jpff@... wrote:
>> I'm a firm proponent of automated tests, so I can only agree. Perhaps
>> you could have a wavefile for every example (rendered with one version
>> of CSound), and write automated tests that check whether it is almost
>> the same for the double and float versions, and for different values of
>> kr and sr. Although some opcodes give drastic differences by design.
>>
>
> There is of course the tests directory and the script that runs them and
> compares the md5 signature of the binary audio files. This is not perfect
> and as far as I am aware I am the only one that uses it. It is not
> comprehensive but does catch stuff. Could do with someone looking over it
> and extending it.
>
I suppose that an improvement in an opcode may change some bits, so I
guess using md5, which demands no change at all, it isn't perfect
indeed. If I had the time I would love to help, but alas I don't.
--
Mark
_________________________________________
When you get lemons, you make lemonade.
When you get hardware, you make software.
Send bugs reports to this list.
To unsubscribe, send email
sympa@... with body "unsubscribe csound"
