don't wanna start a raging fight.......

Press Correspondent said:

Any idea if this has been fixed or still is the same on 5D3? It has a bit higher quantum efficiency and other improvements, one would hope...

Click to expand...

I do not own the 5D3 but I studied a few files posted here, and the flat parts (bokeh, etc.) look very clean to me. I know that DXO does not show improvement but this may be something that they cannot catch.

EDIT: I set the WB to Daylight, cropped the noisy part and looked at the histogram. The dominant color is blue, then red, and very little green. The noise can be explained that the image is based mainly on 1 (the blue) out of 4 pixels in the Bayer array, and on the Canon color filters. When I correct the WB, I suppress the dominant blue color even further, making it worse. So in that particular area, the "push" can easily be several stops.

Anyone who starts out by saying "I don't mean to BLANK..." is actually doing what they say they are not trying to do. Like "not to be rude, but..." is a preface to you thinking you are about to be rude but want to make who you are talking to not feel offended.

Just stop it. You meant to stir things up because dxo mark's "scores" have been discussed here quite a bit. Please do a search before you post what has already been posted several times over a couple of forums.

qianp2k said:

This is from someone who actually owns 5D3 and D800. He said 5D3 is actually better in IQ (assume he doesn't need to print to 60x30"). There are lots of hypes there but you will know when you actually use it.

http://forums.dpreview.com/forums/read.asp?forum=1032&message=41723017
--

Click to expand...

Actually he does not say the Image quality is better with the 5DMKIII he was speaking about recovering blown highlights & Image quality is about much more than highlight recovery.

qianp2k said:

FlyboySA wrote :

I have to wonder about all the noise (pun intended) about how well you can push the images from some cameras. I own a D800E and yes, I can push it maybe 1 stop more than my 5D3, BUT, and this is a big but for me as a landscape photographer, I can recover highlights far better on the Canon than the Nikon (Yes, look at the DXO tests).

Click to expand...

qianp2k said:

As blown highlights (clouds during the day and lights at night) is a far larger problem than shadows (expose to the right!!) I find the files from my Canons better suited to > landscape work than the Nikon's. The Sony's are even worse (overall) than the Nikons.

Click to expand...

In a latter post he adds

qianp2k said:

FlyboySA wrote :

If you look at the DPreview review of the 5D3 you will see that where they compared DR the DR "window" of the 5D3 goes further right (high light) than the D800E.

Click to expand...

I looked & it doesn't they are the same, Its only comparing various tone curves on a Jpeg . Its obvious to most of us when you are processing a raw file all the dynamic range is available to recover highlights or push the shadows.

qianp2k said:

FlyboySA wrote :

When looking at real files I also see less noise at ISO 100 on the 5D3 (and 1Ds2) than on the D800 (I can also see it on the DPreview samples)

Click to expand...

Everything I've seen in reviews & comments shows more noise on the 5DMKIII at ISO 100. Makes me wonder about the creditability of FlyboySA when talking about the better highlight recovery of the 5DMKIII.

HTP on the 5DMKIII loses DR in the shadows . ADL on the D800 expands DR in the shadows & higlights at the same time.

D800 has to turn on "ADL Extra High" to match 5D3's HTP on. I don't know what's the penalty when turning on "ADL Extra High" in D800 that is the extreme extension. It likely causes highlight severely subdued such as white color now becomes grey color that might cause entire photo look a bit "dull". Interesting is that if turn off both HTP and ADL, 5D3 has more shadow range than D800 in the same DPR lab test you quoted above. In real world I virtually never turn on HTP on my 60D and 5D2 that will shift base to ISO 200 and cause a bit of more noises in shadow. I found usually I can recover highlight in LR especially in 5D2 raw.

Anyway he owns both and you probably never experience 5D3. One thing I agree with him is that I also care more in highlight recovery than shadow pulling. In 99% of my photos I don't need to pull 4-5 stops in extreme but only 1-2 stops and very occasionally 3 stops.

--
http://qianp2k.zenfolio.com/

qianp2k said:

D800 has to turn on "ADL Extra High" to match 5D3's HTP on. I don't know what's the penalty when turning on "ADL Extra High" in D800 that is the extreme extension. It likely causes highlight severely subdued such as white color now becomes grey color that might cause entire photo look a bit "dull". Anyway he owns both and you probably never experience 5D3. One thing I agree with him is that I also care more in highlight recovery than shadow pulling. In 99% of my photos I don't need to pull 4-5 stops in extreme but only 1-2 stops and very occasionally 3 stops.
--

Click to expand...

If you are shooting jpegs straight out of the camera with no further post processing they are going to look a little flat, who cares just play around with curves in Photoshop and adjust to whatever you like.
Raw it doesn't matter you have all the dynamic range available.

The only comparison I've seen about highlight recovery of the 5DMKIII versus the D800 showed the D800 to be better I think it was a Dutch site,

xenon-mkd said:

xenon-mkd wrote:

For amateurs and ethusiasts (budget minded) there won't be anything from Canon, it has left these waters.

5d mk3 is an insult!

Click to expand...

Really??? An insult?? I'm thinking that if you can't good good pics from a 5D3 than maybe you should find another hobby. Besides, no one's forcing you to buy the thing....so why be insulted?

Clint
http://www.flickr.com/photos/60455482@N00/

joema1 said:

Those same tests show the D3200 has better dynamic range than the D4 which costs 10 times as much.

Click to expand...

That's because the D3200 has a Sony EXMOR sensor, while apparently the D4 does not. So, no surprise here at all. The read noises on the D7000, D5100, D3200, & D800 sensors are nothing short of spectacular.

joema1 said:

And they show the Canon 70-200 IS II is significantly worse than the older, cheaper 70-200 IS USM.

Click to expand...

It's well known that lenses can exhibit extreme copy variability. So until sites do comprehensive tests across many copies of lenses, I only take those reviews/tests as very generic information. This sort of variability does not, however, generally exist across copies of sensors, which makes the sensor analyses quite reliable.

onemoremile said:

DxO does not take into account accuracy and speed of autofocus, quality of the metering system, quality of construction, weatherproofing, handling, ...the important question for me is whether it will help me make better pictures....

Click to expand...

Indeed, answer to the question is most of them don't help you make "better" picture.

onemoremile said:

... noise levels,

Click to expand...

They do. It is called sports score. I.e. at what ISO the camera still delivers a set standard iq.

onemoremile said:

I think that many modern digital SLRs have sensors that test out very well;

Click to expand...

You think wrong. By 2012 standard, most Panasonic, olympus, canon, Samsung sensor dont make it into "ver well" category.