Okay, I can see both points - I Think. I would rather this be a discussion where each argument is stated and some form of supporting evidence cited than to get to a point where each of you is questioning the other's intellect. I really want to learn something here and I'll bet both of you can help me do that.
If you have differing opinions on something, I am okay with that, too.
On that note...
One, I can see getting resolutions of 200 Hz if conditions are right inside.
Two, if you get a gate time of 200 Hz, then your resolution from that point onward will be 200 Hz.
200 Hz is not really bad, but can hide some data or give you an incomplete image of the real response.
At least that is what I think.
My goal is to understand the off-axis response of the speaker I built. I want to determine what the off-axis response is for the woofer and ribbon tweeter around the 1.2 kHz crossover point.
I don't know if a ground plane measurement will tell me that, but I can clearly see the benefit of at least knowing the on-axis response over as much of the audible spectrum as I can get.
This is essentially my desires. It may not be the right thing to do, but I am open to any and all comments so I can learn.
If you have differing opinions on something, I am okay with that, too.
On that note...
One, I can see getting resolutions of 200 Hz if conditions are right inside.
Two, if you get a gate time of 200 Hz, then your resolution from that point onward will be 200 Hz.
200 Hz is not really bad, but can hide some data or give you an incomplete image of the real response.
At least that is what I think.
My goal is to understand the off-axis response of the speaker I built. I want to determine what the off-axis response is for the woofer and ribbon tweeter around the 1.2 kHz crossover point.
I don't know if a ground plane measurement will tell me that, but I can clearly see the benefit of at least knowing the on-axis response over as much of the audible spectrum as I can get.
This is essentially my desires. It may not be the right thing to do, but I am open to any and all comments so I can learn.
GP testing was pioneered and used by the majors for years , even those with anechoic chambers... I have used gated in the past to measure down to 800 hz , mostly focusing on mid/tweeter phase ,once you get below 800hz room reinforcement becomes an issue without an anechoic chamber.
Of course you can gate and then nearfield measure the woofer in it's pressure zone , cut and splice, difficult to get right ...
regards,
One way to get it done .....
Of course you can gate and then nearfield measure the woofer in it's pressure zone , cut and splice, difficult to get right ...
regards,
One way to get it done .....
Attachments
Last edited:
You and I disagree already on in-room EQing so no need to discuss that topic (other then Im right)
Doug first bit is ok, the "other then Im right" bit is just asking for trouble. Please don't stir the pot. Tony.
Doug20 always turn the conversation Personal when he runs out of gas...
GP will tell you on and off axis response and you will not achieve 200 hz resolution inside IMO, why not go ahead and do both and compare GP to 200 hz gated indoor response , you can post the results ...
regards,
I'm really after off-axis response at 0°, 15°, 30°, 45°, 60°, and maybe 75° from about 500 Hz and up.
I take it I just need to rotate the speaker or should I move the mic? I think the issue is going to be interesting because the speaker is going to be tilted forward. So on what axis do I rotate the speaker?
Like this:
An externally hosted image should be here but it was not working when we last tested it.
The arc with the arrows illustrates the placement of the microphone on the blue 4' by 8' sheet. The cabinet is tilted about 10°*forward and upside down.
The radius of the arc is 48", the limit of the 4' by 8' sheet.
The thing is that the relationship of the arc arrow along the blue 4' by 8' sheet changes with angle. Even though the arc is constant radius, the distance between any point of the arc and that of any driver does change.
How does this impact the measurement?
As I found out!
This post http://www.diyaudio.com/forums/mult...d-farfield-does-look-right-2.html#post2430391 was how I tried to deal with it. I don't know if it was valid or not but seems to have worked. Also good point about the resolution a.wayne I can get the gating down to 300Hz but the resolution isn't great (I believe that dropping the sampling rate helps here, higher isn't always better!) . I've always wanted to try some ground plane measurements but I'd need to get a laptop with USB sound card to do that, lugging around my hulk of a desktop to somewhere suitable (perhaps the work carpark on a weekend) would be difficult.
Loren on the question of what frequency to start the sweep, the last measurements I did I started at 100Hz, when I made it higher than that I got VERY poor results. I'm not sure why but the FR and phase were all over the shop!
Tony.
My only point is that you can remove the reflection issues in the impulse measurement by using OC703 portable walls placed around and below the speaker.
I never posted outdoor GP isnt good because its obvious it can be but its also obvious that not everyone has that controlled environment outside, not everyone has portable equipment abilites and not everyone can move speakers that can be heavy (sure I can move 200lbs speakers but why). Therefore I suggested how to help with in room measurements, you and I know that some people would rather argue splitting hairs because they think they need to be right.
Zilch, Augerpro and many others have built more 2-way waveguide designs then most (very similar XO points), I have built 6 myself. I have used HOLM with great success with all my active designs. Im lucky to have a high end custom room full of treatments so Im impulse measurements are clean out past 6ms.
The science says if you remove the reflections you can gate lower for better resolution. Getting down below 400Hz will give you ENOUGH accuracy to do what ever you want with XO point about 1KHz. You can EASILY see this if you slide the gate on HOLM. For off axis response you really are not going to care what the woofer is doing below 500Hz anyways because the room will start to dominate that response no matter what.
As for your question about 10Hz starting point. I start at 20Hz but you should know, that gating is independant of the starting point and you can move the gating easily in HOLM by just clicking on the gate line and moving. Heck, even in room the meaurement of a tweeter does not change that much from gated to non-gated if the room is a controlled room (Like mine). I can posted many, many references.
In your impulse you can expand your impulse out to see where the driver measurement ends and where room interaction starts and that is where you should set your gate point. I would also change your impulse to ms instead of cm.
btw, I own many of those TD series drivers they are pretty incredible drivers!!
Last edited:
Many successful speaker designs are done from in room measurements. I seem to remember even Dr. Geddes does in room measurements with his designs. Hmmm...he cares about accuracy and his XO is below 1KHz.
So there is a mountain of experience out there that says done right in room measurements are just fine. IMO there is no need to acheive 200Hz resolution with this XO point. Nothing will change if that resolution exists or does not exist.
When finding the largest reflection free zone, you'll probably decide to place the speaker and the mic at half of the ceiling height. The reflection free time will improve with the mic closer to the speaker, but too close is naturally not a good thing. Once you decide the closest you're willing to go, work out the floor/ceiling bounce time, which should be the same, and put the speaker in one corner far enough to make the reflections at least the floor/ceiling time. Leave at least that space around the mic as well. 500Hz isn't usually a problem.
hmm I'm wondering if we are all on the same page with respect to the 200Hz resolution... Are we talking about the lowest frequency usable in the gated measurement, or the minimum step frequency across the entire measurement range? Probably my complete lack of understanding showing
Tony.
Yup, and that is my question, too. I think it may be both. At least that appears to be how the software programs handle it.
If you manually increase the gate time the displayed resolution of the data is reduced.
I'm not sure why that would be the case, unless you think there are some diffraction-related effects. The woofer is a monopole, so it will basically be omni at ~500Hz, and the tweeter will not be significantly contributing to the overall response.
I agree with Penn. Characterize the off-axis response from maybe an octave below the nominal woofer-tweeter Fc (i.e. 500-600Hz) using either gated or ground-plane measurements. Which one shouldn't really make a difference, as long as they're done well. I would pick based on convenience and practicality, which for me means indoors almost every time. Below that point, take spatially-averaged measurements in your room, because after all what matters is not how the upper bass and below looks outside, but how it sounds in your specific seating positions with the speakers placed as you intend to place them.
I Return!
Sorry for the long delay, but I have been so busy lately and I still am.
I did a quick on and off-axis set of plots (three). Holm does not display more than 3 at a time and I am having such a horrible time getting half-way usable measurements in my room that it makes the process complex.
Getting measurements outside is not a good option as it is very noisy right now. Most of it appears to be insects and a steady Florida breeze.
My final attempt was to lay the speaker on its back on a 15" stool in the middle of the room. Bear in mind this is a typical Florida house with ceramic tile floors everywhere. The mic is moved above the speaker in a 41" arc and directly in-line with the Beyma ribbon tweeter.
I took 3 horizontal dispersion readings (0 degrees on axis, 30 degrees, and 60 degrees off-axis). My goal is to get a feel as to how bad the TD15h is off axis at the crossover point of 1.2 kHz.
The actual data is better than expected. I had expected a huge hole at the crossover, but it was not as horrible as I surmised. Your opinions may be different.
I have no explanation as to the suck-out at 12 - 14 kHz is from, but it is way above the crossover frequency. Maybe it is diffraction?
Anyway, please feel free to comment on the data or the procedure used to get the data. I'll answer all questions I can.
Sorry for the long delay, but I have been so busy lately and I still am.
I did a quick on and off-axis set of plots (three). Holm does not display more than 3 at a time and I am having such a horrible time getting half-way usable measurements in my room that it makes the process complex.
Getting measurements outside is not a good option as it is very noisy right now. Most of it appears to be insects and a steady Florida breeze.
My final attempt was to lay the speaker on its back on a 15" stool in the middle of the room. Bear in mind this is a typical Florida house with ceramic tile floors everywhere. The mic is moved above the speaker in a 41" arc and directly in-line with the Beyma ribbon tweeter.
I took 3 horizontal dispersion readings (0 degrees on axis, 30 degrees, and 60 degrees off-axis). My goal is to get a feel as to how bad the TD15h is off axis at the crossover point of 1.2 kHz.
The actual data is better than expected. I had expected a huge hole at the crossover, but it was not as horrible as I surmised. Your opinions may be different.
I have no explanation as to the suck-out at 12 - 14 kHz is from, but it is way above the crossover frequency. Maybe it is diffraction?
Anyway, please feel free to comment on the data or the procedure used to get the data. I'll answer all questions I can.
An externally hosted image should be here but it was not working when we last tested it.
I think the issue with that is blankets will probably absorb some of the high frequencies, but will be pretty transparent to bass, which is probably the primary culprit that is causing the high gate times.
I have some 2' by 4' panels of 3/4" ridged fiberglass insulation that I might try stacking up as one mechanism to absorb some of the reflected sound.
Thanks for the compliment on the woodworking.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.