Is the PEN is mightier than the Nikon?

[goldenlight]

I've just spent several hours (how sad am I?) having a very interesting play with Imaging Resource's Comparometer (link http://www.imaging-resource.com/IMCOMP/COMPS01.HTM) which, amongst other things, features high ISO comparison between different cameras on their database. Frankly, I was shocked by what I found.

First, I compared the E3 against the latest Olympus model, the E-PL1, to see what progress has been made in the last two years. No surprise here, the newer camera was comfortably better at ISO 3200. So I thought I'd see how the little Pen compares to the mighty Nikon D700. I was expecting it to be crushed, but I was in for a surprise.

In all cases except the last two I compared the Pen at ISO 3200, firstly to the D700 at ISO 25000, which would give the Nikon a three stop advantage. Quite frankly, the Pen wiped the floor with it's bigger adversary in terms of both noise and detail. How about ISO 12500 then? That would give the Nikon a two stop margin. Nope, there was very little between the two in terms of noise, indeed the Olympus seemed to fare better in certain areas of the test image and the Nikon in others, but the Pen still held a clear advantage in terms of sharpness and resolved detail.

At ISO 6400 the D700 finally displayed a clear advantage over the Pen in terms of noise control, but still the Pen image was sharper and more detailed! Finally, comparing like for like with both cameras at ISO 3200, the Nikon shakes the Pen off it's back with an equally detailed image and significantly better noise control (noise? What noise?). In fact, it has to be said that the D700 has a slight overall edge at ISO 3200 over the E-PL1 at ISO 1600. Finally, I compared both at ISO 800, where the Nikon advantage is marginal and of little consequence – sure, it's a bit cleaner but the Pen image at that speed can still be classed as excellent.

Puzzled by the surprising better detail and sharpness from the Pen I also checked the database images for the D3s in case there had been a focusing error in the D700 test but results were pretty much the same.

So what did I conclude from all that pixel peeping?

Well, in terms of ISO performance and noise control the D700 appears to have about a two stop advantage, which narrows when ISO values fall below ISO 3200. However, this advantage is at the expense of detail and sharpness. I'm thinking that unless they have something really special up their sleeve Olympus could do worse than upgrade the E-3 to this sensor and processor – and do it now. When fitted with SHG zooms that would reduce the D700 advantage by at least a stop, unless Nikon fast primes were used.

In term of what the E-3 could be, right now, the D700 would carry very little advantage – except for those specialists who really need ISO 6400 or more, whatever the compromise in quality.

That's rather a bold statement, so are there any caveats? Most certainly! One fly in the Olympus ointment is the D300s, which I also took a crafty peek at. Noise at ISO 3200 is comfortably better than the EPL-1, although once again at the expense of detail and sharpness. This is also quite an old camera (despite a facelift) and yet it still seems to match the best and most current that Olympus can offer.

Also, I don't know how consistent those tests are. Do the Nikons fare so badly against the Pen in terms of sharpness and detail because they were set with a higher level of noise reduction? Mind you, if that were the case it would also mitigate some of the Nikon advantage in terms of noise at high ISO. How accurately were the Nikons focused, or is it down to superior Olympus optics (even in the kit zoom) or better DOF from the small sensor?

I should also mention that my observations were conducted at high magnification on the screen and with at normal size the Nikon images look far sharper – but then, the Olympus images look less noisy so once again we have a level playing field. Also, of course, the results may look very different on large, high resolution prints than they do on my monitor. Nevertheless, however you look at it, Olympus are making giant strides towards overcoming their one Achilles Heel and without doubt the Four Thirds format will soon equal APS-C in terms of image quality – if it doesn't already.

 

[Mr_Sukebe]

John,
Some interesting thoughts. Just one point. You're suggesting that the D300s is better than the EPL-1, and that the EPL-1 is close to the D700. By inference, that's suggesting that the D300s is going to probably be better than the D700.
Does that sound right to you?
That sets a few alarm bells ringing for me.

 

[ryanyboy]

This all sounds mad to me John I'm afraid. The D300s is a million miles away from the 700 in terms of noise. Sorry to be so blunt but I really think you're findings are flawed

 

[Flash In The Pan]

So, what you're saying is that a 4/3 sensor produces better results than a full frame one and that an APS-C sensor beats both of those?

 

[dalex_257]

...
First, I compared the E3 against the latest Olympus model, the E-PL1, to see what progress has been made in the last two years. No surprise here, the newer camera was comfortably better at ISO 3200. So I thought I'd see how the little Pen compares to the mighty Nikon D700. I was expecting it to be crushed, but I was in for a surprise.

Are you sure you looked at the right pictures? Looking at E-PL1 ISO 3200 image shows significantly more noise and smudged details than D700 at ISO3200. Look at a hair of the model on full res images to see the difference. Even D700 at ISO6400 looks better than E-PL1 ISO3200.

In addition to that, I would take any resource that does not publish how they produced resulting JPEGs with a degree of scepticism. Did they even use the same focal length lens on all cameras? Sure they described their lighting setup to great extent but I have not found anything saying how the final comparison images were produced. I.e. was it JPEG right out of the camera or RAW file processed, what were the processing options or JPEG settings and so on.

In the end to me it looks like they used straight JPEGs out of the camera because E-PL1 images of higher ISO have that plastered look with smudged details that can only be attributed to noise reduction (hence the guess that it was not RAW). You can clearly see it in their E-PL1 ISO 3200 full res image on the skin of the model face and especially her hair.

 

[Bendihossan]

Sorry but a 4/3rds sensor doesn't hold a candle to a full frame sensor. A FF sensor records more detail and inherently handles noise better. Just look at the size difference.

This review doesn't add up to me.

 

[goldenlight]

I take on board all the comments made so far - I, too, was rather shocked at what I was seeing with my own eyes. I downloaded the still life test scene at various ISO settings and spent considerable time comparing the images greatly enlarged. It was quite clear that although the Nikon had a definite advantage in terms of noise at high ISO it was not as substntial as I would have expected.

It was also plain to see that the Pen was significantly sharper and captured more detail (in most areas of the image) unless the shooting conditions were flawed. I was particularly surprised that this seemed to hold true at high ISO - low ISO would not have been quite so unexpected because when Amateur Photographer tested the Panasonic G1 (essentially the same sensor) they found that at ISO 100 in good light the G1 out-resolved the D3.

Now I do agree that this comparison I stumbled across is far from authoratitive or conclusive and I did acknowledge this in the caveats that I made at the end of my piece. I think we would all expect the FF Nikon to perform significantly and consistantly better in real life situations. Indeed, at the risk of upsetting Canon users I would say that Nikon FF cameras represent the current peak of DSLR performance. However, I believe this does illustrate that Four Thirds sensor technology has vastly improved over the last two years and, whilst still some way off FF in many respects, it is definitely banging on the door of APS-C cameras.

 

[dalex_257]

It was quite clear that although the Nikon had a definite advantage in terms of noise at high ISO it was not as substntial as I would have expected.

I must be looking at some other photos but on that link you specified I cannot see this at all - advantage is quite substantial to my eyes. Also comparing well lit scenes using high ISO does no justice to it really. If you want to see real benefits try low light, poorly lit scenes which is where D700/D3 sensor really excels.

For their still setup it is even more pronounced.

Here is their E-PL1 sample (full res) of ISO3200

And here is their D700 sample (full res) of ISO3200

I think the noise difference is quite significant.

It was also plain to see that the Pen was significantly sharper and captured more detail (in most areas of the image)

Have you paid attention to what I said about how the images were taken? For out of camera JPEGs this tells you nothing really as it can be down to camera default sharpening. For RAW they had to use the same RAW converter with exactly the same settings. Plus for the different crops (lenses, sensor sizes) the images should really be brought to lowest common denominator (i.e. downsized to the smallest size and crop) to make a reasonable comparison. In addition the choice of lens has a significant impact on this and I doubt the lenses were exactly the same in both cases.

Looking at their EXIF does not tell much either and it is a bit suspicious as they stripped out all data except focal length, aperture, shutter speed and whether flash was fired. No ISO, lens data or camera settings.

The D700 image of the house does look soft but looking at the above D700 ISO 3200 samples shot with the same focal distance, and aperture and yet much much sharper I would say that something went wrong for them somewhere (even camera shake as the house images were taken at 1/25 sec).

low ISO would not have been quite so unexpected because when Amateur Photographer tested the Panasonic G1 (essentially the same sensor) they found that at ISO 100 in good light the G1 out-resolved the D3.

This really is laughable - their statement is based on what? Do you know how they tested, in what conditions, what they used to process images (raw converter etc) and how they measured resolution? It would not be the case as it is mostly with pretty much all mass produce magazines that they would not differentiate resolution from image size (and the two are not the same at all)? The resolution has a direct dependency on demosaicing (RAW conversion) whether it happens in PP or directly in camera and a whole load other factors like lens, flare, filters etc (have a look at DPreview there are a couple chaps who made a few RAW converters posting about that from time to time and the). To get a comparison of the resolution, you would ideally need to place the same high quality glass on both cameras (so it does not compromise the resolution), use deep hoods to avoid reflections and flare, use optimal aperture to avoid diffraction and then use the best possible quality RAW converter to avoid buggering up all resolution (RPP for example). Now do you know if Amateur Photographer went through all of this to reach their conclusion or do they simply shot some scenery and judged by the eye comparison using JPEGs from camera?

I don't doubt that 4/3 and micro 4/3 cameras are good (great even) - it's just that the bland statements of the popular publications like this create all the misnomers and fool people. It was the same with DPreview when their D3s review stated that 12 Mp was not up to standards of when D3 was released and was considered low resolution. This caused a few objective replies from the likes of Thom Hogan and as a result Thom is now avoiding that site.

 

[woof woof]

I have a GF1 which should be pretty similar to the Pen. I only shoot RAW and I find that the noise performance doesn't compare to that of my Canon 20D but noise isn't everything and is sometimes given too much weight IMHO.

Looking at the overall image the quality of MFT has really shocked me as has the amount of post capture processing that can be done on my GF1's images without them going bad, I can't perform the same processing on my 20D images without them looking bad.

Test results aside, I think that anyone who hasn't tried a MFT may be genuinely shocked at how good they are once they do and if noise is the only issue I'm pretty sure that careful post capture processing can sort that out to a large degree.

 

[goldenlight]

I must be looking at some other photos but on that link you specified I cannot see this at all - advantage is quite substantial to my eyes. Also comparing well lit scenes using high ISO does no justice to it really. If you want to see real benefits try low light, poorly lit scenes which is where D700/D3 sensor really excels.

For their still setup it is even more pronounced.

Here is their E-PL1 sample (full res) of ISO3200

And here is their D700 sample (full res) of ISO3200

I think the noise difference is quite significant.




Have you paid attention to what I said about how the images were taken? For out of camera JPEGs this tells you nothing really as it can be down to camera default sharpening. For RAW they had to use the same RAW converter with exactly the same settings. Plus for the different crops (lenses, sensor sizes) the images should really be brought to lowest common denominator (i.e. downsized to the smallest size and crop) to make a reasonable comparison. In addition the choice of lens has a significant impact on this and I doubt the lenses were exactly the same in both cases.

Looking at their EXIF does not tell much either and it is a bit suspicious as they stripped out all data except focal length, aperture, shutter speed and whether flash was fired. No ISO, lens data or camera settings.

The D700 image of the house does look soft but looking at the above D700 ISO 3200 samples shot with the same focal distance, and aperture and yet much much sharper I would say that something went wrong for them somewhere (even camera shake as the house images were taken at 1/25 sec).



This really is laughable - their statement is based on what? Do you know how they tested, in what conditions, what they used to process images (raw converter etc) and how they measured resolution? It would not be the case as it is mostly with pretty much all mass produce magazines that they would not differentiate resolution from image size (and the two are not the same at all)? The resolution has a direct dependency on demosaicing (RAW conversion) whether it happens in PP or directly in camera and a whole load other factors like lens, flare, filters etc (have a look at DPreview there are a couple chaps who made a few RAW converters posting about that from time to time and the). To get a comparison of the resolution, you would ideally need to place the same high quality glass on both cameras (so it does not compromise the resolution), use deep hoods to avoid reflections and flare, use optimal aperture to avoid diffraction and then use the best possible quality RAW converter to avoid buggering up all resolution (RPP for example). Now do you know if Amateur Photographer went through all of this to reach their conclusion or do they simply shot some scenery and judged by the eye comparison using JPEGs from camera?

I don't doubt that 4/3 and micro 4/3 cameras are good (great even) - it's just that the bland statements of the popular publications like this create all the misnomers and fool people. It was the same with DPreview when their D3s review stated that 12 Mp was not up to standards of when D3 was released and was considered low resolution. This caused a few objective replies from the likes of Thom Hogan and as a result Thom is now avoiding that site.

You rightly draw attention to the significant advantage of the D700 when both cameras are set to ISO 3200. I thought I'd already said pretty much the same thing; "comparing like for like with both cameras at ISO 3200, the Nikon shakes the Pen off it's back with an equally detailed image and significantly better noise control (noise? What noise?)."

It's really quite laughable that you should critisize the Amateur Photographer test, It shows you don't read the magazine otherwise you would know how they conduct their tests, it's quite transparent. AP is not renowned for praising Four Thirds cameras - quite the opposite - and they of all people can hardly be accused of bias in favour of Four Thirds.

Reading the rest of your post it is clear that there are a lot of if's, maybe's, something must have gone wrong and the tests must be done in the right light to suit the Nikon. I really wouldn't have thought that many excuses were needed to re-assert its authority. Had you read my post carefully you would have seen that I agree with much you say and that the real point I'm making is that Olympus and Panasonic are closing the gap, whilst readily acknowledging that there is still a gap. I find it quite amusing that the need should be felt to defend the D700 so robustly - I thought we all took it for granted that it is a great camera in a class of it's own.

Looking ahead, I would expect small sensor technology to improve to the point that sometime in the next ten years the advantage of full frame becomes pretty meaningless for the vast majority of applications. Full frame may even disappear as a result. For the time being it still reigns supreme in terms of absolute quality but the gap is closing.

 

[boyfalldown]

Looking ahead, I would expect small sensor technology to improve to the point that sometime in the next ten years the advantage of full frame becomes pretty meaningless for the vast majority of applications. Full frame may even disappear as a result. For the time being it still reigns supreme in terms of absolute quality but the gap is closing.

surely simple physics will tell you that'll never be the case. The gap in overall image quality, may (or more likely may not) close.

But simply thinking about crop factors and you know that statement is wrong - ok if you want a wide shot you can make shorter lenses for crop - but what about applications that call for a shallow depth of field - to get the same DoF as a f1.4 lens on full frame, on a 4/3rds sensor you need a lens thats somehow capable of working at f0.7 for a given focal length

 

[dalex_257]

It's really quite laughable that you should critisize the Amateur Photographer test, It shows you don't read the magazine otherwise you would know how they conduct their tests, it's quite transparent. AP is not renowned for praising Four Thirds cameras - quite the opposite - and they of all people can hardly be accused of bias in favour of Four Thirds.

I actually do read them as well as other photo magazines - they never go into details I mentioned in my messages above - which is not the same as how they conduct their tests. And I was not stating that they are praising 4/3 or bashing it - where in my message have I said that? I was only referring to blunt statement about "outresolving". Comparing two different cameras is not an easy task at all and I have yet to see a serious approaches to this by any media or magazine other than just shoot a few shots and do what is best called an subjective guess.

Reading the rest of your post it is clear that there are a lot of if's, maybe's, something must have gone wrong and the tests must be done in the right light to suit the Nikon. I really wouldn't have thought that many excuses were needed to re-assert its authority.

Those if's, maybe's are there because it is impossible to infer precise testing conditions and equipment used from those images you referred to on that site. Even more some of their images made on D700 with the same focal length and aperture (the only data they provide on EXIF) seem radically different which in itself raises a point of concern about consistency of their tests and approach to taking test photos.

Looking ahead, I would expect small sensor technology to improve to the point that sometime in the next ten years the advantage of full frame becomes pretty meaningless for the vast majority of applications. Full frame may even disappear as a result. For the time being it still reigns supreme in terms of absolute quality but the gap is closing.

You omitting a very simple fact - sensor technology advances are not limited to small sensor sizes and technology advances for the larger sensors as well as for the smaller ones. The larger sensors however will still have their benefits simply because they will always be capable to be pushed further than smaller ones (sensitivity, noise, resolution, colour accuracy etc).

 

[goldenlight]

surely simple physics will tell you that'll never be the case. The gap in overall image quality, may (or more likely may not) close.

But simply thinking about crop factors and you know that statement is wrong - ok if you want a wide shot you can make shorter lenses for crop - but what about applications that call for a shallow depth of field - to get the same DoF as a f1.4 lens on full frame, on a 4/3rds sensor you need a lens thats somehow capable of working at f0.7 for a given focal length

Ah, now that's a moot point; is DOF a factor of image quality? If it is it cuts both ways. Sure, Full Frame will always have an advantage for shallow DOF but likewise Four Thirds will always have an advantage for maximum DOF.

 

[goldenlight]

You omitting a very simple fact - sensor technology advances are not limited to small sensor sizes and technology advances for the larger sensors as well as for the smaller ones. The larger sensors however will still have their benefits simply because they will always be capable to be pushed further than smaller ones (sensitivity, noise, resolution, colour accuracy etc).

But there is a law of diminishing returns. Once image quality gets to the point where further increases will not be discernable to the human eye then the advantage of Full Frame over Four Thirds will be negated. That may or may not happen, of course. It's pure conjecture, I'll readily admit, but eventually I think it will. That is pure opinion and I make absolutely no claim that it can be substatiated with fact.

 

[Flash In The Pan]

I'm just chuffed that by your reckoning my D300 is better than both the D700 (and therefore the D3) and the EP-1 and here was me going to "upgrade" too

 

[goldenlight]

I'm just chuffed that by your reckoning my D300 is better than both the D700 (and therefore the D3) and the EP-1 and here was me going to "upgrade" too

Well, the D300 is a fine camera with very refined technology, by any standard. But I was only commenting on what my eyes (imperfect with age though they may be) were seeing. I have repeatedly stated my doubts about the true value of the comparisons but nevertheless consider it to be an interesting resource worthy of discusion. It is important to acknowledge that it is just one set of results from one source which can hardly be regarded as definitive and, as I said, I was rather surprised at what I was seeing.

I think the true moral of this is that there are now many fine cameras representing a number of serious formats. It is fantastic that we have this level of choice and it must be welcome news to all photographers that Four Thirds, with all its advantages of compactness and portability, is now comparing so favourably with larger formats. We've never had it so good.

 

[boyfalldown]

Ah, now that's a moot point; is DOF a factor of image quality? If it is it cuts both ways. Sure, Full Frame will always have an advantage for shallow DOF but likewise Four Thirds will always have an advantage for maximum DOF.

it is if you shoot portraits or anything else that needs a shallow DoF - and on a 4/3rds sensor you'll start to hit diffraction issues long before you max out DoF - so the |FF advantage cuts both ways

 

[dalex_257]

But there is a law of diminishing returns. Once image quality gets to the point where further increases will not be discernable to the human eye then the advantage of Full Frame over Four Thirds will be negated. That may or may not happen, of course. It's pure conjecture, I'll readily admit, but eventually I think it will. That is pure opinion and I make absolutely no claim that it can be substatiated with fact.

Define "image quality" then? I don't think sensitivity for once will ever diminish or get better for that matter for smaller sensors - larger sensor with larger photosites with other things being theoretically equal will always result in better sensitivity. The human eye does not have much to do with it but more the ability to shoot in even lower light conditions.

 

[goldenlight]

Define "image quality" then? I don't think sensitivity for once will ever diminish or get better for that matter for smaller sensors - larger sensor with larger photosites with other things being theoretically equal will always result in better sensitivity. The human eye does not have much to do with it but more the ability to shoot in even lower light conditions.

This is all getting rather theoretical, but there must come a point when meaningful improvements are no longer possible. Once you are able to shoot a black cat in a coal cellar at midnight on a moonless night with results as good as if it was at ISO 100 on a bright day where else do you go? Once you can make poster size prints with detail as fine as the human eye is able to discern how do you improve on that?

Fanciful? Maybe. But all it needs is for some bright engineer to develop a way to bypass the sensor signal amplification connundrum and you have a whole new technology. Ten years ago how many people honestly thought digital imaging technology would be as advanced as it is today?

Surely if smaller sensors are achieving consistant and meaningful improvements it has to be a good thing for all photographers? Wouldn't you like it if you could eventually get present day Full Frame quality in a smaller, lighter and more cost effective package? I don't understand what I am perceiving to be a general prejudice against Four Thirds with what almost seems to be a hope that it doesn't ever improve enough to rival Full Frame. A similar attitude existed around five years ago amongst some film users who begrudged the progress digital photography was making and sincerely hoped that silver halide technology would always remain superior. At least that battle seems to be over!

Please correct me if I am wrong.

 

[Mr_Sukebe]

This is all getting rather theoretical, but there must come a point when meaningful improvements are no longer possible. Once you are able to shoot a black cat in a coal cellar at midnight on a moonless night with results as good as if it was at ISO 100 on a bright day where else do you go? Once you can make poster size prints with detail as fine as the human eye is able to discern how do you improve on that?

Fanciful? Maybe. But all it needs is for some bright engineer to develop a way to bypass the sensor signal amplification connundrum and you have a whole new technology. Ten years ago how many people honestly thought digital imaging technology would be as advanced as it is today?

Surely if smaller sensors are achieving consistant and meaningful improvements it has to be a good thing for all photographers? Wouldn't you like it if you could eventually get present day Full Frame quality in a smaller, lighter and more cost effective package? I don't understand what I am perceiving to be a general prejudice against Four Thirds with what almost seems to be a hope that it doesn't ever improve enough to rival Full Frame. A similar attitude existed around five years ago amongst some film users who begrudged the progress digital photography was making and sincerely hoped that silver halide technology would always remain superior. At least that battle seems to be over!

Please correct me if I am wrong.

I think you've missed where we're coming from.
I'm sure that a lot of people (myself included) would love to have the option of going with an ultra compact SLR camera, which is what a micro 4/3rds camera is.
However, present day technical limitations are that full frame has advantages in noise levels and will always have an advantage in their ability to have a shallow depth of field. To some, that's not important, but it is to me.

I'd still like to ask you why no alarm bells are ringing over your inference that the D300 is better than the D700 in noise levels.

 

[goldenlight]

I think you've missed where we're coming from.
I'm sure that a lot of people (myself included) would love to have the option of going with an ultra compact SLR camera, which is what a micro 4/3rds camera is.
However, present day technical limitations are that full frame has advantages in noise levels and will always have an advantage in their ability to have a shallow depth of field. To some, that's not important, but it is to me.

I'd still like to ask you why no alarm bells are ringing over your inference that the D300 is better than the D700 in noise levels.

How little DOF do you need?

http://www.talkphotography.co.uk/gallery/showphoto.php?photo=86976&cat=500

I think I've already agreed that "present day technical limitations are that full frame has advantages in noise levels" but I think you've missed my point about current progress and future developments and find your lack of enthusiasm for progress quite puzzling.

I agree that I haven't yet said "alarm bells are ringing" but I have said "I have repeatedly stated my doubts about the true value of the comparisons" and " It is important to acknowledge that it is just one set of results from one source which can hardly be regarded as definitive and, as I said, I was rather surprised at what I was seeing." In case this is not plain and unequivical enough I will now say Alarm bells are ringing.

However that does not detract from the fact that this resource is out there, easily accessed on the Internet, and as such is a legitimate subject for discussion. Which is surely what we are doing.

 

[Mr_Sukebe]

Assuming you're happy to discuss it, then how about actually running the tests yourself. i.e. take a memory card to your local Jessops and take some shots at different ISOs with a D700. Then dig out your E3 and duplicate as closely as possible your shots and see what the differences are.

Ref DOF, this was something I asked about recently when looking at a GF1. From what I can see, the DOF would be roughly double what I can achieve now with my 5D. I value shallow DOF as an artistic option, and don't want to trade that away. I know that it's still possible to achieve a shallow DOF with a 4/3rds camera, but clearly it's not as easy.

I have no problems at all with change and progress. I'm just a little surprised that you seem to believe that 4/3rd technology will necessarily close the gap. The physical limitations will always exist, implying that there will always be a gap.
In however the extremely unlikely scenario that manufacturers develop a noise free sensor, then why stop at using a 4/3rds size sensor, why not use something MUCH smaller, e.g. the equivalent of what's in say a mobile phone or cheap compact?

 

[boyfalldown]

How little DOF do you need?

http://www.talkphotography.co.uk/gallery/showphoto.php?photo=86976&cat=500

I really value being able to shoot at very shallow depth of field - and while thats a great photo - the depth of field isn't really shallow enough. The buildings (covent garden?) in the background are a long way behind him and still giving more than just an impression. While I see where you'd like to go with this, basic physics says that a 4/3rd sensor will never be as 'good' as a full frame one, just as medium format or bigger will have qualities full frame never can

 

[goldenlight]

Assuming you're happy to discuss it, then how about actually running the tests yourself. i.e. take a memory card to your local Jessops and take some shots at different ISOs with a D700. Then dig out your E3 and duplicate as closely as possible your shots and see what the differences are.

Oh, I'm sure I can anticipate the results, especially at high ISO. The E3 is hardly state of the art in this respect even in 4/3 sensor terms. I think I've already made it clear that I acknowledge that. There are other things about the E3 that commend it to my possession.

Ref DOF, this was something I asked about recently when looking at a GF1. From what I can see, the DOF would be roughly double what I can achieve now with my 5D. I value shallow DOF as an artistic option, and don't want to trade that away. I know that it's still possible to achieve a shallow DOF with a 4/3rds camera, but clearly it's not as easy.

That's fair comment.

I have no problems at all with change and progress. I'm just a little surprised that you seem to believe that 4/3rd technology will necessarily close the gap. The physical limitations will always exist, implying that there will always be a gap.

I feel that I've already answered that adequately and in detail in a previous post on this thread.

In however the extremely unlikely scenario that manufacturers develop a noise free sensor, then why stop at using a 4/3rds size sensor, why not use something MUCH smaller, e.g. the equivalent of what's in say a mobile phone or cheap compact?

I agree, why not indeed? As long as it doesn't get to the point that the hardware is just too small to be manageable I say "Bring it on!"

 

[goldenlight]

I really value being able to shoot at very shallow depth of field - and while thats a great photo - the depth of field isn't really shallow enough. The buildings (covent garden?) in the background are a long way behind him and still giving more than just an impression. While I see where you'd like to go with this, basic physics says that a 4/3rd sensor will never be as 'good' as a full frame one, just as medium format or bigger will have qualities full frame never can

Is this any better?

http://www.talkphotography.co.uk/gallery/showphoto.php?photo=87014&cat=500

Kestrel shot on an E510 with manual 300mm f4.5 Nikkor wide open. Note that, for those with a deep pocket, Olympus do a 300mm f2.8 (that's 600mm equivalent)!

 

[grum]

I think you've missed where we're coming from.
I'm sure that a lot of people (myself included) would love to have the option of going with an ultra compact SLR camera, which is what a micro 4/3rds camera is.
However, present day technical limitations are that full frame has advantages in noise levels and will always have an advantage in their ability to have a shallow depth of field. To some, that's not important, but it is to me.

You can get pretty shallow DOF on M4/3

 

[boyfalldown]

You can get pretty shallow DOF on M4/3

you can, and that like the kestrel above they are both fantastic shots, nobobies saying anything else, but as above physics means you'll have more options for this with the bigger sensor

 

[goldenlight]

you can, and that like the kestrel above they are both fantastic shots, nobobies saying anything else, but as above physics means you'll have more options for this with the bigger sensor

I agree, but the point is that a Four Thirds camera can do shallow DOF more than adequately, if not quite to the same degree as Full Frame. I am not saying that 4/3 is better than FF, I am saying that it is catching up (in all respects other than DOF, of course). Whether or not a small sensor will ever totally match (for all practical purposes) a larger one is at present a matter of pure speculation on which there are a wide range of opinions, none of them right and none of them wrong. What I do confidentally predict is that during the comming years the gap will continue to close, which must surely be a good thing whatever manufacturer you may lean towards.

 

[Mr_Sukebe]

You can get pretty shallow DOF on M4/3

I never said you couldn't. Re-read my comment. If you'd taken the same shot with an FF DSLR, the DOF would have been half that.

 

[grum]

I never said you couldn't. Re-read my comment. If you'd taken the same shot with an FF DSLR, the DOF would have been half that.

Fair enough - I wonder how many pics there are in your library that take advantage of that though? Not saying it wouldn't be a nice option, but I don't know if I would use it that much.

 

[goldenlight]

Hi Graham,

That's impressive DOF - or lack of it! How did you incorporate a full size image into the thread on this forum? The best I could manage were links to the gallery.

 

[Mr_Sukebe]

Fair enough - I wonder how many pics there are in your library that take advantage of that though? Not saying it wouldn't be a nice option, but I don't know if I would use it that much.

I bought my 85mm F1.8 for use as a potrait lens. It's shallow DOF is part of it's repotaire, he's an example snapshot:

So I could argue that most shots taken with said lens will take advantage of the DOF afforded by FF.

 

[grum]

We could play this game all day

goldenlight - I uploaded to photobucket and then linked to here with the '

 

[dalex_257]

...just as medium format or bigger will have qualities full frame never can

Re MF vs. FF in terms of DOF - this may be not as simple. This post on DPReview actually goes through the choices and it seems that where shallow DOF is concerned the FF is the best affordable choice of sensor size (at least for now).

 

[dalex_257]

A similar attitude existed around five years ago amongst some film users who begrudged the progress digital photography was making and sincerely hoped that silver halide technology would always remain superior. At least that battle seems to be over!

Please correct me if I am wrong.

You are wrong that battle if far from over. The view cameras and large format still cannot be beaten by any digital technology in terms of resolution. And I don't think the day when they will be is closer as larger sensor technology has not improved over the last decade significantly. There are technological as well as physical limits that makes it extremely difficult to make truly large format sensor even a true medium format one (and this is one of the reasons why we don't have 6x7 MF sensors only slightly larger ones than FF).

 

[goldenlight]

You are wrong that battle if far from over. The view cameras and large format still cannot be beaten by any digital technology in terms of resolution. And I don't think the day when they will be is closer as larger sensor technology has not improved over the last decade significantly. There are technological as well as physical limits that makes it extremely difficult to make truly large format sensor even a true medium format one (and this is one of the reasons why we don't have 6x7 MF sensors only slightly larger ones than FF).

I didn't say that digital was superior to film technology! Joe Cornish still uses LF and that is proof enough for me that film still has advantages in certain applications. What I said was the battle seems to be over. Film users seem to have long accepted that digital is a viable alternative and is indeed now the medium of the masses. There no longer seems to be the resentment, suspicion and predjudice against digital. Film and digital can and do now sit happily side by side in the options available to photographers.

It appears that the same resentment, suspicion and prejudice does however live on in the minds of FF advocates who seem appalled at any suggestion that Four Thirds technology is catching up and is now every bit as capable of serious, quality photography as the FF and APS-C formats. Why is this? Please tell me. :shrug:

 

[alexsmith]

I love the way so many users still think bigger is better with sensors. yet since digital , every month or so it seems more information is being able to be stored on a smaller object.
My usb stick has 8gb, cost a tenner from argos yet it stores more than my first laptop. Does it "compress it inaccurately" . Of course it doesn't.
When Nikon made there sensor, it was probably the size it needed to be. now if they started from scratch it would probably be smaller. Every other single thing in the digital world goes this way.

al

 

[boyfalldown]

I love the way so many users still think bigger is better with sensors. yet since digital , every month or so it seems more information is being able to be stored on a smaller object.
My usb stick has 8gb, cost a tenner from argos yet it stores more than my first laptop. Does it "compress it inaccurately" . Of course it doesn't.
When Nikon made there sensor, it was probably the size it needed to be. now if they started from scratch it would probably be smaller. Every other single thing in the digital world goes this way.

al

I don't think 'bigger is better' with sensors is an arguement that holds water on its own, but look at it another way, bigger brings more options with photography, and thats true if you shoot on LF, 35mm Film or 4/3rds digital. - Your usb stick example doesn't hold water, simply because there is no rule in phyisics that will, by default, change the way it behaves as it shrinks

Hugh

 

[alexsmith]

I don't think 'bigger is better' with sensors is an arguement that holds water on its own, but look at it another way, bigger brings more options with photography, and thats true if you shoot on LF, 35mm Film or 4/3rds digital. - Your usb stick example doesn't hold water, simply because there is no rule in phyisics that will, by default, change the way it behaves as it shrinks

Hugh

actually it entirely holds water.. nikon could once "only" squeeze 4 megapixels out of there sensor, now its what 12? 20? and with no loss in quality. so bearing in mind the world doesn't want 50 mp cameras to make 3 acre prints... if punters wanted d40 performance (ie pretty dam good) nikon doesn't need a sensor the size it is. If the big two sarted from scratch they might not start with the sensor the size they have, its just that that was what was needed at the time.
the only moving goal post here is that the smaller sensor technology will always be playing catchup to pixel peepers who are impressed with 12 mp as apposed to 6mp..
If ever you needed proof of this, I found a catalogue at work from about 8 years ago of digi cameras wow most of the manufacturers still felt the need to justify digi over film and 2mp was hot stuff

By "catchup" I mean my old nikon D40 does not perform as well as my olympus e-520 . And thats not because it has twice as many pixels so the 4/3rds caught up, but by now the new nikon (entry model) will probably outperform the oly in low light and every other department too but then the next 4/3rds.... and so on.. its a pointless argument. My other hobby used to be target shooting and some time ago it reached the point where you could say EVERY gun could outshoot the guy behind it and the same will be true of dslr's if it isn't already....

 

[boyfalldown]

actually it entirely holds water.. nikon could once "only" squeeze 4 megapixels out of there sensor, now its what 12? 20? and with no loss in quality. so bearing in mind the world doesn't want 50 mp cameras to make 3 acre prints... if punters wanted d40 performance (ie pretty dam good) nikon doesn't need a sensor the size it is. If the big two sarted from scratch they might not start with the sensor the size they have, its just that that was what was needed at the time.
the only moving goal post here is that the smaller sensor technology will always be playing catchup to pixel peepers who are impressed with 12 mp as apposed to 6mp..
If ever you needed proof of this, I found a catalogue at work from about 8 years ago of digi cameras wow most of the manufacturers still felt the need to justify digi over film and 2mp was hot stuff

no it doesn't hold water - although you're right in that you can now fit more information onto that sensor (more MP) the basic optical rules that govern photography are affected by a reduction in size. Simply reducing the size of a usb stick won't affect the way it behaves in the same way