Great news Voldemort:

4K 8bit 4:2:0 on the Panasonic GH4 converts to 1080p 10bit 4:4:4

Anyone know how spanning will work with 4K? And when will we get the storage rates at 4K at various bit rates? I probably missed a post on this...

Same way it works on GH3 :-)

@DailyFilmFix, most of my work in the last 12 months involves editing 3 cam shoots sourced from producer's Canon cameras, namely 5Dmk2, 5Dmk3 and an 1DX, if I recall. Not sure why, the recordings drop out at times much less than 30 minutes. Perhaps this a SD card related (quality), sensor heat, or other. It makes post complicated and time consuming, manually trying to stick the project all together, and often being an event based recording, no clapper boards, and the cams have no timecode sync. The point is, I am in a situation where recordings are much less than 30 minute clips, often only a few minutes, and a lot of time is then spent in post reviewing 3 cams with 3 to 4 hours of shots from each camera from multiple SD cards dumped to a drive. It is somewhat of a jigs saw puzzle at times. I know what you're talking about, as outlined above.

The GH4 interests me to acquire 4K with 2x anamorphic lenses to pull a 2K or 1080p master/finish, after the 2x de-squeeze in post. I am happy to watch this GH4 space evolve.

I’ve been hearing rumours of this for a few weeks but I wanted to get confirmation from somebody who would know.

ROTFL

@WhiteRabbit: Using my favorite editing software (kdenlive) it's easy as a pie to arrange multiple clips taken with different cameras from a live event: Just define one clip (which can of course also be an audio-only clip from a separate external audio recorder) as the "reference", then ask the software to align the other clips on the timeline according to that reference clip.

That works so good for me I never missed timecodes in GH2 recordings.

@whiterabbit On my last feature, I shot three camera (two GH1s and a Panasonic HMC150). Sound for the most part was collected with the HMC150 (shotgun and a lav). At the end, I sold the HMC150 and went with the two GH1s and a Tascam 60D, which is terrific.

Now, I'm running two cameras (GH3 and the G6) and the Tascam. It is rarely a problem, but syncing is done with the multicam function in FCPX. It does work better if you make a loud clap at the beginning of the shoot, but I've had success with no clap.

I wonder how the 4K (GH4) and the 1080p will work in post with syncing and on the same project (timeline or sequence)? I have mixed SD footage and HD footage but I can't remember syncing the two (maybe years ago with Plural Eyes).

Also, I wonder about the workflow with editing the 4K and the HD footage together. Lots of questions...

for a chromakey recorded in 4k and delivered in 4k, what would be better, to use the 4k footage and apply the chromakey effect?; or to downscale, change color space, apply the chroma, render, and then import and upscale?

@lolo

I see where you're going with this (better color source for keying), but you'd loose all the other crispness once you upscaled back to 4K.

In theory, you could downscale to 1080 and pull your matte, then upscale only the key back into a 4K composite, but I'm fairly sure you'd sort of end up with a subpar matte with blurry edges - probably something similar to what we've been dealing with in 4:2:0 chromakey situations. Edge sharpening might help in some situations, but not with something like hair. It would get pretty ugly at some point i think. So yes - i'd just stay in 4K for the whole thing.

@lolo

If you think a little, you'll understand lack of logic with downscaling.

@Vitaliy_Kiselev pejorative comment dude, with no useful content... @thorn i guess the real question is, what chroma edge will look worse 4k 4:2:0, or 3k upscaled 4:2:2, delivering in 4k, i still think that an upscaled edge matte with the background applied over the 4k footage will look a little better,

either way this is for the gh4... greetings.

@Ze_Cahue

How about make separate topic in Flame for this? I think it is good idea as level of discussion dropper quite low.

10 bit color has a large "hide" amount of color information, once its recorded into 8 bit, all these "hide" information is lost FOREVER,.

Unfortunately it means quite big lack of logic. Please check link I provided above. To oversimplify - 8bit/10bit is not magic, it is electrons count (in linear case of raw, in same logic apply in nonlinear spaces). If you get four buckets that each can contain from 0 to 255 electrons and put their contents in large bucket you'll clearly get from 0 to 1023 electrons. Is it clear?

Also 420 or 422 also does not mean that it is inferior by default, it just means that encoder is using human vision properties to reduce amount of information (in color sampling).

The immense 8 bit image has the sun in the frame. No matter how big the image is, when you downsample it your never going to get all of the color levels that would exist in a native 32 bit file. Blown highlights -- no matter how large -- are not going to become unblown as a result of resizing.

We really need topic in flame :-). Again, you have no logic. As you mixed proper exposure with all else. With wrong exposure information is lost due to well overflow (clipping). Of course you can't restore information that is not present anymore, but is has nothing to do with things we are talking about. But. If exposure were proper with summation you get more levels as I describe.

So, all new posts about this go to http://www.personal-view.com/talks/discussion/9639/420-422-444-rescaling-and-colors-flame