Sigma Art 28mm F1.4 DG HSM Optical Performance?

Hi,

i do not own a Sigma 28mm 1.4 Art. But I own a Canon 24mm 1.4 L and used it last autumn (fall?) for some shots of the milky way while I was in vacation. I can tell, that the stars in the corners looked nearly the same as yours. As you can see in your sample shots, the abberation gets better when stopping down the lens. This is exactly what I experienced back then. So I guess, your lens is behaving quite as usual. Those wide angle lenses usually have their problems at those high aperture settings and are quite soft in the corners as well. I know, the Sigma lens may be newer and maybe even better than the Canon one, but I assume, my lens is not that bad at all. 

You have to remember, that you are using your lens with the focus at the infinity setting. But a wide angle lens might be used with some subjects in the foreground with the focus set to some nearer distance. I assume, this is the distance the lens designers optimzed the optics for. The infinity setting might be a compromise here. With telescopes, you have other use cases and optimization might be much easier assuming you only use the lenses at infinity.

I might be wrong on this, but in general there should be some truth in my words.

CS
Christian

It should be performing decently at f/5.6 and with an APS-C at around f/4. And yes, there are better lens out there.

this is my Canon RF 24-70 L lens aberrations at F/2.8 (70mm full zoom):


and my lens costed $2,400 

Luka Poropat:

It is clear my artificial star test is greatly exagerating aberations, but I am concerned the sigma is showing a higher level of aberation than it should?

Can you share the full image instead of just the corners.

For instance this is my Sigma 28 DG HSM on APSC sensor wide open at F1.4. 
I am more than satisfied with my copy and all the "issues" can be corrected in software like BlurX.

On APS-C sensors, the corners should look much better which they do on Lukas image. But if you look closely, the aberations are still there. On full frame sensors, the quality decreases rapidly in the field that is missing in the APS-C image. So I am still thinking that the Art lens is ok. The images 2-5 in the first post have a really high contrast and therefore the stars look even worse. They almost look like sharpened this way. If you look at the first image where you can see the noise, it looks smoother (if you can say that). Going from the corners of the APS-C sensor to the corners of the full frame sensor, the look in James first image is to be expected. At least I personally feel this way. And this is what I saw in my own images as well.

Kay Ogetay:
Not certain about your issue, but there is an optical difference between the HSM and DN models and DN is known to perform better. I wouldn't be too surprised if HSM gave me this result in FF.

Thanks, I have read this in a few places. Unfortunately the DN models are not available for canon, and I didnt want a DN model because they cannot be used with astro cameras.

Christian Großmann:

Luka Poropat:

It is clear my artificial star test is greatly exagerating aberations, but I am concerned the sigma is showing a higher level of aberation than it should?

Can you share the full image instead of just the corners.

For instance this is my Sigma 28 DG HSM on APSC sensor wide open at F1.4. 
I am more than satisfied with my copy and all the "issues" can be corrected in software like BlurX.

On APS-C sensors, the corners should look much better which they do on Lukas image. But if you look closely, the aberations are still there. On full frame sensors, the quality decreases rapidly in the field that is missing in the APS-C image. So I am still thinking that the Art lens is ok. The images 2-5 in the first post have a really high contrast and therefore the stars look even worse. They almost look like sharpened this way. If you look at the first image where you can see the noise, it looks smoother (if you can say that). Going from the corners of the APS-C sensor to the corners of the full frame sensor, the look in James first image is to be expected. At least I personally feel this way. And this is what I saw in my own images as well.

Thanks, I just posted some real images above. Do you have a similar image taken on FF?

Luka Poropat:

It is clear my artificial star test is greatly exagerating aberations, but I am concerned the sigma is showing a higher level of aberation than it should?

Can you share the full image instead of just the corners.

For instance this is my Sigma 28 DG HSM on APSC sensor wide open at F1.4. 
I am more than satisfied with my copy and all the "issues" can be corrected in software like BlurX.

Do you have any taken with FF and sigma28?