What do you actually mean by phase coherence? Please explain because it may vary.
so how are you measuring all of this and what evidence is there? Lets not forget its very easy to fool ourselves of an audible change. Consider cables for example.
Agree the speaker still needs to be properly designed, meaning it's accoustically compact in the crossovers' transitione bands and time aligned. I'd also point out planars are arguably the only phase coherent drivers. And, even then, only on axis and only at frequencies where the path length difference over the planar's surface is small. If one uses the usual rule of thumb something must be 1/10 of a wavelength or smaller to be acoustically small that's still a 36 degree phase shift from the near edge of the planar to the far. For cone woofer or dome tweeters there's additionally the front to back depth of the driver to consider. Also, every driver I've measured has had the phase shift around Fs which one would expect from a resonance so I'd be surprised if there are any full rangers which measure out at flat phase; the advantage to full range here is it's only one phase shift to deal with.
I find that when listening critically I can hear differences in phase coherency on the order of 10 or 15 degrees across a few hundred Hertz of bandwidth, as best as I can approximate a blind A/B. So a more desirable design target would be 1/100 of a wavelength. On that definition there's no such thing as a point source speaker---at 100Hz and standard temperature and pressure you'd need a driver no bigger than 34mm across. Multi-way and full range are just different implementation tradeoffs within that constraint, with full range typically being more compact than multi-way within the crossover transition band but often considerably larger above it. So, personally, I tend to see it as a tradeoff between implementation complexity and attributes desirable for upper midrange or tweeter operation such as avoidance of cone breakup, reduced moving mass, lower inductance, and greater dispersion.
I do find the OEM three inch full rangers in my car subjectively preferable to just about every box speaker I've ever listened to under $1500, which is both a bit ridiculous and a testimony to how much complexity multi-way adds. Personally, for a main system I find a digital, linear phase crossover's worth it for the benefits it offers in driver selection and equalization capability. Every other project I'm working on is full range, though.
Yup.
There is an easy thot experiment one can do (this borrowed (paraphrased) from a paper by a B&K engineer).
Take a speaker system with a tweeter. Move the tweeter back by 30 km. Play a 60 sec selection of music. The music ends with silence for 40 sec, and then the treble arrives and plays for 30 sec.
Obviously it makes a difference. The question is how much?
Kunchar's research (duplicated independently, but not really aimed at answering this question) indicates on the order of 2mm max (5 uSec) at 7kHz.
dave
Take a speaker system with a tweeter. Move the tweeter back by 30 km. Play a 60 sec selection of music. The music ends with silence for 40 sec, and then the treble arrives and plays for 30 sec.
Obviously it makes a difference. The question is how much?
Kunchar's research (duplicated independently, but not really aimed at answering this question) indicates on the order of 2mm max (5 uSec) at 7kHz.
dave
well, dont think we all hear the same, or listens to music the same way, because we dont
let me put it this way
why do you think there are music production producers who are able to tell exstremely skilled mucisians how they should play their music
what do you think it is they hear
yes, I can hear even the smallest phase issues quite clear
none of my friends hear nothing of that
severe phase issue makes me creep, cant listen to it
my friends are able to listen to it all day long
and they do
many do
but one thing is certain, they like what I do with speakers
is that evidence enough
they are by far not even close to being enthusiasts, just ordinary simple people without prejustice, or any knowledge about golden ears
I would think that makes their opinions valid, in a strange twisted way
blindtests ? well, I could give them sunglasses, and they would still say the same
really, just because we dont hear it, doesnt mean its not there
and ofcourse, then you could in some sense rightfully claim it doesnt matter
but when the magic happens, everyone can hear it
and thats the point
if you want better speakers, then phase matters
Im not sure, but I guess those who have expertice in designing phono/riia amps might be the ones who know about phase issues
let me put it this way
why do you think there are music production producers who are able to tell exstremely skilled mucisians how they should play their music
what do you think it is they hear
yes, I can hear even the smallest phase issues quite clear
none of my friends hear nothing of that
severe phase issue makes me creep, cant listen to it
my friends are able to listen to it all day long
and they do
many do
but one thing is certain, they like what I do with speakers
is that evidence enough
they are by far not even close to being enthusiasts, just ordinary simple people without prejustice, or any knowledge about golden ears
I would think that makes their opinions valid, in a strange twisted way
blindtests ? well, I could give them sunglasses, and they would still say the same
really, just because we dont hear it, doesnt mean its not there
and ofcourse, then you could in some sense rightfully claim it doesnt matter
but when the magic happens, everyone can hear it
and thats the point
if you want better speakers, then phase matters
Im not sure, but I guess those who have expertice in designing phono/riia amps might be the ones who know about phase issues
Phase may be just something one would have to specifically train the ear to identify.
If someone heard a loudspeaker with poor phase coherence they may be able to tell something is wrong but it would be very hard to say exactly what. Within reason of course... Phase shift is not something we are likely to experience in our natural day-to-day life.
If someone heard a loudspeaker with poor phase coherence they may be able to tell something is wrong but it would be very hard to say exactly what. Within reason of course... Phase shift is not something we are likely to experience in our natural day-to-day life.
Before I had test gear I built some OB speakers for my brother (and family!
).They were sort of a line source, with one of the mids (FR) driver near the top of a baffle about 1 meter high.
They were a bit dull, and I had a couple of Linaeum dipole tweeters laying around. I high passed them with a single cap. By calculation should have been around the 8 kHz mark.
The baffle top was thick enough that I had a bit of freedom to move the tweeter forwards and back.
No test gear, so I simply asked everyone to tell me where it sounded best.
There was consistent agreement to a position plus or minus about 3 mm (over a range of about 8 cm).
This approximately coincided with voice coils lining up.
FWIW, EVERYONE was in agreement.
That was funny! You are talking about getting the phase alignment point correct at a crossover point. If it's not right you are going to have a null or ripples in the FR response that you will hear. You can "hear" the phase differences becasue they are generating frequency response anomalies.
I don't see how that is an example of the audibillity of phase. I don't think anyone would argue that it has to be "correct" so it sums properly through the crossover region. Just becasue it sums properly doesn't mean the driver phases are the same. Just because the voice coils are aligned does not mean the phase is aligned.
Rob
I am a firm believer in striving to achieve linear phase, particularly at frequencies below 2kHz.
For one thing, in the most basic sense, steep crossovers discard musical information - i.e. the sum of low and highpass filters does not equal the input. That is a fundamental issue that is at odds with the goal of high fidelity. Basically 1st order filters are the only passive way to achieve it.
I have found that phase issues are most noticeable and objectionable in the range of the human voice, say below 1 kHz. Vocal coherency suffers with poor phase response in that region.
For one thing, in the most basic sense, steep crossovers discard musical information - i.e. the sum of low and highpass filters does not equal the input. That is a fundamental issue that is at odds with the goal of high fidelity. Basically 1st order filters are the only passive way to achieve it.
I have found that phase issues are most noticeable and objectionable in the range of the human voice, say below 1 kHz. Vocal coherency suffers with poor phase response in that region.
Tinitus all this talk of magic is subjective.
can someone provide a clear explanation of phase and time coherency? These are words with no meaning and we might not be talking about the same thing.
Don't forget too, just because we hear it doesnt mean its there.
Which isnt consistent with this statement:
Surely it ought to stick out if it's not typically enountered?
The trouble with this experiment is that no speaker has such distances involved. It's like saying you will hear the effect of jitter if you increase it to very high levels. It's not meaningful.
can someone provide a clear explanation of phase and time coherency? These are words with no meaning and we might not be talking about the same thing.
So what about all the high end manufacturers who dont use 1st order and dont have phase/time alignment as a design feature? LIke B&w, wilson audio, etc
Many people praise the sound quality and they get good reviews. What do we make of that? Being at odds with hi fidelity wouldnt matter if it wasn't audible.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.