Hmm, my girlfriend is having very similar problems with her D200. She's been going crazy checking lenses and focus and cleaning her sensor. It will be interesting to see how many others chime in on this. I'll be following this thread.
Fairly certain (this is all speculation) that the CCD v CMOS might be the culprit here?
In my experience as of late, I'm finding that better sensors need great glass. I have a 17-40L that I've had for years and worked like a charm on the 5D, but on the Mark II it's a little disappointing... so glass can be an obvious factor - the photosite 'microlenses' require better glass - allegedly
You make a good point that the shooting parameters are a bit different (the compositions aren't exactly the same either in the test!). As I explained in my blog, this is not meant to be a truly rigorous test; too many variables aren't controllable in the field with a test such as this. With variance this great, I don't think exact shutter speeds would have changed the muddiness and poor contrast--the final differences between the cameras are too extreme. And for me, this test more accurately reflects the conditions I shoot in (eg. aperture priority) and so are instructive.
I visited your website and see you shoot with a D200; I would welcome you posting results from a more controlled experiment on this thread if you are interested. Perhaps it will shed new light on this problem I've encountered. I also note that you shoot avian subjects sometimes with fill flash--these might be quite different subjects than what I shoot. If you do test the two cameras, I encourage you to shoot vegetation in natural light so we can look at that micro contrast.
To answer each of your points:
1. Actually, the ISO on the D300 is set at the camera's lowest default of ISO 200. We thought it fairer to compare the two cameras on their manufacturer's optimal ISO defaults (if anything, shouldn't the D200 have a performance advantage at a lower ISO?)
2. You are absolutely correct that the images are both technically 'processed' otherwise they could not be displayed from RAW. All were converted with identical parameters from Camera RAW to jpg format.
3. The slight difference in shutter speed is probably more attributable to the fact that one image, the D300 one, was shot four minutes later. There is a difference in intensity of light that might have contributed to the faster shutter speed (along with the ISO speed).
Yes, I shoot mostly with a D200. I also have a D70 that is now a backup, and also used for "piny and shoot" duties. I shoot in Aperture preferred mode which I started using on my first Nikon with auto exposure back around the time of the Nikkormat EL. I almost never use fill flash, preferring natural light, but resort to fill flash when needed.
I don't have a D300 (wish I did) so I can't make any D200 vs D300 comparisons.
For any meaningful comparison test you must control all variables and only change the one variable you are trying to evaluate. In this case that is the camera body. So, you have to have exactly the same subject in exactly the same light at the vary least; not "aren't exactly the same" framing or with enough time different to allow the light to change.
I understand the choice of shooting each camera at the lowest (native) ISO setting rather than using the same ISO on both. But when I normalized the exposure of your D200 shot you gave that exposure 1/3rd less light than you did with the D300. If you underexposed by 1/3rd stop that could partially explain the darker muddier result, so it is not valid to conclude that the D200 images are in general more muddy than those of the D300.
I have a 17-40L that I've had for years and worked like a charm on the 5D, but on the Mark II it's a little disappointing
Samantha,SAME THING happened with me as with Christopher. My 70-200mm/2.8 is more gorgeous on the 5D2, but my 17-40mm went south on the 5D2. I sold it and bought the 16-35mm/2.8 (with teeth clenched! $$$!), but WOW was it worth it, much better. These mega-sensors do push the limit of glass! Darwin did a comparison of the 17-40 and the 16-35 a while back and the differences weren't huge. The 5D2 changed all that, and it looks like the D300 did also.
Tell Darwin that when he's done with the commodes, maybe he can do another excellent comparison article on those Canon lenses with the 5D Mark II... maybe Christopher and I are looney and aren't seeing the image data right.
BTW, I thought your test & blog were very thorough on the Nikon side.
Discussion
10 repliesBut we found out the truth....
Check out my blog entry for details.
In my experience as of late, I'm finding that better sensors need great glass. I have a 17-40L that I've had for years and worked like a charm on the 5D, but on the Mark II it's a little disappointing... so glass can be an obvious factor - the photosite 'microlenses' require better glass - allegedly
ye, me too.
From the EXIF in the images posted on the blog:
D200: ISO 100, 1/10th second at f/11
D300: ISO 300: 1/15th second at f/11
So, the ISO was different. And if you adjust the ISO values you get different exposures (1/20th vs. 1/15th at f/11 for ISO 200).
The posted images are jpegs, not raw, so they've been converted in some manner. How? Were the conversion parameters identical for each conversion?
Darwin
You make a good point that the shooting parameters are a bit different (the compositions aren't exactly the same either in the test!). As I explained in my blog, this is not meant to be a truly rigorous test; too many variables aren't controllable in the field with a test such as this. With variance this great, I don't think exact shutter speeds would have changed the muddiness and poor contrast--the final differences between the cameras are too extreme. And for me, this test more accurately reflects the conditions I shoot in (eg. aperture priority) and so are instructive.
I visited your website and see you shoot with a D200; I would welcome you posting results from a more controlled experiment on this thread if you are interested. Perhaps it will shed new light on this problem I've encountered. I also note that you shoot avian subjects sometimes with fill flash--these might be quite different subjects than what I shoot. If you do test the two cameras, I encourage you to shoot vegetation in natural light so we can look at that micro contrast.
To answer each of your points:
1. Actually, the ISO on the D300 is set at the camera's lowest default of ISO 200. We thought it fairer to compare the two cameras on their manufacturer's optimal ISO defaults (if anything, shouldn't the D200 have a performance advantage at a lower ISO?)
2. You are absolutely correct that the images are both technically 'processed' otherwise they could not be displayed from RAW. All were converted with identical parameters from Camera RAW to jpg format.
3. The slight difference in shutter speed is probably more attributable to the fact that one image, the D300 one, was shot four minutes later. There is a difference in intensity of light that might have contributed to the faster shutter speed (along with the ISO speed).
I don't have a D300 (wish I did) so I can't make any D200 vs D300 comparisons.
For any meaningful comparison test you must control all variables and only change the one variable you are trying to evaluate. In this case that is the camera body. So, you have to have exactly the same subject in exactly the same light at the vary least; not "aren't exactly the same" framing or with enough time different to allow the light to change.
I understand the choice of shooting each camera at the lowest (native) ISO setting rather than using the same ISO on both. But when I normalized the exposure of your D200 shot you gave that exposure 1/3rd less light than you did with the D300. If you underexposed by 1/3rd stop that could partially explain the darker muddier result, so it is not valid to conclude that the D200 images are in general more muddy than those of the D300.
Samantha,SAME THING happened with me as with Christopher. My 70-200mm/2.8 is more gorgeous on the 5D2, but my 17-40mm went south on the 5D2. I sold it and bought the 16-35mm/2.8 (with teeth clenched! $$$!), but WOW was it worth it, much better. These mega-sensors do push the limit of glass! Darwin did a comparison of the 17-40 and the 16-35 a while back and the differences weren't huge. The 5D2 changed all that, and it looks like the D300 did also.
Tell Darwin that when he's done with the commodes, maybe he can do another excellent comparison article on those Canon lenses with the 5D Mark II... maybe Christopher and I are looney and aren't seeing the image data right.
BTW, I thought your test & blog were very thorough on the Nikon side.