It's official: kit lens blows (back from reservicing - samples)

[Greg M]

I think i'm being harsh too, but when i compare my new $100 kit
lens to my bought-used $70 28-80, the 28-80 wins every time. that's
why i'm upset.

A cheap 28-80 is easier to build then a cheap 18-55. Wide angle lenses are softer wide open. This is true even with primes. A good wide angle prime will not be as sharp as normal prime. I had a 16-35L and at 28mm my 28-135 IS was just as sharp wide open. The 28-135 IS is a fraction of the cost of the 16-35L. I still say that I have a very good 28-135 IS and I had a bad 16-35L because like your situation my 16-35L didn't produce what I saw that others did. I returned my 16-35L.

Is there a reason why you didn't return your kit with a sample and get another kit? On a side note this is the reason that when I get the kit for my wife I'm going to get it at Sears. They will exchange an item without question unlike some of the experiences that I've heard about at places like BB.

you are right i'm very sure. that's the other reason i'm upset, i
sent it back and had it serviced. got it back, except it suffers
all the same problems i sent it back for, plus a few extra
fingerprints :-~

That's why I always try and return or exchange an item before going the service route.

--

• lev

--
Greg M
http://www.mocanu.com/gallery/index.php
http://dslr.mocanu.com

 

[DavidP #28649]

How many peanuts do YOU get for $999?

And I wouldn't really call the 300D the equivalent of the 10D. If so, anybody wanna trade him for one?

You've bought the equivalent of a 10D for peanuts.

--
http://www.outboundmusic.com
Your link to independent music!

 

[ScottC12]

Looks like all of these shots are pretty dark! Do you only come out at night? Sleep all day?

As some of you may recall, i believed i got a dud 18-55 EF-s lens
with my rebel. It had very soft corners, especially noticable in
the bottom right. remember this picture?

[1.8MB]
http://www.orionsector.com/qstuph/testphotos/181719.jpg

that was before servicing. i was so disgusted with the soft corners
at wide angles, it practically made the lens useless. sooo... i
sent it off to canon with this very same picture describing the
problem and so on. Well, i just got it back today. here is the
result:

[1.9MB]
http://www.orionsector.com/qstuph/testphotos/img_1017.jpg

see the difference? not much. notice the bottom right corner. same
ol' smear as before. This occurs at wide angles and wide apertures.
It pretty much ruins the photo IMHO.

OK, so maybe it's not the greatest lense at f/3.5 @ 18mm. i can
live with that. so i did some tests. look how drasticly ugly this
lens is at anything less than f/8 (tripod shot):

1/6th @ f/3.5:



verses this one at 1sec @ f/11



with some 100% crops off both:

1/6th @ f/3.5, bottom right corner:



1sec @ f/11, bottom right corner



1/6th @ f/3.5, center area [ notice the orange flag in the
background]:



1sec @ f/11, center area [ notice the orange flag in the background



i now have 4 lenses: kit lens, cheap non-USM 28-80 canon, 28-135 IS
USM, and the 50/1.8. None of them are really high calibur, but the
kit lens performs the worst by far . Maybe this is canon's way of
getting people to buy better lenses and make money, or maybe their
servicing dept was a bunch of trained monkeys. Either way, i am
strongly recomending anything but the kit lens at this point. It
would be great if it worked at wide angles, but sadly it fails
pretty bad. Not only is it soft with bad corners, it also has
HORRIBLE chromatic aberation, the worst of the 4 lenses i own (even
the el cheapo ones). I'm looking forward to the 17-40L now. If i
sell this one on ebay for $100, i only have $600 more to go

sorry guys, i just had to vent a little and give a word of caution
to everyone else. maybe i'm just cursed, but i am happy with my
other 3 lenses.

--

• lev

 

[misha]

My honest opinion is that you are being rather harsh on a 100$
lens. I think it is a great value for the money. And when you stop
down to f/8 it becomes pretty usable.

i dont' mind stopping down a bit, since i usually shoot composed
tripod shots, but all the way down to f/8-f/11 is pretty far IMNSHO.

I think i'm being harsh too, but when i compare my new $100 kit
lens to my bought-used $70 28-80, the 28-80 wins every time. that's
why i'm upset.

If you shoot with the kit lens at 28mm (instead of 18mm), do you still get these smeared corners? It's much more difficult to make a sharp 18-55 zoom than the basic 28-90.
--
Misha

 

[Stan Richard]

As some of you may recall, i believed i got a dud 18-55 EF-s lens
with my rebel. It had very soft corners, especially noticable in
the bottom right. remember this picture?

[1.8MB]
http://www.orionsector.com/qstuph/testphotos/181719.jpg

that was before servicing. i was so disgusted with the soft corners
at wide angles, it practically made the lens useless. sooo... i
sent it off to canon with this very same picture describing the
problem and so on. Well, i just got it back today. here is the
result:

[1.9MB]
http://www.orionsector.com/qstuph/testphotos/img_1017.jpg

see the difference? not much. notice the bottom right corner. same
ol' smear as before. This occurs at wide angles and wide apertures.
It pretty much ruins the photo IMHO.

OK, so maybe it's not the greatest lense at f/3.5 @ 18mm. i can
live with that. so i did some tests. look how drasticly ugly this
lens is at anything less than f/8 (tripod shot):

1/6th @ f/3.5:



verses this one at 1sec @ f/11



with some 100% crops off both:

1/6th @ f/3.5, bottom right corner:



1sec @ f/11, bottom right corner



1/6th @ f/3.5, center area [ notice the orange flag in the
background]:



1sec @ f/11, center area [ notice the orange flag in the background



i now have 4 lenses: kit lens, cheap non-USM 28-80 canon, 28-135 IS
USM, and the 50/1.8. None of them are really high calibur, but the
kit lens performs the worst by far . Maybe this is canon's way of
getting people to buy better lenses and make money, or maybe their
servicing dept was a bunch of trained monkeys. Either way, i am
strongly recomending anything but the kit lens at this point. It
would be great if it worked at wide angles, but sadly it fails
pretty bad. Not only is it soft with bad corners, it also has
HORRIBLE chromatic aberation, the worst of the 4 lenses i own (even
the el cheapo ones). I'm looking forward to the 17-40L now. If i
sell this one on ebay for $100, i only have $600 more to go

sorry guys, i just had to vent a little and give a word of caution
to everyone else. maybe i'm just cursed, but i am happy with my
other 3 lenses.

--

• lev

--
Stanzman

 

[hm]

I'm not questioning whether your lens is soft in the corners, I'd fully expect it to be.

But your comparison is a little confusing to me. You're comparing the corners at f/3.5 with f/11 without specifying where in the scene you've focused. If you want to test the corners, you need to use a scene where the corners are the focus point, otherwise the softness of the lens and softness due to out-of-focus are being mixed together, and it's hard to guage which is the predominant source.
-harry

 

[cameras]

This is why people for years have shelled out large sums to get fine lenses that are made to last, a digital sensor does not change that requirement.They should replace the lens for you in this case I would think. If it sells for $100 how much did it cost to make and what can be expected in terms of quality control. I have a number of plastic sigma and other brand lenses in a junk box, just didn't hold up well even if they performed well when new.

 

[Mark Marshall]

Before I got my Reb I saw a few early kit lens examples on this forum that almost made me sick (about the money my wife was about to spend). Especially horrible CA. Thankfully I get good shots with mine, but even if yours is hopelessly duff it's still better that you pack it away for the possibility down the road that you sell the body. I completely agree that Rebels w/ lens will sell 4x faster than w/o. In the meantime, until you can save for your L glass, I would only shoot stopped down (kick the ISO as necessary) and RAW. Lower the contrast and bring it back on 12bit conversion. You should eliminate a lot of your CA that way. Won't help that corner but hey, frame it so you can always crop there. ~ m²

pretty bad. Not only is it soft with bad corners, it also has
HORRIBLE chromatic aberation, the worst of the 4 lenses i own (even
the el cheapo ones). I'm looking forward to the 17-40L now. If i
sell this one on ebay for $100, i only have $600 more to go

--

'Brothers and sisters, we've learned that there's some bad bokeh going around out there. So like, just be careful man, alright?' (If Wavy Gravy emceed PMA) http://rhodeymark.instantlogic.com

 

[leiavoia]

Looks like all of these shots are pretty dark! Do you only come
out at night? Sleep all day?

no, i live in Oregon

 

[leiavoia]

Before I got my Reb I saw a few early kit lens examples on this
forum that almost made me sick (about the money my wife was about
to spend). Especially horrible CA. Thankfully I get good shots with
mine, but even if yours is hopelessly duff it's still better that
you pack it away for the possibility down the road that you sell
the body. I completely agree that Rebels w/ lens will sell 4x
faster than w/o. In the meantime, until you can save for your L
glass, I would only shoot stopped down (kick the ISO as necessary)
and RAW. Lower the contrast and bring it back on 12bit conversion.
You should eliminate a lot of your CA that way. Won't help that
corner but hey, frame it so you can always crop there. ~ m²

i'll keep that in mind and pack this little guy away for a few years then. thanks for the tip

 

[seafan]

I was thinking the same thing. is f3.5 enough DOF for this scene? even if you change the focus point to the rightmost square, it's not on the corner for sure.

a more comparable test may be shooting a brick wall, then compare the corners.

just a thought.

Jason

I'm not questioning whether your lens is soft in the corners, I'd
fully expect it to be.

But your comparison is a little confusing to me. You're comparing
the corners at f/3.5 with f/11 without specifying where in the
scene you've focused. If you want to test the corners, you need to
use a scene where the corners are the focus point, otherwise the
softness of the lens and softness due to out-of-focus are being
mixed together, and it's hard to guage which is the predominant
source.
-harry

 

[dchao]

I was thinking the same thing. is f3.5 enough DOF for this scene?
even if you change the focus point to the rightmost square, it's
not on the corner for sure.

a more comparable test may be shooting a brick wall, then compare
the corners.

just a thought.

Jason

 

[dchao]

Don't pack it away, this little kit lens is perfect for indoor party shots where 28mm might even be too long.

So far, for me, the only problem on the kit lens is CA, resolution is fine. In fact, the resolution in the center is almost as good as the 17-40L ($800).

In fact, shooting a brick wall is still not the definitive test, since the far end of the wall is still longer to the lens than the center wall, and f/3.5 has fairly shadow DOF.

It should not be that often you need to push the the aperture to f/3.5 at the 18mm end. Since the shutter speed can be slowed down to 1/20s without introducing handshakes, and then after that, you can change the ISO speed, and even if that fails, pop the built-in flash up or use a noise removal program or shooting in RAW for greater DR.

Don't give up on the kit lens yet.

i'll keep that in mind and pack this little guy away for a few
years then. thanks for the tip

 

[mjmetts]

I think that its pretty obvious that your lens is a defect. Mine has not had any problems in the two months that I've owned it.

Here are a couple of shots that I took the other day with my kit lens. Aperature is wide open. Very little PS work.





--MJ

http://www.mjmetts.com/

 

[PHGN]

Is it Canon's worst lens? Posibly.
Do Canon do bad lenses? Not realy.

It's a kit lens, not gonna set the world alight, but cheap, light, and... reasonable. Nothing special, but serves its puropse.

Use it well, it will yeild good photos. Badly it wont. Probably will never yeild 'technicaly' excellent pictures, but that is often over-rated IMHO. For many types of photography, 'tecnical excellance' in the image is less important than the photograph .

Get other lenses if you need them, That is part of the reason why you payed £XXX more for a DSLR after all isnt it? (Certainly was for me.)

 

[hm]

a more comparable test may be shooting a brick wall, then compare
the corners.

A brick wall is fine, but even then I suspect that it's still important to focus on the corner. The "plane of focus" of a lens isn't necessarily really a perfect plane, but typically has some curvature, so it helps to really focus on the thing you want to look at. And it wouldn't hurt to do a little "focus bracketing", i.e. tweak the focus a bit fore and aft and take multiple shots. You just want to grab the best shot you can grab, and make your judgement based on that.

And a tripod doesn't hurt. And enough light to use a fast shutter speed.

Again, the idea is to eliminate as many possible sources of softness as possible, to ensure you can attribute what you see to aberrations in the lens itself.
-harry

 

[jose garcia]

For $100 the kit lens is not bad , I mean is not L glass but I find it acceptable for some shots.

18-55mm shots

http://www.pbase.com/image/21590918

http://www.pbase.com/image/20078950

http://www.pbase.com/image/21601142

http://www.pbase.com/image/21590315

http://www.pbase.com/image/21633373

by the way none of these images have been sharpened .

 

[RATPhoto]

a more comparable test may be shooting a brick wall, then compare
the corners.

A brick wall is fine, but even then I suspect that it's still
important to focus on the corner. The "plane of focus" of a lens
isn't necessarily really a perfect plane, but typically has some
curvature, so it helps to really focus on the thing you want to
look at. And it wouldn't hurt to do a little "focus bracketing",
i.e. tweak the focus a bit fore and aft and take multiple shots.
You just want to grab the best shot you can grab, and make your
judgement based on that.

And a tripod doesn't hurt. And enough light to use a fast shutter
speed.

Again, the idea is to eliminate as many possible sources of
softness as possible, to ensure you can attribute what you see to
aberrations in the lens itself.
-harry

He can do all that testing if he wants to. But I'll tell you right now that he is going to find that it is soft in the corner. Even Canon's Theoretical MTF curves make that clear. In my own testing in comparing the 17-40 F4L and the 18-55 EF-S, the difference is easily visible at f/4.

Also at 18mm, even at f/3.5 the DOF is actually pretty wide on this lens. Because of that, the fact he isn't focused in the corners is less of a factor than you might off the top think. (I'm not saying it isn't a factor - just that it is less important than you might think).

--
If you are a new user chances are good your question is answered in the FAQ at:
http://www.marius.org/eos300dfaq.php

For a small gallery of my photographs, see:
http://ratphoto.home.comcast.net/
See my profile for my equipment

 

[leiavoia]

I'm not questioning whether your lens is soft in the corners, I'd
fully expect it to be.

But your comparison is a little confusing to me. You're comparing
the corners at f/3.5 with f/11 without specifying where in the
scene you've focused. If you want to test the corners, you need to
use a scene where the corners are the focus point, otherwise the
softness of the lens and softness due to out-of-focus are being
mixed together, and it's hard to guage which is the predominant
source.
-harry

that's a very good point, and i did consider that when i did the above shots. that is why i also shot some flat fields when i was out doing my testing along with the other shots. Lacking any brick walls, i shot the ground straight down and a few other similar shots. Obviously with it being flat, it was sharper in the corners, but still exhibits the same basic bottom-right corner "smearing" flaw and overall corner softness.