SPL Full Ranger / Free Ranger false EQ freq info

[Lesha]

I have received a monthly Plugin Alliance newsletter yesterday and noticed there's a link to a video introducing their 4 free plugins.
I have installed them all and went to C.Budde's VST Plugin Analyzer to check out the EQ curves in SPL Free Ranger. Lo and behold, one of them was wrong.

I have sent an email to their support, stating: "Why does one of your EQ bands state "1K8", but the spectrum analyzer shows 900 Hz ?"
And the reply I got was: "This is modeled analog hardware and not a decision from our site."

WTF? What kind of answer is that?

Their info clearly says:
1,8kHz center frequency – this is THE EQ test frequencies. An EQ that sounds good here is a good EQ. Therefore it makes a lot of sense to compare EQs in that range. Applications examples: Boost for more bass definition and presence, cut reduces metal for brass.

What I see here looks more like 900 Hz to me...



Way to go Plugin Alliance

[camsr]

Do the center frequencies change, if you change your DAW's sampling rate?

[Lesha]

Do the center frequencies change, if you change your DAW's sampling rate?

Why should they?

[toothnclaw]

Yes, they are inaccurate on purpose, not by chance or omission. Same goes for their paid 'analog' EQs. Long story short, Plugin Alliance EQs are a waste of time and money.

[Loki Fuego]

I have sent an email to their support, stating: "Why does one of your EQ bands state "1K8", but the spectrum analyzer shows 900 Hz ?"
And the reply I got was: "This is modeled analog hardware and not a decision from our site."

WTF? What kind of answer is that?

If you read product description then, you should have noted that:

The Free Ranger is based upon the Full Ranger-EQ from our EQ Rangers series

It models SPL RackPack Eqs. Here's the quote from SOS review of hardware:

The Full Ranger has frequency centres of 40Hz, 90Hz, 150Hz, 500Hz, 1.8kHz, 4.7kHz, 10kHz and 16kHz, and the cut and boost is controlled by a set of miniature sliders. Unlike a traditional graphic equaliser, where the filters are set one octave apart with identical response curves, this SPL design tailors the response of each filter band based on musical principles (rather than mathematical ones). Essentially, the filter curves get wider (lower Q-settings) the further you go up the audio spectrum.

Basically, the idea is that SPL Free Ranger behaves as original hardware, not as a digital EQ.

[Lesha]

Basically, the idea is that SPL Free Ranger behaves as original hardware, not as a digital EQ.

So, let's mislead our customers and make them believe they are using 1800 Hz instead of 900 Hz.

Very clever.

[Loki Fuego]

Unlike a traditional graphic equaliser, where the filters are set one octave apart with identical response curves, this SPL design tailors the response of each filter band based on musical principles (rather than mathematical ones).

[Lesha]

That still does not answer my question, being it hardware or not.

Why does 1K8 slider work on another frequency?

[Loki Fuego]

Most probably, because hardware SPL Full Ranger behaves that way. Try boosting less, try analysing other bands. Try measuring response to different material. It's hardware modelled EQ, its response is made to behave like original hardware, not like digital EQ.

[Lesha]

Other Free Ranger bands behave normally, it's only this one that works on a wrong frequency.

[Lesha]

I have just downloaded a demo of Full Ranger and it gets even better
Look at the differences.

16K = 17 KHz
10K = 5500 Hz
4k7 = 2500 Hz
1K8 = 900 Hz
500 = 350 Hz
150 = 150 Hz
90 = 70 Hz
40 = 38 Hz

The only thing you can trust here are your ears.

[antto]

what camsr suggested...
perform your test again at a different sampling rate (88.2kHz for example)
if the resultant EQ frequencies change - then that's your answer.. most likely the IIR frequency tuning algorithms in that plugin were lousy

lots of the IIR filter/eq code examples from the internet (musicdsp archive for example) especially if they are old (from the times where CPUs were magnitutes slower) don't come with decent tuning formulas, or don't have tuning formulas at all
that's where the dev has to figure out some formula.. if the formula is lousy - changing sampling rate would change the cutoff frequency mapping of the filter

[Lesha]

I have already tried 88.2 KHz, the behaviour is exactly the same.

[WOK]

If you read product description then, you should have noted that:
It models SPL RackPack Eqs. Here's the quote from SOS review of hardware:

The Full Ranger has frequency centres of 40Hz, 90Hz, 150Hz, 500Hz, 1.8kHz, 4.7kHz, 10kHz and 16kHz, and the cut and boost is controlled by a set of miniature sliders. Unlike a traditional graphic equaliser, where the filters are set one octave apart with identical response curves, this SPL design tailors the response of each filter band based on musical principles (rather than mathematical ones). Essentially, the filter curves get wider (lower Q-settings) the further you go up the audio spectrum.

Basically, the idea is that SPL Free Ranger behaves as original hardware, not as a digital EQ.

I can't believe that the original hardware is that wrong, manipulating mostly 900Hz when moving a 1K8 slider....
And if the algos are wrong, it's strange that not all frequencies are wrong in the same way.
Anyway the answer of the vendor is crap ("it's not a bug, it's a feature...")

[antto]

I have already tried 88.2 KHz, the behaviour is exactly the same.

if the frequencies stay the same when running at different sampling rates - then the tuning is consistent, and the wrongness of the labels vs actual frequencies must be intentional