Just comparing the two on paper, I would have bet the Sigma to have more detail at the pixel level due to the lack of an AA filter and the better lens.
Just curious that if both pics were taken from the same spot, and the same framing was achieved despite the focal length difference. Are we looking at crops?
CVickery wrote:
I understand what you were saying and was not intending to imply this was or should have been a more precise comparison. I('m sorry if it came across that way.
What I was trying to point out is that at 100% it is pretty much a given that a Foveon sensor will outperform a Bayer sensor for detail. If you look back there were similar comparisons with the DP1... the nature of differences between the sensors leads to this result. In addition, the effective focal length differences put the 5DIII at a further disadvantage...so the results aren't surprising.
No worries, I wasn't offended at all. I agree with your points. While it shouldn't be surprising, I truly was surprised at how prominent the detail differences were firsthand. It's an interesting $800 camera.
jfwoodman wrote:
...using same technique from same spots, and even given the focal length differences (31mm to effective 45mm), the results were very consistent: The Sigma always provides more detail, well beyond what a traditional dslr sensor delivers...
I would not make much conclusions of 5DmkIII crops compared to Sigma crops. In real life the pixel level quality doesn't matter that much - try resizing (with Lanczos or some other proper method, which doesn't make images soft like bicubic) your 5DmkII 50mm f/5.6 (almost any 50mm fixed lens is good enough at f/5.6) images to 14.6Mpix (=same as DP2M), and compare them and suddenly there is not much difference in the image detail.
I'm not saying DP2M isn't good camera, just that your testing method won't give any useful information. I'm actually considering to buy one when I next time visit B&H PhotoVideo unless they have Zeiss 2/135 at stock. It's sad that DP1M lens isn't as good as DP2M and DP3M, because DP1M and DP3M would make great small versatile camera system for $1500.
Jochenb wrote:
The resolution of the Merrills is fantastic, but my problem with foveon is it's colors. You often get this gray/greenish tint. I've owned the DP2 in the past and I'm still seeing it in a lot of Merrill photos now.
I'm not sure how much of this is inherent to the sensor and how much is simply the processing.
For detail, this looks like no contest. True, it's an unscientific comparison, I'm sure the 5d3 could do a lot better than what's shown here in a more controlled test. However, it's pretty amazing considering that the DP2M costs $4000 less than the 5d3 and 16-35.
itai195 wrote:
I'm not sure how much of this is inherent to the sensor and how much is simply the processing.
For detail, this looks like no contest. True, it's an unscientific comparison, I'm sure the 5d3 could do a lot better than what's shown here in a more controlled test. However, it's pretty amazing considering that the DP2M costs $4000 less than the 5d3 and 16-35.
True, but it's really an 'apples to oranges' comparison. You don't view an image @ 100% ...well most people don't
If you print the same framed image from both cameras to the same size the Sigma will end up being magnified to reach the same size....it is in this sense a 15mp sensor vs a 22mp Canon. In this scenerio much of the advantage we see here is lost. If I recall when the DP1 was introduced (which was a 4.7mp sensor) it was estimated that it equated to ~8-10mp bayer. Extending this reasoning the Merrill should equate to ~30mp in practical use, still better than the Canon, but not as much as it would seem from the above images. Naturally, this argument neglects many other factors and a proper test would be needed to estimate the size of a Bayer sensor that would be equivalent.
I agree, I'm sure in more practical viewing situations and with a more controlled test the results would be closer. But this is still an $800 compact camera we're talking about. Foveon, at least from a technological standpoint, clearly has merit.
itai195 wrote:
I agree, I'm sure in more practical viewing situations and with a more controlled test the results would be closer. But this is still an $800 compact camera we're talking about. Foveon, at least from a technological standpoint, clearly has merit.
While this is true, don't forget that the Sigma is the closest thing to a one trick pony there is in the photographic universe (no interchangeable lenses, no high ISO worth speaking of, not so great skin tones, worse colour for product photography, no fast AF, no high FPS shooting, no extensive flash system, no true macro, no narrow depth of field, no zoom, no good video capability, and a million other things I have forgotten), whereas the Canon is a full-fledged system camera. There are loads and loads of reasons why the Canon combo is more expensive. The sharpness of the Sigma is impressive, but that and the size are the only advantages. Beyond that, the Canon wipes the floor with it.
No doubt the Sigma DP Foveon camera are excellent cameras in many respects. The Foveon sensors bring a certain clarity to an image that is unique. I owned a DP1 for a while and loved this aspect of the camera. However, other issues led me to explore different options, and the DP1 is long gone. I am watching the Merrill series with interest and at $800 it presents a very interesting choice.
p.2 #10 · Sigma DP2M v Canon 5D3 - resolving power
Don't forget Carsten that Sigma does have the SD1 Merrill that at least partially addresses the issues you raise with the DPxM series and probably is a more appropriate camera to compare with the Canon 5DIII. Perhaps the more appropriate comparison for the DP is the Canon M Camera
p.2 #11 · Sigma DP2M v Canon 5D3 - resolving power
The sigma looks great, but there is also a huge amount of false detail. Look at the first crop from the waterfall. There are like "bands" of sharp and unsharp fields where detail seem to be enhanced when it's above a certain contrast treshold, but stays unsharp if it is below the treshold.
p.2 #12 · Sigma DP2M v Canon 5D3 - resolving power
carstenw wrote:
Price comparisons make zero sense here.
Horses for courses.
They make sense within the context of the comparison presented here. Obviously the Sigma would be worth approximately $0 in many other comparisons, for example if the OP was shooting sports.
p.2 #13 · Sigma DP2M v Canon 5D3 - resolving power
itai195 wrote:
They make sense within the context of the comparison presented here. Obviously the Sigma would be worth approximately $0 in many other comparisons, for example if the OP was shooting sports.
Well, if you are just looking for a small camera which makes sharp images, then the 5DIII makes no sense. The only time it makes sense to compare at all is when you already own both and are a bit bored or curious.
p.2 #14 · Sigma DP2M v Canon 5D3 - resolving power
CVickery wrote:
Don't forget Carsten that Sigma does have the SD1 Merrill that at least partially addresses the issues you raise with the DPxM series and probably is a more appropriate camera to compare with the Canon 5DIII. Perhaps the more appropriate comparison for the DP is the Canon M Camera
p.2 #15 · Sigma DP2M v Canon 5D3 - resolving power
Sigma is good for somebody after that crisp look. I never feel it is represent the real world.
The file from good 5DII, 5DIII photographer still looks more pleasing to me. But I agree those require more attention to bring all the detail out without too much. Sharpening itself can be a book.
I fully agree IQ is just 'A' part of system requirement as tool. (let assume it has decent IQ) It is all about priority. I won't buy it even at $400, so $800 is not a good value to me at all. It is just a market price set by demand and supply, not because of its $800 make $3000 5DIII a poorer value. ot itself a good value just because of it has crisp file to the extreme corner.
p.2 #16 · Sigma DP2M v Canon 5D3 - resolving power
carstenw wrote:
While this is true, don't forget that the Sigma is the closest thing to a one trick pony there is in the photographic universe (no interchangeable lenses, no high ISO worth speaking of, not so great skin tones, worse colour for product photography, no fast AF, no high FPS shooting, no extensive flash system, no true macro, no narrow depth of field, no zoom, no good video capability, and a million other things I have forgotten), whereas the Canon is a full-fledged system camera. There are loads and loads of reasons why the Canon combo is more expensive. The sharpness of the Sigma is impressive, but that and the size are the only advantages. Beyond that, the Canon wipes the floor with it.
Relative price (to the 5DIII) doesn't matter, but the absolute price of the DP2M does matter to me, because $800 is likely less than it would cost to buy a lens for your Canon FF for this type of shooting. I'm actually considering this right now. I was thinking about purchasing a relatively wide lens to use for stopped-down landscape shooting with my Sony a850, something like the ZA 24/2, or maybe adapting the C/Y 28/2.8. But another way I could go would be to just buy a DP1M to complement my DSLR. It would be about the same size in my bag as one of those lenses, and I'd only really use that lens for the sort of shooting that seems to be the DP1M's strength.
p.2 #18 · Sigma DP2M v Canon 5D3 - resolving power
theSuede
Exactly what I meant by my comment. If I add better demosaicing and some good sharpening it can be improved even further. Detail is there, it just needs proper PP.
Main difference between any Sigma and Canon is that with Canon files (5DMK3 lets say) you can do whatever you wish. With Sigma, what you got is pretty much everything you have.
p.2 #19 · Sigma DP2M v Canon 5D3 - resolving power
zhangyue wrote:
Sigma is good for somebody after that crisp look. I never feel it is represent the real world.
The file from good 5DII, 5DIII photographer still looks more pleasing to me. But I agree those require more attention to bring all the detail out without too much. Sharpening itself can be a book.
I fully agree IQ is just 'A' part of system requirement as tool. (let assume it has decent IQ) It is all about priority. I won't buy it even at $400, so $800 is not a good value to me at all. It is just a market price set by demand and supply, not because of its $800 make $3000 5DIII a poorer value. ot itself a good value just because of it has crisp file to the extreme corner.
Again, as someone said above, horses for courses. The 5D3 is a great camera. I use it for everything from wildlife to portraits to landscapes. I have no plans to replace my 5D3 with the SigmaDP2M. But, so far, I'm impressed with many aspects of the Sigma and find it an interesting new tool to learn. I think it could be a very capable travel camera to bring on wildlife shooting trips to have on my hip and complement my dslr and supertele.
And while I focused on resolution / detail in this post, it is not just a high-detail, one trick pony. But I like how many of the files look. Additionally, while it is amazingly slow to process images, eats batteries, etc.., I like it's menu system, it has a high quality lens, and I find it's AF to be quite good. If I can figure out how to master the colors, I will like it even more.
Playing around with waterfalls in shade during mid-day lull
A I think it actually does fairly well with more narrow DOF as you open it up
p.2 #20 · Sigma DP2M v Canon 5D3 - resolving power
I do feel that I have to add that the sample scene is one of the cases where the Sigma really excels...
It's quite easy to live with some color discrepancies and slow handling, in cases when image detail and compactness (not to mention difference in price!) has to take precedence. The detail result from such a compact camera at that price-point is truly mindblowing.
What bothers me with the DP-x series is the color in more difficult light, color when it's important (product shots) and high-ISO performance. If none of those points has any weight for the main part of your usage scenarios, they're a steal. Optical quality is really good to, if quite limited in flexibility... (why couldn't they make the DP-series with interchangeable lenses? That it would be a totally new lens mount wouldn't bother me squat, they're specific use cameras anyway... are the microlenses on the sensors on the models with different focal lengths really that different?)
Neither is real..
