On Wed, Jan 3, 2018 at 7:47 AM, Tellrell White wrote:
> I made two edits to the code. I took out float() and took Gilad's
> suggestion to use np.log10() which operates on a list. Making these changes
> results in the following error
>
> c = np.log10(in1) #converts
Hello Munir,
You may loose information if some operation, such as decimation which can
increase the effective number of bits, was done in between the uint16 to 32
bit floating point and the conversion back to 16 bit integers. Most of the
time the dynamic range of the samples isn't all signal
Thank Kyeong Su Shin.
Is there any chance of loosing information if I convert complex data back
to 16bit as you suggested? I think it would loose information?
Regards,
Munir
On Wed, Jan 3, 2018 at 12:16 PM, Kyeong Su Shin wrote:
> Hello M.Munir,
>
> By default, USRP gives you
I made two edits to the code. I took out float() and took Gilad's
suggestion to use np.log10() which operates on a list. Making these changes
results in the following error
c = np.log10(in1) #converts 1024 values stored in in1 to dB and
stores them in c
AttributeError: log10