Richard, for the crop, could you try upressing the same way as you did for the 5D? I expect it will look better that way, although it wouldn't catch up with the 5D.
KJbruin wrote:
(snip) ... 100% uninterpolated files that have been postprocess exactly the same.
It's always difficult to compare different types of cameras. Personally I don't care what has been done to a file, for me it's the end result that counts. To use the same PP on fundamentally different cameras (sensor size, AA filter, pixel count, IR filter) is of no interest and certainly something that the clients of those who shoot commercially would never ever inquire about.
nice photos. Nice subtle pastel lighting. But that blue stuff at the top of the pictures, I didn't think you had that in Old Blighty?
As for the 5D stuff, I am selling the 5D , an old Zeiss Flektogon 20mm (great condition, but needs lubrication), Leica Summicron 50 (current version), Leica Summilux 80, Elmarit 135 (newest one), Elmarit 180 (newest non-Apo one), and my 17-40L. I have decided on prices for some items but not all. PM me if you see anything interesting. As far as prices, I aim for them to be fair for both sides, but not some kind of superbargain. I can't afford that All lenses are in very good to great shape, some near perfect.
Edit: Oops, nearly forgot: 3 Extreme III cards (I might keep one), Camdapter+handstrap, Brightscreen focusing screen, Ee-S focusing screen, CF card wallet (more like a tank!), extra battery, a handful of Leica-R-to-EOS adapters, and so on... bits and pieces which won't fit the M8....Show more →
I also forgot: I am selling a Brightscreen 5D viewfinder magnifier. It is a really nice piece, which I would have loved to use, were I keeping the 5D, but to be honest, it clarifies more than it magnifies. Nonetheless, I am selling that too.
If anyone is interested in anything, please email me or PM me. I don't want this to spill over into this forum.
Jack Flesher wrote:
Guy Mancuso, Robert Stevens, Jack Flesher, Mark Kay, and I are up at Yosemite for the weekend. Lots of M8 shooting going on. We're having a blast.
quote]niklasl wrote: KJbruin wrote:
(snip) ... 100% uninterpolated files that have been postprocess exactly the same.
It's always difficult to compare different types of cameras. Personally I don't care what has been done to a file, for me it's the end result that counts. To use the same PP on fundamentally different cameras (sensor size, AA filter, pixel count, IR filter) is of no interest and certainly something that the clients of those who shoot commercially would never ever inquire about.
So because of the lack of an AA filter the M8 files cannot benefit from a bit of USM? Would the same interpolation have yielded different results? These are questions that would help us understand the image quality better from these two excellent cameras. Your point is taken but my question is more whether using the tested workflow the true potential of the M8 was shown.
Anyone with an M8 and has seen firsthand the files that this camera can produce will know that the images being used in this comparison fail to represent what the M8 is capable of. My own comparison tests between the M8 and the excellent Canon 1DS showed that the M8 can hold its own in resolution and really shines in DR. For reference the 1DS files are not too far off the 16 megapixel 1DS2 and too see this type of difference between the 10 and 13 megapixel files of the M8 & 5D in my opinion calls into question the testing methodology (once again). At any rate, as others have mentioned before it seems that this tester has found exactly what he was determined to find from the start. So I really hope that these tests though potentially flawed make all 5D owners as comfortable and happy with their camera as I am with mine.
Guy Mancuso wrote:
Crap once again a flawed test , would you please go back and look at your two images your focus on the M8 is NOT on the rear window like the 5D . The M8 focus point is CLEARLY on the front case . Look at the lettering on the right side that is were you focused. Not at the same spot as the Canon so how can you compare a missed focus comparision. LOOK AT THE FILES The focus point is clearly NOT on the front case but looks a little nearer than the 5D but not enough to throw the the test IMO as you have to deal with 3d objects in real world shooting and focus points are a choice we have to make. Additionaly focus planes are rarley flat except with specialist copy lenses and if you look at the edges of the frame it's possible to find areas where the focus planes intersect - LOOK AT THE FILES. The biggest difference is not the sharpness but the awful edge artefacts on the M8 images compared to the 5D. All this proves however is that the M8 does not produce very good jpegs from it's internal processor. All tests are flawed and therefore crap - both the M8 and 5D can produce fine images and either one in the hands of a competent photographer will deliver the goods.
KJbruin - I'm not sure whether you misunderstood me or not. What I'm saying is that when testing a camera against another it is my opinion that what is necessary for each camera model in PP should be done in order to arrive at an optimal end result, which might then be compared. That means that USM, to take one example, must be applied according to what is needed for each camera, and this certainly does not exclude the M8, if the files can benefit from that. After all, that's what we do in real world shooting. Since the M8 and the 5D are so different, one of them would suffer from applying the same kind and amount of USM and none would be the wiser after such a "test" and it would have nothing to do with actually using the camera.
Personally I have no reason to bash one camera or the other. I really enjoy reading about the M8 and I have seen quite a few impressive photos on the web. I regard the M8 vs 5D tests as interesting but nothing more and see no reason to be upset about them. Errors are pointed out and hopefully these will be corrected in future comparisons.
I understood your point as it is not a complicated one. But we may be dicussing details of the test that you are not aware of. Perhaps you can reread the methodology section of his comparison. He applied no sharpening to the M8 file.
niklasl wrote:
KJbruin - I'm not sure whether you misunderstood me or not. What I'm saying is that when testing a camera against another it is my opinion that what is necessary for each camera model in PP should be done in order to arrive at an optimal end result, which might then be compared. That means that USM, to take one example, must be applied according to what is needed for each camera, and this certainly does not exclude the M8, if the files can benefit from that. After all, that's what we do in real world shooting. Since the M8 and the 5D are so different, one of them would suffer from applying the same kind and amount of USM and none would be the wiser after such a "test" and it would have nothing to do with actually using the camera.
Personally I have no reason to bash one camera or the other. I really enjoy reading about the M8 and I have seen quite a few impressive photos on the web. I regard the M8 vs 5D tests as interesting but nothing more and see no reason to be upset about them. Errors are pointed out and hopefully these will be corrected in future comparisons....Show more →
shirozina wrote:
The biggest difference is not the sharpness but the awful edge artefacts on the M8 images compared to the 5D. All this proves however is that the M8 does not produce very good jpegs from it's internal processor. All tests are flawed and therefore crap - both the M8 and 5D can produce fine images and either one in the hands of a competent photographer will deliver the goods.
I think the edge artifacts are caused by the method of upressing. The 5D images were done with a nice smooth filter, whereas the M8 was done by pixel-doubling. We really need to see both upressed with the same filter, to some non-multiplier size.
Thanks Carsten - I didn't read the bit about different interpolation methods. Using 'nearest neighbour' on the M8 is the culprit and a sure way of messing the image up.
I didn't realize the test was done with jpegs out of the cameras. That's not ideal, and using RAW would have allowed each camera to perform at their best. The M8 should perform better, but so should the 5D to some extent. Re-doing the test in RAW (and properly focused, as that might be an issue here) might shrink the difference between the 2 cameras somewhat, but I think it is unlikely that the conclusion will be reversed.
If I had an M8, and not just an 5D, I would do a comparison to see if the results would be repeated. If I can get my hands on one in the Washington, DC area (any volunteers?) I'll give it a try.
I am intending to do a similar test when I get my M8 back next week. I also doubt that the conclusion would be reversed, but I expect the resolution test to be much closer, and the M8 to have nicer pixel-level characteristics, when not upressed. I am not wedded to that conclusion though, in the sense that this is what I want to see. It is just what I expect, based on my experience with the 5D over the last year and nearly 10000 images, and my little experience with the M8. Lotus, if you are going to be in Berlin soon, we can test together
I doubt that the JPG choice really changes anything here. Neither the M8 nor the 5D do significant sharpening that I know of, and as long as the exposures are reasonably close, there should be little difference. Not much processing was done on these images. In general shooting, of course, this could change.
Lotus, isn't it Victor who is also in DC? Doesn't he have an M8 at this point?
I think Richard tests are overall fair except the upres methods. I don't understand the reasons to use different interpolation schemes. Contrary to what he said, the nearest neighbor method doesn't seem to benefit M8 file. It actually exaggerates the harshness of the file.
It makes good sense to apply different sharpness settings to factor in the the AA filter. My experience with M8 DNG files is that you do not want to apply ANY sharpening at all. If you do, the files start falling apart shortly with too harsh edges.
There seems to be confusion about resolution, detail and sharpness. Smaller pixel will result in higher resolution. Camera with more pixels will capture more details. And AA filter strength determines the sharpness. There is no complex technology or magic behind it.
There seems to be still confusion about resolution, detail and sharpness. Camera with more pixels will capture more details. There is no complex technology or magic behind it.
- but there is - without an AA filter and with sophistaicated image processing a lower pixel camera can compete with a higher pixels camera with a detail destroying AA filter - THIS IS THE WHOLE POINT OF THIS LONG AND DRAWN OUT DEBATE!
I suggest you play with an M8 for a little bit longer before making judgements that M8 files fall apart after sharpening. In my experience though less sharpening is needed a little bit of USM can go a long way with M8 files. And the confusion that you seem to suggest with my comments is not that USM improves or creates resolution or detail but that when applied appropriately can certainly reduce the appearance of softness and bring out the appearance of increased detail. This WILL bias any comparison test expecially when viewing at 100%.
Pondria wrote:
I think Richard tests are overall fair except the upres methods. I don't understand the reasons to use different interpolation schemes. Contrary to what he said, the nearest neighbor method doesn't seem to benefit M8 file. It actually exaggerates the harshness of the file.
It makes good sense to apply different sharpness settings to factor in the the AA filter. My experience with M8 DNG files is that you do not want to apply ANY sharpening at all. If you do, the files start falling apart shortly with too harsh edges.
There seems to be confusion about resolution, detail and sharpness. Smaller pixel will result in higher resolution. Camera with more pixels will capture more details. And AA filter strength determines the sharpness. There is no complex technology or magic behind it.
Edited by Pondria on Dec 09, 2006 at 12:46 PM GMT...Show more →
shirozina wrote:
- but there is - without an AA filter and with sophisticated image processing a lower pixel camera can compete with a higher pixels camera with a detail destroying AA filter - THIS IS THE WHOLE POINT OF THIS LONG AND DRAWN OUT DEBATE!
Sophisticated image processing ? I don't know what it is but it doesn't sound like something that I would like to have
AA-filter is a low-pass filter. It cuts off the high-frequency that your pixel-pitch ( Nyquist Frequency ) can't support. It doesn't "kill" the detail but softens the edges. It makes square edges rounded.
, an old Zeiss Flektogon 20mm (great condition, but needs lubrication), Leica Summicron 50 (current version), Leica Summilux 80, Elmarit 135 (newest one), Elmarit 180 (newest non-Apo one), and my 17-40L. I have decided on prices for some items but not all. PM me if you see anything interesting. As far as prices, I aim for them to be fair for both sides, but not