6d2 is here.

Apricane wrote in post #18404493
No one is really doubting whether or not the 6D2 can have same/different performance, but just being concerned that said performance seems worse (even if in only one area and marginally so) than that of the previous generation. That's rather concerning.

I find it especially troubling because this was a weak spot for canon for a long time but they seemed to be making great improvements. I could live with the gimped features on this camera, but if they have also gimped the sensor it is going to really shake up how I feel about canon.

Apricane wrote in post #18404493
No one is really doubting whether or not the 6D2 can have same/different performance, but just being concerned that said performance seems worse (even if in only one area and marginally so) than that of the previous generation. That's rather concerning.

Especially since the competition still out-performs Canon and they have the technology to at least get close. I just makes no sense to put out a camera 5 years in development and all they could do is put in the AF from another body, add a few megapickles, put in a fancy flippy screen, and bump from 30fps to 60fps video. It just makes no sense so I'm thinking that the leaked raw files were incorrect. I'm holding out hope that the 6DII will perform at least as good as the 80D.

davesrose wrote in post #18404497
But they are all V8 engines. Chevy wouldn't think to go back to the LU5 if they were already in the LB9 cycle of production.

I have had an L99 6.2L and an LT1 6.2L. Both are newer engine designs, but they perform differently. Yes Canon would do that. Again, the raw files are there, take a look. They have already been looked at by others much better at this technically.

Stregone wrote in post #18404506
I find it especially troubling because this was a weak spot for canon for a long time but they seemed to be making great improvements. I could live with the gimped features on this camera, but if they have also gimped the sensor it is going to really shake up how I feel about canon.

gjl711 wrote in post #18404507
Especially since the competition still out-performs Canon and they have the technology to at least get close. I just makes no sense to put out a camera 5 years in development and all they could do is put in the AF from another body, add a few megapickles, put in a fancy flippy screen, and bump from 30fps to 60fps video. It just makes no sense so I'm thinking that the leaked raw files were incorrect. I'm holding out hope that the 6DII will perform at least as good as the 80D.

I couldn't agree more with the both of you. Those are my feelings at this point, especially if the decision to gimp the sensor was a deliberate market move to protect 5D4 sales.

TeamSpeed wrote in post #18404513
I have had an L99 6.2L and an LT1 6.2L. Both are newer engine designs, but they perform differently. Yes Canon would do that. Again, the raw files are there, take a look. They have already been looked at by others much better at this technically.

So you're saying Canon would backtrack and have a new FF sensor that had inferior DR performance to the 80D? How could that not insinuate Canon using different sensor tech then their current lineup? As for engaging straw men, I don't have Linux and can't run DCRaw. I did try downloading one of your CR2 links and found it wasn't working with RawDigger (it would open but had a purple cast). I've got the patience to wait for official reviews.

Yep, completely agree.

If Canon asked my opinion -- which they have NOT and probably never will - I would say to never, never, never, skimp on image quality if at all possible.

davesrose wrote in post #18404529
So you're saying Canon would backtrack and have a new FF sensor that had inferior DR performance to the 80D? How could that not insinuate Canon using different sensor tech then their current lineup?

The 5DS is a more expensive camera than the 5DIV - and yet, its DR performance at base ISO is worse than both the 5DIV and the 80D.
And the reason for that is that the 5DS is still using external ADCs - as opposed to the on-chip ADCs in the 5DIV and 80D.

Based on the pre-production RAWs, the 6DII DR performance is consistent with the use of external ADCs, just like the 5DS.

Note that that 5DS is still a current model in Canon's lineup (and will likely remain so at least until Photokina next year).
So, technically, Canon is not 'backtracking' with the 6DII sensor.
This sensor is still made on a current Canon tech - just not the very latest.

But the 5DS used Canon's latest available technology when it was released. Same with the 5DIV, 6D and most other Canon offerings I'm aware of.

If the 6D2 is released with an older technology sensor, it seems Canon would be breaking with their own well established tradition of putting the most up to date sensors in their new releases.

I've never used one myself, but doesn't the 5DSR have pretty good shadow recovery? I remember seeing a few reviews were people had exposed for the highlights and still had a very clean image after pulling the shadows in post.

x-vision wrote in post #18404597
The 5DS is a more expensive camera than the 5DIV - and yet, its DR performance at base ISO is worse than both the 5DIV and the 80D.
And the reason for that is that the 5DS is still using external ADCs - as opposed to the on-chip ADCs in the 5DIV and 80D.

Based on the pre-production RAWs, the 6DII DR performance is consistent with the use of external ADCs, just like the 5DS.

Note that that 5DS is still a current model in Canon's lineup (and will likely remain so at least until Photokina next year).
So, technically, Canon is not 'backtracking' with the 6DII sensor.
This sensor is still made on a current Canon tech - just not the very latest.

This seems logical to me. Perhaps Canon didn't meet their 5D3 sales target after introducing the 6D. This time around they want to create even more separation between the 5D4 and 6D2 by using the 5DS sensor tech. Logical but a real bummer for potential 6D2 buyers.

VooDooZG wrote in post #18404282
So Canon Europe is lying in his official promo video ?? 0:14 sec big letter "INCREASED DYNAMIC RANGE"

https://www.youtube.co​m/watch?v=_i7O5lppFPw

Yes, it has increased from original Canon FF camera, the 1Ds.

Really, all that really needs to mean is any mode and picture style that puts more scene range into a JPEG, compared to some previous camera, to claim "increased DR". For one example, pulling up more shadows and pulling down more highlights into the conversion becomes more palatable if they also increase local contrast, like applying an unsharp mask at a very wide radius. Something like that could qualify for "increased DR", even if everyone can do it even better in software with another camera.

davesrose wrote in post #18404431
Some people aren't claiming that Canon is "exaggerating" claims about improved DR. They're saying Canon has crimped the 6D2 sensor, and that Canon is lying that the 6D2 has similar sensor tech as all current Canon models (even claiming it has slightly less DR then the 6D). That's quite different then subtleties in marketing, and what I think is an example of fallacy (the 6D2 is a known product and is part of the current line of Canon products). Well the final verdict of the 6D2's DR will be shown with published results from DPR and DxO.

I'll need about 10 seconds alone with a release-firmware ISO RAW to know the final 6D2 RAW DR truth. What DxO will be most useful for, IMO, is that they will measure the actual meaning of the RAW levels - that is what their "measured ISO" indirectly means, and that is why the data points when you look at their DR charts can slide about in an analog manner on the X axis. This allows the charts to show noise relative to signal (the trendlines), and "DR" (the dots), in the same chart.

We don't need DxO to know the limit of DR. DxO does not have any deep knowledge of the meaning of RAW data that other people don't have; they are just consistent and methodical in their testing. The most pixel-level DR an ISO on a camera can have is the highlight clipping level minus the blackpoint, then that divided by the standard deviation of the black area on the left of the RAW image. Hundreds if not thousands of people know how to do this. Every time a new Canon comes out, I download an ISO 100 RAW, load it into a program that preserves the real RAW values and the black border pixels, draw a rectangle in the black border, and then derive the statistics, and the standard deviation tells me instantly the limits of pixel-level DR by the "engineering" definition, which is the most generous one in common use. I did that with the 6D2 RAWs I found, and fully expect the DxO results to be very close to that, within about 0.1 stops, as has been the case with every camera I ever got a base-ISO RAW from before DxO tested it.

The 3 files I linked to are at ISO 100 for the 6D2. What do you see with those, despite it being a pre-production camera, sitting at a store?

TeamSpeed wrote in post #18404486
The 6D2 can have the same tech as the newer bodies, but it doesn't have to have the same performance. That is what people are suggesting. Chevy has made several variations of the 350, but they all perform differently.

The numbers are uncannily typical of off-sensor ADC noise, with perhaps a little bit of extra gain wasting a small amount of highlight headroom. Having an on-sensor ADC that gives post-gain noise almost exactly like the older cameras is highly unlikely, IMO.

The hopeful part is that if there is extra gain, that means that the high-ISO noise may be in the 1Dx neighborhood, but with a possibly finer noise character, putting the 6D2 near the top for high-ISO in the Canon camp. That's why I wish that the guy who had the camera on FM had thought to take the exact same shots with the 6D (or 5D4) and the 6D2, with the same lens, same lighting, and the same Av and Tv values. Using different lenses and allowing the cameras to meter ruins any chance of getting a truly equitable comparison.

Canon could play around with the code that is reading the data off the sensor and storing it, though. Highly unlikely, but it is in the realm of possibility. They could use the new tech, but cripple the readout in some way to make it a bit worse than the 5D4, but still have better high ISO, where most of consumers would want it to be anyways. Again, I doubt it, but it could happen. Canon has done silly things in the past.

There are most likely a larger set of people that care about higher ISO performance than eeking out about 1 to 1.5 stops of DR in high DR scenes. This means that the 6D2 will be a huge hit still, with a smaller group that prioritizes DR over high ISO still unhappy.