Canon is getting owned in sensor technology

[V8Beast]

Canon is to old slide film as nikon is the negative film... The slides just had better pop, better and vibrant colors, more "wow" factor... Negatives COULD have more latitude but in the end, a lot more work/care was needed to get negatives in print to get remotely close to the slide image. The same is being proven true with D800 files and the 5d3 files...

That's the best damn analogy I've read in a while. It almost makes me want to pop some Kodak E100vs slide film and go shooting. Oh wait, I can just shoot with my 5DIII - with damn near the same IQ and pop of color slides - and get instant gratification without paying for film and processing

 

[V8Beast]

90% of all visitors of DXOmark have NO clue how DXO is measuring sensor performance.
and from the 10% who know, only half understand why they measure that way.

This is EXACTLY why Canon has an enormous marketing problem on their hands.

I seriously doubt DxO has that much clout. Outside of forums like this, most consumers have never even heard of DxO. For the sake of argument, let's say DxO's test results influenced purchasing habits as much as people suggest. Canon could easily create a puppet company to rival DxO with test standards biased to favor Canon sensors. Hell, it would probably be cheaper than investing in the R&D necessary to actually improve the sensors. Tech savvy consumers would see right through it, but the "dumb rich people" that people in this thread are referring to wouldn't have a clue.

 

[V8Beast]

But, let's be honest. The cheapest DSLR is way more than adequate for 90% of situations, regardless of brand. The 5D II and the D700 were more than adequate for probably 98% of situations. Now, the 5DIII and D800 are going to be more than adequate for 98.4% of situations.
People need to keep a little perspective here.

Well said. I'll take it one step further and say that a $500 Rebel is more camera than 95% of buyers can handle. IMHO, some people enjoy marveling at tech specs more than actually using their gear. Nothing is more pathetic than having a high-dollar camera, and getting smoked by someone with far lesser gear, yet it happens all the time.

 

[V8Beast]

I suspect thats a little out there, ,and that DXO's reputation would preclude this from happening. It might not be trivial to accomplish, but some independant test or even a careful real-world evaluation could show evidence of error and DXO would be toast.

D800 vs Medium Format with Roth and Ramberg

According to DXO, the Nikon should have won in dynamic range big time (14EV) but the hasselblad won. (12.5EV)

Stuff that in DXO's lens mount and smoke it.

I have to give credit where credit is due. The D800 performed extremely well going up against a medium format body in this vid. The shadow recovery is very impressive. I still don't like it's colors, though

 

[Bosman]

But, let's be honest. The cheapest DSLR is way more than adequate for 90% of situations, regardless of brand. The 5D II and the D700 were more than adequate for probably 98% of situations. Now, the 5DIII and D800 are going to be more than adequate for 98.4% of situations.
People need to keep a little perspective here.

Well said. I'll take it one step further and say that a $500 Rebel is more camera than 95% of buyers can handle. IMHO, some people enjoy marveling at tech specs more than actually using their gear. Nothing is more pathetic than having a high-dollar camera, and getting smoked by someone with far lesser gear, yet it happens all the time.

Lets take it a step farther, numbers and "Facts" are being spouted and regurgitated yet most of those doing the science vomit don't actually have either camera. Nor have they looked at the capture and marveled at how well it captured the scene and how pretty much no post work need be done to the 5d3 images. Oh, never mind that silliness, the numbers prove i am talking out my A**. Only numbers say whats true about my experience with the camera. What was i thinking???

 

[Viggo]

But, let's be honest. The cheapest DSLR is way more than adequate for 90% of situations, regardless of brand. The 5D II and the D700 were more than adequate for probably 98% of situations. Now, the 5DIII and D800 are going to be more than adequate for 98.4% of situations.
People need to keep a little perspective here.

Well said. I'll take it one step further and say that a $500 Rebel is more camera than 95% of buyers can handle. IMHO, some people enjoy marveling at tech specs more than actually using their gear. Nothing is more pathetic than having a high-dollar camera, and getting smoked by someone with far lesser gear, yet it happens all the time.

Lets take it a step farther, numbers and "Facts" are being spouted and regurgitated yet most of those doing the science vomit don't actually have either camera. Nor have they looked at the capture and marveled at how well it captured the scene and how pretty much no post work need be done to the 5d3 images. Oh, never mind that silliness, the numbers prove i am talking out my A**. Only numbers say whats true about my experience with the camera. What was i thinking???

I'll jump on this as well. I use the color checker passport to calibrate my colors, and whilst the 5d2 showed a MASSIVE improvement, the 5d3 adjusts a little, but waaaaay less. So for those how doesn't bother with the ColorChecker, the colors from the 5d3 vs the 5d2 is crazy better.

And I saw something about slides and negative compared Nikon vs Canon. Could not agree more. I took home with me some D800 and D4 files, and as soon as I touched them in Lightroom they started to fall apart, I was actually very surprised. Guess I'm spolied with the rubberband/wet clay files from Canon.

 

[hoghavemercy]

Okay, we can cry bias and be in complete denial, but Canon has inferior sensor from the early 2000! Their tech is a decade behind than Nikon. When will they start accepting the fact and invest a little more? It used to be Canon could at least claim "we have high megapixels", but now they can only claim "DXO is biased".

We need competition, otherwise Nikon will become stagnant, just like Canon has become. Sorry for the rant, not trying to rub salt on our collective Canon wounds.

at the end of the day it all comes down to my 5DIII ;D, really surprising a lot of these hatred for Canon came out after DXO reviews, you could've complained weeks earlier. you could always return yours and cross over to Kenro and be his bitch, get off the pc and shoot, relax it's a long way for the perfect sensor.

 

[RuneL]

http://nofilmschool.com/2012/04/markii-vs-markiii-vs-d800-candlelight/ check out this, it's video, but Canon is miles ahead.

 

[LetTheRightLensIn]

http://nofilmschool.com/2012/04/markii-vs-markiii-vs-d800-candlelight/ check out this, it's video, but Canon is miles ahead.

Yeah the 5D3 has much better SNR for video because it uses the whole sensor while the D800 only uses 1 out of every three lines of the sensor. 5D3 is a bit soft, perhaps because the 3x3 mapping means they need to apply lots of extra AA because the built-in filter is too small scale. Adding a 1.6x cropped video might make for a sharper option with more reach, quite a shock it didn't ship with such a mode. One can they will still add it (alogn with zebra stripes and, if possible, focus peaking).

 

[PhilDrinkwater]

But, let's be honest. The cheapest DSLR is way more than adequate for 90% of situations, regardless of brand. The 5D II and the D700 were more than adequate for probably 98% of situations. Now, the 5DIII and D800 are going to be more than adequate for 98.4% of situations.
People need to keep a little perspective here.

Well said. I'll take it one step further and say that a $500 Rebel is more camera than 95% of buyers can handle. IMHO, some people enjoy marveling at tech specs more than actually using their gear. Nothing is more pathetic than having a high-dollar camera, and getting smoked by someone with far lesser gear, yet it happens all the time.

Lets take it a step farther, numbers and "Facts" are being spouted and regurgitated yet most of those doing the science vomit don't actually have either camera. Nor have they looked at the capture and marveled at how well it captured the scene and how pretty much no post work need be done to the 5d3 images. Oh, never mind that silliness, the numbers prove i am talking out my A**. Only numbers say whats true about my experience with the camera. What was i thinking???

I'll jump on this as well. I use the color checker passport to calibrate my colors, and whilst the 5d2 showed a MASSIVE improvement, the 5d3 adjusts a little, but waaaaay less. So for those how doesn't bother with the ColorChecker, the colors from the 5d3 vs the 5d2 is crazy better.

I think this just means lr has got better at producing correct profiles over the years. The d3 profile showed a lot of difference too whereas the d3s was much closer to reality.

Ill check the 1dx but I might not use a profile if it's close enough.

 

[Hillsilly]

But I must say the d800 score really shocked me since they even beat some of the medium format cameras.

Well, the Sony Alpha 900 has scores that beat the Leica M9, and Hasselblad H3DII 50 and so on.

And the Pentax K5 (with crop sensor) beats the 5D3.

I've no problem acknowledging that the Nikon might have a "better" sensor as tested with the DXO methodology. But it doesn't mean it produces better pictures. Canon have rarely been competitive with DXO rankings, but I still prefer the look of the photos taken with their cameras.

 

[dlleno]

I suspect thats a little out there, ,and that DXO's reputation would preclude this from happening. It might not be trivial to accomplish, but some independant test or even a careful real-world evaluation could show evidence of error and DXO would be toast.

D800 vs Medium Format with Roth and Ramberg

According to DXO, the Nikon should have won in dynamic range big time (14EV) but the hasselblad won. (12.5EV)

Stuff that in DXO's lens mount and smoke it.

Assuming the above is itself true and un-adulterated, it doesn't necessarily mean that Nikon Paid DXO off, but a series of results similar to the above would at least strengthen the case that DXO test methodology doesn't perfectly and alwasys correlate with real results for the purposes of comparing two sensors of different design heritage. Whats astonishing to me is to read comments like "numbers don't lie" when the real question is "what do the numbers mean and how do they correlate with things that are important?". DXO can measure whatever they want, their test methodology and scoring policy may tend to favor one sensor over another I dont' know; Their staff may be even filled with Nikon fanboys for all I know, but it doesn't mean Nikon paid them extra to deliberately falsify information or contrive a test that favors the company who paid them the most to do so.

what we really need, in order to interpret the DXO numbers, is to show if and to what extent the numbers (especially the differences between two sensors) correlate to the final result, how the photographer either benefits from the difference or not, or at least how the difference will affect what the photographer does.

 

[jrista]

I suspect thats a little out there, ,and that DXO's reputation would preclude this from happening. It might not be trivial to accomplish, but some independant test or even a careful real-world evaluation could show evidence of error and DXO would be toast.

D800 vs Medium Format with Roth and Ramberg

According to DXO, the Nikon should have won in dynamic range big time (14EV) but the hasselblad won. (12.5EV)

Stuff that in DXO's lens mount and smoke it.

I have to give credit where credit is due. The D800 performed extremely well going up against a medium format body in this vid. The shadow recovery is very impressive. I still don't like it's colors, though

Having watched a few of that guys video reviews now (and a couple direct comparison videos that pit Nikon cameras against Canon cameras), it seems that Nikon cameras overexpose the scene a bit. That makes their shadow DR and the recoverability there make a lot of sense...Nikon cameras cater to the shadows, rather than to the highlights, where as medium format cameras and Canon cameras cater to the highlights, rather than the shadows...which would also indicate why there is a shadow DR noise floor problem on those types of cameras. I am not sure I could say that Canon cameras had quite as much highlight recovery as a Hassey, but they do seem to have more than the D800 (even something like the 7D has a LOT of highlight headroom.) I think you could improve shadow recovery with either a Hasselblad or a Canon camera by exposing to the right more. Not sure you'll reach the same DR performance as a Sony Exmor sensor, but you can probably improve things a bit over not exposing to the right.

 

[Bosman]

I think this just means lr has got better at producing correct profiles over the years. The d3 profile showed a lot of difference too whereas the d3s was much closer to reality.

Ill check the 1dx but I might not use a profile if it's close enough.

I haven't used lightroom but to try out the beta firmware. This has nothing to do with LR. My tests have mostly been jpeg too. I havent processed my images in LR. As Viggo stated after me, he used his color checker and finds the 5d3 to be a massive improvement also.