Measurement system errors, need help...

[romanbednarek]

I'm not sure how much you guys can help me with this problem because I've exhausted a lot of options already and also posted this on the SE user's group but here it goes.

I'm using SoundEasy to take MLS measurements and am getting a boost above 10kHz of about a dB or two. The last time I measured was just over a month ago and everything worked fine. I've compared the new measurements to older measurements of 4 different speakers and all of the new measurements have this boost.

I've tried different cables, different amps, bypassing my jig by just using cables, different mixer inputs/outputs. Nothing fixes the problem. John k. suggested that it had to do with resistance/capacitance in the reference signal lead going back to my computer which would attenuate the amp's response and compensate for it by boosting the measurement. It makes a lot of sense but I think I've eliminated any hardware issues unless there is a bad connection on my soundcard's line-in.

I've uninstalled and reinstalled SE v10. I've checked all of the M-Audio Delta 410 soundcard settings. All that I can figure now is that some kind of Windows/software issue is going on where it is somehow filtering the line-in or some other audio input/output because Windows gets less and less reliable the longer it is loaded and the more stuff that has been added or removed. This is a bit of a stretch though, but I don't know where else to look.

I'm getting to the point where I might just have to modify my mic calibration file to compensate for this new boost but I don't like using "band-aids" like this. I'm not sure how much anybody will be able to help me with this issue but any suggestions would be appreciated because this is getting frustrating.

[dlneubec]

Hi Roman,

Have you tried rebuilding/importing your mic calibration file into SE? Maybe the one you are using has been corrupted insome way.

Just a thought.

Dan

> I'm not sure how much you guys can help me
> with this problem because I've exhausted a
> lot of options already and also posted this
> on the SE user's group but here it goes.

> I'm using SoundEasy to take MLS measurements
> and am getting a boost above 10kHz of about
> a dB or two. The last time I measured was
> just over a month ago and everything worked
> fine. I've compared the new measurements to
> older measurements of 4 different speakers
> and all of the new measurements have this
> boost.

> I've tried different cables, different amps,
> bypassing my jig by just using cables,
> different mixer inputs/outputs. Nothing
> fixes the problem. John k. suggested that it
> had to do with resistance/capacitance in the
> reference signal lead going back to my
> computer which would attenuate the amp's
> response and compensate for it by boosting
> the measurement. It makes a lot of sense but
> I think I've eliminated any hardware issues
> unless there is a bad connection on my
> soundcard's line-in.

> I've uninstalled and reinstalled SE v10.
> I've checked all of the M-Audio Delta 410
> soundcard settings. All that I can figure
> now is that some kind of Windows/software
> issue is going on where it is somehow
> filtering the line-in or some other audio
> input/output because Windows gets less and
> less reliable the longer it is loaded and
> the more stuff that has been added or
> removed. This is a bit of a stretch though,
> but I don't know where else to look.

> I'm getting to the point where I might just
> have to modify my mic calibration file to
> compensate for this new boost but I don't
> like using "band-aids" like this.
> I'm not sure how much anybody will be able
> to help me with this issue but any
> suggestions would be appreciated because
> this is getting frustrating.

[dlr]

> I'm getting to the point where I might just
> have to modify my mic calibration file to
> compensate for this new boost but I don't
> like using "band-aids" like this.
> I'm not sure how much anybody will be able
> to help me with this issue but any
> suggestions would be appreciated because
> this is getting frustrating.

One thing I'd suggest looking into is the mic cal file itself. I was going to ask if you still had it being used by SE, since some mic capsules can have some peaking before rolloff. If you've somehow changed SE so that it is not using the cal file, you'd see a change.

The only other thing that comes to mind is the mic/wand. If the what holds the capsule is changed, such as shrink wrap (as I use) shifting forward and closer to the edge of the mic or the mic sliding into a brass tube a bit farther, the mic will show a bit of elevation at higher frequencies. I think that Kim Girardin said that it can boost the top end by as much as 0.5db.

dlr

[romanbednarek]

> Hi Roman,

> Have you tried rebuilding/importing your mic
> calibration file into SE? Maybe the one you
> are using has been corrupted insome way.

> Just a thought.

> Dan

Thanks for the suggestion but that didn't fix the problem.

[romanbednarek]

> One thing I'd suggest looking into is the
> mic cal file itself. I was going to ask if
> you still had it being used by SE, since
> some mic capsules can have some peaking
> before rolloff. If you've somehow changed SE
> so that it is not using the cal file, you'd
> see a change.

No, the cal file is fine and active. Also the response of the cal file transitions from a slight peak to a rolloff within this region which doesn't coincide with the boost that I'm getting.

> The only other thing that comes to mind is
> the mic/wand. If the what holds the capsule
> is changed, such as shrink wrap (as I use)
> shifting forward and closer to the edge of
> the mic or the mic sliding into a brass tube
> a bit farther, the mic will show a bit of
> elevation at higher frequencies. I think
> that Kim Girardin said that it can boost the
> top end by as much as 0.5db.

> dlr

I'm using an ECM8000 mic. I don't know if mic responses can change over time maybe if exposed to a magnetic field from a driver. Getting a new calibration curve would determine if this is the issue but it is an awful lot to do just on a hunch.

This one really has me stumped and I'm running out of things to try but thanks for the suggestions.

[djarchow]

Roman,

Have you tried doing a loopback test with your mic preamp in the loop (assuming you can turn off the phantom voltage on your mic preamp). Then do the same with the amp in the loop. You should be able to rule everything out other than the mic this way.

Regards,

Dennis

> I'm not sure how much you guys can help me
> with this problem because I've exhausted a
> lot of options already and also posted this
> on the SE user's group but here it goes.

> I'm using SoundEasy to take MLS measurements
> and am getting a boost above 10kHz of about
> a dB or two. The last time I measured was
> just over a month ago and everything worked
> fine. I've compared the new measurements to
> older measurements of 4 different speakers
> and all of the new measurements have this
> boost.

> I've tried different cables, different amps,
> bypassing my jig by just using cables,
> different mixer inputs/outputs. Nothing
> fixes the problem. John k. suggested that it
> had to do with resistance/capacitance in the
> reference signal lead going back to my
> computer which would attenuate the amp's
> response and compensate for it by boosting
> the measurement. It makes a lot of sense but
> I think I've eliminated any hardware issues
> unless there is a bad connection on my
> soundcard's line-in.

> I've uninstalled and reinstalled SE v10.
> I've checked all of the M-Audio Delta 410
> soundcard settings. All that I can figure
> now is that some kind of Windows/software
> issue is going on where it is somehow
> filtering the line-in or some other audio
> input/output because Windows gets less and
> less reliable the longer it is loaded and
> the more stuff that has been added or
> removed. This is a bit of a stretch though,
> but I don't know where else to look.

> I'm getting to the point where I might just
> have to modify my mic calibration file to
> compensate for this new boost but I don't
> like using "band-aids" like this.
> I'm not sure how much anybody will be able
> to help me with this issue but any
> suggestions would be appreciated because
> this is getting frustrating.

[romanbednarek]

> Roman,

> Have you tried doing a loopback test with
> your mic preamp in the loop (assuming you
> can turn off the phantom voltage on your mic
> preamp). Then do the same with the amp in
> the loop. You should be able to rule
> everything out other than the mic this way.

> Regards,

> Dennis

I can turn the phantom voltage off but I'm not exactly certain how to connect everything to perform this test. Could you describe the setup using the Line-in (L/R) and sound card output (L/R)?

I've ruled out the amp because I've tried two different amps with identical results.

[dlneubec]

> Thanks for the suggestion but that didn't
> fix the problem.

Well, I guess you could upgrade to V14, giving you a reason to install a new software to see if that would help.

Good luck.

Dan

[djarchow]

Roman,

It is pretty simple. Using the actual cables you normally use for your setup connect your line out from the soundcard to the soundcard line in. Be sure to uncheck the mic calibration file in SE. Then do an MLS measurement and you should see a straight line in SE. If not there is something up with the cable you used or your setup in windows. Swap out each cable and see if you get flat response with each file.

Once you have done that insert your mic preamp in that same loop with the soundcard line out hooked to the mic preamp mic input (no phantom voltage)and the mic preamp line out hooked to the soundcard line in. if the line isn't flat thes something is up with your preamp.

Hope this helps.

Dennis

> I can turn the phantom voltage off but I'm
> not exactly certain how to connect
> everything to perform this test. Could you
> describe the setup using the Line-in (L/R)
> and sound card output (L/R)?

> I've ruled out the amp because I've tried
> two different amps with identical results.

[romanbednarek]

> Roman,

> It is pretty simple. Using the actual cables
> you normally use for your setup connect your
> line out from the soundcard to the soundcard
> line in. Be sure to uncheck the mic
> calibration file in SE. Then do an MLS
> measurement and you should see a straight
> line in SE. If not there is something up
> with the cable you used or your setup in
> windows. Swap out each cable and see if you
> get flat response with each file.

> Once you have done that insert your mic
> preamp in that same loop with the soundcard
> line out hooked to the mic preamp mic input
> (no phantom voltage)and the mic preamp line
> out hooked to the soundcard line in. if the
> line isn't flat thes something is up with
> your preamp.

> Hope this helps.

> Dennis

Thanks for the info.

The cable test reveals a perfectly flat response.

The mic preamp test reveals a slight deviation from flat that is about +/-0.3dB but the broadest peak is in the midrange and there are a few peaks/dips in the region of concern but nothing that matches the errors I'm getting.

So right now I think it is either something to do with my computer or my mic but I'm not sure how to troubleshoot either of them at this point.

[romanbednarek]

> Well, I guess you could upgrade to V14,
> giving you a reason to install a new
> software to see if that would help.

> Good luck.

> Dan

I upgraded to V11 a while back but installed my base software which is V10. Right now I just use SE for measurements so I'm not sure how much I would gain with V14 in that area. It may entice me to try to use some of the other features other than measurement if the user interface is improved with the newer version.

If it is something related to my computer I have a feeling that it is independant of the SE software but something that alters my soundcard's operation. All that I know is that I am running out of options.

[dlr]

Are you sure that you have all audio options in Windows turned off? There may be some other software accessing the sound card.

dlr

[romanbednarek]

> Are you sure that you have all audio options
> in Windows turned off? There may be some
> other software accessing the sound card.

> dlr

All "system sounds" are turned off if that is what you are referring to. I'm not sure how to adjust any other audio options. The control panel for the M-Audio Delta 410 has an option to disable other program's access to the mixer but this didn't change anything.

Is it common for microphones to suddenly change frequency response? I'm just running out of options.

[johnk]

Are you using the same frequency range? The cal file is tied tot he frequency range set in the preferences screen when it was crreated. If you change the frequency range you must make a new cal file consistent with that frequency range.

[romanbednarek]

> Are you using the same frequency range? The
> cal file is tied tot he frequency range set
> in the preferences screen when it was
> crreated. If you change the frequency range
> you must make a new cal file consistent with
> that frequency range.

I believe so because I've never changed that setting. The cal file I got from Kim G. covered a range of up to around 40kHz and I extended the tails and calculated the phase according to the instructions in your manual so that it now extends beyond 80kHz. I'm using a 96kHz sampling rate for my measurements.

I am certain that the cal file is working because I've processed measurements with and without it and the differences match the cal file response.

This one really has me stumped.

[romanbednarek]

> Are you using the same frequency range? The
> cal file is tied tot he frequency range set
> in the preferences screen when it was
> crreated. If you change the frequency range
> you must make a new cal file consistent with
> that frequency range.

If it is any help at all, I just looked at the entire response up to 48kHz. The high frequency boost gradually rises from 6kHz to 20kHz where it is about 1.6 dB. Then it remains there for the rest of the response but there is a slight convergence in the 30-40kHz range where the boost is minimal. These are rough estimates because the resolution of the plot is poor.

[romanbednarek]

> One thing I'd suggest looking into is the
> mic cal file itself. I was going to ask if
> you still had it being used by SE, since
> some mic capsules can have some peaking
> before rolloff. If you've somehow changed SE
> so that it is not using the cal file, you'd
> see a change.

> The only other thing that comes to mind is
> the mic/wand. If the what holds the capsule
> is changed, such as shrink wrap (as I use)
> shifting forward and closer to the edge of
> the mic or the mic sliding into a brass tube
> a bit farther, the mic will show a bit of
> elevation at higher frequencies. I think
> that Kim Girardin said that it can boost the
> top end by as much as 0.5db.

> dlr

I'm starting to really think that the response of my mic changed. It may have been dropped some time between my previous set of measurements and the current ones. I can surmise that the higher frequencies would be more susceptable to changing if this occurred. I just haven't dealt with this issue before.

I suppose that I have two options. One is to send it back to Kim G. and get a new set of cal curves for it. This is the safe way to go and solve the problem for sure if the mic response did change. The other option is to take my before and after measurements and modify my current cal file to compensate for the change. The only problem with that option is that if it isn't the mic and the problem fixes itself elsewhere my measurements will have a dip in this region.

I would like to thank everybody for their help and I'm still open to any future suggestions if anybody has any ideas.

[romanbednarek]

Provided Link: http://www.pesupport.com/cgi-bin/config.pl?read=398732


> Are you using the same frequency range? The
> cal file is tied tot he frequency range set
> in the preferences screen when it was
> crreated. If you change the frequency range
> you must make a new cal file consistent with
> that frequency range.

I'm starting to think that the my mic response changed. See my other post for details.

[djarchow]

Roman,

Since the loop back test shows flat response with your mic preamp and cables, then that rules out your cables and PC.

I suppose it could still be the preamp when you turn on the phantom voltage, or of course it could be your mic. Have you made certain that you are using the exact same mic cal file as the previous measurements?

Regards,

Dennis

> Thanks for the info.

> The cable test reveals a perfectly flat
> response.

> The mic preamp test reveals a slight
> deviation from flat that is about +/-0.3dB
> but the broadest peak is in the midrange and
> there are a few peaks/dips in the region of
> concern but nothing that matches the errors
> I'm getting.

> So right now I think it is either something
> to do with my computer or my mic but I'm not
> sure how to troubleshoot either of them at
> this point.

[romanbednarek]

> Roman,

> Since the loop back test shows flat response
> with your mic preamp and cables, then that
> rules out your cables and PC.

> I suppose it could still be the preamp when
> you turn on the phantom voltage, or of
> course it could be your mic. Have you made
> certain that you are using the exact same
> mic cal file as the previous measurements?

> Regards,

> Dennis

I'm certain that it is the same mic cal file that I had before because I only have one set up for SE and I didn't change the settings for SoundEasy from the time I measured before.

I'm thinking more and more that it is the mic because I think that I dropped it a while back, after I took the last set of "good" measurements. I'm not sure how mics react to drops but I suppose that the element is susceptable to damage this way. Now I guess I have to make sure that the low frequency response is still okay but that will be a bit tougher to test.

I'm probably going to make a modified mic cal file that compensates for the change in the response. I guess I'll have to take reference measurements of a known speaker on occasion to make sure that things don't revert back to the original state somehow (like if the issue wasn't the mic after all).

I really appreciate your help. At least I learned something in the process including the ability to use SoundEasy to take response measurements of things other than speakers.