Quantization of effect parameters

[emagli]

Hi there,

 

I am trying to set up an eq block where one of the center frequencies is 1120 Hz (because that's the exact center frequency of my analog mid boost, which I am trying to emulate using a parametric eq block). However, with HX edit I can only choose minimum increments of 100Hz (and increments of 0.1 for the Q), so I can select either 1100 or 1200 Hz. I can type any integer number in the frequency field, but then it is rounded off to the nearest 100 Hz increment. Do you know how to set the parameter to any number? I guess it should be feasible, as I have the Mbritt package and his IR blocks have very precise and not rounded-off frequencies.

[phil_m]

I don’t really understand your comparison of the EQ block to the IR block. They’re two different things, and they work very differently. They just don’t the available frequencies broken up to that precision in the EQ blocks. Even with the narrowest Q setting, there’s hardly going to be any audible difference between selecting 1100 and 1120 Hz.

[emagli]

Sorry that I phrased this poorly... For the IR block I mean that the low and high cuts can be set to any value. I have a third-party IR block with high cut set to 11648.35 Hz. Agreed that the difference with 11.5 kHz may not be audible, but just for the sake of understanding how this works, I have tried to edit parameter values directly in the .hlx preset file, and then reload the preset. For the eq block, doing that allows me to change the center frequency (but not the Q) and level to any value.

 

In my application, for the same value of Q,  quantizing the center frequency to 1100 with respect to 1120 Hz leads to a difference in the frequency response up to 0.5 dB at 5 kHz. I doubt that I am able to hear that (and certainly not in a mix or live context), but for the sake of matching a given frequency response as accurately as possible, it would be nice if HXedit allowed to input an arbitrary value. At least for the parameters for which there are no constraints in the algorithm implementation, it should be just an UI thing.

[DunedinDragon]

Hmmmm...EQ'ing by eye.  Do you also watch TV by sense of touch???

 

[phil_m]

6 hours ago, emagli said:

Sorry that I phrased this poorly... For the IR block I mean that the low and high cuts can be set to any value. I have a third-party IR block with high cut set to 11648.35 Hz. Agreed that the difference with 11.5 kHz may not be audible, but just for the sake of understanding how this works, I have tried to edit parameter values directly in the .hlx preset file, and then reload the preset. For the eq block, doing that allows me to change the center frequency (but not the Q) and level to any value.

 

In my application, for the same value of Q,  quantizing the center frequency to 1100 with respect to 1120 Hz leads to a difference in the frequency response up to 0.5 dB at 5 kHz. I doubt that I am able to hear that (and certainly not in a mix or live context), but for the sake of matching a given frequency response as accurately as possible, it would be nice if HXedit allowed to input an arbitrary value. At least for the parameters for which there are no constraints in the algorithm implementation, it should be just an UI thing.

 

I know people have edited those text files to change those parameters, but one thing I’ve always wondered about that, especially in regards to the EQ, is if the actual effect is changing. It may be just be that you’re only really changing the label and not anything in the underlying effect.

[emagli]

2 hours ago, phil_m said:

 

I know people have edited those text files to change those parameters, but one thing I’ve always wondered about that, especially in regards to the EQ, is if the actual effect is changing. It may be just be that you’re only really changing the label and not anything in the underlying effect.

 

Good point, I was also afraid that this might be the case, but it is not. I have my Helix LT with my analog eq pedal in one of the effect loops, and a software on my PC that can test the impulse response by sending a test signal through Helix via USB, recording the output and deconvolving. That way, I can observe tiny differences in the impulse response as I change the parameters (this approach is only technically sound if the system being tested is linear, which is the case of an EQ). I can tell you without any doubt that for the parameters that I managed to change (center frequency and level) I measured the exact change that I was expecting. On the other hand, I do not necessarily feel very comfortable using a feature that is not documented, but so far so good. 

[emagli]

5 hours ago, DunedinDragon said:

Hmmmm...EQ'ing by eye.  Do you also watch TV by sense of touch???

 

 

I see your point :-) To put this in perspective: I have been able to perfectly replicate the behavior of a mid boost stompbox whose tone I love, but which occupies a lot of real estate on my board connected to Helix. Now I can get rid of the stompbox. In the process I have also tried to analyze and replicate my compressor stompbox. Long story short, I have learned that (at least from my findings), the LA2A works as a sustainer in addition to providing compression, which allows me to do away with my compressor stompbox (which I was using as a compressor/sustainer) and free one effects loop. And yes, I did not get to play any real music as I was doing this, so that's all for the experimentation, and now I'm back to playing guitar.