"Do you want to replay the echoes in the speaker or what"
Yep that. And import it into ARTA for CSD and the like.
Any idea how to get the text file into wav in the meantime?
Audacity seems useless in importing raw regardless how it it formatted so no luck for me there.
"For me its not a big deal to convert the ASCII to whatever format I want... but for other (non-programmers) it is."
Do you have a simple comm line app that would work?
I have not been able to find one.
Yep that. And import it into ARTA for CSD and the like.
Any idea how to get the text file into wav in the meantime?
Audacity seems useless in importing raw regardless how it it formatted so no luck for me there.
"For me its not a big deal to convert the ASCII to whatever format I want... but for other (non-programmers) it is."
Do you have a simple comm line app that would work?
I have not been able to find one.
New Release Version 1.1.6.0
New Release Version 1.1.6.0 (2009-06-17)
Features/Changes:
* ASIO support
* DirectSound support
* Export Impulse response as wave file
* TimeWindow start adjusted for logsweep
* Impulse domain: No autozoom in time after measurement
http://www.holmacoustics.com/downloads/HOLMImpulse/ChangeLog.txt
Impulse to wave-file
It was quite easy to implement export to wave file, so soon we will have the other wave-file save/load utilities
MME, DS (DirectSound), ASIO
I have made a test with my USB soundcard (analog loopback)
It seems that ASIO is slightly better regarding higher order harmonic distortions. (See attachment where the THD is shown)
But MME and DS are the same.
DS is also controlled by the same volume settings, but might have improved latency, which does not matter in this application.
http://msdn.microsoft.com/en-us/library/ms790010.aspx
My measurements were reproducible, but I have no explanation why the ASIO did perform slightly better
New Release Version 1.1.6.0 (2009-06-17)
Features/Changes:
* ASIO support
* DirectSound support
* Export Impulse response as wave file
* TimeWindow start adjusted for logsweep
* Impulse domain: No autozoom in time after measurement
http://www.holmacoustics.com/downloads/HOLMImpulse/ChangeLog.txt
Impulse to wave-file
It was quite easy to implement export to wave file, so soon we will have the other wave-file save/load utilities
MME, DS (DirectSound), ASIO
I have made a test with my USB soundcard (analog loopback)
It seems that ASIO is slightly better regarding higher order harmonic distortions. (See attachment where the THD is shown)
But MME and DS are the same.
DS is also controlled by the same volume settings, but might have improved latency, which does not matter in this application.
http://msdn.microsoft.com/en-us/library/ms790010.aspx
My measurements were reproducible, but I have no explanation why the ASIO did perform slightly better
Attachments
I have had some time to play around with this and its ease of use is impressive since I would not be able to use it if it was not easy 
I believe I was able to get a good sound card calibration and my Mic calibration file imported succesfully.
Now I need to figure out how to do gated measurements on my current speakers in my HT Room.
Anyone want to step me through that? Maybe its automatic??
Thanks
Doug
I believe I was able to get a good sound card calibration and my Mic calibration file imported succesfully.
An externally hosted image should be here but it was not working when we last tested it.
Now I need to figure out how to do gated measurements on my current speakers in my HT Room.
Anyone want to step me through that? Maybe its automatic??
Thanks
Doug
Re: New Release Version 1.1.6.0
BTW, I checked the 32bit FP .WAV export, no problems. So now I can easily convolve a step response etc. without further ado. Ya-Hoo!
- Klaus
I found the same (internal semi-pro soundcard, so jitter from clock regen in the USB-card is not the cause, it seems). The error term is in the odd harmomics, 3rd especially. It's weird, isn't it?askbojesen said:
BTW, I checked the 32bit FP .WAV export, no problems. So now I can easily convolve a step response etc. without further ado. Ya-Hoo!
- Klaus
Attachments
Semi-Automatic, more or less. Zoom out, in the impulse window, and follow the info that pops up each time you move your curser from one window into another. Which tells you to simply drag the right edge of the gate window.doug20 said:
Or go via Options-->TimeWindow. You have to calculate the gate time into samples at first:
samples = time * sample_freq, say for 20ms: 0.02s * 44100 = 882 samples
Klaus
Doug,
You could use a simple Radio-Shack SPL-meter nearby (as close as possible to) the mic position to check the sensivity of the mic+ADC-chain before the actual measurement, say with at 400Hz test tone. Don' use any weighting for the level-meter (or C-weighting if that is an option).
Then shift the response in the FR graph at 400Hz to that level --> scaling is then in absolute dB-SPL. Is is more robust when you use a set of frequencies (not to high, say 250, 400, and 600Hz or something). Alas, at the moment Ask has restricted the top range to 20dB, so shift to 0dB instead (will be pretty close anyway) -- not to the SPL reading --, and then just add that value to everything with your brain (or photo-chop the graph output accordingly).
Otherwise I don't know if any mic sensivity given in a calibration file is used by Ask, so we better ask Ask
- Klaus
You could use a simple Radio-Shack SPL-meter nearby (as close as possible to) the mic position to check the sensivity of the mic+ADC-chain before the actual measurement, say with at 400Hz test tone. Don' use any weighting for the level-meter (or C-weighting if that is an option).
Then shift the response in the FR graph at 400Hz to that level --> scaling is then in absolute dB-SPL. Is is more robust when you use a set of frequencies (not to high, say 250, 400, and 600Hz or something). Alas, at the moment Ask has restricted the top range to 20dB, so shift to 0dB instead (will be pretty close anyway) -- not to the SPL reading --, and then just add that value to everything with your brain (or photo-chop the graph output accordingly).
Otherwise I don't know if any mic sensivity given in a calibration file is used by Ask, so we better ask Ask
- Klaus
You don't ... see:doug20 said:
http://www.holmacoustics.com/downloads/HOLMImpulse/Issues.txt
+-----------------------------------------------------------------------+
HOLMImpulse FEATURE REQUESTS
+-----------------------------------------------------------------------+
* Absolute gain (calibration dependent)
If you were to give absolute SPL then you must know:
1. The sensitivity of your soundcard
2. The sensitivity of your microphone
(Like Klaus is telling how we can get)
Future
I'll implement two options for absolute SPL.
A. Enter sensitivity for one frequency
B. Simply have absolute values in the mic-cal file
OK?
KSTR said:Doug,
You could use a simple Radio-Shack SPL-meter nearby (as close as possible to) the mic position to check the sensivity of the mic+ADC-chain before the actual measurement, say with at 400Hz test tone. Don' use any weighting for the level-meter (or C-weighting if that is an option).
Then shift the response in the FR graph at 400Hz to that level --> scaling is then in absolute dB-SPL. Is is more robust when you use a set of frequencies (not to high, say 250, 400, and 600Hz or something). Alas, at the moment Ask has restricted the top range to 20dB, so shift to 0dB instead (will be pretty close anyway) -- not to the SPL reading --, and then just add that value to everything with your brain (or photo-chop the graph output accordingly).
Otherwise I don't know if any mic sensivity given in a calibration file is used by Ask, so we better ask Ask
- Klaus
Thank you, I will do as you suggested.
Calibration using a pistonphone
Using a pistonphone gives you sensitivity for one frequency (Typical 250 Hz)
So that A. "Enter sensitivity for one frequency" covers this
http://www.google.com/search?q=pistonphone+250+Hz
Then how is this one frequency sensitivity measured?
I could of course make a bandpassed levelmeter to make this calibration
good idea!
jcga said:
Using a pistonphone gives you sensitivity for one frequency (Typical 250 Hz)
So that A. "Enter sensitivity for one frequency" covers this
http://www.google.com/search?q=pistonphone+250+Hz
Then how is this one frequency sensitivity measured?
I could of course make a bandpassed levelmeter to make this calibration
good idea!
Thanks very mmuch for the new features! You've been working hard. I like the DS and ASIO support. Below you will find my M-Audio USB Duo in loop-back with all 3 modes. Don't know why the THD shows lower in ASIO, as the impulses look identical - at any zoom.
The zero time works very well now, too. Even with filters. Great!
I did manage to export the wave in 32bit floating, but ARTA does not like it. Had to run it thru a wave editor and go to 16 bit PCM. Anyone else find that?
Important Question
The 3 measurement modes -
Swept Sine
Chirp
MLS
Shouldn't they sound different? They all sound the same to me now - all are a log sweep sine. Am I doing something wrong?
The zero time works very well now, too. Even with filters. Great!
I did manage to export the wave in 32bit floating, but ARTA does not like it. Had to run it thru a wave editor and go to 16 bit PCM. Anyone else find that?
Important Question
The 3 measurement modes -
Swept Sine
Chirp
MLS
Shouldn't they sound different? They all sound the same to me now - all are a log sweep sine. Am I doing something wrong?
Attachments
Here's a cool feature request. Don't know if it's possible, but it would be nice to see all of the harmonics graphed at once.
It's nice to see the harmonics/freq graph. Getting to see each broken out into its own graph line ought to be very interesting. Might relate to the "sound" of a certain driver or other DUT.
It's nice to see the harmonics/freq graph. Getting to see each broken out into its own graph line ought to be very interesting. Might relate to the "sound" of a certain driver or other DUT.
Bugs an future features
(Notice the harmonic distortion pre-impulses, that are marked)
(Just one filter at a time perhaps)
- 32 bit float wave
- 64 bit double wave
- 16 bit PCM wave (ARTA compatible)
- 24 bit PCM wave (ARTA compatible)
- 32 bit PCM wave
- sorry... that is how software development works - its called Regression Testing 
Thanks for telling me
I'll be back with a fix soon
Not at any zoom? - Try the dB scale in the impulse domain and scale to lets say -130 db and zoom out like I have done in:panomaniac said:
(Notice the harmonic distortion pre-impulses, that are marked)
Thanks! can you post your DCX2496 filters meaurement again?
(Just one filter at a time perhaps)
Yes I tested myself in ARTA. In next release you can save as:
- 32 bit float wave
- 64 bit double wave
- 16 bit PCM wave (ARTA compatible)
- 24 bit PCM wave (ARTA compatible)
- 32 bit PCM wave
That's a bug I have introduced somewhere while making features
- sorry... that is how software development works - its called Regression Testing Thanks for telling me
I'll be back with a fix soon
The "Sound"
Yes, I thought about that also. We also want the phase of the distortion
Let us study the phase of the harmonic distortions - it's not irrelevant if considering the masking effects, that Earl are taking talking about also - or...?
My problem is how to visualize this in a consistent way so the plot is not overcrowded by curves.
panomaniac said:
Yes, I thought about that also. We also want the phase of the distortion
Let us study the phase of the harmonic distortions - it's not irrelevant if considering the masking effects, that Earl are taking talking about also - or...?
My problem is how to visualize this in a consistent way so the plot is not overcrowded by curves.
hi...askbojesen
i never translate the page more than 3 pages. but in this time i think its sure more than 100 pages. i would like to cry really...kiki(thai baby has a litle lol) ok i just install it. askbojesen i have a question. i want to know this program can export the the txt file(freq and phase respond)? my xo loudsoft wants the txt file to work with.
askbojesen thanks for all your post.
kiki
i never translate the page more than 3 pages. but in this time i think its sure more than 100 pages. i would like to cry really...kiki(thai baby has a litle lol) ok i just install it. askbojesen i have a question. i want to know this program can export the the txt file(freq and phase respond)? my xo loudsoft wants the txt file to work with.
askbojesen thanks for all your post.
kiki
I had no problem with the Wave export into ARTA. Although The old measurments had a DC offset, but I can't reproduce that now so it might have been an old glitch.
The Lister plugin in Total Commander had some problem playing it, sometimes it would sometimes not. So it might be a problem with the header?
Here is a burst decay with all 7 B&C 15PE40 combined. Mic was 1m in front of the speaker. The volume was turned up "enough".
Any comments?
The Lister plugin in Total Commander had some problem playing it, sometimes it would sometimes not. So it might be a problem with the header?
Here is a burst decay with all 7 B&C 15PE40 combined. Mic was 1m in front of the speaker. The volume was turned up "enough".
An externally hosted image should be here but it was not working when we last tested it.
Any comments?