musclepics wrote:
From those samples, the 7D looks to be about 1 stop nosier than what I get from my 1D MkIII at higher ISOs. In fact the ISO800 sample pic looks like what I get at ISO3200 from the 1D MkIII, so about 2 stops behind there (yes even when resized to the same resolution).
I doubt it. Even the 50D is only like 1.5+ behind the 5D2 and the 5D2 does better than the 1D3.
Here's a decent sized 3200 ISO 7D file from IR for a bit more balance: - again, not too shabby, although it's not heavily shadowed, but over to the right there's a plant with shadow areas - it's worth a poke around that section with the image at full size, it still looks good.
It's quite illuminating to compare that with the 50D sample here.
Not only is the noise better controlled and the file a lot more organic looking, but the colors are much better (though this was no doubt captured before the 50D firmware magenta fix).
I think this bodes really well for the Mark IV cameras when this new sensor mojo shows up with larger pixel pitch. Wowza!
You can only compare noise between photos taken under IDENTICAL circumstances -- Same subject, sensor size, shutter speed ambient temperature and file type. Light coloured subjects will show less noise than dark ones. A 20 MP sensor will have 2 x the number of noise spots than a 10 MP sensor. A photo taken at 1/30 will have more noise than one taken at 1/60 and there be more noise at 40 deg C than at 0 deg C (both phenomena are well known in Astrophotography). RAW files may appear different from JPEGs that may have sharpening and/or NR applied.
How about this. To standardize, take some shots in the dark with the lens cap on and with the same shutter speed and at the same temperature. Take them all in RAW and apply no NR or sharpening. Crop them all at the center of the frame with a 1,000 x 1,000 pixel crop (1 MP identical sensor size). Measure the amount of noise per 1 MP. Do appropriate statistical tests on the data.
garyvot wrote:
Thanks for the instruction manual. Now, will somebody please post a link to a DPP 3.7 download?
That would be nice. I downloaded some RAW files a couple of days ago, but can't process them! Plus, I'd like to see what modest improvements were made to DPP this time around.
7D looks fine to me. I would prefer a 1.3 sensor but you know, under many circumstances it will be fine. I use FF but this would also work very well to balance everything out.
I personally feel RG should be lauded for his review and it should become the standard review for the 7D, with any luck it will cause prices to drop and I'll be able to afford one.
Overall, it really looks like RG's samples intentionally make both the 5 and 7 look bad. Not saying RG has an agenda, but those samples are really bad.
timbop wrote:
Overall, it really looks like RG's samples intentionally make both the 5 and 7 look bad. Not saying RG has an agenda, but those samples are really bad.
Actually, I'm sure that is RG's intention. He has also been slamming Canon's AF all this while. Not sure what happened... maybe something like what Phil Askey went through (they didn't provide him an advance copy of the 1D3 for early preview).
Yakim Peled wrote:
3. No red AF assist on the body. The Elan 50e was the last Canon to have it.
Not so fast there. The Canon D60 dinosaur had the AF assist lamp too. It was taken out because it's mighty useless. Tried it on the Nikon D90... total junk.
FYI FWIW - I was one of the folks able to get an order in with Amazon before the door closed ... and I just asked if they had any idea when it might ship ... and their response was they didn't know for sure, but they said it may be delivered by November 6th - D'OH! :-(
thw2 wrote:
Actually, I'm sure that is RG's intention. He has also been slamming Canon's AF all this while. Not sure what happened... maybe something like what Phil Askey went through (they didn't provide him an advance copy of the 1D3 for early preview).
I've worked with RG on a professional basis, and while we have not spoken in a couple of years, I do not for a minute believe that he has any kind of anti-Canon agenda. He's a professional journalist by training, and his articles have always been much better researched and more objective in tone than some other sites I could name.
I simply think that in this case, he has applied chroma NR in DPP, but has done little or no luma NR on the files. The results he's gotten would be consistent with that practice in my experience.
Some people just philosophically prefer to do minimal destructive processing to the files, and accept some luminance noise as being similar to film grain. I know for a fact that RG used to turn off his in-camera noise processing and sharpening in favor of a more normalized post-processing workflow.
But until he discloses more details about his workflow it's hard to say one way or another what has occurred. I'd hold out for more data points.
garyvot wrote:
I've worked with RG on a professional basis, and while we have not spoken in a couple of years, I do not for a minute believe that he has any kind of anti-Canon agenda. He's a professional journalist by training, and his articles have always been much better researched and more objective in tone than some other sites I could name.
I simply think that in this case, he has applied chroma NR in DPP, but has done little or no luma NR on the files. The results he's gotten would be consistent with that practice in my experience.
Some people just philosophically prefer to do minimal destructive processing to the files, and accept some luminance noise as being similar to film grain. I know for a fact that RG used to turn off his in-camera noise processing and sharpening in favor of a more normalized post-processing workflow.
But until he discloses more details about his workflow it's hard to say one way or another what has occurred. I'd hold out for more data points.
Alek Komarnits wrote:
FYI FWIW - I was one of the folks able to get an order in with Amazon before the door closed ... and I just asked if they had any idea when it might ship ... and their response was they didn't know for sure, but they said it may be delivered by November 6th - D'OH! :-(
Actually, from a dealer I spoke with today, Canon has not told dealers yet about shipment details- even the price can change, as dealer cost has not been disclosed to dealers yet. The MSRP is more of a guideline. Obviously if the exchange rate were to shift at the time of shipment, the price will change- much like what happened in Canada with the 5D mark II shipment.
On the other hand the IR images (To me anyhow) are quite harsh in an overly contrasty way. That too can be due to PP, but you cannot deny that Canon are extracting a lot of detail with that sensor.
Tim
