EOS Bodies - The ultimate duo & why

[epsiloneri]

Sorry, I was a bit sloppy in the distinction between DR and S/N. They are closely related but not exactly the same. I trust it's still evident from my post what I mean.

 

[WarStreet]

Sorry, I was a bit sloppy in the distinction between DR and S/N. They are closely related but not exactly the same. I trust it's still evident from my post what I mean.

Ironically, this was the ONLY thing I manage to understand. I like your contribution. Now, I only need to do some reading to decipher the above Any suggestion ?

 

[Heidrun]

My favoritt duo is difficult to say. I have to sat 2 bodies and 2 lenses.
First of all 1DS mk III with T&S 24 mm and 1D mk III with 24-70 mm

 

[epsiloneri]

Ironically, this was the ONLY thing I manage to understand. I like your contribution. Now, I only need to do some reading to decipher the above Any suggestion ?

Thanks. Yes, it unfortunately got a bit technical, but what I essentially try to figure out is whether there can be an image quality advantage when it comes to dynamic range for a 5D2 + a 1.4x tele converter, as opposed to a 7D only. In many cases the 7D is better, e.g., frames per second, autofocus speed/tracking, no additional optics (no tele converter) etc, but when it comes to dynamic range (the ability to image both bright and dark features simultaneously) I argue there might be an advantage for the 5D2+TC1.4 combo.

If you are interested to learn how properties of a detector affect image quality properties, I suggest Clarkvision: http://www.clarkvision.com/articles/digital.sensor.performance.summary/ which has a good introductory summary.

 

[Loswr]

In many cases the 7D is better, e.g., frames per second, autofocus speed/tracking, no additional optics (no tele converter) etc, but when it comes to dynamic range (the ability to image both bright and dark features simultaneously) I argue there might be an advantage for the 5D2+TC1.4 combo.

If DxOMark is correct, there is no meaningful difference between DR of the 5DII and the 7D. Would you expect using a TC to improve the DR of the 5DII?

 

[epsiloneri]

If DxOMark is correct, there is no meaningful difference between DR of the 5DII and the 7D. Would you expect using a TC to improve the DR of the 5DII?

Yes, but again, DR per solid angle, obviously not per pixel. The DR per pixel only depends on the sensor properties, not the optics (irrespective of whether there is a TC or not).

Let me give a simple example to show you why. Imagine you are using your 5D2 to image a chess board from a large distance. Each chess square takes up exactly one pixel. Now let's say you use a TC 2x. Each square of the chess board now covers four pixels. With four pixels you can collect four times as many photo-electrons for each square, meaning that the S/N (and essentially DR) improves by a factor of two per chess square, or per solid angle. If I have very a high contrast between black and white squares, I would get better signal in both by using the TC. (This of course assumes that the camera doesn't change position).

I cannot explain why DxOMark does not find a significant difference in DR between 5D2 and 7D. I didn't find a description of how they measure DR in detail, so I don't know what their numbers mean. It also stands in stark contrast to what is found on Clarkvision, which is much closer to my expectation. Perhaps DxOMark doesn't measure DR per pixel, but per sensor area or something. That would make sense. That a large 5D2 pixel would have the same DR as a small 7D pixel definitely does not make sense, so they must mean something different.

 

[Loswr]

Let me give a simple example to show you why. Imagine you are using your 5D2 to image a chess board from a large distance. Each chess square takes up exactly one pixel. Now let's say you use a TC 2x. Each square of the chess board now covers four pixels. With four pixels you can collect four times as many photo-electrons for each square, meaning that the S/N (and essentially DR) improves by a factor of two per chess square, or per solid angle.

I wonder if your simple example isn't mixing a metaphor (or in this case, mixing theory with practical reality). I understand the theory of 'adding pixels' with photon signal being additive while noise combines in quadrature, so the four pixels have twice the S/N over that spatial resolution. But doesn't that assume invariant illumination across the original pixel or the four 'added' pixels? In your example, the photons from that one chess square are spread over the area of four pixels (which is why the effective aperture with a 2x TC is 2 stops narrower). We're not creating a real superpixel, merely spreading out the light with diminished signal at each pixel, while each pixel still has the same read noise. So although the theory predicts double the S/N, that assumes 4 times as many photons input and twice the noise - and in your example, there aren't four times as many photons to go around. Or am I missing something?

I cannot explain why DxOMark does not find a significant difference in DR between 5D2 and 7D. I didn't find a description of how they measure DR in detail, so I don't know what their numbers mean. It also stands in stark contrast to what is found on Clarkvision, which is much closer to my expectation. Perhaps DxOMark doesn't measure DR per pixel, but per sensor area or something. That would make sense. That a large 5D2 pixel would have the same DR as a small 7D pixel definitely does not make sense, so they must mean something different.

I don't see any major discrepancy between DxOMark and Clarkvision, provided you carefully interpret Roger Clark's excellent and informative site. If you look at Figure 4, the 5DII has a 14.7-stop DR, and the 7D has a 13.2-stop DR - a 1.5-stop DR advantage for the 5DII. That's fine, but Figure 4 is based on Table 2, the sensor specifications - i.e., it's modeled/calculated data, not real, empirical data. Some of my colleagues have a word for that sort of data, and while that's a bit harsh (even though I won't repeat it here), there's some truth to the idea. Skip down to Figure 5b, which is measured DR for a few cameras. The 7D is not on the plot, but it's cousin the 50D is (the theoretical data in Figure 4 place it at ~13.4-stops of DR, close to the 7D and still a 1.3-stop advantage for the 5DII). From the empirical data, you can see that the 5DII's measured DR at ISO 100 is ~11.3-stops, and the 50D's measured DR is ~11 stops. The stated explanation for the differences between Figs 4 and 5b, "The dynamic range is often limited by the A/D converter and other electronics in the system, illustrated by the measured data falling below the model at lower ISOs," is not really satisfactory, since both cameras are falling below 12-bits, yet they both have 14-bit A/D converters. Looking at Figure 5b, you can appreciate at low ISOs, there's very little difference between the 5DII and the 50D in actual, tested DR.

Furthermore, it's worth noting that the measured DR's in Figure 5b are much more in line with DxOMark's data - which is consistent with the fact that DxO is empirically testing the cameras and lenses (although it would be nice if they provided some details about the testing procedures!).

 

[aldvan]

I wonder if your simple example isn't mixing a metaphor (or in this case, mixing theory with practical reality). I understand the theory of 'adding pixels' with photon signal being additive while noise combines in quadrature, so the four pixels have twice the S/N over that spatial resolution. But doesn't that assume invariant illumination across the original pixel or the four 'added' pixels? In your example, the photons from that one chess square are spread over the area of four pixels (which is why the effective aperture with a 2x TC is 2 stops narrower). We're not creating a real superpixel, merely spreading out the light with diminished signal at each pixel, while each pixel still has the same read noise. So although the theory predicts double the S/N, that assumes 4 times as many photons input and twice the noise - and in your example, there aren't four times as many photons to go around. Or am I missing something?

If I understand correctly the meaning of the debate - I'm not native English speakers, as you might guess - I believe that no one has yet introduced the basic concept of amount of information. In the example of the checkboard, getting sixteen pixels instead of four per square provides four times the information (I'm proportionally increasing the number of pixels in the example to be clearer). If at the four corners of a single checkboard square there are four grains of rice, getting sixteen pixels they will be visible, with four will see only a white spot. S/N and DR are important for image quality, but in my opinion the amount of information recorded comes before any other parameter. Forgive me for my poor English and if I'm off topic ...

 

[epsiloneri]

But doesn't that assume invariant illumination across the original pixel or the four 'added' pixels? In your example, the photons from that one chess square are spread over the area of four pixels (which is why the effective aperture with a 2x TC is 2 stops narrower). We're not creating a real superpixel, merely spreading out the light with diminished signal at each pixel, while each pixel still has the same read noise. So although the theory predicts double the S/N, that assumes 4 times as many photons input and twice the noise - and in your example, there aren't four times as many photons to go around. Or am I missing something?

You're right that the number of photons recorded per second does not change, so if you expose for a fixed time the S/N doesn't change (indeed gets worse because of the additional read-out noise). But since you are spreading the light from a chess square over four times as many pixels you can integrate four times longer and collect four times as many photons per chess square in one exposure. So I'm not talking about DR / solid angle / time; I'm talking about DR / solid angle for a good exposure (i.e. maximum exposure time without saturation), which in this case means exposing for 4 times longer with the TC.

That's fine, but Figure 4 is based on Table 2, the sensor specifications - i.e., it's modeled/calculated data, not real, empirical data.

Ok, so there are apparently some issues with the A/D converters at low ISO that fail to make use of the full potential (at least until ISO 2000 or so). That's good, because it means there is still room for improvement in future generation detectors! So not much help for the DR, but I expect there to still be a S/N advantage for 5D2 pixels and this seems confirmed by the DxO data, which show a factor of ~2 better S/N for the 5D2 pixels (3 dB at 18% of full well).

If we accept that the DR per pixel is the same for 7D and 5D2, then there would not be any DR advantage for 5D2+TC1.4x compared to 7D, because the 7D would still have more pixels per solid angle compared to the 5D2+TC1.4x combination (more precisely, ((6.39/1.4)/4.16)^2 = 20% as many, not really significant).

The S/N per solid angle on the other hand, should still be better in the 5D2+TC1.4x case, because 20% more pixels per solid angle is not enough to outweigh the factor of 2 S/N advantage.

 

[epsiloneri]

If I understand correctly the meaning of the debate - I'm not native English speakers, as you might guess - I believe that no one has yet introduced the basic concept of amount of information. In the example of the checkboard, getting sixteen pixels instead of four per square provides four times the information (I'm proportionally increasing the number of pixels in the example to be clearer). If at the four corners of a single checkboard square there are four grains of rice, getting sixteen pixels they will be visible, with four will see only a white spot. S/N and DR are important for image quality, but in my opinion the amount of information recorded comes before any other parameter. Forgive me for my poor English and if I'm off topic ...

Many of us (including me) are not native English speakers, so no need to apologise. What you write is true but not relevant to the present discussion (as I interpret it). The question I am interested in is if there can be a dynamic range advantage in using a 5D2 with a 1.4x teleconverter as opposed to a 7D, since they both give angular scales per pixel that are roughly the same (the 5D2+TC1.4x pixels will only be 10% larger per linear angle). The resolution (pixels per angle) should thus be comparable in both cases, but the actual sensor surface covered is larger in the 5D2+1.4x case. Since the number of photo-electrons (electrons generated by incoming photons) a detector can hold is proportional to the area, I reasoned that the 5D2+TC1.4x combo should have a S/N advantage at the same resolution as 7D, and thereby in some ways produce a better image (but not in general, since a TC introduces additional optics that may distort the image).

What I wanted to show with the chess board example is why a TC improves the dynamic range / solid angle for a given detector. This "improved" DR is then compared to another detector of similar angular resolution, so no need to take resolution differences into account.

As it turns out, the 7D has better dynamic range per sensor area than the 5D2 because of limitations in the A/D electronics that make them comparable on a per pixel level (where the 5D2 pixels are larger). So the 5D2+TC1.4x combo does not improve the DR over 7D (at least not until ISO~2000). The S/N per solid angle, however, should improve (at the cost of 1.4^2 = 2 times longer exposures), which still makes this an interesting comparison (I think).

 

[epsiloneri]

I tested the 7D vs 5D2+TC1.4x combination on the moon tonight (on a 400/2.8 lens). I exposed the best I could for the two settings, and focused with live view. Aperture was wide open (f/2.8 and f/4.0 with the TC). The moon was only 10 degrees above the horizon, so atmospheric turbulence was visible and limited the resolution somewhat, so this was not the best test. Here is the result anyway, with two 1-to-1 pixel scale crops lightly processed by DPP (default settings, i.e. sharpening set to 3), and a full moon more strongly processed and re-sampled (just for orientation). I find the 5D2+TC1.4x combination has a slight edge, but the difference is small and circumstances could play a role (e.g., instantaneous turbulence, precise focus). Looking carefully, the noise seems slightly better as well, but again the difference is small, and could just as well be due to camera sample variations.

 

[aldvan]

To my eyes your example shows a clear superiority of the 5D on the 7D, no matter what the parameter makes it better. It is interesting that it was exactly Moon photography the reason to purchase a 7D. But after some tests I realized that I had not found the right solution and that's why I bought a Meade LX200 ...

 

[jrista]

I do not have 5D II images to compare, however I believe that either you encountered some atmospheric distortion or camera shake blurring in the 7D image (something that is easy to do with the 7D, and not quite as easy to do with the 5D II, due to the much smaller pixel size). I've taken many moon images myself with the 7D, and it produces astonishing results. You do need to keep the camera very stable, and it is best if you shoot when the moon is high in the sky. You also need to keep it center of the lens for best results. On a per-pixel detail basis with the exact same lens, the 7D will always win vs. the 5D II, simply because it puts more pixels on subject than the 5D II can. Even vs. a lens with a 1.4x TC on the 5D II but not the 7D, the 7D should still have a slight advantage (about 5%), as on a pixel-size basis the 7D has a 45% (1.45x) reach advantage.

http://500px.com/photo/13321795 (B&W)
http://500px.com/photo/13321799 (Color)

http://500px.com/photo/13008253 (B&W)
http://500px.com/photo/12964733 (Color)

http://500px.com/photo/13008271 (B&W)
http://500px.com/photo/13008277 (Color)

 

[epsiloneri]

Yes, I agree that the relatively softer image from 7D could very well be due to atmospheric distortion, since they were not simultaneous and due to the low altitude of the moon. This was in the summer, however, and from my location the moon did not rise much higher that night (I live on 59 deg northern latitude). If I find the time and a clear night I will repeat the experiment this winter with the moon much higher up (and a 5D3).

 

[jrista]

Yes, I agree that the relatively softer image from 7D could very well be due to atmospheric distortion, since they were not simultaneous and due to the low altitude of the moon. This was in the summer, however, and from my location the moon did not rise much higher that night (I live on 59 deg northern latitude). If I find the time and a clear night I will repeat the experiment this winter with the moon much higher up (and a 5D3).

Thats a nice northerly latitude you live at there. While it may not be the best for the moon, I bet you get some amazing auroras. I've never even seen an aurora in person myself...can't wait to get my butt up to Alaska sometime to photograph at night in the winter...preferably shortly after a partially earth-directed X-type Solar Flare. ;-)

 

[RLPhoto]

The Ultimate Duo would be the 1Dx + 5D3. I just can't afford that kind of combo.

 

[mrsfotografie]

For me: The 7D (sports/nature/AF speed) and 5DII (general photography) fits the bill perfectly. This combination definitely beats having a 5DIII (or 7D) only, and the 7D compensates for the 5DII shortcomings enough that I don't feel the need to upgrade my MkII.

 

[bdunbar79]

The ultimate I think is yet to come. Suppose we get a 7D Mark II with APS-C, then you will see a lot of 1DX owners happily purchasing the 7D2 as a second body. I do like the 1D4 though; the IQ is good enough that you still effectively get more reach than some of the older 1.6x crop factor cameras.

If you have the lenses, my choice was the 1DX/5D3 combo. It's expensive but I really enjoy the clean performance of these cameras where I shoot a lot: high ISO.

The most common best duo: Obviously the 5D Mark II/7D combo currently.

 

[emag]

PS: I always believe that there's a reason why Canon make these two models use the same battery With the exception of the 1D/D's, I don't think there's a shared battery for two seperate lines of product.

60D also uses the same battery, and the 20D-50D used the same battery that the 300D and 5D (mark 1) used as well...

Sorry, my mistake. I should have checked the batteries for those cameras.
Didn't realize that the older XXD series, 5D and other models use the BP-511 series battery.

Canon used the BP-511 in a number of cameras. In my own case, I had at one time a G2, 300D, 40D and DV20 all using the same batteries! I still have and use the G2 and 40D, gave the 300D to my brother years ago and the DV20 is destined for trade in with CLS.

 

[Chuck Alaimo]

First off, wow, love all the technical data, but, much of it is a bit over my head. I read and do my best to understand though.

As to the perfect combo, as with everything it really depends on what your shooting. Right now I'm riding the 5d3/7d combo, but find that the 7d more often than not just sits to the sidelines. I am mostly shooting weddings, portraits and events though. Lots of lighting conditions, from well lit to fairly unlit - and more times than not in tight spaces. Even shooting concert from the soundboard I have found that the mk3 with 70-200 and no TC is fine (especially because on the mk3 I have much more room to play with ISO, which means I'm not as limited in DOF options, where with the 7d I'd want to stay as close to 2.8 as possible. But, given that I am usually shooting from right in front of the stage (i get soundboard shoots like 2-3 times a year), investing in a better way to cover that range just doesn't make sense (also, I'm not shooting sports all too often).

For my purposes, the best combo for me would actually be the 5d3/5d3 combo. But I'd also like to pick up a 16-35 this spring too. I have the 10-22, and I have been enjoying using the 7d-10-22 partnered with the 5d3-70-200. At a wedding that gives me close intimate shots, and Wider fun shots. But yeah, I am pushing the ISO on the 7d higher than I'd like in that situation.

Other options are snagging a 5d2, or, possibly a 6d when we find out more about how that body works (I kind of like this option mostly because if I went this direction I could potentially keep the 7d, if i went mk3 I would have to sell the 7d (unless I can book more than I am expecting to book this fall). I have also been giving a good look at the 1d4 too.

All that said, my perfect body combo is probably not the same as everyones. It all depends on what you shoot. If I were offered a big contract to shoot sports this year, yeah, I'd be re-evaluating my setup big time. But as it stands for me, the advantages of FF outweigh 1.6 crop.