EF 70-300 IS USM - Artifacts or user error or both?

Hi all,

First post and it might be preemptive. I'm a relatively new owner of the EF 70-300 IS USM, using it with a T1i. Both are original purchases. Some of my shots have had roughly diagonal "streaking" from top left to bottom right in the out of focus background. I uploaded three of the bad shots to flickr so I can post them below (not sure what posting limits there are here yet).

I have a couple of thoughts as to what could be happening, one idea gleaned from the threads here, but I wanted to fish first and see if anyone has seen this before and what they think it is. (And what should be done.) Thanks for your thoughts!

Noticeable to a small extent at 500x333, and much more so at any larger size:

1
2
3

Looks like handshake to me (camera moving diagonally left to right or vice-versa).

do you have a filter on the lens?...i think that diagonal background blur has been caused by a filter in the past

That's plausible.

FYI images also look soft, especially considering the last one is at F/10 and a good shutter speed. Filter could perhaps also be causing this.

It's not the DO lens is it?
That just looks like nasty bokeh to me, but maybe there's something to the filter theory. I'd also see if there's any difference with IS off.

Thanks all. I was starting to suspect the filter after reading threads, especially after seeing some comments about "Newton's Rings" and some artifacts that occur most prominently on green light. Though these don't look like Newton's rings. I have a cheapo Canon Philippines filter that I purchased used for about $5 to protect the lens. (Way to devalue my investment I know.) I was associating these streaks with general image softness as well. Sometimes this lens has great, crisp shots, and sometimes they're soft as hell even at a pretty good speed and aperture, as noted by phreeky. And not in a way that matches up with any other tests done on this 70-300. I felt like I was getting some of this effect when I tried removing the filter but I wasn't testing it in a controlled way -- these might have been in lower light situations with my aperture wide open at 300mm with camera shake too great for IS to deal with. I think my first step is to do more controlled testing with and without this filter, and especially with green backgrounds at distances.

My other suspicion is camera shake like TheBurningCrown mentioned from pressing the shutter and my camera moving down some, but in reviewing these photos taken from different scenarios I realized that the lines only seem to appear on green, forested areas.

xarqi, it's not the DO lens (I read about some artifact issues with that). I'll do some more tests with the IS off though, I think I also tried a few test shots that in situations that were too low-light.

Definitely ditch the filter, I have the lens and was initially very unhappy with performance. Once I got rid of the medium priced uv filter the difference was amazing. The lens is a light hog anyway and I see no use for more glass for it to have to shoot light thru.

All who guessed the filter seem to be right. I did a series of tests including IS vs. non-IS, w/ filter vs. w/out-filter, different shutter speeds (translating a lot of camera shake or not), 190mm vs. 300mm, and the only constant that produced the streaking was:
a) the filter being on
b) the background being not 100% out of focus or close to in focus, and being verdant

I also tried low shutter speed vs. high shutter speed and it was there with more and less camera shake translated to the image.

I tested it on a treed background vs. a brick building background with a green hedge underneath it. The treed background showed the streaking 100% of the time with the filter on and the blur being within a certain range. The brick building didn't show the streaking at all but the hedge below it did under the same circumstances.

DavidE I'm inclined to agree with you. I got the filter to begin with -- ending my brief 1-month long vow not to use filters just to protect my lenses -- after I dinged a tiny bit of coating off this lens, but a filter's not worth the sacrifice in performance and I don't know if it's worth the $100+ I keep hearing about for 100% clear glass covers, which might give me the same effect if it is an interaction between the filter at that point in space and lens elements, like with Newton's Rings.

Thanks again all! Canon was ready for me to send in the lens for servicing based on my brief description to them and this will save a lot of time and trouble, and shipping costs. I just wish I had bothered to try and figure this out before.

figgen wrote in post #10450558
All who guessed the filter seem to be right. I did a series of tests including IS vs. non-IS, w/ filter vs. w/out-filter, different shutter speeds (translating a lot of camera shake or not), 190mm vs. 300mm, and the only constant that produced the streaking was:
a) the filter being on
b) the background being not 100% out of focus or close to in focus, and being verdant

I also tried low shutter speed vs. high shutter speed and it was there with more and less camera shake translated to the image.

I tested it on a treed background vs. a brick building background with a green hedge underneath it. The treed background showed the streaking 100% of the time with the filter on and the blur being within a certain range. The brick building didn't show the streaking at all but the hedge below it did under the same circumstances.

DavidE I'm inclined to agree with you. I got the filter to begin with -- ending my brief 1-month long vow not to use filters just to protect my lenses -- after I dinged a tiny bit of coating off this lens, but a filter's not worth the sacrifice in performance and I don't know if it's worth the $100+ I keep hearing about for 100% clear glass covers, which might give me the same effect if it is an interaction between the filter at that point in space and lens elements, like with Newton's Rings.

Thanks again all! Canon was ready for me to send in the lens for servicing based on my brief description to them and this will save a lot of time and trouble, and shipping costs. I just wish I had bothered to try and figure this out before.

Hey Figgen,

I myself own the EF 70-300mm IS, and never had any trouble with it, so i decided to do a little test, and put my EOS 350D with lens attached on my tripod (turned IS off) and shot two pictures with the exact same aperture/exposure/iso, AF at the exact same spot. The only difference was that in between the shots i took the UV-filter off (i have been using UV filters ever since i bought my dslr to protect my lenses). I found absolutely no difference between my picture with uv-filter on and the one without the filter. The filter i have on their is a HOYA (58mm). I bought it at Black's photography (in Canada) and it cost me 49.99 CAD (excl. tax). I do not know if you would like to spend that amount of money on a filter, but for me it is worth it for protecting the lens (And i am in no way rough with my camera, but just want to be on the safe side).

Cheers

Thanks Nederlander, that's encouraging. I spent about $70 for a filter for my more expensive lens, so I'm definitely not opposed to it all together, and it sounds like it's worth giving it a shot if others aren't having this same issue with a different filter on the 70-300. (Plus I could just take it to the camera store and test a filter I'm sure.)