40 Years is a Long Time...
jackgiff said:
After 40 years of listening to Valencias. and trusting the marketing pitch of how great they are, I am pissed at myself for falling for all the BS. Now after some research and measuring, we know that they aren't really all that great. But they were designed to follow the X curve, which doesn't cut it for home stereo, but is useful in large theaters. I have had friends over who listened to my Vals, and thought they were really great. So maybe the "trust your ears" thing can be good. But knowing how much better they can be is simply amazing...
snip...
Jack Gifford
Don't you think you’re being a little hard on yourself when you say you're pissed for falling for all the BS. You ended up with what you had by listening and isn't that what's really important? The situation couldn't have been too bad if it took 40 yrs to realize something better might have come along. Speaker sound is such a personal thing I sometimes wonder how there can be any consensus as to what sounds good and what doesn't.
I've been following this thread (and lurking on the LH board) because I've got a pair of Valencias and a pair of Carmels and both exhibit the “flaws” being discussed (shouty, colored midrange and no real VHF). OTOH they have some characteristics I find really attractive such as high efficiency and a dynamic sound. I also like the basic concept of a two-way design. IOW, I really want to like these speakers but it hasn’t been easy.
Over the years I've coped with the worst part of the flaws by pairing the speakers with vintage equipment that have "loudness" controls or thru judicious use of tone controls but I consider this really a band-aid. Ideally, I’d like to see the speakers perform more neutrally on their own so I could have more latitude in the paired electronics. I had never considered changing HF drivers or making xover mods as a viable solution but now I see that this could indeed be the best solution. Also, I never realized how haphazard Altec was when it came to xovers. It really bothers me that the N800-F came in at least two undifferentiated versions AND that Altec specified the N800-F for use with systems containing 16 ohm LF and HF drivers AND for use with systems containing 8 ohm LF and 16 ohm HF drivers. Sloppy or cavalier and certainly not a means of showcasing high quality drivers. Same goes for Altec cabinets to a lesser extent (one box, many drivers spec’d for it w/o sig change)
Zilch has done a good job diagnosing the root causes of the flaws and has provided some interesting suggestions for fixing them. However, I think this journey is still a ways from being completed.
As I allude to above, I find the two-way concept much more attractive than the three-way, add-on tweeter approach. So… As a direct of Zilch’s work, I decided to acquire a pair of 8 ohm BMS 4550s that I plan to use with the Carmels. Due to lack of free time, I haven’t had much of a chance to play with the BMS drivers yet but I can offer these preliminary listening observations based on a 511 horn and compared to Altec 808-8B also on a 511:
1. The 4550 driver is definitely more efficient than the 808-8B. No meas yet but I’d guestimate 6-8dB mid band.
2. The 4550 goes higher than the 808 but the difference is not as pronounced as I was expecting.
3. The 4550 does not go as low as the 808. The specs state that the low end of the 4550 is 500Hz with rec xover of 800Hz. Based on my listening, I believe the 4550 would do a lot better @ >=1KHz.
4. Driven full range at low power (~50mW) the 808 sounds much fuller, relaxed and effortless than the 4550. The 4550 sounds “pinched” and maybe strained. I realize that full range operation is not a designed-for condition for either driver but I find it revealing. I wonder if the unstrained full-range sound is somehow related to the tendency of devotees to the Altec sound aesthetic to overlook the obvious measured flaws.
The next step will be to make the driver comparison using more-or-less optimized xovers. For reasons I note above, the stock N800-F won’t cut it, even with compensation network(s). Since the Carmel LF section is nominally 8 ohms, a starting point might be Zilch’s “N800-8K Modified”. However, since I’m going to be going for 1KHz, modifications will be required. First thought would be to simply scale Zilch’s design from 800Hz to 1KHz. As I think about it, that may be too simplistic an approach. The HF cutoff will be getting awfully close to the midrange notch filter and I expect to see non-intuitive interaction betw the two filters to occur. I would also like to use a 2nd order L-R network which the N800-8K really isn’t. Lastly, I assume Zilch’s design is optimized for the non-conventional Altec practice of making the HP cap 1/2 the value of the LP cap even though inductors are equal. I think my personal journey is just beginning. Guess I’m going to have to start making my own acoustic measurements. Or maybe I’m thinking about this too much.
As a closing comment, It sure would be cool if we all were physically closer to each other. I know my learning curve would be steeper. I’d also like to have been able to participate in the “Valencia auditions” about 12 hours from now but physical distance prevents it. The internet may have eliminated the distance between us in some respects but when it come to things like speakers, I guess you’ve just gotta be there. Good luck on Saturday. Looking forward to listening session results.
