ such as Parallel and Linear Phase that are not comparable at all.

Now to test this point in proper peer-review fashion, null tests would be helpful, but there are some insurmountable difficulties for many EQs. You can't "turn off" the enhancements on many of them, so no proper comparison can be made. Null tests could still be useful to test if the enhancements are actually doing anything, but that's not the author's original point. He doesn't claim that the enhancements are useless, indeed he acknowledges them as the difference that people are hearing between different EQs.

I hate to say it but this has been much ado about nothing.