Hello Ask!
I am wondering what would be the upcoming improvements on HOLMImpulse in 2010 ?
- use of a reference signal (before the DUT)
- impedance measurements
- automatic minimum phase
- CSD graph
- crossover simulation
Am I dreaming ?
I like your software a lot for its stability, rapidity and userfriendlyness.
At the moment, I still use others to complete my needs...
Happy new year to you and everyone here !
I am wondering what would be the upcoming improvements on HOLMImpulse in 2010 ?
- use of a reference signal (before the DUT)
- impedance measurements
- automatic minimum phase
- CSD graph
- crossover simulation
Am I dreaming ?
I like your software a lot for its stability, rapidity and userfriendlyness.
At the moment, I still use others to complete my needs...
Happy new year to you and everyone here !
I have been a little quiet...
I have been a little quiet... I have had my reasons...
I know that. Thank's a lot for not giving up on this piece of software.
Well, Regarding new features - I don't expect much
I rather go for stabilizing and work on the manual, maybe start a wiki.
But It will not be a wiki here at DIY
I might revisit the THD calculation
I'll be back
I have been a little quiet... I have had my reasons...
I know that. Thank's a lot for not giving up on this piece of software.
Well, Regarding new features - I don't expect much
I rather go for stabilizing and work on the manual, maybe start a wiki.
But It will not be a wiki here at DIY
I might revisit the THD calculation
I'll be back
Comparasion with ARTA (Distortion)
Thank you very much for this testing
Can you attach or sent the measurement file where you have measured the distortion, then I can see what is going on ?
I don't think that we can use the logsweep for this, I will re-read and think again
Yes I could do this, I'm not much an analog-filter guy myself, but I could implement some formulas
Thank you very much for this testing
Can you attach or sent the measurement file where you have measured the distortion, then I can see what is going on ?
I don't think that we can use the logsweep for this, I will re-read and think again
Yes I could do this, I'm not much an analog-filter guy myself, but I could implement some formulas
Could you please re-explain this feature?
I guess it's something to do with FR text export?
Yes, it's to do with the FR test export.
HOLM is normalized near 0dB. Thus the exported values are near or often below 0dB.
In some speaker design software this 0dB reference is far too low. The software expects to see 80-100dB, like a real speaker SPL at 1M. So adding 80, 90 or more dB to the exported FR measurement makes the file usable in other software.
We found a work around. Add the offset to the measurement via the "Measurement Options" - "offset dB" window. That will force the graph off the chart (max is 20dB) but it works OK for export.
A "Normalize to Xdb" dialog would be handy in the export. But for the moment the work around is OK.
HOLM is normalized near 0dB. Thus the exported values are near or often below 0dB.
In some speaker design software this 0dB reference is far too low. The software expects to see 80-100dB, like a real speaker SPL at 1M. So adding 80, 90 or more dB to the exported FR measurement makes the file usable in other software.
We found a work around. Add the offset to the measurement via the "Measurement Options" - "offset dB" window. That will force the graph off the chart (max is 20dB) but it works OK for export.
A "Normalize to Xdb" dialog would be handy in the export. But for the moment the work around is OK.
I've hit the same issue as panomaniac. Used the same workaround.
Another little convenience feature that would be nice would be to use drop down, multi-select edit boxes instead of checkboxes for picking I/O channels. I have a 20 input, 20 output audio interface so I end up doing routing in the interface's mixer rather than in HOLMImpulse (Jack would be the software equivalent for routing). A little more flexibility in HOLMImpulse would make it play a bit nicer with the system.
I sent Thomas side by side data for HOLMImpulse and SoundEasy (and said nice things about HOLMImpulse) a few weeks back; toss me a PM if he didn't pass it on to you. Will probably have some more in the future, too.
Another little convenience feature that would be nice would be to use drop down, multi-select edit boxes instead of checkboxes for picking I/O channels. I have a 20 input, 20 output audio interface so I end up doing routing in the interface's mixer rather than in HOLMImpulse (Jack would be the software equivalent for routing). A little more flexibility in HOLMImpulse would make it play a bit nicer with the system.
I sent Thomas side by side data for HOLMImpulse and SoundEasy (and said nice things about HOLMImpulse) a few weeks back; toss me a PM if he didn't pass it on to you. Will probably have some more in the future, too.
Hello Ask,
I did not have the original file any more, but I made I new one (results are consistent).
Did you read my post in 580 (fade-in time)? Could it be that this is causing the problem?
Greetz,
Wim
No it's not the fade in-time - it is the algorithm which I use to calculate THD from the log-sweep impulse response.
See the attachment with the impulse zoomed out and on a dB scale - No Harmonic pre-impulse can be seen
From my point of view the harmonic distortion does not make sense if the noise floor is greater than the harmonics
But HOLMImpulse is misleading here... I agree and want to do better
Attachments
WM61A Capsules & Amplifier Calibration
I have seen the results of other tests on the web of the WM61A on the Internet. While all show some rise in the high end of the audible frequencies, the size and extension of the rise has varied. Some show only a 3 db rise, and some show the rise returning to the mean before 20 khz. Do you have knowledge of how much variation there may be in this part?
Also, I used the loopback to adjust for the variation in the sound card. I'd like to adjust for the complete loop variation. In other words, use a 1Khz tone and an oscilloscope to set the output at the speaker terminals to 1 watt, and then feed the voltage at the speaker back through a resistor divider to the microphone input, and run the D/A calibration on the entire loop. Do you see any problem with that?
I have seen the results of other tests on the web of the WM61A on the Internet. While all show some rise in the high end of the audible frequencies, the size and extension of the rise has varied. Some show only a 3 db rise, and some show the rise returning to the mean before 20 khz. Do you have knowledge of how much variation there may be in this part?
Also, I used the loopback to adjust for the variation in the sound card. I'd like to adjust for the complete loop variation. In other words, use a 1Khz tone and an oscilloscope to set the output at the speaker terminals to 1 watt, and then feed the voltage at the speaker back through a resistor divider to the microphone input, and run the D/A calibration on the entire loop. Do you see any problem with that?
Line-in as input device?
I have a microphone, Audyssey APM1, with the corresponding calibrated preamplifier ACP2. Unfortunately my laptop does only have a microphone input, so I need to buy myself an external soundcard with a line-in input. I intend to buy a Creatives USB Sound Blaster X-Fi Surround 5.1 but before I do I would like to be sure about this setup. Am I able to use a line-in input as the input device in HOLMImpulse?
Thanks in advance
Peter
I have a microphone, Audyssey APM1, with the corresponding calibrated preamplifier ACP2. Unfortunately my laptop does only have a microphone input, so I need to buy myself an external soundcard with a line-in input. I intend to buy a Creatives USB Sound Blaster X-Fi Surround 5.1 but before I do I would like to be sure about this setup. Am I able to use a line-in input as the input device in HOLMImpulse?
Thanks in advance
Peter
This app has a very nice user interface. Perfect for FR and IR.
One question, can HOLMimpulse use a loop-back setup where one output channel is fed back to the input to reduce the effects of a marginal signal path, similar to what WinMLS and RPlus do? I did not find this in the "manual/theory" document, nor in this thread.
Thanks, Michael
One question, can HOLMimpulse use a loop-back setup where one output channel is fed back to the input to reduce the effects of a marginal signal path, similar to what WinMLS and RPlus do? I did not find this in the "manual/theory" document, nor in this thread.
Thanks, Michael
You bet. I can't understand why so much effort would be put into a single channel tool.
John
Make no mistake about it, and I know that you won't like this, but even one channel, HolmImpulse is the best loudspeaker measurement tool that I have used to date. Its all I use anymore. Sure, it could be improved, but it is still exceptional as it is.