1DC vs 1DX II Shootout

[Kino]

11 minutes ago, Luke Mason said:

This is a better test:

C-log as is:

std gamma, -4 contrast:

That is awesome. Thanks for posting. It's obvious those first four bars are clipped on standard gamma, whereas only two are clipped on C-Log. The distribution of the stops is also very different: C-Log provides a much greater "usable" range (10 stops vs. 8 stops).

Again, for me the 1DC footage looks like it came from a high-end cinema camera (and is often used alongside such cameras), whereas the 1DX II just has too much contrast to cut with such cameras.

[Andrew Reid]

43 minutes ago, Kino said:

I posted this amazing Xyla on the other thread not knowing you had already posted it here. Anyway, the reason for the different apertures is because the 1DX II has better ISO performance and thus the 1DC needs to open up for more light on the Xyla test. Moreover, ISO 800 is chosen to maximize the DR performance on both cameras.

I can't understand why ISO 800 was chosen, it seems to have been plucked out of thin air as some kind of 'native' ISO.

In fact if you look at nearly all the stills cameras, they do their best dynamic range at ISO 100-200.

ISO 800 is not where the 1D C and 1D X Mark II give the most dynamic range.

Also a LOG profile and a standard image profile need to be exposed differently.

Furthermore they have set the test up in a way that doesn't test how the roll off is to the highlights or even how much latitude you can pull out of the LOG vs rec.709 image.

43 minutes ago, Kino said:

With Technicolor Cinestyle, the 1DX II is thus achieving the exact same DR as the 1DC with C-Log. If I can see some Cinestyle footage from the 1DX II that demonstrates this, it will have saved me thousands of dollars as the 1DX II is also far superior to the 1DC for rolling shutter.

Really? Don't think so - rolling shutter in 4K/24p looked unimproved on the 1D X II I tried out at the BVE show in London.

43 minutes ago, Kino said:

It's actually even better than many cinema cameras.

I suppose there is no need for C-Log or any "1DC II" when we can install Cinestyle for a similar effect.

CineStyle is not a true LOG profile.

I know for a fact that Canon LOG gives me more dynamic range than the standard Canon profiles, let's not pretend otherwise!

Quite odd how bad people are saying the HD image is. Bodes badly for the 5D Mk IV. They could have intentionally put a mediocre pixel binning 1080p mode on the sensor, which will probably end up in the 5D Mk IV with the 4K mode disabled.

[Kino]

I thought that C-Log must be shot at ISO 400 or above to avoid banding on the 1DC. ISO 800 was probably chosen for what they thought may be the best performance, but it does seem slightly odd considering the fact that they had to use different apertures. As for the rolling shutter, they claim the 1DX II achieves an incredible number: 14ms.

I do agree that the Xyla test is questionable considering everything we have seen from these two cameras. It also does not demonstrate the highlight roll-off of the cameras and is more informative for what is going on in the shadows (where C-Log is also preserving more color information based on other tests we have discussed here). And, yes, of course, Cinestyle is no real substitute for C-Log on the 1DC for reasons I have already stated (no access to RAW sensor data like C-Log).

However, we should also not forget that filmmakers like Abraham Joffe have shot beautiful footage on the 1DC without using C-Log:

In fact, Joffe has stated that he prefers shooting the 1DC using the picture profile settings and not C-Log:

http://www.untitledfilms.com.au/2013/11/shoot-edit-deliver-4k-now/

You can check out all his 1DC films including the National Geographic TV series, Tales By Light, to get an idea of what is possible outside of C-Log:

 

Now if I owned a 1DC, I'm sure I would use C-Log for most situations, but it doesn't mean that great results are not possible without it or that it is ideal for every lighting condition. C-Log or not, the 1DC is just a magical camera.

[DBounce]

2 hours ago, Kino said:

That is awesome. Thanks for posting. It's obvious those first four bars are clipped on standard gamma, whereas only two are clipped on C-Log. The distribution of the stops is also very different: C-Log provides a much greater "usable" range (10 stops vs. 8 stops).

Again, for me the 1DC footage looks like it came from a high-end cinema camera (and is often used alongside such cameras), whereas the 1DX II just has too much contrast to cut with such cameras.

You do realize that this test is not comparing the 1DC to the 1DX Mkii... it is comparing the 1DC to the 1DC using c-log. So all this proved is that the 1DC has more DR than the 1DC... rolleyes.

[Luke Mason]

4 minutes ago, DBounce said:

You do realize that this test is not comparing the 1DC to the 1DX Mkii... it is comparing the 1DC to the 1DC using c-log. So all this proved is that the 1DC has more DR than the 1DC... rolleyes.

It is very clear that 1DX II in neutral profile performs the same, if not worse, as 1D C.

[DBounce]

6 hours ago, Luke Mason said:

It is very clear that 1DX II in neutral profile performs the same, if not worse, as 1D C.

It's very clear you are stating the above as a comparison to a camera that is not in the comparison. The 1DX Mkii uses a different sensor, so nothing other than the aforementioned is "very clear". In my personal testing I am seeing considerably more DR by using Technicolor.