On the matter of a FF Nikon, I tend to think it will happen. I base this on the feeling that Nikon is very proud of its optics and would not like to put out a camera system where they, the optics, are found to be wanting.

However, there is an undeniable loss of market share because of this (probable) high-minded attitude and it will have to be addressed. How to do that without blowing the reputation of the optics? I think the answer will be in a firm, clear declaration from Nikon that getting results on digital as good as on film, at the extreme wide-angle end of the spectrum, is simply not possible. This will be explained as an unavoidable result of the physical shape of any sensor know to date.

As none of the other, competing manufacturers have been able to produce outstanding wide-angle results on FF either, there will be no shame in declaring that a FF body is being introduced in order to allow Nikon photographers to make use of the FF advantages within the focal lengths where it functions well.

No loss of face and huge sighs of relief.

Just an idea, but I think it is the future; are we all that enslaved to the super-wide? And if so, why?

Ciao - Rob C

Good question! Perhaps it's because the super-wide gives us a perspective in one glance that we cannot achieve by looking at the real scene, without turning our head from side to side.

There are certain things the eye can do that the camera (in a single shot) cannot.

When the eye gaze moves from a bright part of a scene to a dark part of the same scene, there's an almost instantaneous enlargement of the pupil to take in the detail in the dark part of the scene. The camera has a fixed aperture. To mimic this effect of the eye, we have to take multiple shots at different exposures and digitally blend the images.

When viewing a really wide-angle scene, the eye cannot take in the full width of the panorama, in full detail, without some head turning. We get an experience of the vast width, without turning our head, but detail is lost (significantly) in our peripheral vision.

Dynamic range limitations of DSLRs fall short of what the eye can see clearly. Wide-angle lenses do not.

When trekking recently in Nepal, I started out with the Canon 24-105mm lens attached to my 5D. This seemed to be the most versatile combination.

However, the landscape scenes in Nepal are sometimes so vast and wide that 24mm on FF 35mm is not wide enough.

Fortunately, I was carrying the Sigma 15-30mm zoom, so when the occasion arose, I would change lenses. However, having changed lenses for a particular scene, it did not make sense to change back to the 24-105mm untill the occasion required it.

Here's a problem. When trekking in Nepal, one can never predict what's round the bend.

I'll use a few shots here to illustrate my point.

The first is a scene of a suspension bridge. I changed from my 24-105 to the Sigma 15-30 to take this shot. I'm quite pleased with it. I used f16 to get maximum DoF without serious loss of resolution. I guessed the hyperfocal distance and my guess appears to be fairly accurate. I took only one shot, I was that confident.

Click to view attachment

Let's examine this shot. It's a Sigma lens on a FF camera. How are the corners?

Not too bad! There's some resolution loss in the extreme corners, but really not a major issue. Below is a 100% crop of the bottom left corner.

Click to view attachment

What about the distant horizon, I hear you claim?

It's also quite good. And to think, I did all this without reference to DoF charts.

Here's a 100% crop of the distant horizon. If you think this crop is better than the whole scene, I will understand.

Click to view attachment

However, having an ultra-wide lens on a FF camera can be a disadvantage. What happens if 5 minutes later a photographic opportunity occurs which requires a longer focal length?

Basically, you're stuffed. This was a major problem in Nepal, so I fixed the problem by carrying 2 cameras around my neck, the 20D with 24-105 lens and the 5D with 15-30 lens.

As a matter of interest, shortly after I took the shot of the bridge, I came across the following scene, still having the 15-30mm attached to my 5D.

I remember the internal decisions I went through. If I change lenses, I might miss the shot. This guy is running at a fairly fast pace.

I decided to stick with my inappropriate lens and wait till the horseman reached the right distance. The trade-off is, I only got one shot. And here it is, at 30mm, the maximum focal length of my lens.

Click to view attachment

For those of you who are very concerned about composition, you might have noticed in the horeseman shot there's a boring, blank area of sky in the top left corner.

Not so! There are interesting power lines there.

Click to view attachment

Oops! I forgot to mention. The first shot of the brdge was at f16 and 15mm. The second shot of the horseman was at f30 and f8.

Perhaps there's another consideration here. Now that photoshop CS3 has provided a superb 'auto stitch' feature which certainly seems to handle 2, 3 or 4 images with great aplomb, perhaps Nikon users can simply stitch images to get the really wide-angle effect.

On the other hand, if one adopts the same procedure with a FF camera, one gets an even wider effect.

Ultimately, it's pixel size and pixel quality that counts, and there full frame has the advantage.

I'm afraid BJL is in a state of denial.

Count me in as well, I have never seen a 5D shot topping my d2x at low ISO in terms of detail.

Cheers,
Bernard

There is such a huge number of factors that influence perceived image quality - many whose impact would vary according to the kind of image and its final output format, that it would take quite an ingenious amount of well-directed testing to determine exactly the right circumstances in which "FF" necessarily delivers superior perceived image quality relative to the D2X format - and I say that objectively as a Canon 1Ds user. The THEORY of the full frame - larger pixels, less noise etc. makes sense on the surface of it, but there is clearly much more under the hood than that. One thing though that is PROBABLY beyond dispute - the larger the sensor the more pixels of a given size can be crammed into it, so FF has the on-going potential of delivering more information than smaller sensors, but from that point onward one gets into the murky waters of quality in respect of what usage and viewing conditions.

Ah! But what about at high ISO, eh!?

They are all yours. Less than 3.141592654% of my images are shot above ISO100.

Cheers,
Bernard

Ray, what facts have I denied according to you?

I have never denied that, all things being equal, larger photosites have advantages such as greater dynamic range, so that larger sensors can offer some combination of higher DR and greater total image detail ("lines per picture height") --- in exchange for increased sensor costs and such.

Nor have I denied that at equal ISO, large pixels will likely give higher S/N ratios when used with equal exposure time and aperture ratio --- in exchange for needing a lens of greater focal length and larger aperture size, meaning a lens of greater size and weight that gives (possibly disadvantageous) lower DOF.

If I denied all that, I would probably be using a format like 1/1.8" or smaller instead of far larger, bulkier, 4/3" format gear.

These are essentially same differences as seen between different film formats, but digital technology makes the differences easier to quantify and measure.

Some people (naming no names) seem to be in state of denial about the possible disadvantages and costs associated with achieving those larger format advantages, such as greater size, weight and price, and seem to ignore the balancing of pros and cons that go into format choices, both by camera users and camera makers. (For example, Canon, Nikon and Olympus have never pursued the ultimate in IQ by making cameras in formats larger than 24x36mm).

Good! Perhaps I misunderstood you.

If you had a 5D (and especially a 1D3), you might increase that percentage.

I also have very few Canon D60 images taken at ISO 800. I was even reluctant to use ISO 400 with that camera.

Ray, I would like to know what focal lengths and apertures you are commonly using at high ISO speeds with the 5D.

I ask because I can see the speed advantage of using a larger format and larger focal lengths to achieve good IQ at the combination of high ISO with low f-stops, but see no point if one ends using long, slow lenses, getting combinations like 400mm, f/5.6, ISO 3200. Because smaller formats and pixels can offer about equally good options like about 300mm, f/4, ISO 1600 or 200mm, f/2.8, ISO 800.

Likewise for any combination of ISO 200 or faster with f/4 or slower, compared to the alternative of half the ISO speed and twice the lens speed (f-stop), using with a format and pixels about 1.4x smaller and a lens that is about 1.4x shorter and one stop faster.

In general, I see little or no high ISO IQ advantage in using a larger format (larger pixels and longer focal lengths) in situations that then involve using telephoto lenses slower than f/2.8, and even more so, slower than f/4, which leads to declining AF performance.

BJL,
I haven't done an analysis, but my most used lenses with the 5D would be the Sigma 15-30 and Canon 24-105. Many of the shots I took in Italy where I was using the 20D required a high ISO because of the restrictions on use of tripod and flash in museums and churches.

It's always an advantage to have a faster lens if you can afford it, but generally lenses for FF 35mm are just as fast as lenses for smaller formats such as the Olympus 4/3rds, are they not?

If we're talking about different size sensors with roughly equal pixel count, such as the D2X and 5D, then it's true you can use a shorter focal length, a wider aperture and a lower ISO to get a similar picture quality provided you have a lens with a wider aperture.

I doubt that I've taken any shots at 400mm, f5.6 and ISO 3200. On the 5D, picture quality at ISO 3200 is apparently no different to ISO 1600 underexposed one stop, and my 100-400 IS zoom is noticeably soft at f5.6. F8, f11 are what I use most with this lens.

Of course I'm aware of the telephoto advantages of the cropped format and I still use my 100-400 zoom with the 20D because of the 20D's higher pixel density. However, the day I get a FF Canon 22mp camera will be the day my 20D becomes fully redundant.

The wide end is certainly where I would expect larger sensors (and the related ability to use lenses of larger entrance pupils that gather light faster) to be beneficial. Then again, I have found ISO 800 adequate for hand-held church and museum photos even without IS, so I would think that with systems that offer stabilization of wide angle lenses through in-body IS, even ISO 400 or less should be plenty. Which is to say, every DSLR system except the two that headline this thread!.



No, they are not for equal FOV. As you should well know, once one allows for the longer focal lengths needed for equal telephoto reach (equal angular resolution) when using "FF" sensors with their larger pixels, the available minimum f-stops are in general higher. (And in the cases where equally low minimum f-stop is available in a suitably longer telephoto, the lens is far heavier and more expensive.)

You cannot claim the better performance at equally high ISO of larger pixels and sensors while ignoring the longer focal length needs that go with it.

Specifically, Canon's current 35mm format lens system does not offer equally low minimum f-stops at the focal lengths needed to get the same FOV, or the same angular resolution after possible cropping (i.e. cropping to roughly equal pixel counts). Not only does equal uncropped FOV require focal lengths at least 50% longer than with mainstream DSLR formats and 30% longer than with Canon's own 1.3x crop models, but with cropping to equal pixel counts, the 8.2 micron pixels of the 5D require using focal lengths
- 36% greater than with a D40x, D80 or D200 ( or A-100 or K10D),
- 49% greater than with a D2X.
- 75% greater than with 10MP 4/3 format cameras
Even if you could afford the huge price step up to the 1DsMkII, the focal length increases needed are 20%, 30% and 50%.

Bearing this in mind, I see no equal f-stop Canon "FF" lens alternatives to using those smaller format DSLR bodies with lenses like
- Nikon 85/1.4, 135/2, 200/2, 300/2.8, 400/2.8, 500/4, 600/4, or zooms matching a 200-400/4, 70-200/2.8 or 80-400/4-5.6 near the long end. Or even a zoom matching the long end of a cheap 70-300/4.5-5.6 telephoto!
- Olympus 150/2, 300/2.8 or zooms lenses matching the minimum f-stops of the longer part of the zoom range of the 50-200/2.8-3.5, 35-100/2, 90-250/2.8. Or even the forthcoming low end 70-300/4-5.6.

BJL,
You seem to have forgotten we're talking about trends and the future. There seems to be a significant lag between the pixel density of the cropped formats and the same pixel density implemented in the FF formats. The 1Ds2 has the same pixel density as the D60, my first DSLR. Even if the D60 had the same sterling noise performance of the 1Ds2, which it doesn't, there would be no circumstances whatsoever in which the D60 could offer any image IQ advantage whatsoever, with any lens, over the 1Ds2.

Go outside, take picture, print picture, show picture, enjoy picture, screw the brand ...

Repeat as desired ...

Actually, we were not: you were talking about the good ISO 3200 performance of your actual 35mmFF camera, with your actual lenses, and I responded by asking and talking about the actual demonstrated advantages and disadvantages of actual cameras and sensors in different formats. In particular, I am not interested in comparisons between imagined future Canon products (28MP! clean ISO 6400!!) to the performance limitations of products from several years ago in other brands (the Nikon D2Xs with its sensor from 2004).


Indeed there is:
1. the pixel spacing of 35mm format sensors is persistently larger than in smaller format sensors at the same time, or even from several years earlier, and
2. the pixel size gap in percentage terms is increasing with time, or in other words, the pixel count gap is decreasing in percentage terms.

The 1Ds and 14/n had about twice the pixel count of the smaller formats, which were then at 6MP. In fact the 14/n has the same pixel size as the D100. Now the pixel count gap from 10MP mainstream models is only about 25% (5D) to 65% (1DsMkII), and relative to the maximum pixel count in APS-C, the D2Xs, the gap is only about 5% to 30%.

So the gap in focal lengths needed is growing, getting fairly close to the factor expected based on sensor size alone. Comparisons between lenses of equal focal length become less and less relevant, if they ever were relevant (I doubt that it is at all common for people to use larger format DSLRs with higher pixels counts and then crop down to the pixel count possible with a far less expensive smaller format camera). Using the 5D with the same focal length as gets the job done in DX or 4/3 and cropping to get the same FOV gives only about 5.5MP (DX format) or 3.4MP (4/3 format).

I can see why the arguments in favor of 35mm format digital are consolidating on wide angle to normal performance and high dynamic range at low ISO, more like the traditional advantages of larger film formats, which were never much touted for telephoto and high shutter speed situations.


P. S. In deference to djgarcia, my points are addressed to comparisons of formats; brands only appear when specific examples must be mentioned.

BJL,
Have you ever tried simplifying your arguments?

What I understand from your above post is that you are merely re-stating the traditional advantage of the smaller format, ie. the opportunity to get the same DoF plus faster shutter speed by using a wider aperture with a shorter focal length.

Comparing the D2X with the 5D with a specific example; if I'm in a low light situation and need to use the widest aperture of my lens, say f1.4, then the D2X will give me approximately the same DoF and the same general image quality at ISO 800 as the 5D at f2 and ISO 1600, using a lens 1.5x the focal length on the 5D.

However, the advantage here is not always clear-cut. It's dependent upon the possession of lenses of equivalent focal length (for the smaller format) that actually are faster, and it's also dependent on the requirement (in relation to the subject) of as much DoF as possible.

For example, if I'm photographing (without flash) a section of the ceiling of the Sistine Chapel where DoF may not be an issue, I can also use f1.4 at ISO 800 with the 5D, or use a faster shutter speed that's maybe needed because it would be a hand-held shot, in oder to get better image quality than the D2X can provide; either better in respect of less noise or sharper because I was able to use a faster shutter speed.

(This argument is based on remembered evidence that the 5D at ISO 1600 produces images that are at least as sharp and clean as the D2X at ISO 800).

There are also a couple of other points that are worth mentioning in circumstances where the same zoom lens is used with both FF and APS-C cameras. Lenses are generally not sharpest at full aperture. Even expensive primes like the Canon 85/1.2 are not too hot at f1.2. At apertures faster than f5.6, sharpest results are usually achieved by stopping down. Furthermore, when the shallowest DoF possible is required, FF is generally better (using the same zoom lens or equivalent prime which is equally fast.)

Finally, it's pretty obvious that pixel counts continue to increase on both formats. As they do, the telephoto advantage of the smaller format diminishes as the greater pixel density of the smaller format struggles to extract more detail from existing lenses. Of course, new lenses are being developed, but the quality is not increasing at the same rate as the increases in pixel count.

"Even expensive primes like the Canon 85/1.2 are not too hot at f1.2."

No but it is f/1.2! If a lens is slower its image quality may be better (?) but its still slower. Without wanting to appear rude (I really don't) I've seen this argument too often and its not valid. If you buy an ultra fast lens to use wide open then at least you CAN use it wide open because it has the fast maximum aperture - I have L primes from 24 to 135 and ALL are used wide open when I want/need to.

Secondly, its worth also pointing out that smaller than FF cameras may also be at a lens design disadvantage when it comes to fast wides simply because they still carry forward the design of 35mm film lens mounting dimensions - which are probably not the ideal requirements for lens designers and fast wides (I exclude the Leica M8 which should have a clear advantage here - but of course isn't an SLR).

Does anyone remember the Olympus Pen series of solid but diminutive half frame SLRs? I wonder.....

No intend to get involved in a Canon vs Nikon debate, but I just cannot agree with this last point. This just isn't proven.

How do you explain that Michael concluded after comparing the Leica M8 with a Canon G7 that both cameras deliver sharp results at low ISO? The sensor of the G7 has a surface that is about 10 times smaller, and there a pixel density 3 times higher.

This shows very clearly that there are lenses able to resolve at least 3 times more than current 35 mm lenses.

Regards,
Bernard

.........and that very crowded sensors with small pixels can still deliver very high quality images?

But even so, there must be a point at which bigger is better; and at that point will the lenses or the sensor be the critical constraint on resolution?

Why should the point not be valid? If there's a trade-off in image sharpness then that's a disadvantage. Are you saying you don't care about image resolution and sharpness?

My remarks were made in the context of comparing a low noise (at high ISO) FF 5D with a slightly higher-noise, cropped format D2x. With the D2X you can use a faster aperture and lower ISO to get, according to BJL, the same image quality and same DoF. My point of dispute was that you wouldn't if the sharpness of the lens at full aperture is less than it is when stopped down one stop. Seems a perfectly valid point to me, in relation to different format cameras that fit the same types of lenses.

Bernard,
You are shifting the debate to comparisons of format extremes here. It does appear to be the case that very small lenses can be made that are diffraction limited at apertures of f4 to f5.6. A lens that is diffraction limited at f4 has double the resolution (at the same MTF) of a lens that is diffraction limited at f8.

For some reason, it doesn't appear to be either possible or economical to produce 35mm lenses or APS-C lenses that are even nearly diffraction limited at f4, perhaps because there's so much glass for the light to pass through.

There are also severe limitations of such lenses in small P&S cameras. Shortest equivalent focal length of the G7 lens is 35mm. Widest 35mm equivalent apertures for DoF purposes are f11-f19. You can't get a shallower DoF than f11 (35mm equivalent) on the G7.

"Seems a perfectly valid point to me, in relation to different format cameras that fit the same types of lenses."

So what equivalent exists for a 24mm f/1.4 L lens on a 1.5x or 1.6x camera (from any manufacturer? I accept that a 50/1.2 will equate to the same fov as an 85/1.2, but shorter, fast, FF lenses are difficult to equate to on smaller sensors. Of course I care about sharpness (and bokeh, etc.) but if you haven't got a fast enough lens to shoot the image 'cause it doesn't exist, then you have to use what is available - and believe me the Canon L fast primes are pretty good wide open (I haven't lost a sale from any, lets put it that way!). And test resolution doesn't tell you everything about a lens either!

We tend to forget that cameras and lenses are tools to do a job (ie they are not an end in themselves) and if you ask me, we never had it so good! As I've said before, Nikon and Canon both produce superb cameras and lenses - both have their advantages and disadvantages, and I would class neither as better, just different. As I've also said before, if I were a student (I think that's where this thread started) I'd lean towards Canon as more commercial photographers are using them than Nikon, and to get an assistant's job this might be the better camera to be familiar with

Given how you so massively misinterpret what I have been trying to say above, perhaps I should. I suggest in particular you re-read my post #111 in this thread, in which I say in part


It should be clear that I am talking about claiming a high ISO advantage for larger formats in situations where that high ISO speed is used in conjunction with a relatively slow lens (or a lens stopped down to a slow aperture for adequate DOF), meaning apertures of about f/4 or slower. This applies to all the zoom lenses that you have mentioned using (Sigma 15-30, Canon 24-105 and 100-400) except for the slightly faster f/3.5 at the wide end of the Sigma.

In that situation, the claimed high shutter speed/low light advantage is illusory, as a somewhat smaller format can achieve very similar results at the same shutter speed with a suitable combination of shorter focal length, lower aperture ratio and lower ISO speed. For example changing from 35mm format, f/4 [or f/5.6] ISO 3200 to "APS-C" format f/2.8 [or f/4], ISO 1600.

I have already said that the larger format has a speed advantage when it can use the combination of high ISO and low f-stop. So why do you use low aperture ratios like f/1.4 vs f/2 in your attempted counter-example?

Finally, note that I am not claiming in this thread any image quality advantage for a smaller format, nor denying that larger formats have some IQ advantages; my point is only to dispute _some_ claims of IQ advantages for larger formats, based on the persistent and often false assumption of using equal f-stop and ISO speed with a larger format and greater focal length.


I am avoiding comparisons of imaginary future products, so will simply say this: regardless of how sensor and photosite technology develops, the only way to get any advantage from using a larger format is to use a larger focal length, and that requires either (1) higher f-stops, and thus either higher ISO speed or lower shutter speed, or (2) larger, heavier and probably more expensive front lens elements and less DOF. Option (3) of using equal pixel size and equal focal length and cropping is of course simply a more bulky and expensive way of reproducing what could be done with a smaller format.

Firstly, please learn to use the quote button, so we know who you are quoting and from which post.

Secondly, I wonder what you are wondering about: why very cheap 35mm film cameras, film and processing eventually displaced smaller formats that had no significant cost advantage, while persistently very expensive 35mm format DSLR's are still confined to about 0.3% of the total digital camera market, under 3% of the DSLR market, and probably under a third of the professional digital camera market? (The 1.3x crop format 1D models dominate the latter.)

Two factors seem to explain it: different price and performance trade-offs with digital than with film. With digital
(1) the cost disadvantage of 35mm format vs the dominant smaller formats like DX, EF-S, 4/3 etc. is far greater than that between 35mm film and smaller film formats,
and
(2) the image quality is far better at any given format size, so the visible IQ differences between 35mm format digital and smaller DSLR formats is far less than between 35mm film full frame and 35mm film half-frame.

My god look what you've started.

BJL,
You're obfuscating here. One uses the lenses one has or can afford to buy. If you are buying into a new system, whatever the format, then the range and type of lenses available for that format should also be a consideration.

I understand where you are coming from. You own an E1 and you probably have a lens or two (probably zoom) that is faster than a Canon zoom of equivalent focal length. If so, that clearly helps towards removing any IQ disadvantage of the smaller format. I've always thought it a pity that Olympus has decided against making their Zuiko lenses compatible with Canon cameras. A bit of vignetting in the corners of a 400D image could be tolerated for the sake of a sharper image elswhere.

Are you arguing with me or against me? You should ask BJL this question.

Who cares about extremely subtle differnces of image quality esp. compared to missing elements of a pictue due to the crop sensor.
I find this argument aboutminor pixel differnces a bit farcical to an extent becuse if you look at a lot of the great photographs of the past, most would be rejected by the image quality fascism that can occur on LL.

Unlike you I shoot at higher ISOs and recently, a Nikon using collegue struggled to get pics with his D2X, whereas I managed to get shots OK with my 5D. I had a more usable high speed ISO and a faster wideangle lens, so I had at least 2 stops advantage over him - we were shooting dancers and I was using 1/160thish at f2.8. He was considering buying a Canon after that.

Good point, JJJ. We should consider the whole system. Available lenses, maximum aperture, effective DoF (in the standard 35mm terms) as well as ultimate image quality.

There are pros and cons for all formats.

How so: I am simply stating a fact about using lenses of relatively high minimum f-stop with larger formats, compared to the option of using shorter but "brighter" lenses with smaller formats. I think the obfuscation comes form people who insist over and over again on comparing low light ability solely on the basis of comparisons at equal ISO, ignoring likely variations in f-stop choices.


I completely agree, or as you say in another post,


And if buying 35mm format limits one to doing most of one's photography with slowish lenses like a 15-30/3.5-4.5, 24-105/4, and 100-400/4-5.6, whereas with a somewhat smaller format one could afford lenses a stop or so brighter (especially with over a thousand dollars more to spend on lenses from savings on the camera body!), then the high ISO advantage of the larger sensor does not translate into any IQ advantage at equal shutter speed under equal lighting conditions when using those relatively slow lenses.

At the risk of sounding snobbish, I hope never again to buy a lens slower than f/4 at any focal length, for the sake of good AF performance and the ability to use a TC without losing AF entirely, as happens with a TC on the 100-400/4-5.6, doesn't it?

On the other hand, if I was more interested in in-door low light wide angle photography without flash, a tripod, or stabilization, I would be more interested in a larger format. But once even wide angle lenses are stabilized by in-body stabilization, I see little need for high ISO in that situation.

In response to a comment about 24mm f/1.4 lenses in 35mm formats having no equivalent in smaller formats


Ray, are you really still ignoring the fact that I completely agree with you that larger formats have an advantage when they can be used with fast, low f-stops lenses?

If I saw no advantages to larger formats over smaller ones, I would not be using a format more than four times larger in sensor area than 90% of digital cameras now sold.

We both agree that the 5D has cleaner high ISO, that the 5D is overall a very good camera and that FF has some advantages over APS for some applications. With today's technology, my view is on the other hand clearly that APS sensors have advantages for landscape.

All I am saying is that, at low ISO that I use 99% of the time, the D2x has best in class image quality and is in many ways an even better camera than the 5D.

I am not sure how trying to use what I see as the best tool around for my type of shooting gets to be image quality fascism... I understand that having different views on an issue can be felt as being disruptive, but "promoting diversity" and "facism" are concepts that are 180 degrees from each other.

For me, high ISO quality is only next to strap color in the irrelevant features list, and that it because I don't use a strap.

Cheers,
Bernard

Ah! Now I see your point. It's purely an economics one. Instead of splashing out on a 5D body, you're implying that I might have been better off buying a faster lens, such as a 70-200/2.8 IS with 2x extender, or a 300/2.8, or 24-70/2.8 or a 16-35/2.8.

As a matter of fact, when the full details of the 5D were released after much speculation as to what format it was, I was rather disappointed to learn that it was 12.8mp on a FF sensor. I was sort of hoping the camera would be a 12.8mp cropped format because I'd already bought the EF-S 10-22mm lens.

However, 12.8mp are better than 8.2mp when you have a wide format printer (as I have) and the EF-S 10-22mm, although roughly equivalent to the Sigma 15-30, is no faster and marginally less sharp.

As I've mentioned before, the availability of the right lenses can affect one's choice of camera. If the EF-S 10-22mm lens was a sharper and faster lens, say f2.8, and, if Canon had offered in addition to the 5D a 12.8mp cropped format body at a lower price than the 5D, I would have bought the cropped format. But they didn't so I couldn't.

While for me the biggest issue that pushed me into retiring my faithful 10D in favor of the FF 5D was that BIG, GLORIOUS, VISIBLE FINDER!!!

The extra megapixels were really just a bonus. I was tired of looking down a long tunnel to see the image. And, fortunately, I hadn't yet invested in any of the S-series lenses. My existing lenses all work fine on my 5D.

I have nothing slower than f2.8 and a big sensor, so not sure what your point is. Finding Fast apertures to go with smallers formats is quite hard, esp. at the w/a end.

IS has no effect on subject movement, so high ISO will still be essential.

Still pokey compared to my OM4, which is also half the size.

If you stick to the brighter tones, they might be more similar, but once you get deep into the shadows, the D2X is an order of magnitude noisier than the 5D. The D2X does not have any high-ISO readout optimization. The read noise is the same level in electrons at all ISOs on the D2X. It starts out about twice as strong as the the 5D at ISO 100, and is about 12x as strong at ISO 1600. Nikon seems to be short on ideas about how to tame read noise.

I've read comments on how Nikon has caught ip to Canon in the low-light area, comparing the D80 to the XTi, yet examining their RAW files, the D80 has 4x-5x times the read noise as the XTI at ISO 1600. Most of the samples in the comparisons are well-exposed, and overlook the fact that the D80 tends to expose very high, and the XTi very low, so you need to use -EC on the D80 and +EC on the XTi to properly compare them in the same lighting.

Deleted due to double post.

Didn't realise there was such a big difference in noise levels between the two cameras, John.

It's no wonder Bernard has taken so few high ISO shots with his D2X .

Not clear to me. As APS is restrictive in certain areas without doing anything a FF sensor cannot do.

I can take images you cannot with your camera, the converse is not true. So hardly a better camera. I've a D2x using friend who was struggling at a dingy event and was considering a 5D afterwards. As I got shots, where he couldn't. Though Nikon does have a better flash system it seems.

Wanting the best tool for the job is fine, but I was generally referencing a certain trend on here towards ever finer pixel counting over simply taking pics, without worrying about the minutae of image quality. And sometimes the technically poorer quality picture is actually better. I used to use Recording film in Acuspeed, a recipe for grain, but it gave amazing images in the right conditions [dingy!]. A graining picture that actually catches the moment is often a lot better than a less graing but very blurred image. There are different types of sharpness!

Most people use a strap and shot at higher than 100ISO, so your needs are in the minority I'd say.

Well, I can tell you that I see a very clear difference in DoF when using my Mamiya ZD compared to the D2x. If the D2x had the same image quality as the Mamiya, the Mamiya would gone in seconds no questons asked.

But OK, I am not trying to convince anyone, great for you if you have found the perfect tool for your applications.



That's where I feel comfortable, never been much of a crowd follower.

This being said, most photographers with a speciality belong to one minority or another. Out of those many specialities, there is a very significant portion where high ISO is just not needed.

Cheers,
Bernard

Well, let's just understand this in context. It is the *read* noise that is so much higher. If you're not down in the shadow areas, read noise is not all that significant. The difference is in pushability. The 5D would seem slightly cleaner in a well-exposed 1600, but in a push to ISO 25000 on both cameras, the D2X is going to be much worse.



The old electric fence phenomenon, perhaps. Or, he may just do landscapes, architecture, or studio work where high ISOs (and pushing them further) may not be necessary.

Personally, I have to shoot things that move in real light, hand-held, and must have high-ISO performance.

Just out of curiousity, could the sensor be made any larger than 36*24 in dslr's without vignetting or lens redesign?

Hi,
Some interesting info on spectral response of a couple of dslrs can be found in http://digitalkamera.image-engineering.de/...070213153138180. Especially D80 looks quite different from the majority. I believe the measurements were done from dng-files.

To be speculated,
J

I have been wading through this forum and am glad to read the many differing viewpoints from you peoples
I am also looking for my first prodidge.slr and if these questins haven't been covered yet ....I guess I'll find out soon
Budget matters, I take my camera into wild places, so weather and dust seals are important, I would also love to create fine art prints ,
including studio shots
Have heard that full frame sensor ican be a challenge to a lens, and non ffs catch the sweet spot.
plenty of you have said go for the 5D, The dpreview users consistently call it a dust trap .
2DXs seems attractive on many counts incl image quality the only down seeming to be noise at and above 400 iso.
The Canon IDs mk2 seems affordable to me yet seems more for action and sport than fine quality poster sized images
Any Comments
p.s what does it mean to be trolled ? sounds like some of you enjoy
regards Sukh

Apology I meant that the Canon ID mk2 as being affordable Not the IDs Mk2
sukh

The chart for the 400D is TOTALLY wrong. All DSLRs with RGB bayer CFAs and a hot cut mirror are least sensitive to red. ALL. This graph is totally backwards.

There's some bad reverse-engineering at image-engineering.

Hi,



What's really interesting is that the graphs are basically very similar except for the D80. M8 shows less pronounced difference between the channels. Are the sensors so different or is the signal processing to the raw file different? If they all happened to be wrong, might they still be comparable between each other?

Cheers,
J

Many do, for weight and/or cost reasons, with lenses like the 24-104/4 and 100-400/4-5.6. Like Ray, with whom I was originally discussing this. (And many situations call for stopping down to f/4, f/5.6 and beyond.)

With (fast) moving subjects at wide to short telephoto in low light, yes, I see that is relevant to some people. But not to many others like me.

As I have said several time before, I am not dismissing all the advantages of larger formats over smaller ones (or I would be using 1/2.5"!), just some exaggerated and misplaced claims of advantages.