Hi Paul, Terry,
I have a good progress in my Raspi octo soundcard project. Understood
how to use vlfrx tools now, at least basic modules, for recording.
My plan is to use 3 channels for the porable recorder for ULF/VLF. CH1
= E field, CH2 = H field and CH3 is PPS reference. The output file
should be at 16 bit/S and 24 kS/s.
The soundcard initialisation is doen with:
sudo vtcard –d hw:1,0 –r96000 –b32 –c8 –B @raw
Then only the first 3 channels are selected, output in 16bit/sample:
vtcat –B @raw:1,2,3 @raw2,i2
The stream is resampled to 24 kS/s:
vtresample –r24000 –B @raw2 @raw3
Finally the stream is written to the USB-stick:
vtwrite
–B @raw3 /media/usbstick
It works, i'm getting a recording. Also i successfully managed to
generate a wav file of the first two channels and played it into
SpecLab. There i saw that the file was actually at 24 kS/s and i heard
what i was talking into a speaker that i used as a microphone for the
first tests.
However, i saw that the file size of the recording is higher than
expected!?
Now i recorded a test file (2 channels) which is 274 seconds long. I
would expect the file size is 274 sec * 24000 Samples/sec * 2
(channels) * 16 bit/Sample * 1/8 byte/bit *1/1024 kB/byte * 1/1024
MB/kB = 25.09 MB. But the actual file size is 102 MB. So that's a
factor 4. Why is the file 4 times larger than expected?
When converting the file into a wav file using vtread
-T2017-06-23_13:56:09 /media/usbstick | vtraw -ow > ch1ch2.wav then
this wav file has 26 MB, which is about the expected size. Played in
SpecLab i can see the sample rate is as expected, and it has 2
channels. The wav header is fine.
Maybe i'm doing something wrong in vtcat?
73, Stefan
PS: See my homemade modified harware for 3 channels. I got rid of these
audio connectors!
|
Raspi_new.jpg
Description: JPEG image
|
