FR: Analysers that go below 20hz and above 20khz

[vectorwarrior]

All the analysis tools in melda plugins go from 20 to 20,000Hz. While this is useful for tweaking the sonic character, there are lots of times where you want to know what is happening above and below this range.

For example, recordings made with most a microphones are going to pick up material below 20hz, and any dynamic processing is going to be influenced by this. If you can't see it, you dont know how best to filter it. I have one mic that picks up a LOT of low frequency information and understanding where to place the filter is vital. Similarly, the slope of any high pass filter you chose is often influenced by the shape of that low frequency information, perhaps you need a steeper slope than you realize because you can't see below 20hz. Finally, the different types of processing can create sub 20hz differences, e.g. linear phase and analog filters create very different slopes in these low frequencies, so it can be helpful to see the effect in analyzers.

Here's a spectrogram of some audio i just recorded when testing a new sound recorder I just bought. The red is room tone without any high pass filtering. The orange is with a 75hz high pass applied in the recorder. Using Audition's frequency analyzer I could see that adding the high pass filter actually ADDS some signal around 0-2hz (perhaps a slight DC offset). Melda analyzers would not have shown this because they stop at 20hz. Hopefully this shows the usefulness of an analyzer that goes below 20hz.

filt1.jpg

My suggestion is that the default remains 20-20000, but you are able to shift or zoom out to display an extended range.

[MeldaProduction]

I'm afraid that just won't be possible. It would indeed be useful for some scientific stuff, but the engine is entirely designed for the range 20Hz - 20kHz and the many features just cannot comply, that would be a brutal development .
Anyways what kind of HP you used in your example. It's quite odd really. On one hand it depends a lot on the window size as there will pretty much always be some DC component, which may not actually be there, it's just a mathematical phenomenon raising from the finite data size. In your case it's not exactly clear what the level is for the DC component, but if it is high, then something seems wrong with the analyzer or the filter.

[vectorwarrior]

Okay, i figured you'd say no but I understand it's used everywhere so has to make sense.

As for the filter used above, it's an analog hardware filter on the mic input of a a sound recorder, so clearly has some inconsistencies or DC offset going on. Your filters dont do this.

Thanks

[MeldaProduction]

As for the filter used above, it's an analog hardware filter on the mic input of a a sound recorder, so clearly has some inconsistencies or DC offset going on. Your filters dont do this.

Thanks

Good! You got me scared!