Super-Tweeters: Acoustic Information or Phase Information?
- Thread starter Ron Resnick
- Start date
Both the TheSonusFaber (Fenice) and the Aida included two medium and a tweeter back firing units. We can adjust their level according to our preferences, and in the case of TheSonusFaber rotate the whole block of the back units.
Unfortunately, as these units are wired in parallel with the main speaker, increasing their level decreases the impedance in their bandwidth.
Dave Wilson says something similar - in his opinion accurately timed (phase) high frequency units are needed for "authentic sounding" - see this article: Life In the Rearview Mirror - Wilson Audio for an explanation of what is "authentic sounding". Fortunately , as he says" (I quote) "However, alignment errors of as little as 10 ?s between the tweeter and the WAMM MC’s drivers are quite audible to listeners over 50 years of age (even over 70!). So the ear is very sensitive to even low microsecond range errors. Furthermore, this sensitivity to timing is much less diminished by age than is presbycusis (loss of high frequency spectrum hearing). This is good news for those who want to continue to enjoy recorded music during their retirement years! "
Phase at supertweeter frequencies is somewhat audible in some cases but mostly inconsequential IME.
The reason why... unless you place your head within a fraction of an inch of the same place and never move it (head must literally be in a vice), phase will change depending on head placement since the wavelengths are so short. Coax drivers may be an exception but they have other issues that are even worse than the issue they are trying to solve.
In my own testing it simply doesn't matter despite many speaker designers coming up will all sorts of reasons it doesn't work, but it seems this reasoning is made without any sort of confirmation via actually listening.
There is certainly a range of frequencies where phase relationship matters A LOT, but this is over a few hundred Hz and under 10 kHz and it makes the most difference near xo frequencies where drivers overlap. People claiming it matters so much at extreme high and low frequencies may have heard some circumstance where it does matter, but IME this isn't for the reasons they assume. With more testing under a wider variety of conditions, room, systems, etc all that stuff simply goes away.
For example the intuitive thought that a rearward-firing tweeter would simply add constructive and destructive interference that would sound horrible in the same way bass frequencies might when you get nulls and peaks is correct in some ways but also almost certainly inaudible despite being able to demonstrate comb filtering using measurements. These measurements paint a bleak picture but luckily this is not how we actually hear. Once again measurements are incorrectly interpreted due to the failure to simultaneously understand psychoacoustics.
There are MANY reasons a supertweeter may not fully integrate in a system but phase and comb filtering are not valid reasons however many times people may claim it to be so.
The reason why... unless you place your head within a fraction of an inch of the same place and never move it (head must literally be in a vice), phase will change depending on head placement since the wavelengths are so short. Coax drivers may be an exception but they have other issues that are even worse than the issue they are trying to solve.
In my own testing it simply doesn't matter despite many speaker designers coming up will all sorts of reasons it doesn't work, but it seems this reasoning is made without any sort of confirmation via actually listening.
There is certainly a range of frequencies where phase relationship matters A LOT, but this is over a few hundred Hz and under 10 kHz and it makes the most difference near xo frequencies where drivers overlap. People claiming it matters so much at extreme high and low frequencies may have heard some circumstance where it does matter, but IME this isn't for the reasons they assume. With more testing under a wider variety of conditions, room, systems, etc all that stuff simply goes away.
For example the intuitive thought that a rearward-firing tweeter would simply add constructive and destructive interference that would sound horrible in the same way bass frequencies might when you get nulls and peaks is correct in some ways but also almost certainly inaudible despite being able to demonstrate comb filtering using measurements. These measurements paint a bleak picture but luckily this is not how we actually hear. Once again measurements are incorrectly interpreted due to the failure to simultaneously understand psychoacoustics.
There are MANY reasons a supertweeter may not fully integrate in a system but phase and comb filtering are not valid reasons however many times people may claim it to be so.
I have ordered the Aperion Audio Dual AMT Ribbon Super Tweeters from Amazon to try with my Acoustats. I have always felt that they were a bit rolled of in the highs (and lows as well of course) and hopefully these will add to the presentation. I went with these because they are dipole super tweeters which should help them integrate with the speakers. They will be mounted on small adjustable stands on the insides of the speakers. This will allow me to experiment with tweeter height, depth (flush with panel, behind, in front) and tilt. Will report on the results.
Cheers!
Cheers!
Just starting to fine tune. Have the crossover at 14kh and the volume at -4db. Like subwoofers less seems to be more, you don’t want them to stand out, just blend in. So far more detail, better attack, more holographic soundstage with better centre fill and less noticeable speaker location. Not brighter but more illuminated maybe. More listening and tuning to come. So far I like what I’m hearing.
Attachments
Similar threads
