Hi I would like to know if others have this problem. I was using LIMP successfully for the first time today and it crashed my Focusrite Saffire Mix Control ASIO driver. A crash so bad I had to reboot the computer as "hardware" was not detected, the crash was so bad I had to use the power switch to shutdown. I also experience random pops with LIMP and more so with the new REW.
@@@@@@@@
Is this common ?
****
I have a Focusrite Saffire Pro24 and I have had no problems with the drivers or pops . . . Maybe a couple years ago (older driver) I would hear the occasional pop when mixing in Protools 10 (usually during a "bounce"). I have not noticed any noise with Protools 11 however I do not mix very often since it's release....My system is also a HT PC and I have noticed no noise listening to music or movies.....
thanks
erick
@@@@@@@@
Is this common ?
****
I have a Focusrite Saffire Pro24 and I have had no problems with the drivers or pops . . . Maybe a couple years ago (older driver) I would hear the occasional pop when mixing in Protools 10 (usually during a "bounce"). I have not noticed any noise with Protools 11 however I do not mix very often since it's release....My system is also a HT PC and I have noticed no noise listening to music or movies.....
thanks
erick
Last edited:
I was indeed sampling at 96k in LIMP, I only had one pop or two. 48K in REW and it was corrupting all but one sweep. ( I am sticking with LIMP for the time being - REW seems problematic - I am a noob to both )
Buffer.... okay.... Last time I was in my control panel I think I set it at 256 samples, a bad habit from tracking/ recording full bands with a PT-HD3 - TDM system. I'll try increasing it . Thank you very much....
Buffer.... okay.... Last time I was in my control panel I think I set it at 256 samples, a bad habit from tracking/ recording full bands with a PT-HD3 - TDM system. I'll try increasing it . Thank you very much....
Last edited:
I was able to finish measuring a Jensen JPW128 for VAS with settings changed to 48k and a 1024 sample buffer. However today I had another fatal crash when I went to do an added mass sweep on a Pyle-Pro PDW158.
I noticed that with the new settings I am having very frequent pops.
I initially had the sampling at 96K as somewhere in the manual it states to use the native sampling rate of your soundcard for accuracy ( "no sample rate conversion" - however I am not sure that is an accurate concept with an ASIO pro-sumer device). I know very well that I don't need the freq. resolution, which is also set in the frequency sweep. But the sample frequency certainly would not hurt in any way. People often think sample frequency just dictates high frequency resolution. It does not, or we would not have 192k HD recordings in the studio and movie sound tracks. It also determines the number of times a sample is taken or read and generated. Creating more accurate samples through shear quantity.....If I recall correctly at 96k sample frequency I am taking 96k samples per second as opposed to 48k or 41k. The audio frequency response is controlled and limited in my "setup". The number of times I sample is the sampling freq.
I will stick with 96k sampling on my next sweep as I had fewer AD/DA & Driver errors. (but my sweep is limited to 20kHz). I will also reduce the buffer back to 256 samples or perhaps lower, as I had significantly fewer pops at 256 samples. 96K@256 sample buffer 1-2 pops per sweep. 1024 sample buffer at 48k I had pops every few "sweep samples".
I noticed that with the new settings I am having very frequent pops.
I initially had the sampling at 96K as somewhere in the manual it states to use the native sampling rate of your soundcard for accuracy ( "no sample rate conversion" - however I am not sure that is an accurate concept with an ASIO pro-sumer device). I know very well that I don't need the freq. resolution, which is also set in the frequency sweep. But the sample frequency certainly would not hurt in any way. People often think sample frequency just dictates high frequency resolution. It does not, or we would not have 192k HD recordings in the studio and movie sound tracks. It also determines the number of times a sample is taken or read and generated. Creating more accurate samples through shear quantity.....If I recall correctly at 96k sample frequency I am taking 96k samples per second as opposed to 48k or 41k. The audio frequency response is controlled and limited in my "setup". The number of times I sample is the sampling freq.
I will stick with 96k sampling on my next sweep as I had fewer AD/DA & Driver errors. (but my sweep is limited to 20kHz). I will also reduce the buffer back to 256 samples or perhaps lower, as I had significantly fewer pops at 256 samples. 96K@256 sample buffer 1-2 pops per sweep. 1024 sample buffer at 48k I had pops every few "sweep samples".
Last edited:
You see it works like this:
samples are taken of a waveform (recording). playback = the analog section fills in the blanks between samples and you get an approximation of the original analog waveform. The more samples you have the better the analog waveform is represented/recorded/digitized/manipulated and read for playback....I am distilling a hour long college lecture.
samples are taken of a waveform (recording). playback = the analog section fills in the blanks between samples and you get an approximation of the original analog waveform. The more samples you have the better the analog waveform is represented/recorded/digitized/manipulated and read for playback....I am distilling a hour long college lecture.
You do not need to use 88.2 kHz more so 96 kHz to measure a woofer. If accuracy is your quest, drop down to 48 kHz and increase your “FFT Size” in addition to the “Frequency Increment” in LIMP.
I use 48 kHz with a 1/48 per octave and 32768. This gives me resolution of 1.46 Hertz which will be equivalent to you using 96 kHz, with a FFT size of 65536 without my computer’s clock becoming overloaded during the process.
You could even use 48 kHz with an FFT size of 66536. That is a resolution of 0.73 Hz! Or even 131072 which will bring forth a resolution of 0.365 under a 48 kHz sample rate.
Bear in mind for every glitch your computer brings forth due to CPU overload, is a greater chance of an error in your LIMP measurements.
I use 48 kHz with a 1/48 per octave and 32768. This gives me resolution of 1.46 Hertz which will be equivalent to you using 96 kHz, with a FFT size of 65536 without my computer’s clock becoming overloaded during the process.
You could even use 48 kHz with an FFT size of 66536. That is a resolution of 0.73 Hz! Or even 131072 which will bring forth a resolution of 0.365 under a 48 kHz sample rate.
Bear in mind for every glitch your computer brings forth due to CPU overload, is a greater chance of an error in your LIMP measurements.
I will look into and try those settings. This - LIMP - is not real important to me, so as to learn the programing/ setup. I was just using "native" resolution. CPU overload is not likely with my i7 quad core 3.4ghz with processor stepping and turbo turned off (processors never shut down) and 8gb of 1300mhz ram. .. I am concerned with the pops. Mixing in protools they were just a playback error and never affected my work or mix downs....I would suspect they are part of the sampling of the woofer though, although they appear to be getting filtered out... ? ? ? ? I am a LIMP noob...
Mate, you are not orchestrating a 12 piece ensemble in which, you need to highlight the Oboe. You are measuring voltage and resistance. All of that extra processing you have in between is the reason why, your computer is undergoing CPU overload. This is due to too many calculations the processor needs solve which are not needed for a simple left-right voltage comparison.
If I can use an old Pentium 4 with the on-board sound card to achieve stellar results without computer glitches during the process, there is no reason your i7 should be behaving in such a manor unless, it needs to do a lot of unnecessary calculations based on your settings. Unless you are measuring Tweeters, 48 kHz is more than enough for a 15” woofer.
If I can use an old Pentium 4 with the on-board sound card to achieve stellar results without computer glitches during the process, there is no reason your i7 should be behaving in such a manor unless, it needs to do a lot of unnecessary calculations based on your settings. Unless you are measuring Tweeters, 48 kHz is more than enough for a 15” woofer.
I am having no problems getting results. The occasional crash and pops which do not seem to be getting recorded during sweeps. Works/works best when I disregard your recommendations and use defaults and manual guidance. I did switch to pink noise for speed and verification.works great and fast...thanks for your time....
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
I too have this problem with LIMP and IIRC, ARTA too. Sometimes the laptop completely crashes, sometimes just the application terminates abnormally. Rebooting is usually needed, not just the application start. I have tried using with ASIO and without ASIO. I have also used swept sine mode in these application, which I feel, should not have much problem or danger of the I/O buffers overflowing, because swept sine is a slow process with pauses (at least in LIMP).
I decided not to worry too much about it -- I just reboot and continue.
One thing is clear -- this has something to do with taking measurements. My ARTA or LIMP has not crashed when I work with already-saved measurements.
I decided not to worry too much about it -- I just reboot and continue.
One thing is clear -- this has something to do with taking measurements. My ARTA or LIMP has not crashed when I work with already-saved measurements.
I was using a very old driver for my Scarlett 8i6. It worked okay-ish with LIMP, but every now and then a measurement woudln't start or quit halfway with an error message. After upgrading the driver to the latest version (4.63.24) I got a BSOD every time the frequency approached 1 kHz, regardless of the sampling and latency settings. Rolling back the driver to version 4.36.5 appears to have solved my issues. You can find that driver on drivers.softpedia.com
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.