MSpectralDynamics in Pro Audio DSM style

[Jopmanajop]

It's tempting to see Pro Audio DSM 3 for $25 (with coupon), but why if there is such an awesome tool like MSpectralDynamics?
So guys, could you share some tips and tricks how to configure MSD to work in a similar way DSM works? It works already, just adjust attack, release, RMS, smoothness etc., but from what i understand DSM always captures the average level of the signal, which is THE key, especially for master bus correction, and capturing process doesn't depend on setting, it's always the average level. While MSD capture is highly settings dependent, adjust attack or smoothness and it is completely different curve already
I've been getting some great results already in the past, but now i want to be sure that i have stop spending and keep asking and learning. Maybe it can do better than DSM, i'd like to know about that

I tried putting MUtility with RMS on before MSD, but that quickly turned into the noise, not the best idea

[MeldaProduction]

I don't really know that plugin and I'm afraid no time to check (though it's cool to finally see after so many years to try to clone MSD ), but what you are describing as average level is actually the detector. They just probably use slower values, since these are easier to handle without artifacts. We are allowing also very fast settings, but on master you probably don't want to do that (though even that is questionable). Personally on master I simply use the Flattening device (which I think is default anyways) and touch the detector settings only if needed .

[Jopmanajop]

You might be right (duh), i tried to increase attack, release and RMS to the max, set slowest envelope detectors, and now it looks like average, needs further investigation obviously
DSM can capture that average curve of reference track and apply spectral compression based on that reference average with completely customizable compression settings, that's why it's so good, it doesn't capture peaks, but, let's say, spectral energy. "One-knob" fixer for bad mixes (i don't wanna mix, i wanna FIX )
I used to use Flattening, it's awesome, should get back to it sometimes

[vectorwarrior]

Spectral dynamics already allows you to capture a threshold, have to tried that?

[Jopmanajop]

Spectral dynamics already allows you to capture a threshold, have to tried that?

Sure, the question was how to capture the average in a similar way DSM captures it. Got some thoughts to share?

[DarkStar]

Out of interest, exactly how would the average level of a signal be calculated?

How about the levels measured in MLoudnessAnalyzer?

[Jopmanajop]

Out of interest, exactly how would the average level of a signal be calculated?

How about the levels measured in MLoudnessAnalyzer?

Not a signal, but a spectrum. Let's say it is similar to a matching EQ, only it captures the average spectrum curve of the FFT analyzer inside that EQ. Then that curve is applied to your track, but it isn't static cuts or boosts, it also pushing down the frequencies above that captured curve. It is easier to understand after watching some demo of DSM in work
So we don't need any analyzers outside MSD, we need to configure internal analyzer to show us average level of reference, capture it, then apply that curve to our track, so yeah, like using MAutoDynamicEQ, but only FFT dynamic EQ
I know i can just match curves with some matching EQ, but that's not really an improvement, peaks don't reflect how we percieve music. So what if i capture peaks of a really good sounding track, where the hi-hats are tight, compact, not so high in peaks, then apply it to my track, where the hi-hats too transienty, dynamic, uncontrolled. They will be much higher in peaks of course, and when they will be matched to our reference curve, they will be simply lower in volume, the same not tight, uncontrolled hi-hats. Here is why it's much better to capture the average, even better to apply that average dynamically
Really, it is easier to understand by comparing them in action or by watching some demos

[DarkStar]

OK, now I understand better.

[MeldaProduction]

Just click Capture and keep it going as long as you want and that will capture the average.

[stripealipe]

Is there any way of saving the Captures, or is this something that could be implemented in the future? This would be cool, especially when using MSD in this way.. ta.

[vectorwarrior]

Is there any way of saving the Captures, or is this something that could be implemented in the future? This would be cool, especially when using MSD in this way.. ta.

There is, although it doesn't work like presets. Right click the threshold graph, it allows you to export as a text file. You can then import them too. So just save them out to somewhere on your computer.

Before you ask, the reason they can't be stored as presets is because this graph editor is for ALL graphs, and the scale can be completely different, so they are not compatible across all graphs... therefore having presets doesn't make a lot of sense, and this import/export system is more suitable.

[jmg8]

Is there any way of saving the Captures, or is this something that could be implemented in the future? This would be cool, especially when using MSD in this way.. ta.

In the meantime as a workaround, you can:
Right-click in the threshold shape window and select export CSV, save the file somewhere.
Then when you need the shape back again you can import CSV.

[stripealipe]

Nice one, thanks for the replies you two, that'll work just fine for my needs.

[jmg8]

Oh, haha. I only just noticed that I pretty much duplicated the same as Vectorwarrior. I started writing mine before he posted his. Oh well.

[aMUSEd]

It's tempting to see Pro Audio DSM 3 for $25 (with coupon)

Where can I find this?