Sanity check for first measurement

[sandyh]

Following are a few screen shots of my first measurement. Equipment is as follows:

Laptop with Core 2 Duo processor
M-Audio Mobile Pre
EMC8000 mic (circa 2005 - uncalibrated)
Outlaw Audio Receiver
BR-1 Speaker with Dennis Murphy crossover

I am just getting back into the hobby a bit and with the large amount of software that wasn't available when I took a hiatus, I thought I'd give it a try again. I tried Holm impulse in this example and wanted a quick sanity check to see if my results made sense given my measurement set-up.

I gated the far response right before the wiggle in the impulse window and I used the ungated data for the near response (blue and red). I stitched near to far at 300hz with a 25 hz span with the match phase option selected. The result is the green trace.

I have attached pictures of the measurement set-up as well. The overall picture shows the general area. I happened to be measuring the small speaker on top at the time, but the same arrangement was made for the far measurement of the BR-1 with the mic lower. The second picture is the near measurement of the BR-1.

Observations:

I had to have the output volume of the mobile pre most of the way up, mic gain most of the way up and receiver volume most of the way up to get to 78db. Seemed way too much to me. . .
The wiggles on the right side of the far response were an artifact from something. I re-ran the measurement and it looked more flat, but I didn't save it, apparently.
The room set-up was far from ideal, but I am not sure how much impact (or what things are the most impact) on the measurement quality.

Any thoughts would be appreciated. I would like to repeat the measurements soon with whatever improvements are suggested.

Thanks!

Sandy.

[rpb]

Turn up the signal output in the HOLM software if you didn't already.



Here's some more HOLM information.

http://techtalk.parts-express.com/sh...-starter-guide

[sandyh]

Turn up the signal output in the HOLM software if you didn't already.



Here's some more HOLM information.

http://techtalk.parts-express.com/sh...-starter-guide

Thanks for the reply. Actually, your tutorial is the only reason I've made it this far. I did the loopback/calibration last weekend and set the output low as you recommended, but I never changed it to a higher value when I did the actual measurements a week later.

I'll go back over the tutorial again to see if I missed something else. Is there anything that is obviously right or wrong with the measurements I posted?

Sandy.

[rpb]

Everything looks fine to me. I would move the speaker off of the BR1 though. It may have a small effect on the response. It would be interesting to see the effect of the grill when you put it back on. Be careful when you turn the level up. On my setup, I can exceed 100 dB easily. You can select a smaller impulse window, and it will be easier to see. Your gating is around 3 or 4, there's no need for the window to show much past ten. Try auto zoom after reducing the window.

If you set the speaker stand on a 24" box, and move the desk about 5 feet behind the mic, you can gate even lower. If you wanted to.

[sandyh]

Well, I re-read the Holm Starting guide thread and made some changes to the set-up. I moved the table, put the speaker on a ladder and added some pillows to the floor and a blanket to the ladder. I made a bunch of measurements at 1meter (tweaking pillows vs no pillows, blanket vs no blanket etc). I made at least 10-15 measurements, many in the same configuration and they were all very similar (not the same, but minor variances). I then decided to measure another speaker. That's when things got strange.

I switched to the second speaker (the other one of the pair) and ran a measurement. Nothing came out. I made sure I tuned the receiver back on (I did) and tried a few more times. I closed out of Holm and restarted. Same thing, no sound - but Windows did make a 'boing' sound when I closed. I played around some more, trying to figure out what had happened and during that process, I clicked the check box for the right channel output as well. Sound! I never switched wires anywhere and I swear I had been using the left in and out channels only, but with the right clicked I got a sweep, but with only the left checked I didn't. With both left and right, I got sound and right only I got sound. See question 1 near the end.

Anyway, figuring I have accidentally clicked something somewhere when I switched the speakers, I just went back to measuring. The curve for the second speaker was way different than the first. Thinking it is possible that I botched the crossover (7 years ago. . .) I switched back to the first speaker and the curve looked the same as the previous run on the second speaker. (i.e. measurement A was the first speaker before the left/right channel issue, B was the second speaker after the left/right issue and C was the first speaker again after the left/right issue. B & C look the same and A looks very different).

Somewhat miffed, I ran a few repeat sweeps with each speaker and the results are similar for both, but very different from the measurements I got during the first run. Attached is a screen shot showing the comparison of the exact same speaker before (blue) and after (red) the output issue.

So, questions:

1: Has anyone ever had an issue where they lost an output channel with Holm? Was it software or hardware?
2: Does the DAC-ADC calibration occur on only 1 channel (i.e. I loop calibrated left out to left in)? By switching to the right out, did I eliminate my loop calibration? If so, that could explain the differences of measurements from before and after the switch.
3: Unrelated to the problems I had, where would you set the gating based on the screen shot? When I did my first measurements last week, there was a pretty obvious wiggle in the impulse window around 3-4ms. Now, I don't see a wiggle with nearly the same magnitude and not much of a wiggle until 8-9ms.

I'm having a bit of fun, but I don't have much faith in my set-up yet, with the differences I see. It's repeatable as heck until the left/right output deal happened.

Any suggestions are appreciated.
Sandy.

[rpb]

Well, I re-read the Holm Starting guide thread and made some changes to the set-up. I moved the table, put the speaker on a ladder and added some pillows to the floor and a blanket to the ladder. I made a bunch of measurements at 1meter (tweaking pillows vs no pillows, blanket vs no blanket etc). I made at least 10-15 measurements, many in the same configuration and they were all very similar (not the same, but minor variances). I then decided to measure another speaker. That's when things got strange.

I switched to the second speaker (the other one of the pair) and ran a measurement. Nothing came out. I made sure I tuned the receiver back on (I did) and tried a few more times. I closed out of Holm and restarted. Same thing, no sound - but Windows did make a 'boing' sound when I closed. I played around some more, trying to figure out what had happened and during that process, I clicked the check box for the right channel output as well. Sound! I never switched wires anywhere and I swear I had been using the left in and out channels only, but with the right clicked I got a sweep, but with only the left checked I didn't. With both left and right, I got sound and right only I got sound. See question 1 near the end.

Anyway, figuring I have accidentally clicked something somewhere when I switched the speakers, I just went back to measuring. The curve for the second speaker was way different than the first. Thinking it is possible that I botched the crossover (7 years ago. . .) I switched back to the first speaker and the curve looked the same as the previous run on the second speaker. (i.e. measurement A was the first speaker before the left/right channel issue, B was the second speaker after the left/right issue and C was the first speaker again after the left/right issue. B & C look the same and A looks very different).

Somewhat miffed, I ran a few repeat sweeps with each speaker and the results are similar for both, but very different from the measurements I got during the first run. Attached is a screen shot showing the comparison of the exact same speaker before (blue) and after (red) the output issue.

So, questions:

1: Has anyone ever had an issue where they lost an output channel with Holm? Was it software or hardware?
2: Does the DAC-ADC calibration occur on only 1 channel (i.e. I loop calibrated left out to left in)? By switching to the right out, did I eliminate my loop calibration? If so, that could explain the differences of measurements from before and after the switch.
3: Unrelated to the problems I had, where would you set the gating based on the screen shot? When I did my first measurements last week, there was a pretty obvious wiggle in the impulse window around 3-4ms. Now, I don't see a wiggle with nearly the same magnitude and not much of a wiggle until 8-9ms.

I'm having a bit of fun, but I don't have much faith in my set-up yet, with the differences I see. It's repeatable as heck until the left/right output deal happened.

Any suggestions are appreciated.
Sandy.

It's hard to tell for sure, but it looks like the mic height is a little lower than the tweeter axis. That could be picking up a x-over related null. The other thing that you should verify, is which speaker sound is actually coming from. Is it possible the speaker on the ladder is not actually creating the sweep?

[sandyh]

Wow, I was just about to comment on how I'm sure which speaker was making what sound, but when I thought through all of the possabilities, you nailed it.

The ONLY logical conclusion that matches the variables and symptoms is that I was indeed measuring the speaker that was near the wall until I disconnected it. I just verified that the left and right speaker wires are crossed. . .

Man, you're good. I was intentionally sitting on the far side of the room, behind the speaker right up against the wall when I ran the sweeps so I didn't affect it, so I didn't hear which speaker was making the sound. . .

On the plus side, at least I can believe the measurements I made the second round have some validity.

I feel really stupid right now.

Where do you think I should set the gating in the attached picture? (blue is raw/nearfield, red is at 4ms/3ft distance).

Thanks again for your help!

Sandy.

[rpb]

When I use my ladder outside, I think I can set it at 200 hz. Maybe a little lower. I can't remember. I consider 300 hz low enough for anything I do. You may need to zoom in on the impulse to see the first reflection. Since you added absorbtion on the floor, it may be hard to see. Don't feel bad, your just getting experience like the rest of us. It looks like the response is nice and flat.