10D's 1st day

[Pekka]

Got my 10D on tuesday afternoon. I bought also the good old twin battery loader for I have need to load several batteries in shot notice.

I sold my D30 a while ago and D60 went out as trade-in for 10D so I can not provide image tests. Nevertheless they are not needed: 10D image quality is better.

As soon as I got out from the shop I started checking it out. Here are some findings in no particular order:

- More comfortable to hold than D30/60.

- Button layout is not how I would have done it. In order to change ISO you have to raise camera from shooting position, seek the button, change the value while looking the top display and then returning to shooting position. This is even more uncomfortable when shooting in vertical with grip as grip does not have the ISO change button. I hoped it was about time Canon put ISO display into viewfinder where it belongs, but apparently not. Of course it is possible to train to change ISO on shooting position.

- ISO change can not be assigned to thumb wheel.

- Back wheel is flimsy, D30/60 had more distinctive steps.

- Shutter noise is excellent (dark and quiet). Even animals don't seem to be bothered by it: http://photography-on-the.net/10D/CRW_0011_crop.jpg

- Shutter response and lag are very good. Now you can match release to very fast moments.

- LCD zoom works just great.

- I wonder why there has not been more raving about having a separate FEL button now.
(C.Fn-13-4)

- have not tested flash or ST-E2 yet.

- Tested "front-focusing" with 70-200 and there is none: it focuses spot on. I'm sure the front focusing issue is not "10D" issue but simply lens vs. camera calibration issue. I sold my D30 to a fellow who has a 70-200/2.8L too - but his lens with my old D30 backfocused heavily and he had to get the camera adjusted for the lens in Canon repair. In my use that D30 focused where it should focus. Odd stuff.

I have not tested all my lenses yet. Actually I have shot only with 70-200/2.8L and 35/2 yet.

- The focus issue (soft images) they rave about in dpreview is easy to understand: if you use AI servo or any other focus mode where you don't define and lock focus point it is easy to get the focus to some undesired position. This is not 10D's fault: the AI Servo works really well with fast moving subjects, it's just you have to be careful to check the focus in viewfinder at the moment you press shutter release. As soon as I got the camera out of package I assigned the C.Fn-04 to 1 which means you press * for focus lock (hold for constant AI, release to lock) and exposure lock happens when you press shutter half way. This is how I always shoot as it gives easy locking control for exposure and focus and lets you concentrate on composition.

- Canon conversion software is mediocre - I use now BreezeBrowser 2.6 for RAW conversions. As I wrote on dpreview I found a problem in 16 bit linear conversion which hopefully gets reported to Canon and fixed in dll update. I'm sad to notice that Canon has not learned anything from past and has not hired a new software developement team - other companies like Kodak excel in this area.

- 10D handles exposure dynamics much better than D60. I'm not sure if it is just more dynamic range: it seems and feels more that the "usable dynamic area" of CMOS is used in more intelligent manner - I believe this was achieved by slight histogram compression where shadows and highlights have more detail or "spread" and middle area is truncated a bit (the bit resolution is so big that there theoretical color shade gradation steps produced by this truncation are not visible). Note that this is just a gut theory based on experience after doing Linearsharpen histogram optimization for a long time.

- Image sharpness is very good. 10D has enough detail to produce beautiful and smooth DoF representation and it delivers very "analog" feel to photos - on D30 the resolution made DoF edge too "violent" and on D60 the lines felt "rougher". Perhaps this is so because 10D has less noise and less aggressive gamma curve and also better interpolation algorithm which produces smoother jagged lines. Acutance is better. With BreezeBrowser normal sharpness output you need only a .2/500/0 USM once or twice to bring plenty of crisp detail out (if the lens has delivered such detail). If you zoom deep down to the (sharpened) images you see the D30 "heritage" there - but in a much improved form. With proper techniques you can dig out a lot more fine detail from 10D than from D60. See a very strongly plain (overall) sharpened example (BB conversion as no sharpening + .3/150/0, .2/300/0 and .2/500/0 USM) which holds great fine detail, and note that this is ISO 400 photo: http://photography-on-the.net/10D/CRW_0087.jpg (full size)

- Exposure metering is slightly different from D60. Generally it is more accurate and in some situations where it feels there is a backlit scene it may tend to overexpose - like bright sky over brick building. This is in my opinion good way to do it (you have to decide when designing the camera the exact way how it reacts to specific scene templates - this is not fuzzy logic). Whatever changes there are you can learn there as you learned D30 and 60.

- Noise levels are low and smooth - ISO 800 is very usable and the light noise it has is easily removed. ISO 1600 needs relatively bright exposure to be really usable in noise levels but in any way it will save the day many times to come I'm sure. ISO 3200 is a noisy curiosity which will be needed only in extreme situations where movement _has_ to be stopped.

- Viewfinder coverage is slightly bigger than in D60, at least in my copy of 10D (I've read there are slight variations there) - and best of all it is perfectly centered (at last). I found it very difficult to shoot straight with both D30 and D60, but 10D is somehow very easy in that aspect - you could speculate if the sensor in D60 was not perfectly in place and 10D has better design structure there. :) I talked about D60 viewfinder slant with Canon repair and they had a simple "shoot a target 3m away to check it" type of procedure to verify it and he said there is an adjustment for straightening it. Well, I don't have that D60 any more so that's that.

White balance is much improved. Great! I just wish Kelvin that the auto WB has fixed is displayed in conversion software (BB). This would make fine tuning faster.

10D can take a lot of saturation (+5 twice in LS):

http://photography-on-the.net/10D/CRW_0038.jpg

Indoor auto white balance works great in very difficult lighting - notice also non-burned-out forehead - new gamma curve of 10D works great in scenes line this. (Conductor Sakari Oramo cheerful):

http://photography-on-the.net/10D/CRW_0053.jpg

At last spring is here:

http://photography-on-the.net/10D/CRW_0057.jpg

ISO 800 lacks red spotty skin tones on shadows, the main concern in D60 photos. This is also noise reduced with LSD60's ISO400 noiseblur, which works great on 10D ISO 800 (btw: this is auto WB which could be easily finetuned to desired tone by kelvin WB):

http://photography-on-the.net/10D/CRW_0067.jpg

Extreme backlight is handled well - the house wall is not black and holds still all detail if you adjust curves to reveal it.

http://photography-on-the.net/10D/CRW_0080.jpg

To my surprise AI focus was not fooled when the runner ran between bushes:

http://photography-on-the.net/10D/CRW_0085.jpg

More later....

Pekka

[julivalley]

Do you use CF4/1 just with AI Servo or all of the time?
I tried it tonight with still subjects and it seemed to work great. Thanks for the tip.

[koalafarm]

You can set the ISO to the thumb dial.
It is CF 13, option 3 I believe.
The only problem is I can't figure out what that does to the assist button.

[mishkin]

No, ISO can only be set using ISO button and main dial.

[koalafarm]

you are right.
sorry.
I use that for the focus point selection.
Got it confused for a second there.

[hugodrax]

Flimsy wheel like it will fall off after a few weeks of use?

[Mickey]

Thanks for the review, Pekka! I just got my 10D last week (moved up from a G2) and I too am VERY impressed. I'll have to try that custom function for the focus lock button...sounds like a good idea and an easy way to switch from AI Servo to One Shot.

As I write this post, my 10D is in the process of taking some 10 minute star trail shots. Couldn't do that with my G2. ;)

Are you planning on releasing a LinearSharpen for the 10D? If so, I look forward to testing it out!

I did notice a tendency to overexpose in certain situations like you said. You're probably right in that it's the new exposure analyzation algorithms that came along with the DIGIC processor. It's just something I'll have to get used to, and remember to dial in extra negative EC if I want to prevent blowing the highlights.

Have fun (I know I will),

Michael

[Mickey]

I forgot to add that I was also very impressed with the AI Servo mode. They didn't put the "AI" in there for nothin'! I tried tracking fast moving cars, bicyclists and kayakers and the thing that really surprised me is how it continued to lock focus on the subjects even when bushes, people, telephone polls, etc. would move between the camera and subject. I can't wait to try shooting some sports and/or motorsports with this baby! :)

[Mickey]

I tried that custom function setting and I like how it separates the AE lock and AF lock, esp. since I usually do the AF lock and recompose thing. I think I misinterpreted you about how it works though. It sounded like you were saying that while you're pressing the AE lock button the camera focuses using AI Servo and if you let go of the AE lock button it performs a One Shot focus lock. This isn't right, is it?

[Pekka]

Mickey wrote:
I tried that custom function setting and I like how it separates the AE lock and AF lock, esp. since I usually do the AF lock and recompose thing. I think I misinterpreted you about how it works though. It sounded like you were saying that while you're pressing the AE lock button the camera focuses using AI Servo and if you let go of the AE lock button it performs a One Shot focus lock. This isn't right, is it?

C.Fn-04-1

When you're in "one shot" focus mode:
- press * to focus
- when focus lock is achieved you can stop pressing * and the camera keeps that focus until you press * next time

When you're in "AI Servo" focus mode:
- keep * pressed down to keep AI Servo working
- release * to stop AI Servo (i.e. lock focus).

[spark]

Agreed the 10D is impressive...

[Mickey]

Pekka wrote:


C.Fn-04-1

When you're in "one shot" focus mode:
- press * to focus
- when focus lock is achieved you can stop pressing * and the camera keeps that focus until you press * next time

When you're in "AI Servo" focus mode:
- keep * pressed down to keep AI Servo working
- release * to stop AI Servo (i.e. lock focus).



Ah, ok, I see what you're saying. It would be cool if it actually did a real focus lock once you let go of the * button though. I find if I keep holding the button in AI Servo, it tends to want to keep adjusting back and forth even though the camera and subject are stationary, which says to me that it wouldn't be as accurate as using One Shot. What do you think? Or does it in fact make a final adjustment to the focus after pressing the shutter to get a solid focus lock? Maybe I'll do some multi shot tests using both methods and see which one is more consistent.

[Dale]

Pekka,

When you use Al Servo do you use center AF point?

[bartvo]

Hi Pekka,

Great review, thx for sharing.

Happy shooting!

Greetings, Bart

http://www.bartvo.com

[Pekka]

Dale wrote:
Pekka,

When you use Al Servo do you use center AF point?

I use center point most of the time, if there is a need for other points i'll use them, e.g. when recomposing with very shallow DoF (or close with 14mm), or tracking moving objects and you want to place that off center.

[Griffin]

OMG! I cannot afford one fttb so be patience...patience is a virtue. ;)

About exposure, I know a local birder who has been using film (and EOS-1V!) lately jumped digital with 10D. He finds it is consistently over by 1/3 to 1/2 stop. I guess digital cameras need some re-learning.

Dear Pekka, by the time I got mine, the beloved new 10D profile would be ready, right? :p

Enjoy and keep us posted.


Griffin.

[achic]

I am not so enthusiastic about the 10D yet. I am not a hobbyist and have shot thousands with the D60 in the last year. The files out of the D60 never failed to impress me right out of the camera and they sharpened up nicely with very little USM. I have not yet got a file out of the 10d that has that wow factor or is as sharp.
Also, the colors seem cartoonish at times. I have tried USM, but the 10D files seem to halo much quicker than the D60 ones. So far, I feel the D60 colors are more real and more satisfying.

[Rudi]

Griffin wrote:
About exposure, I know a local birder who has been using film (and EOS-1V!) lately jumped digital with 10D. He finds it is consistently over by 1/3 to 1/2 stop. I guess digital cameras need some re-learning.

I don't know... how does he know that it's the digital that is "out", rather than the 1V, or the film stock he was using? ???

Every new camera I have ever bought, I always test the exposure meter - it will always differ from the previous model! As long as it's a consistent difference, what is the problem?

[Griffin]

Well Rudi, I see no problem at all. When I want to use fill-flash with my D30, I know I should turn down 1 stop in evaluative mode. I guess it is just some kind of "adaptation".

As a matter of fact, the individual (which I adire very much of his photography) uses incident light meter. :)


Griffin.

[Rudi]

Griffin wrote:
As a matter of fact, the individual (which I adire very much of his photography) uses incident light meter. :)


Well, great minds think alike! :D

I *love* my Sekonic L-358 meter, and I can see how useful it would be for bird photography! As long as the bird is in the same light as you, an incident reading will avoid all the traps of the background or a white or black bird throwing your exposure meter off... Of course, you can compensate for these things, but when there is an easier way, why not use it? :)

As you mentioned, working *around* your camera is one of the most important things that each photographer should learn - no two cameras will be exactly the same, and one of the skills is to recognise your camera's "personality", and use it to your advantage! Of course, if your camera consistently under/overexposes by several stops, or if it's NOT exposing consistently, it's time to get it fixed! :eyes

[pigasus]

Hi Pekka,

I know your 10D seems to have been excellent right out of the box. But, out of curiosity, have you tried the reset that seems to be needed on some not such good examples? I know that some people are ridiculing this fix, but I've seen too many reports (and examples) of success, from people whose opinions I respect, not to believe there's something to this. As to why not every camera seems to have a problem, current theorising would lead me to believe either:

1. Only some cameras have been set up improperly from the factory. Thus others are OK. Or..

2. All cameras have been set up improperly, but the default (ie unadjusted) settings they arrive with are closer to the body's proper adjusted settings, thus giving good results.

In any event, I'm curious whether you've been tempted to try the resets to see what happens.

Cheers,

[Pekka]

pigasus wrote:
Hi Pekka,

I know your 10D seems to have been excellent right out of the box. But, out of curiosity, have you tried the reset that seems to be needed on some not such good examples? I know that some people are ridiculing this fix, but I've seen too many reports (and examples) of success, from people whose opinions I respect, not to believe there's something to this. As to why not every camera seems to have a problem, current theorising would lead me to believe either:

1. Only some cameras have been set up improperly from the factory. Thus others are OK. Or..

2. All cameras have been set up improperly, but the default (ie unadjusted) settings they arrive with are closer to the body's proper adjusted settings, thus giving good results.

In any event, I'm curious whether you've been tempted to try the resets to see what happens.

I doubt that 10D's are "adjusted" at all in factory. There would be no point spending time and money to adjust each 10D to certain lenses when you don't even know which lenses will be used by the client. More likely an adjustment chart for ALL Canon lenses (if there is such a chart) is predefined and then put identically into each 10D's ROM (as a chip). Same way as computer shops build computers: they don't install Windows manually on each machine, the copy a prefedined drive image instead (takes 1 minute instead of 30).

So what goes on here? I think firmware bug is most likely reason.

I think the only explanation that when you first boot up 10D it reads data from ROM to RAM - in some occasions some internal values are messed up. To speed up bootup time that data stays in RAM and is powered by internal battery. Now when you reset custom functions the camera reads the ROM again and replaces old values with preset ones.

What baffles me is that why should resetting custom functions affect internal lens or focus calibrations done with Canon maintenance software? I mean, if you know there is a possibility that a lot of time is spent in repair to calibrate 10D to certain lenses would you to add a menu command that would reset all internal calibrations just like that?

I will make a before/after test soon - I just hope it does not make my 10D worse :)

[pigasus]

Pekka wrote:
I doubt that 10D's are "adjusted" at all in factory. There would be no point spending time and money to adjust each 10D to certain lenses when you don't even know which lenses will be used by the client.
...
What baffles me is that why should resetting custom functions affect internal lens or focus calibrations done with Canon maintenance software? I mean, if you know there is a possibility that a lot of time is spent in repair to calibrate 10D to certain lenses would you to add a menu command that would reset all internal calibrations just like that?

Good question! But from what I've been reading I get the impression that this fix isn't specifically about lenses and focus calibration. This is about 'softness' of image. Someone has said that it's kinda like the pixels are misaligned. There is an interesting thread on DPR speculating as to what's happening technically. http://forums.dpreview.com/forums/read.asp?forum=1019&message=4752557

I will make a before/after test soon - I just hope it does not make my 10D worse :)

So far I haven't heard anyone complain that things got worse - even when they tried the reset on a 10D that seemed fine to start with. Let us know what happens, please. :)

[mishkin]

pigasus wrote:
Good question! But from what I've been reading I get the impression that this fix isn't specifically about lenses and focus calibration. This is about 'softness' of image. Someone has said that it's kinda like the pixels are misaligned.


This can easily be verified if someone has RAWs before and after. Your point is that RAWs are the same, while JPEGs are not.

I have a tool to look at RAW directly, so let me know.