Announcement

Collapse
No announcement yet.

I have my work cut out for me KEF LS50

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • stephenmarklay
    started a topic I have my work cut out for me KEF LS50

    I have my work cut out for me KEF LS50


  • Pete Basel
    replied
    Originally posted by Brian Steele View Post

    ​Jeez, my 29-year old Mordaunt-Short MS3.2s with a modified x-over and replacement cheap Audax mylar tweeters have a smoother measured response than that. I suspect that the LS50 has a lot more going for it than just the measured FR. It actually makes me wonder a bit how important getting a flat FR actually is in loudspeaker design...
    I've been listening to a pair of LS-50s for over a year and on the first listen, without having seen that*
    curve, my impression was that they did not have enough baffle step, yet it can be seen from that anechoic*
    curve that they have more than enough.* My theory is that the elevation from 2K to 5K in a sense "un-does"
    the baffle step seen in the curve, it is all relative after all.* They are decent but not anywhere near the best.
    I've also spent a lot of time listening to Celestion SL700s and like them much more, they are laid back, yet
    provide the feeling of hearing real music and not sound out of a box.* Neither have enough max output for
    me as I prefer (much) larger speakers.
    I'm surprised that the LS50 gets such good reviews, then again Harman found that reviewers were not*
    very good listeners.
    It'd be interesting to see a curve on the active version of the LS-50 to see if they made it smoother, I'd hope
    so EQ is cheap.
    They've really come a long way on that coincident driver and I'd like to use it as a midrange/tweeter in a*
    larger design.* One could do a much better version of the Thiel CS7.2 than the real thing with it.

    Leave a comment:


  • stephenmarklay
    replied
    Lots of cool feedback here. Thanks everyone.

    Leave a comment:


  • fpitas
    replied
    Originally posted by DDF View Post
    IME, its at times necessary to trade off a bit of on-axis flatness for a neutral perceived response. This often takes time to hear and work through. I posted one experienece with that here. I started with textbook perfect on paper crossover and eneded up varying fairly significantly to get neutral perceived response (mine and others listeners perceptions FWIW):
    http://www.diyaudio.com/forums/multi...-redesign.html

    From my experience I believe this is due to the variation in off axis response and can only be really objectively glimpsed at using many off axis measurents that extend to broad angles. 0, 30, 60 deg isn't enough IME.

    Reading on this as much as I can, more than a few professional designers opinions align that on axis deviations from flatness are often needed to make a design sound neutral. For example, some designers like to add a bit of lift to a 2 way woofer where the driver starts to beam before crossover, to combat this.

    Problem with short auditions in large rooms like a diy event:
    - there isn't enough time to hear some issues because they are only exposed with the right music. Working out the nit picky details requires allot of time with allot of music
    - large rooms hide problems with off axis response due to the long delays of the reflections (the ear doesn't fuse them as tonal aberations as much as short reflections), Problems that would stand out more in the home just aren't so audible in the big rooms and flat on axis speakers usually do better at events even if there are some issues to still work out off axis.

    Case in point, I brought my first cut at a diy design using the W4-1720 to a diy event. I designed it for textbook on paper responses. I had never heard it before. At least 4 people decided right there to build them and ordered parts, it sounded really good in that setting for what it was (just saying). Large room, just an hour of listening. Over time and several iterations, I redisigned the crossover. 3 times. It takes time to hear problems and time to fix them. This is one area I think profesional designs probably exceed most diy. My perception is that the best commercial designs, objectively designed for neutral response also are the products of allot of seat time time (hours and different listeners) listening and adjusting.

    Stephen, as an aside, if you really want a very high level of resolution from the W4-1720, I think it has to have a double notch. They open right up and the grain recedes (but doesn't completely disapear) once that's done. I listen to these daily, near field. My experiences with that are here http://www.diyaudio.com/forums/multi...-weight-5.html I don't actually recommend this design for your goals (it was designed for a completely different set of trade offs) but I don't think you'll get the level of resolution you're looking for without these issues notched.
    Exactly. The LS50 has a horn, and horns can be especially tricky; they often have wavey directivity with corresponding wavey on-axis response. For the best results, one has to be very careful EQing things flat on-axis. Often that doesn't give optimum results. It helps a lot to know the DI of the horn vs. frequency.

    Leave a comment:


  • DDF
    replied
    IME, its at times necessary to trade off a bit of on-axis flatness for a neutral perceived response. This often takes time to hear and work through. I posted one experienece with that here. I started with textbook perfect on paper crossover and eneded up varying fairly significantly to get neutral perceived response (mine and others listeners perceptions FWIW):
    http://www.diyaudio.com/forums/multi...-redesign.html

    From my experience I believe this is due to the variation in off axis response and can only be really objectively glimpsed at using many off axis measurents that extend to broad angles. 0, 30, 60 deg isn't enough IME.

    Reading on this as much as I can, more than a few professional designers opinions align that on axis deviations from flatness are often needed to make a design sound neutral. For example, some designers like to add a bit of lift to a 2 way woofer where the driver starts to beam before crossover, to combat this.

    Problem with short auditions in large rooms like a diy event:
    - there isn't enough time to hear some issues because they are only exposed with the right music. Working out the nit picky details requires allot of time with allot of music
    - large rooms hide problems with off axis response due to the long delays of the reflections (the ear doesn't fuse them as tonal aberations as much as short reflections), Problems that would stand out more in the home just aren't so audible in the big rooms and flat on axis speakers usually do better at events even if there are some issues to still work out off axis.

    Case in point, I brought my first cut at a diy design using the W4-1720 to a diy event. I designed it for textbook on paper responses. I had never heard it before. At least 4 people decided right there to build them and ordered parts, it sounded really good in that setting for what it was (just saying). Large room, just an hour of listening. Over time and several iterations, I redisigned the crossover. 3 times. It takes time to hear problems and time to fix them. This is one area I think profesional designs probably exceed most diy. My perception is that the best commercial designs, objectively designed for neutral response also are the products of allot of seat time time (hours and different listeners) listening and adjusting.

    Stephen, as an aside, if you really want a very high level of resolution from the W4-1720, I think it has to have a double notch. They open right up and the grain recedes (but doesn't completely disapear) once that's done. I listen to these daily, near field. My experiences with that are here http://www.diyaudio.com/forums/multi...-weight-5.html I don't actually recommend this design for your goals (it was designed for a completely different set of trade offs) but I don't think you'll get the level of resolution you're looking for without these issues notched.

    Leave a comment:


  • Pete Schumacher
    replied
    Originally posted by dkalsi View Post
    Wow - how high can the anarchy be crossed? Sent from my iPhone using Tapatalk
    Maybe 2000Hz, but you need to have a tank circuit to take care of a cone resonance. I have them crossed at 1300Hz to a Wavecor 30mm tweeter.

    Leave a comment:


  • dkalsi
    replied
    Put the Anarchy 7" 4 Ohm woofer in a .5 cubic foot enclosure tuned to 42Hz and you can expect over 109dB out of it while never exceeding Xmax from 50Hz on up. You can do that for less than $70.
    Wow - how high can the anarchy be crossed? Sent from my iPhone using Tapatalk

    Leave a comment:


  • fpitas
    replied
    That 2 - 5kHz ripple in the LS50 frequency response might be due to the horn dispersion changing. If the total radiated power remains constant, it may not be particularly audible.

    Leave a comment:


  • Drjay
    replied
    Now that I can measure FR, I've measured a bunch of commercial and DIY speakers I have on hand. I have observed that some of the designs with very similar FR sound significantly different. Two of the commercial designs I have, which are Stereophile C rated, sound quite different, but one DIY design is barely distinguishable from one of the high rated commercial designs and a fraction of the price. So, I think what is going on with the KEF LS50 is due to something other than just FR. I suspect that fairly flat FR is necessary, but not sufficient, to having an excellent sounding speaker. Perhaps there is some " magic" in the concentric driver topology...
    It would be a bit scary to all DIY "believers" (including me), but a blind comparison between the Kef LS50 and a couple of top quality DIY speaker of about the same size would be very, very interesting.

    Leave a comment:


  • Face
    replied

    Put the Anarchy 7" 4 Ohm woofer in a .5 cubic foot enclosure tuned to 42Hz and you can expect over 109dB out of it while never exceeding Xmax from 50Hz on up. You can do that for less than $70.
    It'll also sound good in a sealed enclosure that size.

    Leave a comment:


  • stephenmarklay
    replied

    Put the Anarchy 7" 4 Ohm woofer in a .5 cubic foot enclosure tuned to 42Hz and you can expect over 109dB out of it while never exceeding Xmax from 50Hz on up. You can do that for less than $70.
    OK I wil try

    Leave a comment:


  • djg
    replied
    https://www.youtube.com/watch?v=-iGD0p0E42A

    Leave a comment:


  • Pete Schumacher
    replied
    Originally posted by humphreyblowdart View Post

    I can't think of a single 4 ohm, 165 mm woofer that can do that with 88 dB system sensitivity and hit 110 dB peaks.

    Even if you assume the specs below 500 Hz are exaggerated, and knock maybe 3 dB off everything, pretty much the only 7ish" woofer that could give you 107 dB and still net 85 dB with baffle step loss would be an SS 18WU and the only coax with comparable features (shorting rings, prominently) would be the SEAS C16. That's over $1100 in drivers, as a starting point. $165 for cabinets and 3-way crossovers would be a tight squeeze.
    Put the Anarchy 7" 4 Ohm woofer in a .5 cubic foot enclosure tuned to 42Hz and you can expect over 109dB out of it while never exceeding Xmax from 50Hz on up. You can do that for less than $70.

    Leave a comment:


  • Brian Steele
    replied
    Originally posted by fpitas View Post
    Our ears adapt pretty quickly to response irregularities, so it's sometimes not easy to spot mild deviations in frequency response.
    ​I agree. Today I was measuring the frequency response of my Mini-Mes and the rebuilt Mordaunt-Short MS3.2s. Both of them have a measured frequency response that's pretty close and slightly tilted upwards from about 5kHz (my preference), but the MS 3.2 sounds brighter than the Mini-Me when played side by side. However, played separately they both sound fine and nothing "stands out" as far as I'm concerned. So, while gross frequency response aberrations may be an issue, I don't think minor ones really are.

    Or maybe I'm just not gifted with "Golden Ears", LOL. Cased in point - part of my "rebuild" of the MS3.2s involved finding out why the cabinets were apparently leaking around Fb and causing gross and measurable distortion when a pure tone is played at that particular frequency. I traced the problem to a design flaw with the tweeter flanges that resulted in them being a bit leaky. I sorted out the issue with a bit of hot-glue, but to be honest I really can't hear any difference in response when playing music through them.

    Leave a comment:


  • jhollander
    replied
    Originally posted by humphreyblowdart View Post

    I can't think of a single 4 ohm, 165 mm woofer that can do that with 88 dB system sensitivity and hit 110 dB peaks. I know manufacturer's specs are 'optimistic', but that seems like an apostasy against Hoffmanism. If you had a second woofer, an 8", or floor-loaded the woofer I could maybe see it, but that wouldn't be quite the same thing. I also won't take for granted that any of the coax drivers available to DIYers are as good as the KEF units. Maybe the newest SEAS units are, but those don't come cheap.

    Leave a comment:

Working...
X