Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[oliverg]

Hi all,

I have a pair of Bowers Wilkins 801 Matrix S2s which comes with a "Bass Alignment Filter"

(https://kenrockwell.com/audio/b-w/800-series-bass-alignment-filter.htm)

It was advertised as a "high pass filter" and when I put it on my CRO, it seems to simply filter out anything below 20Hz. The explanation above seems to indicate that its a lot more complicated than just that?


I'm  wondering if I can ditch it and simply put a rule in the DSP Studio of "High-pass at 20Hz (Left,Right) and if I did, what slope should I use? Or is it not that simple?

I was also thinking of just using the JRSS Subwoofer feature and diverting anything below 20Hz to my dual subwoofers. would that function the same as the Bass Alignment Filter?

Thanks in advance

[mwillems]

The description suggests that the "bass alignment" filter is a second-order peaking high pass with a Q of 2.  The frequency response of the filter causes an increase in low frequency response as frequency goes down peaking at the cutoff point (20Hz in your case), and then falling very quickly below the cutoff point.  So, for example, the curves show that, set to 20Hz, it attenuates frequencies which are well below 20Hz, but it greatly boosts the frequencies right around 20Hz.  That's what they mean about extending the bass response a half octave lower, in addition to filtering out the subsonics.

None of the built in high pass filter options in JRiver have that kind of peaking topography, and the Q of the high-pass is not adjustable.  You could potentially hack something together using a combination of filters and/or a linkwitz transform, but you'd need a measuring microphone and some patience.

[JimH]

None of the built in high pass filter options in JRiver have that kind of peaking topography, and the Q of the high-pass is not adjustable. 

If you can describe to Matt what you'd like, we'll see if we can make one.  Thanks.

[mwillems]

If you can describe to Matt what you'd like, we'll see if we can make one.  Thanks.

Jim, thanks for the offer.  If the high and low pass filters permitted specifying the Q factors of the poles in addition to the slope of the filter, that would potentially solve this issue and allow for a lot more flexibility in crossover creation (i.e. it would allow creation of peaking low and high pass filters, bessels, and more complex topographies).  It would be nice to be able to dial in the Q for high pass and low pass filter options.

If it were implemented, I'd recommend putting it behind a check box or splitting it off from the main low pass and high pass options (e.g. custom high pass, etc.). At minimum, if it stays in the main low pass and high pass filter dialogs, the Q needs to default to a sensible value.  Most people will just want a butterworth with a Q of .707 (which is the current baked in setting).

There's lots of info in this article; Q is addressed in the second section, but there's useful info throughout:
http://sound.whsites.net/lr-passive.htm

mattkhan may also have some thoughts.

[mattkhan]

apologies to the OP but I think a variable Q high/low pass filter is a pretty niche thing to add, particularly if it were the only change you made. It would benefit a pretty small number of users.

I/we created this thread years ago on the subject of changes to dsp studio - https://yabb.jriver.com/interact/index.php/topic,99096.msg685880.html#msg685880 - and I think all of those are still valid feature requests.

IMO the 2 most valuable/useful things you could do are

1) add a new filter type "advanced biquad" to PEQ
2) provide some easier, less labour intensive, way to add/remove PEQ filters (e.g. to paste in a batch of filters in a text format)

With the first one, you give the user the ability to implement pretty much anything they want including any order of Butterworth or Linkwitz-Riley high or low pass filter (as you can simply create those by stacking biquads)
With the second one, you may it feasible to let people easily use external filter design software and import that directly into jriver quickly and easily.

Even better if you make point 2 available via MCWS so it can be automated easily. If you did this then I could integrate it with an app I wrote (https://beqdesigner.readthedocs.io/en/latest/) which is used over at avs (https://www.avsforum.com/forum/113-subwoofers-bass-transducers/2995212-bass-eq-filtered-movies.html). atm jriver is widely recommended as the way to do this on a computer but it is pretty tedious and error prone to do (because of the manual data entry and because of the way jriver uses S as the parameter for a shelf filter)

FWIW I rarely use PEQ in jriver because it is so tedious and time consuming to use if you want to add a lot of filters.

On the other hand, if you just want to add this particular feature and need the biquad formula for it then you can refer to https://github.com/3ll3d00d/beqdesigner/blob/master/src/main/python/model/iir.py#L555 and https://github.com/3ll3d00d/beqdesigner/blob/master/src/main/python/model/iir.py#L607 for a working implementation of how to calculate the biquad coefficients.

[mwillems]

apologies to the OP but I think a variable Q high/low pass filter is a pretty niche thing to add, particularly if it were the only change you made. It would benefit a pretty small number of users.

I/we created this thread years ago on the subject of changes to dsp studio - https://yabb.jriver.com/interact/index.php/topic,99096.msg685880.html#msg685880 - and I think all of those are still valid feature requests.

IMO the 2 most valuable/useful things you could do are

1) add a new filter type "advanced biquad" to PEQ
2) provide some easier, less labour intensive, way to add/remove PEQ filters (e.g. to paste in a batch of filters in a text format)

With the first one, you give the user the ability to implement pretty much anything they want including any order of Butterworth or Linkwitz-Riley high or low pass filter (as you can simply create those by stacking biquads)
With the second one, you may it feasible to let people easily use external filter design software and import that directly into jriver quickly and easily.

Even better if you make point 2 available via MCWS so it can be automated easily. If you did this then I could integrate it with an app I wrote (https://beqdesigner.readthedocs.io/en/latest/) which is used over at avs (https://www.avsforum.com/forum/113-subwoofers-bass-transducers/2995212-bass-eq-filtered-movies.html). atm jriver is widely recommended as the way to do this on a computer but it is pretty tedious and error prone to do (because of the manual data entry and because of the way jriver uses S as the parameter for a shelf filter)

FWIW I rarely use PEQ in jriver because it is so tedious and time consuming to use if you want to add a lot of filters.

On the other hand, if you just want to add this particular feature and need the biquad formula for it then you can refer to https://github.com/3ll3d00d/beqdesigner/blob/master/src/main/python/model/iir.py#L555 and https://github.com/3ll3d00d/beqdesigner/blob/master/src/main/python/model/iir.py#L607 for a working implementation of how to calculate the biquad coefficients.

Thanks matt, I knew you had addressed this in long thread, but couldn't immediately find it.  Thanks for the additional material and the implementation details!

[oliverg]

Thanks for the informative and excellent responses. I'll keep watching this space and hope the relevant features are implemented.

[Matt]

Coming next build:
NEW: The Q factor can be set for low and high pass filters in Parametric Equalizer.

Thanks for the idea