Soft photo - any help appreciated

[davidedric]

Hi everyone,

On a recent holiday, I took a number of photos of flocks of cranes in flight and coming in to land. When I got them home, I was very disappointed how soft they were, and I would be grateful for any comments.

The photos were taken handheld, at 300mm, 1/400sec, f7.1, ISO 100 (you can see all that from the Exif below) and VC on. I was panning the camera to follow the birds

I can think of a number of explanations:
1. A problem with the camera or lens. Very unlikely, other photos on the trip were sharp
2. Poor technique. Perfectly possible, of course, but the birds were not moving very fast
3. Too slow a shutter speed. Again, a possibility. 1/1000sec would undoubtedly have been better, but if this were the problem then I would have expected to see more softness on the wings than the bodies
4. Problem with autofocus. The camera/lens can struggle when tracking a flock of birds, and there is a narrow depth of field. However, I can't see any part of the shot in focus, and there is material pretty much all the way from the foreground to the background.
5. Misbehaving Vibration Control. The lens only has an on/off switch, and it was on. Reading other posts this seems quite a likely source.

I can't go back tomorrow and try again! (Though I will try and find some birds locally).

However, I wonder if someone more experienced than I can look at the image and suggest what the problem or problems might be. I that is too much to ask, thanks for reading.

Here is the image. The only adjustment I have made to the RAW file is some foreground cropping (all of which was out of focus).

Thanks again,

Dave

[JPS]

Hi Dave,
a couple of thing come to mind here:
1. You were shooting at the maximum focal length for this lens, 300mm? Normally lens are soft at both ends of the range.
2. At 300mm with a crop camera; your one being 1.6 should the minimum shutter speed not have been 480mm? 300 X 1.6
3. I think you are correct about the VC, it would have been better 'off' in this situation.

That my best shot, it will be interesting to see what others think.
Cheers
John

[revi]

Could it be a autofocus problem?
As you were panning, you'd expect the background to be unsharp. If the autofocus fails to latch on the birds,
you'd end up with nothing sharp: birds out of focus and background movement blur. That should be visible on
the full-sized image

[Manfred M]

I would agree with John and Remco as likely sources of the problem. I find that when I shoot with a long lens, at maximum focal length the camera's AF will often have trouble locking and when it does, it will often lock on the wrong subject. Add to that the fact that the birds are moving and you are panning at a relatively low shutter speed, and I think you have a likely explanation.

While lenses can be soft at longer focal lengths, they are not that soft. I have heard the VC argument before but I have doubt it. It is far easier to blame the equipment than the shooter's technique. On some of my lenses I technically have to change stabilization modes when panning; one doesn't want the lens to compensate for the panning motion, just the up and down motion.

[binsurf]

I'm not convinced it's shutter speed or VC. The blurriness is consistent and that's also the result of max FL on the lens as John suggested. Try backing the lens off 100 mm and see if you have the same issue. This is a great shot, by the way. If you can find a way to salvage it, it would make a great poster.

[xpatUSA]

Pixel Peeper here with a technical view:

At f/7.1, the diffraction of about 9.5 um Airy disk diameter may be having some effect, depending on your cam's pixel size. The smaller the pixels, the more the blur. Yep. The birds are quite far - one of their legs measures only about 8-10 px wide but the edge spread (measured with ImageJ) is about 4px 10-90% and edges are uniformly soft around all the birds.

Probably no help at all ;-)

As a product photographer, I continue to be amazed that AF ever works on flying birds!

[Hero]

I don't know about the canon 70-300, but the sigma and nikon ones tend to get a bit soft above the 200-250 mark. So take a few test-shots between 200 and 300mm and see if it's the lens.

[William W]

For this particular image –

Panning? Which way? The birds are flying in different directions and at different speeds and many have landed.

Best guess is that ‘panning’ and using Tv = 1/400s is the answer.
Depending upon how much movement, ‘panning’ actually was: this is the likely explanation of the FUNDAMENTAL cause of the GENERAL SOFTNESS in THIS PARTICULAR IMAGE - because the VC just couldn’t cope with all that.
My first adjustment would be shutter speed – and for a flock of birds moving in different directions I would not be panning.

This answer does NOT exclude other CONTRIBUTING factors such as the general softness of the lens at 300 mm and note nailing focus exactly spot on.

However - in this particular (low res) image there does appear to be a zone of sharp focus albeit with what I see as a blur: and this zone of sharp focus is around the birds, compared to the background, which is Out of Focus.

WW

[Andrew1]

For some instances such as birds and race cars I find the auto-focus sometimes insufficient. In some cases it's speed while in others it's the focus point being confused between the subject and the background. In close shots even the wing tips can conflict with body focus. What I've been doing for shots of birds under 50 feet is to manually focus where I've seen them pass then pan and wait until they get there. Same thing where they land. It works very well when you've got lots of birds around. Panning imparts softness problems of its own so I need to take lots of shots to get one sharp enough. A monopod helps until I can get a Gimbal. If you do test for softness at the extremes of any lens use a tripod. Hand held at any length will not be a solid base for which results and decisions can be finitely made.

[rpcrowe]

First off I want to say that this is a lovely image. I love the diffused colors of the background compared with the colors of the flock of birds. The positioning of the birds are great with some area into which they can move.



I sharpened the image in Photoshop CS6. I also straightened the image's relative horizon and removed the partial bird on the right and the vertical branch on the bottom left, using content aware fill.

The frame is totally optional but, I think a frame of some type would help to assign a definitive border to the image.

Maybe you couldn't make a great enlargement of this image but, I think that this is adequate for say; an 8x10" size image...

[jprzybyla]

Hello Dave, since seeing your post this morning I have thought about why it would be soft focus. I think the shutter speed was too slow, that maybe 1/1000 would have worked better. Panning with a slower shutter speed to have the body/head in sharp focus with the wings blurred showing motion works with a single bird but not necesarily with several birds. Depth of field becomes a problem with several birds. In this image considering the large crop I think the birds were far enough away for 7.1 aperture to be fine. The birds and scene being far away and the focal lenght of 480 FFE will show any motion as softness unless the shutter speed is high enough to freeze it. The whole scene being in soft focus would seem to indicate camera motion at the time of exposure. If I am correct the Cannon 600d has 18 megapixels on the small sensor, that is very high definition. where the old adage of 1/the focal length of the lens used to apply with high definition sensors today faster shutter speeds are needed for tack sharp images. I use 1/400 or 1/500 for stationary birds but with any motion at least 1/1000. I have my camera set up with Aperture Priority for stationary birds at f/8. I have the Manual set to 1/1000 aperture f/8 for slow moving subjects and Shutter Priority set to 1/2000 sometimes increasing this for fast flying birds like ducks. For all of these setting I use the Auto ISO feature of the camera basically turning the camera into a $1000 point and shoot. I switch back and forth between aperture, manual or shutter to suit the conditions at hand.

[William W]

“Sharpenss” of an image has many facets.

On a recent holiday, I took a number of photos of flocks of cranes in flight and coming in to land. When I got them home, I was very disappointed how soft they were, and I would be grateful for any comments.

Notwithstanding the fact that I believe that Tv =1/400s was too slow a Shutter Speed to use for that shot and that is the main cause of the apparent lack of sharpness as it captured:
SUBJECT MOVEMENT and also,
PANNING was probably too aggressive for the Image Stabilization to cope and that caused MOTION BLURRING . . .

Here are several other comments which are segregated into topic headings.

. . . Misbehaving Vibration Control. The lens only has an on/off switch, and it was on.

On the question of Image Stabilization:
You were using a Canon lens and not a Nikon lens - yes?
What Lens, specifically?
The EXIF states that the lens was 70mm to 300mm zoom. Canon has three of those EF lenses with Image Stabilization.
Each of those lenses has ‘Mode 2 IS’ (Panning IS):

EF70-300mm f/4-5.6L IS USM
EF70-300mm f/4.5-5.6 DO IS USM
The EF70-300mm f/4-5.6 IS USM

***

PANNING:
But as a generalization ‘panning’ would not be at Tv = 1/400s.
Tv = 1/400s is kind of a ‘no man’s land’ for a shutter speed where the Subject is in TRANSVERSE movement, no matter what the shooting distance and what the Subject’s Speed.
At 1/400s you are neither following the pack with a fast Tv to freeze motion nor do you have a slow enough Tv, to blur the background and sustain an appropriately crisp image in ONE LINE OF Subject Motion.
IF HORIZONTAL panning were to be used AND you are using one of those three Canon lenses, then IS should be switched ON and ‘MODE 2’ Selected.

***

DoF is not an issue:
The detailed EXIF states, that the approximate SD (Shooting Distance) is well in excess for there to be concern of any shallow DoF being of any issue contributing to softness, provided the AF nailed somewhere on one of those birds in the pack – and by observation the area of the pack and the grass underneath it, is sharper than the background (and assumed is also sharper than the foreground which you cropped out).

***

FILTER?
Upon detailed interrogation the Purple Fringing and general CA seems a bit excessive for any one of those lenses . . . difficult to make an absolute call, but my suspicion is there is a UV filter on that lens: if so, remove it for shots like that.

***

ISO SELECTION:
ISO100: I suggest you forget ISO100 in low light like that. Bump to ISO 400 or ISO 800 and enjoy the gain in Shutter speed and crib to around F/8~F/11, unless you are using the L Series 70 to 300 Lens.

***
EVALUATIVE METERING – its Limitations and KNOWING HOW THE TTL METER WORKS / (NOT) USING the TTL METER for AUTO EXPOSURE control of your camera / ‘CORRECT’ Exposure / Knowing the Camera’s limitations and Exposing to bring back THE HIGHLIGHTS and to KEEP THE SHADOW DETAIL:


The scene (the ‘scene’ at the birds) appears to me to be about EV ≈ 12~13 (daylight - no distinct shadows). If this is so then you are UNDERexposed, (for the birds). The APPEARANCE of sharpness will be increased as the ACUTANCE (edge sharpness/contrast) increases. For LOW CONTRAST SCENES (the Birds) it would have been better to have more exposure, IMO. The 600D could be pushed about ⅔~1 STOP over the edge of what the camera declares ‘maximum exposure’ (as determined by the ‘blinkies’ or the histogram) and I am confident that the highlight detail could be recovered in Post Production, from the raw file.

***

POST PRODUCTION:
In any case, POST PRODUCTION SHARPENING is necessary for all digital images to attain the best ‘sharpest’ resultant image.

As seen by the Post Production work Richard did and also indicated here in this quick example:


The original is on top.

The image has been sharpened.
The dynamic range (tonal range) extended, to ‘bring out’ the shadow detail.
This is a lesser quality facsimile of what would have been better - making a ‘correct’ exposure and then bringing out the highlight detail in post production.

View the A/B comparison 'large' by clicking on the collage and then selecting the 'actual size' option.

WW

[davidedric]

First a huge thanks for all the feedback, and especially for re-working the image, too. There's a lot to think about there, and lots of options to try. It's a shame I can't nip back out there tomorrow and have another go!

Just for clarity, I was actually using a Tamron 70-300mm, which doesn't have the "Mode 2 IS", nor was there a filter on the lens.

A couple of comments:
- I'm pretty sure it's not a problem with a soft lens. I have many other shots from the trip with broadly similar settings, and they are certainly sharp enough
- I was a but puzzled by the comments that it was a heavy crop - in fact the image that I set out to upload was hardly cropped at all, just a slice of the bottom (it was 5170 x 2209). Then I realised that TinyPic had downsized it dramatically. I guess I should have used a different method to get the full image up there. However, the truth is that it didn't actually make a lot of difference- the softness was just as apparent.
- I did realise that the image could be improved, but I thought it best to upload the original in all its fuzziness for your feedback.

So, thanks again - and back almost to the drawing board

Dave

[xpatUSA]

Motion blur has been mentioned a lot in the this thread. Sorry, but I don't see any in the image. Compare the heads of the flying and standing birds below (up-sized 4X, nearest neighbor method), living proof:



The standing birds would have horizontal blurring - I see equal blur all around the head. The entire image is just soft, IMHO.

[revi]

Except that your excerpts are from a down-sampled image, with no info about the treatment...
(posted image @ 1599px hor., original 5170)

Might a deconvolution sharpening be useful here?

[Dave Humphries]

OK I'll lob in something no one has yet mentioned

Heat haze? - the birds look quite distant and the flat plains were probably quite warm.

Dave, Can I ask; was that a an unsharpened jpg from a RAW capture?

Dave, Can you give us a couple of 1500 x 1000 px crops from the full image, so we can see them at 100% without TinyPic downsizing them? these should be central and one above the other, so we get the depth range. SO one 1000 px down from centre of image and the other 1000 px up from the centre of the image - I don't think I'm explaining that well, so here's an equally bad 'diagram'



Obviously I hope you'll give us the missing bits I have covered over

I did actually think the foreground grass might be sharper than the birds, etc., but I may change my mind if we see the 100% views.

Cheers,

[William W]

Just for clarity, I was actually using a Tamron 70-300mm, which doesn't have the "Mode 2 IS", nor was there a filter on the lens.

Thank you, understood.

I don't know the Tamron Lens and that would explain my ‘CA’ and my 'filter' comments.

***

As best as I can interrogate the low res sample image: -

I detect different leading edge blur on the wings when comparing birds in flight.
I conclude it to be SUBJECT MOVEMENT captured:


***

I detect equal leading edge blur in many parts of the grass and the birds, though not consistently apparent across the image I expect this is due to the downsizing and/or angles of the various subjects to the camera.
I conclude this to be some CAMERA MOVEMENT:

*



There could be other explanations either separate or accumulative.

Taking one sample of a small section and concluding something is ‘not there’ is not ‘living proof’ that issues do not exist: Analysis to ‘Proof’ just doesn’t work that way.

A set of larger, cropped samples, will obviously always be better for any detailed analysis.

WW

[jcuknz]

Something I learnt with a bird photo I took some time back that 'sharpness' is relative and if you apply sharpening to the subject but at the same time blurr the background it makes things look a lot better. I did a quick job and duplicated the image and blurred the top layer, then I erased the top layer where the birds are. I applied sharpness until the white surround was noticeable and then backed off a bit and then thought of trying 'my blurr b/g trick

One could make a third layer and blurr it more to give a suggestion of out of focus increasing with distance so say the far trees wwere softer than the near ones. With this I usually take the easy road and just use a soft edged erase tool rather than selecting out the actual subject ... it is an illusion I am creating rather than exactness for the pixel peepers

This obviously doesn't solve your operating problem but maybe is a rescue of what you have.

In my haste I now see I missed erasing the blurred birds on the right

[xpatUSA]

Taking one sample of a small section and concluding something is ‘not there’ is not ‘living proof’ that issues do not exist: Analysis to ‘Proof’ just doesn’t work that way.

Thank you for your clearly superior exhaustive analysis, Bill, and so subtly put with all those big old capital letters in your post. Almost like being yelled at . . .

. . . I'll get my coat.

[William W]

Ted Cousins:

Capital Letters are used to emphasis and differentiate key words or technical words, like "SUBJECT MOVEMENT" - that is not shouting. One will note capital letters in many of my posts to emphasize key or technical words - that is a common technical journal and written procedure.

You asked the question "what movement motion blur" and concluded that you had proof there was none . . . I responded to that with the images I used for my previous analysis and comments

***

Analysis is analysis and proof is proof.

If you do not like dealing with factual analysis and best practice to conclusion and want to paddle in procedures suited more to guesing than actually finding out what is/are the problem(s) - then that s fine by me:

But there is no need to revert to sarcasm and direct it at me by quoting me and then taking aim at my words.

And if you ask a question then expect to get it answered -and if the answer doesn't suit you then don't revert to sarcasm - and DO NOT direct that sarcasm to me. (And that WAS SHOUTING)

This forum is usually above your type of sarcastic response and you have lowered the tone - best you do get the coat and leave - yes I AM CRANKY - as your comments resemble those I would expect from an arrogant schoolboy.

William