myrjola

I wondered whether the problem could be caused by some dirt below the power switch, so I took a piece of thread and guided it between the power switch, tugging it gently back and forth, a bit like flossing teeth. Now only pushing the power switch gently and then leaving it alone is enough to keep the wheel working as long as I keep the switch in the "L" position, so that might be a partial explanation for the issue.

 

The current state of the switch is still not quite perfect, but clearly an improvement, so I guess I'll live with it for now and check about the service costs at some time when I know I won't be needing the camera much (or sooner if the fault gets worse, of course).

 

(Peter, I agree that it would be unpleasant to part with the 20D; it's been my first DSLR and even though I suppose it's only a tool, I have many warm and fuzzy memories of using it. I suppose that even when I eventually upgrade, I'll keep it as an emergency backup as long as it functions at all)

Yesterday, I noticed that my venerable 20D seems to have developed a small

malfunction. When I turned it on as usual (power switch to the usual topmost, or

L, / or whatever one wants to call it - position) and tried to change aperture

in manual mode, the quick dial wheel did absolutely nothing. A quick test

revealed that the wheel itself was working, as menus etc. still reacted to

rotating it. However, it now works just like the power switch would be in the

"On" position, even though it's at the position where the wheel should work also

while shooting. I also noticed that touching the power switch gently while

rotating the wheel makes the wheel to work until I lift off my finger from the

power switch, which might indicate some kind of worn contact.

 

So, I'm wondering whether anybody has encountered the same issue with his/her

20D? If so, is there some easy and safe DIY fix for it, or did the camera need

service? In case of service, what kind of expenses are we talking about? The

issue is not truly crippling, but nonetheless slightly annoying, so I'd want to

eventually get rid of it and I'm interested to know what my options are. (Well,

they are "live with it", "get it serviced" and "upgrade to a new body", but

making decisions is nicer when there's some information :)

Sorry, I was a bit tired while writing the comment (and English is not my native language), which does tend to affect quality of my grammar sometimes.

 

I meant slow in the sense of AF performance. However, it's not really important, just something I wondered about while writing the message.

Brian, I agree that the difference isn't exactly huge; slow improved by 1.8x obviously isn't still fast. It's just noticeably less slow, but slow nonetheless. I recall that one review of the Mark II stated that it's still about 4x slower than the non-L 85mm f/1.8 as far as the focusing goes, which I believe has a focusing speed quite comparable to most other USM-equipped Canon lenses. I was just commenting about the 1/8 designation, which might be interpreted as being 12.5% increase.

 

(As a related item of curiosity, anybody has an idea about what's the second slowest lens after the 85mm L in Canon lineup?)

Regarding the mk II and mk I focusing speed difference, I'm under the impression that the newer model should be about 1.6-1.8 faster, not 1/8 faster, making the difference quite apparent. While that probably still doesn't beat any other USM-equipped Canon lenses, it's a nice improvement.

 

(Personally, I own only the older model. Once I got used to the lens, the focusing speed isn't actually as limiting as I thought first.)

Two cats here. I used to think I was a dog person, but they have converted me. Perhaps their gracefulness is another factor that makes them attractive to photographers, although most shots I've seen are indeed of stationary cats. Then again, almost all of my shots of cats in motion do fail due to their speed.

I'm mostly using a Linux machine for my photography-related things. On the software side, I mostly use Bibble for RAW-related processing and GIMP for the general photo retouching, although there are of course other applications that I occasionally use (Hugin for panorama creation etc).

 

(Yes, there's more to choose for image processing on Windows/Mac, but so far, the tools available for Linux have been good enough for an amateur like me.)

I bought an (used, but I had eventually purchased a new one if I wouldn't have had the oppoturnity to get a secondhand one) 200mm f/2.8L last year for the purpose of occasional astrophotography; the 70-300 DO works well enough for my general tele needs, but it's somewhat slow for that kind of usage. And even though the DO image quality wasn't too bad, the prime will do a bit better.

 

(Naturally, the weather has been pretty horrible most of the fall and the early winter, so I've had very few oppoturnities to actually use it, Murphy strikes again...)

The DO has also a more sturdy build quality. I'd say that the non-DO is quite a bit better bang for the buck if the size, AF speed and sturdiness are not significant factors, as the non-DO has better image quality in some situations and costs quite a bit less.

 

(I bought the DO at the time when the non-DO - 70-300 was not yet released; while I do not regret my decision, I'm not sure which one I'd choose now personally; I value robustness, compact size and good AF, but I'm not sure if I'd want to pay that much more about it...)

Regarding the DO bokeh; to me it seems that the 'weird' bokeh tends to occur occasionally near or at the 300mm end, but not really on the shorter focal lenghts. (I've had the lens for a couple of years now and not really seen the slightly doughnut-like bokeh very often)

Any EF lens you buy will work with all Canon DSLRs. However, there are also EF-S lenses, which won't be compatible with all Canon DSLR bodies, as they are designed for the crop sensor camera bodies (which the 350D also is). So, if you upgrade later to a camera with a larger sensor, the EF-S lens won't be compatible. If your upgrade will still have a sensor with 1.6x crop (which the successors for the current xxxD and x0D cameras will probably retain for a while), then you can buy EF-S lenses without compatibility worries.

 

How wide angle do you want? The Canon's widest lens (10-22 zoom, which gives the approximate field of view equivalent on full-frame/35mm film of 16-35mm lens) is unfortunately an EF-S lens, so you can't use that if you move to a camera with a larger sensor. Moving slightly up from that, there is the 14mm f/2.8L. However, it's very expensive. There's also a much cheaper 15mm fisheye lens, but I doubt most people don't find it a very general-purpose photographic tool due to the fisheye effect.

 

After that, the options open up a bit, with several EF zoom lenses such as the 17-40L f/4 (field of view is about equivalent to 27-64mm on film/full frame, I have this lens myself and find it pretty good, slightly expensive) and 20-35mm f/3.5-4.5 (quite a bit cheaper, FOV equivalent to 32-56mm, apparently decent image quality). There are quite a few more lenses available if minimum focal length of 24mm is OK, but I'd personally think them as no longer really covering wide angle when used on 350D.

 

Another option is to look for a EF mount-compatible 3rd party lens; many of them are perfectly fine. The widest truly EF-compatible lens I know is Sigma AF 12-24mm f/4.5-5.6, which is an ultra-wide lens even on non-fullframe Canon cameras and rather amazingly wide on full-frame cameras.

One way to eliminate any NR that occurs during RAW conversion would be to decode images with something like dcraw. While its newer versions appear to support optional noise reduction, by default it uses very little metadata for the conversion (only camera WB info, I think) and definitely isn't doing any NR based on any tags in the image. Dcraw can be found at:

 

http://cybercom.net/~dcoffin/dcraw/

 

(the page has a separate link related to Windows executables)

At least for me the 1.6x crop factor seems to be a valid factor to include in the aforementioned rule of thumb. There is some dependency on the lens used though; for example, I seem to be able to get comparatively better results with exposures exceeding the suggested time with my 17-40L than with my 50mm f/1.4; probably because the shape of 17-40L allows a more stable grip to hold the camera/lens combination steadier.

One downside of turning the sharpness all the way up on the camera body is that the image might be a bit more prone to gain artifacts if it's sharpened again (which you might want to do after e.g. downscaling the image), or otherwise postprocessed on a desktop computer.

While I have the 70-300 DO instead of the non-DO, I'd say that if you're eliminated camera shake and other potential not-really-the-fault-of-the-lens issues, the lens isn't quite performing as it should. I'm saying this because both the non-DO and DO 700-300's should provide better image quality than the old 75-300 (with some caveats, such as the flare-proneness of the DO) according to the comparisons I've seen.

I think the idea of 20D having slightly more noise at low ISO settings has originated from the Dpreview test of the 20D; at least that is where I personally remember seeing it for the first time. More specifically, from this subpage:

 

http://www.dpreview.com/reviews/canoneos20d/page20.asp

Although this isn't really an answer to the original question (I don't have 24-105 f/4 L, nor the 5D), I'd say that zooms can work quite well with the extension tubes. Of the four lenses I own, the 70-300 DO IS provides the best image quality when used with the tubes, which surprised me somewhat, as in normal use the three other lenses have generally better image quality.

 

The other three lenses are 50mm f/1.4 (quite heavy chromatic aberration and some distortions near the edges with the tubes), 85mm f/1.2L (quite a bit of CA near the edges as well, although better image quality than with the 50mm) and 17-40L f/4 (surprisingly better than the two primes, but the _really_ minimal working distance makes it somewhat painful to use with the tubes).

 

I guess my point is, checking whether you have better luck with some other lenses might also be worth it; not all lenses seem to be suitable to use with the tubes equally well.

I for one hope that the fast primes have a future, despite sensor improvements and image stabilization. It's not that rare for me to run into situations where the combination of ISO 1600 and f/1.2 is bordering on the edge of insufficient light. Sure, a tripod would help, but lugging it around is quite a bit more cumbersome than having an additional lens in my bag. Additionally, I doubt that sensors keep up their current rate of improvement; while there are improvements to be gained, at some point the laws of physics will start to get in the way.

I have firsthand experience only about the DO so feel free to take my opinion with a grain of salt, but here goes anyway:

 

DO is a bit quirky lens; occasionally it delivers excellent image quality, but can give rather muddy results as well. While there are quite a few rules of thumb (for example, image quality DO is more likely to suffer when shooting into light than most other lenses), it's still somewhat more unpredictable than my other lenses. I bought mine mostly because of its portability; when locked into the 70mm position it's really short and compact compared to other lenses with roughly similar focal length coverage. (Also, I bought mine before the new 70-300 non-DO was released - that's also one lens you might want to consider, if you can live with somewhat less sturdy build quality than the lenses you mention do have, as well as with the rotating front element. On the plus side, it's significantly cheaper and has at least as good image quality)

 

I'd _roughly_ categorize your candiates something like this:

 

- A compact carry-easily-around lens; the DO (also applies somewhat to the non-DO 70-300) It's small compared to the other two.

 

- The best image quality and low-light performance: 70-200 f/2.8 IS L

 

- Least need to swap lenses: 28-300 (you might get a couple of other lenses with the same money that one costs, though...)

I'm using Av most of the time. If I use flash or want to play with > 30 second exposures in Bulb mode, I use manual mode. I use other modes very, very rarely.

My secondhand 85mm f/1.2L that I bought recently appears to be made in September, 1990. Despite 16 years of age, it seems to work just as nicely as a brand new one, although it's of course possible that it's been serviced at some point during its life.

Hi, ExifTool might be a possible answer. However, it doesn't have a graphical user interface and you'll also need a Perl interpreter to run it. If you're an experienced user of the command prompt neither of these points will probably be a significant problem, but otherwise there may be an initial learning curve. But anyway, ExifTool is available at:

 

http://www.sno.phy.queensu.ca/~phil/exiftool/

Hi,

 

I can also hear the IS starting up when using my 70-300 DO and the lens has worked nicely for me about year and half by now, so I think it's normal for that lens.

Regarding the 3rd party extension tubes and aperture, I think most 3rd party extension tubes do include a pass-through for the electric connections from the camera body to the lens, so aperture should be controllable just fine. The Kenko 3-tube set I got last year certainly allows the lens to function just as usual (other than the penalty to AF performance due to light loss, but then again, I tend to use manual focusing anyway).

 

But sure, it's a good idea to check before buying in case.

This might be at least partially caused by the fact that when using tungsten white balance, the camera attempts to correct the effect of predominantly reddish light and thus has to amplify especially the blue channel significantly, making any noise from that channel more visible.

 

I'm not sure if that's all in this case though; with ISO 100 the effect should be fairly minor if the exposure is not completely off the mark. Do you have the long (i.e. 1.3 seconds and up) exposure noise reduction turned on; if not, that might also be worth experimenting with.