LetTheRightLensIn said:
jrista said:
LetTheRightLensIn said:
zlatko said:
LetTheRightLensIn said:
And failed to mention one thing he'd bring over from Nikon other than the shutter feel.... the dynamic range difference at low ISO where the Nikon actually is much better. So then you start thinking about all the money dangling above his head again.
It's quite possible that Nikon's dynamic range at low ISO makes no difference to Scott Kelby. It makes no difference to me at all.
Yes that is certainly quite possible, but what is suspect is that he makes a big deal about a high iso improvement compared to his D4 which doesn't actually exist while ignoring a low ISO improvement that is there. Not suspect would have been to mention that the loss of low ISO DR alone or to have mentioned nothing about the sensors whatsoever (high iso is the same for D4 and 1DX pretty much and if he doesn't care about the low ISO DR he doesn't care).
It's certainly quite possibly he did end up liking the 1DX better than his D4 no doubt at all (although whether he would have put out tens of thousands of his own to add a top sports level Canon setup to his arsenal is less certain), but with all the sponsorship you can be sure it will be tougher for him to bring up any negative points about Canon (as the high ISO/low ISO thing hints at), it's just natural for it to be a bit tougher when people are being nice and giving you this and that.
The high ISO difference doesn't exist on paper, or in artificial tests.
Read my previous response to you...I provided some links with visual evidence that there IS a difference between the D4 and 1D X at high ISO, giving a clear edge to the 1D X. The key is that the 1D X has less apparent color noise, especially in the shadows...by quite a visible margin. The D4 at anything over 12800 is doing a digital boost, so it's lifting read noise along with everything else, so more color noise is expected. The 1D X, on the other hand, is doing amplification at the pixel before readout up to ISO 51200, so only ISO 102400 and 204800 are doing a digital boost and lifting read noise. This real-world difference has a meaningful impact on real and perceived IQ at very high ISO settings.
Artificial tests and paper specs don't tell you everything.
How are you processing though? Because direct measurement BEFORE RAW converters do under the hood whatever, which may vary camera to camera even with sliders set to the same thing, the D4 has less chroma noise. Under natural lighting the D4 seems to have getting near 1 stop LESS chroma noise at very high ISO. It does seem that under tungsten lighting they may become about the same though.
It does seem that the 1DX might have somewhat more than 1/3 stop better DR once you get well above ISO6400 though.
So maybe under certain weird artificial lighting the Canon has similar luminace and chroma SNR but somewhat more than 1/3 stop better DR and 2 extra MP so maybe it does slightly better in that case.
I still find it hard to believe that somehow adds up to the 1DX looking way better at high iso. I haven't heard anyone but Scott and you say that. Maybe if you are reallllly pushing it way up there under weird types of artificial lighting it's enough to at least notice the difference in 1DX favor though. It still would seem a bit curious he makes such a huge deal about it though and ignores all the other sensor aspects.
Also he also implied the 5D3 blows away his D4 at high ISO, but the 5D3 has worse SNR, worse chroma noise and more noticeably worse DR at very high ISO than the D4. Granted the 23MP vs 16MP might the 'grain' tighter which helps make up some of the measured differences since the eye prefers smaller grain. But I still find it dubious to think the 5D3 would appear to blow the D4 away at high ISO.
It isn't me processing. It is other people, who have written well-known 1D X reviews, from whom I linked example images. They tend to process with LR, although some may process with DPP.
I think this is part of the problem with DXO results. That they are solely based on pre-demosic results. NO ONE uses RAW without demosaicing them! Every time a person views a RAW image, it has been demosaiced. And, given the plethora of visual examples on the web, the 1D X, regardless of whether sample images are produced with DPP or LR/ACR or anything else, clearly has lower noise at higher ISO settings. Especially color noise.
I think the notion of testing pre-demosaic RAW is rather misleading in many respects. I do not think that demosaicing will really change the difference between Canon and SoNikon DR all that much, maybe by a third of a stop or so. But however things work, however the algorithms in DPP and LR work, Canon results, without any additional downsampling, clearly improve AFTER they have been demosaiced. I think that is an important factor.
Also, I never expected you to take my word for it. I linked several examples in my previous answer. Just take a look. The results speak for themselves. (And those are not my images, they were taken from other well-known 1D X tests, from other people, besides Scott and myself, who ALSO say that the 1D X has less color noise at high ISO.) I don't believe there is "funky artificial lighting" or anything like that that has anything to do with the 1D X results. That's illogical, as the D4 comparison images are captured under that same light. Lighting isn't going to change how deep the shadows are post-digitizing...your really grasping at straws there. It's just a bunch of reviewers pointing the camera at something that is normally lit an taking some photos, then performing a visual analysis of the results. Just LOOK at the examples...the 1D X is clearly superior with color noise and in the shadows.
Artificial tests don't tell the whole story. Real-world tests, across the board, indicate that the 1D X has rather exceptional high ISO performance all the way up to ISO 51200. Those results FIT with the THEORY as well...it is EXPECTED that the 1D X would have less color noise and deeper shadows, because the pixel value are being amplified at the pixel, CDSed, then digitized. The D4, on the other hand, is amplified & CDSed at the pixel only up to ISO 12800, then it is digitally amplified by another stop or two to achieve ISO 25600 and 51200. It is EXPECTED that the D4 shadows would not be as rich and deep or have as low color noise as Canon's. So, it really shouldn't be all that surprising that real-world visual examples correlate with what I've said here.