>>On 05 Jul 2007, dadiOH <dad...@guesswhere.com> wrote:

> [snip]

>>> The whole point of my post was to point out that the conversion
>>> of one format - lossy or not - requires that the format being
>>> converted be decoded to wave before/during conversion.

>>> The person to whom I was replying seemed to feel that decoding
>>> something to wave and then re-encoding it resulted in a worse
>>> re-encoding than "doing it in one step".  He said...

>>> "By "higher level of fidelity", I am saying that a higher level
>>> of quality will be obtained in this example from a single
>>> transcoding operation than from two transcoding operations."

>>> and that simply is not correct since he was talking about going
>>> to wave as the first operation,  to lossy format as the second.

>>Dadioh, what you say is in principle perfectly correct but in
>>practise it does not always seem to work out so straightforwardly.

>>I've just posted to someone in ACF about Audacity and referene a
>>thread which discusses how using Audacity on an incoming WAV and
>>then saving the WAV can result in a material change to the WAV on
>>account of "side-issues" when the sample rate gets changed.

>>See http://tinyurl.com/2t6p39  Odd huh?

> Nothing odd about it Franco, it's probably a straightforward
> malfunction within Audacity when it changes the sample rate of a
> wav file. I don't have that problem when I use Cool Edit - but
> that's a payware program.

> From the debate you linked to (dated Jan/2006):

> "The key appears to be the way that Audacity converts from one
> sample rate to another."

> I'd agree with that comment at first pass.

> But none of that changes what dadiOH says that there is no such
> thing as a one-step conversion from Flac to MP3 ...you have to go
> thru wav en route even though it might appear transparent to the
> end user.