Super Tweeters

[mypasswordis]

Anyone have experience with them? If you do, please let me know the details and what you think of them. What you're using, how they're setup physically, crossover frequency, dispersion characteristics, horn vs. ribbon, etc. Oh, and uh, how they sound. Thanks.

[Fitz]

They're super.

[mulveling]

I had some Tannoy ST100 on my Kensington SE speakers for a while. I got them last year "on special" from Stone Audio UK, and they arrived looking like B-stock-ish, re-branded ST200.

Overall, I'm still undecided. They can be a bitch to setup, because it's VERY bad if you can hear them distinctly in your seating location - they shouldn't be noticeable until you remove them, so they say. I had mine crossed over at 18KHz and at 91-93dB sensitivity (speakers were 93dB, go up to 22k on their own). The crossover isn't a brick wall - they definitely extend way into the audible range. For the pros - they added a bit more punch to the sense of dynamics, and definitely GREAT extension plus a bit of sparkle on top. For the cons - coherence took a ding and the result was a bit less musical, also in my system things got a bit hot in the upper mids & treble region. However, the cons could very likely have been my fault (setup). Tannoy has a nice gauge that specifies precise placement on virtually every Prestige speaker EXCEPT the Kensington (which is also a rather tall Prestige speaker, proportionally). It was a guessing game I didn't have a lot of patience for. As for the hot treble, that may have been spotlighting/exacerbating a foible in my analog setup. Lately I've switched from a Benz PP-1 to a smoother-sounding PS Audio GCPH (on loan), went from 100 ohm loading to 50 ohms, and lowered the tonearm (VTA adustment) by a notch. This combination of tweaks has helped immensely. I also remember seeing a review with Kensington measurements (older non-SE version) where they had a bit of a bump in that region, so that would indicate they won't combine well to an upstream system with a similar bump. However, by then the ST100 were already sold to fund my amp upgrade I may give them a 2nd shot in the future.

This weekend I heard a properly setup ST100 on Glenair 15s with my Rogue amp/preamp and excellent sources. The ST100 were set to 18K crossover and 91dB sensitivity into the 95dB speakers (go up to 25k on their own). So, clearly they were dialed way down to avoid any hot spots up top. It sounded glorious, but I know for a fact it would still be glorious without them. IMO these are going to be fairly subtle differences. After hearing Glenair 15s with ST100, and both Glenair 10s and Kenstington SEs without, IMO the Apollo monoblocks make a VASTLY greater improvement than any effects of the supertweeter. Then again, the 'blocks are a lot more money. Always keep in mind that 2K or so could go to a super cartridge upgrade

[tkam]

Yeah I really don't see the point behind super tweeters especially in speakers that have no problems getting up to 20khz on their own. Now for single-driver speakers that can't effectively produce volume in that range they make sense.

[mulveling]

For what it's worth, my local tube/speaker tech/designer guy has previously posted (GordonW on audiokarma) that having extension beyond 20k reduces phase error, and that the cumulative phase error for a system adds up from each component. That would be the argument for having speakers extend well past 20k, though as of now I don't understand the physics involved. If that is the case, then I would think that placement of the supertweeters on a speaker would be very precise, and should also take seating position into account.

[guzziguy]

Maybe he was referring to a phase error in the speaker seller's bank account?

[mypasswordis]

Yeah I really don't see the point behind super tweeters especially in speakers that have no problems getting up to 20khz on their own. Now for single-driver speakers that can't effectively produce volume in that range they make sense.

That's exactly what I'm thinking of using them for, if I decide to. I'm thinking crossover point maybe around 14-16kHz. Thanks for sharing your experiences. Any of you have problems with dispersion/beaming? Ie does the supertweeter pretty much have to be level with your head and pointed directly at you?

[mulveling]

That's exactly what I'm thinking of using them for, if I decide to. I'm thinking crossover point maybe around 14-16kHz. Thanks for sharing your experiences. Any of you have problems with dispersion/beaming? Ie does the supertweeter pretty much have to be level with your head and pointed directly at you?

Yeah, super-tweeters could make a lot of sense in that setup. Crossover will be critical. I'm pretty strictly a sweet-spot listener at home, so I'm not the best reference for the dispersion characteristics. I would test the super-tweeters by standing behind them, then taking a CD case and angling it in front of the dome and back towards me - you can hear the super-tweeters quite clearly like this, even with the main speakers playing simultaneously. It's kinda interesting to hear what parts of the music make it above 10k. You could wiggle the angle of the CD case a bit, and still hear them, but any more than that and it drops off rapidly. So I get there impression that there is SOME dispersion, but it's limited, at least in the case of the ST100/200.

You don't have to have the tweeters sitting at ear level, but I've always seen/used the ST100/200 angled (usually down) towards ear levels for the seating location.

Again, FWIW - Gordon has also said that properly setup super-tweeters expand the system's sweet-spot. I'm not sure how this works, but he does know an awful lot more then me about acoustics & electronics.

[Craig Sawyers]

I haven't used or listened to supertweeters, so this thread is kind of interesting to me. I did however come across a reference in The Journal of Neurophysiology Vol. 83 No. 6 June 2000, pp. 3548-3558, entitled "Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect". The web link is Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect -- Oohashi et al. 83 (6): 3548 -- Journal of Neurophysiology . The full text link works perfectly.

Basically these guys used Gamelan music, which has a lot of spectral content from 20kHz to 50kHz. They tested subjects with full range, up to 20kHz and just 20kHz to 50kHz. They used EEG and PET scanning. What they showed was that areas of the brain lit up specifically when the full bandwidth music was played - so something is happening when there is music coherent frequencies present outside the range of normal human hearing. In their conclusion they say that they have no idea what the actual mechanism is.

[Duggeh]

I've got a pair and unless you set them to obvious mode, you don't spot what they're doing until you take them away. They help in building air and space in the sound. Sort of filling in blanks that you hadn't noticed before. Tricky to get them right and inappropriate for use with some setups. I cannot get them to agree with the Soundlabs but they could play nice with the Quad 12L in nearfield or the Quad 21L in a room setup. I have hopes of building a set of J-Lows over the summer and I'll be keen to try them with those.

The hypersonic effect isn't well understood at all, but Dr Takei at TakeT found that a supertweeter used outside of a closed headphone affects the perception of the sound too.

[Bigguy]

Man, I thought this was a thread for prolific Twitter users.

[mypasswordis]

In their conclusion they say that they have no idea what the actual mechanism is.

The thing is, you need enough bits for that kind of treble extension. With the CD brickwall of 22kHz there's no point in going past, well, 22kHz, and in fact may be a bad thing if you do if the DAC sucks. But I would be very curious trying supertweeters out on hi-rez (outside of actually bringing out the top audible octave).

The hypersonic effect isn't well understood at all, but Dr Takei at TakeT found that a supertweeter used outside of a closed headphone affects the perception of the sound too.

A supertweeter used... outside of a closed headphone? That's pretty out there, I must say.

Again, FWIW - Gordon has also said that properly setup super-tweeters expand the system's sweet-spot. I'm not sure how this works, but he does know an awful lot more then me about acoustics & electronics.

If I understand correctly, the smaller the driver (horizontally and/or vertically), the higher the frequencies the driver will start to beam. The ribbons I see are typically long but narrow which means they have good horizontal and bad vertical dispersion. The fact that the full-range panels are so big on my Acoustats mean they beam starting from a very low frequency.

Here's an example of the opposite, a short ribbon with good vertical dispersion:

[luvdunhill]

The reason I'd never consider a super tweeter in the traditional application as an "add-on" would be the problems that it would add in the time/phase domain. My ears say that this is important.

I'd consider one if the speaker was designed for one, depending on the design decisions. It's very difficult to integrate multiple drivers and my experience designing crossovers says the higher the frequency that you have to integrate, the harder it is, so I'd expect it to be a very hard nut to crack.

[skullguise]

I used to use the Tannoy's with a Lowther in a bass reflex cabinet. They worked well in that system, though it was still tough to get them integrated well. The Lowther's had a higher efficiency, and I had to play with crossover and physically moving the tweeter back and forth on the top of the speaker to get the best integration. Some music didn't seem to be affected, but other music did not sound natural.

But when it snapped together sound-wise, it made some very nice music. It DID seem to grow the sweet spot (which can be nasty for some single-ended drivers, their off-axis response gets ugly at times), and of course added all that good hi-fi stuff: "air," "transparency," "detail," etc.

BTW, I crossed mine over closer to 17-18kHz IIRC.....just sounded better to me.