40D v MkIII

[Cashie]

I've just found this review, so if anyone is thinking of buying a MkIII over a 40D, have a look here I found it very interesting.


http://iwishicouldfly.com/iwishicouldfly/journal/html/093007.html

[Birdingben]

Thanks for this link,Cashie.

Reassuring as I've just (yesterday) bought a 40D & today looks like a good day (weather-wise) to get it out of the box!

[Roy C]

I read this over at WPF, very interesting read.

[Keith Reeder]

I could have written the same review (without the MK III references, of course) because he sounds as pleased with his 40D as I am with mine!

;)

[Words]

Interesting read. I plan to get a 40D fairly soon.

[Lisa W]

Wish you had posted this about 3 weeks ago! My Mark III came in the mail this week, but still think I will be extremely happy with it - once I've learned to use all of the features, in about 10 years.

[Keith Reeder]

Oh, I'd still have a Mk III if I could, Lisa - even better noise performance, and it would work with my 100-400mm and a 1.4x.

Even so, the 40D is a gem, and I'm happy as a clam with mine.

[Cashie]

Wish you had posted this about 3 weeks ago! My Mark III came in the mail this week, but still think I will be extremely happy with it - once I've learned to use all of the features, in about 10 years.


Dont wory Lisa, they have ironed out any problems with the MkIII now, with the latest batches & firmware, you will be over the moon with your purchase.

Anyway I will be getting one soon myself to go with my 500 lens. ( cant wait ) :-O

[Lisa W]

Keith and Paul, thanks that makes me feel somewhat better.
Keith, I put the 1.4 on with the 100-400 lens today, hopefully I'll get some shots with it soon.
Paul, if you ever tire of that 500 lens you can donate it to a veteran and school teacher here in the states if you would like :-O

[Keith Reeder]

Well if ever you want a straight swap Lisa, just gimme a shout - I'll pay for the shipping both ways!

;)

[Cashie]

Ok Lisa, I will keep that in mind. 8-P

[Lisa W]

Thanks, Paul - had to say that before I reply to Keith.

Keith, let me think about that. (.1 second pause) uhhhhh no thanks ;)

[Malcolm Stewart]

Cashie, thanks for the reference. If only life was that simple and all things could be given a number.

I own 3 of the bodies on his list, 5D, 30D and my latest toy, 1D MkIII. I bought the last two specifically for birding. For birding lenses I've used EF 300 f4L, 300 f2.8L, 300 f2.8L IS and EF 70-200 f4L. With the primes I've frequently used a 2x MkII Canon extender and a 1.4x MkI extender. I've had good results with all combinations of lens, extender and body. I've also had plenty of inexplicably off results which has taken much time (and money) to sort out.

My feelings now (may change with more experience) are that if I can get close enough to the bird, my 5D is going to get the best results, but for birds in flight my 1D MkIII has given a real beginner in this field some memorable images. On the other hand, I did waste about a year trying to find whether my 30D or my old 300 f2.8L were to blame for disappointing results. Getting Lehmanns to adjust the 30D and repair the AF on the old 300 f2.8L was one of my better decisions! (If my 300 f2.8L had been off (front-focusing) from day one, I wouldn't have wasted so much time checking on my technique, but it crept up on me.)

[Tannin]

The "review" linked to above tests resolution under unrealistic circumstances - few competent bird photographers in real-world situations limit themselves to 200 ISO when the 1D III is more than capable of delivering quality, low-noise images at much higher sensitivities. He blandly claims that the 400D and the 40D "will out-resolve the Mark III in a focal-length-limited scenario" and passes on, as if that matter is settled.

His claim is arguable in perfect light, though by no means proven, but is quite clearly not the case under more typical real-world conditions.

When I first got a 1D III I naturally did some test shots of my own, comparing it to the best-performed other camera I had at that time (a 20D). It was a no-brainer to use an ISO that was representative of typical real-world bird photography - on a 20D, that is ISO 400, which is equivalent to ISO 500 on a 400D, 40D, or 1D III. (I also had a 400D at the time, but didn't compare it directly to the 1D III as I had previously confirmed that the 20D was equal or better.) The result was plain: at 400 ISO, the 1D III produced a sharper, higher-quality image.

I haven't compared the 1D III directly to my new 40Ds. I probably should, as the 40D is a very impressive little camera. It would be interesting to see how the identical-circumstances comparison turned out, but really only as a matter of curiosity. Certainly there won't be much change to observe in it, as the 20D and 40D are very, very close in their performance. Can a 40D match a 1D III under representative focal-length-limited conditions? Some fairly extensive real-world use of both says "no" - the 1D III consistently delivers better results - but maybe I should set aside a little time to play with them in the backyard instead of in the field.

[citrinella]

- on a 20D, that is ISO 400, which is equivalent to ISO 500 on a 400D, 40D, or 1D III.
Just out of curiosity, why ? What do you mean ISO400 on a 20d is equivalent to ISO 500 on those newer models ?

TIA, Mike.

[rentoncharman]

I have both 40D and the MkIII. In my experience the 40D autofocus is very consistent compared with the MkIII. I've compared shots in the field and also in a test situation using an angled chart. With my MkIII I had to use microfocus adjustment to reduce front focussing with all my lenses, even on my 500mm which had been back to Canon to be "matched" with the MkIII. The 40D always found the focus point correctly on my test rig with the same hyperfocal point, ie dof 1/3 in front and 2/3 behind the focus target. The MkIII although adjusted for front focussing was less consistent but accepatable unless the lens was wide open.

Noise is a different problem. MkIII images are much better and I think they show less noise at 1600 iso compared with the 40D at 400 iso. This limits the suitability of the 40D in poor light situations.

Hope this is useful.

[Tannin]

Hi Mike,

ISO is a measure of the sensitivity of the recording medium (film or digital sensor) to light. When we dial in (say) 400 ISO on our cameras, we tend to assume that this is the end of the story.

Let's say you use two different cameras to take the same picture under carefully controlled conditions. You meter the scene and then manually set the first camera to ISO 400, f/8, 1/500th and the result is correctly exposed.

If you now set the second camera to ISO 400, f/8, and 1/500th, you will get exactly the same exposure, yes?

No!

At least not always. Different cameras are differently sensitive to light. Let's say the first camera was a 40D and the second camera was a 20D. The second shot will be over-exposed by about a third of a stop.

Why? Because the 20D's actual base ISO isn't exactly 100, it's about 125. Canon labeled it "100" because that was the nearest "standard" ISO setting to the actual sensitivity of the sensor. Similarly, the setting labeled "ISO 200" on a 20D is actually about ISO 250, and the "400 ISO" setting is actually about ISO 500.

In contrast, the 40D's base sensitivity really is 100 ISO (or close to that - it might be 97.6 or 103.31, I'm not sure, but in any case it is near enough to exactly 100 as makes no difference).

The upshot is that if you dial in "ISO 400" on both a 20D and a 40D, and set both for f/5.6 in aperture priority mode, both will expose the scene correctly - but where the 20D might use 1/1000th of a second to get the right exposure, the 40D will need 1/800th of a second.

Or, putting the same thing another way, to correctly expose that scene at f/5.6 and 1/1000th of a second, you need a true ISO 500, which is indeed ISO 500 on the 40D dial, but is labeled "ISO 400" on the 20D.

This is not a new thing, there have been a number of cameras that have shipped with "incorrect" ISO calibration in the past (I forget which ones now, but I think there have been quite a few). Some were pessimistic (like the 20D) and give you a bit extra, others were optimistic and gave you a bit less than they claimed to.

A little while ago Canon decided to make a point of calibrating new camera models correctly, so that 400 ISO on the dial really is 400 ISO. I think the other majors have done the same thing.

Anyway, the practical import of this is that if you want to compare two cameras, you need to make sure that the actual ISO settings are the same. So when you compare a 20D with a 40D (or a 20D with a 1D III) at 400 ISO you should set the dial on the 20D to "320 ISO" - except that the 20D ISO adjustment only goes in whole stops (100, 200, 400, 800, etc.) so that isn't possible. But you can set the 40D (or 1D III) to 500 ISO and that will give you a fair comparison.

(By the way, I wasn't fully aware of this when I made my test shots a while back, and I compared the 20D to the 1D III at a nominal 400 ISO - in other words, I gave the 1D III a 1/3rd of a stop unfair advantage. But that does not invalidate my original point - although I haven't bothered to repeat the test, the difference was clearly great enough to swamp my error with a little left over.)

Perhaps another way of looking at these issues is to consider the "standard" ISO I have learned to dial into different cameras under "standard" conditions - i.e., birding in good but not outstanding light, using either the 100-400 or the 500/4.

On the 20D, I use 400 ISO (actually 500). Lower makes the shutter speed to slow, higher is too noisy.

On the 40D I use either 400 or 500 ISO, going to 640 if I have to. 400 is best, but that leaves me a little tight on the shutter speed side.

On the 1D III I use 640 ISO as my starting point. That gives me slightly cleaner images than the 20D can do at 400(500), or about the same as the 40D at 400.

Obviously, if the light is better than good, I'll drop the ISO down a bit; typically to 400 on the 1D III (gorgous images!), 400 or sometime 320 on the 40D, and ... er ... 400 on the 20D because (failing fantastic light) 200 is too low and there isn't any 300!

And if the light is bad, the 1D III is just amazingly good. Here is a 100% crop from a picture I took the other night in very dim indoor lighting: 1D III, 50/1.8 at f/2.5, 1/200th, and 6400 ISO. Look at that feather detail! (Err .... I mean "facial hair".)

[gymell]

I have both the 40D and the Mark III. I usually use the Mark III, but have the 40D as a 2nd body/backup. Even though I did have the Err99 problem with the Mark III, I have been very happy with it. Put a 500 f/4 on it, and that's an unbeatable bird photography setup. Add the 40D as a backup and you can't go wrong!

[gmax]

Well Tannin, I must admit that when I first glimpsed your comment, I wondered what this long narrative was about and if it ended up by turning my beard into something similar to the shot you uploaded ... o:D

Just kidding, it's a heap of useful advice and info many of us may not know (at least I didn't)!
One off-topic question: I usually find myself underexposing by 1/3 of a stop with my 20D, when compared to, say, my 350D: is this regular in your experience?
Cheers,
Max

[Tannin]

Max, I'm not quite sure what you mean there. There would seem to be two possibilities:

1: take the same shot under the same lighting conditions using automatic exposure on both cameras. (For example, f/8, ISO 200, aperture priority.) One camera uses 1/500th, the other uses 1/450th. Both shots look correctly exposed.

2: Take a number of different shots with both cameras, using automatic exposure. Your eye says that the shots from one camera consistently look a little over-exposed, so you start making a habit of dialing in 1/3rd of a stop negative exposure compensation on that camera as routine. Once you start doing this, the results look good.

Possibility #1 is what I'm talking about in my post above: one camera is more sensitive than the other but the ISO calibration doesn't take this into account. Both expose correctly, but they will use different shutter speed/aperture combinations for the same scene at the "same" ISO because the actual ISO is not the same. We tend not to notice this one: the shots end up correctly exposed either way, so you don't spot the difference unless you are looking for it.

Possibility #2 is quite different. Here we are talking about calibration of the exposure system, and variation in true ISO sensitivity is not relevant. "Correct exposure" is, of course, a myth. There is no such thing! Where the shot is way over or way under (let's say two or three stops out) most of us will agree that it is wrong. But once we are somewhere close to "correct", it all becomes a matter of opinion.

Often you will see a shot posted somewhere for comment, and one poster will say "nice shot but a bit over-exposed" but a couple of posts later someone else says "pity it is underexposed"! A friend of mine, using pretty much the same equipment as me, dials in 1/3 -ive EC on most of his shots. I prefer to shoot about flat or with 1/3 +ive EC! To me, most of his shots look a little too dark and muddy. He says my shots look a little colourless and over-bright! This isn't equipment, it's just disagreement over what constitutes "correct" exposure.

Given that there is no such thing as "correct" exposure, it should be no surprise to discover that different cameras don't always agree with each other either. My old 400D, as an example, fairly consistently produced results that were, to my eye, a little too bright, so I wound up using 1/3rd -ive EC on it a lot of the time. Was this just individual sample variation (i.e., was my particular 400D just miscalibrated a little) or is it variation between different models? Individual variation would be my guess - they are both Canons, after all, and I've never noticed any consistent difference between my two 20Ds, two 40Ds, and the 1D III; it was just the 400D.

[Adey Baker]

It's option 2, I think, Tannin. I do the same with my 20D as Max and I used to dial-in -0.5 on my old EOS50E 35mm camera as well. Canon's evaluative metering isn't as spot-on as they'd like to think! If there's some sky in the shot, then it's fine, but with the kind of subjects that we're dealing with here on BF, then it tends to over-expose unless there's a good area of light tone somewhere else in the shot to compensate for the lack of sky. It's a simple matter to flick the compensation dial round a bit when you've got a bird against the sky but I always re-set to -0.3 as a 'default' setting.

I'm hoping to add a 40D to my set-up soon, so it'll be interesting to see how that performs.

[Tannin]

Cheers Adey. I don't think I can really comment on evaluative metering as .... no, wait .. why not start a new thread for the thought? We are wandering further and further off topic, and this is interesting, so I'll do that.

[gmax]

It's option 2, I think, Tannin.
I'm hoping to add a 40D to my set-up soon, so it'll be interesting to see how that performs.

Option # 2 indeed ... I apologize for not having been clear enough ;) ... I agree it may be just a matter of personal taste or long time habits, but I usually find myself dialling in a neg comp ... it's also an added bonus, providing some more "speed" to my shots, since I'm constantly working with my f/5.6 lens + 1.4x TC = f/8 ...
I agree also on the topic led astray, so we may follow up in another thread :t:
Cheers,
Max

[K-Lex]

I concur with the review. The 40D is a totally fabulous camera and should in no way be seen as the inferior 1Dmk3. I have no plans to trade mine in anytime soon.

[Tannin]

The 40D is indeed an excellent camera. But have you used a 1D III?

[Roy C]

I concur with the review. The 40D is a totally fabulous camera and should in no way be seen as the inferior 1Dmk3. I have no plans to trade mine in anytime soon.
I too have a 40D and am very happy with it but having never used a 1Dmk3 I could not make a statement like this - I very strongly suspect that the mk3 is probably a superior camera. I am a great believer in the old adage - 'you get what you pay for'.

[gymell]

I have both. The 1D III is definitely a superior camera. The 40D is a great camera in its own right, but the 1 series is a noticeable difference. Anyone who has used both will immediately know it.

[K-Lex]

I agree, the 1Dmk3 is obviously the better camera and if I owned one I'd pick it up every time (after the 1DsMk3 obviously!!). What I do say though is the 40D is a good camera in it's own right and can more than hold it's own in a fight. I have used pretty much every camera going in my years as a pro photographer and the 40D genuinely suprised me when I used it. The reasoning is simple, the 40D, you just don't expect it to be as good as it is. I'd still have a 1D3 given the choice though!!!

[gymell]

Agreed about the 40D. A pro told me once that the 40D was the first non-1 series camera that he'd used where he felt he wasn't giving up too much from the 1D. It's a very good camera and that's why I have one as a 2nd body/backup to my Mk III.

[K-Lex]

I had a 1D mk3 to play with on a recent trip to Tuscany (along with a 1DsMk3 and a Mamiya ZD - I felt a lucky guy that trip!!! My shoulders hurt like hell but hey ho!!!)
Just for the record, ZD isn't a great camera (for anything - let alone wildlife!) - buy a Mamiya 645 and stick a digital back on it instead.
Anyway, the difference I noticed was with the 1D3, when the files were put into Lightroom and PS, just now more, what's the word, 'vibrant' and 'alive' the pictures looked from the 40D comparison. They needed extensive work in PS because the sensor of my test camera was filthy but if it had been clean, there was almost no post processing work required. That's the difference - the 40D image needs work, the 1D3 doesn't.

[Tannin]

What I do say though is the 40D is a good camera in it's own right and can more than hold its own in a fight.

Indeed the 40D is an excellent unit whichever way you look at it. Hell - I bought one a while back, and was happy enough with it to buy a second one afterwards. My usual rig these days is 1D III and 500/4 as the primary birding rig; backed up by a 40D and 100-400 for various odds and ends, notably larger wildlife like kangaroos, the odd telephoto landscape, and small birds up too close to focus on with the 500; then the other 40D with a 24-105; and my old faithful 20D with an ultra-wide (either a Canon 10-22 or a Tokina 10-17 fish).

While my 1D III was off having the sub-mirror replaced, I fell back on a 40D and found it quite slow, awkward, and inferior .... for about 10 minutes. After that, I got used to it (just like my old 20D & 500/4 rig really, only faster and with better controls) and used the 40D quite happily for a couple of weekends.

Just the same, bolting the 1D III back onto the big white lens again was ... well, like sleeping in your own bed after a couple of weeks in a motel, or like driving a proper car again after a hundred kilometres in a delivery van. I have never yet seen anyone describe the difference between using a 40D (or similar) and a 1 Series properly. There are all sorts of technical differences, the odd one major, most of them really quite trivial in the overall scheme of things, and you can list features and compare them all day long, but that won't go close to explaining it. In the end, I have to fall back on analogy: both do the job perfectly well, but the difference is like wearing silk vs wearing nylon.

Anyway, regardless of my inability to describe the differences properly, the 1D III is better.

But this raises an interesting question: how much better is the 1D III?

Current $AU price for the 40D (body only) is $1350. A 1D III is $5800. That means, for the same money, you could have:

A 1D III (body only)
Four 40Ds and $400 left over. Yes, four of them.
A 40D with a Canon 100-400L, Tamron 17-50/2.8, Canon 28mm f/2.8 prime, Canon 60mm f/2.8 macro, Sigma 10-20 ultra-wide, and $40 left over.


Is the 1D III, then, a four-and-a-bit times better camera than the 40D? No way! Not even close. Is it twice as good? Nope. Probably not even one and a half times as good. Maybe, trying to put a number on the very real but quite intangible, something like 30-odd percent better. And for that last realatively small improvement on what is, after all, already a very fine camera, you pay more than four times the price. Looked at rationally, for most people, a 1D III is really a pretty senseless extravagance, to be viewed in the same light as you view Mercedes-Benz cars and tailor-made designer suits. Well, I don't want a Mercedes-Benz, and I wouldn't wear the suit even if you gave it to me .... but if I didn't have one I'd buy another 1D III without even blinking.

[Cashie]

Tannin,
You say your usual rig is a MkIII & 500 f4 and backed up with a 40D with a 100-400, being as the crop factor on the MkIII is 1.3x & 1.6x on the 40D, making it 650mm & 640mm equivalent, is there a lot of difference in IQ at the long end of each lens ??

[Tannin]

Actually, the figures are a bit different, Cashie. The crop factor on a 1D III is 1.28, which makes them 641mm vs 640mm in theory, but that's not right either, as the 100-400 is actually about 107-382 or something like that. In other words, the 1D III and 500/4 puts you noticeably "closer" than the 40D and 100-400 @400 does.

Secondly, I have a 1.4 converter on the 500 about half the time.

No matter, let's compare the 1D III & 500 vs the 40D & 100-400 anyway. The short answer is that the big iron has more effective reach, by quite a margin. You can crop the 1D III image harder and still get an equal quality picture, or crop them the same and get a better picture. It's a double whammy: you get the better "per-pixel" resolution of the 1D III, and the quality of the big prime lens.

Put both cameras on the same lens and it gets harder to say. I haven't done enough formal testing to comment in any detail beyond what I wrote 15 posts or so above, but I have done a moderate amount of field testing. One of the most interesting questions I asked myself was: "which is better when really pushed for length: 1D III & 500 & 2X, or 40D & 500 & 1.4X, or maybe 1D III & 500 & 1.4X?". The short answer is that there isn't a whole lot in it, but I'd lean towards the 1D III & 2X converter. The differences are small enough that I'd need to do proper formal testing to nail them down for sure. Informal field tests tell me to use whichever of those three combinations I have handy right now before the damn bird goes somewhere else!

Sometimes I swap the cameras around. The other day, for example, I was playing games with some Eurasian Coot (I haven't photographed Coot since digiscoping days and have been making a point of revisiting the old "easy" birds this year, looking to improve on my ancient Coolpix 4500 shots) in the beautiful late afternoon sunlight and because they seemed very approachable I put the 100-400 on the 1D III, but kept the 40D & 500/4 handy in case something else came along at a more conventional distance. That was a lot of fun! Before too long I was at 100mm on the 100-400 and still couldn't get the whole bird in frame. I tend to think (quite incorrectly) of the 100-400 as a slow-focusing, not-so-sharp lens, but neither of those are true. But the f/5.6 at the long end bugs me. Ohh how I wish that Nikkor would make their 200-400 f/4 VR in a Canon mount! I've even thought about buying one anyway and getting a Nikon camera to go with it, but the D300 is just an overpriced 40D and the D3 must be pretty marginal for reach. I like the bigger pixel elements of the 1D III (as compared to the 40D or 20D) but there are practical limits to everything. I rather suspect that the ultimate good-light birding camera today is the 1Ds III - same pixel density as the wonderful old 20D, but full frame with a 1D III body, controls and autofocus. On the other hand, when the light is less than perfect (which is most of the time) the superb high ISO of the 1D III is very hard to beat.

Errr ... what was the question?

[Cashie]

Thanks for that Tannin,

Very useful info as I am just about to buy a MkIII.

As for the rest of the kit I have it all 40D, 100-400, 500 F4, & 1.4X extender, all I need now is the new MkIII body.
I will let you know when I'm about to post some shots with it in the near future.

Thanks for your help

Cheers B (:

[gymell]

Sounds like a killer setup - I have the same! ;)