• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Converting Klispch RP-160M to Active Crossover - REW/Equalizer APO

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
2,051
Likes
2,673
Location
NL
As far as your comment about "You should be running a simple high-pass on the tweeter, and low-pass on the woofer, at roughly 1500-2500Hz", the simple answer is: I am!
I’m not sure you are. Your current filter will still send < 500Hz content to your tweeters won’t it?
 
OP
P

paulgyro

Active Member
Joined
Nov 21, 2020
Messages
144
Likes
100
I’m not sure you are. Your current filter will still send < 500Hz content to your tweeters won’t it?
All I can do is show you what I'm doing on REW. See the highlighted setting on the right. Speaker driver, 1500 hz HP cutoff, Butterworth 8th order.
tweeter with crossover.png
 

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
2,051
Likes
2,673
Location
NL
If so, please share a screenshot of the graph at the bottom of its Configuration Editor window.
 

alex-z

Addicted to Fun and Learning
Joined
Feb 19, 2021
Messages
910
Likes
1,684
Location
Canada
All I can do is show you what I'm doing on REW. See the highlighted setting on the right. Speaker driver, 1500 hz HP cutoff, Butterworth 8th order.
View attachment 246266

Those are the REW target curve settings that define your frequency slope. What you have actually applied in EqualizerAPO is not the same, you shared a list which included 18 PK filters, but not a single high-pass filter.

REW expects you to have externally applied the low and high pass filters before the measurement process.

8th order is unnecessarily steep. 4th or even 2nd order will be perfectly adequate for most bookshelf speaker designs and offer you a smaller group delay peak.
 
OP
P

paulgyro

Active Member
Joined
Nov 21, 2020
Messages
144
Likes
100
Have you imported your filter in EQ APO yet?
I did and I could see the proper slops and then the lower frequencies rose back up, this is one of the primary reasons I posted here. I can't upload a current screen shot as I'm messing around with Peace and it over wrote everything.
 

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
2,051
Likes
2,673
Location
NL
Right I see. I don’t know how I missed the highlighted question at the end of your opening post, sorry.

One thing you could do is open the REW EQ Filters window and add one [Control=Manual, Type=HP, Frequency=1500] filter line. To see what this does, clear the other lines first with the red cross at the top. In the main EQ window this should result in a first correction towards your target curve. From there you can match the rest of the response to your target the way you did initially. You'll see that it will leave the Control=Manual line in tact. Then export from REW and import into EQ APO. The LF section should now be properly attenuated.

Hope this helps!
 

CristianoLO

Member
Joined
Aug 18, 2022
Messages
28
Likes
31
please review your procedures
you need windowing the response before applying a high pass filter and some eq adjustment for the driver before applying room eq
 
  • Like
Reactions: JRS

CristianoLO

Member
Joined
Aug 18, 2022
Messages
28
Likes
31
and try to use a softer slope, because something as aggressive as an 8th order increases the group delay a lot, which can blur the sound in this region
 

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
Please show us the measured frequency response of the driver in REW. I don't think you've measured the driver yet.
 

ROOSKIE

Major Contributor
Joined
Feb 27, 2020
Messages
1,915
Likes
3,394
Location
Minneapolis
Please show us the measured frequency response of the driver in REW. I don't think you've measured the driver yet.
If you mean the driver outside of the enclosure there is no point. The final enclosure is already fixed and so design work should be with the drivers installed in said enclosure.
 

ROOSKIE

Major Contributor
Joined
Feb 27, 2020
Messages
1,915
Likes
3,394
Location
Minneapolis
Op, I suggest working on the baffle step for the woofer 1st.
Get your baseline for the speakers sensitivity and the general slope of the woofer then build your tweeters slope off of that.
Make sure you time align the drivers based on voice coil position and get the best response match you can by just using the baffel step and crossover slopes, ignoring eq-ing anything else for now.
Take some on and off axis measurements and start to tweak the crossover and maybe eventually adding in some specific PEQ for response issues that are present both on and off axis.
8th order may or may not give you a good blend in the mids. I'd experiment with shallower slopes as well.
 

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
If you mean the driver outside of the enclosure there is no point. The final enclosure is already fixed and so design work should be with the drivers installed in said enclosure.
No, I mean the drivers response in the enclosure. I haven't seen that yet. The OP has posted an EQ screen and I'm getting the impression that he hasn't done a complete measurement yet ... maybe doesn't know how to use REW. Let's let him respond shall we?
 
OP
P

paulgyro

Active Member
Joined
Nov 21, 2020
Messages
144
Likes
100
How are you faring @paulgyro?
Going ok now that I have some Windows 11 PC issues worked out. I've implemented the basic 2nd high pass / low pass filters in EQ APO using the PEACE UI which makes it much easier (though PEACE itself isn't the easiest to use). I'll start testing soon!
 
OP
P

paulgyro

Active Member
Joined
Nov 21, 2020
Messages
144
Likes
100
Op, I suggest working on the baffle step for the woofer 1st.
Get your baseline for the speakers sensitivity and the general slope of the woofer then build your tweeters slope off of that.
Make sure you time align the drivers based on voice coil position and get the best response match you can by just using the baffel step and crossover slopes, ignoring eq-ing anything else for now.
Take some on and off axis measurements and start to tweak the crossover and maybe eventually adding in some specific PEQ for response issues that are present both on and off axis.
8th order may or may not give you a good blend in the mids. I'd experiment with shallower slopes as well.
Thanks for the guidance, will do!
 
OP
P

paulgyro

Active Member
Joined
Nov 21, 2020
Messages
144
Likes
100
No, I mean the drivers response in the enclosure. I haven't seen that yet. The OP has posted an EQ screen and I'm getting the impression that he hasn't done a complete measurement yet ... maybe doesn't know how to use REW. Let's let him respond shall we?
The response of the tweeting in the enclosure is in the original post ( I incorrectly called it the "RAW Driver"). Here are the non-gated measurements with distortion. Gated coming soon.
 

Attachments

  • woofer FR with distortion.jpg
    woofer FR with distortion.jpg
    346.6 KB · Views: 55
  • tweeter freq response with distortion.jpg
    tweeter freq response with distortion.jpg
    343.2 KB · Views: 49

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
At what distance and how are you taking the measurements? There's too much hash in the response to come up with any kind of EQ/crossover solution .. at least in my mind.

When I measure drivers for crossover work it's done @around .5meter distance. I measure both drivers that way .. in this case woofer and tweeter at exactly the same .5meter distance. Shouldn't need to use a gate on the measurement or any smoothing. You match gain for each driver here as well. The response of the driver itself is independent of the room this way. If you need to EQ the driver response for the xo work, you can see what is needed easily this way. You overlay both the woofer and tweeter responses in REW and work out the xo frequency from that overlay. Use the overlay to check your xo work.

Once that is done, you can move the mic to a meter distance and check driver time alignment. If measurements at the listening position need EQ for the room .. go for it but remember this set of EQ solutions are for the room interaction .. those you did for the drivers are different. You shouldn't need to change those at all .. so don't.
 

ROOSKIE

Major Contributor
Joined
Feb 27, 2020
Messages
1,915
Likes
3,394
Location
Minneapolis
At what distance and how are you taking the measurements? There's too much hash in the response to come up with any kind of EQ/crossover solution .. at least in my mind.

When I measure drivers for crossover work it's done @around .5meter distance. I measure both drivers that way .. in this case woofer and tweeter at exactly the same .5meter distance. Shouldn't need to use a gate on the measurement or any smoothing. You match gain for each driver here as well. The response of the driver itself is independent of the room this way. If you need to EQ the driver response for the xo work, you can see what is needed easily this way. You overlay both the woofer and tweeter responses in REW and work out the xo frequency from that overlay. Use the overlay to check your xo work.

Once that is done, you can move the mic to a meter distance and check driver time alignment. If measurements at the listening position need EQ for the room .. go for it but remember this set of EQ solutions are for the room interaction .. those you did for the drivers are different. You shouldn't need to change those at all .. so don't.
I really think you have build in your baffel step for the woofer 1st or the whole thing is going to be way off.

Respectfully you should be gating (I look at both gated and ungated measurements here) or you will likely not have matching SPL levels as even @.5m the room reflections are summing into the responses in different amounts - especially in the bass and lower mids. This will leave the design top heavy /bass light.
Plus even with gating there is still plenty of resolution for the typical basic Xover region.
When designing active in room I start @18"@ look at gated measurements (& also smoothed ungated measurements) and get a fairly accurate sense of what I have building the basics. OP you choose the design axis now, either directly on tweeter or exactly halfway between the woofer and tweete are 2 good ideas. I set the time alignment on my driver's based on physical measurements though I suspect Klipsch already set them well with the deep waveguide so in this speaker it may be done, then I go setting my baffel step and then basic Xover ideas. I usually build in 4-6db of BSC here but some use less(1-3db) if the speakers will be close to the walls. Mine end up out on stands away from walls so the full amount or nearly full amount is best for me. (6db is 'full' BSC). Also make sure you kill that strong looking woofer resonce in the treble so that it is down enough not sum with the tweeter.
Then I move 1meter or more away (2m would be better when designing for Fairfield, if you can use a stand outside or in a massive room with high ceilings. For near, 1m is great) and start to fine tune.
Make sure your mic at 1meter or whatever distance it is, is closer to the drivers than any boundary including the floor.
When fine tuning all my adjustment measures are gated and check your off axis responces. You can add PEQ here to help correct some of the obvious speaker and driver issues you discover but be very conservative with it as the accurate resolution just isn't there like with the Klippel scanner.
I also cross test the results by placing the speaker in the room in situ and from the listening position/listening window I do moving mic steady state energy captures and then compare them with previous MM's from known speakers with similar dispersion. This helps me check the overall energy level balance against a known good example, but works best when they have similar dispersion and the rp160m is very narrow so keep that in mind.
Once place in room in situ you can try room correction.
I also design a high pass for the woofer based on frquency measurements, HD testing, impedance sweep and modeling the woofer in simulation and determining the best way to get maximum output while limiting excursions on deep bass.
I design for Fairfield by the way and as well this strategy would be appropriate for near field.
Others will have ever better approaches and I could add more details s/complexity but the above plan is a good basic minimal start.
Eventually if you make more you can start looking at software that can help to make Spins and design the Xover and a rotating stand that allows you to take many off axis responses easily.
 
Last edited:

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
I don't worry about BSC in an active application. I allow for it once the loudspeaker is completed and in room measurements indicate a need for it. A variable low shelf filter takes care of that nicely.
 
Top Bottom