Interesting. So, my homemade D2 'protection network' would not be needed. So there is no risk in feeding the D2 such a signal right out of the amp coupled to an active crossover (in my case actually a DBX 360 signal processor)? My system is still not built (cabinets are the bottleneck) but I like the idea of removing complexity and 1 fairly powerful amp, Marantz 250.Originally Posted by pos
There would be some modifications to make to the settings, as the passive network also integrate a 1kHz high pass filter, and linearize the impedance.
The whole network could be replaced with a simple protection cap for something like a 200Hz high pass filter (under the impedance peak).
I plan on doing that with measurements when I receive the amp (of course we would have to agree on a common value for the cap if we want to share settings).
I already did something similar in the past but cannot find the post right now.
A 200Hz filter would bring less protection than a 1kHz one, so of course the driver would be more exposed to user error or crossover failure.
How are you planning on replicating the settings into the DBX?
I looked it up a while ago and cannot remember exactly, but I am not sure it has enough EQ points, or of the right type.
I followed that thread closely. John Schuermann and his crew worked hard to put it together, and they should be lauded for seriously working to do what many people flap their gums about but never attempt. Floyd Toole had some input into the process, too, but since I've done human subject testing for three decades and also had the opportunity to sit in the Harman MLL a few times, I had misgivings about some things.
Sean Olive states that to make this kind of speaker testing double-blind, the test subjects cannot not know what loudspeakers are being tested. Since the event was billed as a shootout between the S2 and M2, it's pretty clear that every subject knew what was being tested before the test ever started. Additionally, human subjects, especially untrained ones, tend to be susceptible to binary processing, i.e., one is good and one is bad, and even if both are good, one is better and one is worse. That tendency needs to be mitigated by introducing a random factor or factors to establish a hierarchical challenge. Once a third or fourth unknown speaker is inserted, the listener must actually work to discriminate degrees of performance rather than a simple binary analysis of two known speakers.
Also, it doesn't matter if a listener has previously heard both, one, or neither of the (S2 and M2) speakers. Once they are announced as the objects of the study, he or she will already have preconceived ideas about each speaker based on others' opinions, available data (like spinoramas), subjective reviews, and even speaker design. One is a two-way horn, one is a Ti-Be multi-driver, and thus the binary compartmentalization begins.
The spread of points or ratings was quite narrow in the shootout, which means there was insufficient diversity to establish a useful hierarchy. Having that third or fourth unknown speaker (as well as not knowing the S2 and M2 were used, too) in the shootout would have increased the spread top to bottom, and more than likely would have narrowed the S2 and M2 scoring. Also, because of the greater complexity of listening to three or four speakers, the lower ranking models would have acted as a baseline or reference point for the higher ranking ones, and rationally could have flipped some of the preferences when the binary processing was removed from the exercise.
Another difficulty is the use of well-known, oft-used demo tracks. I know people love to take their favorite music or music recommended by an expert with them to audition speakers. The lists tend to be duplicative, as all of us share "here's a great track for..." with our equally enthusiastic friends. It's part of the ritual. However, it defeats or at least diminishes the listener's ability to discern differences in the speakers' performance, because the focus is on the music they know and how it is being rendered, often filtered by familiarity with the tracks in their current environment and filtered by emotional attachment to experiences related to the music. There's also the intense desire to "hear more" in the recording than they've heard before, and when they do (or imagine they do) it becomes an epic moment. However, hearing more is not the same as hearing better. These kinds of subtleties in untrained human subjects testing are critical, as they can render the accumulated data virtually meaningless.
There are other nits to pick, but unfortunately they all add up to making an already complicated undertaking nearly impossible to pull off. Again, props to John et al for the yeoman's effort, which is more than any of the rest of us likely have done.
Now back to your regularly scheduled topic.
Out.
I wouldn't attempt to replicate the D2 usage in the M2. My thought was since the D2 is successfully used in the passive 4367, it probably could be made to work nicely with the DBX unit. My project morphed long ago from M2 clone and is now a d-i-y. Planned to be a 3 way, 2245/2123/d2430k in a 4345 enclosure with modified baffle and no Lpad. I've got all the major parts but the cabs. I sold my table saw a long time ago and am hoping to find existing clone cabs that need a home, within a day's drive.
Dear all,
POS, totally agree, I did not use the stock network and hence the impedance would be different in the original M2 system.
-And the 2451 cores are way easier to mount. The D2’s and the 2450’s with the small bolt circle are a challenge to the fingers and the vocabulary. I use pin-bolts and nuts to mount mine.
I see no need for a cap in series with the 4” drivers with normal listening levels and good equipment. The dia’s are pretty sturdy so you would have to have a major “incident” to damage them. Remember that we are only feeding them a fraction of the power they are getting four hours in a row at concerts. And they survive.
Besides short turn off/on spikes what seem to damage the performance is moisture and heat. I have had a few Ti ones that had turned blue from heat. They all worked but the measurements gave at hand that their best days had passed.
Often a restistor can be good to use in series even in an active setup to reduce the negative gain in the XO (dropping bits) or the power amp and also to reduce noise/hiss from the amp, if indeed audible.
The 4” Be’s produces a fair bit cleaner curves and with my amateur a/b testing, sounds slightly more “here/now”. But the draw-back of having to include a UHF (read 045) makes the SL’s more attractive and easier to integrate in my opinion.
I have nothing to add to the M2 vs Revel shoot out and I think Titanium Dome summed up the pros and cons very well. The "sound of listening with the eyes" can be deceptive, but it is probably difficult achieve something significantly better in a non-lab environment. Very good reading and “hat’s off” to the people putting in a lot of time and effort to pull it off.
On a side note.
I really appreciate that we in this community, that deals mainly with hobby projects, can keep the acceptance and appreciation for different opinions and combinations that may be slightly (or very J ) off “stock”. We are not competing with JBL but rather having fun on our spare time and having fellow “sound nuts” chipping in on also “odd builds” to offer sincere help is priceless .
A construction can, and must be, allowed to be very satisfying for the builder even it is does not rival the M2 system performance. It's a hobby for most of us.
Most of us can never challenge the experience and resources of JBL and their premium products but we sure are allowed to have fun trying! J -And having a few beers getting there.
Kind regards
//Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Very well put
Hi POS,
Is it PT-H or PT-F horn type (long pr short) ?
Regards
ivica
I owned the 4367's, then the Salon 2's and now the M2's. In my small listening room they each excelled in different areas.
I am happiest with the M2's, though I haven't heard the bass tonality that I heard on the 4367's yet. I don't have the right amplification it would seem.
Nice! Keep us posted about this.
Hi ivica,
Pretty short: PT-F95HF
Hi POS,
So it is very much the same as STX825 horn type...
I have seen some measurements of the JBL 2453-SL driver in the STX825 horn, and round 12kHz, there has been very ugly deep notch width about 500~1000Hz. I wonder if any spacer between 2453 driver and the horn throat has to be put in order to eliminate such notch.
Some users have made an improvement using 1/2-inch spacer there.
Regards
Ivica
I don't know for the 2453, but the phasing plug in the 2452 is markedly shorter than the one in the 2451/2450SL/476.
That might be a reason why, like with the 2435.
http://www.audioheritage.org/vbullet...l=1#post147356
The 2435H is finicky as to what wave guides it works with because the phasing plug is so short. - Courtesy of Doug Button
FWIW, I couldn't understand why they didn't at least use the same power amps on the Salons and M2's, the preferences displayed could've been down to the amplification imho.
Hi,
Yes, the 2453 is nowhere near as deep as the 2451/2450SL so the phase plug would have to be simplified.
I have not been able to get the smooth 2447/2451/2450SL/476 curve with 2452, 2453 or the 243X drivers. That says absolutely nothing about the sonic quality but the curves are not as even.
Kind regards
//Rob
The solution to the problem changes the problem.
-And always remember that all of your equipment was made by the lowest bidder
Hi Ivica,
I use a 2453h-SL on a JBL 2384 waveguide and have found the need for a 1/2" spacer. Here is a before and after:
In addition, if I rotated the CD a quarter position, the response would change, in each of the 4 possible positions. I have no explanation for that, other than the spacer is not perfect and neither is the fit - some variance in mounting. The one shown in red is the best response.
Hope that is useful.
Kind regards, Mitch
There are currently 2 users browsing this thread. (0 members and 2 guests)
Posting Permissions
Reply With Quote