NatDeroxL7 wrote:
I had a good overnight recently where I got to try my 12-24 on some stars. I did like it better than my previous Sigma 14-24 DG DN. Slightly less coma and little less fringing on the stars. 12mm also adds a little extra dynamic to the scene with a incrementally larger expanse of sky. However, 12mm also makes it very obvious when objects are near the edge of the frame and show perspective distortion from an un-level camera. Had to be extra careful to avoid this.
Got some cloud cover initially (below, still kind looks cool) but eventually I got a clear window.
I forgot to turn off the in camera vignetting correction on my second camera body and unfortunately that one ended up being used for the 12-24 GM so it has the red color shift rings issue....ugh. Was correctable to an extent, and I fixed the issue so next time both cameras are properly set to mitigate this issue as much as possible. its hard to make a perfect master flat frame @ 12-18mm to fix this, especially at higher apertures because the DOF makes details come out AND its very hard to achieve perfectly even illumination. Still working on finding a solution for that. This would hopefully provide total rectification of this issue for critical images using Adobe 'flat field' correction. I may go full-astro mode and buy a flat frame LED light setup.
Interestingly the red colour ring issue is worse with the shading correction turned off with the 12-24GM. I've always used all corrections off for astro, with all lenses, but something is seemingly baked into the files with the 12-24GM regardless of corrections on vs off (other Sony lenses reportedly do this too, including the 20mm f1.8). It's fairly pronounced at 12mm, a little better at 13mm, and not too bad by 14mm. You'll only see it when pushing files fairly heavily though. I do wonder if covering the lens contacts would stop this happening...
I've finally managed to do an in-depth comparison of the 12-24GM vs the Sigma 14-24 DG DN for Milky Way images, and star shape is better on the Sigma at the wide end; stars are round corner to corner whereas they lose shape in some areas of the frame with the Sony (if corners are focused for round stars, some mid-field stars are slightly stretched, and vice versa). The GM is still superb for astro, but the Sigma 14-24 really is a monster at the wide end. Slightly better light transmission with the Sony, but the 'ring issue' and the mild star deformation is enough to put it behind the Sigma for astro. I didn't expect to be writing that, having owned the Sigma for a while but recently 'upgrading' to the Sony. I'll be keeping both as I shoot both at once. 12mm is nice for Milky Way timelapses too.
TimWildAstro wrote:
Interestingly the red colour ring issue is worse with the shading correction turned off with the 12-24GM.
This is awful news. Sony has got to fix this.
I really want to purchase this lens in particular, but I think I'm done with Sony until they make it possible to truly disable shading/vignette correction.
There have been numerous threads about this, and yes you're correct--taping the contacts solves the problem, at the expense of losing AF, exif, and the corrections you might want that are tied to it (like distortion correction.)
Fred Miranda, how about running with this Sony shading/vignetting madness on the front page? This is much, much worse than star eater, lossy compression artifacts, or most of the other image processing bugs and stupidity we see from other vendors.
Lots of threads at FM and DPR about it. Snapsy has done a deep dive.
TimWildAstro wrote:
Interestingly the red colour ring issue is worse with the shading correction turned off with the 12-24GM. I've always used all corrections off for astro, with all lenses, but something is seemingly baked into the files with the 12-24GM regardless of corrections on vs off (other Sony lenses reportedly do this too, including the 20mm f1.8). It's fairly pronounced at 12mm, a little better at 13mm, and not too bad by 14mm. You'll only see it when pushing files fairly heavily though. I do wonder if covering the lens contacts would stop this happening...
I've finally managed to do an in-depth comparison of the 12-24GM vs the Sigma 14-24 DG DN for Milky Way images, and star shape is better on the Sigma at the wide end; stars are round corner to corner whereas they lose shape in some areas of the frame with the Sony (if corners are focused for round stars, some mid-field stars are slightly stretched, and vice versa). The GM is still superb for astro, but the Sigma 14-24 really is a monster at the wide end. Slightly better light transmission with the Sony, but the 'ring issue' and the mild star deformation is enough to put it behind the Sigma for astro. I didn't expect to be writing that, having owned the Sigma for a while but recently 'upgrading' to the Sony. I'll be keeping both as I shoot both at once. 12mm is nice for Milky Way timelapses too.
These were with all in-camera corrections off, one at 12mm, one at 13mm and one at 14mm.
I've included some jpegs showing the files very heavily pushed for illustration purposes only. You'll notice that the red ring issue is significantly worse at 12mm than 13mm and 14mm. In the real World, the rings will rarely show up, but for astro it will be an issue when pushing files, especially at 12mm.
goto_dengo wrote:
This is awful news. Sony has got to fix this.
I really want to purchase this lens in particular, but I think I'm done with Sony until they make it possible to truly disable shading/vignette correction.
There have been numerous threads about this, and yes you're correct--taping the contacts solves the problem, at the expense of losing AF, exif, and the corrections you might want that are tied to it (like distortion correction.)
Fred Miranda, how about running with this Sony shading/vignetting madness on the front page? This is much, much worse than star eater, lossy compression artifacts, or most of the other image processing bugs and stupidity we see from other vendors.
Lots of threads at FM and DPR about it. Snapsy has done a deep dive.
It's still an excellent lens, but I would now recommend the Sigma 14-24 DG DN over the Sony if astro is a main use. For landscapes etc the red ring issue will rarely show up, and the Sony is the better lens in most circumstances except when really pushing files hard (like with astro).
I’ll have to try taping the contacts sometime. In reality, doing Astro is the only time it really matters and that is all manual anyhow so aside from Lightroom not having all the metadata it isn’t a big deal. I would like to see more user-choice and transparency on what is happening behind the scenes with their RAW files though.
Yeah I've just bought some low-residue tape to give it a go.
A little update: I've just had a play around with taking some flat frames, and doing the flat-field correction in Lightroom does a pretty good job of removing the rings, even the really obvious one at 12mm. It'll be a pain in my workflow as I stack a lot of frames for astro images, but at least it's a workable solution. Still, Sony should not force us into this situation.
TimWildAstro wrote:
Yeah I've just bought some low-residue tape to give it a go.
A little update: I've just had a play around with taking some flat frames, and doing the flat-field correction in Lightroom does a pretty good job of removing the rings, even the really obvious one at 12mm. It'll be a pain in my workflow as I stack a lot of frames for astro images, but at least it's a workable solution. Still, Sony should not force us into this situation.
How did you set up and take the flat frames, I was having a hard time finding something evenly lit across an entire 12mm FOV. It’s so huge ha
Just purchased one. Crossing fingers for a good copy. I have tried 2 copies of the Sigma 14-24mm f/2.8 DG DN Art on my Sony a7R IV and wasn't impressed. The first one was decentered in two corners. The other one was very consistent at any focal length and aperture, but it just never seemed to reach excellent sharpness but just very good. Maybe it's just high expectations from such a wide lens, but my Voigtländer 21mm f/1.4 clearly reaches another level at 21mm. Didn't have anything to compare at wider focal lengths.
Frederik0711 wrote:
Just purchased one. Crossing fingers for a good copy. I have tried 2 copies of the Sigma 14-24mm f/2.8 DG DN Art on my Sony a7R IV and wasn't impressed. The first one was decentered in two corners. The other one was very consistent at any focal length and aperture, but it just never seemed to reach excellent sharpness. Maybe it's just high expectations from such a wide lens, but my Voigtländer 21mm f/1.4 clearly reaches another level at 21mm. Didn't have anything to compare at wider focal lengths.
No your expectations of possible IQ with the Sony or Sigma aren't too high, but the variance of individual copies in my experience makes it extremely hard to find a copy that comes close to that potential at all focal lengths. The Sony I eventually decided to keep is still a little asymmetric between 12-14mm, but nevertheless sharper than any WA/UWA prime I've ever owned, including Loxia 21, Loxia 25 and Voigtlander 21/1.4.
TimWildAstro wrote:
Another good copy first try here. The corners aren't 100% equal but they're very, very close, at all focal lengths. I bought this primarily for astro & landscapes (I have a 14-24 DG DN and 24mm GM too) and it certainly does an excellent job with both. Slight star bloat in the corners at 12mm but they're still nice and round.
A nice surprise was the sharpness at 24mm & f2.8 - it outperforms both of the 24mm GM lenses I've had at f2.8. Sharper in the centre, sharper mid-frame, but a tiny bit softer in the extreme corners (tested on an A7Riv).
The only drawback I've found in real-world use is that optimal focus for stars and landscapes etc is within the infinity zone on the electronic readout. For my 14-24 DG DN and 24mm GM there's a specific number where they perform best for infinity focus, but I'm going to have to nail it manually every time with the 12-24 GM. That'll cost me some time in the field with astro, but it's not a big problem really.
Superb lens....Show more →
A late reply to your post. I do a lot of astroscapes and getting sharp stars without aberrations is really tough at wide apertures. I tried the 16-35GM and it wasn't that good (it was also decentered after 24mm so I returned it).
Between the Sigma and the Sony zooms, who is the star performance around 14-15mm?
zuru wrote:
A late reply to your post. I do a lot of astroscapes and getting sharp stars without aberrations is really tough at wide apertures. I tried the 16-35GM and it wasn't that good (it was also decentered after 24mm so I returned it).
Between the Sigma and the Sony zooms, who is the star performance around 14-15mm?
After some more nights out and a bit of pixel peeping, the Sigma 14-24 DG DN is better at 14mm for astro than the Sony is in the 12-14mm range (with my copies at least). Taken with an A7Riv and viewed at 50%, both look superb. Zoom to 100% and the Sony exhibits slight stretching on stars midframe if corner focused, and stretching in the corners if focused for midframe. With my Sigma at perfect focus, pretty much all stars are round corner to corner at 14mm (focus accuracy is absolutely critical though, but the Sigma is easier than the Sony there too).
The above is mainly relevant if you're using a tracker though; if you're doing untracked images with slight trailing, then the tighter stars from the Sigma are no longer as beneficial. At 14mm, the Sony has slightly less vignette, and seems to have very slightly better light transmission. For tracked astro, the Sigma is slightly better than the Sony at the wide end.
TimWildAstro wrote:
I don't use a tracker but do stack images so I stay well below visible trailing (using the stricter NPF rule). So I take it that in this case, the Sigma would be better?
Before I moved to Sony for astro, I had a dedicated prime for MWs in the Fuji system. I'd love to have something like that for Sony but the choices aren't great. (Venus Laowa is not that good, and the other choice would be the Batis 18mm, but wider would be better).
Being a prime only shooter I'm beginning to question myself it the 12-24 GM might be the best choice for an ultra wide. I'm eagerly waiting for the confirmation/release of the 14mm 1.8 GM, but if the price is around half of the 12-24 and performance not as stellar as this one I might add my first zoom to the collection.
Thanks a lot for the hard work putting together all of the amazing reviews.
In this case even 12mm wasn't wide enough and I had to stitch two 12mm image stacks to get the AOV I wanted. Not long ago I thought I would virtually never want a lens shorter than 14mm for landscapes. But this lens has certainly changed that. At this rate, I might even try a Laowa 9mm sometime...
I'm looking to buy a "brand new in box" 12-24GM at a decent discount from a local seller later today. If it is as good as I hope, it will replace my Laowa 12mm, and 16-35GM. But after reading this thread I'm nervous as hell that I'll be stuck with a decentered or soft copy that I can't return.
As this is an in-person meet-up, the seller will give me 15 minutes max to inspect and test it on my A7RIV before buying it. I've seen Fred's decentering test, but given that I'll be in person with the seller, I'm not sure I'm comfortable inverting the camera handheld with his BNIB lens mounted, so I'll probably pick a distant subject and compare the 4 corners at f2.8 normally, then chimping on the LCD or EVF.
Any other quick and dirty tests that you folks would recommend?
Good luck! Just sticking a distant object in each corner of the frame then checking on the screen should be good enough without rotating it etc. If you get time, I'd check with both corner and centre focusing (field curvature). I'd check various focal lengths just to be safe. Mine is absolutely spot on, pretty much perfectly centered at all focal lengths. Fingers crossed the one you're looking at is too.
TimWildAstro wrote:
Good luck! Just sticking a distant object in each corner of the frame then checking on the screen should be good enough without rotating it etc.
Thanks for suggestion.
I was planning to pick a distant subject (like a sign with some text), focus it in the centre at f2.8, then switch to MF and rotate the camera shooting the same subject in each of the 4 corners. I've done this when buying used lenses from local sellers before, and it's fairly quick and reliable.
Just so I understand, your suggestion instead, is not rotating and just shooting a single image and comparing all 4 corners (on a flat/uniform subject like a building or wall I assume)?
