Hi Phase accurate
The way we (a little known company) have been able to grow is on the strength of the sound in previous installations and side by side comparisons.
The largest companies generally depend on marketing the image of technology and fortunately the stuff tends to sound pretty dismal side by side.
Clarity is the things most comment on, in commercial installations.
I do like transients though, play the fireworks recording on our web site through some good headphones (first).
All of the full range speakers take a turn in my living room as the main system.
These (SH-50’s) are too large, too heavy and too ugly for the typical home and while not as good as the Manger in time, unlike the Manger having a huge headroom over average listening level means very low distortion too and they have real directivity, I could never go back to regular speakers.
Best,
Tom
The way we (a little known company) have been able to grow is on the strength of the sound in previous installations and side by side comparisons.
The largest companies generally depend on marketing the image of technology and fortunately the stuff tends to sound pretty dismal side by side.
Clarity is the things most comment on, in commercial installations.
I do like transients though, play the fireworks recording on our web site through some good headphones (first).
All of the full range speakers take a turn in my living room as the main system.
These (SH-50’s) are too large, too heavy and too ugly for the typical home and while not as good as the Manger in time, unlike the Manger having a huge headroom over average listening level means very low distortion too and they have real directivity, I could never go back to regular speakers.
Best,
Tom
Svante said:
Hi Svante,
I asked about using headphones, because one of the journal papers, from many years ago, includes a similar experiment to yours. And yes, I agree, that we are trying to determine if preserving the time domain waveform is important. An allpass filter is just fine as long as it is comparable to the crossover type in question. Indeed, some crossovers are referred to as allpass types. That old journal article, I don't have the reference handy, found similar results to yours, that it was impossible to hear with speakers in typical rooms with reflections, but that through headphones with certain limited source material it was audible but very difficult to hear. I'd say that your findings are valid, however we must keep in mind that it is generally impossible to hear in most rooms. I'll take a linear phase design if it does not compromise other characteristics, however most do compromise polar response and system distortion.
Pete B.
When citing old JAES papers one must not forget that most often they were only testing for the perception of timbre changes with static signals and not music.
Since most recordings are made for loudspeaker reproduction (intensity stereophony) - the validity of a test using HEADPHONES for checking the audibility of LOUDSPEAKER crossover introduced allpass distortions is at least questionable.
Regards
Charles
P.S.: The latest paper i know of was in the March issue of JAES:"On the Audibility of All-Pass Phase in Electroacoustical Transfer Functions "
Since most recordings are made for loudspeaker reproduction (intensity stereophony) - the validity of a test using HEADPHONES for checking the audibility of LOUDSPEAKER crossover introduced allpass distortions is at least questionable.
Regards
Charles
P.S.: The latest paper i know of was in the March issue of JAES:"On the Audibility of All-Pass Phase in Electroacoustical Transfer Functions "
phase_accurate said:
I disagree.
What I examine in the experiment in NOT introduced by the loudspeakers but by the filter. All aspects of directionality etc are not included for this specific question.
On the contrary, I would argue that headphones is better since they have no inherent crossover frequency mid-band. The alternative would be to listen through fullrange, single driver speakers.
If phase was an issue, one would of course have to test the audibility of it through a system that does not introduce (and possibly mask) any audible differences in the test object. Even if it is possible to maka a loudspeaker that is phase accurate on axis, there is still the problem off-axis.
I am not saying that testing with loudspeakers is wrong, but I do believe that the sensitivity of such a test is much lower.
Anyway, let's not argue over this, instead I encourage everyone to build an allpass filter and try a blind test for themselves. With headphones or speakers, or why not both.
phase_accurate said:
I'm going to drop this after this post because you don't seem to read what I write and you make sweeping generalizations. They did indeed use music, and they did use speakers in an untreated room (acoustically), and the conclusion was that it was not audible. They also tried headphones where it was also very difficult to hear with nearly all music source material. Certain percussive sounds, played alone so there was no masking taking place, through headphones was the only source where it was audible.
Was this an exaustive test, no, obviously not, but it does prove that those who claim they can easily hear phase non-linearity are generally wrong.
Pete B.
phase_accurate said:
Quite a bold statement. Did they test with all of those who claim it ??????
Fun aside: What kind of speakers did they use ? Where they of the transient-accurate kind ???
Regards
Charles
You're going to pull the old golden ears card?
Too bad you are not closer here in the US, perhaps then we could do some testing.
I don't recall the speakers, Quads, B&O filler driver based perhaps.
Pete B.
I'm late to this thread here. I have four Alpair 12PW to use in a simple folded smallish open baffle.
I'm planning a pair of 1 and a half ways, using parts I already have- starting with one run full range, and the other filtered 2nd order around 250 hertz or so, to get the lower OB limit closer to meeting a subwoofer.
I'll be starting with the passively filtered one wired in reverse polarity relative to the unfiltered one, unless I run them in push pull mounting.