MAutodynamicEQ control interaction for dynamics?

[mesaone]

So I'll admit, after working with MEqualizer for a while, I made an impulse buy. So now I have MAutoDynamicEQ. I won't have the opportunity to work with it for several more days, so I have been reading through the manual.

From my understanding:

1. Threshold, Attack, and Release controls act how they normally do in a compressor
2. Dynamics knob sets the maximum boost/attenuation for the band
3. Level Gain acts as a pre-emphasis for the detector circuit only, much like an internal sidechain would, and does not affect the audible gain
4. There is no ratio knob, but there is a Level Transformation graph that can be used to control the in:out ratio

Is the above correct? I'm having a hard time understanding how the amount of gain reduction is dictated, aside from the maximum applied with the Dynamics knob.

1. If my dynamics knob is at -12 dB, and my threshold is -40 dB, at what input level will the maximum gain reduction be applied? How is this determined?
2. If my Level Transformation graph shows a straight line from 0,0 to 100,100 that would mean my ratio is 1:1. How, then, is gain reduction handled if I set the Dynamics knob to, say, -12?
3. What are the advantages and disadvantages (sonically) of Logarithmic, Squared, and Linear shape settings?

This might be more simple than I am making it, but I'd like to know all I can about my new purchase.

[MeldaProduction]

Rule #1 - do not overcomplicate things
Rule #2 - try not to think about it as if it would be compressor .

I'd suggest simply using the main view, no threshold, no attack... just the basic eq bands, X frequency, Y gain, X slider Q, Y slider dynamic gain. By default it works like a normal eq with very very awesome filters . If you then change the Y slider (dynamic gain), you could say it gets an expander (up) or compressor (down), but it's better to simply think that the band gain is changing with increasing level - towards the Y slider, the dynamic gain. It has proven the best results simply with threshold, attack and release default, hence everything auto, minimum threshold.

Check this tutorial:

http://soundbytesmag.net/meldamdynamiceqpart1/

And trust me, you won't need transformation shape that soon .

There's no "quality comparison" for linear/low/sqr, they just respond differently, my personal favourite is sqr, if you want really strong processing use log, if you want just some mild corrections use linear.

[mesaone]

I appreciate your reply, Vojtech. I'm probably thinking about it too much. I have now had a chance to work with it for a bit, and it's definitely not very complicated to just do things by ear... Although I do usually prefer to know what's going on behind the GUI, at least a little bit.

I'm really impressed with Melda. I initially tried the free bundle when it first was made available for AAX, and it quickly became clear that your work is top notch. I should probably stop buying so many plug-ins, but you've made that difficult with all these discounts.

[MeldaProduction]

Thank you!
And please please buy more!!!