Contraversial method for testing quality of a plugin ...

[Kiwburger]

Here's another cat to throw in amongst the pidgeons ... (developers, if your plugin fails this test, don't get mad at me for exposing it) ...

I was experimenting with the excellent Voxengo impulse response Deconvolver. This has a feature that creates impulse wave files. These are a swept sine wave, normalised at -0.3dbFS, and with fade in and fade out. They are ideal for taking impulse responses - but i've found they are also ideal for exposing weaknesses in plugins. Especially headroom/distortion issues.

Using one of these files, I started testing a bunch of plugins, and found that this really exposes some interesting faults or issues with many of the popular plugins. I am now realising I have been expecting far too much in terms of headroom from most plugins. Because most material has transient peaks, the damage done is not always apparant. We all know that some plugins sound better than other, but it's not always obvious why. This stress test - of strong, sweeping, test tones, really let's you hear clearly which plugins really distort and damage your sound.

I'm not going to name any - because i'm already extremely unpopular here. But I expect a few developers will be extremely mad at my suggestion of testing their plugins via this method. Just do it, and make up your own conclusions.

There is one very popular plugin that distorts and sounds very bad unless you restrict the output to -20dBFS! That's waay lower than most people would use it. It's also easy to prove this is internal clipping, because you can lower the master fader, and the distortion doesn't clear up.

Some people might argue that it doesn't really matter, because this is not a real world test. But in my opinion, if a plugin can't handle this test, it will clip on the transients of normal audio. It might not be extremely obvious, but distortion like this accumulates in a mix, to the point where you know something is wrong but can't quite identify it.

If you have choice of two plugins - one passes this test, and the other fails: then why not use the one that passes.

Let the slaughter commence ...

[sophonomoth]

Well you forgot to explain how you notice these distortions on impulses.

There is a far more revealing soft. http://www.savioursofsoul.de/Christian/ ... alyser.zip
Also, the frequency resonse is measured with 64 bit scaled (impulse like) sweeps. So transient related issues are better lined up than pure sweeps.
Probably you know it yet.

And these distortions are for many a feature, apparently dev seems to studdy their own THD decay & shape to match some analog reality.