So yeah - the Sigma 50mm kinda sucks

Bought the Sigma 50mm today and am dissapointed. Tried 1 copy at McBain and it was horrible wide open (super soft and crazy CA), then tried a second copy at another location and it was mis focussing lots. Then ended up buying one at another store (not able to try it out in store) and it's soft as well. Thats pretty ****ty odds 3 out of 3 were bunk. Here is a small comparison I did against my Canon 50mm 1.8. The Canon seems sharper all the way to f4.0 which is weird. Also looks like the Sigma is front focussing a tiny bit. Not the best test but I did it with a tripod, auto focus and a remote shutter. Did 2-3 shots for each aperture on both lenses.


FOCUS WAS ON THE RED 16 MARK.

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

instead of going straight to a focus test chart/measure tape. go test it in the real world and see how it performs unless your main subject is actually a ruler. most sigma complaints are user error expecting the quality of a 50L when taking a picture of a focus chart. for that price point you are not going to get perfection. even the 50L isnt perfect but it still produces some excellent quality when used properly. no lens is going to be perfect for you every time, ever lens is different and performs different form the others. you are going to have to either learn that lens to get the most out of it or buy something else because honestly i have seen people go through tons of copies of a sigma 50 never trying it out on anything but a focus chart.

your other option is to mail it in and get it calibrated and then it will work just fine. i know its frustrating but no one ever said photography is easy. its art and it takes patience and mastery of your tools not just your camera but each lens as well

also on a side note on my monitor it does not look like your 50 1.8 is spot on either

Probably the worst object for a focus test....where did the lens actually focus? Was it at 16, 16 and 1/8, 16 and 1/4, etc, etc?

Fact is, with so many potential contrasting lines to focus on, there is absolutely no way to know which one the camera/lens chose. Also, the concave surface is another potential source of error. Did the camera focus on the close edge of the tape, or the center of the tape? Makes a difference with such a narrow DOF.

A lot of faster lenses are a bit softer than their slower alternatives. It's a design compromise.

It looks like similar sharpness to me. Maybe a bit less contrast. The AF thing is a pain, but I'd check AF performance at actual normal usage distances before writing it off.

Brian_R wrote in post #13147632
instead of going straight to a focus test chart/measure tape. go test it in the real world and see how it performs unless your main subject is actually a ruler. most sigma complaints are user error expecting the quality of a 50L when taking a picture of a focus chart. for that price point you are not going to get perfection. even the 50L isnt perfect but it still produces some excellent quality when used properly. no lens is going to be perfect for you every time, ever lens is different and performs different form the others. you are going to have to either learn that lens to get the most out of it or buy something else because honestly i have seen people go through tons of copies of a sigma 50 never trying it out on anything but a focus chart.

your other option is to mail it in and get it calibrated and then it will work just fine. i know its frustrating but no one ever said photography is easy. its art and it takes patience and mastery of your tools not just your camera but each lens as well

also on a side note on my monitor it does not look like your 50 1.8 is spot on either

Completely understandable that real world photos are the real test. I tokk about 20 or so portraits today and in most the focus is slightly off. I am at work right now but will post some later. At 6' away from te subject and AF focussed on the eyes the sharpest point seems to be the bangs so it looks like it's front focussing about 2-3" at that distance. The other copy I tried in the store was laughably soft. I am not copmparing it to the 50L but to a $100 lens so I would hope it would at least be as sharp as that.

mannetti21 wrote in post #13147646
Probably the worst object for a focus test....where did the lens actually focus? Was it at 16, 16 and 1/8, 16 and 1/4, etc, etc?

Fact is, with so many potential contrasting lines to focus on, there is absolutely no way to know which one the camera/lens chose. Also, the concave surface is another potential source of error. Did the camera focus on the close edge of the tape, or the center of the tape? Makes a difference with such a narrow DOF.

Valid points. I tried to focus right the red 16 square. At 1.4 for the sigma it consitently missed focus. Out of about 10 tries I don't think even one was focussed on the 16. Most were at the 17/18 mark. It got better stopped down. I'm going to be exchanging this copy and seeing if another will be better. I am just suprised that all the 3 copies I tried today had issues.

N.O.R.E. wrote in post #13147774
Completely understandable that real world photos are the real test. I tokk about 20 or so portraits today and in most the focus is slightly off. I am at work right now but will post some later. At 6' away from te subject and AF focussed on the eyes the sharpest point seems to be the bangs so it looks like it's front focussing about 2-3" at that distance. The other copy I tried in the store was laughably soft. I am not copmparing it to the 50L but to a $100 lens so I would hope it would at least be as sharp as that.

Valid points. I tried to focus right the red 16 square. At 1.4 for the sigma it consitently missed focus. Out of about 10 tries I don't think even one was focussed on the 16. Most were at the 17/18 mark. It got better stopped down. I'm going to be exchanging this copy and seeing if another will be better. I am just suprised that all the 3 copies I tried today had issues.

If you must start with a focus test like this, I strongly suggest you download a real test chart before you draw any conclusions. Like I said before, you have no way of knowing where the camera intended to focus. The focus sensor extends beyond the focus square that you see in the viewfinder. You may have intended for the 16, but the camera actually chose the 17 (or one of the other many lines in the focus area)...this doesn't mean there was any malfunction, all it means is that the camera cant read your mind.

You need to test on isolated, high contrast lines with a solid background surrounding so the camera doesn't get confused.

My Sigma 50 is brilliant. I'd consider the point that the vendors are now trying to fob off the poorly performing units rather than returning them or having them serviced.

looks fine to me. focus is just off. exchange or get it calibrated

I would say that if you have tried three lenses, from different sources, that your technique is probably the problem. There is actually quite a learning curve to using a lens with a DOF that narrow.

The fact that you used a tape measure for a focus test tells me that you aren't well versed in how a camera's focus system works. Your response to mannetti21 only confirms that you don't understand how the camera focuses.

One main thing to consider is that your camera's focus sensor may not be exactly where the square or dot is in your viewfinder. On most camera's, the marker in the viewfinder is MUCH smaller than the actual focus sensor. Therefore, if you are aiming the center point at the 16, the focus sensor may be covering all the way from 15 to 17. What the camera does is lock on to the area in the sensor area that has the greatest contrast. If you are hoping the camera will focus on the red 16, you are going to be wrong most of the time because there are a lot of black numbers and lines in the area that provide greater contrast against the yellow than the red. Because of this, the camera may choose to lock on to any one of those numbers or lines, even if your focus point is directly on the 16.

This is why a true focus alignment tool will have a large target that is parallel to the sensor. You focus on that because focusing on anything that is angled away from the camera like that is going to be inconsistent at best.

These same problems will also show in real world photos. If you are shooting a person and the camera and lens are front-focusing on the bangs, you might want to consider that the camera may have locked on to eyebrows or an area on the subject's nose. Sometimes, I will focus on an area of the face that is in the same plane as the eye. Like I said, shooting with this type of lens takes PRACTICE. Don't take it back just because you haven't figured it out on the first day.

mannetti21 wrote in post #13147818
If you must start with a focus test like this, I strongly suggest you download a real test chart before you draw any conclusions. Like I said before, you have no way of knowing where the camera intended to focus. The focus sensor extends beyond the focus square that you see in the viewfinder. You may have intended for the 16, but the camera actually chose the 17 (or one of the other many lines in the focus area)...this doesn't mean there was any malfunction, all it means is that the camera cant read your mind.

You need to test on isolated, high contrast lines with a solid background surrounding so the camera doesn't get confused.

I know that the tape was not the best but it was the closest thing I had around. I didn't even think of doing a test until I noticed almost all my initial shots with this lens had teh focus a bit off.

drumnut01 wrote in post #13148062
I would say that if you have tried three lenses, from different sources, that your technique is probably the problem. There is actually quite a learning curve to using a lens with a DOF that narrow.

The fact that you used a tape measure for a focus test tells me that you aren't well versed in how a camera's focus system works. Your response to mannetti21 only confirms that you don't understand how the camera focuses.

One main thing to consider is that your camera's focus sensor may not be exactly where the square or dot is in your viewfinder. On most camera's, the marker in the viewfinder is MUCH smaller than the actual focus sensor. Therefore, if you are aiming the center point at the 16, the focus sensor may be covering all the way from 15 to 17. What the camera does is lock on to the area in the sensor area that has the greatest contrast. If you are hoping the camera will focus on the red 16, you are going to be wrong most of the time because there are a lot of black numbers and lines in the area that provide greater contrast against the yellow than the red. Because of this, the camera may choose to lock on to any one of those numbers or lines, even if your focus point is directly on the 16.

This is why a true focus alignment tool will have a large target that is parallel to the sensor. You focus on that because focusing on anything that is angled away from the camera like that is going to be inconsistent at best.

These same problems will also show in real world photos. If you are shooting a person and the camera and lens are front-focusing on the bangs, you might want to consider that the camera may have locked on to eyebrows or an area on the subject's nose. Sometimes, I will focus on an area of the face that is in the same plane as the eye. Like I said, shooting with this type of lens takes PRACTICE. Don't take it back just because you haven't figured it out on the first day.

Possibly as well even though I regularly use a Canon 35mm f1.4 wide open and a 135mm f2 and have never had any problems with it not focussing properly. I will try this lens out a bit more tomorrow in real life situations and see how I like it. Just the initial impression was a bit dissapointing (especially when I compared it to the cheapest lens Canon makes).

in tests like that your not comparing lenses your just testing how your camera cant decide what to focus on because there is too much to lock onto

I'll retry it tomorrow with a proper test and some actual shots - thanks for the info/advice.

If you continue to have problems with it and you don't have problems with other fast primes with the same body, you might want to try microadjusting if you have a body you can do that with. If not, a simple trip to Sigma will fix it. My Sigma 85 1.4 needed +4 microadjustment, which is really very little. Now it is dead on. I just can't believe that three of them in a row didn't meet your expectations.

My Sigma 50 is superb on my 1D2n and was hopeless on my 7D, but then all lenses were hopeless on the 7D. And yes, at f1.4 the DOF is amazingly thin, theres no way all features of a face in a portrait shot would be in focus at this aperture. I do realise that QA on Sigma products is a bit hit and miss, my first 10-20 was absolutely useless, a piece of junk and even the Sigma engineers accepted that.

Worst subject ever, my eyes can barely focus reliably on a number on a ruler, how do you expect your camera to?