B&C DCX464

Member
Joined 2007
Paid Member
These results are very interesting the harmonic distortion at 130 dB seems similar to the JBL cone midrange horn loaded drivers in this tech note:
https://adn.harmanpro.com/site_elements/executables/19573_1571427998/CMCD_TechNote_original.pdf
Absolute SPL capability is of course lower, but as your results show the harmonic distortion is getting a bit much at 130 dB anyway.

Just realised that I misread the graphs. The CMCD harmonic amplitudes have +20dB added! so they are much lower distortion.
 

Attachments

  • CMCD.png
    CMCD.png
    185.8 KB · Views: 667
Yeah, 10% is kinda an industry standard huh?....
Mark,

As one definition of reaching Xmax, yes.

As for high frequency drivers, where excursion is limited to under 1mm before contact with the phase plug, there seems to be no agreed upon standard.
It is rare to see distortion results above 5 watts (nominal impedance/voltage, not actual watts) as current HF compression drivers always exceed 10% THD well below rated power in the top octaves.

Art
 
Mark,

As one definition of reaching Xmax, yes.

As for high frequency drivers, where excursion is limited to under 1mm before contact with the phase plug, there seems to be no agreed upon standard.
It is rare to see distortion results above 5 watts (nominal impedance/voltage, not actual watts) as current HF compression drivers always exceed 10% THD well below rated power in the top octaves.

Art

Hi Art, i see your point. Food for thought......

Hey, semi related question for you....i keep perceiving a spectral tilt as i turn SPL up, keeping all else equal, that just sounds brighter.
It's with any of my various speaker builds, which all use either the bms or dcx CD's.
Do you think this perceived tilt could be due to a sensitivity to increasing THD, or should I be looking elsewhere? like maybe a psyche ward :D
 
Member
Joined 2007
Paid Member
This is normal and due to how sound is perceived at different volumes, You should change the EQ to suit. There is an excellent document written by Rog Mogale that I have attached that tells you what to do.

*quick question for the more experienced; is the harmonic distortion performance we are seeing in the upper range of these comprehension drivers the best that can be achieved by single drivers?
 

Attachments

  • SetupTipsVoid.pdf
    429.4 KB · Views: 208
This is normal and due to how sound is perceived at different volumes, You should change the EQ to suit. There is an excellent document written by Rog Mogale that I have attached that tells you what to do.

*quick question for the more experienced; is the harmonic distortion performance we are seeing in the upper range of these comprehension drivers the best that can be achieved by single drivers?

Very good paper....very good recommendations ime/imo. Thx :)

My question though, is the perception due to equal loudness contours, or increased high frequency content due to rising THD levels?

Because it seems to me, the loudness contours level out at higher SPLs, giving us relatively less sensitivity to high freq content ????

Unless i'm reading the loudness contours' effect backwards, i'm left thinking 'where is the hf perception coming from'? THD is all i can come up with...
 
Hey, semi related question for you....i keep perceiving a spectral tilt as i turn SPL up, keeping all else equal, that just sounds brighter.

Do you think this perceived tilt could be due to a sensitivity to increasing THD, or should I be looking elsewhere? like maybe a psyche ward :D
Mark,

Since all the "good" HF compression drivers seem to have increasing THD in the upper range, and distortion sounds "louder", sounds like it could be at least a partial explanation.

You could record the HF horns playing the same tune at widely different levels, then play back at the same level to confirm if the distortion sounds "brighter".

Or you could listen to the music recordings I did at different levels in High Frequency Compression Driver Evaluation
and be the first to report any difference you heard between the recordings!

Art
 
Mark,

Since all the "good" HF compression drivers seem to have increasing THD in the upper range, and distortion sounds "louder", sounds like it could be at least a partial explanation.

You could record the HF horns playing the same tune at widely different levels, then play back at the same level to confirm if the distortion sounds "brighter".

Or you could listen to the music recordings I did at different levels in High Frequency Compression Driver Evaluation
and be the first to report any difference you heard between the recordings!

Art

Wow Art, what a test . Have not seen that before...
Will study and try to digest....thx!
 
Hi Mark,

I suspect the best thing to do to get some idea of how these perform in the real world is to use a signal that approximates music such as IEC 268 – 5 which starts to roll off at 2kHz and is about 20 dB down at 20kHz.

The problem is that you really need to use a broad band sign of this shape and with the power being shared between the MF and HF section in accordance with the selected crossover points ... but that’s no good for measuring distortion levels.

The 464 has a large overlap between the MF & HF sections. This will allow the use of crossovers with slopes typically from 12 dB - 24 dB per octave. What I don’t know is what happens through the crossover region where some of the same signal is shared with both drivers, it may reduce the distortion through this region … don’t know. In the case of the 464 this will occur between 2-4kHz – ish and if there was a reduction and given the shape of IEC 268 / music this could be significant with respect to how we perceive these two drivers.

In your case you have used brick wall filters so I’m not sure how the results compare. FWIW I found I preferred using 24dB per oct. filters over brick wall filters when playing with the 464. With the 4594 because there is more or less no over lap at the crossover point of 6K3 there is an advantage with the brick wall approach.

The other thing to note is how we perceive distortion. We tend not to notice second harmonic distortion because its masked by the fundamental, however we are much more sensitive to third harmonic distortion, so its contribution to THD is much more important.
 
Hi Peter,

That all makes sense, thx.
I had already started playing with lower sloped crossovers, given the dcx464's wider overlap. And may end up closer to 4kHz, it seems.

I know second harmonic is supposedly less offensive, and i see how it dominates the two CD's tests i made.
Yes, it would be nice to be able to make some distortion tests made off a signal more closely aligned with music than swept sine.

The THD measurements were more about curiosity than anything else, although i truly would like to know if THD can be heard.
Listening though Art's extensive test files now.

I may be just about done traveling this THD trail....
 
Hi Peter,

That all makes sense, thx.
I had already started playing with lower sloped crossovers, given the dcx464's wider overlap. And may end up closer to 4kHz, it seems.

I know second harmonic is supposedly less offensive, and i see how it dominates the two CD's tests i made.
Yes, it would be nice to be able to make some distortion tests made off a signal more closely aligned with music than swept sine.

The THD measurements were more about curiosity than anything else, although i truly would like to know if THD can be heard.
Listening though Art's extensive test files now.

I may be just about done traveling this THD trail....

There is a masking effect in our hearing; they take advantage of it in data compression for MP3 etc. The tones that are masked don't need to reproduced and can be discarded - less data to record ... in terms of distortion it matters how close it is to the main signal and at what frequency the distortion occurs. i.e. second harmonics are more easily masked than others.
 

Attachments

  • masked frequencies.png
    masked frequencies.png
    23.8 KB · Views: 826
Last edited:
Member
Joined 2007
Paid Member
Industry standard for scattering birds from airport

Overview

The 4599ND has two most outstanding features:

It is the loudest audio transducer ever made
The frequency range is optimized for human voice

These capabilities make it best suited for:

Communication devices for long distance
Security systems
Emergency devices
Mass notification systems
Systems for scattering birds from airports
Marine applications
Military and police communication system
Noise cancelation systems
High End Audio Loudspeakers

If you compare it to the BMS4592ND-mid-8 on the same horn it has more output, ~-6dB harmonic distortion and a smoother frequency response. I guess the only disadvantage if you wanted a dedicated mid-range driver is the cost!

http://www.bmsspeakers.com/fileadmin/bms-data/product_data_2012/bms_4592-mid_t_data_2012-01.pdf
http://www.bmsspeakers.com/fileadmin/bms-data/product_data_2014/bms_4599nd_preliminary.pdf

I wonder how these drivers (coaxial and dedicated midrange 2" throat) compare to the midrange specific drivers from community that had a 4" throat?
 
The 4599ND has two most outstanding features:

It is the loudest audio transducer ever made
The frequency range is optimized for human voice

These capabilities make it best suited for:

Communication devices for long distance
Security systems
Emergency devices
Mass notification systems
Systems for scattering birds from airports
Marine applications
Military and police communication system
Noise cancelation systems
High End Audio Loudspeakers

If you compare it to the BMS4592ND-mid-8 on the same horn it has more output, ~-6dB harmonic distortion and a smoother frequency response. I guess the only disadvantage if you wanted a dedicated mid-range driver is the cost!

http://www.bmsspeakers.com/fileadmin/bms-data/product_data_2012/bms_4592-mid_t_data_2012-01.pdf
http://www.bmsspeakers.com/fileadmin/bms-data/product_data_2014/bms_4599nd_preliminary.pdf

I wonder how these drivers (coaxial and dedicated midrange 2" throat) compare to the midrange specific drivers from community that had a 4" throat?

The problem with 4599 is finding a VHF driver that will keep up - I was looking to build a super verion of my PM90/60 with the 4599 but I could not find a suitable VHF driver that could match its output.
 
Pano's explanation of the tests in post #23 of the High Frequency Compression Driver Evaluation thread is helpful for "digestion" ;^).



Hi Art, yep, Pano did (always does) do a good job of explaining.
I understood the test quick enough...although it did take a little to figure out the filename code structure...but like you said, it was all there on the first page.

If I've heard any difference on the test tracks i've listened to, it's not any I'd be willing to bet i could hear in a double blind. Even pulled out the Stax phones to give a best listen...

So far, it's just been the music tracks for the two bms and the EV.
I went straight for the uh-oh 1Ap1763, figuring if that one didn't sound strained in comparison, none would ;)

Very nice work :)
 
If I've heard any difference on the test tracks i've listened to, it's not any I'd be willing to bet i could hear in a double blind. Even pulled out the Stax phones to give a best listen...
Thanks for listening and commenting!

While conducting the HF driver tests, I thought that the distortion at high levels sounded quite horrible, though in retrospect, seems most of the discomfort was in my own hearing response- there is a fairly hard SPL limit in the upper range that sounds bad whether clean or distorted when my ears are “fresh”.

The “too bright” sensation at louder levels which seems contrary to the equal loudness contours that flatten out with increased SPL may be due to the manner in which hearing TTS (temporary threshold shift) occurs.
TTS is the reduced sensitivity to noise after exposure to loud noise, it’s temporary effect diminishes after time in a quiet environment.
Evidence of TTS is apparent when you get into your car the morning after coming home from a loud concert the night before, then must turn the stereo down to half the level that seemed “normal” the night before.

William Melnick has presented some evidence that indicates TTS is less when exposed to loud low frequency <125 Hz than higher frequencies.
With loud music played back with a low frequency “haystack” of +10 dB or more below 125Hz, the TTS threshold won’t “kick in” as hard, fast or with as deep of a reduction as it would from the same SPL at high frequencies. As SPL increases, TTS reduces sensitivity to the HF, but at some point your threshold of pain (or annoyance) is reached, and you will desire protection from the offensive “too bright” high frequencies.

Looking at Melnick’s charts, we see as much as 20dB shifts from 95 dB SPL of 1.4-2kHz noise, while increasing the level to 108 dB with 500-710Hz noise reduces TTS to about 10dB. The range at which TTS between subjects is also quite a lot, “too loud” is very subjective.

The decay time from TTS can be as much as 24 hours, it can be quite a while before your hearing normalizes after exposure to loud music or noise, what sounded good (or bad..) yesterday might be quite different after a break, and at a different level.

Art
 

Attachments

  • Growth and decay of TTS.png
    Growth and decay of TTS.png
    92 KB · Views: 674
  • W Melnick,TTS & Noise.png
    W Melnick,TTS & Noise.png
    653.9 KB · Views: 658