JBL DMS-1 Experience and Measurement

[Dave M]

I finally finished installing JBL DMS-1 monitor. It is too bad the digital technology was very primitive when JBL released DMS-1. I used mastering grade FIR filter, DMG Equilibrium, which is considered to be one of the best sounding digital EQ on the market today, and this monitor sounds absolutely stunning with it. The software is on my music computer, so no extra ADA conversion.

It is as accurate as the best pro monitors or high end speakers today, while it still retains THAT classic JBL horn monitor tone. I compared it with my trustful Barefoot MM27 monitor. With DMS-1, JAZZ, Classic, Rock, Dance music, all sound excellent. I agree with Floyde Toole, it has a tendency to play it extremely loud, because it doesn't sound loud at all even at 110dB. The compression driver (2450SL-A) and horn (2332?) can be changed to D2 and more contemporary waveguide without modification, so I may try it in the future, but I’m already more than happy with current configuration.


The attached screen shot is from REW measurament (0 degree, 1 meter). I tested a lot of different crossover methods, and I found Keele Linear Phase method (Linkwitz Lab) at 600Hz sounds the best, compared to any other non linear phase filter method we tested. We all can clearly hear the annoying phase shift with analog style non linear crossover in blind tests performed here. It was almost too easy to detect. Slight pre ringing is visible in the step response, of course, but I honestly can't hear it. One person says he does, but he also prefers linear phase. Measured response is unusually clean both vertical and horizontal for this size monitor. The amp is Aleph J (DIYAudio). I always think Aleph is the best amp for JBL monitors.

Frequency+ Phase, Step, Distortion 2nd, 3rd..., Waterfall.

[ivica]

I finally finished installing JBL DMS-1 monitor. It is too bad the digital technology was very primitive when JBL released DMS-1. I used mastering grade FIR filter, DMG Equilibrium, which is considered to be one of the best sounding digital EQ on the market today, and this monitor sounds absolutely stunning with it. The software is on my music computer, so no extra ADA conversion.

It is as accurate as the best pro monitors or high end speakers today, while it still retains THAT classic JBL horn monitor tone. I compared it with my trustful Barefoot MM27 monitor. With DMS-1, JAZZ, Classic, Rock, Dance music, all sound excellent. I agree with Floyde Toole, it has a tendency to play it extremely loud, because it doesn't sound loud at all even at 110dB. The compression driver (2450SL-A) and horn (2332?) can be changed to D2 and more contemporary waveguide without modification, so I may try it in the future, but I’m already more than happy with current configuration.


The attached screen shot is from REW measurament (0 degree, 1 meter). I tested a lot of different crossover methods, and I found Keele Linear Phase method (Linkwitz Lab) at 600Hz sounds the best, compared to any other non linear phase filter method we tested. We all can clearly hear the annoying phase shift with analog style non linear crossover in blind tests performed here. It was almost too easy to detect. Slight pre ringing is visible in the step response, of course, but I honestly can't hear it. One person says he does, but he also prefers linear phase. Measured response is unusually clean both vertical and horizontal for this size monitor. The amp is Aleph J (DIYAudio). I always think Aleph is the best amp for JBL monitors.

Frequency+ Phase, Step, Distortion 2nd, 3rd..., Waterfall.

HI Dave,

Very interesting results, especially for the 2nd, 3rd distortions on 115dB spl.
Showing abou -50bB level, not to mention necessary CD horn compensation, especially over 10kHz.

Regards ivica

[Dave M]

HI Dave,

Very interesting results, especially for the 2nd, 3rd distortions on 115dB spl.
Showing abou -50bB level, not to mention necessary CD horn compensation, especially over 10kHz.

Regards ivica

I'm sorry, but the 115 is a wrong value. I was too lazy to set REW.
The data was actually measured at around 100-105 dB.

[Dave M]

Impulse Response for DMS-1 (The same measurement data used as before), it looks as it should be.
Until recently, I didn't know how to show it this way with REW .

I'll upload UREI 813C data in the future for comparison.

[Dave M]

I don't know if anyone is interested in this, but for reference, I just update my current setting.

1. Crossover is 810Hz with a lot of room EQ correction point. JBL official setup is LR4@1K.
2. I find Classic D'appolito BW3 (3rd order butterworth) sounds overall better than linear phase. Linear phase has its own advantage, but more I use it, more I'm annoyed by pre-ringing and some other things that I initially ignored.
3. I'm still using DMG Equilibrium, FIR 260,000 steps impulse response.
4. -2dB dip around 2-4KHz @1m sounds more balanced in my studio.
5. I went back and forth between 2332 and STX-825 waveguide since 2017, but I decided to stick to 2332, and STX was sold.

http://www.audioheritage.org/vbullet...l=1#post420049

I still have Urei 813C, and their 2215 woofers were replaced with vintage alnico woofers. 813C has been sitting in our living room for casual listening. If requested, I can upload casual in room measurement of 813C @1m.

[Ian Mackenzie]

Great thread 👌

[dubkarma]

I don't know if anyone is interested in this, but for reference, I just update my current setting.

1. Crossover is 810Hz with a lot of room EQ correction point. JBL official setup is LR4@1K.
2. I find Classic D'appolito BW3 (3rd order butterworth) sounds overall better than linear phase. Linear phase has its own advantage, but more I use it, more I'm annoyed by pre-ringing and some other things that I initially ignored.
3. I'm still using DMG Equilibrium, FIR 260,000 steps impulse response.
4. -2dB dip around 2-4KHz @1m sounds more balanced in my studio.

I've (finally!) assembled components for the DMS-1 (2450SL, 2332, 1400ND reconed from 1400Pro), and would appreciate learning more about the "Classic D'Appolito BW3" Butterworth crossover.

I'm restricted to using the dbx DriveRack 4820 speaker processor, and so can't do linear phase type of crossovers, but would like to try implementing the crossover arrangement you've mentioned.

Any details appreciated! Thanks, Joel.

[Ian Mackenzie]

Hi Joel,

It looks like you are having fun.

Are you using REW impulse response to locate the driver acoustic centres and then applying time correction?

With the measurements one thing l would try is some vertical off axis at 5, 10, 15 degrees to confirm the main lobe is correctly focused towards the listening position.

[dubkarma]

Hi Joel, Are you using REW impulse response to locate the driver acoustic centres and then applying time correction?
With the measurements one thing l would try is some vertical off axis at 5, 10, 15 degrees to confirm the main lobe is correctly focused towards the listening position.

Hi Ian, thanks for your comments.

If by "time correction," you mean delay, there isn't any applied, according to JBL's "Settings" document for the DMS-1, concerning which see the earlier discussion on this forum:

http://www.audioheritage.org/vbullet...ings-for-DMS-1

My plan is to first of all reproduce as exactly as I can what JBL did. After that, listen, measure, and decide further. . .

1. Components: 4 x 1400Nd woofers, 2 x 2450SL-A drivers, 2 x 2332 horns — I'm using the 2450SL hoping the difference is minimal

2. Cabinet dimensions and tuning — I've only got external measurements, so if anyone can supply internal and port dimensions, much appreciated!

3. Crossover according to the "DSC 280 Recommended Array Settings DMS-1" document:

LF: LR12 high pass @ 18 Hz, LR24 low pass @ 1 kHz; HF: LR24 high pass @ 1 kHz, low pass, out.
Gain Settings: LF 0 dB, HF –13 dB

I'll be using a dbx DriveRack 4820, arguably a much better processor than the DSC 280, but of course the settings won't necessarily "translate" perfectly. But, again, that's the starting point.


4. EQ: According to same document, only four EQ settings:
4.1. +6 dB @ 183 Hz with a "bandwidth in octaves" of "12 Lo" — not sure what "12 Lo" means here
4.2. –1.5 dB @ 378 Hz with a bandwidth in octaves of 0.25
4.3. +10 dB @ 8.28 kHz with a bandwidth in octaves of "12 Hi" — as with "12 Lo," not clear on this specification
4.4. –1.5 dB @ 3.24 kHz with a bandwidth in octaves of 0.25

[Dave M]

Hi Ian, thanks for your comments.

If by "time correction," you mean delay, there isn't any applied, according to JBL's "Settings" document for the DMS-1, concerning which see the earlier discussion on this forum:

http://www.audioheritage.org/vbullet...ings-for-DMS-1

My plan is to first of all reproduce as exactly as I can what JBL did. After that, listen, measure, and decide further. . .

1. Components: 4 x 1400Nd woofers, 2 x 2450SL-A drivers, 2 x 2332 horns — I'm using the 2450SL hoping the difference is minimal

2. Cabinet dimensions and tuning — I've only got external measurements, so if anyone can supply internal and port dimensions, much appreciated!

3. Crossover according to the "DSC 280 Recommended Array Settings DMS-1" document:

LF: LR12 high pass @ 18 Hz, LR24 low pass @ 1 kHz; HF: LR24 high pass @ 1 kHz, low pass, out.
Gain Settings: LF 0 dB, HF –13 dB

I'll be using a dbx DriveRack 4820, arguably a much better processor than the DSC 280, but of course the settings won't necessarily "translate" perfectly. But, again, that's the starting point.


4. EQ: According to same document, only four EQ settings:
4.1. +6 dB @ 183 Hz with a "bandwidth in octaves" of "12 Lo" — not sure what "12 Lo" means here
4.2. –1.5 dB @ 378 Hz with a bandwidth in octaves of 0.25
4.3. +10 dB @ 8.28 kHz with a bandwidth in octaves of "12 Hi" — as with "12 Lo," not clear on this specification
4.4. –1.5 dB @ 3.24 kHz with a bandwidth in octaves of 0.25

I'm sorry I didn't see your post. I visit this forum not very often...

Port dimension is, 3" depth x 3" diameter. The port has a large flare at the opening.
I also wondered what was 12 LO. I guessed 12 LO / 12 High means low/high shelf boost with Q 0.7 (12dB/par octave?).
"bandwidth in octaves of 0.25" probably means Q=12 or so.

I have less top boost, and slightly more bass boost, and lower cut off frequency than JBL's suggestion.

[Dave M]

Answering private message from dubkarma.

MDF panel is 1" thickness, bracing (2"x4") . Yellow glass fiber (1"), attached to top, bottom, side, Blue and Yellow bracing.
Sorry for poor image, it's not very accurate, so use it as a reference.

Red. 4 side panel bracing from front to back.
Green 2 off center bracing from front to back.
Yellow 1off center bracing from bottom to top, attached to Green
Blue 1 bracing from left to right, attached to Yellow. This bracing is also holding the back of the driver unit via 1/3" thick small white packaging foam.

[dubkarma]

Answering private message from dubkarma.

MDF panel is 1" thickness, bracing (2"x4") . Yellow glass fiber (1"), attached to top, bottom, side, Blue and Yellow bracing.
Sorry for poor image, it's not very accurate, so use it as a reference.

Red. 4 side panel bracing from front to back.
Green 2 off center bracing from front to back.
Yellow 1off center bracing from bottom to top, attached to Green
Blue 1 bracing from left to right, attached to Yellow. This bracing is also holding the back of the driver unit via 1/3" thick small white packaging foam.

Much appreciated! Many thanks!

[Dave M]

Good luck with your project. I hope you'll like this monitor as I do.