Archive 2014 · Sony FE 24-70 lens
p.4 #1Sony FE 24-70 lens
Ok. The FE55/1.8 is 281 grams.
On this shot the FE24-70 stands to the right of the FE55. The FE55 is not very big but it aint small either.
EDIT: link didnt work. New try, 5th image from the top:
http://www.ephotozine.com/article/sony-alpha-7r-vs-olympus-om-d-e-m5-comparison-23206
On this shot the FE24-70 stands to the right of the FE55. The FE55 is not very big but it aint small either.
EDIT: link didnt work. New try, 5th image from the top:
http://www.ephotozine.com/article/sony-alpha-7r-vs-olympus-om-d-e-m5-comparison-23206
Jan 26, 2014 at 04:42 PM
p.4 #2Sony FE 24-70 lens
Sony went to work on weight but can only do so much for dimensions, it seems.
Here is a direct comparison of the 55/1.8 and 24-70/4 on the a7r:
http://camerasize.com/compact/#488.395,488.393,ha,t
So to swap, you go from 750 grams to 900 grams plus an inch in retracted lens length and slightly broader lens width.
A better guide for other cameras is 'like for like' although many D800 users prefer their a7r output:
http://camerasize.com/compact/#488.393,291.327,ha,t
I couldn't find a 24-70/4 in Nikon's available lenses on that camera size site - so either they don't make one yet or it is not listed. I put the 24-70/2.8 on it therefore, you get 1900 grams versus 900 grams that way.
The Canon 6D plus Canon 24-70/4 is here, against the a7/r plus 24-70/4:
http://camerasize.com/compact/#488.393,380.367,ha,t
It is more compact than Nikon's offering, but weighs 1370 grams all up, for a penalty of 470 grams = 50% more weight. [Then more weight delta for batts, pod/head, bags, accessories.] Canon setup is $3400 at BH, Sony a7 plus 24-70/4 is $2900.
Another early 24-70/4 user report here:
http://www.dpreview.com/forums/thread/3613392#forum-post-52987578
Here is a direct comparison of the 55/1.8 and 24-70/4 on the a7r:
http://camerasize.com/compact/#488.395,488.393,ha,t
So to swap, you go from 750 grams to 900 grams plus an inch in retracted lens length and slightly broader lens width.
A better guide for other cameras is 'like for like' although many D800 users prefer their a7r output:
http://camerasize.com/compact/#488.393,291.327,ha,t
I couldn't find a 24-70/4 in Nikon's available lenses on that camera size site - so either they don't make one yet or it is not listed. I put the 24-70/2.8 on it therefore, you get 1900 grams versus 900 grams that way.
The Canon 6D plus Canon 24-70/4 is here, against the a7/r plus 24-70/4:
http://camerasize.com/compact/#488.393,380.367,ha,t
It is more compact than Nikon's offering, but weighs 1370 grams all up, for a penalty of 470 grams = 50% more weight. [Then more weight delta for batts, pod/head, bags, accessories.] Canon setup is $3400 at BH, Sony a7 plus 24-70/4 is $2900.
Another early 24-70/4 user report here:
http://www.dpreview.com/forums/thread/3613392#forum-post-52987578
Jan 26, 2014 at 07:13 PM
p.4 #3Sony FE 24-70 lens
I generally like the rendering of SLRCLUB's review images (even if Jpeg smoothing is a bit too strong), though more so when there is ample subject-background separation. When not, then there is some tendency to double-line effects. Perhaps a bit disappointing though is that at 24mm there is still fairly noticeable sharpness falloff towards the edges in higher detail scenes, even when stopped down past f/8.
Philip - you asked about the Canon 24-70/4. It's generally perceived as a relatively disappointing release compared to the optical quality of most of their recent refreshes, such as the 24-70/2.8 and 70-200... I've briefly used one and my opinion is it has a too smooth quality lacking fine detail bite. I think the Sony looks better, though do wish it performed better at 24mm..
Philip - you asked about the Canon 24-70/4. It's generally perceived as a relatively disappointing release compared to the optical quality of most of their recent refreshes, such as the 24-70/2.8 and 70-200... I've briefly used one and my opinion is it has a too smooth quality lacking fine detail bite. I think the Sony looks better, though do wish it performed better at 24mm..
Jan 26, 2014 at 07:31 PM
p.4 #4Sony FE 24-70 lens
I had a chance to play with one at one of the major camera retailers here in Japan yesterday. It is a nicely built lens and really doesnt feel all that large on my gripped a7r. My only problem as others has mentioned the sharpness at 24mm was not as good as I was hoping. I will need to play around with the files a little bit, but so far I think I will give this a pass. Shame I was hoping this would be a good solution for hiking but for now I will keep looking. I hope the wide angle zoom will be better, if and when Sony releases it.
Ryan
Ryan
Jan 26, 2014 at 08:03 PM
p.4 #5Sony FE 24-70 lens
Is a 16-35 confirmed to released - and when - or is it just speculation?
This one is linked from sonyalpharumors and called "almost certainly a fake"...
This one is linked from sonyalpharumors and called "almost certainly a fake"...
Jan 27, 2014 at 02:07 AM
p.4 #6Sony FE 24-70 lens
The fake roadmap above is someone´s wishful thinking, the real lens roadmap is more indicative, but a wide-angle f4 zoom is on it for the first half of 2014:
http://www.sonyalpharumors.com/the-full-frame-e-mount-lens-roadmap-15-lenses-will-be-launched-until-late-2015/
http://www.sonyalpharumors.com/the-full-frame-e-mount-lens-roadmap-15-lenses-will-be-launched-until-late-2015/
Jan 27, 2014 at 02:57 AM
p.4 #7Sony FE 24-70 lens
Steve Spencer wrote:
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.
Jan 27, 2014 at 04:37 AM
p.4 #8Sony FE 24-70 lens
One more link for the shots taken with f/8.0.... 55/1.8 and 24-70.
http://www.slrclub.com/bbs/vx2.php?id=minolta_forum&no=822358
http://www.slrclub.com/bbs/vx2.php?id=minolta_forum&no=822358
Jan 27, 2014 at 10:25 AM
p.4 #9Sony FE 24-70 lens
alundeb wrote:
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.
Unfortunately, I think this is too optimistic. What CA correction does is turn the coloured fringing grey, it does not recover the lost resolution so instead of a coloured blurry image you will get a grey blurry image. This is one of the reasons that CA is still an important issue even though it is easily "correctable" with software. Whether the astigmatism is caused by the CA or the CA causes the astigmatism (I think it is the former and not the latter), the point is that the image in the test chart shows both and although you can get rid of the colour from the CA you can't get rid of the blurriness by changing the colour to grey.
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.Unfortunately, I think this is too optimistic. What CA correction does is turn the coloured fringing grey, it does not recover the lost resolution so instead of a coloured blurry image you will get a grey blurry image. This is one of the reasons that CA is still an important issue even though it is easily "correctable" with software. Whether the astigmatism is caused by the CA or the CA causes the astigmatism (I think it is the former and not the latter), the point is that the image in the test chart shows both and although you can get rid of the colour from the CA you can't get rid of the blurriness by changing the colour to grey.
Jan 27, 2014 at 10:42 AM
p.4 #10Sony FE 24-70 lens
Steve Spencer wrote:
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.Unfortunately, I think this is too optimistic. What CA correction does is turn the coloured fringing grey, it does not recover the lost resolution so instead of a coloured blurry image you will get a grey blurry image. This is one of the reasons that CA is still an important issue even though it is easily "correctable" with software. Whether the astigmatism is caused by the CA or the CA causes the astigmatism (I think it is the former and not the latter), the point is that the image in the test chart shows both and although you can...Show more →
Steve,
Looking at the SLRClub test chart at 24mm and f/2.8, I see the astigmatism as the faint ghost below and to the right of the subjects which is rather visible near the "18" and the single line below it; this isn't visible to me at f/5.6. The lateral chromatic aberration only becomes sharper, though, and that would be easy to fix in software. Even in the converging lines, where the green from one line overlays the red from the next, would clear up in software because it is working on each color independently. You're right that CA can only be desaturated when you're looking at longitudinal CA (from spherical aberration) or purple fringing (from internal reflections), but lateral CA is conveniently different.
But then, I'm self-taught when it comes to optics, so I could be extraordinarily misguided. I'd appreciate any corrections to my understanding that you might offer.
Cheers,
Jon
alundeb wrote
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.Steve,
Looking at the SLRClub test chart at 24mm and f/2.8, I see the astigmatism as the faint ghost below and to the right of the subjects which is rather visible near the "18" and the single line below it; this isn't visible to me at f/5.6. The lateral chromatic aberration only becomes sharper, though, and that would be easy to fix in software. Even in the converging lines, where the green from one line overlays the red from the next, would clear up in software because it is working on each color independently. You're right that CA can only be desaturated when you're looking at longitudinal CA (from spherical aberration) or purple fringing (from internal reflections), but lateral CA is conveniently different.
But then, I'm self-taught when it comes to optics, so I could be extraordinarily misguided. I'd appreciate any corrections to my understanding that you might offer.
Cheers,
Jon
Jan 27, 2014 at 11:02 AM
p.4 #11Sony FE 24-70 lens
That's a mighty big lens for a camera system that is trying to extol the smallness factor that is supposed to be inherent to a mirror-less system.
I hope Sony manages to release some small lenses that let in plenty of light. So far, the 55/1.8 does neither as does the 24-70/4
I hope Sony manages to release some small lenses that let in plenty of light. So far, the 55/1.8 does neither as does the 24-70/4
Jan 27, 2014 at 11:16 AM
p.4 #12Sony FE 24-70 lens
kezeka wrote:
That's a mighty big lens for a camera system that is trying to extol the smallness factor that is supposed to be inherent to a mirror-less system.
I hope Sony manages to release some small lenses that let in plenty of light. So far, the 55/1.8 does neither as does the 24-70/4
Sony chose performance with the 55/1.8, and, while being better than any DSLR standard lens this side of the OTUS, it is still about the same size as any other DSLR standard prime + adapter, which is an accomplishment.
That's a mighty big lens for a camera system that is trying to extol the smallness factor that is supposed to be inherent to a mirror-less system.
I hope Sony manages to release some small lenses that let in plenty of light. So far, the 55/1.8 does neither as does the 24-70/4
Sony chose performance with the 55/1.8, and, while being better than any DSLR standard lens this side of the OTUS, it is still about the same size as any other DSLR standard prime + adapter, which is an accomplishment.
Jan 27, 2014 at 11:30 AM
p.4 #13Sony FE 24-70 lens
JonPB wrote:
Steve Spencer wrote
alundeb wrote
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.Looking at the SLRClub test chart at 24mm and f/2.8, I see the astigmatism as the faint ghost below and to the right of the subjects which is rather visible near the "18" and the single line below it; this isn't visible to me at f/5.6. The lateral chromatic aberration only becomes sharper, though, and that would be easy to fix in software. Even in the converging lines, where the green from one line overlays the red from the next, would clear up in software because it is working on each color independently. You're right that CA can only be...Show more →
Hi Jon,
I am self taught or rather I have tried to learn from some of the smart folks around here sand I believe Toothwalker who knows his stuff has commented negatively on the possibility of really correcting for even lateral CA. He has written an excellent guide on CA as well. I do see lots of astigmatism at F8 or at least very different sharpness between horizontal and vertical lines which I take as astigmatism and I stand by my assessment that the sharpness will not improve with CA correction.
Jan 27, 2014 at 11:46 AM
p.4 #14Sony FE 24-70 lens
Steve Spencer wrote:
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.Unfortunately, I think this is too optimistic. What CA correction does is turn the coloured fringing grey, it does not recover the lost resolution so instead of a coloured blurry image you will get a grey blurry image. This is one of the reasons that CA is still an important issue even though it is easily "correctable" with software. Whether the astigmatism is caused by the CA or the CA causes the astigmatism (I think it is the former and not the latter), the point is that the image in the test chart shows both and although you can...Show more →
The CA correction software you describe, is how Longitunidal CA or bokeh fringing is corrected. For that type of CA the software cannot know the radius of the color fringing and has to analyze the image and remove color.
Correction of lateral CA works differently. Even without a lens profile, simple lateral CA can be modeled as if some wavelengths project a slightly larger, but otherwise identical image, and some a slightly smaller. The correction is made by decomposing, rescaling, and recombining color channels. When I apply lateral CA correction in Capture One, and look at the edge og the image, I can actually see that the image details moves a litte back and forth when I toggle the correction.
Still software correction is not perfect and I do not believe a software corrected image is as good as an image from a lens with better color correction.
alundeb wrote
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.The CA correction software you describe, is how Longitunidal CA or bokeh fringing is corrected. For that type of CA the software cannot know the radius of the color fringing and has to analyze the image and remove color.
Correction of lateral CA works differently. Even without a lens profile, simple lateral CA can be modeled as if some wavelengths project a slightly larger, but otherwise identical image, and some a slightly smaller. The correction is made by decomposing, rescaling, and recombining color channels. When I apply lateral CA correction in Capture One, and look at the edge og the image, I can actually see that the image details moves a litte back and forth when I toggle the correction.
Still software correction is not perfect and I do not believe a software corrected image is as good as an image from a lens with better color correction.
Jan 27, 2014 at 01:59 PM
p.4 #15Sony FE 24-70 lens
alundeb wrote:
Steve Spencer wrote
alundeb wrote
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.Correction of lateral CA works differently. Even without a lens profile, simple lateral CA can be modeled as if some wavelengths project a slightly larger, but otherwise identical image, and some a slightly smaller. The correction is made by decomposing, rescaling, and recombining color channels. When I apply lateral CA correction in Capture One, and look at the edge og the image,...Show more →
Ok, I will be more precise in my language and I think we can agree. CA correction cannot create resolution that was lost when the CA was present in the initial capture. That is why an image corrected by software can never be as sharp as an image taken with a lens that does not produce the CA.
Jan 27, 2014 at 02:09 PM
p.4 #16Sony FE 24-70 lens
Steve Spencer wrote:
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.Unfortunately, I think this is too optimistic. What CA correction does is turn the coloured fringing grey, it does not recover the lost resolution so instead of a coloured blurry image you will get a grey blurry image. This is one of the reasons that CA is still an important issue even though it is easily "correctable" with software. Whether the astigmatism is caused by the CA or the CA causes the astigmatism (I think it is the former and not the latter), the point is that the image in the test chart shows both and although you can get rid of the colour from the CA you can't get rid of the blurriness by changing the colour to grey.The CA correction software you describe, is how Longitunidal CA or bokeh fringing is corrected. For that type of CA the software cannot know the radius of the color fringing and has to analyze the image and remove color.
Correction of lateral CA works differently. Even without a lens profile, simple lateral CA can be modeled as if some wavelengths project a slightly larger, but otherwise identical image, and some a slightly smaller. The correction is made by decomposing, rescaling, and recombining color channels. When I apply lateral CA correction in Capture One, and look at the edge og the image, I can actually see that the image details moves a litte back and forth when I toggle the correction.
Still software correction is not perfect and I do not believe a software corrected image is as good as an image from a lens with better color correction.Ok, I will be more precise in my language and I think we can agree. CA correction cannot create resolution that was lost when the CA was present in the initial capture. That is why an image corrected by software can never be as sharp as an image taken with a lens that does not produce the CA.
We may agree, lost resolution cannot be restored. The question is how much of the resolution is actually lost, and how much is only apparently lost. I do observe that some of the loss in observed acuity caused by CA can actually be restored by software correction of colors. A "perfect" lens, with the exception of lateral CA, will project perfect images if the light is monochromatic. If the light is white and the subject black and white, and we take images filtered with narrow spectrum color filters, the lens will produce perfect images with different colors but slightly differently scaled. If these images are then perfectly re-scaled and perfectly realigned, we will get an image in black and white that is as sharp as a perfect image from a lens without CA.
The reasons we don't get perfect correction, is in the accuracy of the model, and in the implementation of the correction. I do observe that the correction process itself, smooths out some very fine detail.
Here is an example from the edge of the Sony E 10-18 lens, without and with CA correction. The view is at 300%.
One interesting side-effect, is that false detail from aliasing has been removed as well.
http://i306.photobucket.com/albums/nn242/Overflate/CaCorrection_zps8d15fad9.jpg
alundeb wrote
Steve Spencer wrote
alundeb wrote
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.Correction of lateral CA works differently. Even without a lens profile, simple lateral CA can be modeled as if some wavelengths project a slightly larger, but otherwise identical image, and some a slightly smaller. The correction is made by decomposing, rescaling, and recombining color channels. When I apply lateral CA correction in Capture One, and look at the edge og the image, I can actually see that the image details moves a litte back and forth when I toggle the correction.
Still software correction is not perfect and I do not believe a software corrected image is as good as an image from a lens with better color correction.
We may agree, lost resolution cannot be restored. The question is how much of the resolution is actually lost, and how much is only apparently lost. I do observe that some of the loss in observed acuity caused by CA can actually be restored by software correction of colors. A "perfect" lens, with the exception of lateral CA, will project perfect images if the light is monochromatic. If the light is white and the subject black and white, and we take images filtered with narrow spectrum color filters, the lens will produce perfect images with different colors but slightly differently scaled. If these images are then perfectly re-scaled and perfectly realigned, we will get an image in black and white that is as sharp as a perfect image from a lens without CA.
The reasons we don't get perfect correction, is in the accuracy of the model, and in the implementation of the correction. I do observe that the correction process itself, smooths out some very fine detail.
Here is an example from the edge of the Sony E 10-18 lens, without and with CA correction. The view is at 300%.
One interesting side-effect, is that false detail from aliasing has been removed as well.
http://i306.photobucket.com/albums/nn242/Overflate/CaCorrection_zps8d15fad9.jpg
Jan 27, 2014 at 02:45 PM
p.4 #17Sony FE 24-70 lens
kezeka, you may know that aberrations in fast lenses, such as f1.4 primes, are an order of magnitude higher than 1-2 stops slower lenses. Designers also have other design techniques at their disposal to induce focus fade and bokeh qualities in much better behaved slower lenses. It applies especially in lenses of normal and longer FLs.
Very fast lenses made real sense in the era of ISO limited film emulsions, where high quality colour might end at ISO 400 or so, and people just wanted to be able to get the shot. The figure for Sony FF cameras is now ISO 3200-6400, so letting more light in is a much less pressing issue.
The FE55 is a mere 281 grams, has D/L of 64mm x 71mm, costs $1000 and loses but a ~half stop to the standard fast lens speed for the class - f1.4.
Because this is a competitive industry, let's look at the latest Nikkor, the 58mm f1.4G - so it's one that can 'let more light in'. It weighs 385 grams (+100 grams); is D/L of 85mm and 70mm (+21mm for D, same length) and costs $1700 (+$700). Has much less raw performance also, and got a poor reception on the Nikon forum:
https://www.fredmiranda.com/forum/topic/1247893
nice test showing this one:
http://slrgear.com/reviews/showproduct.php/product/1660/cat/12
It's nowhere near as good as the FE55 in almost all photographic aspects, and is larger, heavier and much more costly. As an example the FE lens is virtually free of CA, one of the remaining pernicious aberrations. Distortion too. It seems Sony Zeiss made some very good optical choices in the design of its 55mm prime lens for FE, no?
Or is it only mirrorless system lenses that need to be exemplars of fast aperture, light weight, small size, low cost, high performance?
That could be interpreted as a double standard, just maybe. Many consumers might see it that way too, given time, since quality generally wins out.
Very fast lenses made real sense in the era of ISO limited film emulsions, where high quality colour might end at ISO 400 or so, and people just wanted to be able to get the shot. The figure for Sony FF cameras is now ISO 3200-6400, so letting more light in is a much less pressing issue.
The FE55 is a mere 281 grams, has D/L of 64mm x 71mm, costs $1000 and loses but a ~half stop to the standard fast lens speed for the class - f1.4.
Because this is a competitive industry, let's look at the latest Nikkor, the 58mm f1.4G - so it's one that can 'let more light in'. It weighs 385 grams (+100 grams); is D/L of 85mm and 70mm (+21mm for D, same length) and costs $1700 (+$700). Has much less raw performance also, and got a poor reception on the Nikon forum:
https://www.fredmiranda.com/forum/topic/1247893
nice test showing this one:
http://slrgear.com/reviews/showproduct.php/product/1660/cat/12
It's nowhere near as good as the FE55 in almost all photographic aspects, and is larger, heavier and much more costly. As an example the FE lens is virtually free of CA, one of the remaining pernicious aberrations. Distortion too. It seems Sony Zeiss made some very good optical choices in the design of its 55mm prime lens for FE, no?
Or is it only mirrorless system lenses that need to be exemplars of fast aperture, light weight, small size, low cost, high performance?
That could be interpreted as a double standard, just maybe. Many consumers might see it that way too, given time, since quality generally wins out.
Jan 27, 2014 at 04:40 PM
p.4 #18Sony FE 24-70 lens
phillip, this is otherwise known as 'having your cake and eating it too.' Complaints like that are pretty common for every system. Nothing is going to please everyone or be perfect in every possible way.
The FE 55 is a good balance between size and performance. It's small enough to fit my definition of small, and its general performance is as good as, if not better than, any other normal prime I've used. The tradeoff for those benefits? It's on the pricey side.
The FE 55 is a good balance between size and performance. It's small enough to fit my definition of small, and its general performance is as good as, if not better than, any other normal prime I've used. The tradeoff for those benefits? It's on the pricey side.
Jan 27, 2014 at 04:56 PM
p.4 #19Sony FE 24-70 lens
Steve Spencer wrote:
To me it looks like the subjective difference in sharpness in horizontal and vertical lines is caused by CA and not astigmatism. If it was astigmatism in the strict sense (different focus sag / tan), it should be worse at wider apertures and almost disappear at smaller apertures. This is not the case. Lateral CA, however, is not dependent on aperture, and it looks like it is what we see here. IMO there is hope that the lens will deliver good performance after software correction of CA, and the type seen here (red / green lateral), is easily correctable.Unfortunately, I think this is too optimistic. What CA correction does is turn the coloured fringing grey, it does not recover the lost resolution so instead of a coloured blurry image you will get a grey blurry image. This is one of the reasons that CA is still an important issue even though it is easily "correctable" with software. Whether the astigmatism is caused by the CA or the CA causes the astigmatism (I think it is the former and not the latter), the point is that the image in the test chart shows both and although you can get rid of the colour from the CA you can't get rid of the blurriness by changing the colour to grey.Steve,
Looking at the SLRClub test chart at 24mm and f/2.8, I see the astigmatism as the faint ghost below and to the right of the subjects which is rather visible near the "18" and the single line below it; this isn't visible to me at f/5.6. The lateral chromatic aberration only becomes sharper, though, and that would be easy to fix in software. Even in the converging lines, where the green from one line overlays the red from the next, would clear up in software because it is working on each color independently. You're right that CA can only be desaturated when you're looking at longitudinal CA (from spherical aberration) or purple fringing (from internal reflections), but lateral CA is conveniently different.
But then, I'm self-taught when it comes to optics, so I could be extraordinarily misguided. I'd appreciate any corrections to my understanding that you might offer.
Cheers,
JonHi Jon,
I am self taught or rather I have tried to learn from some of the smart folks around here sand I believe Toothwalker who knows his stuff has commented negatively on the possibility of really correcting for even lateral CA. He has written an excellent guide on CA as well. I do see lots of astigmatism at F8 or at least very different sharpness between horizontal and vertical lines which I take as astigmatism and I stand by my assessment that the sharpness will not improve with CA correction.
Toothwalker.org does have some nice articles that I'll have to look more at--thanks for mentioning that.
I agree insofar as lateral CA cannot be perfectly corrected (though you can get pretty much all the way there if you know the lens' and sensor's exact performance, which off-the-shelf software won't).
I did some of my own tests, like alundeb's, and what I found was fairly good removal of the problem but also, most interestingly, a strong addition of color noise. The camera presumably removes color noise in its AD conversion, but correcting for lateral CA undoes that by turning luminance noise into radially-oriented color noise. I haven't done further tests to see how easy that is to correct for, but it certainly agrees with the idea that you want the best possible capture before applying any image processing.
Ah, the joys of camera geekery...
Cheers,
Jon
JonPB wrote
Steve Spencer wrote
alundeb wrote
Steve Spencer wrote
The astigmatism is clearly evident as well. If you look at the test chart for 24mm and f/8 in the corner you will see that the horizontal lines show pretty good resolution, whereas the vertical lines have pretty poor resolution. This discrepancy between resolution in different orientations is exactly what astigmatism is and can be measured on MTF charts from a K8 machine as differences between tang/sag lines. At 24mm there is some pretty strong CA as well. This may be caused by the astigmatism, but I am not sure.Looking at the SLRClub test chart at 24mm and f/2.8, I see the astigmatism as the faint ghost below and to the right of the subjects which is rather visible near the "18" and the single line below it; this isn't visible to me at f/5.6. The lateral chromatic aberration only becomes sharper, though, and that would be easy to fix in software. Even in the converging lines, where the green from one line overlays the red from the next, would clear up in software because it is working on each color independently. You're right that CA can only be desaturated when you're looking at longitudinal CA (from spherical aberration) or purple fringing (from internal reflections), but lateral CA is conveniently different.
But then, I'm self-taught when it comes to optics, so I could be extraordinarily misguided. I'd appreciate any corrections to my understanding that you might offer.
Cheers,
Jon
I am self taught or rather I have tried to learn from some of the smart folks around here sand I believe Toothwalker who knows his stuff has commented negatively on the possibility of really correcting for even lateral CA. He has written an excellent guide on CA as well. I do see lots of astigmatism at F8 or at least very different sharpness between horizontal and vertical lines which I take as astigmatism and I stand by my assessment that the sharpness will not improve with CA correction.
Toothwalker.org does have some nice articles that I'll have to look more at--thanks for mentioning that.
I agree insofar as lateral CA cannot be perfectly corrected (though you can get pretty much all the way there if you know the lens' and sensor's exact performance, which off-the-shelf software won't).
I did some of my own tests, like alundeb's, and what I found was fairly good removal of the problem but also, most interestingly, a strong addition of color noise. The camera presumably removes color noise in its AD conversion, but correcting for lateral CA undoes that by turning luminance noise into radially-oriented color noise. I haven't done further tests to see how easy that is to correct for, but it certainly agrees with the idea that you want the best possible capture before applying any image processing.
Ah, the joys of camera geekery...
Cheers,
Jon
Jan 27, 2014 at 06:07 PM
p.4 #20Sony FE 24-70 lens
Thanks for the posts alundeb and JonPB. I don't think we are that far apart, I guess I am just a bit more pessimistic about CA removable. My own experience is that I have never found it to be that useful and that it robs details and I think that can be seen in alundeb's example as well. The example is interesting because what I like best about the CA removal is how it reduces moire and I am not fond at all about what it does to colour (look at how the windows in the front structure now have a green tint) or what it does to detail (I agree with alundeb that it smooths out fine detail--more that I personally would like). Anyway, I have enjoyed the discussion and I thank you both. I feel I have learned a bit more today and that is always a great party of FM.
Jan 27, 2014 at 08:52 PM