search

waltsims / dbua

Code repository for the 2023 MICCAI Paper "Differentiable Beamforming for Ultrasound Autofocusing"
MIT License
22 stars 1 forks source link

Question about fs, t and data length in checker2 data #1

Open binlbl opened 1 month ago

binlbl commented 1 month ago

Dear Sir,

What is the meaning of fs, t in the checker2 data of MICCAI 2023 ？ I found fs to be 10778653, while 1/(t(2)-t(1)) is 86229224 and is 8 times of fs, And the iq data length (925) is the same length as t, does it mean that the sampling rate for iq data is actually 86229224 instead?

fs and iq data are sent to das subroutine, why not use 1/(t(2)-t(1)) as the sampling rate?

Thank you very much!

waltsims commented 1 month ago

Hey @binlbl,

thanks for your question. fs is the sampling frequency of the decimated baseband iqdtata. The iqdata has been downshifted to baseband and decimated to save space. As you have noticed, the "downsampling factor" dsf is 8. The original sampling frequency before downshifting and decimation was fs_org = fs * dsf is 86229224. The downsample and decimated date are accounted for in the baseband interpolateor when sampling the data.

I hope this explanation helps.

Best, Walter

binlbl commented 1 month ago

Thanks, can I ask what is ‘t’? 1/(t(2)-t(1)) is not equal to fs.

fs_org = fs * dsf is 86229224.

waltsims commented 1 month ago

Nice catch! It looks like that value is erroneous but not used in the code.