Hi Everyone,
First a big thank you to the dev(s), EQAPO ist really a powerfull tool!
I need some help, I'm trying to put a Linkwitz-Transform on my subwoofer output, but when I activate the filter, the sound is muted. The filter was built according to the example biquad, which works just fine.
I also tried to specify the filter for different sample rates, but this wont work either. For better analysis I tested it globally, so the problem wont be in my channel-copying configuration.
My code:
Thanks for the fast reply.
I'm not into biquad to understand, what a division by 8 means - tested it though: It does not mute anymore, no measurable effect between 20 an 200 Hz :D
You meant it like this, didn't you?
Maybe MiniDSP scales the coefficints automatically. Here's an image showing one of the filters from your 1st post (is this what you're after?):
Blue line is the original filter. Red line is based to the coefficients which were dvided by 8.0. It looks like you need use much bigger divider to get the filter down to suitable dB area.
Be careful with the '-' sign of the coefficients.
Last edit: Juha 2016-05-21
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Ok, I get that the filters provided by the sheet are probably for use with miniDSP only. Is there a workaround or a possibility to program a Lowshelf or Biquad with a behaviour similar to this (for example)? For this is, what the Filter above will do in miniDSP
I have not checked the functions used in MiniDSP's spreadsheet but IIRC, someone mentioned few years ago that MiniDSP uses some kind of reverse signing (or ordering) for biquad coefficients.
'-' means just the minus sign in front of coefficient.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I do know what a minus sign looks like :'D I have not yet worked myself into the depths of biquad, so I don't know how a change to the coefficients will affect the filter :/
I think I just have do take a measurement of the mini-DSP Filter for the Linkwitz-Transform and then program the lowshelf in APO which comes the nearest (Sounds awkward to me, sorry if I messed up the grammar^^)
Thanks for your advice! Feel free to write if a similar situation comes across.
Cheers
Last edit: johein 2016-05-21
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
OK, had some time to look this closer. I prepared a c++ function (based on coding found in spreadsheet from minidsp site) for to make the calculations.
Coefficients you already have there produces this curve:
when you use this order for the coefficients (this is the order in SynthEdit i used for to prepare a plug-in for the measure):
Hi Everyone,
First a big thank you to the dev(s), EQAPO ist really a powerfull tool!
I need some help, I'm trying to put a Linkwitz-Transform on my subwoofer output, but when I activate the filter, the sound is muted. The filter was built according to the example biquad, which works just fine.
I also tried to specify the filter for different sample rates, but this wont work either. For better analysis I tested it globally, so the problem wont be in my channel-copying configuration.
My code:
The given coefficients work just fine with my miniDSP, so they are viable.
Any help would be greatly appreciated!
Greetings from northern Germany :)
Probably just unstable filters.
Does it work if you divide those 1st three coefficients of both sets by lets say 8.0?
Thanks for the fast reply.
I'm not into biquad to understand, what a division by 8 means - tested it though: It does not mute anymore, no measurable effect between 20 an 200 Hz :D
You meant it like this, didn't you?
As mentioned above, the filter is stable using miniDSP.
Edit: Is there a way to convert a Linkwitz-Transform into a lowshelf with identical behaviour?
Last edit: johein 2016-05-21
Maybe MiniDSP scales the coefficints automatically. Here's an image showing one of the filters from your 1st post (is this what you're after?):

Blue line is the original filter. Red line is based to the coefficients which were dvided by 8.0. It looks like you need use much bigger divider to get the filter down to suitable dB area.
Be careful with the '-' sign of the coefficients.
Last edit: Juha 2016-05-21
Ok, I get that the filters provided by the sheet are probably for use with miniDSP only. Is there a workaround or a possibility to program a Lowshelf or Biquad with a behaviour similar to this (for example)? For this is, what the Filter above will do in miniDSP
edit: And whats the matter with the '-' ?
Last edit: johein 2016-05-21
I have not checked the functions used in MiniDSP's spreadsheet but IIRC, someone mentioned few years ago that MiniDSP uses some kind of reverse signing (or ordering) for biquad coefficients.
'-' means just the minus sign in front of coefficient.
I do know what a minus sign looks like :'D I have not yet worked myself into the depths of biquad, so I don't know how a change to the coefficients will affect the filter :/
I think I just have do take a measurement of the mini-DSP Filter for the Linkwitz-Transform and then program the lowshelf in APO which comes the nearest (Sounds awkward to me, sorry if I messed up the grammar^^)
Thanks for your advice! Feel free to write if a similar situation comes across.
Cheers
Last edit: johein 2016-05-21
OK, had some time to look this closer. I prepared a c++ function (based on coding found in spreadsheet from minidsp site) for to make the calculations.
Coefficients you already have there produces this curve:
when you use this order for the coefficients (this is the order in SynthEdit i used for to prepare a plug-in for the measure):
Notice the sign change b1/b2!
So, coefficients got from the spreadsheet are valid but you need to do some changes before using.
EDIT: Here is a Linkwitz Transform script: https://sourceforge.net/p/equalizerapo/discussion/general/thread/e8a40cc8/
Last edit: Juha 2016-05-22
Awesome, thank you so much! This kind of script would be exactly what I was looking for. :)
Great work.