(edit: I should have said "I am trying to measure them" in the title, because that's what I meant. There are now measurements posted later on in this thread, and there will be more.)
Much discussed, rarely measured. I've totally stumped myself measuring these, as I'm getting a mystery reflection 2.45ms out in the impulse, exactly the same every time, seemingly regardless of whether the woofer is on a baffle, in a box, in free air, near a boundary, or whatever you like. I am a total beginner at driver measurement, but it's only these woofers giving me this weirdness - not other drivers. If there's anyone out there that wants to compare some graphs here, please speak up! Thanks.
Much discussed, rarely measured. I've totally stumped myself measuring these, as I'm getting a mystery reflection 2.45ms out in the impulse, exactly the same every time, seemingly regardless of whether the woofer is on a baffle, in a box, in free air, near a boundary, or whatever you like. I am a total beginner at driver measurement, but it's only these woofers giving me this weirdness - not other drivers. If there's anyone out there that wants to compare some graphs here, please speak up! Thanks.
Last edited:
Too funny as I have often thought the same thing.
I have a pair coming but can't help you yet.
So, if you sweep the TD and then change nothing else but the driver the mystery blips gone?
Yes, it's just this driver (both of them - I have a pair). I have posted the details elsewhere, and I will copy them over here when I'm feeling less lazy (so, tomorrow).
I recall there being an issue with the cone on the 15M ringing that would show up that way---you should be able to dig it up the relevant DIYA thread with some searching. Is this MLS or swept sine/chirp measurement?
Hi there
Have a look at the thread 'Is the AE TD15M the ultimate mid driver?'. There are several measurement graphs posted there.
Enjoy,
Deon
Have a look at the thread 'Is the AE TD15M the ultimate mid driver?'. There are several measurement graphs posted there.
Enjoy,
Deon
It shows up in either type, or at least I'm 99% sure it does. I will see if I can find that thread.
Those are mine.
Well, I couldn't find the thread you are talking about. To clarify about MLS vs log sweep, I saw the issue in a long log sweep measurement, and then switched to MLS for a bunch of fast measurements trying (and failing) to eliminate the issue. Unless there was some massive coincidence where the log sweep had a normal reflection in the same spot, it is the same for both types of measurement.
If you are showing me that thread because there is something in it about this, you'll have to be more specific because I didn't see it. I have several posts in that thread, and I am the only one that posted measurements in that thread (I think).
Dumptruck,
MLS and sweptsine (or chirp) should show the same as far as real reflections go. But MLS can sometimes show phantom impulses from nonlinear distortions, while log swept sine is immune to that kind of phenomenon (with log swept sine, phantom impulse occur before the real impulse, so they don't get into the result). So if the second impulse shows with swept sine, it is probably real.
It isn't a reflection off your mic stand, is it? The TD15M has a lot of HF output, so it will show impulse peaks more sharply than other big drivers, and its size gives it lots of opportunity to excite nearby objects or surfaces. Look for things that are 14 to 24 inches from your mic or the woofer.
MLS and sweptsine (or chirp) should show the same as far as real reflections go. But MLS can sometimes show phantom impulses from nonlinear distortions, while log swept sine is immune to that kind of phenomenon (with log swept sine, phantom impulse occur before the real impulse, so they don't get into the result). So if the second impulse shows with swept sine, it is probably real.
It isn't a reflection off your mic stand, is it? The TD15M has a lot of HF output, so it will show impulse peaks more sharply than other big drivers, and its size gives it lots of opportunity to excite nearby objects or surfaces. Look for things that are 14 to 24 inches from your mic or the woofer.
Last edited:
You can sometimes set up your measurement to repeat and move a pillow around the area and figure out what direction a discrete reflection is coming in from.
If you have a reflection in the measurement at 2.45ms you should also have a corresponding first cancellation notch in the frequency response at about 200 Hz.
Here is a copy of what I posted on AudioKarma about this:
(minor note: not that it matters, but you may notice the time window on the graph says 2.417ms - that's because I accidentally bumped the start line up a tiny bit before I grabbed the image)
Well I'm just absolutely mystified here. Not exactly the first time in this project, but still... Below, you can see I have marked the two reflections I am trying to figure out. Neither occurs with other drivers. I am assuming them both to be reflections because they introduce a regular ripple to the response right when the gate approaches their peak.
The first, I suspect is cabinet-related. Either baffle diffraction, or a reflection from inside the box, or something. The 2nd is the real puzzle for the moment. Check out the notes of what I just did below. How can this be?? "Near" means about 30", "far" means 1-2m as the location allowed. Everything on-axis.
speaker on floor pointing up, mic near or far
1: 1.7ms
2: 2.45ms
on stand in middle of room facing forward (~halfway between floor and ceiling), mic near or far:
same
woofer in free air on mattress on floor pointing up:
1: gone
2: 2.45ms (same)
woofer in free air on two paint cans above cement pointing up:
1: gone
2: 2.45ms
speaker flat on floor facing forward (mic ~9" above floor), near:
1: would not be able to find in this impulse, new nulls at 625Hz & 2.2kHz, to early to identify by gating
2: 2.45ms
speaker flat on floor facing forward, far:
1: gone
2: 2.45ms
speaker on floor, angled up 20 degrees, ~1m
1. 1.7ms, but distorted by new reflection(s) at almost same time
2. 2.45ms
When I say 2.45ms every time, I mean EXACTLY the same every time. 😕
(minor note: not that it matters, but you may notice the time window on the graph says 2.417ms - that's because I accidentally bumped the start line up a tiny bit before I grabbed the image)
Well I'm just absolutely mystified here. Not exactly the first time in this project, but still... Below, you can see I have marked the two reflections I am trying to figure out. Neither occurs with other drivers. I am assuming them both to be reflections because they introduce a regular ripple to the response right when the gate approaches their peak.
The first, I suspect is cabinet-related. Either baffle diffraction, or a reflection from inside the box, or something. The 2nd is the real puzzle for the moment. Check out the notes of what I just did below. How can this be?? "Near" means about 30", "far" means 1-2m as the location allowed. Everything on-axis.
speaker on floor pointing up, mic near or far
1: 1.7ms
2: 2.45ms
on stand in middle of room facing forward (~halfway between floor and ceiling), mic near or far:
same
woofer in free air on mattress on floor pointing up:
1: gone
2: 2.45ms (same)
woofer in free air on two paint cans above cement pointing up:
1: gone
2: 2.45ms
speaker flat on floor facing forward (mic ~9" above floor), near:
1: would not be able to find in this impulse, new nulls at 625Hz & 2.2kHz, to early to identify by gating
2: 2.45ms
speaker flat on floor facing forward, far:
1: gone
2: 2.45ms
speaker on floor, angled up 20 degrees, ~1m
1. 1.7ms, but distorted by new reflection(s) at almost same time
2. 2.45ms
When I say 2.45ms every time, I mean EXACTLY the same every time. 😕
Attachments
Last edited:
sorry! i wasn't attentive enough. this is not the one that
dumptruck,
you're not the first individual that got puzzled by this wiggle at 2.5ms...
... look here :
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-691.html#post2258879 (Gary Dahl describing the phenomen)
After some long and interesting talk with much speculating of what it might be (I was a top speculator there and erred big time 😉), "John.K" and and "Few" arrived at this conclusion of what is the most likely root cause :
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-176.html#post2265030 (impulse response)
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-704.html#post2265073 (frequency response and explanation).
you're not the first individual that got puzzled by this wiggle at 2.5ms...
... look here :
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-691.html#post2258879 (Gary Dahl describing the phenomen)
After some long and interesting talk with much speculating of what it might be (I was a top speculator there and erred big time 😉), "John.K" and and "Few" arrived at this conclusion of what is the most likely root cause :
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-176.html#post2265030 (impulse response)
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-704.html#post2265073 (frequency response and explanation).
dumptruck, in Arta you can put the marker at the start of this 2.45ms sub impulse and see what the frequency response looks like down there. Maybe this will give you a clue about its origin.
Thanks so much for the links KSTR. I had looked at that thread earlier but didn't find the relevant posts.
pos, I would be happy to try anything suggested tonight, although it looks like the links above may have it covered pretty well - I'll have to read them a bit more.
pos, I would be happy to try anything suggested tonight, although it looks like the links above may have it covered pretty well - I'll have to read them a bit more.
Okay, done reading. Anybody have thoughts on John K's conclusion (in what seems to be the most definitive post on this in the thread) on just crossing over well below this breakup? I kind sorta planned on 1.4k+... Do I see what living with it is like, try to filter it...? http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-706.html#post2266386
So it's a mechanical resonance inherent to the TD15M that shows up in the impulse response at 2.45ms (or there abouts)? Very interesting.
- Status
- Not open for further replies.