Looking at them again, it would seem that they are using converted images, where the response is affected by white balance, because the area under the curves is basically similar, in the visible light range. The narrower peaks go higher, and the wider responses don't go as high.

These figures are not very meaningful, then, as they don't reflect the sensitivities of the cameras, but the cameras after they are compensated for their sensitivity differences!

Yes, mostly economics, but sometimes of weight as well as cost: many of us would prefer not to carry a 400/2.8 (or even maybe a 400/4 or 300/2.8) even if someone else were paying. And at long enough focal lengths, f/2.8 (or even f/4) is not available at any price. So once each of us reaches the longest focal length at which about f/4 or faster is possible, I tend to think that increasing telephoto reach (angular resolution) might most often be best done by increasing sensor resolution (reducing pixel size). Maybe by cropping from a still large sensor, maybe with a smaller format body at least for the telephoto end of things.

Both very valid factors. I was thinking a bit before your post that 12.7MP vs 8.2MP (within the Canon options at least, or within the price level of the 5D) is probably more important overall to the desirability of the 5D than the much talked about high ISO and high DR. And to be slightly cynical, Canon's possible holding back of EF-S format a bit to maintain unique selling points for 35mmFF. Will there be a 10-12MP high end EF-S model to replace the 30D once there is a higher resolution replacement of the 5D? (maybe 16.5MP, using "1DMkIII pixels").

That seems pretty obvious, doesn't it? And I bet that this 12 MP APS sensor will have high iso noise and DR very similar to those of the 5D today.

People will say that this is the result of improved sensor technology and processing.

Regards,
Bernard

I think threads like this are useful in 'fleshing out' many of the factors that have a bearing on technical image quality, the pros and cons of particular formats with differing pixel densities and differing noise performance in relation to different camera/lens combinations, but the one unknown factor which tends to throw a spanner in the works, especially with regard to making the best personal economic decisions, is what's round the corner.

Many of us either cannot afford to buy (or cannot justify the purchase of) whatever system is currently on the market that best meets our requirements for the particular type of photography we may be interested in at a particular time, so we have to make some sort of assessment as to the significance of such pros and cons in relation to the general type of photography we think we'll be engaged in.

For example, right at the moment, a good combination of cameras for a wide range of (effective) focal lengths from wide angle to telephoto, would be the 12.8mp 5D and the 10mp 400D. But how would one feel if in 6 month's time Canon released a 22mp upgrade to the 5D with ISO 6400 capability and no more noise than the 5D currently has at ISO 3200?

That one 22mp camera could do everything the other 2 could do, and some. There would be no need to mess around changing cameras with the 100-400 IS zoom in order to get a longer effective reach and possibly lose the photographic moment in the process, and the extra ISO performance would allow use of f8 in circumstances where f5.6 would previously have been used.

And of course, needless to say, at shorter focal lengths 22mp is always going to be better than 12.8mp for large prints. The difference in pixel count is great enough to be noticeable. It's unlikely we'd have the same confusion that sometimes existed when the 5D resolution was compared to that of the 1Ds2. 16mp versus 12.8 is subtle. 22mp versus 12.8mp is not, I would think.

You should probably make tha 16.7 MP vs 12.8 MP, if you're going to use decimals ...

You're right, but I can't always remember the differences between the total number of pixels and the effective number of pixels. I tend to forget things easily that I don't think are significant .

That is more like it: choosing amongst actual options. (For my more modest print size goals, a good current choice would be a 10MP body, giving me an effective built-in 1.4x TC when cropped to the 5MP that I find quite adequate for many wild-life shots.)


But if we must speculate about future products, an equalization of pixel sizes available in Canon's EF-S and EF mount offerings seems very unlikely; there has been no sign of Canon moving in that direction, or of any format from "1.3x" up matching the pixel size of smaller formats. You are imagining a jump from the 5D/30D pixel count ratio of about 1.5 and pixel spacing ratio of 1.3 to roughly equal pixel size and pixel count ratio of around 2.5. And that with the 30D sensor being the older of the two, and so probably with more room for progress.

The actual pattern in pixel sizes and counts has a curious split at "1.3x" (EOS-1 D series, Leica R back and M8).
- From digicam formats like 1/1.8" up to 1.3x, roughly the same maximum pixel count, with larger formats putting the extra real estate almost entirely into larger photosites.
- From 1.3x upwards, a reversal to almost invariant pixel size, with larger formats offering higher pixel counts.

Since the ".8" in 12.8 is significant but the ".7" in 16.7 is not, I take it you own a 5D?

I don't see how you can justify such claims, BJL.

The 1Ds followed closely on the heels of the D60. The difference in pixel density (and the significance of that) was duly noted by owners of the D60 and politely ignored by owners of the 1Ds who simply didn't want to admit that a much cheaper Canon DSLR might in some circumstances produce a sharper or more detailed image than the flagship model.

However, that discrepancy of pixel density was soon rectified in the 16.7mp successor to the 1Ds, the 1Ds MkII. Of course, in the meantime, the pixel count of one of the cropped formats has increased to 10mp and the others to 8mp.

Whilst I don't think it likely that the pixel density of the 400D will be matched in the next upgrade to either the 5D or the 1Ds2, it will not be atypical if Canon matches the pixel density of the 20D with a 22mp full frame.

The interesting question is, will Canon produce two FF models with 22mp? Perhaps the successor to the 5D will be a mere 16.7mp, in which case it's unlikely I'll be getting one.

Correct. However, from a purely objective point of view, the 0.8 in 12.8 really is more significant than the 0.7 in 16.7. We're looking at 6.25% for the 5D against 4.19% for the 1Ds2.

Completely agree. Doesn't mean I'm going to let you emasculate my hard-earned 4.19% ...

We agree of course.

Two and a half years after the D60; not so "soon".

Exactly my point: by the time the 1DsMkII arrived, 2 1/2 years after the D60, EF-S format already had the 20D with higher pixel density than the 1DsMkII. EF-S (and other sub-35mm formats) have always offered a higher maximum pixel density that 35mm offers at the same time, and I see no trend of this gap closing.

I agree that it would not be atypical for the next 1Ds model to do so. But a 22MP 1DsMkIII was not what you speculated about in your previous post, and nor is the current 8MP of the 30D what it would probably be competing with for pixel density available in high end EF-S models. The 30D still being at 8MP is an aberration given that everything around it is at 10MP and up, so 10MP or more in a high end EF-S format body seems likely soon. (Unless Canon shoots itself in its "EF-S" foot in an effort to prop up sales of entry-level 35mm format, accelerating Canon's recent decline in DSLR market share.)

Ah good; back to reality. The idea of jumping from 12.7MP in the 5D (it is not 12.8 you know!) to 22MP in a replacement only about two years later would be unprecedented, and to match a likely 10MP+ 30D successor would need 26MP, more than double the pixel count of the 5D. That would be a truly unprecedented increase in a single Canon upgrade step, or within two years.


For reference, here are some Canon models in chronological order of announcement date,with each new high of pixel density marked ***.
May 2000: D30 with 10.5 micron pixels
Sep. 2001: 1D with 11 micron pixels
Feb. 2002: D60 with 7.6 micron pixels ***
Sep. 2002: 1Ds with 9 micron pixels
Aug. 2004: 20D with 6.4 micron pixels ***
Sep. 2004: 1DsMkII with 7.2 micron pixels
Jan. 2004: 1DMkII with 8.2 micron pixels
Aug. 2005: 5D with 8.2 micron pixels
Aug. 2006: 400D with 5.7 micron pixels ***
Feb, 2007: 1DMkII with 7.2 micron pixels

In each case it takes two years or more from when a new smaller pixel spacing arrives in EF-S format before a 35mm format model matches or passes that pixel spacing, and by the time that happens, EF-S format already has a more recent sensor with smaller pixel spacing than that new 35mm model. And 1.3x lags behind 35mm. And "entry level 35mm" (5D) lags behind everything else.

Okay! Point taken. However, (there's always an however ) trends do not necessarily continue forever. The big attraction of Canon's APS-C format is that it uses existing 35mm lenses. The fact that greater pixel densities have been able to capture more detail from the central part of the image is a clear indication, up till now, that good 35mm lenses still are not the weak point in the chain. But, I would suggest that from now on, they will increasingly become so.

If one were to check dpreview records, one would probably find that the 10mp 400D is capable of close to double the resolution of the 3mp D30, using a good prime. Is it conceivable that a 40mp cropped 35mm format would be capable of double the resolution of the 10mp 400D? Not even nearly, I would suggest.

At some point, Canon will have to decide if it's going to create a separate format for its APS-C cameras with high quality and expensive EF-S lenses that don't fit FF 35mm bodies, because increasing pixel densities of the smaller format will become increasingly irrelevant with current 35mm lenses.

When we get to these levels of resolution, if we ever do, the main limiting factor is going to become the shooting environment, rather than the body-lens combo itself.

Shooting handheld will stop to be an option, and very sturdy tripods and heads will become mandatory. In the end, all the advantages of a small and light format will be lost... The bodies with few pixels but more DR/less noise will provide better images 95% of the time.

Regards,
Bernard

Bernard,
As regards resolution there can be no disadvantage in having too many pixels. It may be the case that in order to realise the greater potential for sharper images, faster shutter speeds may be required for handheld shots, or use of a tripod, but under the same shooting conditions a 40mp camera cannot produce less sharp images than a 10mp camera, assuming everything else is the same, such as sensor size, RAW capability etc.

My concern would be that any increase in resolution flowing from such large pixel counts would be too slight with current 35mm lenses to justify the inconvenience of slower speeds in processing such large files.

Dynamic range and noise issues are less clear because they are dependent upon technological advances. Also, whilst the noise and DR of individual, small pixels might invariably be less than that of individual larger pixels, for any given size print the noise and DR might be the same.

We should not forget how good the Canon G7 is. A sensor the size of Canon's 20D, with the pixel density of the G7 would be around 88mp. For such a camera, we would need EF-S lenses significantly better than BJL's Zuiko lenses .

Ray,

You are correct, but considering how many people judge cameras by looking at images 100% in PS, the 40MP will probably be perceived by many as being less sharp... even if the actual detail ends up being exactly the same or slightly better at a given print size.

Regards,
Bernard

The problem with lenses and making comparisons with them in the sense of different format sizes is that this is nothing new or confined to digital.

I remember well that when I bought into the last Pentax 67 system I had high hopes for much more impressive transparencies than I was getting with Nikon 35mm; this was, in fact, despite knowing that 35mm lenses are so sharp because only a tiny image circle is being used.

Well, I should have saved my money. Though much more impressive transparencies resulted on 6x7, this was only because of tonality in the sense of smoothness; as I've said here before, the colour of Velvia through 35mm Nikon glass was, to me, far superior to the colour of Velvia through Pentax 6x7. Worse, the things were simply never as crisp, despite always being shot on a huge Gitzo tripod. There was no way, if you chose to cut out a 36mmx24mm piece from the 6x7 that it matched the straight 35mm frame from 35mm Nikkor lenses.

This was also borne out for me doing the same thing in the darkroom with Hasselblad lenses and Nikkors.

I think that we are in danger of asking for more than can be delivered by the optical manufacturers.

Ciao - Rob C

Lens designers will have to consider more than just the possibilities of optical design - the economics and quality control of lenses needed to provide sufficient data for higher Mega Pixel chips will, I'm sure, become a limiting factor very soon.

And why do we need ever increasing MPixels? We seem to be expecting a 35mm type body to deliver an image that would have required a far larger format in the past (on film). Why? Are we now entering a similar phase to that of cars - most can exceed national speed limits and certainly existing road conditions, so why do we really need ever faster cars? Could we be in danger of falling into the same trap with digital cameras? How many users really need the 1DS MkII and really push its capabilities - and I mean in reality, actual output of images that is. I've said it before, and no doubt will again, I have never lost an image sale due to inadequacies in the image (either digital or film). This could of course be down to my abilities with inferior equipment (but no, I'm not THAT good!. Or is it because I taylor my images to meet my market (this is more like it). The same has to be true of MPixels, lenses, etc. If you really, truly need a better system (rather than just thinking that you do) then move to something bigger!

This has been a fascinating thread, but to my mind the gist is that both Canon and Nikon make exellent cameras each of which has its pros and cons. What the future will hold we can only guess, but I for one would prefer to believe that the final viewed image will still be more important than the technology behind it.

That's a very paradoxical statement, Paul. What do you mean by this? The final viewed image, which wouldn't exist without the technology behind it, is more important than the technology that made it all possible???

Lenses of similar quality to the Zuiko lenses designed for the Olympus 4/3rds system, with just a slightly bigger image circle, would be fine for the Canon 30D and 400D format. Unfortunately, EF-S lenses of that quality would result in the 400D producing images on a par with (and perhaps better than) the more expensive 5D with lesser quality 35mm lenses.

Paul

I think you are quite correct suggesting that quality control might prove to be more of a problem for manufacturers. It seems that variations of quality within the same lens type can be very wide - naturfotograf.com's proprietor has made some startling statements about the number of lenses of the same type that he has tested prior to finding one worth keeping. That he has had the opportunity to go this route is great, for him, but don't we all wish we could just ask the local friendly dealer for a box of lenses so that we could try them all and select the one that worked!

I'm told that Leica lenses go through a greater, more severe set of tests prior to release for sale - I'd have thought that existing legislation on the theme of fitness for purpose would have forced them ALL to play fair, but then again, how do you fight Goliath? Don't mention slings - ths is 2007 and the lawyers are far more powerful than that!

You have a point with cars, but really, I think that that's more just a matter of male stupidity and the need to impress other, equally dumb males. If these guys have the money and are willing to blow it, then other than the fact that they are helping us die earlier, what's the harm? With photography, I think it's somewhat different. The quest for mega pixels is more to do with the fear of reaching a point when the available/affordable printers start to outstrip the ability of the camera to keep up with them. This, again, I think is probably more of a 'digital' mindset than it ever was with film; it's quite possible that with film, and the reality of the wet darkroom, one was quite happy to accept that a smaller film would print up to whatever size with accepted characteristics and that a larger film would do so in a different manner. Within those long-established norms, there wasn't such an anxiety or fear of inadequacy, of getting left behind, unlike in this digital medium where something newer and better is always expected 'next show'.

Frankly, I have long been of the belief that film and digital photographers are powered by different motives and inclinations. I, for one, doubt if I would ever have considered a career in photography if I'd been confronted from the start with the digital world. Yes, I've had to adapt, as far as it goes, but that's not to say that I believe the medium now holds the same fascination for me as it once did. Could just be age - but I don't really think it's as simple as that.

Ciao - Rob C

Rob,
I guess it all boils down to what the final cost of the lens is going to be and what the public perceives as good value. Before finally accepting an EF-S 10-22mm zoom for my 20D, I tested a copy of that lens in 3 different countries, Australia, Singapore and Malaysia. The first lens I actually bought in Australia before thoroughly testing it; just a brief test in the shop. I later discovered it had an autofocussing problem at close distances; returned it but the store didn't have another copy of the lens available to offer me and I'd already arranged to travel overseas and couldn't wait for a replacement.

Whilst in Singapore, I tested another EF-S 10-22mm in the shop, using my laptop to view the results. The store manager could clearly see that the lens was not nearly as sharp as my Sigma 15-30 at 22mm. I didn't buy it of course.

The third lens that I tested in Kuala Lumpur seemed reasonably okay. No focussing problems and the lens was almost as sharp as my Sigma, so I took it.

Who knows, if I'd continued testing lenses from different batches in different countries, I might have eventually found one that was decidedly sharper than my Sigma. But at what cost to myself in time and trouble?

Ray -

Yes, testing many lenses in the wan hope of finding a good one is not on; also, it is almost criminal that one should know that it is the way things are but also feel powerless in the face of it all. I have no sympathy with any manufacturer who has the ability, but not the desire, to weed out the trash before going to market with his wares.

What do they call these things, class actions? Quentin?

By the way, doesn't it seem sort of strange that a thread started by a possible troll has turned out to be one of the longer ones?

Ciao - Rob C

Just goes to show that we don't need much encouragement to express our views, eh?

Tooo trooo!

Rob C

What's a "2Dxs"?

How could you comment on a camera which isn't yet released (1D Mark III)? Have you beta tested it, or just commenting based on specs?

Actually if you mean D2X, it compares very well with Canon's best depending on what you do with your camera. I have Canon D30, 10D, 1D, 1DS and 1D Mark II along with Kodak DCS-760, Sigma SD10, Sigma SD14 and Nikon D2X.

The D2X is my first choice for wildlife because of the autofocus speed and accuracy, 1.5x crop, excellent buffer and versatility. My 1D used to be my favorite action camera - still very good. My 1Ds I like for some landscape work but it never goes in my bag for shooting widlife. My favorite dSLR for IQ is my Sigma SD14.

Different strokes for different folks and use the tool which works best for the task at hand.

The choice of tool among the best from Nikon and Canon pretty much boils down to ergonomics and personal preference. CMOS does better at high ISO but for many of us high ISO is not terribly important. Personally, if I have to shoot wildlife at over ISO 400 I don't waste my time. On the other hand when I was shooting action and sports my favorite tool was my Nikon based Kodak DCS-720X for low light high ISO work. So for action sports in natural no-flash low light conditions, high ISO is very important. For the event or wedding photographer who shoots perhaps 50% of his work where flash isn't allowed then high ISO thus Canon would prevail. For a wildlife shooter who prints large high ISO is pretty much wasted as it is for a landscape or portrait photographer.

There is no "right" answer to the OP's question - just choose what works for you because the skills as a photographer far outweigh the minor differential in equipment.

Lin

I really don't think it matters for a photography student whether they choose Nikon or Canon. They should try each and choose the one which feels right to them. Each has it's own set of advantages and disadvantages

Why is that? With the 5D there's a slight loss of resolution at ISO 1600. With the 1Ds2 there's vitually no loss of resolution but slightly more noise than the 5D at ISO 1600. I'm not sure about the 1D3 which has upped the ante to ISO 6400, but I'd be surprised if there's any objectionable noise or loss of resolution with that camera at ISO 1600.

What is more objectionable than a slight trace of high ISO noise is loss of resolution due to a shutter speed that's too slow.

Earlier in this thread I tried to give some reasoning as to why Canon may be a more useful choice currently. These were not based on whether they were better or worse than Nikon, but on the potentiality of which it might be more useful to be fully conversant with if trying to get assisting positions after studying. I know of assistants who have specifically gone for Canon precisely to get work (which they have).

Whilst the thread has been very interesting and has shown just what a superb choice of equipment we have today, many of the choices we have to make are based on not merely our desires to own specific equipment (I'd love an M8 - despite all its apparent flaws - as I've owned several M Leicas, but am realistic enough to appreciate that it would be a substantially underused and expensive luxury which I have trouble justifying), but on factors over which we have less control. In an ideal world it would be nice to own one or even several of each cameras and lens systems, but I have enough trouble carrying all the equipment I want to use now, to say nothing of the workflow implications. In reality most of us are keyed into a single system for a very long time once we have made that initial, important decision.

On the ISO point - I have had substantial trouble changing my mindset on using low ISOs (a hangup which derives from film days I know!), and do find the low noise, high ISO settings to be surprisingly good when I actually steel myself to use them. I wonder how many others feel the same?

Like my 1DS, neither the 5D nor the 1DS2 are suitable for my wildlife photography use but there is "objectionable" noise at even ISO 800 with both. As far as ISO 1600 or higher, it's fine for event photography where it's possible to use noise reduction in post and where print size requirements are not critical but useless when tight crops and enlargement are desired. ISO 6400 is rarely useful even with the finest low light cameras such as the Kodak DCS-620X. How the 1D3 will fare is yet to be seen, but judging from Canon's own released shots from this camera I sure hope it can do better.

Best regards,

Lin

Interesting perspective. It reminds me of the situation I was in when I first acquired a Canon D60. The ultimate resolution of a 6mp camera is less than fine grained 35mm film can offer, with a good scanner.

But the freedom from noise or grain that ISO 100 can offer, with a mere 6MP DSLR, gives the impression that the digital image is somehow closer to MF.

I was very reluctant to use ISO 400 with my D60, not because it was bad compared with film, but because the noise at ISO 400 reminded me of the grain with even low ISO film.

When I jumped from the D60 to the 20D (I don't believe in intermediate upgrades), I was surprised that the 20D produced better images at ISO 1600 than the D60 at ISO 400. Better resolution and better color saturation.

I've completely got over this reluctance to use high ISO. Some people haven't. It's probably a hangover from past experiences.

Good high ISO is great, but I will always keep trying to use low ISO if allowed by the image's type. That will remain the better option.

Cheers,
Bernard

Don't some Nikon DSLRs have some sort of ISO bracketing, Bernard? This is a long overdue feature missing on Canon cameras. It is of course better to use the lowest ISO consistent with appropriate shutter speed for the scene but correct f/stop for desired DoF. Choosing the appropritae f stop is often easier than ensuring an adequate shutter speed. F stop is consistent in its effect on DoF. Shutter speed is not consistent in its effect on image sharpness. Take 6 handheld shots at 1/30th (with a 50mm lens, no IS), and some will be sharper than others. One out of the 6 might even be perfectly acceptable.

ISO/shutter speed bracketing would solve this problem, ie. 3 automatic shots at ISO100 and a 30th, ISO 200 and a 60th, and ISO 400 at a 125th would be a better option than 3 shots at a 30th.

Using a trpod doesn't always solve the problem when there is subject movement, but I understand that your shots, Bernard, seem to be mostly of static landscapes, which is why I guess you are not particularly interested in high ISO performance.

Sure, they do and I use the function when I need to.



That's true, most of my images are of static landscape where long exposure time are often not a problem, or can sometimes even be a desired feature.

It is true that excellent high ISO would help in some occasions though.

Regards,
Bernard

This is really stupid, and I can't believe I'm actually posting to this. However, I just can't resist.

To say that generally one manufacturer is better makes no sense. If you don't start with some pretty well defined parameters or requirements, then the entire discussion is meaningless. I suppose someone could look at relative market shares between Canon and Nikon, and at any point in time, it will change. If you look at comparable cameras for a given set of features and price, you could identify what camera is selling more. Without some kind of framework, it means nothing.

Next, take all the comments with skepticism. You have "hacks" from both companies posting their hype. Some people actually search for this kind of forum and try guerrilla-marketing techniques to spread their FUD (fear, uncertainty and doubt). Others are trying to rationalize their decision to buy one over another.

I have experience with both Canon and Nikon products. They’re both excellent. I switched from a Canon EOS 850 to a Nikon D70s and was very happy, even though it meant new lenses. I recently added a Nikon D2Xs after trying several other products. I have the resources to buy any product that offered a value, and I was prepared to spend over $8000 or $9000 for a camera body. The D2Xs was well below my budget and met all my requirements better than anything else. That analysis was only relevant for me.

By the way, one of the highly respected experts in the field of photography is Ken Rockwell. Before the guerilla marketers overwhelm the reader with their hype, you should read Ken’s piece here:

http://www.kenrockwell.com/tech/nikon-vs-canon.htm

Tom

Nikon as the industry standard? Not on your life. Perhaps in the past it was the popular choice. I too was a Nikon user since 1976. F, F2, FE-2, F3, F100, etc., etc. My last set-up right before digital consisted of a Nikon kit, MamiyaRZ kit, and an Arca-Swiss 6x9/4x5 kit.

In 2003, I sold all of that and went with the Canon 1Ds. Never looked back.



Respected? Your'e being sarcastic I hope. Look here to see Ken using what he refers to as "left handed camera Nikon custom made for him". Yeah- right.
http://www.kenrockwell.com/

We have started with some well defined parameters. We're talking about 35mm photography; not P&S, not MF or LF.

Of course Nikon produces high quality equipment. I'm sure you can find some Nikkor lenses that are better than Canon equivalents. Each manufacturer has it's strengths and weaknesses, but in my view Nikon has got itself into a bind for opting for the sub-35mm format.

You might have noticed there's a general principle running through all photography; the larger the format, the higher the quality.

Nice use of irony.

That article is uninformed and useless.

That was the case in the days of film, when grain was obvious in A4 prints from 35mm film. But these days your statement is too simplistic to have much meaning.

Not at all. The principle is similar but the format differences across the spectrum of digital cameras are now often much smaller than was the case in the days of film. We now have a plethora of different digital formats varying in size to such a small degree that other factors, such as pixel density and noise handling, might be more significant than format size.

These other factors also used to minimise the differences between film formats. One would not see much differences in image quality between 35mm used with a fine grain, high resolving film and 6x4.5 format used with a lower resolving, coarser grain film.

The full frame 35mm format is 2.5x the area of the Nikon cropped format; great enough to make a difference. For Nikon to match the image quality of the 1Ds3, expected later this year and expected to have a 22mp sensor or thereabouts, it will either have to move to a larger format or pull some rabbits out of the hat.

Can you show me some examples of a current, latest model, digital camera that produces higher quality images than a current, latest model larger format digital camera?

I think that you will find that assumptions were made early on in the discussion that a student would be best advised to gain experience in the camera system most likely to be encountered when seeking work (perhaps as an assistant) after graduating. In my own experience (and other posters) this leads to the Canon 1D series, simply because these are the cameras most likely to be encountered in this situation. I'm not sure that many posts actually say the Canon are better than Nikon, most point out that there are differences and that there are more advantages in the circumstances in learning the Canon 1D series inside out rather than the Nikon system.

On the note about formats, the more I use my FF Canons, the more I am appreciating them as filling a niche midway between 35mm and 645 film - in my own rather simplistic and arbitrary way!

M8 vs. D2x or 1DsII (only at low ISOs in the latter)?

JC

Well first of all, John, these 3 cameras are an example of the small differences in format which, from an image quality point of view, can be over-shadowed by technological factors which blur the advantage of the larger format.

Whilst there's a significant difference in format size between the D2X and the 1Ds2, the M8 is about midway between the two formats. Nevertheless, it's understandable that the M8 might produce marginally higher image quality than the D2X, but I think any claims that the M8 produces better image quality than the 1Ds2 would be very contentious.

Can you point me to some reliable comparison images which illustrate the magnitude of such differences?

From a theoretical standpoint, it is possible for a smaller format camera to best a larger format camera if the smaller format has more resolution in terms of pixels/picture height than the larger format and the lens of the smaller format camera has sufficient resolution to match the sensor. Of course, to avoid diffraction limitations, you might have to shoot at a larger aperture with the smaller format camera.

Bjorn Rorslett did a shootout between the Nikon D2X and the Canon EOS 1Ds MII where he claimed that the Nikon came out on top even with less resolution, but this test is disputed by some. Since his web site does not support direct links, go to the D2X review and then to section 8, taking on the competition.

In practice, the larger sensor has an advantage in lower noise at high ISO. At base ISO, noise is not a problem with the D2X. However, I agree that if Nikon wants to match the resolution of the much anticipated Canon MIII, they will have to use a large sensor and improve its noise characteristics.

Bill

The point I was making is that your statement was too simplistic, and hence not very meaningful. If it was meaningful, then no-one would buy the Nikon D2x and everyone would instead buy the Canon 5D.

In many respects the Nikon D2x can produce 'higher quality images' (whatever that means) than a Canon 5D. Clearly at high ISO the 5D wins hands down. If dynamic range is your requirement, then the Fuji S5 Pro can outdo both of them. For many people the 'telephoto factor' of the smaller format provides higher quality images 'cos they can't get close to the subject.

I would not along with his claims 100%, but he certainly makes a good argument that the two cameras are in many respects comparable. But at high ISO the Canon walks over the Nikon. Well, actually at moderate and high ISO.

http://www.l-camera-forum.com/leica-forum/...est-studio.html

I really don't think the M8 is better than the D2x or the 1DsII. I just think they're different. The M8 is 10mp compared to 12+ for the D2x and 16 for the 1DsII. You could make a very good argument that the 1DsII is at least a full stop better than the M8 at high ISOs; of course, you could make the agument that the top Leitz glass is at least a stop better than the top Canon.

JC

Both Nikon and Canon are great, and its personal taste what you choose for a tool (I, for example, could never go with Canon, not even in digital video, I couldn't stand the XL-1, but that doesn't mean it is not good).

Pentax, a much smaller company, doesn't have the marketing punch, and so its new entry and midlevel SLRs do not get the attention they deserve in a Nikon-Canon bi-polar photography world.

Too many people shoot test charts instead what is out there (in the world) through what is in there (the vision in your mind).

What counts most is the photographer and his vision.

Ray,
since Nikon does not make 35mm format DSLR's, I dispute your first statement. We are talking about DSLR photography, and two companies that offer a total of four different formats of DSLR.

As to the "bigger is better" idea: surely, in choice of an SLR brand, bigger format size alone is not the decisive factor; there are many other factors beside format size. After all
1. If format size alone were decisive, neither Canon nor Nikon would be in the race, but only Hasselblad, Mamiya and maybe the new Rollei based systems.
2. Offering a choice of 35mm, 645 and 6x7 format in the film era did not make Pentax the first choice of professionals.
3. You at one point chose Canon DSLR's with 15.1x22.6mm sensors over Nikon DSLR's with somewhat larger sensors, and before I believe the announcement of Canon's first 35mm format DSLR.

You will probably be quick and correct to point to compensating advantages of the smaller format options from Canon (and perhaps even Nikon) over those larger format options in areas such as cost, lens selection, frame rate, shutter lag, portability and in particular telephoto reach, but a lot of those same factors also often weigh against 35mm format in favor of smaller DSLR formats.

P. S. Perhaps I should take it back about the telephoto reach disadvantage of DMF: current DMF backs with 6.8 micron pitch Kodak sensors offer higher resolution in l/mm than any Canon 35mm for 1.3x format DSLR, and new Dalsa sensors at least match Canon's, so DMF cameras can probably match or slightly outperform those Canon models for resolution by using the same focal length and cropping. But any of the D2Xs, 40D, D40x, D80, D200 or 30D is ahead on telephoto reach with a given focal length. (Not to mention the current DSLR l/mm resolution leader, the Olympus E-410.)

I think you are wrong. My statement is simple but true. I'll repeat it. "There's a general principle running through photography; the larger the format the better the quality." The word general qualifies that it's not an immutable law without exception, but a tendency.

I've even seen comparisons between 35mm film and 4x5 format film where the 35mm image quality was virtually as good at the same enlargement, just a bit more obvious grain. How was that achieved? By using a very sharp and expensive 35mm lens at f5.6 with the 35mm format, as opposed to a standard lens at f22 with the 4x5 format in order to get equal DoF for both shots, a reasonable requirement I think when comparing image quality from different formats. Different film was also used; T-Max 100 with 35mm and Tri-Ex 400 with the 4x5 format.

I think it is understood that there are many factors other than format size that influence image quality; lens quality being just one of them. You didn't think, Slough, that I was trying to say that format size is the only thing that matters, did you?

My remark was directed only at raw image quality with which this site seems to be mostly concerned. When I first came across LL, Michael's contentious review of the Canon D30 was the main topic. How could a 3mp sub-35mm format produce 8x12 prints that actually looked better and sharper than 35mm film?

Since those days it seems that Michael and half the contributors to this site have moved up from sub-35mm to FF 35mm to sub 6x4.5 format, and the reason is, I would suggest, because the larger formats (tend to) offer better image quality.

You'll also notice that the great majority of cameras models out there, the very small format P&S, hardly get a mention on this site. Why is that? Because the format is too small to allow the standard of image quality that readers of this site are mainly concerned with.