True. Like I said, even though the Olympus software said NR was off, I think there is a little NR going on somewhere. I had the same feeling--the noise was very similar to my Rebel XSi.
I have a few more shots up there now, at full size. One shows the macro capabilities of the kit lens, and how sharp it is stopped down (it's sharp wide open, too, but this is stopped down):
All of these are out of camera jpegs, with the camera at default settings. The jpeg engine is really quite good. To see the full-sized images, click on "All Sizes" above the image, then select "Original."
jhapeman wrote:
I have a few more shots up there now, at full size. One shows the macro capabilities of the kit lens, and how sharp it is stopped down (it's sharp wide open, too, but this is stopped down):
All of these are out of camera jpegs, with the camera at default settings. The jpeg engine is really quite good. To see the full-sized images, click on "All Sizes" above the image, then select "Original."
Jeff...Show more →
Regarding the ISO 100 shot:
Seems like you would actually get more detail below F5.6 because diffraction starts to kick in with the smaller sensor above 5.6 or so. Contrast looks good and makes it appear that things are actually sharper than they are. There is a surprising amount of rainbow color artifacts/noise for an ISO 100 shot. I think noise reduction would get rid of that without hurting the detail. That is almost certainly due to the weaker AA filter.
This is a very nice review J, thank you for posting.
I have two questions I would like to ask: how is viewing the LCD in low light for focusing, and how noisy is the shutter, say compared toa DSLR or a film RF?
thanks, George.
Thanks for the review. I can hardly wait for my pre-ordered E-P1 17mm kit to arrive (we're at least two weeks behind in Canada). Also, I look forward to seeing how my Voigtlander rangefinder lenses work on it. It appears that I'm looking for the same features and performance from this camera as you are.
Are you ditching the G10 then? I bought the G10 a month plus back and have just about discovered the main shooting / post processing techniques to get the best IQ from it. But to me it's main asset is I can take it anywhere in a pocket. I guess the Pen is too large to do this?
I like the images from the G10 but they are not in the same class as my 40D...do you feel the Pen images are substantially better than the G10? Forget high ISO settings ...I rarely shoot above ISO 200...must be my film upbringing...very much a Pan F person (50 ISO to the uninitiated!!).
The Pen would be outstanding if it matched 40D / L lens qualities.
Should have mentioned that I bought the G10 for landscape / cityscape etc.
Tariq Gibran wrote:
Regarding the ISO 100 shot:
Seems like you would actually get more detail below F5.6 because diffraction starts to kick in with the smaller sensor above 5.6 or so. Contrast looks good and makes it appear that things are actually sharper than they are. There is a surprising amount of rainbow color artifacts/noise for an ISO 100 shot. I think noise reduction would get rid of that without hurting the detail. That is almost certainly due to the weaker AA filter.
You would think, but its always more complicated than just looking at f-stop and pixel size. The lens comes into play, contrast comes into play, the AA filter comes into play. Theory is all good, but in reality, you just have to test and see. I find discussions about diffraction softening to mostly be impractical. In the real world, I just want the shot to be relatively sharp and have a certain depth of field, and its the DOF that usually determines my aperture choice. Diffraction softening can be tweaked out with micro contrast adjustments and mild sharpening.
As for the color effects, that has nothing to do with diffraction at the sensor level, but it is all about diffraction. The dog's hairs are white and hollow, and those hairs have their own distinct diffraction effect that is highly visible in daylight. Every camera I have ever used, regardless of pixel size, aperture, etc, will record that effect in bright sunlight. In regular diffused light, it is not visible. I didn't think that anyone would subscribe that to the AA filter, but it does make sense why you would do so.
dasrocket wrote:
This is a very nice review J, thank you for posting.
I have two questions I would like to ask: how is viewing the LCD in low light for focusing, and how noisy is the shutter, say compared toa DSLR or a film RF?
thanks, George.
How do you mean? It's a nice bright LCD so it's great in low light. If you mean for manual focus, it has a nifty feature where it will zoom in when you start manually focusing, so you have a nice enlarged image to focus, and then when you are done, it snaps back to the full sized picture.
As for shutter noise, the shutter is really quiet. It's quieter than any of my DLSRs.
anthonygh wrote:
Are you ditching the G10 then? I bought the G10 a month plus back and have just about discovered the main shooting / post processing techniques to get the best IQ from it. But to me it's main asset is I can take it anywhere in a pocket. I guess the Pen is too large to do this?
I like the images from the G10 but they are not in the same class as my 40D...do you feel the Pen images are substantially better than the G10? Forget high ISO settings ...I rarely shoot above ISO 200...must be my film upbringing...very much a Pan F person (50 ISO to the uninitiated!!).
The Pen would be outstanding if it matched 40D / L lens qualities.
Should have mentioned that I bought the G10 for landscape / cityscape etc....Show more →
I will be keeping the G10. It's the classic "horses for courses" argument. If I really want outstanding image quality in a compact body, the G10 is impossible to beat at ISO 200 and below. It will always stay in my work bag, always with me.
The Pen would fit in a jacket pocket. In fact, that is how I have been carrying it around when its not around my neck. In terms of overall quality, it's not quite up to the 40D+L combo, but that shouldn't be surprising: the sensor is smaller, and the lenses are not even purported to be top of the line like the L lenses. Still, it comes remarkably close to that level of quality in a tremendously compact package that has few compromises.
The big advantage it has over the G10 at lower ISOs is dynamic range. I would guess that the dynamic range is very similar to the 50D/450D. It also seems to have a slight edge in resolution, but I will need to do some side-by-side shots to confirm that general impression. If I am correct, I would guess its because of the weak AA filter that Olympus appears to be using on the Pen.
jhapeman wrote:
You would think, but its always more complicated than just looking at f-stop and pixel size. The lens comes into play, contrast comes into play, the AA filter comes into play. Theory is all good, but in reality, you just have to test and see. I find discussions about diffraction softening to mostly be impractical. In the real world, I just want the shot to be relatively sharp and have a certain depth of field, and its the DOF that usually determines my aperture choice. Diffraction softening can be tweaked out with micro contrast adjustments and mild sharpening.
As for the color effects, that has nothing to do with diffraction at the sensor level, but it is all about diffraction. The dog's hairs are white and hollow, and those hairs have their own distinct diffraction effect that is highly visible in daylight. Every camera I have ever used, regardless of pixel size, aperture, etc, will record that effect in bright sunlight. In regular diffused light, it is not visible. I didn't think that anyone would subscribe that to the AA filter, but it does make sense why you would do so....Show more →
I did not mean to imply that the color aliasing is due to diffraction. It is due to the use of the bayer color filter array and can show up in areas where both high contrast and high frequency detail exist. The Anti Aliasing filter reduces the effect which is why it could show up more with the use of a weaker filter. If you have seen the effect before, chances are it was with other cameras that also use bayer sensors. Have you seen it with film for instance? If it was some sort of natural light effect through the dogs hair, it would not be the typical red and green dots. Furthermore, you can also see the same effect on the dogs black nose. I have never seen the effect so dramatic before as in the hair of the dog in this instance.
Tariq, I don't see anything on dog's nose. I also don't see red and green dots anywhere. I do some colored specular highlights. I have seen the diffracted light effect before with this dog, and it showed up even with the Sigma DP-2, which as you know does not use a Bayer filter.
Nice, thank you for the high ISO shots. I like the type of noise it makes (), it is round and film-like. Convert to B&W and it will look like pushed Tri-X.
jhapeman wrote:
Tariq, I don't see anything on dog's nose. I also don't see red and green dots anywhere. I do some colored specular highlights. I have seen the diffracted light effect before with this dog, and it showed up even with the Sigma DP-2, which as you know does not use a Bayer filter.
I have seen other aliasing issues with my DP-1 (diagonal jaggies for instance) but not the christmas tree effect. That typically is directly related to the RGB bayer filter color interpolation. If you have a DP-2 image that shows the same effect, that is surprising. I have seen DP-2 examples that show moire.
I see the red and green aliasing just near the dogs nose and of course dramatically in the dogs hair, even on my little netbook screen. The 100% version of course. I doubt you would see this in normal shooting in any case.
Rob Riley wrote:
not sure this is a totally kosher thing to do, but it is a bit special
if you want a peak at EP1 with a CZ 35/2 Biogon aboard check out the links
Wow, that is truly camera porn. I have ordered a Voigtlander 50/1.1 lens and the Cameraquest adapter. Hopefully I will have it at the end of this week.
jhapeman wrote:
How do you mean? It's a nice bright LCD so it's great in low light. If you mean for manual focus, it has a nifty feature where it will zoom in when you start manually focusing, so you have a nice enlarged image to focus, and then when you are done, it snaps back to the full sized picture.
As for shutter noise, the shutter is really quiet. It's quieter than any of my DLSRs.
Thanks J. In terms of the LCD, does it get really grainy under low light when it has to gain a lot to compensate? I found that to be the case with live view before.
OK, understood. It does turn up the gain, but it doesn't get very grainy. I just gave it a shot. It depends on how you define low light, but the shot I just tried was f/3.5, 1/20s, ISO 1600, so pretty low light. I could see just a little grain on the LCD.
Tariq Gibran wrote:
I see the red and green aliasing just near the dogs nose and of course dramatically in the dogs hair, even on my little netbook screen. The 100% version of course. I doubt you would see this in normal shooting in any case.
I think that, besides well confined CA, you may be looking at iridescence from the dog's various reflective microstructures, rather than colour aliasing. SEM imaging of its hair would likely show washboard patterns at the correct frequency to work as diffraction gratings. Compare the (very slight but consistent) CA around reflections caused by water droplets around its nose and eye versus the rainbow effects in light reflected off its hair and some of its dry skin.
The market (demanding weaker AA) appears to have spoken, and in torture scenarios some colour aliases do manage to get imaged on the E-P1. But I've yet to see anything like the dog's result in a controlled test.
), it is round and film-like. Convert to B&W and it will look like pushed Tri-X.