MautoDynamicEQ in PluginDoctor

[Boone777]

Hello! I recently watched Dan Worall video comparing Reaeq and Fabfilter ProQ3.


So I wanted to test some of my Eqs as well.

Something came up that I was'nt expecting. Maybe someone can explain. When upsampling MadEQ there is a dip at the end of the curve can someone explain why is that ?

I don't get the same results from DDMF IIEQPro for exemple.
This is not really scientific just trying to wrap my head around some behaviors I was not trying to match the curves just wanted to know if there is some theory to take into account behind various behaviors.

Can someone comment if there is any technically possible audible difference using these 2 EQs when multiple instances are used in a project ?

MadEQ 1x

MadEQ 4x

DDMF IIEQPro

[Burillo]

MADEQ uses oversampling to compensate for cramping. apparently, IIEQPro uses something else that doesn't involve oversampling (or its oversampling does not involve brickwalling stuff above 20k). that's my half-educated guess.

[sirmonkey]

The graphs visually look different, but:
1. Consider that many people can't even hear frequencies above 15K Hz, especially those middle age or older. Same for some unfortunate younger people who have listened to too much loud music.
2. While even those with great hearing can hear up to 20,000 Hz, even their ears are more sensitive to frequencies far below that...like in the mid-range.
3. While the graphs look different, look at 10,000Khz at the bottom of each graph, then go straight up to see the gain boost. Compare the amount of boost for each graph. Then do this for 15,000Hz and 20,000Hz.
The difference in db's between the graphs isn't that much. And even for "super-hearers", those high frequency ranges are still far above optimal hearing, as mentioned.

So, I think that any differences would be minuscule to non-existent for 99% of people.
* Also, many speakers and headphones won't reproduce some of those high frequencies anyway.
** Earbuds suck! (Just thought I would throw that in)

[Boone777]

Sure it's not going to make a difference on one instance I was wondering if there could be any difference on the high end when dozens are used, theorically. Cause I'm not confident I can tell difference but it's nice to know what to look for to help spot the difference if any.

[Fourstepper]

I wonder about this too pretty cool

[MeldaProduction]

Simple: "Peak analog"
For the record oversampling needs the high end LP filter you can see there. Personally I wouldn't worry about this anyways, nobody hears 20k , but we did the analog peak anyways long time ago.

[vectorwarrior]

Personally I wouldn't worry about this anyways, nobody hears 20k , but we did the analog peak anyways long time ago.

I of course agree with you, but please don't forget those of us that work in non musical jobs. In game audio our audio is regularly slowed/pitched down so you will hear any nastiness in those high frequencies.

A VERY common technique is to randomize the sound pitch each time a sound plays, this causes those normally inaudible frequencies to suddenly become audible when the pitch randomisation makes it lower pitch.

We also tend to do a lot of slowing/pitching down audio when creating/designing sounds too, so again, this high frequencies become audible.

Sorry, not trying to be pedantic but I often see the sentiment online that 'you can't hear it so it doesn't matter' and that is not true for some of us for these reasons.

Thanks!

[sirmonkey]

I of course agree with you, but please don't forget those of us that work in non musical jobs. In game audio our audio is regularly slowed/pitched down so you will hear any nastiness in those high frequencies....

Interesting....I never knew that.
So now I'm wondering, purely out of curiosity, why not just use a high-cut filter to get rid of the highs on your original samples? (Just a hobbyist here, who wonders about stuff)

[vectorwarrior]

I of course agree with you, but please don't forget those of us that work in non musical jobs. In game audio our audio is regularly slowed/pitched down so you will hear any nastiness in those high frequencies....

Interesting....I never knew that.
So now I'm wondering, purely out of curiosity, why not just use a high-cut filter to get rid of the highs on your original samples? (Just a hobbyist here, who wonders about stuff)

Hmm not sure I follow. What I mean is we'll take a sound and slow it down to pitch it down, so freqs at 22khz are now, say, 15khz and very audible. This means that high frequencies should not be messed around with because they could become audible. If an eq is messing with high frequencies then it could sound awful when slowed down.

[Svama]

But you didn´t hear this problems anyway until now? Maybe it´s only a problem of plugin order then. First pitch down and after it the eq doesn´t need to work in this higher range anymore or am I wrong?
But I think Vojtech will take a look at it now!?

[ozonepaul]

...Maybe it´s only a problem of plugin order then. First pitch down and after it the eq doesn´t need to work in this higher range anymore or am I wrong?...

You're right, the plugin and editing "chain" order will change what an EQ does to a sound.
But it's still true that how an EQ behaves at the highest frequencies can have a major effect on the sound (and not just on the highest frequencies) for example if the sound is heavily distorted/ time stretched/ pitch edited etc. To understand this watch eg. this video about distortion from 8:23 (really important bit is at 10:35):


(If Aliasing and Nyquist etc are alien terms for anyone, I suggest this guy's explanation:)

Hello! I recently watched Dan Worall video comparing Reaeq and Fabfilter ProQ3....

Whether there could be situations where the differences on your graphs would translate to real audible differences that can not be resolved by different settings on the actual plugin(s), I don't have a clue but I think it's an interesting question. Whatever the answer is I'll continue using MAutoDynamicEq a lot though!

[sirmonkey]

I of course agree with you, but please don't forget those of us that work in non musical jobs. In game audio our audio is regularly slowed/pitched down so you will hear any nastiness in those high frequencies....

Interesting....I never knew that.
So now I'm wondering, purely out of curiosity, why not just use a high-cut filter to get rid of the highs on your original samples? (Just a hobbyist here, who wonders about stuff)

Hmm not sure I follow. What I mean is we'll take a sound and slow it down to pitch it down, so freqs at 22khz are now, say, 15khz and very audible. This means that high frequencies should not be messed around with because they could become audible. If an eq is messing with high frequencies then it could sound awful when slowed down.

Actually, what I mean is that if you cut out the high frequencies (with a high cut filter) - the higher frequencies will be eliminated, or a least almost eliminated. So then, even if pitched down, the (formerly) high frequencies will still be very quiet.

[MeldaProduction]

vectorwarrior, good point. But I think it's still a specific small part of the "industry". For 99.999% audio production you just don't do that. Anyways the peak analog does that. Btw. the normal peak would work fine if the sampling rate was like 88k or higher (or oversampling was used). In 44/48k you'd still be limited to 22/24k maximum frequency, so for lowering pitch and expecting some cool stuff to get "unburried" on top it wouldn't probably make too much sense.