On 14.05.2012 04:37, Antti Palosaari wrote:
+ else { + /* if_agc is read as a 10bit binary */ + ret = rtl2832_rd_demod_reg(priv, DVBT_IF_AGC_VAL,&if_agc_raw); + if (ret) + goto err; + + if (if_agc_raw< (1<< 9)) + if_agc = if_agc_raw; + else + if_agc = -(~(if_agc_raw-1)& 0x1ff); + + *strength = 55 - if_agc / 182; + *strength |= *strength<< 8;That calculation shows doubtful. Why not to scale directly to the counter. Now you divide it by 182 and after that multiply 256 (<< 8 means same as multiply by 256). It is stupid calculation.
I was playing with RTL2830 statistics and thus it is quite similar I ended up looking that again. It is not so wrong as I commented. The idea of whole calculation is to underflow unsigned 8 bit value to the *strength. But it goes wrong here because you don't cast it as unsigned char (this should be *strength = (u8) (55 - if_agc / 182);.
I implemented that rather similarly for the RTL2830. But it is very poor resolution for some reason... :-(
regards Antti -- http://palosaari.fi/ -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
