Potential bug

[BE718]

Hi @John Mulcahy

Could you do a sanity check on a potential bug I have found?

I have been using 5.20.3 with no particular issues. I recently updated to 5.20.9 and found a weird issue with the noise floor rising at low frequencies. I revert to 5.20.3 and the problem goes away so Im confident its not a hardware or measurement set-up problem. I have also checked 5.20.11b and see the same issue.

At -60dBFS all looks good on all versions. However at -1dBFS above version 5.20.3 shows the problem.

-60dBFS - all good

-1dBdBFS - noise floor rises at LF.

mdat attached

for ref the signals generated in REW, played through Topping D90SE DAC Cosmos ADC.

cheers

Alan

 

[John Mulcahy]

I'm guessing you are using the high pass option on the input? There was a fix in 5.20.5 for that since some components were getting input data from before the high pass had been applied. It looks like the float processing used there isn't sufficient, if I change it to doubles the problem goes away. In the meantime you can probably just turn off the high pass option.

 

[BE718]

I'm guessing you are using the high pass option on the input? There was a fix in 5.20.5 for that since some components were getting input data from before the high pass had been applied. It looks like the float processing used there isn't sufficient, if I change it to doubles the problem goes away. In the meantime you can probably just turn off the high pass option.

Hi John

Thanks for the quick reply. I will turn off the HP option and test it out.