Hathaway said:I continue to find it interesting that the technical people of these forums (Canon Rumors, DPReview, FM, etc.) only focus on Fred's pictures of the shadows from the buildings. I will be the first to say it is extremely impressive what Sony/Nikon has done with the sensor, but if you continue to read Fred's review and look at all of the pictures he takes for the articles, he points out some serious flaws in focus with the Nikon.
I think it's important that we all bitch a bitch so that Canon hears loud and clear. DR is very important to me. I shoot people wearing helmets with brims during the mid-day sun. High DR means being able to turn dark shadows into colorful, natural-looking faces. Hell, just today I sold three photos where I pushed the hell out of the shadows and got really nice looking outdoor event shots that looked like portraits. This is actual money in my pocket that I would not have had shooting with a Canon, and I charge a bit for my photos. Unless you can control the light, there will always be situations where you need to bring color out of shadows.V8Beast said:So if DR is that important to you - whether it's for practical reasons or just to placate your ego - admit that Canon is inferior at the moment, and buy yourself a D800![]()
sarangiman said:I shot a Stouffer Transmission Wedge (13.2 stops DR total) with a 5DII, 5DIII, & a D7000.
The 5DIII required 1/3EV less exposure than 5DII to keep any channel (meaning: green, since that blows first) from blowing out in the brightest patch of the wedge. The D7000 needed 2/3EV less exposure than 5DII (i.e. 1/3EV less exposure than 5DIII) to keep any channel from blowing out.
But regardless of the highlight handling, the D7000 retained a SNR>1 down to the darkest patch, giving it ~13EV DR if your lower limit is SNR of 1. The 5DIII was stuck somewhere around 11.3EV DR (again, assuming lowest acceptable SNR is 1), b/c its SNR went to 1 at patch #35 (whereas D7000 maintained SNR>1 down to patch 42). These patches are 0.1 density increments, so you can calculate the DR by doing the following calculation:
DR = log2[(10^#of patches for which SNR>1)/10)
So that pretty much debunks the whole theory that the Canon sensors might still have really good DR b/c of how they handle highlights.
smirkypants said:I think it's important that we all bitch a bitch so that Canon hears loud and clear. .
I am now seriously considering selling all of my Canon gear.
dilbert said:You misunderstand. The clean shadows means very low (almost non-existent) read noise on the sensor.
That read noise is going to be in every capture, it is only in shadows where it can be most easily demonstrated.
In effect, what this means that even a photo of a blue sky and beach (no shadows or anything) is going to be cleaner on the D800 than the 5D3, even if the signal drowns it out.
dilbert said:V8Beast said:Personally, I very rarely have the need to push the shadows that much, and the 5DIII is a better balanced machine for my needs. I just don't see why I should get upset about one aspect of camera performance (DR) that doesn't affect what I shoot all that much.
You misunderstand. The clean shadows means very low (almost non-existent) read noise on the sensor.
That read noise is going to be in every capture, it is only in shadows where it can be most easily demonstrated.
In effect, what this means that even a photo of a blue sky and beach (no shadows or anything) is going to be cleaner on the D800 than the 5D3, even if the signal drowns it out.
Now I understand why when processing raw files from Canon cameras the "black" value is considered to be 2048 (anything under 2048) on a scale of 0-16383 (for each of red, green and blue).
smirkypants said:V8Beast said:So if DR is that important to you - whether it's for practical reasons or just to placate your ego - admit that Canon is inferior at the moment, and buy yourself a D800![]()
I am now seriously considering selling all of my Canon gear.
I am now seriously considering selling all of my Canon gear.
the way I see it they already have it. D700 does really good at 8fps and it costs about the same as a 5DMKII. And if you don't need 36MP, 12MP is plenty then. The rumored D600 is also interesting. 24MP sensor with the D800 processor could make it very very fast and push the ISO to 5DmkIII levels or beyond. If they keep that 14stops of DR then they could have a true D700 successor.so as i see it on nikon forums none are complaining about the dr. they are saying they wanted a camera more like the mkiii in function
risc32 said:I ALMOST can't believe we are being told to worry about noise on all of our imagines that we admittedly can't see. That reminds me about the release of the DVD format. All the Laserdisc guys went on and on about the invisible noise from the digital mombo gumbo tricks that are going on, and how it's everywhere and our pictures will suck for it. yeah?
Astro said:and has not changed with the 5D MK3.
KeithR said:Yeah, about that...Astro said:and has not changed with the 5D MK3.
It has changed.
astro said:people complaining about the 5D MK2 shadow noise forever, that is nothing new... and has not changed with the 5D MK3.
and none of the 5D Mk II-style pattern noise the LuLa review managed to dig out.
Frankly I've no idea how Michael Reichmann got such poor results
It's night and day, Keith. I'm sorry, it really is. Once again, I'm not trying to just whine or complain needlessly... This is real world stuff here that affects my bottom line.KeithR said:I've downloaded and processed a number of 5D Mk III files, and I've been able to recover a massive amount of data from the shadows with no significant (certainly no unmanageable) noise penalty, and none of the 5D Mk II-style pattern noise the LuLa review managed to dig out.