More Nikon D800 1080P video tests

iso testing , it looks like to me that the 5d mark ii and the d800 is falling apart at 3200 and 6400 compare to the other cameras tested (allot of noise)
 
Last edited:


This camera does produce some of the most filmic image out of any camera that I have seen. I think it is because it has very good DR. If the moire filter can decrease to eliminate the aliasing/moire, then this could be another benchmark in price to quality for DSLR and I would say, large sensor camera at large. If Nikon could do a special camera like the rummored Canon 4k it could be groundbreaking. It does not even have to be 4k. A 3k/4k output with less pixel skipping even for just 2 min.
 
This camera does produce some of the most filmic image out of any camera that I have seen. I think it is because it has very good DR. If the moire filter can decrease to eliminate the aliasing/moire, then this could be another benchmark in price to quality for DSLR and I would say, large sensor camera at large. If Nikon could do a special camera like the rummored Canon 4k it could be groundbreaking. It does not even have to be 4k. A 3k/4k output with less pixel skipping even for just 2 min.

It's DR is definitely excellent. Looking forward to seeing more from Nikon. Nikon is on the right track.
 
Very very nice film like DSLR and I think it is due to the DR which is very very good. In some shots like the helicopter it is very log type. Perhaps it is due to the light reflection on the helicopter mirror but it is very very nice and someone could use some low-con filter to get the same result.

What is the most battling about this camera is the discrepancy of the sample. The youtube one are just horrible, I think it is due to many Nikon users not knowledgeable of video. I think they just crank the sharpness and contrast setting and film without ND at very high shutter speed. Even I saw some video done at the same event and they are just awful compared to these. I am still downloading the 1080p version and will see about the aliasing.
 
from all the bad low light examples i saw from the d800 which examples allot of noise at high iso
this one looks remarkably good and also other nice footage
 
OK I downloaded the 1080p version and I see the aliasing now. Good to see that vimeo's 720 got rid of most of it at least on the 1st view I had.
I can´t wait for the Mosaic Engineering AA filter. It´s going to make this camera just perfect (until a new model comes out of course) XD
 


This is a short film made with the new Nikon D800 DSLR at various locations in New York as a part of press trip named NY Challenge in 48 hours. Almost rough cut, no color grading.
Used lens: Nikkor 14-24 f/2.8G ED, 24-70 f/2.8G ED, 70-200 f2.8G ED VR II.
 
THe test does seem to show the following: lost of apparent DR when recording to an external recorder.
BUT, this is my theory.
The HDMI output of the nikon seems to be full range, and prores is a limited range codec.
You have two options to fix this:
- bring down the levels with 3 way CC in FCP (no CS5 won´t let you recover details lost)
- shoot in DNXHD
 
Last edited:
I tend to agree with this diagnosis. The ProRes codec is clipping the DR range. Better option maybe the Hyperdeck Shuttle 2 which can record full uncompressed 10bit 4:2:2 or as indicated Avid DNxHD, which will be a better codec than ProRes on conversion.
 
THe test does seem to show the following: lost of apparent DR when recording to an external recorder.
BUT, this is my theory.
The HDMI output of the nikon seems to be full range, and prores is a limited range codec.
You have two options to fix this:
- bring down the levels with 3 way CC in FCP (no CS5 won´t let you recover details lost)
- shoot in DNXHD
Third option: set the D800 to respect limited range (in compliance with the HDMI 4:2:2 YCbCr specification, as well as SMPTE and ITU signal standards). See page 257 of the D800 User Manual.

-- Bob Elkind
 
Back
Top