matthias_meixner2

If you shoot landscape, it is probably a better idea to keep the 17-85 and add an EF-S 10-22 to your set of lenses. This will give you real wide angle and is probably more useful than some better image quality shooting wide open in a range that is already covered by another lens.

If you have got problems with the resolution, there are two possilble causes:

 

1) AF is not working correct: Some other part of the image is sharp but not the "right" one. This is easy to check: Activate the center AF field and make some test shots. If something in front of or behind the subject is sharp but not the subject itself, you have a problem with AF. This is probably a case for warranty.

 

2) Your lens is not sharp enough. Try to stop down to f/8 or F/11 and check sharpness. If it is sharp by then, one of the mentioned L-lenses will help to get better sharpness shooting wide open. (I replaced a 28-105 by a 24-150L and it did make a big difference shooting wide open)

Ed wrote:

> As far as smaller and lighter, I think the 24-70 definitely wins.

 

According to the published specs this is wrong: The 24-70 weighs 950g whereas the 24-150 only weighs 670g. This is what the canon website says.

The EF-S 10-22 and the EF 24-105L make a nice combination for an APS-C dslr.

It all depends on the viewing distance. If it is 100m even a 30mx20m print will look sharp.

At f/6.3 your autofocus may have problems, since it is only designed to work up to f/5.6, only some of the (semi-) professional cameras are designed to work with f/8.

The 25-70 is considerably heavier than the 24-105. Therefore, if weight is an issue, the 24-105 is the way to go. I would not want to carry around the 24-70 on my neck all day, the 24-105 is already quite heavy ...

> Sigma 12-24 or 15-30... surprising good, and you can use either one on > your EOS 3 fullframe as well.

 

The Sigma 12-24 seems to suffer heavy sample variations. I have read a report of someone who tested I think 6 lenses. The first were all bad but the last one was reported to be very good. So it seems to be a good idea if you can test the lens before actually buying it.

In-body IS has one advantage and two disadvantages:

 

+ It is working with all existing non-IS lenses.

 

- It does not stabilize the viewfinder

 

- The shifting of the sensor is limited to a fraction of a millimeter while IS in the lens may shift around the image by "any degree". So IS in the lens can be more effective than IS in the body.

Due to the larger sensor pixel density on the 5D is lower than that of the 350D. Therefore, while it may be the case that images taken using the 28-135 on the 350D are less sharp than those taken with the 17-85 on the 350D, while images taken with the 28-135 on the 5D are sharper than those taken with the 17-85 on the 350D.

Simple answer: Get the EF-S 10-22.

> However, never give up speed for IS so instead of the f/4 IS buy the non IS f/2.8 version.. still an awsome lens.

 

Speed is not a substitute for IS. If you want large DOF (e.g. for landscape) a faster lens does not help. So whether IS or a faster lens is better depends on what you are shooting.

I wonder why anyone would care about the internal lens design. If the optical quality was superb I would not waste time on thinking about design details. So just wait and see how it performs.

> 1) is the rule of thumb the lower the f-stop (numerically, that is, im

> not sure the correct lingo for saying a larger opening aperture), the

> sharper the photo can be?

 

Yes and no. If you have got a perfect lens, i.e. it is diffraction limited, then yes, the wider the aperture the sharper the image due to the larger opening generating less diffraction (and the shallower the DOF). However most lenses are not perfect and, therefore, the smaller the aperture the more "defects" of the lens are hidden behind the aperture blades and the sharper the image. So in real lenses there is some f-stop which is the best compromise of the two effects and in most cases this is not the lowest f-stop.

Substitute the 19-35 with an EFS 10-22 and 24-105L and then see what you are still missing.

I have read a posting in another forum about this lens: The poster tested several copies of that lens and was about giving up. But then the 8th lens he tested was sharp and he kept it. So Sigma seems to have quite some quality problems with this lens.

> What is that is so magically about 35x24mm that means that it is the

> optimum size for future sensors?

 

It is easier to achieve high resolution and low noise at high light sensitivity with larger sensors. If you want to have the same amount of light per pixel (at the same resolution), the absolute diameter of the aperture must be the same independent of the sensor size, i.e. you need a 25mm 1/0.7 lens on a 17.5x12mm sensor to gather the same number of photons per pixel that are gathered using a 50mm 1/1.4 lens on FF.

 

So the larger the sensor the better the potential quality. So what is so magic about 35x24mm? It is the largest format that can be used with most existing lenses. OK, you could have square sensors, but that is another story.

The mentioned price is what Canon would like to have. But it all depends on the market if they get it or not. If the sales do not develop as expected, the price will drop. Overall it is probably better for Canon to start with a high price, since then they have less problems satisfying the (lower) demands and it is easier to significantly lower the price than to increase it.

Robin you are right if you refer to a perfect lens. But most lenses suffer from pincushion or barrel distortions. And, therefore, you will most likely see something egg-shaped and not an ellipse.

A simple experiment will explain why IS works better with telephoto lenses. Take a telephoto lens and a wide-angle lens and shoot two pictures of a ball with each lens: One with the ball in the center and one with the ball near the corner of the image. Now look at the pictures. The ball on the telephoto picture will be round in both cases. But on the wide-angle photo it will be round only in the center and look egg-shaped near the corner.

 

Image stabilization can be considered equal to moving the sensor to compensate for camera shake (Canon moves the image by some additional optics instead of moving the sensor, but the effect is the same). This works perfect with a telephoto lens, since the subjects always have the same shape wherever they are in the image. But with a wide-angle lens camera shake results not only in a movement of the image but also in a distortion of the image (the ball became egg-shaped near the corners). This cannot be compensated by moving the sensor, i.e. image stabilization is less useful.

The Secure in SD refers to support of some kind of digital rights management. That is, the card is able to lock in some content so that you cannot copy it. This is only useful for someone who tries to sell some content to you. For the use in a digital camera its useless at best.

Scanning film was the way I started to "digital photography". The problem here is: Flat bed scanners do not deliver high quality scanning film. Either you scan prints or you need an expensive film scanner. Then scanning film is very time consuming. It took me about 3-5 Minutes per slide (preview scan, setting the scan parameters ....). So scanning say 100 slides is real work.

 

And then you miss one of the most important benefit of digital photography: direct feedback. With a digital camera you can immediately see the result and learn from your errors which is not the case with film.

> Paul: assumptions have been made in the creation of the very concept

of "depth of field". The "circle of confusion" that is acceptable to

one person may not be to another. In fact, the laws of Physics and the

mathematics used to describe the focal plane show that depth of field

does not actually exist.

 

Yes and No. There are more things to consider: Resolution of the camera and diffraction limit. If the circle of confusion is well below the pixel size of the camera and/or within the limits of diffraction, you will not be able to see any difference between an in focus area and a slightly out of focus area. However, this also means that with each increase in resolution of the camera, the DOF gets smaller and smaller.

I received a Canon EF 24-105 L today. Now I wonder what this gentle click is,

when IS is activated or deactivated. Is it the IS mechanism being (un-)locked to

prevent damage to it when not active? It sounds very similar to a closing of the

aperture.

Yes, some of you were right: I still have an EOS 50E. That's why I have the 20-35. The 18-55 came along with the 350D in a kit (since it was a good deal, I took it. It may help to better sell off the 350D some day).

 

I have two concerns:

 

- improve image quality over the 28-105

 

- improve low light capability

 

Wide angle is not required, since I already have the 10-22 and am quite satisfied with it. I only rarely use the 100-300 USM, so I can currently live with the quality it provides (when stopped down to f/11 it is OK).

 

For weight reasons I initially thought about the 24-105 IS. The reviews I have read seem to indicate that at f/4 it is on par with the 24-70 and then IS will probably be more useful, since it combines with all apertures.

 

Now that the 17-85 IS has been ruled out for quality reasons (I never really believed what they had told me in the camera store), I guess will decide between the 24-105IS and 24-70.