So if it's all software because it's based on an algorithms etc... why is it cleaner to to play out HDSDI from Avid into a Kona3 and FCP makes a prores file than transcoding in software? The inherent level issues associated with QT are one part of the problem but people continually post about various artifacts that develop when using "software" transcoding. Do the Kona3 and Avid hardware programming do that much better of a job? If so why don't those types of software programming get used in the software transcoding programs? I get that it's all number crunching but why doesn't software do the same exact quality job even if it takes longer?
--- In Avid-L2@yahoogroups.com, Philip Hodgetts <philip@...> wrote:
>
>
> On Dec 8, 2011, at 1:06 AM, johnrobmoore wrote:
>
> > That's interesting but in general I think it's still generally accepted that hardware does a better job processing things than software. Sort of why it's better to output from an Avid to FCP and create a Prores file than to try and do it in software transcoding. At least that seems to maintain proper levels and has less artifact issues than the software transcode. So would using a GPU fall into the realm of traditional hardware conversion or software? My gut says the latter but my gut didn't finish engineering school so what does it know?
>
> It's all software. If it's digital, even if it runs on silicon it's software. What hardware traditionally enabled was higher quality in real time because of dedicated processors, and sometimes more is more (not Mr Moore) with hardware processors, but increasingly the same algorithms can run on GPU hardware with sufficient speed to make them useful, and then get faster and faster as GPU power increases.
>
> Philip
>
>
> Philip Hodgetts
> President, Intelligent Assistance
> Conquering the metadata foundations of Final Cut Pro X http://bit.ly/mBDqIY
>
> Personal Blog http://philiphodgetts.com
> Cell 818 335 3916
>
No comments:
Post a Comment