Tilt filter, "blue-ing" filter (cough cough martin?)

[guyman]

btw that schematic is currently only and not

[martinvicanek]

I uploaded version 2 in the original post above.

@Adam, this was actually my first thought: I nailed the transfer function at 1 kHz. However, I realized that if you tilt whichever direction, you'll overflow the next stage. I found it annoying to have to adjust the volume each time I changed the slope. HTat's why I chose the other normalization, with pink noise resembling (very roughly) a musical source.

@guyman, not quite sure what exactly you want to achieve?

[guyman]

I want to create a smile/frown (low/high boost cut) filter consistent with both the db/octave law of power, and normalization technique you are implementing in these filters. Let's say 0/db per octave is x, or our input signal, unaltered across the spectrum. in one instance we tilt it y db resulting in both + and - gain changes, and in another instance we tilt it -y db resulting in both + and - gain changes. Again like you emphasized before this is not relevant to a changeable "center frequency" but to your approach to normalization (not sure I fully understand - which may be the root of me missing this whole thing.)

In my smile, or low/high boost filter, you tilt the filter both +-y db from x, but only keep the corresponding +gain changes from both tilts, and vice versa with the frown/ low/high cut filter. you keep the -y gain changes from x from -+y tilt.

like let's say you have purple noise (+6db/oct) and brown noise (-6db/oct), on the smile filter you would have the positive high frequency gain changes from the +6db/purple tilt, and the positive low frequency gain changes from the -6db/brown noise tilt, resulting in a boosting filter that is consistent to the spectrum.

perhaps I lack the language to articulate it, or my imagination has run far too rampant.

I'm going to fiddle with your updated schematics and see what's going on.

[guyman]

Actually my imagination is not failing me. This is a hyper-relevant tone nob, consistent with the laws of power and our hearing.

[guyman]

Is something like LU normalization possible?

[guyman]

Do you have a copy of the one that's nailed to 1k? I am assuming it doesn't have normalization.