TOPIC:

Hello,
i'm new to the MiniDSP biquad world so sorry for that (maybe) dumb question.I have a SHD with firmware 1.15. To calculate the biquads I use the spreadsheet v1.2 / v1.3.
I want to create an HPF Filter at 260Hz with an Q of 0,7.
When I copy the values ​​to the SHD I get no curve and absolutely nothing at the outputs.
I tried 96kHz and 48kHz. According to the manual 48 should be used but maybe thats a typo.
As far as i know the SHD works with 96kHz. Both values do not work at all.I don't understand what I'm doing wrong and hope you can help.

The SHD uses a 96 kHz internal sample rate.

I've never tried to manually calculate biquads to help more on this, but I take it that the pre-defined crossover fiters don't do what you'd want?

 

Exactly, I received some special filters from the manufacturer of my loudspeakers which I would like to copy as exactly as possible.  The 260 Hz filter is just one example.  In the end i would like to understand how it works.

@sacha157, did this get resolved? I have the same issue. I entered the biquad values, pressed Calculate & Apply, no error messages, and absolutely nothing happened, the freq response remained flat.. 2x4HD with the new Device Console.

Jan,

Please post your biquad coefficients.

Dave.

Hi Dave, this is the set the online calculator gave me:

a1 = -1.98682529
a2 = 0.98699552
b0 = 4.25576816e-5
b1 = 8.51153632e-5
b2 = 4.25576816e-5

Online calculator is at arachnoid.com/BiQuadDesigner/index.html
The order of the data has to be changed so that the b-terms come first; with the a-terms first as shown, the miniDSP s/w throws an error.

Hopefully you can get me going! Thanks beforehand.

Jan

Well, it's not formatted correctly and you have the wrong sampling rate for the 2x4HD.
Regardless, those coefficients don't look correct for the filter you're trying to achieve.
Is this just a simple 100Hz low-pass filter with Q=1.0?

I suggest to use the info/spreadsheet located here:
www.minidsp.com/applications/advanced-to...d-biquad-programming

Dave.

Thanks Dave, that's helpful! I did play with the sample rate in that online calculator, and got a non-protesting reaction from the 2x4HD but the curves were weird.
Then I switched to the spreadsheet you mentioned, that looks like a cool tool!

After a bit of trial and error and swearing I got it to work with the proviso:
Do not use A0, since it is 1 it is not needed but if you include it it gives an error;
Start with the b-coefficients, then followed by a1, a2 ...

So I think I'm all set now and will continue my adventure with biquads!

PS I don't see any possibility to set or read the sample rate in the Device Console. 
Should that not be matched to the spreadsheet SR? Did I just miss it?

Jan

BTW If you re-arrange the output cells in the spreadsheet a bit you can insert the whole thing into the miniDSP in a single copy-paste.


Jan

Jan,

Good progress.
You don't have the option to change sample rate in the Device Console.  Each of the platforms has a native sampling rate.  It's up to you as the user to generate your biquad coefficients with the proper sampling rate to match the platform you're using.

Yes, on all the tabs of that spreadsheet I reflect/rearrange the coefficients into a blank area of each sheet in the proper format for easy copying/pasting.

Good luck.

Dave.

Can anyone educate me with regards to the BPF1 filter config in the biquad spreadsheet?
The inserted Fc and the filter shape don't seem to match, some times a factor > 10 difference.
I'm also unfamiliar with the use of Q as 'constant skirt gain'.
What is the use of such a filter?

Jan

Jan,

I think there's an error somewhere in the spreadsheet for that tab.  I've never used the BPF1 portion, so hadn't noticed this.  (We normally wouldn't use this type of thing for audio applications.)

The reference for all this filter implementation is Robert Bristow-Johnson.
If you google you can find his pdf's on the subject with the explanation for skirt and skirt gain.

This is a well laid out reference that I've used in the past:
www.musicdsp.org/en/latest/Filters/index.html

Dave.

This is a pretty awkward process it seems. I use the spreadsheet to generate two biquad files, one is a high pass filter, the other a shallow midrange suckout.
When ever I insert one or the other, it is processed correctly and shown in the graph.
But. Whenever I add a second one, as shown here: 

biquad1,
b0=0.989864576240592,
b1=-1.97972915248118,
b2=0.989864576240592,
a1=1.97948404490459,
a2=-0.9799742
biquad2,
b0 =0.955775598361364,
b1 =-1.146235832784200,
b2 =0.230895056452633,
a1 =1.146235832784200,
a2 =-0.186670654813997

After I click Process and Insert, the bottom part disappears and I only get this:

biquad1,
b0=0.989864576240592,
b1=-1.97972915248118,
b2=0.989864576240592,
a1=1.97948404490459,
a2=-0.9799742

Maybe there is a specific syntax for having more than one biquad but this isn't working.
And yes I tried lots of permutation with and without closing comma etc.
Grateful for any advice on how it really works.

Jan

I think I got it.
Although the unit supports up to 9 biquads, you can only enter one in each EQ slot. For the second biquad you need to go to PEQ2 etc.
Is there a way to add a comment or label in each biquad window to keep track of what is what, or is it back to the yellow postits
Onwards!

Jan

There are some bugs in the spreadsheet. There's a link to a descent biquad calculator in this tutorial. It automatically produces 8 biquads in the correct MiniDSP format: