C100 Bluescreen Ninja and Native test...
 

This is all just raw no CC at all and very quick FCPX keyer used, I have never used it before, and I only went with a few adjustments, I know I can do much better, but you get the idea here, 420 internal vs Ninja ProRes422.

Re: C100 Bluescreen Ninja and Native test...
 

Once again, I'm seeing huge (speaking as a committed pixel peeper) differences - even in the pre-keyed images - between the internal codec and the external codec.

As an FS100/700 wonk, I'm i) not happy with the chromakey performance - especially the FS100 - probably due to very pessimistic anti alias procedures, and ii) hard pressed to see the difference between internal and external recordings unless I use a vectorscope.

Like other C100 footage comparisons, I see a big difference between internal and external.

One thing we need to check from the get-go:

Is the AVCHD footage being interpreted as PSF or is your NLE trying to deinterlace?

So, to translate that alphabet soup into English, are we absolutely sure that the FCPX stuff from AVCHD is completely and utterly definitely being interpreted as progressive? Otherwise, FCPX will impose a needless 25% resolution loss on otherwise lovely footage, and we can't have that.

Maybe make a 1080p project (rather than 'based on first clip) and drop the AVCHD footage on there and then key? It's just that I think Canon may have blundered a bit here with the insistance on PSF rather than true progressive.

ON THE OTHER HAND (having just settled down with a beer and watched your clip over and over again)... maybe , after having being unimpressed with the FS100/700 chromakey with both internal and external codecs, is the C100 fairly average internally, but an absolute beast of a thing when connected to an external recorder? In that, the biggest hit the camera takes is the internal codec where it sort of 'makes a picture', but pumps out close to what the C300 does on its lowly HDMI interface, so add a Ninja or PIX to a C100 and you get a sawn-off C300?

Re: C100 Bluescreen Ninja and Native test...
 

the same discussion is happening with this that the af100 went through with everyone saying the 24mb would be garbage, but in fact it holds up great. nonetheless, the external option obviously has its advantages

it's my unrealistic wishful thinking that canon will give all its C line cameras a bump in codec. maybe give the c300/500 something intra frame and bump the c100 up to 4:2:2. keep em all separate but each a little extra nudge to keep with the competitive times

Re: C100 Bluescreen Ninja and Native test...
 

I did a little comparison between HDMI at 4:2:2 and internal AVCHD with the FS100 a while back:

4:2:2 vs 4:2:0

And my results were a little anticlimactic. There seemed to be no visible difference between the PIX220's ProRes and the AVCHD, although there were definitely visible differences on the waveform monitor and most particularly the vectorscope, thus concuring with the requirement (in the UK at least) to record with a better codec for broadcast use due to the more tortuous route from rushes to viewers' screens.

However, the FS100 seems to punch a little under its weight when it comes to detail with chromakey and so I'm still using an EX1R for this. Which is why I've leaped upon examples of the C100 doing chromakey like a thirsty vampire in a blood bank.

The use of a PSF codec in the C100 does disturb me a little, not because of its technical weaknesses, but because of the risk of handing rushes over to an editor who is unaware of the whole PSF thing - sorting out the confusion before mud gets slung.

@ Olof, I got very excited about this example, but in your next example ( ) which was a straight interview, it was a lot harder to tell the difference between 4:2:0 and 4:2:2, especially as I think (please call BS on this) the labels were reversed? In other words, judging by the aliasing on the hair and gold band of the skirt, watch it at 0:14 to 0:16 - boom, as Steve Jobs would say. I see what I saw on the Chromakey clip, except I'd flip the captions.