I'm finally making acoustic measurements with my Wallin Jig II. It's a great program, but...
But I don't know if it's working right! It would be nice if someone could look at this graph and say, "yup, that looks right," or "nope, you done something wrong."
The following is for the Tangband 881S in a TL.
Thanks,
Dave
PS--I've crossposted this to the SW forum, but it's not nearly so populated as this one. Sorry if I offend.
But I don't know if it's working right! It would be nice if someone could look at this graph and say, "yup, that looks right," or "nope, you done something wrong."
The following is for the Tangband 881S in a TL.
An externally hosted image should be here but it was not working when we last tested it.
Thanks,
Dave
PS--I've crossposted this to the SW forum, but it's not nearly so populated as this one. Sorry if I offend.
No, that is probably not correct....
Things that make me say that:
- even nearfield, the response would not likely be that smooth in the upper treble.
- even nearfield, I doubt the response is 3dB down at 20hz.
Try this - hook up your loop cable on both channels and take a nearfield measurement. You should have a perfectly flat response. You should have about the same signal level on both channels (in the VU meters). If that is okay, post more info about your mic, preamp, how you're measuring (where your mic is), levels and settings in SW, etc.
Things that make me say that:
- even nearfield, the response would not likely be that smooth in the upper treble.
- even nearfield, I doubt the response is 3dB down at 20hz.
Try this - hook up your loop cable on both channels and take a nearfield measurement. You should have a perfectly flat response. You should have about the same signal level on both channels (in the VU meters). If that is okay, post more info about your mic, preamp, how you're measuring (where your mic is), levels and settings in SW, etc.
It looks like you either have used a lot of smoorthing in the measurements or you have choosen for a very small time window. Anyway, the result doesn't look like a real lif speaker measurement. What is the microphone distance ? According to the high frequency sloop it looks like a free field measurement, but on the other hand acoording to the flatness of the measurement it looks like a nearfield measurement, both with a lot of smoothing and a very short time gate. Are you sure you are looking at the frequency response instead of the phase response ? It sure does look more like a phase reponse plot.
Here's how I'm doing it:
Microphone: Studio Projects C4 with omni capsule
Preamp: Mackie 1202 VLZ
I have to say, I'm a little confused about the SW settings for acoustic measurement. The "unofficial" manual, as good as it is, and far from me to complain, is not particularly clear here. There are references to non-existent menus, and procedures which are impossible to duplicate. At least for the first 100 times I've tried.
Oh, and Jig II is my jig.
Dave
An externally hosted image should be here but it was not working when we last tested it.
Microphone: Studio Projects C4 with omni capsule
Preamp: Mackie 1202 VLZ
I have to say, I'm a little confused about the SW settings for acoustic measurement. The "unofficial" manual, as good as it is, and far from me to complain, is not particularly clear here. There are references to non-existent menus, and procedures which are impossible to duplicate. At least for the first 100 times I've tried.
Oh, and Jig II is my jig.
Dave
Dave.
I don't think I've read the unofficial manual. Never the less, with both channels looped (ie, no speaker, no mic involved), you should have about the same signal level on both channels (in the VU meters). If one channel does not have the same level, that means something is probably wrong. Ie, if one channel is 16000 max and the other is 100 max when the channels are looped, the 100 channel is probably not getting any signal - just noise or crosstalk between the channels.
For example, open SW, create a driver, make that driver active, have both channels looped, in the View menu have 'VU meter' checked, take a nearfield measurement (so there is no gating to worry about), and the Max and Min numbers in the VU meters should be about the same for the left and right channels. In my case, 32000 is clipping, so I shoot for a max around 16000 to avoid any clipping.
When I actually am taking a measurement, my levels may end up being different between the left and right channel - it depends on how your gain structure is set up, but ideally the levels would still both be about the same and close to 16k or so (in my case) for good signal to noise resolution. For example on my last xover design, I had a compression driver tweeter and two 8" mids where the mids were much less sensitive. I would measure the tweeter first and then when I measured the mids at the same level, the mic channel's level would be much lower than the reference due to the lower sensitivity drivers playing at a lower SPL for the same volume settings. This could be worked around if I cared to take the time, but I do not care - that was good enough for what I was doing.
John
I don't think I've read the unofficial manual. Never the less, with both channels looped (ie, no speaker, no mic involved), you should have about the same signal level on both channels (in the VU meters). If one channel does not have the same level, that means something is probably wrong. Ie, if one channel is 16000 max and the other is 100 max when the channels are looped, the 100 channel is probably not getting any signal - just noise or crosstalk between the channels.
For example, open SW, create a driver, make that driver active, have both channels looped, in the View menu have 'VU meter' checked, take a nearfield measurement (so there is no gating to worry about), and the Max and Min numbers in the VU meters should be about the same for the left and right channels. In my case, 32000 is clipping, so I shoot for a max around 16000 to avoid any clipping.
When I actually am taking a measurement, my levels may end up being different between the left and right channel - it depends on how your gain structure is set up, but ideally the levels would still both be about the same and close to 16k or so (in my case) for good signal to noise resolution. For example on my last xover design, I had a compression driver tweeter and two 8" mids where the mids were much less sensitive. I would measure the tweeter first and then when I measured the mids at the same level, the mic channel's level would be much lower than the reference due to the lower sensitivity drivers playing at a lower SPL for the same volume settings. This could be worked around if I cared to take the time, but I do not care - that was good enough for what I was doing.
John
Don't know if your jig will do it or not. I built a Wallin jig a long time ago, but I only use it for Z measurements. But what I'm talking about is running the left output of the soundcard to the left input of the soundcard and the right output to the right input. So there is just a loop between input and output. I do this with a 1/8" stereo to RCA cable and a 1/8" to RCA converter on the other end - all stuff you could pick up at radioshack. If this procedure does not work correctly (response with both channels looped), it is likely that nothing else will work correctly even though you may get results that look half-way convincing.
Ok,
I may have tracked down an incorrect routing of cables--I know I did, in fact--thanks to your help, John.
This looks like a real driver, though I wonder how I might improve the response!
Thanks,
Dave
I may have tracked down an incorrect routing of cables--I know I did, in fact--thanks to your help, John.
This looks like a real driver, though I wonder how I might improve the response!
An externally hosted image should be here but it was not working when we last tested it.
Thanks,
Dave
Hi Dave,
about mantaining the same channels level (acoustic measurement), it is done by the Mic-pre and the Windows mixer.
About the new curve, it is just a nearfield, right? Did you add the port response? Remember that NF is limited in the upper frequencies. Besides, your mic has a FR starting from 40 Hz, and do you have it's curve? Looks like, to me, it is a recording mic, more than a measurement mic, so be careful in interpreting the measurements.
Regards
Claudio
about mantaining the same channels level (acoustic measurement), it is done by the Mic-pre and the Windows mixer.
About the new curve, it is just a nearfield, right? Did you add the port response? Remember that NF is limited in the upper frequencies. Besides, your mic has a FR starting from 40 Hz, and do you have it's curve? Looks like, to me, it is a recording mic, more than a measurement mic, so be careful in interpreting the measurements.
Regards
Claudio
OK,claudio said:
The mic is a nice recording mic; you are correct, and thanks for reminding me that its curve will make my measurements misleading.
And now that I've actually gone to the trouble of zero-ing out my equipment, everything is reading fine. The port response graph has helped smoothe things out, too.
I'm suddenly very happy!
Thanks
Dave
That looks like the proximity effect of a cardoid microphone and the sharp dip at 8kHz is maybe due to the standing wave between the dustcap and the front of the microphone itself or due to partial swinging of the speakercone. Try to do nearfield measurements at a distance wich is at least equal to the diameter of the speaker cone, this way you still have a good response of the actual low frequency behaviour of the driver. (baffle step not included)
Your advice paid off. I'm not sure it was microphone proximity effect because it was an omni mic; it could have been a variation introduced by such a small driver or something, but that's an argument for another day.Sjef said:
Another adjustment I made was to get the speaker off the ground; that sharply reduced the 200Hz hump.
I've graduated from Speaker Workshop 101, now, and I'm ready to get on to advanced measuring--maybe tomorrow
Thanks,
Dave
To extend my joke:paulspencer said:
Maybe community colleges ought to offer courses in "Understanding Speaker Workshop."
It has proven to be a fine measuring device for me. I wouldn't rest entirely on it, but the investment of time and effort seems--seems--to have paid off. I need to build something bigger and more complicated before I make my final judgment.
Dave
speaker workshop is really as good (and in some respects even better) as any other commercial available speaker mesurement software. The only thing is that people won't believe it because it is freeware. The only thing you have to do is to make sure you have the right hardware around it. Believe me, the software is worth it enough to invest in a serious microphone/preamp and a good measurment jig and soundcard (lesson 1, be aware of ground loops) , without that it's useless. You can build your self a very good measurment tool for the total amount of less than $200,- isn't that something nice ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.