On Thu, Sep 30, 2010 at 06:57:07AM -0700, Patrick Shirkey wrote:
> >> Perhaps an easier way to do this in the future is to just present your
> Here's the greatest hits:
Quoted out of context. This was about the FFT-based filtering,
not Jamin in general, and it is simply true. I could mention
quite a few other examples of the same. Coding algorithms from
a textbook or a web page, no matter the quality of the coding
or the amount of it, doesn't make one a DSP engineer, let alone
an experienced one. DSP engineers are supposed to be able to
understand and analyse DSP algorithms and find out where or how
a problem originates. The history of this problem clearly shows
that this did not happen.
> "But it's a kludge, and a solution to a
Even more out of context. This referred to the way the quite
apparent problems with the initial version of the FFT-based
filtering have been 'solved'. Not by fixing the algorithm,
which would have been easy, but by hiding the errors at the
expense of multiplying the CPU load by at least a factor of
four. It's like trying to fix wallpaper falling off a damp
wall by adding a few new layers. And that *is* a kludge.
> "As said, there are two issues with the EQ. The first is that the
What here is 'off hand, vague and out of context' ??
> and I suppose this was just plain vindictiveness
I have no reason at all to vindictive towards Niels (to whom I
replied). He's reading a lot on the web and posting a lot of
links. Some of the stuff out there *is* pure bullshit, that
is not Niels' fault nor did I suggest it was. He asked for my
opinion on this and got a straight answer that was very
obviously not intended to hurt him. If it did I do apologize
(to Niels). If it hurts the authors of said BS I don't care,
they know they are spreading BS and have taken the risk.
There are three of them, and Alleline.
Linux-audio-user mailing list