Softness on the edge

Started Feb 22, 2021 | Questions
Kendunn Contributing Member • Posts: 880
Softness on the edge

Getting a weird, overly soft edge on rare occasions, not sure whats up.  Canon 5DS with a 17-40.  The example posted shows the top 20% of the frame is way too soft, but the rest of the frame is reasonably sharp, including the bottom.  It was shot at 35mm, f/7.1 off of a tripod, I think it was 1/20 at ISO 100. Most of the time images are fine. Loose element?

ANSWER:
This question has not been answered yet.
Canon EOS 5DS
If you believe there are incorrect tags, please send us this post using our feedback form.
MitchAlsup Veteran Member • Posts: 5,462
Re: Softness on the edge
1

Kendunn wrote:

Getting a weird, overly soft edge on rare occasions, not sure whats up. Canon 5DS with a 17-40. The example posted shows the top 20% of the frame is way too soft, but the rest of the frame is reasonably sharp, including the bottom. It was shot at 35mm, f/7.1 off of a tripod, I think it was 1/20 at ISO 100. Most of the time images are fine. Loose element?

You do realize the stuff at the top is more than twice as far away as where the tree emerges from the ground !! do you not ?

My guess is that the lens accommodation (near-focus to far-focus) is the culprit. Your choices are to push the aperture slower and get diffraction across the whole image or to widen the aperture and deliberately take the top OoF.

-- hide signature --

Mitch

OP Kendunn Contributing Member • Posts: 880
Re: Softness on the edge

MitchAlsup wrote:

Kendunn wrote:

Getting a weird, overly soft edge on rare occasions, not sure whats up. Canon 5DS with a 17-40. The example posted shows the top 20% of the frame is way too soft, but the rest of the frame is reasonably sharp, including the bottom. It was shot at 35mm, f/7.1 off of a tripod, I think it was 1/20 at ISO 100. Most of the time images are fine. Loose element?

You do realize the stuff at the top is more than twice as far away as where the tree emerges from the ground !! do you not ?

My guess is that the lens accommodation (near-focus to far-focus) is the culprit. Your choices are to push the aperture slower and get diffraction across the whole image or to widen the aperture and deliberately take the top OoF.

Thanks for the input, but a 35mm len @ 7.1 focused at 20' will have a DOF to infinity.  This tree was probably 100' away and what was focused on.  You can also look close at right side of the image and see trees that are sharp at their base but very soft at the top.

MitchAlsup Veteran Member • Posts: 5,462
Re: Softness on the edge
2

Kendunn wrote:

MitchAlsup wrote:

Kendunn wrote:

Getting a weird, overly soft edge on rare occasions, not sure whats up. Canon 5DS with a 17-40. The example posted shows the top 20% of the frame is way too soft, but the rest of the frame is reasonably sharp, including the bottom. It was shot at 35mm, f/7.1 off of a tripod, I think it was 1/20 at ISO 100. Most of the time images are fine. Loose element?

You do realize the stuff at the top is more than twice as far away as where the tree emerges from the ground !! do you not ?

My guess is that the lens accommodation (near-focus to far-focus) is the culprit. Your choices are to push the aperture slower and get diffraction across the whole image or to widen the aperture and deliberately take the top OoF.

Thanks for the input, but a 35mm len @ 7.1 focused at 20' will have a DOF to infinity. This tree was probably 100' away and what was focused on. You can also look close at right side of the image and see trees that are sharp at their base but very soft at the top.

If you change "will have" to "should have" in the first sentence I can agree with you.

Using a 17-40 lens at F/7.1 is not the same as using a 35mm prime at F/7.1 ! Plus primes are better off axis than zooms. The 5Ds may be out resolving the edge of what the 17-40 lens renders.

-- hide signature --

Mitch

TrackDayLT4 Regular Member • Posts: 143
Re: Softness on the edge
3

Kendunn wrote:

MitchAlsup wrote:

Kendunn wrote:

Getting a weird, overly soft edge on rare occasions, not sure whats up. Canon 5DS with a 17-40. The example posted shows the top 20% of the frame is way too soft, but the rest of the frame is reasonably sharp, including the bottom. It was shot at 35mm, f/7.1 off of a tripod, I think it was 1/20 at ISO 100. Most of the time images are fine. Loose element?

You do realize the stuff at the top is more than twice as far away as where the tree emerges from the ground !! do you not ?

My guess is that the lens accommodation (near-focus to far-focus) is the culprit. Your choices are to push the aperture slower and get diffraction across the whole image or to widen the aperture and deliberately take the top OoF.

Thanks for the input, but a 35mm len @ 7.1 focused at 20' will have a DOF to infinity. This tree was probably 100' away and what was focused on. You can also look close at right side of the image and see trees that are sharp at their base but very soft at the top.

You have to state the assumptions for DOF calculations and the COC.   Your statement may be true for an 8x10 print viewed at 10", but not for pixel peeping a 50mp full frame sensor at 100%.

marcio_napoli
marcio_napoli Senior Member • Posts: 2,042
misaligned element

Not in any way trying to be rude to other posters, but guys, don't  waste his time denying the obvious.

It's obvious the lens has a misaligned element.

It's showing up even in a reduced web viewing, seriously, it's an obviously faulty lens.

OP, I shoot Nikon (I'm here in this forum because I'm bored :D).

I have a Nikkor 17-55 f2.8 which was somewhat faulty out of factory, but not by much.

I assume it was within QC tolerances, but had one side definitely softer than the other.

That lens got fungus at some point, and was cleaned by a third party assistance, probably without the necessary QC tolerances.

When I got it back it looked like your lens or even a tad worse, so obviously a problem that was introduced by misaligned element, when putting the elements together after cleaning.

Your lens is for sure 100% faulty, that's not even open to debate, I'd exchange it if possible.

Best regards,
Marcio Napoli _ fashion photographer . indie filmmaker
.
NEW video just posted:
https://youtu.be/SZGrVC6J994
.
check it out my You Tube channel:
https://youtu.be/SIO0J3aqLVg
.
Aliens (acclaimed short film_near 700K views on YT):
https://youtu.be/aliscTnlsvg
.
Instagram:
@marcio_user

Ray UK Contributing Member • Posts: 905
Re: Softness on the edge
1

It could be a lens fault but this is the wrong image to use in order to judge the problem.

It needs a shot of a flat detailed surface that is parallel to the sensor, something like a brick wall, then it would be possible to see any sharpness difference across the frame.

OP Kendunn Contributing Member • Posts: 880
Re: misaligned element

marcio_napoli wrote:

Not in any way trying to be rude to other posters, but guys, don't waste his time denying the obvious.

It's obvious the lens has a misaligned element.

It's showing up even in a reduced web viewing, seriously, it's an obviously faulty lens.

OP, I shoot Nikon (I'm here in this forum because I'm bored :D).

I have a Nikkor 17-55 f2.8 which was somewhat faulty out of factory, but not by much.

I assume it was within QC tolerances, but had one side definitely softer than the other.

That lens got fungus at some point, and was cleaned by a third party assistance, probably without the necessary QC tolerances.

When I got it back it looked like your lens or even a tad worse, so obviously a problem that was introduced by misaligned element, when putting the elements together after cleaning.

Your lens is for sure 100% faulty, that's not even open to debate, I'd exchange it if possible.

Best regards,
Marcio Napoli _ fashion photographer . indie filmmaker
.
NEW video just posted:
https://youtu.be/SZGrVC6J994
.
check it out my You Tube channel:
https://youtu.be/SIO0J3aqLVg
.
Aliens (acclaimed short film_near 700K views on YT):
https://youtu.be/aliscTnlsvg
.
Instagram:
@marcio_user

Yeah, I didn't want to be rude either, but there is no way that is a DOF issue!  I've been a photographer for 40 years, shot everything from large format, medium format......M 4/3 and phone cameras, lol.  The lens is not new, not by a long shot, I guess I need to either see about a repair or maybe upgrade to the new 16-35

OP Kendunn Contributing Member • Posts: 880
Re: Softness on the edge

Ray UK wrote:

It could be a lens fault but this is the wrong image to use in order to judge the problem.

It needs a shot of a flat detailed surface that is parallel to the sensor, something like a brick wall, then it would be possible to see any sharpness difference across the frame.

Well, like I said, its pretty random and hard to replicate.

Andy01 Veteran Member • Posts: 4,849
Re: Softness on the edge
5

Kendunn wrote:

Getting a weird, overly soft edge on rare occasions, not sure whats up. Canon 5DS with a 17-40. The example posted shows the top 20% of the frame is way too soft, but the rest of the frame is reasonably sharp, including the bottom. It was shot at 35mm, f/7.1 off of a tripod, I think it was 1/20 at ISO 100. Most of the time images are fine. Loose element?

I assume that you are aware that the EF 17-40L is well known for it's soft edges - especially on a FF camera, and I would image that a 50Mp sensor would greatly magnify this. I chose not to buy a 17-40L when I tried it on a 600D APS-C in 2011 because I thought the edges & corners were not great - I can only image how a 5Ds would amplify this issue People talk about using it at f11 on a 20Mp 6D or 5D ii/iii to avoid softness.

While it could be a lens defect (decentreing) the defect would be there in every image (although some more obvious than others, but it probably wouldn't "come and go" unless something was very loose, in which case you would probably hear something moving), and it should be easy to pick by doing some checks such as a well set up brick wall test.

Have you checked the lens for decentreing ?

The bottom corners are fairly soft as well (especially bottom left), though not as bad as the top portion.

I am inclined to think that this is the inherent softness amplified by the DoF difference (perhaps with a touch of decentreing). While a DoF calculator might indicate that at f7.1 and 100', the DoF is huge, it does not mean that everything within the suggested DoF is in "perfect" focus - my understanding is that it would be in "reasonable" focus.

I would certainly NOT recommend a 17-40L for use on a 5Ds - I actually wouldn't recommend this lens at all unless budget is VERY tight and there is no other choice AND it was going to be used on a APS-C body. A EF 16-35L f4 is a vastly better choice that is likely to do some justice to the great sensor on the 5Ds.

Colin

 Andy01's gear list:Andy01's gear list
Canon EOS M5 Canon 6D Mark II Canon EF 100mm F2.8L Macro IS USM Canon EF-M 22mm f/2 STM Canon EF 35mm F2 IS USM +5 more
leeosenton Forum Member • Posts: 75
Re: misaligned element

I recently sold my 17-40 and replaced it with the 16-35/4 IS. Wow! The difference between the 2 was immediately obvious. I was happy enough with the 17-40 until I upgraded to the 5Dm4 and combining it with the 16-35 was a big step up for me.

 leeosenton's gear list:leeosenton's gear list
Canon EOS M6 II Canon EF 70-300mm f/4-5.6L IS USM Canon EF-M 22mm f/2 STM Canon EF-M 11-22mm f/4-5.6 IS STM Canon EF-S 55-250mm f/4-5.6 IS STM +2 more
MitchAlsup Veteran Member • Posts: 5,462
Re: misaligned element

Thanks for this report.

-- hide signature --

Mitch

OP Kendunn Contributing Member • Posts: 880
Re: misaligned element

leeosenton wrote:

I recently sold my 17-40 and replaced it with the 16-35/4 IS. Wow! The difference between the 2 was immediately obvious. I was happy enough with the 17-40 until I upgraded to the 5Dm4 and combining it with the 16-35 was a big step up for me.

Do you have SxS samples comparing the two?

OP Kendunn Contributing Member • Posts: 880
comparing the 16-35 and 17-40

I think a lot of people have bought into the marketing that the 16-35 is head and shoulders better than the 17-40.  I tested my 17-40 against a friends brand new 16-35 and found the difference to be very small, and only in the corners (stopped down a couple of stops on both).  My 17-40 usually gives very satisfactory results, was just wondering if this was an issue that happened in maybe 1 out of every few hundred pictures (the lens is probably 15 years old).  DXO gives basically the same score.  Ken says they are similar also.  Here is a picture that was taken with the same lens, near the same focal length, and stopped down to f/20 to give the water a little blur so its not at its sharpest (it looks like I could have had the RAW converter set with a little too much sharpening also, this was one of the first times using this camera). Sure, the corners are slightly softer at 100%, but that's with any super wide zoom. In the real world it wouldn't matter, especially doing landscapes and being able to stop down to f/8 or f/11.

Andy01 Veteran Member • Posts: 4,849
Re: comparing the 16-35 and 17-40

I would have expected the corners to be less soft (softer than the rest) at f20. I mean - it is f20 for goodness sakes

The point is that people buy this lens because it is a cheap constant f4 L series zoom lens only to find that the performance is really quite poor on FF (edges & corners at anything close to f4.

Sure it can be stopped down to f11 and crop out the soft edges & corners, but then what is the point of buying a 17mm f4 zoom if it cannot be used (and achieve reasonable IQ) at f4 and 17mm ?

Many argue that in many cases the edges or corners are not important, and sometimes this is true, but not always.

Many will say that they want the lens for daytime landscape or real estate on a tripod, so f4 doesn't matter because they use f8 or f11, and this is probably true as well, but how many people buy a lens exclusively for those purposes and never to use as a walk around WA when on vacation for example.

Personally if I buy a lens (especially an L series lens) I expect it to perform reasonably well at all focal lengths and apertures - there will always be some that are better than others, but I will never buy a 17-40mm f4 that I can only really get decent frame wide performance if I use f11 and cop out the outer 10% of the frame on all sides - what is the point ?

Look at this - the difference between 17-40L f4 and 16-35L f4 at f4 and widest setting is like night and day to me - perhaps you got an exceptional copy of the 17-40L and your friend got the world's worst copy of the 16-35L ?

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=100&Camera=979&Sample=0&FLI=0&API=0&LensComp=949&CameraComp=0&FLIComp=0&APIComp=0

Colin

 Andy01's gear list:Andy01's gear list
Canon EOS M5 Canon 6D Mark II Canon EF 100mm F2.8L Macro IS USM Canon EF-M 22mm f/2 STM Canon EF 35mm F2 IS USM +5 more
leeosenton Forum Member • Posts: 75
Re: comparing the 16-35 and 17-40

I haven't bought into anything except replacing a wide angle that was always decent but never great. I sold it and told myself 24mm was wide enough. I later decided it wasn't wide enough and the 16-35/4 wasn't much more so I bought it and like it much better. I'm glad you're happy with your 17-40/4; I wasn't with mine.

 leeosenton's gear list:leeosenton's gear list
Canon EOS M6 II Canon EF 70-300mm f/4-5.6L IS USM Canon EF-M 22mm f/2 STM Canon EF-M 11-22mm f/4-5.6 IS STM Canon EF-S 55-250mm f/4-5.6 IS STM +2 more
Keyboard shortcuts:
FForum MMy threads