I was at the focus show today, I was allowed to put in a CF card and take RAW and Jpeg images, please note this was a BETA camera. I can't find a program to open the RAW format at this time, but will upload the images as soon as I can, the images will only be available for a short time due to bandwidth allowed. I'll post my opinion on the camera, overall I'm more impressed than I was last week. It's solid, silent shooting is impressive although the FPS decreases with that option on. Acquiring focus is excellent although it was static subjects and didn't search at all. Please note NR is a factor in JPEGs however I'll upload the RAW files to a upload site very soon and even then slight NR is applied which can be switched off. I didn't go through the menu's to find it due to time allowed.
For JPGs, these look atrocious! There's pronounced shadow noise at ISO 100 and 400. This doesn't ultimately do anything for me since I shoot RAW but I'm anxiously awaiting RAW file performance.
snapsy wrote:
It can be done on every other Canon DLSR. What leads you to believe the 5DM3 is different in this respect?
There is no option to switch off RAW NR on any Canon DSLR. Unless you are talking about long exposure NR setting. Where the camera takes a second dark frame to eliminate noise. If that is what the OP was mentioning then whatever. But it didn't sound like this to me.
arbitrage wrote:
There is no option to switch off RAW NR on any Canon DSLR. Unless you are talking about long exposure NR setting. Where the camera takes a second dark frame to eliminate noise. If that is what the OP was mentioning then whatever. But it didn't sound like this to me.
What leads you to believe there is raw NR on Canon DSLRs?
snapsy wrote:
What leads you to believe there is raw NR on Canon DSLRs?
It has been proven that there is (I will try to google and find where I read this, it was a while ago). Nikon has a lot more NR on RAW files. Canon's press release for the 5D MkIII states that they have applied NR before it hits the Digic 5+ so I assume this would be in the RAW file before the Digic 5+ converts to jpeg and does all the other jpeg crap including more NR and picture styles and whatever.
arbitrage wrote:
It has been proven that there is (I will try to google and find where I read this, it was a while ago). Nikon has a lot more NR on RAW files. Canon's press release for the 5D MkIII states that they have applied NR before it hits the Digic 5+ so I assume this would be in the RAW file before the Digic 5+ converts to jpeg and does all the other jpeg crap including more NR and picture styles and whatever.
Thanks, I'd be interested in whatever links you can find. Btw, my threshold for "proof" is rather high, as in an analytical noise power analysis, including an FFT demonstrating pixel correlation.
snapsy wrote:
Thanks, I'd be interested in whatever links you can find. Btw, my threshold for "proof" is rather high, as in an analytical noise power analysis, including an FFT demonstrating pixel correlation.
I don't have the knowledge of physics/engineering to even understand the terms you used. From the above, I gather you know a lot about this already, and maybe you are just pulling my chain and trying to get me to pull up some evidence that you have already seen and already feel isn't valid. If this is true then I find that disrespectful as you may as well just come out and say that. If you haven't actually researched this topic then I will try to search and find the info that I had read (this may have been over a year ago and would have been originally linked from a thread on either FM or DPReview).
I'm not trying to be or come across as an expert on this topic at all. I had just read threads and linked articles in the past. Please let me know if I should try and find the info that I read but if you have read and maybe even discussed this on the forums before then a link to past discussions would be great.
arbitrage wrote:
I don't have the knowledge of physics/engineering to even understand the terms you used. From the above, I gather you know a lot about this already, and maybe you are just pulling my chain and trying to get me to pull up some evidence that you have already seen and already feel isn't valid. If this is true then I find that disrespectful as you may as well just come out and say that. If you haven't actually researched this topic then I will try to search and find the info that I had read (this may have been over a year ago and would have been originally linked from a thread on either FM or DPReview).
I'm not trying to be or come across as an expert on this topic at all. I had just read threads and linked articles in the past. Please let me know if I should try and find the info that I read but if you have read and maybe even discussed this on the forums before then a link to past discussions would be great....Show more →
Not at all. If I was jerking your chain I would ask you for what evidence you have and then tell you why I think it's insufficient. Instead I indicated up front what type of evidence is necessary to prove any claims about raw NR. I've done my own searches in the past and never found anything legitimate for Canon or Nikon DSLRs. There is proof for Pentax and Sony however.
arbitrage wrote:
It has been proven that there is (I will try to google and find where I read this, it was a while ago). Nikon has a lot more NR on RAW files. Canon's press release for the 5D MkIII states that they have applied NR before it hits the Digic 5+ so I assume this would be in the RAW file before the Digic 5+ converts to jpeg and does all the other jpeg crap including more NR and picture styles and whatever.
Not true!! Nikon gives you the ability to turn High ISO RAW off completely. And when it is on it will only kick in at higher ISO.
