I can't get a sharp picture for the life of me! [Archive]

jlau

11-03-2009, 06:36 PM

Camera: 7d

Lens: 17-55mm f/2.8

I bought this lense SO I could shoot in low lighting. The boyfriend thinks its the lense. I think its me. So what am I doing wrong? Some pictures are coming out sharp but very few and inconsistently. The rest are very soft.

I have tried various ranges, shutter speeds, iso's and aperture settings.

Is it because of my low shutter speeds? At 1/50 and 1/20 with apertures of f/5 my boyfriend can get a perfectly sharp shot in similar lighting (5dII with kit lens). Are my hands just not still enough? Am I choosing horrible objects to focus on? I'm getting frustrated because I can't fix what I don't know is wrong.

http://img337.imageshack.us/img337/8914/img0276k.th.jpg (http://img337.imageshack.us/i/img0276k.jpg/)

http://img145.imageshack.us/img145/8830/img01411.th.jpg (http://img145.imageshack.us/i/img01411.jpg/)

Two examples of the sharpest I can get...

Sean Setters

11-03-2009, 06:43 PM

We need more info:

1) What shutter speed/aperture were you using in those shots?

2) What were you focusing on exactly?

3) Was Image Stabilization (IS)on?

It also might help if you found a way to give us access to the full-sized images so we can see what you're seeing.

jlau

11-03-2009, 06:48 PM

I was focusing on the canon lens box, where the words "EFS 17-55MM F/2.8 IS USM" ARE

1/100 sec

f/2.8

iso: 3200

focal length: 55mm

metering mode: center weighted average

yes IS was on

the below is a crop of the area that I was focusing on at 100%

http://img190.imageshack.us/img190/1862/cropr.jpg

Daniel Browning

11-03-2009, 06:55 PM

the below is a crop of the area that I was focusing on at 100%

http://img190.imageshack.us/img190/1862/cropr.jpg

Looks like a depth of field and/or focus problem. Your target is not flat. The text on the left is more distant than the text on the right. And you'll notice that the text on the right is sharper, because it is more within the depth of field.

It's probably not a lens problem, but if you want to make sure, try shooting a flat target with liveview at 10X magnification and a tripod.

jlau

11-03-2009, 07:01 PM

Hmm...Depth of field and or focus...How do I correct that? If I stop down then I lose light in low light settings...=S

Daniel Browning

11-03-2009, 07:20 PM

Hmm...Depth of field and or focus...How do I correct that? If I stop down then I lose light in low light settings...=S

The first thing is to make sure that your focus is dead on. Your image appears to be front-focused slightly.

The second thing is that there is no way to shoot low light with deep depth of field. If you stop down and increase ISO, then you lose the advantage of having a large sensor: the noise will be as bad as a digicam (depending on how far you stop down). You can of course add more light (flash, etc.), but then you're not shooting "low light" anymore. The best solution is to learn how to use the thin DOF to your advantage. For example, if this was a portrait instead of a picture of a box, you would have one eye in focus and one eye out of focus. If you moved to the left, then both eyes would be in focus because they would both be the same distance from you.

Learning to get sharp pictures with thin DOF is difficult, but you'll get better with time.

Good luck!

jlau

11-03-2009, 07:26 PM

Thank you so much! That was very helpful!

Sheiky

11-03-2009, 07:27 PM

He means that it can be both in this case. It is hard to judge seeing what we see. The thing is: the bigger your aperture (smaller f-number), the less DOF you get. So if you want to shoot indoors in lowlight and still have a nice DOF, you need to take a flash or a longer shuttertime. That's the only ways you can correct it...

Also when there isn't a lot of light present, the AF can be somewhat less accurate as well. I suggest you trysome outdoor photo's to judge the image sharpness. And not only at f2.8, but try a few values. Hope this helps a little

I see some people are faster then me :P

By the way, also try lower iso values. iso 3200 isn't really nice and when you use the incamera highiso noise reduction, your photo's will soften a little. Try iso 100 or so with noise reduction off. Then you can really see the true sharpness of this lens. I love this lens too ;)

jlau

11-03-2009, 07:35 PM

I see I see, thank you all for the insight. I will indeed try some outdoor photos at different aperture settings...this is all starting to make some sense.

What I don't understand though is why is my boyfriend (5DII+kit lens) able to capture sharp images at f/5.6, 1/50 shutter speed, iso 2500 without flash in low light, but I can't? When I do similar settings there is little to no light!

Vince

11-03-2009, 08:24 PM

Well~ thats difference between Full frame and 1.6x body...... Full frame sensor getting more lights into, am I right?

Daniel Browning

11-03-2009, 08:33 PM

Well~ thats difference between Full frame and 1.6x body...... Full frame sensor getting more lights into, am I right?

In low light, the FF only gets more light if you shoot it at a thinner DOF. If you use the same DOF, then the light, too, is the same. FF is no more or less difficult to get sharp photos in low light. In the case of the 7D vs 5D2, I would say the 7D has the advantage due to better autofocus.

wickerprints

11-03-2009, 08:50 PM

Well~ thats difference between Full frame and 1.6x body...... Full frame sensor getting more lights into, am I right?
<div style="clear: both;"]</div>

No, this is incorrect as it pertains to exposure. For simplicity*, the amount of light per unit area reaching the image plane is determined by effective f-number and shutter speed. It is independent of the total area of the imaging sensor. It is also independent of ISO, which corresponds to the sensitivity of the imaging medium to incoming light.

That is to say, if you have a 5D Mark II and a 7D, and each is set to take an exposure of f/4 @ 1/100s @ ISO 100 @ 50mm of the same subject from the same distance, the resultant images will have the same exposure level. The only difference is that the 7D will show a narrower angle of view.

*for the pedantic, this should be effective t-number, not f-number.

Daniel Browning

11-03-2009, 09:19 PM

Great post, one minor comment:

It is independent of the total area of the imaging sensor.

You are correct that "exposure" is independent of the total area of the imaging sensor. But exposure is not the only factor to consider when it comes to low light. The most important factor is "total light", which is a home-made term for "light per area times total area". The 5D2 and 7D have the exact same sensor performance per area, but the 5D2 has far more area. So if you give them the same exposure, the 5D2 will end up with more "total light", which results in far less noise (and more dynamic range).

For example, a 3.6x2.4mm mobile phone camera with a 5mm f/2.8 lens has the exact same angle of view and exposure as a 36x24mm sensor (5D2) with a 50mm f/2.8 lens. But the total amount of light is 10 stops higher in the 5D2, and that's why it has far less noise than a mobile phone camera in low light.

You probably know all this already, I only responded because your comment made it sound like exposure was more important than total light when it comes to low light.

Kind regards,

Sinh Nhut Nguyen

11-04-2009, 02:18 PM

Tim

11-04-2009, 09:12 PM

To me, the right side of the box looks plenty sharp for being wide open at 2.8 and at ISO 3200. that high of an ISO should degrade the picture somewhat, although I don't have a new canon model, so I don't know how drastic the improvements are for high ISO.

Deva207

11-05-2009, 11:45 AM

wanna know if it's your lens or your hands, that's easy. Just take a shot with your camera mounted on a sturdytripod, IS off, ISO 100,use the timer to trigger the shutter. If the shots come out tack sharp, it means you need to work on your handholding technique.
<div style="clear: both;"]</div>

finally

rule with image stabilization, if youre shooting with tripod or from any fixed and stable surface IS must be off...

bardinjw

11-05-2009, 04:38 PM

The "kit" lens for a 5DmkII is usually a 24-105L, pretty darn sharp to begin with. The 17-55 isn't terribly sharp at f2.8 to begin with either.

Easy test: try his lens, and see if you get the same results.

jlau

11-05-2009, 07:24 PM

If that jerk would come see me I just might try out his lens...but...lets not get into that.

I thought that the 17-55 is supposed to be sharp wide open.

"This lens is sharp! Wide open and from edge to edge. Unless the distance is close that is - I'm finding that close subjects do not produce the same image sharpness as normal distance subjects." according to the review on this site...

So my question now, is what is normal distance versus close? I tried to gauge using the lens but i'm still a bit lost...

Tim

11-05-2009, 11:31 PM

I thought that the 17-55 is supposed to be sharp wide open.

"This lens is sharp! Wide open and from edge to edge. Unless the distance is close that is - I'm finding that close subjects do not produce the same image sharpness as normal distance subjects." according to the review on this site...

Check out the comparison of these two lenses at 2.8. I know they are on different bodies, but the 16-35L is much sharper at 2.8, yet both are reasonably sharp wide open.

http://www.The-Digital-Picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=398&Camera=474&Sample=0&FLI=4&API= 0&LensComp=412&CameraComp=453&SampleComp=0&FLIComp =4&APIComp=0 ("http://www.The-Digital-Picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=398&Camera=474&Sample=0&FLI=4&API= 0&LensComp=412&CameraComp=453&SampleComp=0&FLIComp =4&APIComp=0)

Daniel Browning

11-06-2009, 12:16 AM

Check out the comparison of these two lenses at 2.8. I know they are on different bodies, but the 16-35L is much sharper at 2.8, yet both are reasonably sharp wide open.

I think you are mistaken. If you were to put the 16-35 on the 50D, you would find out that the 17-55 is sharper than the 16-35. Read Byran's review:
<p style="padding-left: 30px;"]

My Canon EF 16-35mm f/2.8 L USM Lens ("http://the-digital-picture.com/Reviews/Canon-EF-16-35mm-f-2.8-L-USM-Lens-Review.aspx)
is slightly sharper in the center at 17mm wide open, but the Canon EF-S
17-55mm f/2.8 IS USM Lens is sharper at all other tested focal lengths
- and proved much sharper in the corners at all focal lengths and
apertures.
Distortion was also less on the 17-55.
In my opinion, the only reasons to buy the 16-35 over the 17-55 are for
full-frame compatibility (a big reason), better build quality and
environmental sealing (EF-S compatible bodies are not weather sealed at
this time).

Your mistake was using the charts incorrectly. That is why Bryan specifically points out that charts from different bodies cannot be compared for lens performance.

wickerprints

11-06-2009, 12:41 AM

If that jerk would come see me I just might try out his lens...but...lets not get into that.

Sorry, couldn't resist making a comment...I just want to point out that lucky is the man who has a significant other with whom he shares an interest in photography. Very, very lucky.

Can you try posting a test chart? It doesn't have to be elaborate. Just take a photo of a calendar hanging on the wall, something where you can confirm focus on a flat target. Taking photos of three-dimensional objects can be misleading due to shallow DOF.

jlau

11-06-2009, 01:20 AM

You try telling him that! Ugh. a wijajlkd awjds lkjsjlkdalsd.

Anyways, here is a...ugh...my attempt at a test chart.

http://img25.imageshack.us/img25/7687/sdfsdfuu.jpg

jks_photo

11-06-2009, 01:20 AM

hi,

have you ever tried setting to your focus point on the camera?? try to set set you camera to always focus using the center focus point. that way you'll be sure that your camera will focus only on the subject where the center focus point is on. You may may have set it at automatic whereby the camera chooses where to focus. so though it gives you the green light saying its focused, it may have actually focused on a different part of the whole image and not necessarily on what you are focusing at. Try to check this out. The red blinking lights you see when you look into your veiwfinder, the one in the center should always be "the only one" blinking when you half press the shutter and the camera starts to focus.

This doesn't mean that your subject has to be in the center of the shot all the time. You can always focus on the subject first [half press the shutter] then without releasing the half press on the shutter recompose the shot so that the subject is on one side of the shot if that's what you want. What you can be certain of is that the subject will be in focus and judging by the camera and lens you use you should have no problem getting sharp images.

jlau

11-06-2009, 01:25 AM

I am not sure what you mean by focus point? I have tried all metering options on the camera including, evaluative, partial, spot, and centerweighted average...I am assuming that the AF points that I see through my viewfinder are what the camera is focusing on right? Objects are not moving, and the only movement I'm contributing may or may not be my hands so I am not sure...where I am going with this anymore...but...yes...

jks_photo

11-06-2009, 01:35 AM

hi

yes they are called AF points. sorry for the wrong terminology.

yes,try to set your camera so it ONLY uses the center AF point. thats the one in the center of your viewfinder.

jks_photo

11-06-2009, 01:43 AM

the metering options on the camera only controls how the camera meters or calculates the "right" exposure. It doesn't really have much to do [at least for me] with the sharpness of the picture.

I would suggest shooting in the AV mode. That way you control the depth of field which determines which part of the image is sharp which is "out-of-focus" with your 2.8 lens this will be easy. But to get sharpest image try not to use openings higher than f4. so 5.6, 8, 11 all will give you sharp images.

by the way also try to adjust the sharpness control on the picture style you currently use this will also help.

wickerprints

11-06-2009, 02:19 AM

Well, you could post his email address on here and watch dozens of photographers write him messages saying what a dork he's being. (Kidding, of course) [:P][;)]

As for your test chart, it's a good start. Here are a few comments:

Sorry for not telling you earlier, but when you shoot the test target, make sure you force the camera to focus using only the center AF point. This is not a huge issue, for reasons I will explain later.

Post two 100% crops from each shot. The first crop should be from the center, and the second crop should be any of the corners. If your test target is not large enough to fill the entire frame, a simple solution would be to take another object (like a piece of paper with printing on it) and stick it on the wall in the spot where it'll show up in a corner of the photo. These two crops help us determine whether your lens is faulty, but even better would be to compare the performance from corner to corner, because if the lens is misaligned, you could have one half of the image sharper than the other!

From what I can see from your test images, the performance @ 55mm is not as good as @ 20mm. I realize you were shooting in available light so you were going to ISO 1600, but try finding a nice bright lamp to shine on your target so you can hopefully shoot at ISO 100-200. This will enable us to see just how much detail can be resolved by the sensor+lens combo.
Note that if the lens is not up to spec, the solution is simply recalibration by Canon. Judging from the 100% crop of the image, I would say your camera's sensor is capable of resolving enough fine detail @ ISO 1600, and that it is the lens that does look a little soft @ 20mm and more so @ 55mm.
Shooting this test is very, very informal. It's not going to show you subtle defects, only major ones. But that is the goal, to diagnose major problems.

It's hard to say for sure whether the lens is not up to spec. If you can test out the EF 24-105/4L IS, that would be an excellent next step. And if you aren't satisfied with the EF-S 17-55/2.8 I would not hesitate to send it to Canon to have it looked at. Good luck!

jlau

11-06-2009, 02:03 PM

Lol thank you for putting a smile on my face. Anyways...

So I've tested my lens in daylight (FINALLY) and I'm too lazy to post my results but it is indeed sharp with enough light out even AT 2.8. Not so much when about 1-1.5m away from the subject though (no zoom). Is that normal? Also, is it normal that I cannot get a sharp image at 2.8 in LOW light?

Oh and just wanted to thank everyone who's helped me so far =)

jlau

11-06-2009, 02:05 PM

So I changed the focus to only focusing on the centre of the frame. I THINK its helped, but I was shooting in daylight today so I will test this out tonight.

I have found that indeed f/4 range is sharper, but the point if my getting this lens is that I can use it in low light with aperture wide open. Otherwise...what was the point of getting a 2.8 if it can't give me a sharp image?

Where is this sharpness control...??

elmo_2006

11-06-2009, 02:47 PM

Having a lens at f/2.8 will not garantee sharp images, it allows the user to shoot images in lower light without a tripod (better if IS is available for further step-downs). A lens @ f/2.8 allows more light into the lens/camera. You can try taking a photo at f/2.8 and check the aperature within the viewfinder, then change it to f/4 and so on. With each step down, you will notice the aperature time increases. Whereas at f/2.8 you may get 1/13 of a sec, at f/4 you may get 3 seconds. But withf/4 and lower, you will get some really beautiful background blur and a narrower depth of field!

Hope this helps.

jks_photo

11-06-2009, 09:42 PM

I have found that indeed f/4 range is sharper, but the point if my getting this lens is that I can use it in low light with aperture wide open. Otherwise...what was the point of getting a 2.8 if it can't give me a sharp image?

Where is this sharpness control...??

Image sharpness at certain lens openings vary from lens to lens. Try to read the lens review for your lens on this site so you can get an idea on what gives opening gives the sharpest image and and at what zoom range. Like what was said having a 2.8 lens gives you somewhat better options when shooting in low light but that dopesn't necessarily mean that the 2.8 lens would give the sharpest image at 2.8 opening.

as for the sharpness controls... its found in the camera menu. not really sure with the 7d but on my Xti it's within the picture style controls. that is to say you have to know how the adjust parameters oif your picture style to find the sharpness control. You can also post process using Digital Photo Pro that canon supplied but I think you have to shoot in RAW format to be able to adjust sharpness.

bob williams

11-08-2009, 10:45 AM

Like Jlau, I too have had this problem. Many, if not most of my shots are a little to soft for my taste in comparison to what I see other photogs posting, so I have been intently following this post for ideas. One thing I learned last night is that perhaps my expectations are too high for the equipment I am shooting with and my skill level. Based someones suggestion, I took a look at pixel peeper. com and found most of those shooting with the same equipment were getting similar results as I was---especially when viewed at full resolution. I primarily shoot with a 50d and a 100-400L and my tack sharp images are far and few between. When I looked at different combos, i.e. the 100-400 and a 5dII, the results were far different. The 5dII was much sharper than my 50D with the same lens. The same thing applies when Ilooked at the 50D and 70-200L combos, thier shots were much sharper and aestheticlly pleasing. Though like most, I am always considering upgrades, but for the moment I will continue to work with what I have. I have noticed that those shots I take my time with and really pay attention to focus and other settings---they are much better than those I take in a hurry----unfortunately, when shooting wildlife, a hurried shot is sometimes the only shot. I am curious what others think in regards to this.

Thanks for you post,

Bob

Alexander

11-08-2009, 11:13 AM

Are you using one autofocus point? You could get a bit of trouble using all autofocus points.

elmo_2006

11-08-2009, 12:48 PM

To make a long story short, typically the full-frame cameras do a better job at image quality than theFOCV cameras and some lenses are geared for this format moreso than others. I'm unable to locate the link but I have read that plenty of the 'L' lenses from Canon were specifically intended for the full-frame format.

Hope this helps.

bob williams

11-08-2009, 02:04 PM

Thanks, That helps reassure me thatI wasn't just seeing things.

bob williams

11-08-2009, 06:24 PM