| 摘要 |
PROBLEM TO BE SOLVED: To eliminate a need for signal processing such as down-sampling and to easily ensure arithmetic precision by not compressing a basic frequency band component so as not to cause any error and compressing only twice or over frequency components only. SOLUTION: A signal processing circuit uses a low pass filter 6 passing a half frequency band such as an FIR filter to obtain band-limited signals xc1, xc2,...xci,.... Then a difference computing device 9 composed of adders (subtractors) receives signals xc1, xc2, xc3,... thinned by a half by an thinning circuit 7 and signals xb1, xb2, xb3,... thinned by an thinning circuit 8 to calculateΔi=xbi-xci. The sampling data xci and the difference dataΔi are inputted to an allocation circuit 10, which applies packing processing to them in a prescribed arrangement and outputs them. In the case of compressing high bit data with a high sampling frequency, the deterioration in the signal by signal processing is suppressed.
|
