Few questions about the 50D

[HiFiGuy1]

I concur with what Ben said, for the most part. The only point I'm not sure about is "There's a pretty big step up in image quality between the 40/50D and the 7D." There's a big step in resolution between the 40D (10 megapixels) and the 7D (18 megapixels), but not much a step from the 50D (15 megapixels) to the 7D. The 5D, as a FF camera, is a significant step up in terms of image quality especially in the ISO noise performance. But keep in mind that with FF, you lose the 1.6x crop factor - in practice that means that if you find yourself frequently cropping your images, you'll lose the benefit of FF (look over your current shots - if you crop shots from your XSi already, a 5D will be worse than the XSi, for you).


Taking FF out of the picture, ISO noise may be a differentiator - from Bryan's review of the 7D, "I'm particularly impressed at the Canon EOS 7D's noise performance against the 50D. Comparing the gray blocks at the top of the 50D's image shows the 7D to have a slight advantage over the 50D at ISO 800 that builds to a solid advantage at ISO 3200. By ISO 12800, the 7D solidly outperforms the 50D." However, as a 7D user I'm not particularly thrilled with ISO 3200, which still looks pretty noisy to me even after cleanup of the RAW file. But then again, an image that's shot at ISO 3200 on the 7D, not cropped much, and printed 8x10" still looks good. The 7D may 'solidly outperform' the 50D at ISO 12800, but I've never been desparate enough to go that high.


In terms of ISO noise on the 40D vs. 50D, they are pretty similar there (and kudos to Canon for a 50% increase in pixel density without taking a hit on noise!).


Having said all of that, honestly, there are not going to be huge differences in image quality between the XSi that you have and the 50D or 7D. Incremental differences, yes. But not huge. I upgraded from a T1i (very similar IQ to the 50D) to a 7D, and really didn't notice a big difference based on sensor performance. IMO, the biggest differences are in other areas.


Build quality takes a big step up from XSi to 40/50D, and another noticeable step up from 50D to 7D. The ergonomics are also better, IMO - having the dial and joystick on the rear means much faster adjustments of image parameters than on a Rebel.


As Ben mentioned, AF Microadjustment can be a significant benefit. Most of my lenses have some amount of adjustment applied. Beyond the microadjustment feature , the AF performance itself will be a bit better with the 50D (9 cross-type sensors) vs. your XSi (similar 9 points, but only the center is cross-type). The 7D's AF system is excellent, with 19 cross-type sensors and lots of customizable options for AI Servo tracking.


If you do shoot wildlife/birds/sports with your 300mm f/4L that's on your camera 90% of the time, the other big benefit you'll notice is frame rate. Your XSi is 3.5 fps, the 40/50D are nearly twice as fast (6.5/6.3 fps), and the 7D is over twice as fast with 8 fps. Here's an example of two successive frames from an 8 fps burst:


[img]/cfs-file.ashx/__key/CommunityServer.Components.UserFiles/00.00.00.35.15/IMG_5F00_2847.jpg[/img] [img]/cfs-file.ashx/__key/CommunityServer.Components.UserFiles/00.00.00.35.15/IMG_5F00_2848.jpg[/img]


EOS 7D,EF 100-400mm f/4.5-5.6<span style="color:red;"]LIS USM @ 400mm, 1/2000, f/6.3, ISO 3200


On the whole, if your reasons for wanting to upgrade your camera body are primarily around image quality, I'd say keep the XSi and consider lenses (what's that other 10% shot with, and do you want better IQ there?), a Speedlite flash if you don't have one, etc. In general, lenses have a much greater impact on image quality than the camera body.


But, if your reasons also include overall performance, ergonomics, etc., then an upgrade is worthwhile. In that case, you can't really go wrong with the 40D, 50D, or 7D. If you can stretch your budget to cover the 7D, it's an amazing camera. If the 50D fits your budget, I'd get that over the 40D.

Good insight, as usual. I do have one question for you though. You say in one sentence that most of your lenses have at least some AF Microadjustment applied to them, so that seems to imply that your IQ would be improved even with just a 50D over a 40D, all else being equal, regardless of resolution advantage.


Then later you said that if IQ were the only concern, that you'd keep the XSi and get better glass. If I am not mistaken the XSi doesn't have AF Microadjust, and wouldn't it be better to optimize the good glass he already has more than buy more glass? I think the 300/4 is an already excellent lens, and it seems like a clear way to optimize it is by properly adjusting its AF to match the body perfectly.


These are really questions to which I don't know the answer, so please don't take it as argumentative. Since I haven't experienced AF Microadjust, and I don't honestly know if my 40D is off with either or both of my current lenses, I don't know what I'm missing. Based on some shots I see here, when compared with some of my own carefully AFed shots, I wonder if mine is slightly out, though, and the problem is I don't think there is anything I can do about it without changing bodies.


Am I wrong?

[neuroanatomist]

Good insight, as usual. I do have one question for you though. You say in one sentence that most of your lenses have at least some AF Microadjustment applied to them, so that seems to imply that your IQ would be improved even with just a 50D over a 40D, all else being equal, regardless of resolution advantage.


Then later you said that if IQ were the only concern, that you'd keep the XSi and get better glass. If I am not mistaken the XSi doesn't have AF Microadjust, and wouldn't it be better to optimize the good glass he already has more than buy more glass? I think the 300/4 is an already excellent lens, and it seems like a clear way to optimize it is by properly adjusting its AF to match the body perfectly.

Good points. In discussing relative IQ, I was referring to sensor performance for the post part. Certainly, if there's an AF adjustment problem, that's going to have a major impact on IQ.


Although I have an AF microadjustment applied to most of my lenses, I'd be hard-pressed to say it's required.My testing is pretty rigorous, using a [url="http://www.lensalign.com/]LensAlign Pro[/url]. If I was just eyeballing sharpness, I may not have even noticed an adjustment should be applied in most cases. I've tested 10 lenses on my 7D(well, 9 lenses, one combined with an extender), and here are the adjustments applied:


-5 EF-S 10-22mm f/3.5-4.5 USM
+1 EF-S 17-55mm f/2.8 IS USM
+2 EF 24-105mm f/4<span style="color: red;"]LIS USM
+2 EF 70-200mm f/2.8<span style="color: red;"]LIS II USM
+6 EF 70-200mm f/2.8<span style="color: red;"]LIS II USM +EF 1.4x II Extender
0 EF 85mm f/1.8 USM
-1 EF 100mm f/2.8<span style="color: red;"]LMacro IS USM
-1 EF 100-400mm f/4.5-5.6<span style="color: red;"]LIS USM
-2 EF 200mm f/2.8<span style="color: red;"]L IIUSM
+2 EF 300mm f/4<span style="color: red;"]LUSM


<span style="color: red;"]



So, first off for 8 of the 10 lenses, the adjustment was between -2 and +2. One unit of adjustment is 1/8 of the DoF at max aperture (where focusing is done), so at most an adjustment of 1/4 of the DoF was applied in 8/10 cases. The rated accuracy of the AF system is within the DoF for lenses slower than f/2.8, and within 1/3 of the DoF with f/2.8 and faster lenses using the high-precision f/2.8 center AF point only. So, in all those cases, the non-adjusted AF accuracy falls within the rated accuracy of the lens. What that means is that random chance could result in an OOF shot that AF adjustment would have no hope of correcting (since it's not a systematic error). Most of the time, being off by 1/4 the DoF would not be noticed - especially when not shooting wide open on a fast lens. Of the two adjustments larger than 2 units, it's unlikely the 10-22mm would ever have a noticeable issue - it's a landscape lens, almost never shot wide open, and for most shots with that lens, everything from a couple of feet to infinity is in focus. So, that leaves just 1 of 10 cases where it might be an issue. Granted, I'd expect tack-sharp focus with my 70-200 II, even with a teleconverter.


The other thing that seems apparent from the above numbers is that my 7D is pretty well centered - the direction of adjustment is split between front- and back-focusing.


Sorry for the longwinded explanation. I guess my point is that while I find AF microadjustment to be important personally, a lot of that is due to my meticulous nature. Sure, an adjustment of 1/4 of the DoF likely means a slightly lower percentage of OOF shots. But for many people, it's not going to be a huge deal, unless most shooting is done wide open with fast lenses.

Based on some shots I see here, when compared with some of my own carefully AFed shots, I wonder if mine is slightly out, though, and the problem is I don't think there is anything I can do about it without changing bodies.

If you do have a problem beyond the occasional, random, and unavoidable missed AF that occurs, you could send your set of lenses and your 40D to Canon. In the 'old days' that's what pros did. If you had a lens+body combo with an issue, and they adjusted the body to correct the issue, your other lenses would be affected. So you send the whole kit in at once. Probably not cheap, though.


The other thing to keep in mind is how you do your focusing. Focus-recompose is problematic with some lenses and at some distances. Likewise, many people don't realize that the actual AF point is larger than the little rectangle you see through the VF (except in the case of the Spot AF feature of the 7D). That means that if you're shooting a small subject, like a bird that's the size of your AF point in the VF, all those branches around it also are part of what the AF sensor sees, and if they're not all in the same plane it can 'confuse' the AF and lead to an OOF shot.


The true test is do you get these 'slightly out' shots some of the time, or nearly all of the time with a given lens. If the latter, AF microadjustment would likely help. One good example of that - fellow forum member Jan bought a Sigma 50mm f/1.4 and it needed a 9-unit adjustment. That means that it was more than a full DoF off, and every wide open shot would have been out of focus. Of course, if he only shot at f/4 it wouldn't have been an issue because the deeper DoF would have masked the problem.


Hope that helps...


--John

[mikehillman89]

Well I'm sure working with microscopes has gotten you in the habit of having dead-on focus neuro[]

[HiFiGuy1]

Looks like that LensAlign kit is a very useful tool for all the upper-end bodies after mine. Sure wish I could use it with my 40D. Come to think of it, it may be useful at least to know where I stand, even if I can't fix any AF issues. That might put my mind at ease.

[neuroanatomist]

Come to think of it, it may be useful at least to know where I stand, even if I can't fix any AF issues.

Well, short of spending $180 on a LensAlign Pro, you might be able to get some peace of mind (or generate angst!) with a printed chart. Here's [url="http://focustestchart.com/focus21.pdf]one example[/url]. It works best if you tape it to a door at the end of a long hall (depending on the lenses you're going to test), so you can easily put the chart at 45&deg; angle to the camera, and adjust the distance (camera on a tripod) to ~25x the focal length (i.e. ~8 feet per 100mm of lens focal length).


I tried the testing with my T1i and found that one of the three lenses I had at the time (theEF 100mm f/2.8<span style="color: red;"]LMacro IS USM) was front-focusing a small amount (not that big a deal, actually, since macro shots are mostly manually focused). AF microadjustment was one factor in my decision to upgrade to the 7D, but there were others.