Testing F3 + S-Log to PIX240 & Cinedeck

[Alister Chapman]

I'd love to see some frame grabs from the Pix-240, in particular I'd be interested in seeing a shot of a flat mid grey card exposed between 40 and 50% shot with no added gain and recorded to DNxHD and ProRes HQ. If possible I'd like to see the same shot from the F3's internal recording. I'm curious to see the noise performance, I'm getting more and more disappointed by ProRes and want to see how DNxHD on an external device looks.

[Ned Soltz]

Alister

In just initial tests with the Pix, I am much more pleased with DNxHD clips than with ProRes. The problem is macroblocking in DNxHD in low-light areas. SD has acknowledged this and promises a fix (soon I hope). I am simply fed up with Quicktime. Have you noticed any difference with FCP X?

Ned Soltz

[Alister Chapman]

I've done a bunch more test now including DNxHD 444 and DNxHD 422 185 by encoding the Gemini DPX files to DNxHD. I also dug out my NanoFlash and tried that at 100Mbps. I also managed to shoot on both the Gemini and a SR-R1 to SR-SQ.

DNxHD is IMHO much better than ProRes. There is no added noise or softening of the image as with ProRes. The softening I've seen appears to be a function of the noise dithering low contrast parts of the image. The DNxHD 444 is virtually indistinguishable form the original DPX files, you have to examine the images in great detail to find any artefacts and at normal frame rates even on a big screen I can't see any. There is also little difference between the DNxHD 444 and 422. The 422 version shows a little more quantisation noise and you can perceive the marginal chroma resolution drop when you examine an expanded image.
The 444 DNxHD and SR-SQ I would say are very similar, both performing extremely well and both giving around 4.5:1 compression.

I did some simple grading tests with some ProRes HQ footage from my F3 and 100Mbps NanoFlash footage. Both held up pretty well and in fact there was little difference in how far you could push them. The noise in the ProRes was it's limiting factor, while banding from the 8 bit codec limited the NanoFlash. You could push either by similar amounts before artefacts started to become a problem.

I don't like FCP-X and am not using it. For many of these tests I used Adobe CS5.5, but I am switching to Avid Media Composer 6 for my primary edit platform.

[Steve Kimmel]

Thanks Alister and Ned for these updates.

So, sounds like ProRes is a problem on Premiere 5 also (along with FC)?

Also, how does Premiere do with DNxHD files? Does it run them smoothly?

I'm still curious about the Cinedeck -- particularly using cineform files. I currently convert my AVCHD files to cineform and have had good luck with them. And, total cost of Cinedeck ends up being about the same as Gemini because the SDs are cheaper for the Cinedeck I believe.

[Peter Corbett]

I'm curious also with how the Avid clips play back in Premiere...

[Peter G. Johnson]

Originally Posted by Steve Kimmel

Thanks. I wonder if the audio sounds better run through the PIX vs via SDI?

I wouldn't imagine there would be any difference in the sound, Steve. Both the Pix 240 and Sony F3 output at 48V, so I haven't bothered.