Introduction
And so again, SIGMA has gone where no man has gone before… and releasing the worlds fastest 14mm lens for a full frame mirrorless camera that is not a fisheye.
When the engineers designed this lens, they had astrophotography primary in mind and tried their best to address and fix some of the issues astroscapers usually deal with. Attributes like sharpness corner to corner, contrast, optical distortion, chromatic aberration, coma and astigmatism are among other things that we have to deal with. Some issues are fixable in postprocess, but some of are something you just have to live with.
In reality, most of the issues I have mentioned will actually go unnoticed by most of our viewers since most of us, me included, posts our work on social media, where the algorithms compress the images, making the quality less good. We also usually post our work with much smaller resolution than our cameras can produce. So for the majority of those who love to shoot astrophotography and northern lights, might choose cheaper lens with less good image quality.
This review however will be aimed to amateurs and professionals that are pixel peepers and just want the very best that is out there and fits their needs! And in this case, for Astroscapers.
Since I do not own the Sony 14/1.8 GM lens, nor have I ever used it, I am not going to compare or draw any parallels to that lens. I will just make some notes about it. The SIGMA weights 2.5! times more than the Sony version, but it is "just" a f/1.8 and cost about $1600 when released. Don't be surprised if the SIGMA lens will match that price or even exceed it. But remember, Sigma is no longer just any third-party manufacturer anymore, but a premium one. Those seeking one of the very best wide-angle lenses for astrophotography will not be disappointed!
Disclaimer 1: This lens was lend to me by SIGMA with no strings attached
Disclaimer 2: I don't normally do public reviews. I do test all my gears in a similar fashion before deciding to keep it or not. So if I have missed some important facts/tests, misspelled words or grammar error. I do appologize beforehand.
Disclaimer 3: As I test this lens, summer is upon us in Sweden, and that far up north, we do not get any astronomical nights anymore. The darkest sky I could get was when the sun was just 12° below the horizon, which means the border between civil and nautical twilight. Easier put, the sky is not black, but bluish and the milkyway is barely visible, even with a fast lens like this.
When the engineers designed this lens, they had astrophotography primary in mind and tried their best to address and fix some of the issues astroscapers usually deal with. Attributes like sharpness corner to corner, contrast, optical distortion, chromatic aberration, coma and astigmatism are among other things that we have to deal with. Some issues are fixable in postprocess, but some of are something you just have to live with.
In reality, most of the issues I have mentioned will actually go unnoticed by most of our viewers since most of us, me included, posts our work on social media, where the algorithms compress the images, making the quality less good. We also usually post our work with much smaller resolution than our cameras can produce. So for the majority of those who love to shoot astrophotography and northern lights, might choose cheaper lens with less good image quality.
This review however will be aimed to amateurs and professionals that are pixel peepers and just want the very best that is out there and fits their needs! And in this case, for Astroscapers.
Since I do not own the Sony 14/1.8 GM lens, nor have I ever used it, I am not going to compare or draw any parallels to that lens. I will just make some notes about it. The SIGMA weights 2.5! times more than the Sony version, but it is "just" a f/1.8 and cost about $1600 when released. Don't be surprised if the SIGMA lens will match that price or even exceed it. But remember, Sigma is no longer just any third-party manufacturer anymore, but a premium one. Those seeking one of the very best wide-angle lenses for astrophotography will not be disappointed!
Disclaimer 1: This lens was lend to me by SIGMA with no strings attached
Disclaimer 2: I don't normally do public reviews. I do test all my gears in a similar fashion before deciding to keep it or not. So if I have missed some important facts/tests, misspelled words or grammar error. I do appologize beforehand.
Disclaimer 3: As I test this lens, summer is upon us in Sweden, and that far up north, we do not get any astronomical nights anymore. The darkest sky I could get was when the sun was just 12° below the horizon, which means the border between civil and nautical twilight. Easier put, the sky is not black, but bluish and the milkyway is barely visible, even with a fast lens like this.
Specifications
Product name: SIGMA 14mm f1.4 DG DN I Art
Compatible mounts: L-mount, Sony E-mount
Lens construction: 19 elements in 15 groups (SLD1, FLD3, aspherical lens4)
Number of diaphragm blades: 11 (rounded)
Minimum aperture: f/16
Max magnification ratio: 1:11.9
Dimensions: 101.4*149.9mm (L-mount) 101.4*.151.9mm (E-mount)
Weight: 1170g (L-mount) 1160g (E-mount)
Compatible mounts: L-mount, Sony E-mount
Lens construction: 19 elements in 15 groups (SLD1, FLD3, aspherical lens4)
Number of diaphragm blades: 11 (rounded)
Minimum aperture: f/16
Max magnification ratio: 1:11.9
Dimensions: 101.4*149.9mm (L-mount) 101.4*.151.9mm (E-mount)
Weight: 1170g (L-mount) 1160g (E-mount)
This is the SIGMA 14/1.4 DG DN Art vs Samyang 14/2.4 XP. The Samyang was my lens of choice back in 2017. The SIGMA 14/1.8 DG HDM didn't pass my quality control test while the Samyang ticked most of my boxes.
The new SIGMA weights in at 1160g with the tripod collar and the Samyang 916g without collar but with the MC-11. The latter has also quite a complex optical design with 18 lenses in 14 groups. So I am quite familiar with the size and weight.
The new SIGMA weights in at 1160g with the tripod collar and the Samyang 916g without collar but with the MC-11. The latter has also quite a complex optical design with 18 lenses in 14 groups. So I am quite familiar with the size and weight.
Construction and handling
Large, heavy and built like tank is a very simple way of describing this beast. Outperfoming its predecessor, SIGMA 14/1.8 DG HSM Art lens in size, weight and optical performance.
It weights 1160g (Sony E-mount) and has the dimensions of 101*152mm which means that the setup will become front heavy. SIGMA has resolved this by supplying a tripod foot with it and to my delight, it is Arca-Swiss compatible.
The lens uses an inner focus system meaning, the front lens will not shift position while focusing. It also uses SIGMAs High-response Linear Actuator (HLA) which should in theory be very fast in focusing while shooting fast moving subjects. This is nothing I have tested since my main objective is shooting by night. It supports DMF, AF+MF and also compatible with AF assist function found in Sony camera bodies.
It supports lens aberration correction (camera specific function)... I’m actually going to stop rewriting the spec sheet and just concentrate on things I consider to be significant for my type of work. There will be plenty of this data all over net.
So, yes, it has a manual focus switch and that is what most people will rely on since in pitch black darkness, autofocus will work poorly or not at all. You could illuminate the foreground with a flashlight and then use AF, but then again, I still prefer using MF to better see where to focus.
After achieving focus, SIGMA has put kindly put a manual focus lock (MFL) button, disabling the focus ring of making changes to the focus while shooting longer exposures or rearranging the composition.
It weights 1160g (Sony E-mount) and has the dimensions of 101*152mm which means that the setup will become front heavy. SIGMA has resolved this by supplying a tripod foot with it and to my delight, it is Arca-Swiss compatible.
The lens uses an inner focus system meaning, the front lens will not shift position while focusing. It also uses SIGMAs High-response Linear Actuator (HLA) which should in theory be very fast in focusing while shooting fast moving subjects. This is nothing I have tested since my main objective is shooting by night. It supports DMF, AF+MF and also compatible with AF assist function found in Sony camera bodies.
It supports lens aberration correction (camera specific function)... I’m actually going to stop rewriting the spec sheet and just concentrate on things I consider to be significant for my type of work. There will be plenty of this data all over net.
So, yes, it has a manual focus switch and that is what most people will rely on since in pitch black darkness, autofocus will work poorly or not at all. You could illuminate the foreground with a flashlight and then use AF, but then again, I still prefer using MF to better see where to focus.
After achieving focus, SIGMA has put kindly put a manual focus lock (MFL) button, disabling the focus ring of making changes to the focus while shooting longer exposures or rearranging the composition.
Side view of the buttons. Three buttons are important to me. The first is the Manual focus (MF) switch, the second, the Manuel Focus Lock (MFL) button to lock the focus and the third is the custom button. The custom button could conveniently be assigned for Sonys function called MONITOR BRIGHTNESS. This function is superb for the composition part. It will use high ISO and longer exposure time to better "see" in the darkness. I think this is unique feature for Sony and is not well known function.
Since you can't use front filters, SIGMA has put an option to use rear filters. This can certainly be beneficial for normal landscape photographers, but I haven't seen any specialized filters for astroscape using gel filters. The best thing to use for such a wide lens is actually a didymium filter which are made out of glas. What company will be first to make such a thinn filter for this kind of work? For those who doesn't know what a didymium filter, it is a filter used in the old days to enhance autumn foliage in analog photography. I was one of the first to use them for Astroscape and back then. There weren't many suppliers back then, but now, there are plenty of them and they are sold with names associated to nightsky or lightpollution and advertised to suppress the artificial lightpollution that surrounds us. Didymium filter is a filter which can not be replicated/done in postprocess.
The tripod foot also makes sure that the center of balance is perfect, putting less strain on my small portable tracker and therefor achieving better tracking performance.
Depending where to live/shoot from, dew is one many things Astroscapers have to deal with. To battle the problem, we use dew heaters/straps to make sure that the front elements temperature stays above the dew point. My narrowest dew strap is 5cm wide and I haven't found any narrower than that online. The problem is that it covers the buttons and putting the strap further away will make the strap visible in the field of view and darkening the edges of the image. This is problem since you might alter the focus when trying to switch to MFL. The work around is to push the strap forward, focus, lock and then put it back to the correct position. If there are narrower dew straps for lenses to buy, please let me know.
Image quality
Optical design and image quality - Barrel distorsion
This first test shows a slight barrel distortion, and the horizon is not straight. This will only show when you have clear view of the horizon and is easily fixed in Photoshop. There are currently no lens corrections available in Lightroom, but I am confident that there will be soon.
And here is an image with a small cabin quite near the lens. You can see how the lens distorts all straight lines. This is a trait that is common for wide angle lenses and not really a "defect". This is also quite easy to fix if you do not like the effect.
And this is the final edited image. I have straightened the horizon and fixed the house. I didn't fix it too perfect since I still wanted the wide angle feel these kind of lenses produce. I've also "forced" the milky way to stand out in the sky. I wouldn't normally edit like this on a such bright sky as I had, but this is just to demonstrate how much information the lens can capture in just 30s. This is a two-image composite, 30s tracked and 30s untracked, put together in Photoshop.
Optical design and image quality - Chromatic aberration (CA)
The lens performs very well with respect to aberrations although some purple fringing is visible around the stars. CA normally occurs on high contrast edges so this might be the worst-case scenario and really puts the lens to its pace. What can be higher contrast than a pinpoint white star against a dark sky (well, almost dark sky). Stepping down the aperture will not improve the fringing noticeably.
Manual correction of purple and green fringing inside Lightroom is effective and removes the little fringing there were present before.
Optical design and image quality - Coma
And now comes the part many might have come here to read about, the dreaded COMA!
What is Coma?
"Coma is an aberration which causes rays from an off-axis point of light in the object plane to create a trailing "comet-like" blur directed away from the optic axis (for positive coma). A lens with considerable coma may produce a sharp image in the center of the field but become increasingly blurred toward the edges."
- Georgia State University
More easily put, a bad looking optical flaw that makes my eyes hurt and is quite hard to get rid of in a fast way during postprocess.
So, have SIGMA succeeded in designing a lens without Coma? They have stated before that they have built lenses with Astro in mind. but almost all of them have had issues with this phenomenon.
Let’s take a closer look on their new lens!
What is Coma?
"Coma is an aberration which causes rays from an off-axis point of light in the object plane to create a trailing "comet-like" blur directed away from the optic axis (for positive coma). A lens with considerable coma may produce a sharp image in the center of the field but become increasingly blurred toward the edges."
- Georgia State University
More easily put, a bad looking optical flaw that makes my eyes hurt and is quite hard to get rid of in a fast way during postprocess.
So, have SIGMA succeeded in designing a lens without Coma? They have stated before that they have built lenses with Astro in mind. but almost all of them have had issues with this phenomenon.
Let’s take a closer look on their new lens!
This is an aberration test is used by many astrophotographers that specializes in deep sky objects to review and fix optical flaws they have in the image train. This image shows the center and all four corners and edges in 100% (pixel peeping) at f/1.4. What we are seeing is a slight decentralization of the setup, meaning the lenses are not aligned perfectly. This is very hard to fix and takes time to counteract manually and is not something you fix by yourself on a lens (it is easier on a telescope). This is however nothing that will ever show normal daytime images.
The center of the image has near perfect round stars and the right-hand side shows a slight tendency of coma, the left side however shows more pronounced coma. The top of the image shots almost perfect stars excelt the top left corner which have elongated stars compared to the bottom of the image. Looking how the stars are formed across the field gives us a hint on how the lenselements are tilted. This defects will vary from copy to copy and there's not really much you can do about it. I am not certain how well a technician can fix this problem or even if they have the necessary tools to discover it. I've sent many copies back to the store complaining about thsi kind of problems and the solution has always been by sending me a new copy. Sometimes better, sometimes worse. I am not gonna give out the companys name (not SIGMA), but I sent back 4!!! copies once and the last one was perfect, just like most reviews found online. This is one of the reasons you should buy new lenses instead of used ones, especially if you are concerned about image quality and how it performs at night. As I earlier stated, this kind of defect won't show in normal usage.
The good part is that even if this is visibly my test, I still would say this lens performs very well to excellent considering the nature of being such a fast wide angle lens and I would be much happy if I owned this copy.
I wouldn't mind if SIGMA let me keep this lens for a longer period, so I can really test it when we do have true darkness and the milky way is more visible. This lens would also be superb for the upcoming Perseid Meteor shower in August where I am arranging a star party in south Sweden.
The center of the image has near perfect round stars and the right-hand side shows a slight tendency of coma, the left side however shows more pronounced coma. The top of the image shots almost perfect stars excelt the top left corner which have elongated stars compared to the bottom of the image. Looking how the stars are formed across the field gives us a hint on how the lenselements are tilted. This defects will vary from copy to copy and there's not really much you can do about it. I am not certain how well a technician can fix this problem or even if they have the necessary tools to discover it. I've sent many copies back to the store complaining about thsi kind of problems and the solution has always been by sending me a new copy. Sometimes better, sometimes worse. I am not gonna give out the companys name (not SIGMA), but I sent back 4!!! copies once and the last one was perfect, just like most reviews found online. This is one of the reasons you should buy new lenses instead of used ones, especially if you are concerned about image quality and how it performs at night. As I earlier stated, this kind of defect won't show in normal usage.
The good part is that even if this is visibly my test, I still would say this lens performs very well to excellent considering the nature of being such a fast wide angle lens and I would be much happy if I owned this copy.
I wouldn't mind if SIGMA let me keep this lens for a longer period, so I can really test it when we do have true darkness and the milky way is more visible. This lens would also be superb for the upcoming Perseid Meteor shower in August where I am arranging a star party in south Sweden.
Stopping down from f/1.4 to f/1.8 (left image) show little to no improvement regarding coma, while it is almost gone at f/2.5 (right image). The tilt is still visible though.
Optical design and image quality - Vignetting
Most reviewers would probably say that they are not impressed with respect to vignette. It is quite strong at f/1.4 with fast improvement stopping down to f/1.8 and considerably much better at f/2.5 but then again considering that we are reviewing a very wide angle lens with super-fast aperture, you really can't ask for too much of the light fall-off. I normally don't mind vignetting with this kind of photography since they mostly draw your attention to the middle of the frame where you usually put the Milky way. And even if I from time-to-time use lens correction to get rid of the dark corners in Lightroom, I tend to make an artificial vignette before posting my images anyway.
As you can see from my aberration (Coma) tests, I've already taken care of the varying luminosity throughout the image.
As you can see from my aberration (Coma) tests, I've already taken care of the varying luminosity throughout the image.
Optical design and image quality - Field curvature
This is actually bad, really bad! There is nothing terribly wrong with the resolution, sharpness nor contrast in the edges.... If you focus in the corners!.
The top image is the entire frame so that you can see how I have cropped and how far the bridge is. The second image is when I focused on the center in the first image, locked focus and then recomposited so the bridge is in all rout corners. I then returned to the center and reshot that composition again to confirm focus hadn't changed. The result speaks for itself. Center is good and the corners are bad.
The third is when I put the bridge in the top left corner of the sensor and focused and then made a new set of recomposition without refocusing. This time, we have acceptable resolution in the corners and a very bad center..
Well, what can you do? Shoot multiple images and focus stack? No, for Astroscape, it all depends on the scenery. For the starfield, I would put the focus point in slightly off center. I hope this is a variance in the copies and not a design flaw.
The top image is the entire frame so that you can see how I have cropped and how far the bridge is. The second image is when I focused on the center in the first image, locked focus and then recomposited so the bridge is in all rout corners. I then returned to the center and reshot that composition again to confirm focus hadn't changed. The result speaks for itself. Center is good and the corners are bad.
The third is when I put the bridge in the top left corner of the sensor and focused and then made a new set of recomposition without refocusing. This time, we have acceptable resolution in the corners and a very bad center..
Well, what can you do? Shoot multiple images and focus stack? No, for Astroscape, it all depends on the scenery. For the starfield, I would put the focus point in slightly off center. I hope this is a variance in the copies and not a design flaw.
In the field
Most people might want to stop down a bit to get rid of some of issues mentioned earlier and achieve more resolution. But for me, the light gathering power is such an important parameter that is outshines all the others. The image comparison above shows the star field at various apertures and the brightness between the f/1.4 and f/1.8 is tremendous. As you can see on all my tests, I've have also chosen to test the lens at f/2.5. This is because I have handpicked a similar, but slower lens years ago for Astroscape, which I, back then found to be one of the best for my needs, the Samyang 14/2.4 XP.
Here you can really see the difference and benefits of having such a fast lens as f/1.4. The left is SIGMA at f/1.4 and on the right, the Samyang at f/2.4. There is a clear difference between these two images in terms of brightness.
Earth rotates on its axis and it takes one day to do a full rotation. This is the reason why stars becomes trails instead of pointy dots on our images if you shoot with a too long exposure time. We take long exposures because of the lack of light and to keep the noise to a minimum. There are of course ways to get around this problem, like stacking multiple images, tracking and merging in photoshop, making a panorama with a longer focal length etc. But some times you just have to settle with single exposures. There is an old popular rule called the 500-rule for full frame cameras. This means that you divide 500 with your focal length to get an exposure time that will be ok and keep the stars pinpoint. This is however false for most modern sensors that are high resolution. So nowadays, 500-rule actually means major trailing. Other derivatives of the rule is the 300-rule and that gives you slight trailing while the 200-rule gives you pin-point accurracy.
So let's put that to the test. According to the rules, SIGMA 14mm would be 35s for major trailing, 21s for slight trailing and 14s for pin-point stars.
The left image above is a 5s test shot while the right one is a 10s test shot. If you look very closely, you can see that trailing is already occuring at 10s. Another thing to notice is that the coma has diminished because we now have trails instead. The rules do not comply very well on such a wide angle lens. Remember this when reading other reviews on other wide angle lenses about Coma and CA shot without trackers and exposure times above 5s.
So let's put that to the test. According to the rules, SIGMA 14mm would be 35s for major trailing, 21s for slight trailing and 14s for pin-point stars.
The left image above is a 5s test shot while the right one is a 10s test shot. If you look very closely, you can see that trailing is already occuring at 10s. Another thing to notice is that the coma has diminished because we now have trails instead. The rules do not comply very well on such a wide angle lens. Remember this when reading other reviews on other wide angle lenses about Coma and CA shot without trackers and exposure times above 5s.
And here you see the traiing at 15s on the left and 25s on the right. I would'nt use either for my starfield shots without using a tracker. But for timelapse, northern lights and meteor shower, I would definitely use 15s exposures. And deal with the trailing during postprocess.
This shot is a single exposure without any aid of anything more than a tripod. Shot with Samyang 14/2.4 @f/2.4 during total darkness. Imagine reshooting this scenery with the SIGMA 14/1.4 DG DN Art gives me the chills and even more exciting if I would to track the sky with my star tracker and get even better image. This is just a 30s exposure and f/2.4 and if I was to shoot at f/1.4, it would mean that more than three! times more light would be hitting the sensor. I would guess that even with an unmodded camera, a bit redder nebulosity would shine through. You can already hint some nebulosity in this image, and this is shot with a stock Sony camera (unmodified).
I'll be revisiting this place in august with a high-resolution Sony camera, Sony A7R4, which has been astromodified. We'll see if I have can get hold of this lens when the time comes, recapturing a similar scene during the Perseid meteor shower.
I'll be revisiting this place in august with a high-resolution Sony camera, Sony A7R4, which has been astromodified. We'll see if I have can get hold of this lens when the time comes, recapturing a similar scene during the Perseid meteor shower.
Final thoughts
What I like:
* Superb build quality
* Impressive image quality wide open across the entire frame
* Aperture f/1.4
* Manual focus lock
* Tripod collar
* Superb build quality
* Impressive image quality wide open across the entire frame
* Aperture f/1.4
* Manual focus lock
* Tripod collar
What I dislike:
* Big and heavy
* Some vignette (expected though)
* Not able to use front filters
This lens is truly very interesting and enjoyable to use. It has many features an astroscaper like me wants to have and the optical image quality is really good. It outperforms my Samyang 14/2.4, which I have taken many wonderful images with and if money was of no concern, I wouldn't hesitate having a copy of my own. The bright maximum aperture will help supressing the noise and bring out the faint details that normally isn't visible otherwise and by increasing the signal to noise ratio, you also have more headroom when pushing the postprocess to its maximum before the image quality falls apart. The coma is well controlled even in the corners and the little tilt I see will be a normal variation in all copies. There is two clever ways to counteract coma on field and one is shooting at longer exposure times to get slight startrails and the other way is to slightly have it out of focus. This is a technique I actually tested, but not shown in this review.
The true test is something I couldn't do and that is how it performs when darkness is truly upon us. The astronomical night in Sweden ended long ago and we'll have to wait until mid August/September to shoot the milky way.
I end this review with three images in way I normally would have shot and edite for the purpose of posting them to the public.
* Big and heavy
* Some vignette (expected though)
* Not able to use front filters
This lens is truly very interesting and enjoyable to use. It has many features an astroscaper like me wants to have and the optical image quality is really good. It outperforms my Samyang 14/2.4, which I have taken many wonderful images with and if money was of no concern, I wouldn't hesitate having a copy of my own. The bright maximum aperture will help supressing the noise and bring out the faint details that normally isn't visible otherwise and by increasing the signal to noise ratio, you also have more headroom when pushing the postprocess to its maximum before the image quality falls apart. The coma is well controlled even in the corners and the little tilt I see will be a normal variation in all copies. There is two clever ways to counteract coma on field and one is shooting at longer exposure times to get slight startrails and the other way is to slightly have it out of focus. This is a technique I actually tested, but not shown in this review.
The true test is something I couldn't do and that is how it performs when darkness is truly upon us. The astronomical night in Sweden ended long ago and we'll have to wait until mid August/September to shoot the milky way.
I end this review with three images in way I normally would have shot and edite for the purpose of posting them to the public.
The Tree of Åhus. This is just 5s long exposure at f/1.4 for the tree, background and stars. I shot another image at f/2.5 at 20s and refocused on the sand closest to me and then merged the two images to give me a sharp final imaged from near far.
I drove to another location and found this rock in bit out in the water. 30s tracked @f/1.4 for the stars and then refocused and took another 30s for the rock untracked. No need to get sharpness/focus on the water close to me.
Final image to show how the scene looks like when I acieved the darkest sky I could at this time of the year. Like the other two images, the is a two image composite, one tracked at 30s and one untracked for the foreground. I have straightened the horizon and the cabin to counteract the barrel distorsion.