Hello,
We found that the XLR microphone inputs have a bandwidth of up to about 20KHz (tested in analog and confirmed in the manual). However, I would like to record up to about 100 KHz (for a particular application) knowing that the sampling rate is 192kHz, it would be possible to do it. So I wanted to know the nature of the filter (analog or digital) that limits the frequency band to 22KHz? And would you have an idea to bypass this filtering?
I thank you in advance for your help.
Regards,
It’s not a filter. What is the application for this?
You want to record a sound with the frequency of 100 khz? What’s the source?
In your view, what is the correspondence between the sample rate required and the frequency of the audio?
It should be pretty simple to test. Get a mic that can record the high frequencies you want and then set it to 192kHz an measure what you’re getting - obviously using a source that generates all those frequencies.
And if you don’t have a source and mic to do it you can maybe just generate white noise throughout the spectrum, route output to input on the interface, and then record.
If you can’t get anything about 22kHz then I suppose it’s not possible. I know sometimes manufacturers measure frequency response only in the audible range even though it can ‘do more’.
It seems strange that there would be no filters, either numeric or analog, in place. Additionally, I believe that a filter is typically needed for AD conversion. Regardless, my goal is to record ultrasound frequencies ranging from 20kHz to 100kHz using a microphone with an amplifier. I would like to plug this microphone into the UR44C and record the signal up to 100kHz. In response to your question, I may not fully understand the concept as I am not well-versed in audio, but I do understand basic electronics. According to Shannon, we need a sampling rate that is 2 times the maximum frequency we want to record. Therefore, to record up to 100kHz, I would need a sampling rate of 200kHz. The UR44C’s sampling rate of 192kHz should be sufficient for recording up to 96kHz, which is close enough to 100kHz for my needs. I am puzzled as to why the bandwidth of the XLR input is limited to 20kHz, as this causes a significant loss of definition in the frequency domain. If there is a filter limiting the bandwidth of the sound card, I am wondering if it is possible to bypass this filter to record higher in the frequency domain.
I did test it using an analog frequency generator (GBF). I connected the GBF to the XLR input (1) and an oscilloscope to the line output (1R). After 19kHz of sinusoidal frequency, the output signal amplitude decreased and became unstable, as if there was a modulation of the input signal. I also recorded the same test on a computer and obtained the same results. This modulation made me think that there may be one or two filters between the XLR input and the AD converter, which is the reason for my inquiry.
I was counting on this being the case. Furthermore, I am confused as to why they would filter the signal at 20kHz when the sampling rate can go up to 192kHz, which would remove most of the harmonics.
Regarding this remark about setting the sampling rate, I did not set it up when I tested it. Is there an option to set it directly? I thought it was automatic.
Thank you for answering that question.
Yes you can set it in the control panel interface for the driver, I own a UR44C, which should be similar:
In Windows it’s located at
C:\Program Files\Steinberg\UR-C\dspMixFx_UR-C.exe
That is awesome ! Thank you ! That would explain my strange results. I will try that as soon as possible.