20D vs. 5D for long lens photography

I am wondering about the claims that the EOS 20D provides telephoto imagery of

a quality as good as the 5D, at comparable image size. Since the 20D uses the

1.6 conversation factor for its smaller sensor, zoom lenses are effectively

longer. So my 400mm lens on the 20D equates to a 640mm lens on a 5D. Or, more

to the point, my 20D shot with that 400mm lens is equivalent to the central 60%

or so of a 5D shot with the same lens. According to that math, the 20D ought to

be superior to the 5D for this type of photography, given a choice between the

two. Do your real-world experiments support this? And what about the 5D's

broader ISO range? Does that make the 5D a better option? Thanks for any

opinions based on your actual experience.

Funny, Patrick, but I posted a version of the same question 10 minutes after yours:

 

<a href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00JjuE&tag=">

Image Quality Question: 5D + 1.4x + 500/f4 v. 20D + 500/f4 ?</a>

I'm not sure you understand what the 'crop factor' really does to your images. All that happens is that the effective 'field of view' of your image is smaller. A 400mm telephoto is still a 400mm telephoto...you just see less of the entire image with the 20D. In other words...not more magnification...just less 'usable' image to the sensor.

I have both a 30D and a 5D....the FF sensor is better in my opinion....

Here's one of many links on this confusing subject...

http://www.earthboundlight.com/phototips/digital-crop-factor.html?search=crop+factor&bool=and

 

Hope that helps...

If you are thinking of either getting a crop sensor body with a 400mm lens OR a full frame

body with the 400mm lens, and then comparing the uncropped image from the 20D with the

cropped image (to give the same coverage) from the 5D, the latter will have a lower

megapixel count because of lower sensor density on the 12MP full frame sensor.

 

If you were to compare a shot taken with a 400mm lens on the 20D to a shot taken with a

mythological 640mm lens on the 5D, the 5D image should be "better."

If you are serious sports shooter, you make use of teleconverters anyway, so the 20D with a 400 mm lens and the 70-200L is a standard set up--and significantly cheaper than the 5D with a 600 mm. The 5D and the 20D have the essentially the same ISO range--you are unlikely to use the 50 ISO setting on the 5D with an enormous lens. The only thing to think about in comparing the 20D and the 5D in terms of long-lens photography is the number of shots in the buffer when you shoot with the motordrive. The 20D only holds, what, six RAW images in the buffer? The 30D holds 11, which is a significant improvement. Or you can shoot JPEGs.

 

Your question is about which is superior or better. Can you clarify?

<i>According to that math, the 20D ought to be superior to the 5D for this type of photography,</i><p>

 

WHAT type of photography? you haven't said what you plan to shoot <p>

You might want to look at this article:

 

http://www.bobatkins.com/photography/digital/full_frame_vs_aps-c.html

 

It specifically addresses the 20D vs 5D issue.

 

BTW the expended ISO range of the 5D only adds ISO 50 and at ISO 50 you get reduced dynamic range, so both cameras effectively shoot at ISO 100-3200. I acn't think of any circumstances when you'd want ISO 50 shooting with a long telephoto lens.

 

Bottom line is the 20D will give you better results when you re limited by the length of the lens you have. The 20D will be better thana cropped 5D, and if you add a 1.4x to the lens on the 5D, you could do the same on the 20D and get an even longer (effective) lens.

 

If you're not limited by focal length, then the 5D will be better.

I'm sorry to add confusion to a subject that already lacks clarity. So to explain: Jack, I do understand what you are spelling out. What's important about that is the fact that in all DSLRs, regardless of the sensor, the format is WYSIWIG. What I'm after is an assessment of final image quality. Dan, the 5D/640mm image would be about 50% better, but that fact sort of sidesteps the question a bit. Preston, to streamline my question, I want to know which sensor produces a better image: the 20D with let's say the 400mmL, or the central 60% of the 5D with the same lens. Which final image offers best quality, in terms of sharpness and color and density, etc.? I have seen competing claims about this and although I know that the math favors the 20D, I am interested in real-world experiences. I would probably be better off just reading Adam's forum post :-) But I thank all for their submissions. I am mainly interested in wildlife photography such as bird shots, and also, I was making the assumption that the 5D has a broader ISO range at both ends of the spectrum, which evidently is wrong...

20D and 5D are the same sensor. The 5D is just larger.

Preston, I was under the impression that the 20D has a different sensor than the 5D. I thought it had a higher pixel density. If they actually are the same sensor then my question is answered: image quality should be identical for a 20D and the central 60% of a 5D.

See this for comparison:

 

http://www.dpreview.com/reviews/canoneos5d/

 

The density issue arises because both cameras produce images that will be printed at 35 mm dimensions, regardless of print size. So for an 8x12 inch full-frame print, the 5D offers more pixels per inch than the 20D.

Regarding 20D versus 5D pixel density from Bob's excellent article linked above:

 

"Given the budget, I'd own both a 5D and a 20D and I'd use each in the situation for which it was best suited. Since I do a lot of nature work, the higher pixel density of the EOS 20D is useful for me."

 

It's my understanding that Bob is correct and the 20D does have higher pixel density but I must admit I haven't gone in there and measured 'em myself. ;-)

<I>So for an 8x12 inch full-frame print, the 5D offers more pixels per inch than the 20D</i><P>

 

Well therein lies the rub! You can't make an "8X12 inch full-frame print" from a 20D because it's not a full-frame camera.<P>

 

Yes, the 5D unquestionably has *more* pixels but they are not packed as closely together as they are in the 20D.

Didn't mean to misunderstand...it just worked out that way...

The 20D sensor's physical dimensions are proportional to that of the 5D; the 20D is just smaller. But the ratios are the same as a 35 mm piece of film (ratio of 1.5). If you print a 20D image on an 12x8 piece of paper, or a 5D image on a 12x8 piece of paper, there will no crop and nothing left over with either camera.

 

"Full frame" is really a misnomer. The 20D creates frames that can be printed fully. They are identical to those of the 5D except that the 5D has more pixels.

5D sensor:

35.8 x 23.9 mm (35.8 / 23.9 = ratio of 1.4979)

 

20D sensor:

22.5 x 15.0 mm (22.5 / 15 = ratio of 1.5)

 

These are standard 35 mm "full frame" ratios that will print without crops on 4x6, 8x12, 16x24, etc.

The last few posts do not sound correct.

 

If memory serves, the 5D sensor is physically different from the 20D sensor. The 5D sensor is an enlarged version of the 1D-II sensor material.

 

You can so some simple math on the MP vs surface area of the sensor to verify this.

When I look at the specs on the dpreview site for the 5D and 20D, both sensors are listed as CMOS with DIGIC II processors, with an aspect ratio of 3:2, and with differing total pixel counts (5D is larger).

The pixel density of the 20D is higher as you can easily calculate:

 

5D: 12000000/(35.8*23.9) = 14025 pixels per mm^2

 

20D: 8000000/(22.5*15.0) = 23703 pixels per mm^2

<I>"Full frame" is really a misnomer.</i><P>

 

I would respectfully disagree. "Full Frame" means the imaging sensor is the same size as a piece of 35mm film. I'm not talking about aspect ratios - that's a completely different ball of wax and in that regard, both are the same.<P>

 

A smaller sensor size changes things quite a bit. You'll get more DOF out of a smaller sensor. You also have to deal with the "crop factor" when buying lenses. A "20mm" lens has the field of view of a 32mm lens on a 1.6x DSLR. It's got a 20mm FOV on either a 5D or 1Ds Mark II.<P>

 

So again, IMHO, "Full Frame" is far from a misnomer.<P>

 

And yes, I would agree that the pixel density of the 20D is higher than that of the 5D. Therefore, if one is looking for more "reach" from a long telephoto lens, a 20D will probably be a better choice than a 5D. I haven't done the math but it may even beat a 1Ds II in terms of pixel density. Best wishes . . .

Maybe this is old math but the 20D and the 30D both have 1.6 sensors. That's 1.6 X that sensor size is going to be the same size as 35mm film. 1.6 X 8mp = 12.8mp and that would be the same ballpark pixel density? , Bill

Pixel density is one issue, but there is also a tradeoff between lens resolution. The 20D has higher pixel density and thefore more detail in theory than a cropped 5D. But each of those pixels has to be enlarged more to achieve the same size print. If using the same lens, then the image that has to be enlarged more will be degraded due to lens imperfections more than the image that has to be enlarged less. So pixel density is not the only issue.

 

Otherwise Canon could point a teeny 10 MP P&S sensor in a DSLR to achive an effective focal length of say 5000 mmm (with a 500 f4 lens), Unfortunately the lens would not have the resolving power to withstand having to enlarge the image so much more than a larger sensor.

 

I think Bob's article explains this.

<p><i>Maybe this is old math but the 20D and the 30D both have 1.6 sensors. That's 1.6 X that sensor size is going to be the same size as 35mm film. 1.6 X 8mp = 12.8mp and that would be the same ballpark pixel density? , Bill</i></p>That's completely wrong. 1.6x refers to the Field of View; it is not a megapixel factor. You use the 1.6 multiplier to determine the closest equivalent focal length, if the sensor were full frame. In order for the sensor of the 5D to have the same pixel denisty as the 20D or 30D, it would need to be a 21 Megapixel sensor. Therefore, if you crop an image from a 5D to get a similar FoV crop that you would get due to the inherent design of the 20D or 30D, you would actually end up with a photo that was lower resolution than if you were to have taken it with a 20D or 30D to begin with.

"That's 1.6 X that sensor size is going to be the same size as 35mm film. 1.6 X 8mp = 12.8mp and that would be the same ballpark pixel density?"

 

NO! That math is incorrect. The 1.6 relates to the vertical and horizontal dimensions. NOT the area of the chip. The area of the 5D is 1.6 x 1.6= 2.56 times that of the 30D. This is what is messing up your math.

 

The 1D-II and the 5D sensors are the same, just the sensor sizes are different. The 1D-II came out earlier. . .which is part of the reason that it is smaller (more chips per given area of wafer). . .

 

The other reason the 1D-II chip is smaller is a more practical one: The 1D-II has a substantially higher frame rate requirement. More MP means you have that much more data to move . . .

 

The 5D was made possible because of the lower frame rate. . .and (probably) a reduced cost of production (either Canon got better with the production line efficiency, or the production line could be expanded with minimal cost. . whatever. . .doesn't matter).

 

Nominally, you would expect the 5D and 1D-II to have dramatically improved performance relative to the 20D/30D because of the reduced pixel density. The truth is, however, is that the 20D/30D chips are *newer* and have almost equivelent performance *despite* the higher pixel density.

 

Much the same is true for the XTi. Despite a 20% higher pixel density. . the chip is more modern and therefore produces similar quality to the older, lower density chips.

 

Such is technology. I have a 10 year old PC that came with a hard drive with a smaller capacity than my $30 USB key.

"20D and 5D are the same sensor. The 5D is just larger."

 

No.

 

It is not "just larger."

 

The pixel density is lower as well which, at least in theory, should allow for a lower noise

level. In addition, there are positive effects relative to circle of confusion/DOF when using

a sensor with lower sensor density.

 

To shorten this now quite long thread. Are you going to make really HUGE enlargements of

your photographs? If not, you can probably get very high quality images from the crop

sensor body and you will have more "reach" from shorter and lighter (and less expensive)

lenses.

 

If you are planning to make really big prints of the highest quality it _might_ be worth the

expense to get the full-frame body and the necessary longer lenses.