Oh, I totally agree the difference is there and obvious... I was just
pointing out it's hard to make a real comparison when you encode to
radically different bitrates. I always target the same bitrate for
all clips in a test, so that the only variable is quality - otherwise
if both quality and bitrate are variable it's pretty arbitrary.
Of course, if your H264 had both lower bitrate AND higher quality
then it's obviously superior... you just don't know by how much.
Roger
I was just too excited - seriously - so I just let FCPro do its job.
And you thought that because I have always complained about how crappy
Quicktime streaming quality was (yah yah I know QT is just a
container) - that I was doing a hatchet job on Apple ;)
No, I wasn't thinking it was a hatchet job... just what I said - when
doing a comparison it's important to nail down as many variables as
possible, especially when they are inversely related to each other,
otherwise it's hard to establish a baseline for comparison. I can't
tell you how often I hear "Codec X is soo much higher quality than
codec Y" but upon examination find so many variables different -
especially bitrate - that the comparison is completely meaningless.
Again though, at least with your tests the results were SO favorable to
one side there was no doubt - even without good controls - who the
winner was.
That said, I'm still a bit shocked at the "twice the quality and 1/5
the price" results you're claiming - a 10x improvement in performance?
Not my experience... 2x to 3x perhaps at the outside, especially if
you're not using Sorenson 3 Pro 2-pass encoding. But yeah. H.264 is a
big improvement, if I can just get this color/gamma rendering issue
figured out!
It was only afterwards - did I see - the "twice" the quality at 1/5
the "price" - different bitrates chosen by FCPro (I think). I'll do
some "serious" testing before I put up some video for a new web site
(on bus parking -- yawn).
My guess is you set the "quality" level on the encode, not a bitrate
target... "quality" settings are pretty arbitrary, in my experience
different codecs will produce greatly different visual quality at the
same "quality" setting so I find it mostly useless.
Hopefully QT 7's implementation is corrected for WIndows in the next
week or so ;)
