HELP!! Is my 100-400 a soft one? * link added for picture *

swag72

swag72

Suspended / Banned
Messages
7,969
Name
Sara
Edit My Images
Yes
I have always found it hard to get pin sharp pics with my 100-400 and read with interest the comments that they are often not sharp at 400mm. On the other hand there is someone on here who takes pics of puffins and theirs is tack sharp.

So I have done the focussing test in order to get my micro adjustment as sharp as it will go. It ended up on +1 on my 50d micro adjustment. So I have posted a picture below, at +1 - Do you think this is aceptably sharp from approx 12 feet at 400mm? According to my test this is the sharpest I will get with this lens and camera.

If not, please say, and some tips if I need to take it back to Jessops would be helpful.

3156416945_800405f449_o.jpg


Hope you can tell from this, if not, not sure how I can get it bigger without enduring the wrath of the mods!! Do you need it bigger?
 
I`m not sure what I am looking at here, but if you want to post a bigger image, use Flickr to host a small one then link that to a full size image.
Use the insert picture icon to insert the image then use the insert link icon to link to full size
Allan
 
1/25th shutter speed at 400mm probably isn't going to help...
 
In my opinion, the best way for us to judge the sharpness is for you to shoot in raw and upload the raw file somewhere for people to download and review. That will eliminate any questions regarding resizing or otherwise sharpening and processing the image.

You can upload a raw file to http://rapidshare.de/ for free and then post the download link to it back here.

FWIW, that looks like a focus test pattern that I have and, looking at your image above, I would say it looks sharp to me. The original raw file or even the original unedited JPEG, if you shot JPEG, should confirm the facts.

EDIT : Wow, 1/25 - I hope you used a solid tripod with IS disabled, mirror lockup and a remote release or self timer to fire the shutter. If anything less then all bets are off. For a handheld shot from a 50D and 400mm lens you would beed to shutter speed of 1/640 to (hopefully) avoid camera shake. Even with IS, that would only buy you a couple of stops, so 1/160 if handholding. Tripods wobble, mirrors slap and fingers induce vibration when they touch buttons. At 1/25 you need to take quite strong measures to eradicate camera shake from the equation.
 
Another thought, have you seen samples of the puffins you are comparing to at 100% rather than web processed?
 
swag72 said:
Here's the link to the full size raw picture Clickity click
Click to expand...

I'm not sure how/why your .CR2 file turned into a TIF but I think there is enough there to go on. You shot in standard picture style with sharpening set to 3. Your shutter speed was actually only 1/20 :)

Ordinarily I might expect a bit more sharpness from an image with sharpening at that level but when you consider (a) the reputation the lens has; (b) the punishing demands the 50D's high pixel density places on any lens; I'd say it is not too bad. A 50D file viewed at 100% is the equivalent of a pretty huge print - probably something like 30"x20" as I am viewing it on my laptop - and not one that you would ordinarily view from only 12" away. The only way I could really judge whether your lens is good or bad would be to compare it to an identical shot with my 50D and 100-400, but I don't have a comparable image handy that I could use to make a judgement.

You could try stopping down to f/7.1 (supposedly the threshold before diffraction softening starts to make itself visible) or f/8 and see how much it sharpens up. You could also try the same shot from half the distance with a 200mm focal length and see how that compares.
 
Thanks for taking the time Tim, really appreciate it. Looks as though I should just be content with what I have got! Do you think I'll ever get those pin sharp shots with it? I'm seriously doubting it at the moment. :shrug:
 
Do you have some examples of what you expect?
 
If you want to get really really sharp shots straight out of camera you're unlikely to accomplish that with any zoom lens, especially wide open and at at least favoured focal length. If you want to make the most of what you have, within the practical constraints of optical technology, the sensor itself and the AA filter, then you will need to get down and dirty with some more advanced post processing.

It's a fair bet that the very finest images you see (in terms of sharpness at least) have good lighting, with a nice level of contrast, and most importantly, sharpening after resizing for the web. They may well have been shot with a prime lens as well, rather than a 4X zoom lens. Canon has published a tutorial regarding the importance of sharpening to get the most from your equipment. It might be helpful to take a look at what they say in this Quicktime movie....

http://www.usa.canon.com/content/sharpening/sharpening.html

Here is a sample photo, just a snap, taken with my 100-400 at 390mm (I missed getting it all the way to 400mm) and f/5.6, handheld with my 40D. The first example is straight out of DPP with Standard picture style and sharpening at 3. The second version has had a little extra fine tuning of sharpening in Lightroom, including sharpening after resizing. The differences may seem subtl at first glance but they are there. A subject with more fine detail might illustrate the point better but this is something I happened to have handy.

20080512_170750_4120_DPP.JPG
20080512_170750_4120_LR.jpg


To my eyes the image from DPP seems to lack the pop and crispness that the Lightroom version has. Neither has had any PP other than setting picture style, resizing and sharpening.

Here's another shot that I happened upon, this time taken with my 50D and 100-400 + taped Kenko 1.4X teleconverter at 400mm (= 560mm with the telecon) and f/7.1 (= f/10 with the telecon), tweaked and processed in Lightroom....

Full image :
20081122_112845_1795_LR-3.jpg


100% crop :
20081122_112845_1795_LR-2.jpg
 
This is a shot at 400mm F5.6 Handheld, was taken a long time ago now lol

Has only been changed to jpg from raw no other editing at all ! It is 100% size so almost 2mb in size so will take a while to download !

# Exposure Time (1 / Shutter Speed) = 1/400 second = 0.0025 second
# Lens F-Number/F-Stop = 56/10 = F5.6
# ISO Speed Ratings = 100
# Original Date/Time = 2005:07:30 14:40:41
# Shutter Speed (Exposure Time) = 1/400 second
# Aperture = F5.6
# Flash = Flash did not fire, compulsory flash mode
# Focal Length = 400/1 mm = 400 mm
# Image Width = 3504 pixels
# Image Height = 2336 pixels

http://www.cookster.co.uk/gallery/albums/userpics/10002/Gull.jpg
 
I find on mine reducing the FL a little from 400mm to 375 -385mm makes my shots a tad sharper than 400. I'm sure I read other owners found the same with this lens...perhaps on TP.
 
Although I can't speak for this particular lens I can relate to what you are going through. I too recently got hold of a nice telephoto lens (of the Nikon variety) and went through rigorous testing to see if I had a 'sharp copy'. I had the added worry that mine was a second hand lens which only made matters worse. All of the shots I did in poor light came out looking pretty soft for that particular lens and that was under 'test conditions'. I made sure the camera didn't move and inch, used a remote release, prayed to the Photography Gods and yet STILL they came out soft.

After waiting a few days for some decent light I went outside to give the lens a proper field test and it was only then did I find out that the lens itself is fine. It's all too easy to get paranoid with new or expensive gear which is exactly what happened to me and could be happening to you too. Wait for a really bright day then have a play with it outside taking as many shots as you possibly can and then take a look at the results, hopefully you will be surprised.

I hope it works OK for you and I apologise in advance if this sounded patronising (not what I intended at all!).
 
Some good advice from Mr T above :) We know that this lens is plenty sharp, so what is important is do you have a good copy? Here's a quick test you might want to try.

I also have the 100-400mm and wanted to check that I had a good copy. So I tested two before I bought, as I always do with new lenses. And as usual, I found that both copies were fine. This 'bad copy' stuff has got more to do with bad testing than bad lenses in my opionion, and top of the list is that dangerously irrelevant front/rear focus test, if it's the A4 sheet of paper shot at close range at 45 degs. This test often leads people to believe their equipment is out, and now that cameras can be user-tweaked a bit here, that's what they do, which then puts the lens OUT for most normal situations :eek: Don't use that test (except maybe for macro lenses). Use a real world test if you have a genuine reason to think that the focus is out. If you have adjusted the camera as a result of doing the A4 paper test, it's probably best to re-set it back.

The most common lens fault by far is poor alignment of the elements, either in assembly, or after a knock. If this happens, one side of the image will always be less sharp than the other. It is therefore very easy to test for this.

In good light, choose a distant target subject to minimise focus variables, like a road sign or number plate. Set the lens at its widest aperture (low f/number) and choose a mid range focal length. Set the fastest shutter speed possible by cranking up the ISO, and use a tripod if you can. Focus very carefully and then lock off AF and IS so that nothing can change. Shoot four pictures with the target positioned in each corner of the frame.

Pop your memory card into one of those in-store photo booths, select max enlargement and do a 6x4in print of the target section. (Or do it on your PC of course - it's just that these booths are very handy when you're buying a new lens.) If you've done the test correctly, and if you have a good copy, all four prints will show a SIMILAR level of sharpness. If one or more corner is off, and you are confident that you've done the test correctly, then the lens is out.

What you are looking for is all four corners being the same. However, don't expect them to look great because this is a very tough test and they probably won't. You might find that one image is slightly worse than the rest but unless this immediately leaps out at you, it's probably acceptable. And quite common in complex 20-odd element lenses. In normal use, there's very rarely importat detail in the extreme corners, and this will clear up anyway when you close the aperture down to normal f/numbers.

It may not even be necessary to actually print out the pictures. You can spot major problems just by enlarging the images on your camera's LCD, on your PC, or on the photo-booth screen.

Hope this helps :)

Richard.
 
Remember as well that if shooting in RAW you will need to sharpen your shots any way as there is no processing applied to the image unlike a jpg
 
Cheers for all your help guys - Certainly haven't found any of it patronising. Below is a picture of my 100-400 wide open (I know it's been posted before, sorry). Looking at this, am I just worrying uneccessarily?

I also think I am just in the position of not being able to use the lens out and about too much as the light etc is crap by the time I get home. So when I do try, the conditions are pretty poor. I know that this lens loves light, so I'll just have to wait for the sun!!

3041165675_a6b8192446_o.jpg
 
Sara, was this shot raw originally ? is it a crop of a larger image ? and one more question, what is your post processing regime used on this picture ?

It actually looks fine to me, it might take a little more sharpening but i dont see anything that would scream the lens was at fault ?
 
I've just sat down and looked back though some of my images, found a couple that you can get an idea how it should be. I was never unhappy with my 100-400mm - although I did find a lot of my images were either 370mm of F6.3 (400mm wide open was not easy to find :) )

SHot in good light, high shutter speed and unfortunatly higher ISO (On my old 10D)
img_3971.jpg


and a crop of the wing (Where I suspect I focussed)
penguin_crop.jpg


Another from my 20D, lower light and on a tripod
IMG_1766.jpg


And here is a lonk to the above file full size straight out of Capture One (It is a crop on the 20D going from landscape to portrait)
http://scruffywhippet.com/talkphoto/IMG_1766.jpg

Hope this helps.
 
HoppyUK said:
Here's a quick test you might want to try ...

This 'bad copy' stuff has got more to do with bad testing than bad lenses in my opionion ...

The most common lens fault by far is poor alignment of the elements, either in assembly, or after a knock. If this happens, one side of the image will always be less sharp than the other. It is therefore very easy to test for this.
Click to expand...
Richard speaks a great deal of sense, as ever. I've had a handful of lenses (out of 150) which have been noticeably less sharp on one side than the other, or noticeably less sharp in one corner. I have never had a "soft copy" and I'm starting to wonder whether they really exist.
 
The way I see it is that you either have a working lens or a faulty lens, if it's faulty you really will know about it. What I think is even more of a myth than this "soft copy" are the "sharp copies" that people seem to have :p.
 
StewartR said:
Richard speaks a great deal of sense, as ever. I've had a handful of lenses (out of 150) which have been noticeably less sharp on one side than the other, or noticeably less sharp in one corner. I have never had a "soft copy" and I'm starting to wonder whether they really exist.
Click to expand...

Thank you Stewart. Kind comments mate. Ditto :)

I am new to this place, but I have to say it's a very good website. The level of properly grounded knowldege, and willingness to share, is refreshing. I have posted similar stuff about 'bad copy' lenses on DPReview and got nothing but a deluge of abuse, supported by test chart images that prove how crap the lens really is :shake: Deluded Americans who have thown money at expensive kit, without understanding how it works, or bothering to learn how to use it. I just don't post on DPReview any more :lol:

Cheers,

Richard.
 
StewartR said:
I have never had a "soft copy" and I'm starting to wonder whether they really exist.
Click to expand...

I'm with you on this. I too have had a fair number of lenses in my time (OK, not 150+, but still) and while some have had focus calibration issues I've never had a lens that is softer than I am lead to expect - I've had a few that were sharper than my expectations mind.
 
Sara - I don't think this lens suffers with bad copies like some other Canon lenses. This lens does take sharp photos but it just takes a little getting used to. You'll definitely see an improvement stopping down, I found the wide open aperture of f/5.6 at 400mm narrower than usual
 
swag72 said:
Cheers for all your help guys - Certainly haven't found any of it patronising. Below is a picture of my 100-400 wide open (I know it's been posted before, sorry). Looking at this, am I just worrying uneccessarily?

I also think I am just in the position of not being able to use the lens out and about too much as the light etc is crap by the time I get home. So when I do try, the conditions are pretty poor. I know that this lens loves light, so I'll just have to wait for the sun!!

3041165675_a6b8192446_o.jpg
Click to expand...


This was shot in RAW and was a crop of about a third of the original picture (At a guess) Post processing I would say probably got a curve and a saturation and sharpen at the end. Is this OK then?

I am interested by a previous comment about my CR2 turning into a TIFF file - All of mine do once passed through DPP. Isn't this normal?
 
swag72 said:
I am interested by a previous comment about my CR2 turning into a TIFF file - All of mine do once passed through DPP. Isn't this normal?
Click to expand...

Well if you are supplying an unedited original image file, why are you passing it through DPP at all? You could have simply uploaded the original .CR2 file. It would have been a smaller upload (and download :) ) and would have been a sure guarantee that there was no accidental alteration, or degradation of the file.

If you want to output a processed image, even with no additional edits, you do have the option of JPEG instead of TIF, which would have been far, far smaller, but for these purposes the original CR2 would have been the best option.
 
You must log in or register to reply here.
Back
Top