Canon 7D owners..

[UUilliam]

Fair enough, I considered the fact that Crop sensor = less need to buy larger lens' allowing the photographer to save some cash
and I like the response to my post XD

I am guessing they fixed the fact that crop sensor = more noise than full frame then?
I would have thought they would have made it Canon 60D just to annoy everyone who paid for the 50D and to keep their Pro range with full frame..
then again... they only have 9 cameras in their pro range.. the 1D and the 5D and now the 7D
(I know there is 1D, 1D II, 1D III, 1Ds, 1DsII and 1Ds III and soon to be 1D IV and 5d, 5d ii but yeah.)
ahh just ignore me

 

[Overread]

I am guessing they fixed the fact that crop sensor = more noise than full frame then?

nope. The 5DM2 and other fullframe options will still beat the 7D in the noise department. That is not to say the 7D has bad noise control (from what I have heard it appears to have very good noise control). Its a factor of physical and pixel densities and stuff.
But as pointed out hte crop sensor camera has a different market than the fullframe - which is not to say that either camera cannot do the others job (and do it well) but to say that when given the choice some will go for more "apparent" focal length in the crop sensor options whilst others will want that larger frame capacity of the fullframe cameras.

Heck canon even make a 1.3 crop sensor body at the full 1DM3 range - so crop does have a market purpose.

As for cheaper glass - canon have a policy that thier best (L grade) is all fullframe compatabile and whilst some EFS lenses are very good (optically L grade) they are no where near as large a range as the fullframe lens options. So when people are dropping money on a 7D chances are they are serious and so after some really good glass - they will only find that in thefullframe options mostly (exceptions would be ultrawide angles and things like the EFS 60mm macro).

 

[epp_b]

Why the hell did canon make a Crop frame Pro DSLR anyway? and why did everyone buy it?

I thought the same at first, but after talking to someone who recently bought one, I realized that it competes directly with the Nikon D300.

 

[beni_hung]

I absolutely love my 7D. Definitely worth putting myself further into debt :mrgreen:

 

[musicaleCA]

I am guessing they fixed the fact that crop sensor = more noise than full frame then?

<good stuff here>

Noise is not necessarily worse with higher pixel densities, as is sharpness of any particular lens. The difference is that the 100% crop of a FF sensor of non-equivalent pixel-density simply covers a larger physical area than a high-pixel-density crop sensor, and so shot noise is less apparent. If you down-sample the image from a high-pixel-density crop sensor so that it is roughly equivalent to the pixel density of the FF sensor you're comparing it to, the noise would be roughly equal (I would argue that the 7D would likely outperform the 5D in this area, since the sensor was redesigned to significantly reduce shot and thermal noise).

Larger pixels don't mean less noise, it just means the noise will be less apparent.

 

[Derrel]

Yes, the 7D clearly has more noise than the first-generation 12.8 MP EOS 5D.

Here's a comparison done by an owner of the original 5D and the new 7D.

Image Quality: 7D vs 5D original ? - FM Forums

Take note--the 2.5x larger sensor of the "old" 5D classic provides the same, or better, resolving power,and lower noise, than the new 7D. Why? The sensor of the 7D is 2.5 times smaller than the full-frame 5D's sensor, AND such small (4.9 microns) and densely-packed pixels demand lens performance that,frankly, a lot of Canon glass simply can not provide. it's quite clear to me that the 5D out-performs the 7D in noise performance AND in detail on a like-for-like basis.

The EOS 5D is a slight bit better than the Nikon D3 or D700 at the per-pixel level at ISO's up to an including 1600. The EOS 5D is a 12.8 MP sensor,and it's exceedingly clear that both Canon,and Nikon, understand that the 12 megapixel full-frame sensor offers an amazing combination of high-ISO capabiliy, low noise, easy processing, easy storage requirements,and a very high level of performance that eliminates a lot of potential issues with lenses, diffraction, camera movement, moving subjects, etc. Witness the EOS 5D, Nikon D3, Nikon D3s, Nikon D700. All 12.1 to 12.8 MP models.

It's pretty simple really: a larger capture format, digital or film, will out-perform a smaller capture format, even if the larger-format camera has a lens that delivers only moderate quality. The smaller the sensor, the higher the absolute lens quality required to get good performance. Since lenses for Nikon and Canon are/were designed mostly for 35mm film use, and it's exceedingly difficult to make a lens that is 2.5x better than what's in the lineups today, the sensors that are 2.3x larger than 1.5 or DX (Nikon and Sony) or 2.5 larger than 1.6x APS-C (canon), the full-frame cameras deliver higher resolution,higher acuity, and better noise profiles. The FF cameras cost more than the APS-C bodies,and in the case of Canon, the affordable FF cameras don;t have the feature sets many buyers want; the 5D Mark II's AF system is very centrally-weighted and not all that good for action work.

The Canon 7D has high framing rate at an affordable price, and was aimed to compete directly with Nikon's D300 and D300s bodies, which pretty much set the standard in the mid-priced body category, and which left the 40D and 50D looking kind of flat-footed and behind.

 

[MrLogic]

The EOS 5D is a slight bit better than the Nikon D3 or D700 at the per-pixel level at ISO's up to an including 1600.

The old 5D being better than the D3 / D700 in the ISO department in the 200 -1600 range? That's quite a claim. Can you prove it?


edit: for some reason the "old" 5D isn't listed in the imaging resource comparometer.

 

[Derrel]

Why don't you do some work and dis-prove it?

 

[Garbz]

Why don't you do some work and dis-prove it?

Here we go: It's just something this guy called Derrel said. There disproved. Everyone's claims carry equal weight until some source is cited. The onus is on you to prove your point, otherwise my friend you're just likely to be ignored.

Noise is not necessarily worse with higher pixel densities, as is sharpness of any particular lens. The difference is that the 100% crop of a FF sensor of non-equivalent pixel-density simply covers a larger physical area than a high-pixel-density crop sensor, and so shot noise is less apparent. If you down-sample the image from a high-pixel-density crop sensor so that it is roughly equivalent to the pixel density of the FF sensor you're comparing it to, the noise would be roughly equal (I would argue that the 7D would likely outperform the 5D in this area, since the sensor was redesigned to significantly reduce shot and thermal noise).

Larger pixels don't mean less noise, it just means the noise will be less apparent.

That's actually wrong on all accounts. Firstly the uncertainty of photo measurement is the square of the number of photons counted. So physics introduces an exponential decrease in noise with larger frame, but you don't get the same level when down sampling. Secondly you are assuming that the gaps between pixels are smaller when in fact they stay the same, the sensor isn't one contiguous surface, there are microscopic signals routed around on it. So as you cram more pixels on, less realestate is dedicated to the capture of photons.

So ultimately assuming two otherwise identical sensors one with a higher density, the noise would not be the same after interpolating down, the larger frame will win.

 

[beni_hung]

 

[musicaleCA]

Why don't you do some work and dis-prove it?

Here we go: It's just something this guy called Derrel said. There disproved. Everyone's claims carry equal weight until some source is cited. The onus is on you to prove your point, otherwise my friend you're just likely to be ignored.

Noise is not necessarily worse with higher pixel densities, as is sharpness of any particular lens. The difference is that the 100% crop of a FF sensor of non-equivalent pixel-density simply covers a larger physical area than a high-pixel-density crop sensor, and so shot noise is less apparent. If you down-sample the image from a high-pixel-density crop sensor so that it is roughly equivalent to the pixel density of the FF sensor you're comparing it to, the noise would be roughly equal (I would argue that the 7D would likely outperform the 5D in this area, since the sensor was redesigned to significantly reduce shot and thermal noise).

Larger pixels don't mean less noise, it just means the noise will be less apparent.

That's actually wrong on all accounts. Firstly the uncertainty of photo measurement is the square of the number of photons counted. So physics introduces an exponential decrease in noise with larger frame, but you don't get the same level when down sampling. Secondly you are assuming that the gaps between pixels are smaller when in fact they stay the same, the sensor isn't one contiguous surface, there are microscopic signals routed around on it. So as you cram more pixels on, less realestate is dedicated to the capture of photons.

So ultimately assuming two otherwise identical sensors one with a higher density, the noise would not be the same after interpolating down, the larger frame will win.

Heh. Given the first reply in your post, I find it funny that I'm saying this: Source?

The light-gathering surface of the 7D's sensor is essentially gapless, with microlenses directing all incoming light to one photo diode or another, so I really don't think my comments are baseless at all. Shot noise is going to be less apparent on a sensor with larger pixel pitch and consequently photo diode size, yes, but only because per photo diode, as you said, more photons are being counted. But that still doesn't mean that somehow a FF camera magically produces less noise; if you cram 44MP of photo diodes on a FF35 sensor, all things being equal, you're still going to see the same amount of noise as an 18MP 1.6 APS-C. So as I said, it's not that there's more noise, it's that it's more apparent on the higher-density sensor. (If we're talking thermal noise, well yeah, then larger sensors could certainly have a clear advantage simply because there's more potential real estate to devote to dissipating heat.)

Show me the maths and I'll believe you, Garbz. (Not that belief enters into it when talking math. )

 

[astrostu]

Why don't you do some work and dis-prove it?

Here we go: It's just something this guy called Derrel said. There disproved. Everyone's claims carry equal weight until some source is cited. The onus is on you to prove your point, otherwise my friend you're just likely to be ignored.

Noise is not necessarily worse with higher pixel densities, as is sharpness of any particular lens. The difference is that the 100% crop of a FF sensor of non-equivalent pixel-density simply covers a larger physical area than a high-pixel-density crop sensor, and so shot noise is less apparent. If you down-sample the image from a high-pixel-density crop sensor so that it is roughly equivalent to the pixel density of the FF sensor you're comparing it to, the noise would be roughly equal (I would argue that the 7D would likely outperform the 5D in this area, since the sensor was redesigned to significantly reduce shot and thermal noise).

Larger pixels don't mean less noise, it just means the noise will be less apparent.

That's actually wrong on all accounts. Firstly the uncertainty of photo measurement is the square of the number of photons counted. So physics introduces an exponential decrease in noise with larger frame, but you don't get the same level when down sampling. Secondly you are assuming that the gaps between pixels are smaller when in fact they stay the same, the sensor isn't one contiguous surface, there are microscopic signals routed around on it. So as you cram more pixels on, less realestate is dedicated to the capture of photons.

So ultimately assuming two otherwise identical sensors one with a higher density, the noise would not be the same after interpolating down, the larger frame will win.

Heh. Given the first reply in your post, I find it funny that I'm saying this: Source?

The light-gathering surface of the 7D's sensor is essentially gapless, with microlenses directing all incoming light to one photo diode or another, so I really don't think my comments are baseless at all. Shot noise is going to be less apparent on a sensor with larger pixel pitch and consequently photo diode size, yes, but only because per photo diode, as you said, more photons are being counted. But that still doesn't mean that somehow a FF camera magically produces less noise; if you cram 44MP of photo diodes on a FF35 sensor, all things being equal, you're still going to see the same amount of noise as an 18MP 1.6 APS-C. So as I said, it's not that there's more noise, it's that it's more apparent on the higher-density sensor. (If we're talking thermal noise, well yeah, then larger sensors could certainly have a clear advantage simply because there's more potential real estate to devote to dissipating heat.)

Show me the maths and I'll believe you, Garbz. (Not that belief enters into it when talking math. )

Garbz is right, you're wrong musicaleCA. Though Garbz made a slight mistake (I assume type-o since he and I have explained this concept MANY times on this forum) that the noise is the SQRT (square-root, not square) of the number of photons counted. If you want a source, there are hundreds of articles in the scientific literature. To quote from Wikipedia: "Light intensity from a single source varies with time, as thermal fluctuations can be observed if the light is analyzed at sufficiently high time resolution. Quantum mechanics interprets measurements of light intensity as photon counting, where the natural assumption is to use the Poisson distribution. When light intensity is integrated over large times longer than the coherence time, the Poisson-to-normal approximation is appropriate." And if you do any basic search on Poisson statistics, you will see that the standard deviation is the SQRT of the counts (the mean).

Very straight-forward. And because I've explained at length why down-sampling one sensor to the size of another is not equivalent, I will refer you to this TPF thread, specifically reply #3 and reply #9.

 

[Derrel]

There are quite a few good, somewhat older comparisons between the Canon 5D and the Nikon D3 and D700 in term sof per-pixel resolution, acutance, and noise. Garbz's conttntions about noise are very full of mathematicall blowhard numbers and incorrect assumptions, like his disagreement that noise is noise. The actual facts will tell you that there is chroma noise and luminance noise, and one looks like crap, and the other looks like film grain. Garbz seems to have the mind of a college student who studies physics, but lacks eyes and actual experience in photography to trust his own senses; noise "measurements" are very subjective,and quite often cameras with lower mathematical noise measurements look WORSE than cameras that turn in higher noise numbers on test charts. That is why noise "numbers" are not the same as actual images in terms of judging noise in terms of how the images actually "look".

Shocking as it might be, the Nikon D3 camera out in 2007, with the D700 following not too many months behind, at seven months after. The EOS 5D has a 12.8 megapixel sensor, the D3 and D700 have slightly smaller sensors with 12.1 or so effective megapixels. Anybody who wishes can do a Google search on noise comparisons between EOS 5D and Nikon D3 and D700,and will eventually find some real-world comparisons. Some of them were done on dPreview by a couple of well-known commercial studio photographers who were interested in the best FF d-slr performance. I downloaded the sample images back in 2007 and early 2008,and my own EYES showed me that the "old" Canon 5D delivered a slightly higher-resolution, higher-acutance picture than either the Nikon D3 or the D700. Very simple really--I looked at actual images, not noise graphs. It's a testament to Canon's sensor design, microlens design, and internal software that it took Nikon almost three years to come up with a FF sensor that ALMOST equalled the 5D's overall,total imaging capabilities in the 200-1600 range.

Visually, that the "old" EOS 5D is the equal,or better, of the Nikon D3 and D700 in terms of per-pixel sharpness, which is a combination of resolution and acutance, as well as noise, from ISO 200-1600. It is also very clear, according to the URL I linked to above that at EQUAL IMAGE SIZES, the new 7D loses out in terms of noise,and shows lower detail than the "old" Canon 5D. Both Canon and Nikon have figured out a 12-12.8MP FF sensor is sweet!

I love how a Norwegian Nikon lover challenges me, and how a newbie who has been involved in photography for less than a year challenges me, one no doubt thinking I am a Canon lover basking Nikon, the other thinking I am bashing on the two brand new 7D's he bought,and eager to try and convince me what I have seen with my own eyes is invalid. Funny...I own both systems. I like both, for different things. Everybody sees bias where none exists. Nikon fans and Canon fans see bias where none exists.

Check out MAJOR Nikon fanboy Ken Rockwell's comments here Nikon D3 and D700 vs Canon 5D
"The Canon 5D excels in image quality. It's also the lightest and least expensive of these cameras. It has the sharpest pictures of these three cameras by a slight margin. If you're backpacking or shooting careful landscapes, the 5D could give superior image quality above anything from Nikon, at a bargain price. If you want the best picture quality for landscapes and huge enlargements, the 5D is the best camera. I'm not kidding: I have 20x30" (50 x 75cm) prints here, and the 5D is clearly superior. (Then again, medium format film is superior to any of these in 20x30" prints.)"

Wow,in actual tests of the three, the 5D proves to be the better imager!

Nikon D3, D300 and Canon 5D Sharpness Comparison
Canon 1Ds Mk III vs. Canon 5D vs. Nikon D3
Nikon D3X Sharpness Comparison

One of the most interesting tests was comparing 20x30 inch prints from the D3 and 5D and Nikon D40 here: 20x30" Print Comparison
Where the conclusion about PRINTED versus on-screen comparisons was made,
"I didn't expect such an obvious difference, but if you're too close, the Canon 5D is clearly superior to the D3. I shot the world's best 14-24mm on the D3, and Canon's not-as good (in the lab) 16-35mm L II on the 5D, and the 5D shot still smokes the D3 in sharpness. I shot the D3 on a tripod, and shot the 5D freehand. Both files were the same size.The 5D image is much better, but does show some lateral color fringes which the D3 corrects itself. If I had used the NEF from the D3 (shot but not shared here), many raw converters don't remove the fringes as does the D3 in its own JPGs.The on-print results are more striking than the differences seen on-screen at my sharpness comparison." Wow, yet another real comparison,and the "old" 5D makes a betetr 2030 inch print.

Combine Ken's multiple tests over several years, plus the tests of Fredric Frauhaus,and other commercial photographers, and it's pretty clear that Canon's original 5D sensor and image processor was *far* ahead of its time.
It's not just pixel count that determines how a camera performs, but how the images turn out. One of the more striking comments was from Rob Galbraith about why he prefers the 10.1 MP Canon 1D Mark III over the 12.1 MP Nikon D3: per-pixel image quality of the 1D Mark III is higher than that of the D3.

Last reference. Go here Nikon D3 Review: 18. Photographic tests: Digital Photography Review

and use your own eyes to see which camera has more detail at ISO settings from 200 to 1600. The winner? Canon 5D, by a nose.

 

[UUilliam]

OH GOD LOOK WHAT I STARTED!!!!! ale:ale:

 

[AverageJoe]

:cheers: