What has happened to Canon?

[RedRobin]

I honestly believe that too many people spend far too long looking at review sites and other such internet sites of expertise and correctness, rather than using the cameras and seeing if they like what they produce with there cameras.

That is not leveled at anybody in this thread at all,I hasten to add, just something that bugs me.

....I agree. The time to trawl through reviews etc is when you are deciding what to consider buying.

It doesn't bug me though.

 

[arad85]

Somewhat I have to disagree with this notion, DXO has got little to do with real life photography.

Nikon D3200 > 5D MK3?
Sony A99 >>>> Canon 1DX?

Can't be true

Why can't it be true? Why do people assume numbers are made up or don't have any relevance in the real world?

Noone is saying you can't take a stunning photo with a Canon, just that other sensors are better and you might be able to take an even better picture with an alternative, especially if you need to do heavy post on getting more dynamic range out of the photo.

I also take umbrage about pixel peepers being viewed as "lower than low". I'm one - don't mind admitting it. Does it make my views any less valid?

 

[LongLensPhotography]

Why can't it be true? Why do people assume numbers are made up or don't have any relevance in the real world?

Noone is saying you can't take a stunning photo with a Canon, just that other sensors are better and you might be able to take an even better picture with an alternative, especially if you need to do heavy post on getting more dynamic range out of the photo.

I also take umbrage about pixel peepers being viewed as "lower than low". I'm one - don't mind admitting it. Does it make my views any less valid?

I suggest you test the hell out of 1Dx and then compare the rest.

DxO results are only as good as the set of rules a few guys make up. Like surveys these can be vastly biased and distorted. I could ask some certain questions to portray a certain opinion.

I would be honestly surprised if that 3200 thingie can produce a better large print of a complicated scene than my cheapest body - 9 year old 1Ds II. Any takers?

 

[EspressoJunkie]

Obviously everyone does things their own way, but personally speaking since Ive tried to stop myself pixel peeping I have found myself concentrating much more on (in my eyes) the much more important aspects of composition, colour, subject, and essentially the picture as a whole.

Of course ultimate IQ is of importance, but I can tell you now that in the last few months there's more than one picture I've kept where before it would have been binned.

Basically the only people who really care about what a picture looks like at a pixel level are other photographers. I doubt very much that anyone else will notice a little more noise or a little less resolution or vice versa.

 

[arad85]

I suggest you test the hell out of 1Dx and then compare the rest.

No. I have DxO to do that for me. If you are so interested in claiming DxO numbers as fake, why don't you test the A99 and 1DX against each other and prove them wrong. What people are doing is taking the single figure and not interpreting the wider set of figures that make up the DxO number and looking at what they actually mean. For example, my GH3 has a wider dynamic range than all Canon sensors at base ISO, but this changes as ISO increases. Also the noise levels are way higher at equivalent ISOs (which is what you';d expect when comparing against a FF sensor).

Canon sensors ALL have a prevalence to level off their dynamic range at lower ISOs as far as DxO is concerned. Unless there is a problem in the testing methodology, then the only thing you can conclude is that it is an inherent trend in Canon sensors.

 

[arad85]

Basically the only people who really care about what a picture looks like at a pixel level are other photographers. I doubt very much that anyone else will notice a little more noise or a little less resolution or vice versa.

IMHO, it's not about the pixel level (noise/resolution), it's about a sensor delivering you the ability to do things in post that you can't otherwise do. An extreme example can be found here: http://admiringlight.com/blog/olympus-e-m5-vs-panasonic-gh2-dynamic-range-battle/

 

[LongLensPhotography]

No. I have DxO to do that for me. If you are so interested in claiming DxO numbers as fake, why don't you test the A99 and 1DX against each other and prove them wrong. What people are doing is taking the single figure and not interpreting the wider set of figures that make up the DxO number and looking at what they actually mean. For example, my GH3 has a wider dynamic range than all Canon sensors at base ISO, but this changes as ISO increases. Also the noise levels are way higher at equivalent ISOs (which is what you';d expect when comparing against a FF sensor).

Canon sensors ALL have a prevalence to level off their dynamic range at lower ISOs as far as DxO is concerned. Unless there is a problem in the testing methodology, then the only thing you can conclude is that it is an inherent trend in Canon sensors.

I am not saying the numbers are wrong, but merely questioning what numbers you are looking at. What does DR mean for example precisely? And what the hell is the overall score? I see D3200 has only got any chance to do anything significant at ISO100 compared against 5DIII, and then it is completely down the hill. And have they considered plxel sharpness? Colour and DR accuracy vs sheer scale? NO. Overall it is like saying we should quite EU because some idiot told us.

 

[arad85]

I am not saying the numbers are wrong, but merely questioning what numbers you are looking at.

Which ever one I need to look at for the parameter I'm interested in. The sensor score is just a weighted average of other underlying scores. It's as relevant as quoting a single benchmark figure for a computer...

What does DR mean for example precisely? And what the hell is the overall score?

http://www.dxomark.com/About/Sensor-scores

And have they considered plxel sharpness?

That's a property of the sensor/lens combination and is measured if you look at the lens tests.

I use DxO as a method of getting somewhere close to a level playing field. I'm currently deciding on a new car and which car I will get is a tradeoff of engine (power, acceleration, fuel economy), features and cost. Whilst the features and cost are known, I can only go on standardised manufacturers figures for the engine performance. I don't look at them to say that's what I'll get as they are idealised, but I do look at them to give me some idea of their strengths and weaknesses. The consequence is that I ended up choosing a lower powered car as it is more fuel efficient and I want to minimise monthly ongoing costs....

 

[Gaz J]

DXO call it as they see it. But it's just another piece of information to use when choosing a camera. If you're tied into a system it doesn't matter that a D800 has greater dynamic range that a 5D3 at lower ISO.

There's the other features of the camera to consider depending on what you use it for, the lens and accessory range if you want to do something a bit different with it.

The sensor, whilst important, is not the end product. Otherwise we'd all be on Sony sensors.

 

[specialman]

Who came up with the myth that Nikon cameras are more advanced or better than Canon ones?

Probably the same people who came up with all the other 'myths' about camera equipment, which in reality mostly boils down to personal preference and not actual fact....

 

[EspressoJunkie]

The sensor, whilst important, is not the end product. Otherwise we'd all be on Sony sensors.

I used to think it was! I don't think like that anymore, as you say if I did I'd still be shooting Nikon.

Also as lovely as Nikon cameras sensors are, I genuinely prefer the colours and white balance from Canon ( I realise this is subjective). For me it takes much less tweaking to get a Canon file looking "finished".

I'm sure other people would disagree, but that's what's I've found.

 

[Rudlin]

Don't believe DXO

 

[Gaz J]

D800 is a fabulous bit of kit. Whoever I see images that a friend takes with it I am blown away by the ability to crop into it and still get decent sized prints. But I shoot wildlife and motorsport and am tied into canon too much to contemplate a change now. I also genuinely believe that my 5D3 is a better all round camera for my needs than the D800 would be.

My mate reaches for his D3 when he shoots wildlife as he feels that's it's a far better option. That's not to say you can't use the D800 it's just not primarily designed for that
use.

I was going to buy a 7D simply for it's crop factor but after reading up on it and more importantly talking to people who had used it for the same purpose that I would decided that it's issues with noise at higher ISOs were not for me. I've seen some good high ISO images from the 7D but I don't regret for one second buying the 5D3 instead of it.

I'm a full frame convert now and will shortly be replacing my 1D MkIV with a 1DX and the fact that the D4 gets a higher score from DXO won't leave me feeling I've been short changed.

 

[RedRobin]

Obviously everyone does things their own way, but personally speaking since Ive tried to stop myself pixel peeping I have found myself concentrating much more on (in my eyes) the much more important aspects of composition, colour, subject, and essentially the picture as a whole.

Of course ultimate IQ is of importance, but I can tell you now that in the last few months there's more than one picture I've kept where before it would have been binned.

Basically the only people who really care about what a picture looks like at a pixel level are other photographers. I doubt very much that anyone else will notice a little more noise or a little less resolution or vice versa.

....I wholeheartedly agree. Some people can't see the wood for the trees (or is it the other way around?). Pixels doth not a picture make.

However, anyone involved in commercial print production will be checking the repro quality and also some commercial photo libraries.

 

[Fireproof Creative]

Why is MP so important to you, out of interest?

Quite simply the ability crop and still have a decent size image for repro. Obviously pixels have little bearing on the talent side but that's not the point. If IQ is perfect at 100%, which it is on the d800 and 5D 3 then you are less reliant on reach which for Motorsport stuff can be the difference between £1000 and £8000+

The D800's high pixel count allows it more flexibility in this regard. Banding about pixel peeper as derogatory just means you don't fully understand how some people use their equipment. I love my 5D and am unlikely to change for as many years as it works, but I do see the advantage of a quality sensor with a even higher pixel count.

 

[big soft moose]

Quite simply the ability crop and still have a decent size image for repro. Obviously pixels have little bearing on the talent side but that's not the point. If IQ is perfect at 100%, which it is on the d800 and 5D 3 then you are less reliant on reach which for Motorsport stuff can be the difference between £1000 and £8000+

The D800's high pixel count allows it more flexibility in this regard. Banding about pixel peeper as derogatory just means you don't fully understand how some people use their equipment. I love my 5D and am unlikely to change for as many years as it works, but I do see the advantage of a quality sensor with a even higher pixel count.

while this is true thers also a converse of how much people actually crop , and how big a file you actually need your end result to be , so even with cropping there'll soon come a point where more MP doesnt necessarily mean better.

Also the more you crop in the more sensor noise becomes an issue, so if cramming **** knows how many MP onto a sensor (particular an aps C sized one) increass sensor heat levels and thus digital noise , there will eventually come a point where this degradation of IQ outweights the ability to crop conferred by yet more MP

I would prefer a clean image from an 18MP sensor (which can still be cropped by about half) to a 36MP noisy one resulting from canon trying to double the number of pixels recorded inthe same space

 

[Fireproof Creative]

while this is true thers also a converse of how much people actually crop , and how big a file you actually need your end result to be , so even with cropping there'll soon come a point where more MP doesnt necessarily mean better. Also the more you crop in the more sensor noise becomes an issue, so if cramming **** knows how many MP onto a sensor (particular an aps C sized one) increass sensor heat levels and thus digital noise , there will eventually come a point where this degradation of IQ outweights the ability to crop conferred by yet more MP I would prefer a clean image from an 18MP sensor (which can still be cropped by about half) to a 36MP noisy one resulting from canon trying to double the number of pixels recorded inthe same space

Granted but Nikon has managed it very well at 36, canon improving on IQ at that pixel range would be useful to me. I'm not talking about an endless race to higher counts at the end of the day the IQ is the single most important factor to me.

I could do with an extra 15 and I crop a lot at 400mm

 

[HoppyUK]

Granted but Nikon has managed it very well at 36, canon improving on IQ at that pixel range would be useful to me. I'm not talking about an endless race to higher counts at the end of the day the IQ is the single most important factor to me.

I could do with an extra 15 and I crop a lot at 400mm

It makes hardly any difference.

If you crop to half the total image area (1.4x linear), with the 5D3 (5760x3840) you will have 4114 pixels on the long side. With the D800 (7360x4912) the same crop will leave 5257 pixels.

Assuming you have a very sharp lens in the first place, the D800's pixel advantage is only 20% or so - barely noticeable.

 

[Radiohead]

Or try getting it right when you take the shot rather than relying on cropping. No excuse for you landscape types, you've got hours and it doesn't move.

 

[Professional]

Or try getting it right when you take the shot rather than relying on cropping. No excuse for you landscape types, you've got hours and it doesn't move.

Not landscapes, but sports and birds/wildlife and sometimes aviation if only using full frame camera, 1DX is amazing fast camera for those but it is only 18mp, i don't have any long tele and i may not able to afford one, DSLR with higher mp is cheaper than 400-600mm primes, so in this case i would like to have that ability to crop and retain high quality.

For landscapes i crop very rare and about %5-15 at most cases, so no problem there for landscapes, and wide angle-mid range lenses are far way easier to afford than 400 f2.8 or 500 or 600/800 [Brand new or second hand but latest versions/models MKII].

 

[Fireproof Creative]

Not landscapes, but sports and birds/wildlife and sometimes aviation if only using full frame camera, 1DX is amazing fast camera for those but it is only 18mp, i don't have any long tele and i may not able to afford one, DSLR with higher mp is cheaper than 400-600mm primes, so in this case i would like to have that ability to crop and retain high quality. For landscapes i crop very rare and about %5-15 at most cases, so no problem there for landscapes, and wide angle-mid range lenses are far way easier to afford than 400 f2.8 or 500 or 600/800 [Brand new or second hand but latest versions/models MKII].

Exactly!

 

[Fireproof Creative]

It makes hardly any difference. If you crop to half the total image area (1.4x linear), with the 5D3 (5760x3840) you will have 4114 pixels on the long side. With the D800 (7360x4912) the same crop will leave 5257 pixels. Assuming you have a very sharp lens in the first place, the D800's pixel advantage is only 20% or so - barely noticeable.

that's still 1000px of extra detail which does, believe it or not, make a fair amount of difference to me and what I use images for

 

[Gaz J]

Not landscapes, but sports and birds/wildlife and sometimes aviation if only using full frame camera, 1DX is amazing fast camera for those but it is only 18mp, i don't have any long tele and i may not able to afford one, DSLR with higher mp is cheaper than 400-600mm primes, so in this case i would like to have that ability to crop and retain high quality.

For landscapes i crop very rare and about %5-15 at most cases, so no problem there for landscapes, and wide angle-mid range lenses are far way easier to afford than 400 f2.8 or 500 or 600/800 [Brand new or second hand but latest versions/models MKII].

But you need the rest of the package to go with the ability to crop in to the frame. Big megapixel sensors dont seem to be able to go in hand in hand with 12 FPS. Its all a compromise. You trade one part of the spec against another.

If the D800 could do 8 FPS at that resolution with a 25 shot RAW buffer than it would be really something. But it cant and that is why for me the 5D3 is a better all round camera.