Welcome, Guest
Username: Password: Remember me
  • Page:
  • 1
  • 2

TOPIC: Mixed FIR/IIR filters for a 4-way speaker

Mixed FIR/IIR filters for a 4-way speaker 3 years 2 weeks ago #27478

  • pos
  • pos's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 345
  • Thank you received: 196
  • Karma: 39
Hi Thomas,

Ouch! that is a nasty bug :ohmy:
Thank you for spotting it.

Last question: What does the "Horbach-Keele last" filter mean?
This is the "special tweeter" setting that you are looking for :)
download the last version of rephase here: rephase.org
Last Edit: 3 years 2 weeks ago by pos.
The administrator has disabled public write access.
The following user(s) said Thank You: samoht

Mixed FIR/IIR filters for a 4-way speaker 3 years 1 week ago #27491

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Pos,

Thanks a lot for the quick answer.
I must repeat rephase is a really great tool.

Best regards

Thomas
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 3 years 1 week ago #27507

  • pos
  • pos's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 345
  • Thank you received: 196
  • Karma: 39
samoht wrote:
The ratio 1.0 is used by Horbach-Keele for implementimg a single central tweeter.

Hi Thomas

Where do you get the info from?
I cannot find any reference implementation, and could not find any allusion about this in the 2007 AES HK papers and presentations.
What did you use to calculate the critical frequencies and R values?

I don't remember where I got this [0.5;4.5] range for R from, byt the way :D
download the last version of rephase here: rephase.org
Last Edit: 3 years 1 week ago by pos.
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 3 years 1 week ago #27522

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Pos,

I do not remember exactly where i got this information from.
Let me check my folder and i will come back to you this evening.

Best regards

Thomas
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 3 years 1 week ago #27538

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Pos,

I have just the 2 papers from Horbach and Keele, Part 1 and Part 2, presented at the 32nd AES conference 2007 including the associated power point presentations., nothing more.
I cannot find a written confirmation that R = 1 for the filter calculation of the central tweeter.
But if you look to the Hlp formula on slide 17 of Part 2 ppt and use R=1, you will see that the denominator becomes 0 which is an invalid operation and forbidden. In so far the indicated range for fn and R (1<=fn<=R ) is not correct for R=1. Therefore i thought that the Hlp formula for the single central tweeter on slide 19 of the same ppt was choosen for that reason as it does not use R as a parameter anymore.
Maybe, that this is my wishful thinking but i have no other explanation.
Sorry for that.
But let me ask which value(s) for R are using for your H-K last filter?

Best regards

Thomas
Last Edit: 3 years 1 week ago by samoht.
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 3 years 1 week ago #27539

  • pos
  • pos's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 345
  • Thank you received: 196
  • Karma: 39
Hi Thomas,

OK, so we have the same info and sources.
I don't remember where I got that range from, and it is obvious R=1 will crash things...

Note that there is also an error in this formula, as they state "0 otherwise" where it should be 1 or one side and 0 on the other.
That is what I implemented.

I also used the formula from the paper for the "last" setting, and the proposed taper.
Now I wander if the taper is a good thing as if the presentation states it is meant to avoid rapid change in coverage, the paper states it is only meant to simplify design...
Maybe I should add an un-tapered variant...

How did you calculate the critical frequency and R values for your design ?
download the last version of rephase here: rephase.org
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 3 years 1 week ago #27585

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Pos,

As you maybe read, i had my speakers ready already when i found the Horbach-Keele principle to replace a classic D'Appolito crossover. Meanwhile i Know that this was wrong because you have use H-K filters for the whole frequency range of the speaker. Nevertheless i did it just with one crossover and got at least a better vertikal diagramm than the normal D'Appolito.
First I defined the D/lambda = 0.4 in order to use the lowest xover frequency.
Then i calculated the critical frequency by using the formula fc = c/D x 0.4 with D = 0.24 m. Fc becomes 567 Hz.
I'm currently planning to build a complete Horbach- Keele speaker and i'm looking to find the right small tweeters and midrange tweeter.

Best regards

Thomas
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 2 years 11 months ago #27769

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Devteam,

The work on my 4-way speaker is finished and thanks to OpenDRC and rephase it is a really good speaker now. I'm fully satisfied with the performance of the OpenDRC DA8, it's a phantastic tool.
My next speaker will be a challenge because it will be a 5- to 6-way Horbach-Keele speaker. The design is under progress and so far seems to be feasible.
I have to use two OpenDRC DA8 one for each stereo channel to cover the 5 to 6 frequency bands.
The Horbach-Keele design requires a lot of FIR power also in the low frequency band. Therefore I have only one big wish to Devteam.
Please increase the upper limit of taps from 2048 to 4096 at least for the 48kHz plugin. This would improve the woofer setting a lot.
Total number of taps is okay, but in my case I need more than 2048 taps per channel.


Best regards

Thomas
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 2 years 7 months ago #30316

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Pos,

Long time no conversation about your rephase tool.
The concept for my 5-way Horbach Keele speaker is ready and validated.
Unfortunately Devteam did not react on my wish to increase the max. number of taps per channel, which is a really pitty.
Subwoofer and bass woofer definitely need more taps than 2048.
I' m going to use 2 x OpenDRC DA for the Subwoofer and woofer having 6144 taps per channel. In addition there will be 1 x OpenDRC DA8 for the remaining 6 channels. Critical frequencies and Rs are calculated and I have started to generate the FIR filters.

In order to get the lowest sidelobes I found out that the "albrecht-2term" window obviously is best.
My question: Does the best visual appearance of the filter curve also provides always the best electrical characteristics?

Best regards

Thomas
Last Edit: 2 years 7 months ago by samoht.
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 2 years 7 months ago #30326

  • devteam
  • devteam's Avatar
  • Online
  • Administrator
  • Posts: 8459
  • Thank you received: 1037
  • Karma: 156
Hi Thomas

Sorry we didn't get this feature implemented into the 4x8 plugin.. we've been very busy the whole team lately so it's been a bit stretched to add that new feature (need new DSP firmware + new plugin). It's still about 1.5week work based on our typical process for releasing this..

Need to somehow get a voting system going on here to realize how much people really want a feature.. so we know the pulse..
Will see in the future if we can do something to 4x8 minisharc plugin.

DevTeam
MiniDSP, building a DIY DSP community one board at a time.
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 2 years 7 months ago #30332

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Devteam,

These are really great news.
Thank you very much.
I' m excited to learn more about this feature.

Best regards

Thomas
The administrator has disabled public write access.

Mixed FIR/IIR filters for a 4-way speaker 2 years 7 months ago #30384

  • pos
  • pos's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 345
  • Thank you received: 196
  • Karma: 39
samoht wrote:
In order to get the lowest sidelobes I found out that the "albrecht-2term" window obviously is best.
My question: Does the best visual appearance of the filter curve also provides always the best electrical characteristics?

Hi Thomas,

The Albrecht windows are quite nice indeed.
The more terms you use the lower the ripples, but of course also the slower the slope.
The 2 term version is equivalent to a Hamming window.

The red curves will tell you exactly what you should measure at the output of the convolution (and the larger the FFT the more precise the view down low, with a fixed number of taps).
You can also look at the impulse itself, possibly in dB, to observe the discontinuities at the limits of the impulse, and also the effect of the optimization algorithm if you used it (this is especially true when you are doing minimum-phase corrections, as the optimization algorithm will have a tendency to steer away from minimum-phase relationship and result in pre ringing).
download the last version of rephase here: rephase.org
The administrator has disabled public write access.
The following user(s) said Thank You: devteam, samoht, denver8me4dinner

Mixed FIR/IIR filters for a 4-way speaker 2 years 2 months ago #32520

  • samoht
  • samoht's Avatar
  • Offline
  • Expert Boarder
  • Posts: 111
  • Thank you received: 9
  • Karma: 1
Hi Pos,

Long time since we spoke to each other.
My 5-way Horbach Keele Speaker cabinets are ready and the drivers are in.
Now i'm implementing the Horbach Keele filters using rephase.
As said earlier for the Woofer and bass driver i'm using an OpenDRC-DA with 6144 taps per channel. So far so good.
Part of the optimization is the compensation of the mechanical offset between both drivers. Having done this, the combined amplitude response is very flat, nearly perfect without a dip at the crossover frequency. But if i look to the step response it looks like that the signals of the woofer and the bass have opposite polarity. Is this possible? There are no other filters involved, except that i linearized the closed box rolloff.
Your feedback is highly appreciated.

Here comes a picture. The side coverings have to be painted of course.
It's my big boy: 1.65 m tall, 0.73 m wide with a weight of 80 kg.
9 drivers per cabnet fired by 7 amps.




Best regards

Thomas

This message has an attachment image.
Please log in or register to see it.

Last Edit: 2 years 2 months ago by samoht.
The administrator has disabled public write access.
The following user(s) said Thank You: devteam

Mixed FIR/IIR filters for a 4-way speaker 2 years 2 months ago #32574

  • pos
  • pos's Avatar
  • Offline
  • Platinum Boarder
  • Posts: 345
  • Thank you received: 196
  • Karma: 39
Hi Thomas

Very nice build! B)
7 amp channels per speaker!! :side:

Regarding your issue with the step measurement, if you measure a flat phase curve across the crossover then you should not have any polarity issue.
download the last version of rephase here: rephase.org
The administrator has disabled public write access.
The following user(s) said Thank You: samoht

Mixed FIR/IIR filters for a 4-way speaker 2 years 2 months ago #32634

  • Zas
  • Zas's Avatar
  • Offline
  • Fresh Boarder
  • Posts: 12
  • Karma: 0
devteam wrote:
Hi Thomas

Sorry we didn't get this feature implemented into the 4x8 plugin.. we've been very busy the whole team lately so it's been a bit stretched to add that new feature (need new DSP firmware + new plugin). It's still about 1.5week work based on our typical process for releasing this..

Need to somehow get a voting system going on here to realize how much people really want a feature.. so we know the pulse..
Will see in the future if we can do something to 4x8 minisharc plugin.

DevTeam

Hey DevTeam,

I'm using only 4 channels on my minisharc, and all my channels are using 2048 taps, it would be fantastic to be able to use more taps per channel, especially for low frequencies. Any update on this ?
The administrator has disabled public write access.
  • Page:
  • 1
  • 2
Moderators: devteam