60D 10-Bit RAW @1728x736 25p Overexposure testing

[whyfy]

10Bit Raw module, DeBayered with MLVProducer, Corrected and edited in Resolve.
Set to 1728x736 25p upscaled to 1920x820
Night shots @ 1250 iso
Day shots overexposed by ~2-3EV

[aschille84]

Interesting comparison. Totally different clarity with 10-bit raw. Great night shots!

[D_Odell]

I always appreciate tests! They give me knowledge which is valuable. But when a shooter like you does a good job, how come you don't state the changes in the video, ML and H264..? And the most useful comparison is when doing side by side comparison. And I don't believe I'm different but I can't remember pics after a cut. You need to have them side by side to see and feel the difference.

Hopes for the future!

Glad you make tests!

Best,
David

[maxotics]

Hi D_Odel.  I thought he only missed one captioning of footage.  The most obvious difference between H.264 and RAW is either in the dark or bright parts of the image.  H.264 usually "crushes" dark parts of the image to black, whereas you can see more detail in RAW.  Same for bright parts of the image.  The sky will usually band, or get blotchy in H.264 but more light grays and detail in RAW. 

I thought the test very conservative in that he could have went for a lower contrast image possible in RAW. 

Anyway, if you look at the video again, but then look at the dark and bright parts of the image between each clip you'll get that AH HA moment

[whyfy]

@D_Odell @maxotics
Thanks for the feedback! Better to have something to improve on than nothing. All of the editing and corrections were done in daVinci resolve so the options were limited. And it was my first comparison video. But for next time it will be done differently.

@maxotics Yes you're right about the high contrast in the RAW shots. Whilst correcting in Resolve I had problems when exporting the video. Maybe its to do with video settings truncating levels below 16 and above 235 or I'm bad at Grading lol.