Jump to content

so, i created a 3ms time delay convolution (did this at 400hz for the scala)


tofu
 Share

Recommended Posts

you delayed everything below 2000hz? i thought

the bass bin was already delayed... why delay it more? wouldn't

the point be to delay the high frequencies in order to make them

aligned with the low?

The point of the sound sample was to simulate the effects of

time-alignment...not an attempt to correct for it. I wouldn't suggest

trying to correct for time delay by modifying the waveform because you

will have to figure out the exact transport function of the speaker

you're connecting to and will have to somehow create the exact

opposite, which is essentially an impossible task. Any close

approximations will result in some nasty effects and probably even

sound worse.

Link to comment
Share on other sites

  • Replies 136
  • Created
  • Last Reply

Top Posters In This Topic

Dr Who,

Your experiment applying

different delays to the right and left channels of a stereo

"program" is not a fair test. You must mix the inital and delayed sound

into the SAME channel and listen in mono. Your are now using the

brains ability to perceive phase difference between two ears if you use

two channels. Your ear/brain "team" definitly CAN hear that difference.

That is how it computes a stereo image. What the ear can NOT hear is

phase differences bewteen to seperate sounds with the same

ear. That is, the phase relationships of the components

(harmonics) of a complex waveform. Remember too that there is a point

where phase differences become so large that they actually become two

different sounds in time. At that point, you definitely can again

hear the difference. This is the point where the classic tap dancer

case makes the audable echo. When the phase is on the order

of a few hundred degrees (I expect) or less, you can't hear it. PWK did

testing to demonstrate that the path difference between his woofer and

midrange were below this point where the delay difference could be

heard. I personally have done tests with 3 seperate signal generators

to simulate the harmonics of a square wave. The generators, not being

phase locked to each other, created a sontinuously changeing waveform.

I looked at the waveform on an oscilloscope and listened to it (in

mono) on a set of headphones at the same time. Even when looking at the

scope for visual clues I could hear no changes in the sound. It was a

continuoius raspy tone. I know from this that I (at least) can NOT hear

the phase relationships of a complex (square) waveform.

Al K.

If you read more carefully you will notice that I converted the stereo signal to mono first!

I even applied the same crossover to both sound samples so that

there could be no differences between a slightly different EQ (in case

I didn't get the crossover perfect).

Link to comment
Share on other sites

If the two bursts were of

different frequencies this would not happen. The phase between the two

would then be random and inaudable. Actually,

there will be different points in space where you get perfect

cancellation, not to mention you will also have combinations tones present. The two bursts would have to be vastly seperated

in time before you could perceive that they were two seperate bursts

(an echo).

Modifying the phase of one of the bursts will change the amplitude and

phase of the combination tones. It has to be audible because these

tones are interacting with the rest of the music.

Link to comment
Share on other sites

you delayed everything below 2000hz? i thought

the bass bin was already delayed... why delay it more? wouldn't

the point be to delay the high frequencies in order to make them

aligned with the low?

The point of the sound sample was to simulate the effects of

time-alignment...not an attempt to correct for it. I wouldn't suggest

trying to correct for time delay by modifying the waveform because you

will have to figure out the exact transport function of the speaker

you're connecting to and will have to somehow create the exact

opposite, which is essentially an impossible task. Any close

approximations will result in some nasty effects and probably even

sound worse.

ohh, i thought your sample was an attempt to correct it, hence my crusade to get my time delay convolution working.

sigh...

Link to comment
Share on other sites

What is the frequency of an "A?" Obviously it is in waves/sec. Graph the "A" and graph another "A" on top of the old "A" but adjusted to the right by 3/1000 of a second. The intersections create a new tone.

(no, they create the same tome, but a new amplitude)... Sorry to be blunt, but it's apparent that you don't even understand the most fundamental concepts of wave mechanics...something everyone in Illinois is supposed to learn in high school physics.

Okay, you missed my corrections to my own poor math. Actually, my math is pretty good, it's just going that far back when we studied this takes me a while to get my bearings straight.

I think there's something that both you and I missed so far (so, you're probably not in Illinois, either).

Look at my quote (actually at the later post I made to correct the earlier post, where I said later, the tone's the same), and then, look at your response. You agreed with me that the overlaying creates the same tone, but a new amplitude. Not..... Better go enroll at Illinois U.

I went ahead and physically plotted the "delay thing" using pen and paper. If you will recall from physics, you have 2 sources plotted (being the 2 "A" notes in our example). You will also recall your teacher telling you that the two waves create a RESULTANT wave, whose fequency and amplitude is detemined by the intersections of the 2 sources. Yep! A 3rd wave.

I'm not embarrassed at all to be wrong. I like working through problems, and if there's not folks out there like you to put in their 2 cents, you miss out on some things. You gave rise to a challenge, but YOU and I were both wrong on a point upon which we agreed.

I still question how audible the effect really is in terms of whether or not I "like" the way a song sounds through speakers.

Finally, I have no personal need to get defensive regarding my speakers. Like you said, they are great speakers. Believe me, I have WAY more going on in my life than to define my self-worth by a pair of speakers I own.

I simply found the theory being discussed to challenge my memory of physics to see if I would agree with you. I am not at the point of agreeing, yet, but I will agree both you and I were both wrong for not taking into account the resultant wave that occurs from the intersection of 2 waves spaced apart.

Link to comment
Share on other sites

Dr Who,

"Modifying the phase of one of the bursts will change the amplitude and phase of the combination tones. It has to be audible because these tones are interacting with the rest of the music."

NO! The two bursts I show are the SAME burst and therefore the same frequency. The phase between them is shifted, but remains constant over time. The "rest of the music" is of DIFFERENT frequency and therefore are of random and continuously changing phase. Only an oscilloscope can compare the rapidly changing phase of two different frequency signals if you set it up to triger on just one of the two. The human ear/brain isn't fast enough to do that! It can only compare the phase of the same singal between two ears, not one! A single ear can only percieve amplitude changes of a signal or signals.

Al K.

Link to comment
Share on other sites

Al, buddy, I think you and I were wrong. But you made a valid point about the "noise" that the brain discards.

Now that this "seems" (which I say very carefully) to be coming back to me. Music is a harmony of frequencies that we find pleasing. Among the music are all sorts of frequencies that, by themselves, can be heard. But, since we prefer to filter in only the combinations that please us, we subconsciously discard the waves that don't. These are those resultant waves that don't fit in.

But... If this were the end of the story, then, every person would say every speaker was the same. I might enjoy those Blose more, like DrWho said.

But, I'll go out on a limb (which I bet is pretty stout limb) and say that we naturally do not filter out all "noise." Some of us will filter out different "noise" frequencies, depending upon way too many factors that to consider them, would leave us all in agreement that the whole listening process is subjective. It is objective to a limit, but then goes subjective.

That's why the example of the TV going and the wife talking at the same time applies. If you're focused on the TV, at least some PART (if not all - [:D]) of the wife gets filtered out. Conversely, if you're focused on the wife, at least some part of the TV gets filtered out.

It all depends upon what the BRAIN wants to filter out.

That said, remember that this went back to the "Eleanor Powell two-taps" issue where it was said a sound engineer (maybe not a PhD. engineer - I do not know the man's credentials) supposedly heard two-taps where there was one physical tap by Mrs. Powell. This was alleged to have been heard by the alleged engineer-guy based upon a sound delay between the HF and LF of 7 ms. I still would contend, this far, that nobody.... and I mean nobody... could ever discern 2 taps where there was only 1, by only a 7 ms delay between LF and HF.

Link to comment
Share on other sites

Dr Who,

"Modifying the phase of

one of the bursts will change the amplitude and phase of the

combination tones. It has to be audible because these tones are

interacting with the rest of the music."

NO! The two bursts I

show are the SAME burst and therefore the same frequency. The phase

between them is shifted, but remains constant over time. The "rest of

the music" is of DIFFERENT frequency and therefore are of random

and continuously changing phase. Only an oscilloscope can compare the

rapidly changing phase of two different frequency signals if you set it

up to triger on just one of the two. The human ear/brain isn't fast

enough to do that! It can only compare the phase of the same singal

between two ears, not one! A single ear can only percieve amplitude

changes of a signal or signals.

I was saying two bursts of a different frequency...modifying the phase

or amplitude of one of the tone bursts will change the phase and

amplitude of the resultant combination tones (of which there are

two...so we have 4 sounds total, not 3).

Link to comment
Share on other sites

[sorry to be blunt, but it's apparent that you don't even understand the most fundamental concepts of wave mechanics...something everyone in Illinois is supposed to learn in high school physics. So these misconceptions are hardly those of "common-sense" considering most people are beyond a high school education.

Apparently, you forgot your high school wave mechanics class, like I did. After you digest my follow-up, do you agree that you were wrong, as I was? I think I am right, now. Wondering if you still think I'm wrong. I'd like to see if someone could finally get an "answer" we all agree with.

BTW: Be careful not to undermine common-sense by relating it to higher education. I got my JD from the University of Texas in 1993. It didn't help me much with wave mechanics, though. So, I never jumped out and said "look at my degree." Common sense=logic=scientific approach.

Link to comment
Share on other sites

You will also recall your teacher telling you that the two waves create a RESULTANT wave, whose fequency and amplitude is detemined by the intersections of the 2 sources. Yep! A 3rd wave.

Quoting myself for the purpose of admitting, yet, that I might be wrong again! Is it the intersections of the 2 that define the resultant 3rd? Or is it the mathematical SUM of the 2 that defines the resultant 3rd?

In any event, you still have the 3rd, but depending upon how the 3rd is defined (by intersections or the sum), it could make a whole lot of difference in the conclusion.

Anybody know how the resultant 3rd gets defined?

Link to comment
Share on other sites

Dr,

"I was saying two bursts of a

different frequency...modifying the phase or amplitude of one of the

tone bursts will change the phase and amplitude of the resultant

combination tones (of which there are two...so we have 4 sounds total,

not 3)."

HUH? Run that buy me again! 2,3, 4 bursts, now I'm totally confused!

Al K.

I think I'm confused now too...could you describe again what the pictures you plotted were measuring?

I was trying to say that two different frequencies playing at the same

time generate their own sum and difference tones...if we change the

phase or amplitude of one of the two original frequencies, our sum and

different tones are going to change in amplitude and phase. I was under

the impression that you were saying this wasn't the case...

Link to comment
Share on other sites

Who,

I think I see what your thinking. You are confusing comb anomolies with intermodulation distortion. If two different signals pass through a nonlinear system the sum and difference frequencys are generated. This does not happen in a good linear low distortion system. Two tones remain just two tones. If they interact it's only to add together algebraically. This is what the two photo plots show but of tones of the SAME frequency. They are actual oscilloscope screen pictures taken of the output of a microphone a few feet in front of my Belles. The two bursts were generated with a Wavetek function generator gated by a pulse generator. They are actually the SINGLE burst shown on the bottom trace on moth pictures. The first burst in time is from the tweeter, which is closer to the mike than the squawker, which is the source of the second burst (top trace of each picture). If the duration of the burst is made longer than the delay difference between the driver (the bottom picture), the tones overlap. The overlap time shows the two add together. The is only a single frequency, you simply see it twice in time. If you move the mike a bit you can find a spot where the overlap time actually cancells to nothing!

Al K.

Link to comment
Share on other sites

DrWho, Al.... I surrender. It appears the resultant is the sum, so pitch does not change. I'll leave this alone, but I find it very interesting. Still, can you demonstrate the claimed "two-taps" from only a 7ms delay? From what it appears, the amplitudes are summed, and their resultant peaks are known as "beats." So, DrWho, with your uploaded samples, it seems the difference is in these beats. The only thing is they are so close we do not hear it like "beats" in laymen's terms. Instead, we just notice something a little "different" - whatever the right term might be (out of phase?). But really.... 2 taps of Mrs. Powell's tappies? Can you reproduce what TBrennan said about Powell's 2 taps from a mere 7ms delay?

I watch you guys play it out. You are ahead of me for sure.

Link to comment
Share on other sites

Jeff,

Yep, that nails it when you said "beats"! That is intermod distortion, NOT time delay distortion! To know if 7 mSes is significant you would need to use the speed of sound to know how far sound travels in 7 mSec. Then you need to know how far apart the drivers where in the theater where all this tap dancing was being reproduced. I'm to blasted lazy to do the math!

Al k.

Link to comment
Share on other sites

Jeff,

Yep, that nails it when you said "beats"! That is intermod distortion, NOT time delay distortion! To know if 7 mSes is significant you would need to use the speed of sound to know how far sound travels in 7 mSec. Then you need to know how far apart the drivers where in the theater where all this tap dancing was being reproduced. I'm to blasted lazy to do the math!

Al k.

Al, thanks for allowing me for being all over the place while trying to figure this out for an hour or so. You are a gentleman and a scholar!

Would like to hear more from you to determine whether this makes any difference at all between folded horn and direct-radiating.

Link to comment
Share on other sites

Jeff,

Yep, that nails it when you said "beats"! That is intermod distortion, NOT time delay distortion! To know if 7 mSes is significant you would need to use the speed of sound to know how far sound travels in 7 mSec. Then you need to know how far apart the drivers where in the theater where all this tap dancing was being reproduced. I'm to blasted lazy to do the math!

Al k.

Actually, Al, I think you answered my question, and I overlooked it. You're saying it's intermod distortion. So, I guess I get a little more intermod from KHorns than Corns because of the folded woofer.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share


×
×
  • Create New...