Sony FE 12-24mm f/2.8 GM Review
p.10 #1Sony FE 12-24mm f/2.8 GM Review
I've just received a very good copy of the Voigtlander 15mm f/4.5 III (Perfectly symmetric without any corner smearing). I sold mine a while back and have regretted the decision.
If the weather cooperates I will compare the CV 15/4.5 III and FE 24/1.4 GM to the Sony 12-24/2.8 GM today.
If the weather cooperates I will compare the CV 15/4.5 III and FE 24/1.4 GM to the Sony 12-24/2.8 GM today.
Aug 25, 2020 at 11:37 AM
p.10 #2Sony FE 12-24mm f/2.8 GM Review
Can I be first in line for that cv 15 if you ever sell 
Aug 25, 2020 at 12:11 PM
p.10 #3Sony FE 12-24mm f/2.8 GM Review
johnahill wrote:
These are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...I realise that this is disappointing, genuine question - how much would this affect real life use at say f5.6-f8
It is still apparent at ~F7.1 where I like to focus stack, but it probably wouldn't show in prints by then. But I would want this lens to be good wide open for astro anyway, so this copy is out of the question for me. I will be very disappointed if the sample images I sent to the seller, which are the same as the ones I posted above, go to Sony for assessment and come back as being considered "in spec".
rvh23 wrote
This is the part I dislike most about evaluating these complex zooms. I received my GM 12-24 today and it clearly failed its centering test at all focal lengths with a soft lower left corner across the whole rangeThese are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...
It is still apparent at ~F7.1 where I like to focus stack, but it probably wouldn't show in prints by then. But I would want this lens to be good wide open for astro anyway, so this copy is out of the question for me. I will be very disappointed if the sample images I sent to the seller, which are the same as the ones I posted above, go to Sony for assessment and come back as being considered "in spec".
Aug 25, 2020 at 04:09 PM
p.10 #4Sony FE 12-24mm f/2.8 GM Review
 |
||
 |
||
|
 |
||
 |
||
|
- Distance: Infinity
- Focus: Center - Best of three @ 12.4x magnification
- WB: Daylight for both lenses
- Both Lenses perfectly centered using my decentering test
- Software: Lightroom with FM Default Landscape Sharpening. All other settings set to default
15mm
CENTER resolution/contrast comparison between Sony FE 12-24/2.8 GM @15mm vs Voigtlander 15mm f/4.5 III
Note: Both Voigtlander 15mm f/4.5 III M-mount and E-mount version have identical performance on the A7R IV
At center, both lenses are already optimal at f/4.5 which is a great performance for the Voigtlander 15/4.5 III since it's wide open Very close performance for both lenses at center at all apertures.
Center at f/4.5: Both lenses perform similarly which is a surprise for the Voigtlander since it's wide open
Center at f/5.6: Not much changes for both lenses at f/5.6
Center at f/8: Both lenses diffract
Aug 25, 2020 at 04:23 PM
p.10 #5Sony FE 12-24mm f/2.8 GM Review
15mm
MID-FIELD resolution/contrast comparison between Sony FE 12-24/2.8 GM @15mm vs Voigtlander 15mm f/4.5 III
At mid-field, both lenses are already optimal at f/4.5 which is a great performance for the Voigtlander 15/4.5 III. However, the prime never matches the resolution and contrast of the Sony 12-24/2.8 GM.
Mid-field at f/4.5: Higher resolution and contrast for the Sony 12-24/2.8 GM
Mid-field at f/5.6: Not much changes for both lenses
Mid-field at f/8: Both lenses show signs of diffraction
MID-FIELD resolution/contrast comparison between Sony FE 12-24/2.8 GM @15mm vs Voigtlander 15mm f/4.5 III
At mid-field, both lenses are already optimal at f/4.5 which is a great performance for the Voigtlander 15/4.5 III. However, the prime never matches the resolution and contrast of the Sony 12-24/2.8 GM.
Mid-field at f/4.5: Higher resolution and contrast for the Sony 12-24/2.8 GM
Mid-field at f/5.6: Not much changes for both lenses
Mid-field at f/8: Both lenses show signs of diffraction
Aug 25, 2020 at 04:23 PM
p.10 #6Sony FE 12-24mm f/2.8 GM Review
15mm
EXTREME CORNER resolution/contrast comparison between Sony FE 12-24/2.8 GM @15mm vs Voigtlander 15mm f/4.5 III
At extreme corners, the Voigtlander 15/4.5 III improves slightly at f/5.6 while the Sony is already optimal at f/4.5. The prime never matches the resolution and contrast of the Sony 12-24/2.8 GM.
Extreme corner at f/4.5: Great performance for both lenses but the Sony GM is noticeably ahead
Extreme corner at f/5.6: Here the Voigtlander improves very slightly and the GM shows no improvement. The GM is already optimal at f/4.5 for this area
Extreme corner at f/8: Both lenses diffract
EXTREME CORNER resolution/contrast comparison between Sony FE 12-24/2.8 GM @15mm vs Voigtlander 15mm f/4.5 III
At extreme corners, the Voigtlander 15/4.5 III improves slightly at f/5.6 while the Sony is already optimal at f/4.5. The prime never matches the resolution and contrast of the Sony 12-24/2.8 GM.
Extreme corner at f/4.5: Great performance for both lenses but the Sony GM is noticeably ahead
Extreme corner at f/5.6: Here the Voigtlander improves very slightly and the GM shows no improvement. The GM is already optimal at f/4.5 for this area
Extreme corner at f/8: Both lenses diffract
Aug 25, 2020 at 04:24 PM
p.10 #7Sony FE 12-24mm f/2.8 GM Review
|
||
|
||
|
|
||
 |
||
|
- Distance: Infinity
- Focus: Center - Best of three @ 12.4x magnification
- WB: Daylight for both lenses
- Both Lenses perfectly centered using my decentering test
- Software: Lightroom with FM Default Landscape Sharpening. All other settings set to default
24mm
CENTER resolution/contrast comparison between Sony FE 12-24/2.8 GM @24mm vs Sony FE 24mm f/1.4 GM
Aug 25, 2020 at 04:40 PM
p.10 #8Sony FE 12-24mm f/2.8 GM Review
24mm
MID-FIELD resolution/contrast comparison between Sony FE 12-24/2.8 GM @24mm vs Sony FE 24mm f/1.4 GM
MID-FIELD resolution/contrast comparison between Sony FE 12-24/2.8 GM @24mm vs Sony FE 24mm f/1.4 GM
Aug 25, 2020 at 04:40 PM
p.10 #9Sony FE 12-24mm f/2.8 GM Review
24mm
MID-FIELD resolution/contrast comparison between Sony FE 12-24/2.8 GM @24mm vs Sony FE 24mm f/1.4 GM
MID-FIELD resolution/contrast comparison between Sony FE 12-24/2.8 GM @24mm vs Sony FE 24mm f/1.4 GM
Aug 25, 2020 at 04:40 PM
p.10 #10Sony FE 12-24mm f/2.8 GM Review
Fred
Your copy seems to be cleaning up against all of the others.
Your copy seems to be cleaning up against all of the others.
Aug 25, 2020 at 04:55 PM
|
p.10 #11Sony FE 12-24mm f/2.8 GM Review
Would love the Sony, but at well over $4K in Australia not going to happen and the Sigma is plenty good enough IMO as it pushes the Sony hard and for well under half the money.
Aug 25, 2020 at 10:27 PM
p.10 #12Sony FE 12-24mm f/2.8 GM Review
I'm kind of thinking about the 12mm Voigtlander to go with my 16-35 GM.
Edited on Aug 26, 2020 at 08:55 AM · View previous versions
Edited on Aug 26, 2020 at 08:55 AM · View previous versions
Aug 25, 2020 at 10:31 PM
p.10 #13Sony FE 12-24mm f/2.8 GM Review
Now Sony just needs to apply that to an 14 f/1.4 or even f/1.2 and I'm willing to pay ~2k for a prime!
C'mon you can do it Sony!
@Fred
I disagree slightly, f/2.8 is kind of borderline for astro.
Since the Sigma 14 1.8 exists we are spoiled!
C'mon you can do it Sony!
@Fred
I disagree slightly, f/2.8 is kind of borderline for astro.
Since the Sigma 14 1.8 exists we are spoiled!
Aug 26, 2020 at 12:29 AM
p.10 #15Sony FE 12-24mm f/2.8 GM Review
What an amazing lens, it's even better than I expected when Sony announced it.
Aug 26, 2020 at 01:10 AM
p.10 #16Sony FE 12-24mm f/2.8 GM Review
rvh23 wrote:
These are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...I realise that this is disappointing, genuine question - how much would this affect real life use at say f5.6-f8It is still apparent at ~F7.1 where I like to focus stack, but it probably wouldn't show in prints by then. But I would want this lens to be good wide open for astro anyway, so this copy is out of the question for me. I will be very disappointed if the sample images I sent to the seller, which are the same as the ones I posted above, go to Sony for assessment and come back as being considered "in spec".
I feel your pain Richard. IMHO, that's not an acceptable amount of variation if it exists throughout the focal length range. I know I said earlier in this thread that paying more money for a lens doesn't guarantee less sample variation, but that doesn't mean I think that status quo is acceptable; its the principle of the thing: if you buy a 2.8 lens at this price it should be usable without compromise at 2.8!
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions for most, so best to future proof for those rather than compromising just because current reflective media destroys some detail.
Might be better to wait until you're out of lockdown so you can test the next sample in-store...
Cheers
Peter
johnahill wrote
rvh23 wrote
This is the part I dislike most about evaluating these complex zooms. I received my GM 12-24 today and it clearly failed its centering test at all focal lengths with a soft lower left corner across the whole rangeThese are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...
I feel your pain Richard. IMHO, that's not an acceptable amount of variation if it exists throughout the focal length range. I know I said earlier in this thread that paying more money for a lens doesn't guarantee less sample variation, but that doesn't mean I think that status quo is acceptable; its the principle of the thing: if you buy a 2.8 lens at this price it should be usable without compromise at 2.8!
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions for most, so best to future proof for those rather than compromising just because current reflective media destroys some detail.
Might be better to wait until you're out of lockdown so you can test the next sample in-store...
Cheers
Peter
Aug 26, 2020 at 05:24 AM
p.10 #17Sony FE 12-24mm f/2.8 GM Review
Petegh wrote:
These are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...I realise that this is disappointing, genuine question - how much would this affect real life use at say f5.6-f8It is still apparent at ~F7.1 where I like to focus stack, but it probably wouldn't show in prints by then. But I would want this lens to be good wide open for astro anyway, so this copy is out of the question for me. I will be very disappointed if the sample images I sent to the seller, which are the same as the ones I posted above, go to Sony for assessment and come back as being considered "in spec".I feel your pain Richard. IMHO, that's not an acceptable amount of variation if it exists throughout the focal length range. I know I said earlier in this thread that paying more money for a lens doesn't guarantee less sample variation, but that doesn't mean I think that status quo is acceptable; its the principle of the thing: if you buy a 2.8 lens at this price it should be usable without compromise at 2.8!
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions...Show more →
Thanks Peter, it is indeed disappointing to see a lens perform like this at this high price. And whilst I don't like it, I do concede that it may be somewhat inevitable. Presumably the super high IQ that can be obtained with both the Sony and the Sigma UWA zooms would not be possible without their complex designs with many elements. But that accordingly affords more room for manufacturing defects to manifest. So in that sense advanced expensive designs may be prone to more variation, rather than less. Hopefully that is offset by more stringent quality control for those lenses so that we still end up with a reasonable number of 'good copies' in the hands of fussier users. I doubt we'll stop seeing this trend ay time soon, with new lenses always on the edge of what is possible in terms of manufacturing processes.
Luckily my seller has agreed that my GM zoom is indeed unacceptable, and is already in the process of replacing it for me. Hopefully I'll have a good copy by the time our lock-down ends...
rvh23 wrote
johnahill wrote
rvh23 wrote
This is the part I dislike most about evaluating these complex zooms. I received my GM 12-24 today and it clearly failed its centering test at all focal lengths with a soft lower left corner across the whole rangeThese are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions...Show more →
Thanks Peter, it is indeed disappointing to see a lens perform like this at this high price. And whilst I don't like it, I do concede that it may be somewhat inevitable. Presumably the super high IQ that can be obtained with both the Sony and the Sigma UWA zooms would not be possible without their complex designs with many elements. But that accordingly affords more room for manufacturing defects to manifest. So in that sense advanced expensive designs may be prone to more variation, rather than less. Hopefully that is offset by more stringent quality control for those lenses so that we still end up with a reasonable number of 'good copies' in the hands of fussier users. I doubt we'll stop seeing this trend ay time soon, with new lenses always on the edge of what is possible in terms of manufacturing processes.
Luckily my seller has agreed that my GM zoom is indeed unacceptable, and is already in the process of replacing it for me. Hopefully I'll have a good copy by the time our lock-down ends...
Aug 26, 2020 at 06:16 AM
p.10 #18Sony FE 12-24mm f/2.8 GM Review
rvh23 wrote:
These are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...I realise that this is disappointing, genuine question - how much would this affect real life use at say f5.6-f8It is still apparent at ~F7.1 where I like to focus stack, but it probably wouldn't show in prints by then. But I would want this lens to be good wide open for astro anyway, so this copy is out of the question for me. I will be very disappointed if the sample images I sent to the seller, which are the same as the ones I posted above, go to Sony for assessment and come back as being considered "in spec".I feel your pain Richard. IMHO, that's not an acceptable amount of variation if it exists throughout the focal length range. I know I said earlier in this thread that paying more money for a lens doesn't guarantee less sample variation, but that doesn't mean I think that status quo is acceptable; its the principle of the thing: if you buy a 2.8 lens at this price it should be usable without compromise at 2.8!
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions for most, so best to future proof for those rather than compromising just because current reflective media destroys some detail.
Might be better to wait until you're out of lockdown so you can test the next sample in-store...
Cheers
PeterThanks Peter, it is indeed disappointing to see a lens perform like this at this high price. And whilst I don't like it, I do concede that it may be somewhat inevitable. Presumably the super high IQ that can be obtained with both the Sony and the Sigma UWA zooms would not be possible without their complex designs with many elements. But that accordingly affords more room for manufacturing defects to manifest. So in that sense advanced expensive designs may be prone to more variation, rather than less. Hopefully that is offset by more stringent quality control for those lenses...Show more →
Congratulations!
The worst words I'm afraid is "it's in SPEC" from local service center
oh, no, something could be even worse " my copy in my lab is worse than yours, yours is not bad". I really heard it when i wanted to exchange my 1st copy of 35/1.4 ZA
Petegh wrote
rvh23 wrote
johnahill wrote
rvh23 wrote
This is the part I dislike most about evaluating these complex zooms. I received my GM 12-24 today and it clearly failed its centering test at all focal lengths with a soft lower left corner across the whole rangeThese are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions for most, so best to future proof for those rather than compromising just because current reflective media destroys some detail.
Might be better to wait until you're out of lockdown so you can test the next sample in-store...
Cheers
Peter
Congratulations!
The worst words I'm afraid is "it's in SPEC" from local service center
oh, no, something could be even worse " my copy in my lab is worse than yours, yours is not bad". I really heard it when i wanted to exchange my 1st copy of 35/1.4 ZA
Aug 26, 2020 at 07:06 AM
p.10 #19Sony FE 12-24mm f/2.8 GM Review
rvh23 wrote:
These are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...I realise that this is disappointing, genuine question - how much would this affect real life use at say f5.6-f8It is still apparent at ~F7.1 where I like to focus stack, but it probably wouldn't show in prints by then. But I would want this lens to be good wide open for astro anyway, so this copy is out of the question for me. I will be very disappointed if the sample images I sent to the seller, which are the same as the ones I posted above, go to Sony for assessment and come back as being considered "in spec".I feel your pain Richard. IMHO, that's not an acceptable amount of variation if it exists throughout the focal length range. I know I said earlier in this thread that paying more money for a lens doesn't guarantee less sample variation, but that doesn't mean I think that status quo is acceptable; its the principle of the thing: if you buy a 2.8 lens at this price it should be usable without compromise at 2.8!
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions for most, so best to future proof for those rather than compromising just because current reflective media destroys some detail.
Might be better to wait until you're out of lockdown so you can test the next sample in-store...
Cheers
PeterThanks Peter, it is indeed disappointing to see a lens perform like this at this high price. And whilst I don't like it, I do concede that it may be somewhat inevitable. Presumably the super high IQ that can be obtained with both the Sony and the Sigma UWA zooms would not be possible without their complex designs with many elements. But that accordingly affords more room for manufacturing defects to manifest. So in that sense advanced expensive designs may be prone to more variation, rather than less. Hopefully that is offset by more stringent quality control for those lenses...Show more →
I know it can be frustrating Richard and I'm glad you will be able to try a second copy. The visual discrepancy showing corner asymmetry is amplified by our high megapixel bodies. Therefore, centering tests are less forgiving on the A7R IV compared to lower megapixel bodies.
The higher the lens' number of elements and production complexity, the higher the copy variance. I'm not surprised the FE 12-24/2.8 GM does not have low variance and it seems to be on par with other GM and G lenses regardless of price. Unfortunately the Sony 12-24/4 G and Sigma 14-24/2.8 DG DN's variance is not any better.
My loan copy was surprisingly centered for a ultra wide zoom, so there are good copies out there.
Petegh wrote
rvh23 wrote
johnahill wrote
rvh23 wrote
This is the part I dislike most about evaluating these complex zooms. I received my GM 12-24 today and it clearly failed its centering test at all focal lengths with a soft lower left corner across the whole rangeThese are 100% crops on A7r4, using manual corner focus for one of the corners and then no adjustmeent for the other corners.
http://www.richardvanhoesel.com/test/GMcorners12.jpg
http://www.richardvanhoesel.com/test/GMcorners14.jpg
http://www.richardvanhoesel.com/test/GMcorners24.jpg
Being in Australia, I've been told the provider of the lens may refer to Sony support to decide whether this is reasonable 'within spec' performance for a US$3000 lens...
Also, whilst I think prints will always be around to some degree, I think the future of image display is 8-12K digital picture frames or televisions for most, so best to future proof for those rather than compromising just because current reflective media destroys some detail.
Might be better to wait until you're out of lockdown so you can test the next sample in-store...
Cheers
Peter
I know it can be frustrating Richard and I'm glad you will be able to try a second copy. The visual discrepancy showing corner asymmetry is amplified by our high megapixel bodies. Therefore, centering tests are less forgiving on the A7R IV compared to lower megapixel bodies.
The higher the lens' number of elements and production complexity, the higher the copy variance. I'm not surprised the FE 12-24/2.8 GM does not have low variance and it seems to be on par with other GM and G lenses regardless of price. Unfortunately the Sony 12-24/4 G and Sigma 14-24/2.8 DG DN's variance is not any better.
My loan copy was surprisingly centered for a ultra wide zoom, so there are good copies out there.
Aug 26, 2020 at 10:50 AM
p.10 #20Sony FE 12-24mm f/2.8 GM Review
Lens looks great but throwing 3k at it with what we seen so far with some issues im just not sure. Ive done this a thousand times it seems and been rewarded and been killed too. I usually have something to say about buying or not but im stuck here. Fred I honestly dont think its worth giving up that Sigma 14-24. Yea not as brilliant but not bad at all. You need 12 than its a good argument for it but certainly want to throw your money in the right corner. Seriously im a little baffled by this one.
Aug 26, 2020 at 11:16 AM