ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: I am trying to figure out how to divide frequencies across 20hz to 20khz  (Read 807 times)

Kevin Maxwell

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1492

I am trying to figure out how to divide frequencies across the 20hz to 20k spectrum.

I am trying to setup EQ points for a program that will be controlling the parametric EQ in a mixer. The mixer puts out 960 different data points (from 20hz to 20k) and in the program I am restricted (at the moment) to 256 data points. I am hoping I will be able to use just 256 data points for the channel EQ and I will be close enough that the proper frequencies that will fit in the 256 row limitation. Since there are basically 11 octaves in this range with 12 notes in each that gives me 132 rows and I am thinking of instead of 12 per octave maybe some of the octaves in the middle of the range might be better to divide it up so it is 36 points per octave in those ranges. It looks like I can do 5 octaves at 36 data points per octave and 6 octaves at 12 per octave. I am just trying to figure out which octaves I should do that to, I was thinking of doing the 36 points per octave between about 261hz to 7.9khz. Also I can’t find a chart that shows the note of the highest octave from above 15804.264hz to 20khz.

Can anyone help me with this or point me to where I might find the answers?
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1381
  • Edmonton, AB

I am trying to figure out how to divide frequencies across the 20hz to 20k spectrum.

I am trying to setup EQ points for a program that will be controlling the parametric EQ in a mixer. The mixer puts out 960 different data points (from 20hz to 20k) and in the program I am restricted (at the moment) to 256 data points. I am hoping I will be able to use just 256 data points for the channel EQ and I will be close enough that the proper frequencies that will fit in the 256 row limitation. Since there are basically 11 octaves in this range with 12 notes in each that gives me 132 rows and I am thinking of instead of 12 per octave maybe some of the octaves in the middle of the range might be better to divide it up so it is 36 points per octave in those ranges. It looks like I can do 5 octaves at 36 data points per octave and 6 octaves at 12 per octave. I am just trying to figure out which octaves I should do that to, I was thinking of doing the 36 points per octave between about 261hz to 7.9khz. Also I can’t find a chart that shows the note of the highest octave from above 15804.264hz to 20khz.

Can anyone help me with this or point me to where I might find the answers?

Can you describe your application in more detail? I'm not sure I understand what you're trying to achieve in practice.

-Russ
Logged

Kevin Maxwell

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1492

Can you describe your application in more detail? I'm not sure I understand what you're trying to achieve in practice.

-Russ

The application is that I am attempting to write a Palladium mixer file for the Behringer WING. The Channel EQ is used as Actor tracking EQ. The EQ you set for an Actors mic follows them where ever they are on the mixer. If the actor is wearing a hat and needs a different EQ the trick is to just rename the actor (for example BOB) for those scene or cues to something like “BOB Hat”. And that EQ will follow when “BOB Hat” has the lines. And back again when it is just BOB (no hat). 

The number of OSC EQ commands that the WING outputs when the frequency knob is rotated are different than on the X32 or M32 mixers. Palladium works for many different mixers and I am trying to set things up so it can also work with the WING.

I think the resolution I would get using the frequency spacing I mentioned above would work for this application just fine because when used this way for Actor Tracking EQ is usually just to make the mic sound good for that actor it isn’t usually pinpoint feedback frequency EQing. It is more tonal shaping.
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1381
  • Edmonton, AB

The application is that I am attempting to write a Palladium mixer file for the Behringer WING. The Channel EQ is used as Actor tracking EQ. The EQ you set for an Actors mic follows them where ever they are on the mixer. If the actor is wearing a hat and needs a different EQ the trick is to just rename the actor (for example BOB) for those scene or cues to something like “BOB Hat”. And that EQ will follow when “BOB Hat” has the lines. And back again when it is just BOB (no hat). 

The number of OSC EQ commands that the WING outputs when the frequency knob is rotated are different than on the X32 or M32 mixers. Palladium works for many different mixers and I am trying to set things up so it can also work with the WING.

I think the resolution I would get using the frequency spacing I mentioned above would work for this application just fine because when used this way for Actor Tracking EQ is usually just to make the mic sound good for that actor it isn’t usually pinpoint feedback frequency EQing. It is more tonal shaping.

It's funny; I was wondering if this might have to do with Palladium (or course, I'm pretty sure my old license didn't support anything like 256 frequency points, but it's pretty old). I'm pretty sure I was the first person to get it working with the X32 (before Palladium supported OSC, and before the X32 supported MIDI), but I digress.

Honestly, my suggestion would be to do whatever will be easiest to program (which, I would guess, is a linear-per-octave distribution). I'm not sure how golden your ears are, but mine would have a heck of time telling if a typical bell filter had been adjusted up or down by a quarter-tone (i.e. 24 points per octave), let alone less, so in my mind this probably isn't worth fussing over (the same as why I never worried that, back in the "old days" on the LS9, EQ changes were only ever recalled within 0.5 dB of where they were originally set).

-Russ
Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 16828
  • Hickory, Mississippi, USA
    • Resotune

256 is 8 bit resolution... I'd start with similar fraction of octave spacing like used in GEQ (graphic EQs).

JR
Logged
Cancel the "cancel culture". Do not participate in mob hatred.

Kevin Maxwell

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1492

256 is 8 bit resolution... I'd start with similar fraction of octave spacing like used in GEQ (graphic EQs).

JR

I am probably not explaining the situation well enough. The 256 is the number of rows, each row is an EQ frequency and the associated Hex of the frequency. So I am trying to get the highest resolution possible with those 256 frequencies that can be setup to work with this. Also I am limited by the actual frequency points that the mixer outputs when the frequency knob is turned. I have the ability to monitor and output that data (the frequency and the Hex of that frequency) when I slowly turn the frequency knob. So I can then enter that data in the knob position table that is used to match and store the data sent to and from the mixer.
Logged

Doug Fowler

  • Member since May 1995, 2nd poster on original LAB, moderator on and off since 1997, now running TurboMOD v1.826
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1979
  • Saint Louis, MO USA

I am probably not explaining the situation well enough. The 256 is the number of rows, each row is an EQ frequency and the associated Hex of the frequency. So I am trying to get the highest resolution possible with those 256 frequencies that can be setup to work with this. Also I am limited by the actual frequency points that the mixer outputs when the frequency knob is turned. I have the ability to monitor and output that data (the frequency and the Hex of that frequency) when I slowly turn the frequency knob. So I can then enter that data in the knob position table that is used to match and store the data sent to and from the mixer.

So you're looking to implement a 256 band EQ from 20-20K, similar to a 32 band GEQ except you'll have 256 bands?

I'll bet someone with under-the-hood (under-the-bonnet?) measurement chops can figure this out.  Fractional octave banding or some such.

Or check out this nifty chart from NTI:
https://www.nti-audio.com/portals/0/data/en/Fractional-Octave-Band-Filter.pdf
Logged
"It's got electrolytes.  It's got what plants crave."

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1381
  • Edmonton, AB

I am probably not explaining the situation well enough. The 256 is the number of rows, each row is an EQ frequency and the associated Hex of the frequency. So I am trying to get the highest resolution possible with those 256 frequencies that can be setup to work with this. Also I am limited by the actual frequency points that the mixer outputs when the frequency knob is turned. I have the ability to monitor and output that data (the frequency and the Hex of that frequency) when I slowly turn the frequency knob. So I can then enter that data in the knob position table that is used to match and store the data sent to and from the mixer.

I do think I understand, I just think you're over-thinking it is all. 256 possible frequencies divided evenly over the span of 20 Hz to 20 kHz should be more than granular enough. Heck, the X/M32 only offers ~200 possible frequencies on its parametric EQs, and I've never heard anyone complain about them being "low resolution".

-Russ
Logged

Riley Casey

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1689

ASSuming that the high pass filter doesn't eat into your EQ DSP if you really do need more granularity why not restrict the frequency spectrum to 100 - 10khz. That would allow you more frequency choices in the range of interest.


... The Channel EQ is used as Actor tracking EQ. The EQ you set for an Actors mic follows them where ever they are on the mixer. If the actor is wearing a hat and needs a different EQ the trick is to just rename the actor (for example BOB) for those scene or cues to something like “BOB Hat”. And that EQ will follow when “BOB Hat” has the lines. And back again when it is just BOB (no hat). 

...

Mark Wilkinson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1064

So you're looking to implement a 256 band EQ from 20-20K, similar to a 32 band GEQ except you'll have 256 bands?

I'll bet someone with under-the-hood (under-the-bonnet?) measurement chops can figure this out.  Fractional octave banding or some such.

Or check out this nifty chart from NTI:
https://www.nti-audio.com/portals/0/data/en/Fractional-Octave-Band-Filter.pdf

Thanks Doug,

Those charts make sense to me, because i feel the task is how do i best divide 256 linearly spaced points, into log space like sound works.

It all comes down to log math...which i forgot in HS ... about 50 years ago lol
Logged

ProSoundWeb Community

Re: I am trying to figure out how to divide frequencies across 20hz to 20khz
« Reply #9 on: June 20, 2021, 03:58:38 PM »


Pages: [1] 2  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.044 seconds with 20 queries.