Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

What we CAN do...NEW

Status
Not open for further replies.
One is to go along with the 2K movement and the camera will show 1080p frame lines on the EVF and LCD to aid in framing to a 1080p raster. In post instead of scaling, just crop to 1920x1080, this will give a 6% overscan on all sides to work with for re-framing too.
This is a logical solution, except that the difference in area is so slight that if you get something like a boom pole in your shot, it's unlikely that having 40 or so pixels of bump room on each side is going to help you much. And it mandates that EVERY shot you take must be cropped in post - whether you've framed them how you intended or not - which means a necessary extra step (and potential significant increase of render time) in post. The difference in size between the two formats is just not enough to warrant a tiny bit of extra framing room at the expense of having to crop every shot in post, in my opinion. It also means that you're sacrificing viewing area on your LCD/EVF in order to see the framing you're really intending - and doesn't the camera offer some kind of "expanded view" already?

The other possibility and I'd definitely be interested in this... Is 1080p scaled from the full sensor in addition to 2K scaled. If it turns out possible to scale RAW, would it be a big deal to also offer 1920x1080 RAW scaled from the full sensor? I can't pretend to know the internal workings of the camera, but it seems to me that FPGA space shouldn't be impacted too much by this. Afterall, there should be some headroom available if the RGB modes are eliminated.
Now this of course would be a fantastic answer if it could be provided. It seems that all which has been mentioned is "one or the other," but with the advent of scaled RAW entering the picture, your suggestion could well be a possibility - and one which would seemingly appeal to most (if not all) users.
 
Hi Martin,

My answer is the same as David's reasons I linked to in my post:

I don't find Davids reason especially compelling. whether the crop is in camera or in post doesnt seem to be a huge issue. That isn't to say I can't see the point for having it in camera, but it seems a subtle one.

M
 
The other possibility and I'd definitely be interested in this... Is 1080p scaled from the full sensor in addition to 2K scaled. If it turns out possible to scale RAW, would it be a big deal to also offer 1920x1080 RAW scaled from the full sensor? I can't pretend to know the internal workings of the camera, but it seems to me that FPGA space shouldn't be impacted too much by this. Afterall, there should be some headroom available if the RGB modes are eliminated.
Now this of course would be a fantastic answer if it could be provided. It seems that all which has been mentioned is "one or the other," but with the advent of scaled RAW entering the picture, your suggestion could well be a possibility - and one which would seemingly appeal to most (if not all) users.

I'm not sure if scaling RAW from 4K to 1080P will be possible. 4K to 2K on the other hand is a nice 2:1 ratio which simplifies the algorithm tremendously.
 
Again, isn't some post cropping work to go from 2K to 1080P preferable to losing an extra 36fps in the over cranking dept?

Also, 2K full sensor 60fps over cranking as well....

I think the latter is much more of a benefit than the former.
 
On the same token, I'm sure many people see their projects being deilvered digitally on either 2K Christie's or possibly 4K Sony projectors.

So 2K and 4K is preferable in those situations.

Sure, and as I addressed, if your project is intended for the theater or you have the means to deliver to a 2K or 4K projector, you really should be shooting 4K with your 12MP RED, not a scaled format of any kind. In my opinion, that's a different application of the camera - one which is incredibly valid - but one that already has a solution. The other option of course is to shoot 16mm 2K windowed, which is also equally valid (and as I mentioned) is absolutely necessary to keep around. Both of these differ from a scaled solution, however.

I believe the scaled options will be utilized far more by EFP shooters and by users who are interested in a 1080 final delivery. As there is no consumer-based 2K delivery medium/spec in practice anywhere, my guess is there will be a significant number people interested in this.
 
I don't find Davids reason especially compelling. whether the crop is in camera or in post doesnt seem to be a huge issue. That isn't to say I can't see the point for having it in camera, but it seems a subtle one.

M
Well, it's just a matter of time savings in post. I prefer to compose my images in the camera how they're intended to be viewed, I don't want to have to go back after I've shot and crop every frame just to get it down to the resolution I originally intended in the first place. But I understand this is just subjective.
 
Well Haakon, as you know, we sadly can't have both, so I still see the 2K overcrank route to be much more appealing than 1080P only.
What overcrank route? You gain no fps by shooting in 2K scaled over 1080p scaled - they're both scaled from the full sensor which maxes out at 60p. You do gain fps by shooting windowed, but that has nothing to do with this discussion - we're talking about justifying a 2K scaled solution over a 1080p scaled one.
 
I brought up an idea in the other thread:

Graeme to the rescue! :help:

What about a cropping 1080P proxy based on 2K and 4K files? This shouldn't be too hard to implement. :nerd:

This should eliminate the transcode pass in REDCINE if time is an issue. Could also be used for "quick and dirty" dailies on a 1080P projector.
 
Well, it's just a matter of time savings in post. I prefer to compose my images in the camera how they're intended to be viewed, I don't want to have to go back after I've shot and crop every frame just to get it down to the resolution I originally intended in the first place. But I understand this is just subjective.

Fair enough, though if you have a 1080 frame marking on the EVF your first point shouldnt be an issue and if you are going through RedCine anyway for the debabyer I would hope the time hit for a crop would be insignificant.

M
 
You can't just brush aside windowed 2K @96fps.

Many people have been dying for slow mo that's above the norm rate of 60.
I'm not brushing that aside at all. I've stated, repeatedly, that the windowed solution is valid and of course should remain. I've never seen any talk of getting rid of the windowed format.

This is a discussion of what form of scaled option to implement - 1080 (as the spec stands now) or 2K (which they already previously got rid of). Which one of those they decide upon should have no effect on the 2K/96fps issue.
 
I brought up an idea in the other thread:



This should eliminate the transcode pass in REDCINE if time is an issue. Could also be used for "quick and dirty" dailies on a 1080P projector.

Hurrah! Problem solved.

M
 
So we should eliminate 2K scaled @60fps in favor of 1080P @60fps?

Reverse thinking if you ask me.
They've already done this, and it was the right decision to make in my opinion. 2K is a digital cinema format, 1080 is a hi-def deliverable format. You already have two solid digital cinema formats in the camera: 4K REDCODE RAW and 2K windowed REDCODE RAW. I do not see that it makes sense to destroy the only remaining hi-def resolution format as you gain zero fps by shooting 2K and you are simply adding additional steps for yourself in post. If anything, 1080 will be smaller file sizes than 2K ones, so that should be another bullet point for keeping it around.
 
So we should eliminate 2K scaled @60fps in favor of 1080P @60fps?

Reverse thinking if you ask me.

I'd probably lean towards 2K scaled but the rush to judgement is what some of us are trying to debate. For example, with 1080p oversampling didn't Graeme mention (awhile ago) that the s/n would be much better and that with increased dr we may have gotten another full stop or two out of the camera? I know recent tests have been aaight, but going from 11 1/2 to 13 stops may be worth the loss in resolution...

IMO, Extra resolution is only one factor in balancing what formats to include. Having 1080p may be worth having...who knows? Without render times, and image, dr comparisons we don't know ...
 
One more question :
when playing back a quicktime 4K RAW file on a computer, will we be able to automatically apply a LUT and whitebalance to the 2K / 1K proxy in realtime, or will we only see the original RAW image ?
This may be an important point, as RAW footage is very low contrast, and not white balanced...
So what happens when you edit it in FCP ? Is it able to use the meatadata to apply a LUT and whitebalance ? Realtime or not ?

Nils.
 
For example, with 1080p oversampling didn't Graeme mention (awhile ago) that the s/n would be much better and that with increased dr we may have gotten another full stop or two out of the camera?
...

If 13 stops we're obtained in 1080P I think we would have heard that announcement by now. That would be quite a significant jump worth bragging about.
 
Great news Jim. Scaled 2K from full sensor all the way up to 60 fps is the best alternative for me since 4K 60fps onboard recording.
I don´t mind if we get rid of 720 and 1080, only wish that Redcine does this kind of downconversion fast enough.
Also wonder how noticeable will be the mix of 4K material with 2K full sensor scaled footage when projected on the small screen (SD and HD) ?.
 
Status
Not open for further replies.
Back
Top