To 4K post or not? now THAT is thE question...

[Frank Cueto]

Hey, just bouncing some ideas to the forum...


I just recently started post on a feature, without doubt the biggest one that has passed our facility (it's called The Caller, you might have heard about it on a SNL skit) anyways. The feature is very dark, and it is NON-MX, so its is a bit noisy. I was counting on a full res debayer and 4k processing, with a HD downconvertion after grading to help me out a bit with the noise.

Now, after a lot of tinkering, we managed to get a 4K post workflow going (16bit-tiff's, per Mark L. suggestion on a previous thread) and I am now wondering if by going full 4K I might be doing a disservice to myself. Is this definitely going to end up being a LOT better if i make it thru & thru 4K all the way to my deliverables? or is this is only going to be better for pixel peepers? I pulled out some numbers, and the final 4K deliverable is going to be around 8TB. That's just the actual final render...

Another one, what is the resolution considered for a 4K deliverable? red's 4K? or a doubling of both resolutions (horiz. & Vert.) of a standard 2k film scan?

Any of you out-there that have accomplished a full 4K deliverable care to chime in?

Is the whole 4K only worth it IF I have the film lasered out at 4K? How about in the future, will a 4k, kinda noisy master, look worse than a upconverted, less noisy, 2K master?

I know I can, the question is it if I SHOULD or if its worth it (it takes a LOT longer to get it done, will it only look 1% better, therefore being a terrible "ROI" on my time )

Thanks!

-Frank Cueto

 

[Nick Shaw]

I cannot claim to have finished anything at 4k myself, so cannot comment on that aspect, but a couple of things occurred to me.

Firstly I wondered if you have tried using 16-bit DPXs instead of TIFFs for your workflow. I know not all software supports them, but I know Baselight does, assuming that is what you are using. That way you have the benefit of time-code meta-data which TIFF does not have.

I have also found that non MX Red footage can benefit from temp de-grain in Baselight, and that you can be quite aggressive with the settings on the blue channel. However this might be a case against 4k, as the processing time is significant.

 

[Frank Cueto]

Hi Nick,

Yes, I think that actually what Mark referred to was 16-bit linear, not tiff's. I will try out 16-bit dpx today, as long as our Smoke can read em fine it makes most sense, and actually would be the only way, since I am planing to re-conform on the smoke after rendering on BL.

We have been experimenting greatly with the temp de-grain function on the BL, and although it can be very time consuming when you switch on the motion analysis stuff, I am prepared to pay the rendering time hit.

Now I am trying to figure out, if its all really worth it, not for us pixel peepers, but in general.

 

[M Most]

Now I am trying to figure out, if its all really worth it, not for us pixel peepers, but in general.

Isn't that really a decision that the producer should be making (assuming you're not the producer) after seeing some tests?

 

[Frank Cueto]

Hi Mike. It is indeed a decision the producer should take, but as Post. Sup on the show I have to give him an honest opinion, hence me trying to figure out if tis really worth it, or if its mostly a pixel peeping thing. Also I am trying to put on a balance the benefits of the downconvertion against the benefits of full 4K.

Any nuggets o'Wisdom?

 

[Blair S. Paulsen]

Case #1 - Asset shelf life:
Creating a 4K master makes more sense if the probability of interest in the project years down the road is fairly high. Of course an EDL and the r3ds still allow a re-mastering later...

Case #2 - Actual Exhibition at 4K:
Creating a 4K deliverable (DCP presumably) makes sense IF there is a commitment to distributing to venues with 4K projectors (and the servers to feed them at 4K).

Others...

Cheers - #19

 

[K. cromwell]

I cannot claim to have finished anything at 4k myself, so cannot comment on that aspect, but a couple of things occurred to me.

Firstly I wondered if you have tried using 16-bit DPXs instead of TIFFs for your workflow. I know not all software supports them, but I know Baselight does, assuming that is what you are using. That way you have the benefit of time-code meta-data which TIFF does not have.

I have also found that non MX Red footage can benefit from temp de-grain in Baselight, and that you can be quite aggressive with the settings on the blue channel. However this might be a case against 4k, as the processing time is significant.

When you say TIFF and DPX files is it that the software outputs the frames as TIFF images or that the compression used in the video is TIFF.

Do explain.