jbuhajla

Using a parametric EQ with a very narrow Q will prevent you from cutting out frequencies that you don't wish to attenuate. You'll never isolate one pure fundamental frequency, but using as narrow a peak as possible will minimize that effect. 

If you set the Q first, then boost the output, you can scroll through the frequencies to isolate the spot that's most objectionable, then roll the output back down, and that should give you a fairly narrow cut. 

Did you try playing with the input impedance? 

Most of the time I use a Gibson LES PAul Classic. What I am finding now is that if I use AMP +CAB on more than a few modules the clean tones are good and full until you add any OD or distortion from the helix. It woul be great if they had in the editor a way to trap out specific frequencies because the objectionable sound I would like to remove is reasonably consistent and tends to be bright. If you try to use any EQ you end up eliminating wanted freqeuncies along with the one you want to eliminate

Tried input impedance , but it made matters worse.

You mentioned above using an IR  in your signal chain... if its a third party IR you can apply EQ, but you're stuck with whatever mic they used. Using  the stock cabs, you have a choice of  mic models, and distance. Some mic models (especially the ribbon mics) apply a much darker color to your sound, and may help eliminate some of the harshness you're hearing.