Have I got a soft copy of 24-70L?

onform

onform

Messages
2,860
Name
matt
Edit My Images
Yes
I brought a 24-70 2.8L just before Christmas which I've been mainly using for portraits. I use the centre single point focus on a 5D mkI and I usually focus on the eyes. I am taking jpeg's only for the mean time as my screen is not high enough res.

My portraits all seem to be slightly soft! and the landscapes I have tried to shoot all seem just out of focus!

What do I do?

Matt
 
ok this was hand-held with studio lighting! and it has been sharpened with unsharp mask already!

Exposure: 0.008 sec (1/125) Aperture: f/6.3 Focal Length: 70 mm ISO Speed: 100 Exposure Bias: 0 EV

3371701255_5aa1f5f6bb_o.jpg
 
Best to show us some samples :)

Edit - oh there one is! It doesn't look right, that is for sure! Might be worth popping it on a tripod..remote release, just as a basic test. Might mean you have to pose though!
 
This was on a tripod no wind! and mirror lock up I focused at 3 metres! The only PS ive done is to raise the midtone levels a little.

Exposure: 0.125 sec ( 1 / 8 )
Aperture: f/16.0
Focal Length: 24 mm
ISO Speed: 100
Exposure Bias: 0 EV

3372526544_a7af6e9458_o.jpg


and this is the original for reference:

3371703609_6fc44e7dbe.jpg
 
Sorry just misread, you have a 5D not sure if it has micro adjust.

You could still try the basic setup for the test.
 
No. no micro adjust unfortunately
 
It will sharpen more,


33717012555aa1f5f6bbo.jpg
 
Yes I did try that. but it looks way over sharpened. And aren't I just masking a problem with the lens??
 
I think it does look like its the lens... might be worth doing as Purpleclouds suggests to make 100% sure (and maybe to show the shop you purchased it from if need be).

Anna
 
Its not soft its OOF to my eyes, the portrait is focused on the nose, and the backyard shot is focussing on the grass/balls not the fence
 
You will really have to go and conduct some proper tests on your lens. Using the depth of field calculator http://www.dofmaster.com/dofjs.html And assuming in a studio you shot from approx 10ft away, you should have a depth of field of about a foot fore and aft of the focus. So focusing on the eyes should have meant the full depth of the face in focus so I can see why you are concerned.

It does, however, look like you have focussed on the bridge of the nose. If you did have 2ft depth of field then the whole face should be in focus regardless of the focal point you used.

Find yourself a good static target and spend an hour or two doing a full range of tests and good luck, it's never a fun task but if you think you have a problem, you will be a lot better equipped to prove it.
 
Thanks for the input guy's but in the OP stated that I focused at 3 metres so yes you are right i did focus at the grass but theoretically focusing at three metres at 24mm and F16 should give me an acceptable sharpness from just about a metre from the camera to infinity! as you can see not even the grass is in sharp focus and i'm sorry but neither the fence,bin or those houses do not quite seem to be breaking the infinity barrier to me! Also as stated by AliB even if i did focus on the bridge of the nose at the Fstop i should of had at least six inches of depth of field or in focus.

Plus I have just had an message from someone saying they have the same problem!!
 
It doesn't help with your lens but I've got a Sigma 24-70 you could compare it against.


(If you're near my area of North Herts :) )
 
@ AliB, I did in fact use that very site to base my first set of tests, thanks. In fact I think you guided me to that very site some weeks back when i asked about the sweet spot of the same lens..lol
 
@ mrgubby I'm in Letchworth!
 
I'm sure this is just bad technique in focusing. If you look in the menus on your camera you should be able to enable the AF points to show on the camera preview screen and show exactly where you had that single AF point positioned on the subject.

If the portrait was central in the frame it's highly likely the AF point was positioned between the eyes or on the nose on a plain area of skin with insufficent detail for the AF system to lock on.

The second shot is just a bad example to judge from - presumably you focused on the foreground ball - how sharp is that?
 
I'd suggest it's the lens. Mine is going back to canon this week for adjustment.

I tested mine against my 70-200. Both lens at 70mm. The 70-200 was sharper at f2.8 than the 24-70 at f6.3. I'll be interested to see how my lens comes back.

Have you got enough balls in your back yard.
 
I agree with CT. Portrait is inconclusive (most likely just OOF) and the back yard pic is not a good test. If you want to see if the fence is in focus, then focus on it!

The portrait is shot close up and even at f/6.3 you've not got much depth of field. If either you or the subject moves a few inches after focusing, it'll be out.

Edit: it's very easy to do a good copy/bad copy test, but that's even less likely than usual to be a problem with a new L class lens.
 
I took this a little while ago, again on tripod mirror lock up static object at exactly 3 metres I know because I measured it. And also the lens distance gauge showed exactly 3 metres too, so we now know that bit works..lol Centre focus square dead on top half of bottle the auto focus square was a little smaller than the width of the bottle it's self.

Camera: Canon EOS 5D
Exposure: 0.2 sec (1/5)
Aperture: f/16.0
Focal Length: 24 mm
ISO Speed: 100
Exposure Bias: 0 EV

3372684625_6e024dc531_o.jpg
 
onform said:
I took this a little while ago, again on tripod mirror lock up static object at exactly 3 metres I know because I measured it. And also the lens distance gauge showed exactly 3 metres too, so we now know that bit works..lol Centre focus square dead on top half of bottle the auto focus square was a little smaller than the width of the bottle it's self.

Camera: Canon EOS 5D
Exposure: 0.2 sec (1/5)
Aperture: f/16.0
Focal Length: 24 mm
ISO Speed: 100
Exposure Bias: 0 EV

3372684625_6e024dc531_o.jpg
Click to expand...

OK that doesn't look good. :gag:

Couple of points though..

I note it was a 1/5th exposure. How did you release the shutter?

You do appreciate that any image reduced to this size will need sharpening as it loses resolution int he process? Sharpening images for web publication is mandatory regardless of whether they were sharpened at the larger size. (In fact they shouldn't be).

Also when reducing images it's vital that you have the correct reduction algorithm selected . It varies from one package to another, but generally speaking it's Bilinear for downsizing, Bicubic for upsizing.

I'm just having a bit of trouble accepting that any Canon L lens could be quite this bad, and this lack of understanding on the need to sharpen downsized images causes lots of problems here.
 
Point taken CT. There are a lot of things I have not learned yet and I have never blamed my equipment before thoroughly checking my technique or understanding of a subject.

But I think you will agree a lens of this calibre should be able to auto focus on an object at three metres away and I should be able, after forking out over eight hundred quid for this lens, to expect it to be reasonably sharp.

Is there any other tests I can do? to find out what is exactly wrong. Any excuse to pick up the camera again...lol
 
onform said:
Point taken CT. There are a lot of things I have not learned yet and I have never blamed my equipment before thoroughly checking my technique or understanding of a subject.

But I think you will agree a lens of this calibre should be able to auto focus on an object at three metres away and I should be able, after forking out over eight hundred quid for this lens, to expect it to be reasonably sharp.

Is there any other tests I can do? to find out what is exactly wrong. Any excuse to pick up the camera again...lol
Click to expand...

For over £800 on a lens it should be a lot better than that, regardless of processing (but you still need to do SOME processing :p)
 
If you're happy that you focused on that Autoglym bottle, then what I'd like you to do is post a full size crop from the centre of the full size image if you can. Don't sharpen it! :thumbs:

With regard to further tests - why all this shooting at small apertures?. Shoot at a much larger aperture, don't worry about DOF, just see how well the lens focuses on the actual item you focus on.

Put the ISO up a little and get faster shutter speeds. ;)
 
Something definately isn't right, mine is so sharp that it's scary on my 5D.

I'm with CT, shoot at a normal aperture with a nice fast shutter speed and see what you get.
 
Well with regard to shooting at small apertures, this all started when I decided to give landscape a go. I actually went out into the fields and found my spot and waited a long time for the correct light freezing me proverbials off i may add. I tried alot of different techniques to get the optimum depth of field in the photo only to find out when i got home that they were all slightly soft. of course me being me I assumed that it was me and that i had done something wrong so i went over all the meta data, consulted the on-line depth of field calculator and reviewed the images over again. then I realised that even if I had got it slightly wrong there would still be something in focus surely thats when I started doing tests.

I will do some like for like test shots along side my 50mm 1.4 all at same settings. and larger apertures. will post back when i'm done.

And thanks for everyone's help so far it's been admirable.

Matt
 
Stick with it Matt, I'm sure it's summat daft. At least be sure before you condemn the lens. ;)
 
Can I butt in and say that I don't think that this is a focus issue at all.

24mm at f/16 focussed at 3m should give everything from 86cm to infinity in focus. I can't see anything in the shot that I would describe as being in focus. If the lens was back or front focussing by half a metre then the far distance would still be pretty sharply focussed...and it's not.

I'm more inclined to think that the registration distance is somehow incorrect or one of the elements/groups is out of position.

I'll go now....

Bob
 
To anyone's knowledge, with those settings is there anything that I could be doing wrong to get these results?
 
Maybe a silly question at this stage but have you got a filter on it?

Bob
 
onform said:
To anyone's knowledge, with those settings is there anything that I could be doing wrong to get these results?
Click to expand...

To add to Bob's question about the filter.

(1) How are you releasing the shutter with these relatively long exposure times? It needs to be either the delayed timer or a remote release or you'll certainly get camera shake.. even on a tripod.

(2) What AF drive are you using - Single or Servo?

(3) You say you're shooting jpegs - what sharpening is enabled in the camera.. if any?

It may be a problem with the lens in the end but there are too many unknown elements to condemn the lens on this evidence.

By far the best way to deal with sharpening is to disable it in the camera altogether and use your own good judgement in processing to sharpen the image. It's a common misconception that you should get pin sharp digital images right out of the camera, in fact if you shoot RAW, then no sharpening is applied to the image at all. We're not machines and obviously, some images will appear sharper than others - while some will need more sharpening but sharpen perfectly well.
 
Ok here goes

1. No filter fitted, I keep meaning to buy one for protection purposes only. but the lens is clean!

2. Camera mounted on manfrotto 055proB with 322RC2 grip and ball head all on the pavement for last shot of bottle.

3. I used one shot focus instead of AI or servo

4. shutter triggered via Miror lock up on timer, so 3 sec delay after mirror locks up.

5. Plus shooting in Jpeg highest setting. No in camera sharpening and in faithful mode so as close to unprocessed as you can get as far as Jpeg's go!

And just to clarify I am hoping it is my technique or incompetence as I saved for a long time to get the gear i have now and more expense is definitely not welcome at the moment... But there is a little part of me inside hoping it's not me being a complete **** and doing something so simple that I will look like an idiot...lol plus if it is the lens my images will improve some more on solving the problem..lol but for now i'm dreading the expense! just me being honest..

More tests today. indoors with different lens as comparison.
 
Matt, while I don't like what I'm seeing any more than you do, I would like to see another test. Sending the lens back is a lot of hassle.

What I would like to see is an image where maximum sharpness is definitely obtained. In other words, a pic where there are clear out of focus areas in front of and behind a sharper subject. Try 50mm at f/4, then do the same with your 50mm lens. You shouldn't need a tripod for this as the shutter speed will be okay at ISO400. We're not looking for ultimate sharpness here, just looking to see if under the same conditions your zoom is noticeably worse than the prime. If the lens is faulty, it will stand out immediately.

I would suggest the front of a car, shot at 45degs angle and at a distance where the front fills the frame. Focus on the number plate in the middle, and at f/4 the headlights will both be out of focus. Make sure the light doesn't change - the difference between sunny and cloudy might skew things. Compare the two number plates - pics from both lenses should look identical. If they're not, the lens is a wrongun.
 
O.K an update

I didn't manage to get out in daylight as I had quite a busy day being mothers day and all.

But I did set up a little test indoors to compare the 50mm prime with the 24-70L

This time I used flash mounted on camera on ETTL and Gary Fong diff to give me the light I needed:

Three sets of blocks one at the front of the table, one in the middle and one at back of table 380mm between each stack. Blue brick is halfway between middle and rear stack.

Both lenses had same settings.
Exposure: 0.01 sec (1/100)
Aperture: f/4.0
Focal Length: 50 mm
ISO Speed: 100
Exposure Bias: 0 EV
Flash: On

all around 2 metres away camera on tripod

3376086029_1f961d6b65.jpg


24-70mm 2.8L 100% crop, only PP was auto colour and crop.
3376905460_eec4f1bf98_o.jpg


50mm 1.4 Prime 100% crop, only PP was auto colour and crop.
3376905102_9575c91def_o.jpg
 
Matt I'm not seeing a significant difference between these two lenses at all. In fact in the first pic, I've pasted the two images side by side for easier comparison. The 50mm is the one on the left.

3377372126_32012cd397_o.jpg


Neither of these images is optimally sharp, and both are lacking the sharpening in editing you'd normally expect from a full size RAW image. I've applied equal normal sharpening in the second image.

3376556807_16730f323d_o.jpg


I don't see anything wrong with this lens at all Matt.
 
You must log in or register to reply here.
Back
Top