Vulcan question HF roll off question

[seedlings]

The opamp gain-bandwidth product is a musical feature because higher gains have the treble rolled back, which sounds natural to me.  Right now I have a Vulcan circuit on the breadboard with slight capacitor changes.
http://www.diystompboxes.com/analogalchemy/sch/vulcan.html

I really, really like the sound and response, especially how this circuit does 'clean to mean' very well - the output level sounds the same loudness throughout 90% of the drive knob range.  The one-and-only improvement I feel it needs is some high treble roll-off as the gain goes from noon to full (sort of like an opamp).  A 120pF bright cap on the volume pot along with a permanent NFB 68pF B-C on Q3 almost, almost gets there. I also tried some global high pass NFB.  I considered a dual-gang drive+tone, but do you have any other suggestions?

Thanks,
CHAD

[Paul Marossy]

I really, really like the sound and response, especially how this circuit does 'clean to mean' very well

I don't really have a suggestion to achieve your quest, but I have observed that the many circuits that have an interstage volume control can be very "touch sensitive" and quite amp-like in their response to the guitar player's dynamics.

OK, maybe I have a suggestion. Have you tried a small cap to ground anywhere? If you have more treble than you want, maybe a small cap to ground somewhere might help with that. Maybe something like a 470pF or .001uF on the output might work without affecting the sound too much.

[slacker]

You could try caps in parallel with the 10k resistors from the collectors to 9 volts, some tube amps do that to roll off highs. You could try any of the collectors or a combination of them to roll off highs across the stages.

[seedlings]

I've tried those things (cap across volume pot, cap across Rc), but I only want the highs to roll off as the gain is turned up.  I'm thinking this may require adding 10pF at-a-time to brighten the drive, while also bleeding 10pF at-a-time to ground somewhere past that.

Can the gain-bandwidth of a transistor be reduced?  Is this just a matter of reducing the Rc and Re until the transistor is drawing enough current?  I'm looking at a 2n5088 datasheet and can't decipher.

CHAD