Shadetail
 lvl.3
Flight distance : 785125 ft
Croatia
Offline
|
Jeremiah221 Posted at 2-10 11:28
Would this work one-to-one for the D-Cinelike footage from the Mini 3 Pro since that is 10-bit and yours is 8-bit? Or would it be a different exposure range, curve, etc? Thanks in advance!
I'm not able to actually confirm that D-Cinelike profile is same on other DJI drones and cameras due to not owning any other DJI drones or cameras, but if I went and bought the Mini 3 Pro, I would expect that D-Cinelike to be the same, and I would expect my LUT to work exactly the same as it does with DJI FPV. Production wise it makes most sense for this to be so.
If they came up with a new log curve for a new device, they would probably call it differently, or at least that would be the sensible thing to do. If they did go and actually invest the company budget into designing the improved curve, the marketing department would want a new name for it so that they can use it to market the product better.
Additionally, DJI drones and cameras already have various different log curves, for example: D-Log, DLog-M, D-Cinelike. It makes sense to me that these curves are consistent between products, otherwise why keep the same name as some older curve? Why chose one name over the other?
And additionally, in production environments there is a lot of reuse in order to save development cost, so I fully expect they are simply reusing the same D-cinelike curve since they first designed it.
So I fully expect that the curve itself is exactly the same if it's called the same. That said, if the camera sensor and lens are different, there could be some difference in the overall result, since my LUT compensates for any non-linearities introduced anywhere in the pipeline, regardless if it results from sensor, lens, tonemap curve, or some other unknown. As far as I know, there shouldn't be any other non linearities introduced by these other things, but it's not entirely impossible. If there are such non linearities, chances are that they are small, orders of magnitude less influential than what the tone mapping curve does. But this probably the reason, if a legit technical reason even exists, why DJI themselves offer for download different LUTs for different drones even though they use tone curve with the same name "D-Log":
https://www.dji.com/at/downloads/softwares/transcoding-mavic-3
https://www.dji.com/at/downloads ... ding-zenmuse-x7-rec
I personally wouldn't use my D-Cinelike LUT to turn Mavic 3 Pro into a scientific light measuring instrument (as I did with my DJI FPV) without first confirming that results are exactly the same. But for the purposes of grading, I'd use my own LUT without a second thought, and would only look deeper and go test it's accuracy if I would notice that something feels off. I'm 99% confident that for purpose of grading, my D-Cinelike LUT would do the job correctly on Mavic 3 Pro.
Footage being 10 bit vs 8 bit doesn't influence this at all. Exposure range also has nothing to do with any of this. |
|