Despite my vilification of their sound quality, I have recorded MP3s and enjoyed listening to them (in limited doses) when the material was not available in a higher quality format. And I acknowledge that at 192kb/s or higher, MP3s can often approach 'CD quality', although I much prefer high res digital to CDRB standard.
Wrt the significance of driver performance above 10khz, complaints about metal dome tweeter HF 'harshness' are quite common and most of them appear to be related to a resonance peak that the diaphragm has between 15-25khz. Such comments even extend to compression drivers where aluminum is generally acknowledged to sound better than titanium because of its higher rigidity (as applied in the diaphragm)(unfortunately, aluminum suspensions also tend to work fatigue long before titanium does in an equivalant driver application) and the relative sonic merits and demerits of various suspension constructions such as roll vs diamond pattern or metal vs plastic. Also, the sound of a compression driver I own, the TAD 4001, although overall superior due to its beryllium diaphragm has been criticized for a 'thinness' which appears to be related to a 4 db peak at 16khz. Almost all of the above mentioned factors affect performance predominantly above 10khz and many beyond the sine wave perceptual capability of almost all of us but have resulted in consistent responses from a wide variety of listeners who have had the opportunity to listen to and compare them.
Wrt the significance of driver performance above 10khz, complaints about metal dome tweeter HF 'harshness' are quite common and most of them appear to be related to a resonance peak that the diaphragm has between 15-25khz. Such comments even extend to compression drivers where aluminum is generally acknowledged to sound better than titanium because of its higher rigidity (as applied in the diaphragm)(unfortunately, aluminum suspensions also tend to work fatigue long before titanium does in an equivalant driver application) and the relative sonic merits and demerits of various suspension constructions such as roll vs diamond pattern or metal vs plastic. Also, the sound of a compression driver I own, the TAD 4001, although overall superior due to its beryllium diaphragm has been criticized for a 'thinness' which appears to be related to a 4 db peak at 16khz. Almost all of the above mentioned factors affect performance predominantly above 10khz and many beyond the sine wave perceptual capability of almost all of us but have resulted in consistent responses from a wide variety of listeners who have had the opportunity to listen to and compare them.
Last edited:
What kind of SPL and how much higher?? If you want loud, clean and don't mind them crashing at 22K JBL 2405's/2404's will work just fine. They are also available all over the place.
Rob🙂
Do you run them solo or use a tweeter from say 10K and up?? I have the JBL berylliums and don't use a tweeter above them. They have issues as well above 10K but they sound good as is.
Rob🙂
The JBL 2405's/2404's seem to be the choice in a lot of high end systems.
I've always wondered if it was because they sound good, or because "everyone else uses them." 😉
A surprise to me has always been the Hi-Vi planar tweeters. Not very expensive and they sound a lot better than they ought to.
I've always wondered if it was because they sound good, or because "everyone else uses them." 😉
A surprise to me has always been the Hi-Vi planar tweeters. Not very expensive and they sound a lot better than they ought to.
Hi -
I run the TAD4001s without a supertweeter. With shaping, they appear to be good up to around 20khz. I basically agree with Dr. Geddes regarding the complexities of adding a supertweeter and also feel that to achieve time alignment with a supertweeter is somewhat a losing battle in the real world due to acoustical center/listener position alignment variables, although as long as the objective is restricted to adding 'air', they'll certainly accomplish that. In the future, I'll introduce a mild notch at 16khz to control this characteristic of the TAD4001.
I run the TAD4001s without a supertweeter. With shaping, they appear to be good up to around 20khz. I basically agree with Dr. Geddes regarding the complexities of adding a supertweeter and also feel that to achieve time alignment with a supertweeter is somewhat a losing battle in the real world due to acoustical center/listener position alignment variables, although as long as the objective is restricted to adding 'air', they'll certainly accomplish that. In the future, I'll introduce a mild notch at 16khz to control this characteristic of the TAD4001.
Last edited:
While waiting for the PC to catch up, a PS3 slim is probably the better HTPC.
Best, Markus
Yes it's trade off for sure. I have 2 systems one with and one without, although I would not call 2405's supertweeters. I am in the middle of cloning another system that has real supertweeters, response above 20K, so I am real curious to see if I feel the supertweeters are worth it or not.
LOL Well hard to say. I know you can set them up to sound really good.
Rob🙂
I have to admit that the experiences that you quote are quite common and I myself have had them, which is the only data that I know of to support that above 10 kHz is a significant factor. But here is a counter argument. The DE250 has a very similar resonance at about 16 kHz and yet does not have the characteristics that you describe. Since there are other explainations as to the subjective factors that you quote I am inclined to somewhat discount this widely held belief simply because its not always true.
I am inclined to believe that the strong HF resonances above 10kHz have an effect below 10kHz as sub harmonics or something like that. The whole concept of sub-nonlinearity was not part of our distortion study since all of the distortions that we created in our test would not have created any sub-harmonics. So I don't discount that sub-harmonic distortions - which WOULD NOT be masked - could be a major audible effect. But what needs to be understood is that in the classic Volterra series modeling of nonlinear systems sub-harmonics cannot occur. So the "how" they are generated becomes the issue long before the audiblity aspects.
My testing of this driver showed it to die above about 8-10 kHz. The owner pointed me to the "spec sheet" showing response well above this and I pointed out that on the horn they were using for the test the device would be extermly beamy at those freqs, thus raising the low output by directionality and not sound output.
I do agree that metal dome tweeter harshness are quite common. However, I'm not too sure whether its attributed to the resonance peak between 15-25K.
What I am certain about are:
1) Tweeter Fs
Some tweeters have nasty resonance. When crossed low (<3K), an LCR must be used, otherwise its going to sound terrible. I am particularly sensitive to this.
2) Cone breakup of mid/woofer.
Because this is buried underneath the tweeter, we take it for granted. This is one of the main causes for harshness.
3) Crossover
When its not optimized, it results in peaking and nulling. And this is on-axis. One can only imagine its dispersion properties.
the factors you mention can certainly create harshness, but I have some personal experience regarding the HF resonance contribution with a Focal metal inverted dome tweeter that had a hint of grit. The manufacturer's data sheet showed a significant resonance in the 21-24khz region. When I added a partial notch filter to the xover, the sound, for lack of a better euphemism, became much 'sweeter' without losing HF extension.
This is conjecture, but I wonder if the HF harshness that was noted here was due to intermodulation distortion products in the band of energy around the tweeter's HF resonant region which might typically span a couple to a few khz. Assuming that the diaphragm motion is relatively uncontrolled there, and with the increased amplitude response of up to 10db, it seems reasonable that the HF unit's intermodulation distortion products would become very much more prominent for input in this frequency range, and the low band distortion products would probably land right in the HF passband in the midrange where they are most audible. Seems plausible, at least, and should be subject to verification either way by testing.
This is conjecture, but I wonder if the HF harshness that was noted here was due to intermodulation distortion products in the band of energy around the tweeter's HF resonant region which might typically span a couple to a few khz. Assuming that the diaphragm motion is relatively uncontrolled there, and with the increased amplitude response of up to 10db, it seems reasonable that the HF unit's intermodulation distortion products would become very much more prominent for input in this frequency range, and the low band distortion products would probably land right in the HF passband in the midrange where they are most audible. Seems plausible, at least, and should be subject to verification either way by testing.
Last edited:
Hi -
That's the supposition here. If this were the case, the IM would be probably 'weakly correlated' with the overall musical signal so it might be more audible than, say, IM generated from midband information.
That's the supposition here. If this were the case, the IM would be probably 'weakly correlated' with the overall musical signal so it might be more audible than, say, IM generated from midband information.
Hmmm.... I wonder how you could measure or test for this.
Finding out where those harsh edges live would be the 1st step in getting rid of them.
Finding out where those harsh edges live would be the 1st step in getting rid of them.
I know what you're talking about. I heard in one of my customers' Watt Puppy. I made sure I kept my mouth shut.
Don't have a Focal with me, so can't perform any test. Since a notch helps, it points to some sort of peaking. I wonder whether these extended frequencies like 30-40K are cone breakup in disguise.
Hi Earl,
How do you connect the 3 subs? Daisy chain or start configuration? Are you using the same preamp out to connect to the 3 subs? Any sonic lost?
Listening test
Had a fun morning today. RA7 came over to my place to listen to Pro vs Hi-Fi drivers. I think we had a good and fair test.
The two setups were as follows:
Hi-Fi:
Von Schweikert VR2 towers. A narrow floor standing TWW with rather nice drivers. A good speaker that I like a lot. You can find specs online
Pro:
Altec Voice of the Theater A5. Altec 825 bass cabinets with 15" 416-8A woofers, 803B 300Hz 8 cell horn with Altec 291 drivers, Selenium Ti210 tweeters on "BBC" horns. Crossovers 450Hz active, 5KHz passive.
Electronics:
HP laptop, Windows XP. JRiver player, ASIO drivers. M-Audio DUO USB to SPDIF. 75:110 ohm transformer to CAT5 cable. Berhinger DEQ2496 AES digital input for room correction (slight). AES digital out to DCX2496 for crossover, custom transformers on the outputs. (All filters bypassed to drive the Von Schweikerts)
Icepower amp for Altec woofers and VR2. Modified Tripath amp for horns. 14 gauge power cords for speaker cables. 😉
So both pair of speakers were driven by the same source, same amp and same DAC. Levels matched to within 0.2 dB on pink noise.
My thoughts:
The Von Schweikerts put on a good show, they did not embarrass themselves at all. They did some things better than the Altecs. They sound much bigger than they look. (Good room acoustics, good DAC). At low to medium levels, they are very nice to listen to. Good depth, width and detail. Fine vocals. Even level matched they sound louder than the Altecs. I suspect higher distortion and compression, but could be wrong.
The Altecs are big and sound big. But I don't find they sound like they look. They look like they will pin your ears back and part your hair. But they don't, they mostly just get out of the way. Great depth, width and even some nice height. Can play loud, loud , loud and never strain or get into trouble. But also sound nice at low levels. Sense of space is immense. Even old mono recordings sound 3D. (it's all there in the recording, you just have to decode it).
We listened to many types of music. Female vocal - both old and new, electronic, jazz, classical, classic rock and even Led Zep turned up to 11. Good fun.
For me it was a good demo that pro drivers can sound just as refined as Hi-Fi drivers, and in some ways more so. Better detail, better ambient clues, etc. And that horns do not have to sound "in your face." Ra7 made the comment that they are very easy to listen to. I agree, as long as the crossover is done right. Most of what people don't like about horns and pro drivers is bad crossovers, not bad drivers.
I hope that Ra7 will share some of his thoughts. He will probably disagree with me on some points, I hope he does!
Had a fun morning today. RA7 came over to my place to listen to Pro vs Hi-Fi drivers. I think we had a good and fair test.
The two setups were as follows:
Hi-Fi:
Von Schweikert VR2 towers. A narrow floor standing TWW with rather nice drivers. A good speaker that I like a lot. You can find specs online
Pro:
Altec Voice of the Theater A5. Altec 825 bass cabinets with 15" 416-8A woofers, 803B 300Hz 8 cell horn with Altec 291 drivers, Selenium Ti210 tweeters on "BBC" horns. Crossovers 450Hz active, 5KHz passive.
Electronics:
HP laptop, Windows XP. JRiver player, ASIO drivers. M-Audio DUO USB to SPDIF. 75:110 ohm transformer to CAT5 cable. Berhinger DEQ2496 AES digital input for room correction (slight). AES digital out to DCX2496 for crossover, custom transformers on the outputs. (All filters bypassed to drive the Von Schweikerts)
Icepower amp for Altec woofers and VR2. Modified Tripath amp for horns. 14 gauge power cords for speaker cables. 😉
So both pair of speakers were driven by the same source, same amp and same DAC. Levels matched to within 0.2 dB on pink noise.
My thoughts:
The Von Schweikerts put on a good show, they did not embarrass themselves at all. They did some things better than the Altecs. They sound much bigger than they look. (Good room acoustics, good DAC). At low to medium levels, they are very nice to listen to. Good depth, width and detail. Fine vocals. Even level matched they sound louder than the Altecs. I suspect higher distortion and compression, but could be wrong.
The Altecs are big and sound big. But I don't find they sound like they look. They look like they will pin your ears back and part your hair. But they don't, they mostly just get out of the way. Great depth, width and even some nice height. Can play loud, loud , loud and never strain or get into trouble. But also sound nice at low levels. Sense of space is immense. Even old mono recordings sound 3D. (it's all there in the recording, you just have to decode it).
We listened to many types of music. Female vocal - both old and new, electronic, jazz, classical, classic rock and even Led Zep turned up to 11. Good fun.
For me it was a good demo that pro drivers can sound just as refined as Hi-Fi drivers, and in some ways more so. Better detail, better ambient clues, etc. And that horns do not have to sound "in your face." Ra7 made the comment that they are very easy to listen to. I agree, as long as the crossover is done right. Most of what people don't like about horns and pro drivers is bad crossovers, not bad drivers.
I hope that Ra7 will share some of his thoughts. He will probably disagree with me on some points, I hope he does!
I recommend a Behringer DCX2496, but I got by for along time just driving three subs from one preamp out.
- Status
- Not open for further replies.