Voxengo SPAN vs. SuperVision

[Timo00]

I’m trying to resolve a discrepancy between the frequency spectrum that SuperVision shows me vs. what Voxengo SPAN shows me.

I’ve got both on my Control Room bus, and SPAN tells me that there is no energy below 250 Hz on my main out:

Supervision tells me that there is quite a bit of low end energy there:

FWIW iZotope Tonal Balance Control tells me the same thing:

So, if I were to count votes, it seems like I do have some low end energy in my mix.

But - I’ve EQ’d all that low end out, and if I look at the channel strip analyzer of the solo’d channel that’s producing this main bus signal, that channel also doesn’t have any energy in the low end. And I’ve got nothing else in the signal chain between that channel and the main bus that might add some low end energy.

I really just want to have an accurate analysis of the frequency spectrum of my main bus, esp. on the low end, and I’m not sure what to believe here.

Anybody have any insight and/or tips?

Somebody else has pointed this out in 2021, but that thread died on a vine, so I’m hoping for a different fate for this one

SuperVision vs. Voxengo SPAN

the problem is mainly that SPAN can change the windowing for the FFT but Supervision uses a fixed window function and it’s not clear which one without the knowledge about the way the signal is prepared for FFT and how the FFT is applied it is very hard to match them SPAN is much more flexible in regarding the FFT (spectrum display) all the above says that it is not only the view that’s different, it’s the measurement as well

[st10ss]

Can you provide the particular settings used in the screenshots?

FFT measurement and the settings is a beast of its own.

[glennloopez]

supervision is accurate and so is the Izotope. I am sure Span is too but never used it. Seems to me that Span cuts off the information before the other plugins and that’s why you don’t see it. as for why you have these low rumblings I can’t tell you. Check your signal chain. Simple test would be a lowpass eq filter on the master buss to see the results and work your way backwards

[st10ss]

@glennloopez

Seems to me that Span cuts off the information before the other plugins and that’s why you don’t see it

No. It’s more complex. As already stated in the other thread, it depends on the settings. But for Supervision some parameters changed since then.

[Aivaras]

@Timo00

SPAN tells me that there is no energy below 250 Hz on my main out:

Not quite: it shows that there is no energy in that band at -96 dB. Make sure you scale the EQ interfaces equally before comparing. Also, make sure you understand what type of level (e.g., peak, RMS, averaged, smoothed, etc.) is being shown in the graph.

@Timo00

I’ve EQ’d all that low end out

Filters have slopes. Depending on the steepness of the slope, you may have attenuated the energy in the band without completely removing it.

@Timo00

I really just want to have an accurate analysis of the frequency spectrum of my main bus, esp. on the low end, and I’m not sure what to believe here.

You still have to interpret what the analyzer is showing, given its specific settings.

[Timo00]

Thank you all!

The answer here is that I should have looked more closely at the Y axis and noticed that SPAN doesn’t quite go as low as SuperVision does. I bet once I adjust the SPAN Y axis, it’ll show some of that low frequency rumble as well (I’m currently away from my studio, so I won’t have a chance to do that for a bit).

I’m sure that there are some level averaging/FFT window etc. settings that can be tweaked to make all three plugins show the same (or a very similar) graph, Either way, that low frequency energy is low enough (sub ~100 dBFS) that it’s basically inaudible, so I’m not going to worry about getting the graph presentations to match exactly.

My main take-away here is to keep in mind that the SuperVision default Y axis scale goes below -100 dBFS, so between that, and the dB scale being logarithmic in the first place, it’s OK to ignore some of the really low level stuff being shown in it.