View Full Version : Concerned about my Sigma 70-200
Lightstream
16th of July 2006 (Sun), 08:11
I've been out shooting today with my Sigma 70-200 EX f/2.8 non-DG, but for the first time I'm pairing it with my 350D instead of my 5D, which it usually sits on. The big surprise - it's really soft. And I'm surprised since 1.6X cameras usually bring out the best in lenses!
http://pix.lightrefineries.org/albums/photos/abn.sized.jpg
100% crop of the image, focused on the eyes, and here's the EXIF.
Camera Model Canon EOS 350D DIGITAL
Shooting Mode Aperture-Priority AE
Tv( Shutter Speed ) 1/500
Av( Aperture Value ) 2.8
Metering Mode Evaluative Metering
Exposure Compensation 0
ISO Speed 200
Lens 70.0 - 200.0 mm
Focal Length 70.0 mm
AF Mode One-Shot AF
Parameters Settings Contrast Mid. High
Sharpness Mid. High
Color saturation Mid. High
Drive Mode Continuous shooting
Add that to the fact that I'm shooting with Parameter 1 (some in-camera sharpening already applied) and I'm even more surprised. The horse was standing still, so this is not a movement problem. Plus, shutter speed should be high enough to combat shake. These are some of the better shots that day, most of the action ones are real trash. And I'm really surprised.
I own two other lenses which overlap the 70mm range (17-85 and 24-105 f/4L) and both of them have yielded VERY much sharper results at 70mm in the past. This is definitely not what I expect from those two lenses, if it was those two they'd be drawing fire from me straight away.
The really wierd part is that when I use the Sigma on my 5D, it's fine. It's sharp, it delivers some serious resolving power. When it first arrived, I did my testing with my 5D and pronounced it good.
Or is the 350D's lack of f/2.8 high precision AF sensors finally tripping me up? Maybe I should stick with the 5D.. but I know the 350D is capable of so much more. Or perhaps this is just a bad combo?
Dorman
16th of July 2006 (Sun), 08:30
That's a rather interesting situation Lightstream, you could always send the 70-200 to me for testing? :)
LightRules
16th of July 2006 (Sun), 10:03
Post a few more f2.8 shots with shutter speeds over 1/500th (at least). I don't see anything blatantly bad about this first crop, but it's consistency of results that is key.
Raphael Emond
16th of July 2006 (Sun), 10:03
In my experience, I find using fast aperture glass with the XT is a little tricky. I don't know if it's the AF sensors on the camera that lacks the 2.8 precision or my technique, but I get more blury shots using fast glass. I'd say, 50% comes sharp as tack. But for the rest, it seems to be a guess by the camera.
Lightstream
16th of July 2006 (Sun), 22:36
Yeah, it seems to be pretty consistent delivering this. I shot around 200 frames that day, most of which I can't post (people involved). Kept around 45 of them, but none of them are the amazingly sharp stuff I get from my other glass. It's also interesting to note that my other glass is all f/4 and higher, the Sigma is my first f/2.8.
I may try again with my 5D in the near future, there may be a good reason why Canon implemented the extra-precision focusing sensors.
Also, would continuous drive mode shooting have anything to do with it? Does the 350D operate in 'release priority' mode where they will take the shot even if AF has not been confirmed? (although I did refocus every couple of shots)
mbellot
16th of July 2006 (Sun), 22:46
Was the horse moving? I noticed you were in one-shot focus mode, and I've gotten similar results with my 70-200 IS when in one shot.
I've just chalked it up to subject movement shifting the focus between focus lock and shutter release, since sometimes if I focus several times (trying to time the shot a bit) I get better results.
Also - DOF @ 2.8 is pretty thin, how far were you from the horse?
Raphael Emond
16th of July 2006 (Sun), 22:49
The main problem with 2.8 or faster lens, it's not the AF confirmation. It's that the DOF is so thin, you can't ever try focus/recompose. And the AF must lock on the thing you want in focus. The AF sensors don't always "see" want you want. I find it a little tricky to get full focus image of moving things.
I personnaly don't think it's a lens problem. More the AF on the XT and my technique.
I'm pretty sure that the 1DMkII is better AF wise.
grego
16th of July 2006 (Sun), 22:58
With action, 2.8 on any camera above the Rebel line has these extra sensors that the bodies use when you shoot at 2.8 or faster. The XT does not have these, so that's part of the reason. I think if you tried this lens on the 20D or 30D, you'd have better results like you would on the 5D.
As much as people want to claim that there's not much difference between the 350 and 20/30D, there are things. Not everyone needs them of course. And the camera body is only better if it meets those needs of yours.
Honeybee
16th of July 2006 (Sun), 23:04
I can confirm that shots from my 70-200 are much better on the 20D than the Xt. Until I got the 20D I was actually going to sell the lens. thanks, grego, for an explanation.
Lightstream
17th of July 2006 (Mon), 07:35
Thanks for the replies. It's hard to believe I am actually outgrowing the 350D at this rate, I have always thought of it as a tremendously capable camera - it is! But there are places where it does fall short. If it truly did EVERYTHING, there would be no 20D, 5D or 1-series.
I've never had a f/2.8 lens before apart from my short primes and those are typically used for subjects that don't run around. Friend's daughter was riding the horse and yup.. naturally, it was moving. Around 5-30 meters away when I took the photos.
Think I'll stick with the 5D for this job, I've heard the AI Servo is substantially better (I've heard it called "close to 1-series", though I don't know how credible it is). I'm aware the 5D has the f/2.8 sensors - I've always wondered whether they made a difference, and if they did, I couldn't see it - not for the applications I was using it for. But now I'm starting to believe...
Lightstream
22nd of July 2006 (Sat), 04:08
OK....thread follow-up.
I realize duck shots and gull shots and so on are not necessarily frivolous things. They enable one to quantify the performance of gear during initial shakedown testing before the gear is deployed out into the field. This is not about pixel peeping, but rather about gaining confidence that the gear will do its job if I do mine. Once I feel sure about the gear, it's never tested again unless something (touch wood) catastrophic happens.
100% crop with AF target painted on the gull's face
http://pix.lightrefineries.org/albums/photos/abw.sized.jpg
No postprocessing other than crop.
Camera Model Canon EOS 5D
Shooting Date/Time 7/22/2006 14:47:02
Shooting Mode Aperture-Priority AE
Tv( Shutter Speed ) 1/500
Av( Aperture Value ) 2.8
Metering Mode Evaluative Metering
Exposure Compensation 0
ISO Speed 100
Lens 70.0 - 200.0 mm
Focal Length 175.0 mm
AF Mode One-Shot AF
Drive Mode Single-frame shooting
This is much, much better. I spent the day shooting the 70-200 extensively, I put around 400 frames on it and it is fairly consistent from frame to frame whether at f/2.8 or f/4. Whether the subject is moving has a BIG impact on it (I did a fair bit of AI Servo training with it too). The gull was stationary, having just landed. But I wouldn't worry about that, I test for static shot performance to ensure the lens can deliver the goods and worry about motion later.
Color and saturation in other shots is fairly impressive. I'd grade it on par with my L lenses in this department. However, at the end of the day, even with the improvement seen in the gull shot, it still lacks the ultimate sharpness of both my Ls that makes even a pixel peeper go "wow". Of course, pixel peeping is not the only thing, but it just stops the Sigma short of being a truly exceptional lens like my 24-105 and 17-40. These two will positively *WRING* every last detail from an image. I also believe that the newer Canon L's were designed with the absolutely punishing demands of full frame digital in mind. I think (my speculation) Canon knew what was coming and made the lenses over-spec to cater to things like the 1DsMkII and the 5D. And in the case of their 70-200, they deliberately set out to make a masterpiece professional zoom.
I notice if that I do not pixel peep, the Sigma is a very fine lens indeed.
But as glass has taught me there is nothing truly perfect in life.. we all have compromises, but there are upsides to make up for it - I got this Sigma for slightly less than a new Canon 70-200 f/4L. If I wanted to trade up to the Canon 70-200, no doubt I would be happy. I would also be worse than broke.. :)
grego
22nd of July 2006 (Sat), 04:11
you could trade your 350 + a lens to put towards a new body.
Or you could trade 70-300 + 70-200 and you will then have your 70-200 IS. :)
Lightstream
22nd of July 2006 (Sat), 04:21
Yeah, something like that :) I do love the 70-300's light weight though....
Wait, stop tempting me!! Stop giving me ideas! I shouldn't, I KNOW I shouldn't!! Dang!! :mrgreen:
Actually I quite like the bargain I got from the Sigma. I did find a solution to the softness problem - if I zoom out to 50%, which the 5D's huge amount of MP allows me to do, the image is tremendously sharp. So I guess I should spend the extra $$ on CompactFlash cards and either size or crop as necessary. :D
The only real reason the Sigma's in my lineup is for action shots, HSM and the huge aperture. The 70-300 actually comes pretty close to the Ls, I have no complaints about it.. loved it since day zero, portrait bug or not (interestingly: mine doesn't have the bug even though the serial number is amongst the affected)
lungdoc
22nd of July 2006 (Sat), 10:37
Aren't there easier things to test focus/sharpness on than live animals?
Borderfox
22nd of July 2006 (Sat), 14:53
I have a Sigma 70-200 and shoot polo with it and a 350D. I am over the moon with the quality of it.
Lightstream
22nd of July 2006 (Sat), 20:05
Aren't there easier things to test focus/sharpness on than live animals?
Yes but I hear that over at POTN, if you have been a member for a while, and have not contributed something in the way of a duck shot with a new lens, rumor says they send a whole flock of rabid commando ducks down to your place to beat you up with your lens http://pix.lightrefineries.org/img/crackup.gif
A more technical reason would probably be that AI Servo needs a moving target.. :)
kraterz
24th of July 2006 (Mon), 03:06
My experience with the sigma 70-200/2.8 HSM has been the same. It's a nice lens, well made, etc. but the one I used was not sharp till f/4 or 5.6. At that f-stop I'd rather use a 70-200/4L, which I ultimately ended up buying. However I have come across some sigma 70-200's which are sharper than the one I used. Maybe it is just bad luck.
But as glass has taught me there is nothing truly perfect in life.. we all have compromises, but there are upsides to make up for it - I got this Sigma for slightly less than a new Canon 70-200 f/4L. If I wanted to trade up to the Canon 70-200, no doubt I would be happy. I would also be worse than broke.. :)
Lightstream
24th of July 2006 (Mon), 06:19
Hmm.. you're actually tempting me with that f/4L suggestion, however I can't bring myself to cough up the extra $$ for the tripod collar. I do find the collar very useful when the lens is on a monopod. Maybe I'll stop the Sigma down instead when lighting conditions permit. Then again it always makes me ask, what is the point of fast glass if you have to stop down to get it to perform :( if I wanted a f/4, I'd look at an f/4.
I have a 70-300 that is really close to the performance of the f/4L as well, if I wanted a really sharp image, I'd use that lens. Sharp all the way out to 300, fast, lightweight, plus it has IS.. beautiful!
René Damkot
24th of July 2006 (Mon), 06:33
Do a focus test with the XT. Might be it's backfocussing a bit.
Raphael Emond
24th of July 2006 (Mon), 12:29
Mine is sharp at f/2.8. I'm not afraid to use it wide open. Maybe yours need a little tweaking by Sigma?
Remember that at 200mm and 2.8, having sharp picture with the shallow DOF can be tricky.
Lightstream
24th of July 2006 (Mon), 19:05
Hm, good point, maybe I should get it checked by Sigma. The 350D's been checked in-house and focusing is fine, never let me down before. Thanks for the suggestion :)
nchphoto
25th of July 2006 (Tue), 13:54
I was having the same problem with mine and my Mark II. Except, I sent both of them
in. Teh problem was the camera body. the lens was fine, until I got it back from Sigma after being, ugh, repaired. AF is complete front focusing, Manual is spot on, so I had
to send it back in. I hope they fix it correctly this time around.
Lightchaser
3rd of August 2006 (Thu), 01:48
I have a Sigma 70-200 and shoot polo with it and a 350D. I am over the moon with the quality of it.
Yup, same here. I do quite a bit of equestrian sports and love this setup.
vBulletin® v3.6.12, Copyright ©2000-2012, Jelsoft Enterprises Ltd.