DxOmark just published the Canon 650D sensor results. And once again we can see how Canon is struggling to get on par with Sony sensors. The Sony A65 dynamic range the beats the Canon by one stop! Still the Sony cameras would be even much better without the Translucent mirror: “we know from having tested various types of Translucent Mirror Technology that such systems usually reduce low-light scores by about 1/3 to 1/2EV. The SLT-A65 does in fact divert a portion of incoming light to support its electronic viewfinder, and… this is pretty much the only reason that the 650D can compete with Sony’s DSLR video in terms of low-light sensitivity.”
Anyway, the difference between the A77 and NEX-7 are negligible (here is the comparison). I really want to see how good the new [shoplink 14165]Sony A99[/shoplink] and RX1 sensor will be. And of course, if there is a major difference between the two.
Coming back to Canon, at least on the sensor front they are now clearly behind the competition. And keep in mind that Nikon, Olympus and very likely also Panasonic are now using Sony sensors too!
—
Sony preorder list:
A99 body Amazon, Adorama, BHphoto, Jessops, Wex UK, Wex Germany, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH and more links at Alphacameradeals.
A99 with kit lens links Alphacameradeals.
RX1 Amazon, Adorama, BHphoto, Wex UK ,Wex Germany and more links at Alphacameradeals.
NEX-6 body Amazon, Adorama, BHphoto, Wex UK, Wex Germany Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH and more links at Alphacameradeals.
NEX-6 with kit lens Amazon, Adorama, BHphoto, Jessops, Wex Uk ,Wex Germany and more links at Alphacameradeals.
NEX-VG900 Amazon, BHphoto, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH, Wex Uk and more links at Alphacameradeals.
NEX-VG30 Adorama, Amazon, BHphoto, Wex UK, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH, and more links at Alphacameradeals.
Sony E-mount 16-50mm Amazon, Adorama, BHphoto, Wex UK, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH, and more links at Alphacameradeals .
Sony E-mount 35mm Amazon, Adorama, BHphoto, Wex UK, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH, and more links at Alphacameradeals.
Sony E-mount 10-18mm Amazon, Adorama, BHphoto, Wex UK, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH, and more links at Alphacameradeals.
Sony A-mount 300mm f/2.8 Amazon, BHphoto, Sonystore DE–UK–FR–IT–ES–NL–SE–FI–CH.
37 Responses
Its mine:-)))
“divert a portion of incoming light to support its electronic viewfinder”
No it doesn’t. To support the AF sensor, the EVF is just another form of live view, main sensor data, otherwise EVF and OVF would be almost the same. Or so I see it.
yeah, they made a huge mistake here!
No, they didn’t, Sony SLRs lose a fraction of the light which enters through the lens to support PDAF and metering :
http://www.petapixel.com/2011/08/29/sonys-translucent-mirror-may-reduce-detail-in-photos-by-up-to-5/
From Wikipedia :
Sony SLT cameras have a mirror, but – unlike the one in DSLRs – it does not move, and it is semi-transparent, allowing the majority of the light to pass through to the sensor whilst reflecting a portion of it onto a phase-detection autofocus sensor, this way allowing full time phase detection auto focus.[1]
To support PDAF is not the same as to support the EVF. So yes, they were mistaken there.
Yes, but how does this change the meaning of what they wanted to say? PDAF or EVF it doesn’t matter, Sony loses a percentage of the light entering through the lens and if it weren’t the case, Sony cameras should have had slightly better scores, theoretically. This is exactly what DxO said and I think they’re not mistaken here.
learn to read
Don’t read wikipedia. It’s articles on SLT are one huge pile of crap.
(That said – if there’s any wikipedian reading this – fix this shit, please)
Wow. I’m very surprised too. Especially from DXO… Theses guys are supposed to know very well technical aspects of photography… EVF is, right, just another kind of live-view in a tiny screen, and doesn’t need any “light loss”, light deviation” or anything like that. I hope that it’s just a typo…
Wow, that’s a huge gap between the two. My old a100 had a similar score than in 650D – in 2006!!!!
Andrea, are there any new rumours? A1 perhaps???
Working on many of them. Don’t worry, rumor time will come back soon 🙂
Thank you. Keep up the good work.
722 vs 717. How many stops better does that make the Canon? 😛
The Nikon 3200 only scores 1131 using the same sensor as the 65/77, which again, isnt really that much difference. This whole translucent mirror thing is way over-emphasized.
on par with my a200!! no bad.. 🙂
i mean.. not bad.. 😀
What is it with Canon and dynamic range? Both low-end cameras and their top offerings suck big time when it comes to dynamic range of the sensor.
they’re too busy putting their heads together in their M-series (which will end in rumbles soon anyway)
It’s due to Canon’s noise reduction technique: it sacrifices part to of the available dynamic range.
long live Sony, Canon’s going down… (don’t like that? sue me)
I used to have the Canon 5D and all L primes(the F1.2 ones), I have to say that with Canon, I had to go thru over 20 steps of post processing to get the result(color, dynamic range, SHARPNESS…) I wanted, but with Sony, I only needed less than 5 steps of post processing, some photos just stay the way they are straight out from Sony body, and believe me or not, I’m comparing Canon 5D to Sony A100 image quality!
The number one company never stays number forever. I believe if sony gets there stuff together they will eventually be number one.
Is Canon becoming the 2nd Nokia? wait and see 😉
Apple will not be number 1 forever, for that matter.
BTW, Minolta was the most sold SLR back in the 80´s with the 5000 and 7000 series.
I’ve had it with Canon fanboys gloating when I had my A100 nearly a decade back. Now it’s my turn!
VeRy TruE !!!
The mirror drives the Autofocus Phase Detection – not the viewfinder!
The EOS 550D has a better DxOMark score than the newer 650D! I used my Sony A580 in a dark club one evening alongside a friend using an EOS 550D and they needed to add a couple of stops exposure equivalent to get the same results as me…
i think now a57 is better than canon 650d in every respect and not to mention cheaper too
My A290 has better rate than EOS 650 😀
I used to trust DXO’s scores, but after using a 7 year old Canon 5D Classic, I can say it’s just bullshit.
According to DXO, it has about 1/3 stop better ISO than 5N or 7. Which is just bullshit. It may be better at low ISOs but when you crank it to iso1000 or more, the newer Sonys are clearly superior by at least one step.
The high ISO rating between different sensor sizes can’t be compared directly.
It looks like the ratings depend heavily on sensor size, which is why Canon’s aps-c cameras have slightly lower than average high ISO scores (1.6 vs 1.5 crop factor). The reality is that the sensors from the same generation (and manufacturer) will perform about the same (including high ISO).
But then again, the 1D Mk IV’s 1300’s rating equals to one of the best low light cameras according to dpreview.
It will be interesting to see what dpreview + dxomark partnership will do to the current ratings.
Ok ok sony, i had my first DSLR, the a300 which i am torn between it and the canon XTi but i decided to go with and bought sony because of the articulating screen, live view and fast AF. but the image quality in terms of hiqh iso performance especially in JPG is obvious that canon has the edge. Now that i have the a77, comparing it to the canon 7D, iso 1600 and up, canon has cleaner and less noise than sony’s. and what’s with sony’s weird smudgy JPG engine? it improved over time but competitors also improved.
how come nikon’s DSLR using same sony sensors, have much better JPG engine and better at handling noise at high ISOs? firmware update with improved JPG engine and better noise compensation is all we need.
Few people care, Canon cameras are a cult are have marvellous slapping mirror.
Do you guys really think that DXOMARK’s numbers are 100% reliable, trustable, and dependable? Based on my own experiences, sometimes these DXOMARK numbers are meaningless in real life.
I’m so frustrated with DxOMark. Try in real world and you will not notice any difference. The link below is a test between 550D and NEX-5n (which is better than A65 according to DxOMark). The only difference I realize is if you apply +3 or more EV compensation to shadows you’ll see banding in 550D and Nex-5N’s shadows come out cleaner. Now, how many of you guys are actually applying +3 or more EV to shadows regularly?
http://halkboyleistiyor.blogspot.no/2012/10/canon-550d-ve-sony-nex-5n.html
Plus, 550D’s shadows have more light in “normally exposed” shots whereas 5N’s shadows get clipped: http://1.bp.blogspot.com/-xqHBGiqxwHc/UGy5phoS1WI/AAAAAAAAEPk/gly1xkuwOBo/s1600/1.jpg
DxOMark is overrated. I love my NEX-5n but 550D is not a failure as DxOMark shows.
In real life, there isn’t a big difference in output. The more important thing is that you know how to use your equipment and have the skillset to bring life to the picture. Most of us should spend more time learning how to shoot interesting photos, rather than reading blogs to decide if we should upgrade our cameras. A little noise doesn’t ruin any pictures, sometimes it adds a bit more character. A camera that is 1 or even 2 stops better in low light certainly won’t improve your pictures much in the real world.
I do use shadow pulling with DRO option 😉