DPreview adds 6D sample images to their preview

[onkel_wart]

for someone like me coming from a 50D this looks like a dream come true.

 

[nightsky87]

Did a quick and dirty comparison using shots from dpreview. Canon 6D against the 5D mkIII at ISO 102400. I'll let you draw your own conclusions.

http://i.stack.imgur.com/VVoQ6.jpg - or attachment

Not to burst anyone's bubble but the 5D3 sample I saw in the DPR galleries was with no NR applied. So unless I missed something, this comparison is like comparing the in-camera NR'd 6D images vs no NR 5D3. Hardly a fair comparison. But these images are certainly encouraging compared to the ones on Flickr.

Okay, I just looked around a bit more and I did see the right samples. So unless there was a significant different between the pre-production 5D3 and the final version, the 6D samples look great! Time to really pre-order this one.

 

[bvukich]

The plot thickens... (dramatic pause) dum dum duuummmm...


But seriously, I have no doubt that the IQ will be acceptable; I, and I'm sure many others, are more concerned about the AF. I fully understand that this is a compromise camera, spec'd and priced for a certain bracket in their lineup; but I'd be very disappointed if the AF performance is less than the 60D. I know spec wise, it looks to be lesser, but it could still perform better, especially in low light with the -3EV sensitivity.

I really want to be able to like this camera so I can get it plus a decent lens, instead of having just a 5DIII and having to make due with my current mediocre glass.

 

[verysimplejason]

The plot thickens... (dramatic pause) dum dum duuummmm...


But seriously, I have no doubt that the IQ will be acceptable; I, and I'm sure many others, are more concerned about the AF. I fully understand that this is a compromise camera, spec'd and priced for a certain bracket in their lineup; but I'd be very disappointed if the AF performance is less than the 60D. I know spec wise, it looks to be lesser, but it could still perform better, especially in low light with the -3EV sensitivity.

I really want to be able to like this camera so I can get it plus a decent lens, instead of having just a 5DIII and having to make due with my current mediocre glass.

For my usage, AF-wise, if it can get AF in low-light even if it's not so fast, I think that's enough for me. I agree with you. At least it should be able to surpass 60D AF or else, this is just another glorified 5D2. If that's the case I'd also consider 5D2.

 

[Marsu42]

for someone like me coming from a 50D this looks like a dream come true.

Given that nowadays the camera upgrades are incremental, it pays to wait a little to wait ($$$ and iq-wise) ... and even with the 60d it's the same, if the 6d has 1 usable stop more than the 5d2 that's good news and a reason to get the newer gear.

it's nice to know that Canon has given DPReview a per-production camera to play with.

Where did you get this information? Everywhere I look it has the "The camera used for this gallery was pre-production, and image quality should be considered 'Beta' standard" notice, though that probably only means the in-camera raw-converter / jpeg-engine and not the sensor hardware.

 

[verysimplejason]

Previously in another thread, I've raised the possibility that 6D is built on an entirely new generation of sensor. After seeing the sample pictures even if they're just pre-production, I'm thinking that 6D's sensor might be it. Maybe Canon got tired from giving Nikon a lot of advantage in the sensor department. I'm still crossing my fingers... I hope they got/nail it with 6D because soon it will be my first FF (forgive me for my enthusiasm, I've waited a lot already).

 

[dtaylor]

Here's the 6D studio shot at ISO-6400. And here's the 5DIII shot.

There's either an exposure or a levels difference between the two. I don't see any real difference in noise. If there is a difference, it is well below the threshold of differences introduced by post processing (NR software and technique).

 

[verysimplejason]

Here's the 6D studio shot at ISO-6400. And here's the 5DIII shot.

There's either an exposure or a levels difference between the two. I don't see any real difference in noise. If there is a difference, it is well below the threshold of differences introduced by post processing (NR software and technique).

The 6D shots might be a little bit underexposed compared to 5D3 but nevertheless, I can see it a little bit better than 5D3. Besides, generally underexposure usually means more noise when processed. Is this NR or is this really a new sensor technology? Let's wait till December.

 

[dtaylor]

Did a quick and dirty comparison using shots from dpreview. Canon 6D against the 5D mkIII at ISO 102400. I'll let you draw your own conclusions.

Again, there is a lighting / exposure / processing difference at play. It's hard to discern real noise differences with these confounding factors, but any real noise differences are likely less than those introduced by user post work.

Side note: I hate the new DPReview "low light" test. Brightly lit objects against pure black is not the situation where DSLRs struggle. With this test a 7D would produce good 6400 shots. No, the problem is when a lot of your detail falls on the shadow side (not black), especially if you have to lift in post. It's these situations which force me to cap 7D shots at 3200 and 5D2 shots at 6400.

 

[dtaylor]

There's either an exposure or a levels difference between the two. I don't see any real difference in noise. If there is a difference, it is well below the threshold of differences introduced by post processing (NR software and technique).

The 6D shots might be a little bit underexposed compared to 5D3 but nevertheless, I can see it a little bit better than 5D3. Besides, generally underexposure usually means more noise when processed. Is this NR or is this really a new sensor technology? Let's wait till December.

You're assuming underexposure vs. the tonal curves of the cameras (or some related setting). At 6400 there's no actual resolution difference, and little noise/contrast/color difference which disappears with a levels adjustment.

At 102k there's more of a difference at first glance. But it could still be due to the camera tone curves.

Sorry, I think people are getting excited about JPEG processing. And not even JPEG NR processing, but small differences in image parameters between a consumer and a pro camera.

 

[Marsu42]

Previously in another thread, I've raised the possibility that 6D is built on an entirely new generation of sensor.

Definitely not, this was already ack'ed by Canon - it's the same sensor tech, though the 6d has a bit little larger pixels for the loss in resolution. The reason probably is that after the criticism about the lack of advancement 5d2->5d3 Canon had a 3/4 year time to optimize everything else.

The 6D shots might be a little bit underexposed compared to 5D3 but nevertheless

If you look at the exif data, all 6d shots have a shorter exposure time then the 5d3...

but small differences in image parameters between a consumer and a pro camera.

However, this shows how desperate the Canon crowd is for sensor advancement ... the Nikon competition obviously has left its mark :-> ... and again, imho there is no fixed distinction between "pro" or not except for marketing, why not use the 6d if it is at least as good or even better (low light af) for some shots?

 

[nightsky87]

However, this shows how desperate the Canon crowd is for sensor advancement ... the Nikon competition obviously has left its mark :-> ... and again, imho there is no fixed distinction between "pro" or not except for marketing, why not use the 6d if it is at least as good or even better (low light af) for some shots?

Agreed! If the low-light AF functionality of the 6D is as advertised, then it certainly caters to a different group of photographers. Body is still magnesium alloy reinforced after all. What's really missing is the lack of the advanced AF system of the higher-end models. Maybe eventually, the 6D would also be used as a backup body for when low-light is needed.

Personally, as long as the 6D doesn't perform poorer in the IQ area compared to the 5D3, then its certainly a viable option for people like me who don't really take much pictures of moving subjects.

 

[Gothmoth]

Re: dpreview 6D samples

What are you talking about?

I just compared the 6D at ISO-6400 vs the 5DIII at ISO-6400 ... and the 6D is cleaner and with better colors.

Overall, the 6D image quality is shaping up as better than the 5DIII.

yeah well if details in the images don´t count for you.... the images look good.

but im only interested in low iso.. (100- max 1600) im not needing these incredible high isos for my shots.

and like the 5D MK3 i see no real improvement compared to the 5D MK2... not in these samples.

 

[Marsu42]

Re: dpreview 6D samples

i only speak about low iso.. im not needing these incredible high isos for my shots. and like the 5D MK3 i see no real improvement compared to the 5D MK2... not in these samples.

And that wasn't to be expected - for lower iso shots the 5d2 should be great (actually even little better than the 5d3). But most people who upgrade to ff want to have the medium-high iso capability, meaning ~6400 which is exactly what is really missing from the crop sensors and in this region there is an improvement over the 5d2. I agree crazy-high iso speeds are mainly for marketing.

 

[Gothmoth]

Re: dpreview 6D samples

And that wasn't to be expected

yes it was not... but i really hope we will see a new process (180nm) soon.

then the great camera bodys will be joined with even better sensors.
and for low iso, high quality landscape and studio photographer, who don´t care much about AF speed or AF tracking, there will be a reason to upgrade to a new body.

 

[RLPhoto]

These are out of camera Jpg's. They look identical to the 5D3 files but only with more NR in camera.

 

[Area256]

it's nice to know that Canon has given DPReview a pre-production camera to play with.

Where did you get this information? Everywhere I look it has the "The camera used for this gallery was pre-production, and image quality should be considered 'Beta' standard" notice, though that probably only means the in-camera raw-converter / jpeg-engine and not the sensor hardware.

It's the logical conclusion that DPReview got a pre-production 6D from Canon. How else would they have been able to take sample images with it? These aren't samples Canon took themselves and sent to DPReviews. How closely this camera comes to the production camera is anyone's guess. Although if Canon intends to have these in stores world wide for Christmas, I'd bet they are already in production, and the only thing likely to change is firmware.

 

[Marsu42]

It's the logical conclusion that DPReview got a pre-production 6D from Canon.

Sorry, of course you're correct - I mis-read your post :-o

These are out of camera Jpg's. They look identical to the 5D3 files but only with more NR in camera.

Look again, for example @iso6400 on the fur of the mouse on the left - there are more details in the 6d even though it has a cleaner background, so it's not just more nr. But still doesn't mean much since it's a comparison of two beta in-camera jpeg engines with too high nr settings...

 

[RLPhoto]

These are out of camera Jpg's. They look identical to the 5D3 files but only with more NR in camera.

Look again, for example @iso6400 on the fur of the mouse on the left - there are more details in the 6d even though it has a cleaner background, so it's not just more nr. But still doesn't mean much since it's a comparison of two beta in-camera jpeg engines with too high nr settings...

Meh. If there is any tiny difference, it will be irrelevant in real world use.

 

[Marsu42]

Meh. If there is any tiny difference, it will be irrelevant in real world use.

Sure, though we'll only know for sure by looking at studio shots after a 6d raw converter is out. And concerning the dpreview shots: Actually I looked again and the focus on the 5d3/6d shots is a little different, so even @f11 that explains some differences.