Has anyone tried using the DCX2496 with the LeCleach Butter18 approach?
I was going to just use the stock LR48 and auto-align functions and then use an RTA to do basic EQ. Now I'm not sure this is the best way.
I have read the data and the LeCleach spreadsheet but I am not sure how to proceed using a two-way system (dual 15-inch crossed over to a 1-inch B&C compression driver in a 10-inch waveguide).
TIA
I was going to just use the stock LR48 and auto-align functions and then use an RTA to do basic EQ. Now I'm not sure this is the best way.
I have read the data and the LeCleach spreadsheet but I am not sure how to proceed using a two-way system (dual 15-inch crossed over to a 1-inch B&C compression driver in a 10-inch waveguide).
TIA
OK, I will throw out my first guess in hoping that it will be easier for someone to correct me rather than guide me.
Setup:
2 Alpha 15As in parallel on the bottom, OB design
1 B&C DE250 compression driver in 10-inch DDS waveguide
Assuming a 1.2K xover target, I would do the following with the DCX2496:
1. Set both high and low pass filters as third order butterworth.
2. Invert polarity on the high pass compression driver.
3. Using a target 1.2K crossover, I move the low pass point to 1,044 (x 0.87) and change the high pass point to 1,368 (x 1.14).
4. Low pass driver delay set to 0.957 ms (1044 Hz = 33 cm, 0.957 ms delay)
5. High pass driver delay set to 0.73 ms (1368 Hz = 25 cm, 0.73 ms delay)
6. Set all driver phase to 0 degrees.
Is the above process the correct implementation for the LeCleach approach? Am I mission something else? Do I need to run auto-align and use those corrections as a starting point PRIOR to steps 4 and 5?
TIA
Setup:
2 Alpha 15As in parallel on the bottom, OB design
1 B&C DE250 compression driver in 10-inch DDS waveguide
Assuming a 1.2K xover target, I would do the following with the DCX2496:
1. Set both high and low pass filters as third order butterworth.
2. Invert polarity on the high pass compression driver.
3. Using a target 1.2K crossover, I move the low pass point to 1,044 (x 0.87) and change the high pass point to 1,368 (x 1.14).
4. Low pass driver delay set to 0.957 ms (1044 Hz = 33 cm, 0.957 ms delay)
5. High pass driver delay set to 0.73 ms (1368 Hz = 25 cm, 0.73 ms delay)
6. Set all driver phase to 0 degrees.
Is the above process the correct implementation for the LeCleach approach? Am I mission something else? Do I need to run auto-align and use those corrections as a starting point PRIOR to steps 4 and 5?
TIA
Horizons said:
I have a similar setup but my bass driver is 12" not 15". This combo works quite well, I have a measuring device(ie microphone and software) and I don't have to reverse polarity of the compression driver. I cross around 1.7Khz. I don.t use any delay the sound seems to be good enough. Owing to quality of my 12" driver the midrange sound is not as good as if you add another huigh quality midrange.
cheers.
Probably! 
But it should depend on Q and how close the EQ is the x-over point.
Be sure to save a normal and corrected setting so that you can compare. Or use the compare button. You should hear it. Imaging should be better with the corrected version.
Your dual woofer setup might muddle things a bit, I don't know.
But it should depend on Q and how close the EQ is the x-over point.Be sure to save a normal and corrected setting so that you can compare. Or use the compare button. You should hear it. Imaging should be better with the corrected version.
Your dual woofer setup might muddle things a bit, I don't know.
Horizons said:
More info, I use 3rd order Butterworth filter 1.7Khz, not reversal of polarity required.I suggest you view the FR curve on your s/w before you apply reversal. BW filter sounds better?
The horn on my DE250 is RCF horn H100 I got from Loudspeaker Plus very cheap less than $30, no much equalization required, you can get away with no equalization. I am happy with it much better than PE 10" spherical horn.
Hello Horizons,
Thank you to have given some informations about the Le Cléac'h crossover setting.
There is no need to delay both the LF and HF channel in a 2 ways loudspeaker. As you know a Low-Pass filter induce more group delay than a High-Pass filter. To perform the alignement, only the HF channel should be delayed (if acoustic centers initially vertically aligned).
For the Le Cléac'h crossover you have 2 solutions to perform the aligment in order to obtain a quasioptimal pulse response.
1) starting from a geometrical alignment of the bass and mid-high frequency loudspeaker, the mid-high frequency being kept in its position, move the bass loudspeaker in direction of the listener position of a distance equal to 0.22 times the wavelength at the common attenuation frequency between the low-pass filter and the high-pass filter. In your example, for a common attenuation frequency of 1,2k (this common attenuation corresponds to the frequency for which both the LPF and HPF give -5dB) you have to move forward the bass loudspeaker on 63millimeters in direction of the listener position. Don't use any delays on the diigtal crossover.
2) starting from a geometrical alignment of the bass and mid-high frequency loudspeakers you can apply a delay of 0.183millisecond on the high pass filtered channel, this is equivalent to a distance of 63millimeters travelled at the speed of sound.
Then if you start from an initial chosen configuration of the loudspeakers you'll have better to perform the alignement using a method based on pulse measurements (by example using ARTA or MLS or the Aurora plugins GENSWEEP and CONVOLVE put in Adobe audition.
Please note that many people fear the electrical polarity inversion of the tweeter (or mid-high) versus the bass loudspeaker. They think that when the membrane of the bass loudspeaker will move forward, the diaphragm of the tweeter will move backward. This is not the case in a correctly settled Le CLéac'h crossover a goal of which is to have both loudspeakers quasi operating in phase at all frequency from 0Hz to 4kHz (with a resulting maximal phase difference lesser than 20°).
This is the same as with the Linkwitz-Riley of the second order which requires the polarity of the tweeter to be inverted to obtain the bass loudspeaker and the tweeter operating in phase at all frequency...
Also you have to know that my spreadsheet was written long ago before the DCX2496 was introduced so it is not exactly the same logic and this can be the source of improper settings (specially many people have problem with the delays by example when a compression driver is 20 to 30 centimeters backward of the bass loudspeaker... ).
Also they can be some problem with the autoalignment... I prefer not to use it and to optimize the alignement using a pulse response measurement.
Finally, measurement methods only based on the frequency response curve (RTA and so on) are unable to help someone to perform the alignment of a Le Cleac'h crossover (or every crossover for which an excellent pulse response is expected).
Best regards from Paris, France
Jean-Michel Le CLéac'h
Thank you to have given some informations about the Le Cléac'h crossover setting.
There is no need to delay both the LF and HF channel in a 2 ways loudspeaker. As you know a Low-Pass filter induce more group delay than a High-Pass filter. To perform the alignement, only the HF channel should be delayed (if acoustic centers initially vertically aligned).
For the Le Cléac'h crossover you have 2 solutions to perform the aligment in order to obtain a quasioptimal pulse response.
1) starting from a geometrical alignment of the bass and mid-high frequency loudspeaker, the mid-high frequency being kept in its position, move the bass loudspeaker in direction of the listener position of a distance equal to 0.22 times the wavelength at the common attenuation frequency between the low-pass filter and the high-pass filter. In your example, for a common attenuation frequency of 1,2k (this common attenuation corresponds to the frequency for which both the LPF and HPF give -5dB) you have to move forward the bass loudspeaker on 63millimeters in direction of the listener position. Don't use any delays on the diigtal crossover.
2) starting from a geometrical alignment of the bass and mid-high frequency loudspeakers you can apply a delay of 0.183millisecond on the high pass filtered channel, this is equivalent to a distance of 63millimeters travelled at the speed of sound.
Then if you start from an initial chosen configuration of the loudspeakers you'll have better to perform the alignement using a method based on pulse measurements (by example using ARTA or MLS or the Aurora plugins GENSWEEP and CONVOLVE put in Adobe audition.
Please note that many people fear the electrical polarity inversion of the tweeter (or mid-high) versus the bass loudspeaker. They think that when the membrane of the bass loudspeaker will move forward, the diaphragm of the tweeter will move backward. This is not the case in a correctly settled Le CLéac'h crossover a goal of which is to have both loudspeakers quasi operating in phase at all frequency from 0Hz to 4kHz (with a resulting maximal phase difference lesser than 20°).
This is the same as with the Linkwitz-Riley of the second order which requires the polarity of the tweeter to be inverted to obtain the bass loudspeaker and the tweeter operating in phase at all frequency...
Also you have to know that my spreadsheet was written long ago before the DCX2496 was introduced so it is not exactly the same logic and this can be the source of improper settings (specially many people have problem with the delays by example when a compression driver is 20 to 30 centimeters backward of the bass loudspeaker... ).
Also they can be some problem with the autoalignment... I prefer not to use it and to optimize the alignement using a pulse response measurement.
Finally, measurement methods only based on the frequency response curve (RTA and so on) are unable to help someone to perform the alignment of a Le Cleac'h crossover (or every crossover for which an excellent pulse response is expected).
Best regards from Paris, France
Jean-Michel Le CLéac'h
Horizons said:
Jean-Michel:
Thank you very much for the detailed response.
Due to my physical baffle design, I am unable to move the drivers for alignment and must do this in the DCX2496.
I assume this requirement forces me to perform the alignment "using a method based on pulse measurements (by example using ARTA or MLS or the Aurora plugins GENSWEEP and CONVOLVE put in Adobe audition."
Are you saying that the DCX2496 auto-alignment feature is not accurate?
FYI, I really enjoyed your PPT presentation and find it has merit. I only wish the application was a bit easier. Your approach seems to virtually eliminate phase distortion.
http://freerider.dyndns.org/anlage/LeCleach.htm
Thank you very much for the detailed response.
Due to my physical baffle design, I am unable to move the drivers for alignment and must do this in the DCX2496.
I assume this requirement forces me to perform the alignment "using a method based on pulse measurements (by example using ARTA or MLS or the Aurora plugins GENSWEEP and CONVOLVE put in Adobe audition."
Are you saying that the DCX2496 auto-alignment feature is not accurate?
FYI, I really enjoyed your PPT presentation and find it has merit. I only wish the application was a bit easier. Your approach seems to virtually eliminate phase distortion.
http://freerider.dyndns.org/anlage/LeCleach.htm
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.