Bad luck with Canon lenses?

[telyt]

It would be a great if Leica started to make lenses which fit on Canon mounth. And starting with the wide angles. Why they don't do it?

Leica has approached both Nikon and Canon with this idea and both C and N responded with a "not interested".  There's certainly the possibility of reverse-engineering but for the premium you pay for a Leica lens are you willing to take the risk that C for example will make some hidden tweak to the software that would render the Leica lens inoperative?

There's also the mirror-clearance issue.  Several of the Leica wide-angle lenses have serious issues with Canon full-frame mirrors.

Best of both worlds for me is the Leica lens on a 16-bit DSLR with a decent viewfinder.  I can focus a lens, and I can keep a 560mm lens steady; I don't need software to do that for me.  The color rendition I get from native 16-bit capture and processing is not easily replicated when you start with 12 or 14 bits.

[dwdallam]

I sent my 24-70L lens in for calibration after I looked at the same image taken with my 70-200IS L lens both wide open. The 24-70 was noticeably soft. They calibrated it free of charge, and they test the lens to make sure it needs to be calibrated--or you pay for their time. So mine did need calibration and it was less than a year old, I think around 8 months. It came back extremely sharp.

I got mine serviced in the USA with no problems, except it took them around three weeks to get my lens back. After that, I was tempted to send in all my lenses for calibration. However, my 16-35L and the 70-200 were very sharp wide open and stopped. SO I just used them and forgot about it.

[Ray]

Leica has approached both Nikon and Canon with this idea and both C and N responded with a "not interested".  [a href=\"index.php?act=findpost&pid=148727\"][{POST_SNAPBACK}][/a]

That's interesting. Are you sure about this? What's your source?

However, if that is true I can readily appreciate why both Canon and Nikon would reject the offer. They're in the business of producing lenses and most photographers will spend more on lenses than camera bodies. The last thing that either Nikon or Canon would want is someone revealing in an undeniable practical sense, with real comparisons, just how deficient their lenses sometimes are.

A streamlined form of MTF testing and categorisation of each and every lens would be good for both Canon and the customer. It would help Canon to improve its quality control and it would give them some valuable insights into market demand in relation to price/quality. For example, how popular are grade C lenses of a particular model compared to grade A at double the price?

[Ray]

I sent my 24-70L lens in for calibration after I looked at the same image taken with my 70-200IS L lens both wide open. The 24-70 was noticeably soft. They calibrated it free of charge, and they test the lens to make sure it needs to be calibrated--or you pay for their time. So mine did need calibration and it was less than a year old, I think around 8 months. It came back extremely sharp.
[a href=\"index.php?act=findpost&pid=148728\"][{POST_SNAPBACK}][/a]

Well, that's just amazing and also rather disturbing, the idea that Canon lens performance can be so significantly affected by calibration.

I wonder what's going on here. There's an implication that all lenses undergo a calibration prior to shipping and that some lenses are either not calibrated properly in the first instance or are knocked out of calibration during handling and shipping.

Anyone got the inside story here?

[telyt]

That's interesting. Are you sure about this? What's your source?

The late photojournalist Sal DiMarco, who had close contact for many decades with Leitz and later Leica.

[willow]

Ray

Hi. I thought I was the only member in Thailand!

Have you ever used Canon’s Service Center in Sathorn.
If so, what do you make of the standard of workmanship?

I need to repair my 17/85usm lens – for a second time.
The first was something to do with the aperture ring (?) – I was returned the broken part. 4 months (and not a lot of use) later the barrel jammed, now has completely given up.
Could this have been mis-aligned in the original repair?

Will / Should the lens be calibrated on its return as part of the service - it didn't seem so well the 1st time!


Willow

[Ray]

Sorry! Can't help you here. I'm not a permanent resident. Just visiting. Fortunately I've never needed to have any repairs done whilst in Thailand. In fact I've never needed any camera repairs at any time in my entire life, so far. Cameras (and lenses) in general seem to me to be surprisingly robust pieces of equipment.

I once dropped my Sigma 15-30 on moss-covered rocky ground in a rain forest, whilst attempting to change lenses, and watched it with dismay as it rolled and clattered down a slope for several metres. I assumed I'd have to write it off. But, no! It's still fine apart from a few scratches on the outer surface of the barrel. And a damned sight sharper on my 5D than the EF-S 10-22 on my 20D.

This is one reason I'm very curious about these significant improvements in resolution that some people seem to get after their Canon lens has been calibrated.

If it's just the autofocus system that's being calibrated, then I can understand it.

[Mark D Segal]

Ray,

The usual problem with the zoom lenses is allignment, not auto-focus. To improve the allignment apparently requires time and specialized people and equipment. From what I am reading here such resources exist in Canon USA, but froim what I'm told, not in Canon Canada [and perhaps not in Australia? - I have no idea.] The Canon people here told me they cannot improve on the alignment of my 17~40L eventhough it is not within the top 10% of total output; but because it is within their definition of an acceptable spec, they would not send it back to Japan, nor would they exchange it, because that is only for retailers to do.

I have been to Canon on Sathorn Rd in Bangkok. It looks like a substantial operation, but they didn't have to do anything for me that would test their technical abilities. (They were kind enough to quick-charge my battery at a time when it ran out of juice, I had no spare and forgot the charger in Toronto - wasn't that clever? never again!)

Mark

[Ray]

The usual problem with the zoom lenses is allignment, not auto-focus. To improve the allignment apparently requires time and specialized people and equipment. From what I am reading here such resources exist in Canon USA, but froim what I'm told, not in Canon Canada [and perhaps not in Australia? - I have no idea.] The Canon people here told me they cannot improve on the alignment of my 17~40L eventhough it is not within the top 10% of total output; but because it is within their definition of an acceptable spec, they would not send it back to Japan, nor would they exchange it, because that is only for retailers to do.
[a href=\"index.php?act=findpost&pid=148807\"][{POST_SNAPBACK}][/a]

Mark,
Interesting! I'll check on this when I get back to Australia. When I returned my Canon 400/5.6 prime to the store, which incidentally claims to be the largest photographic store in the southern hemisphere, it was sent to the Canon service department for a calibration. I don't know what sort of calibration. Perhaps it was just an autofocus tweak. When I tested the lens for a second time, I found a marginal improvement which brought the performance closer to that of my 100-400 IS zoom, but it still wasn't sharper than the zoom, so I got my refund.

I have a Canon 50/1.4 which is not as sharp as my el cheapo 50/1.8. I'll see if I can get that calibrated in Australia and what's involved.

[Rob C]

We might be getting ahead of ourselves here, in the rush to praise Leica: even within their own world, I understand that the glory belongs with the RF set of lenses, not the SLR ones, which would mean that there probably wouldn´t be much in it for either C or N bodies to migrate...

Trouble seems to lie in having to have a back focus longer than ideal, which isn´t required in RF cameras due to the happy lack of mirrors! Sensor construction might make this problem even more acute.

Rob C

[Conner999]

RobC

The higher prices (for Leica) and popularity due to the M8 belong to the M lenses. If you ever test a Leica (or Zeiss) lens against it's C or N counterpart you'd realize the praise was justified. I was once a sceptic .... then like my wife and the idea of eating sushi, I actually tried a piece.

Once the R10 is released at Photokina 2008, R lenses may share some of the limelight (and $$$$ sadly) once again.

[Rob C]

That´s part of the trouble at the Leica farm: they need to make all their profit at one time!

Rob C

[telyt]

We might be getting ahead of ourselves here, in the rush to praise Leica: even within their own world, I understand that the glory belongs with the RF set of lenses, not the SLR ones, which would mean that there probably wouldn´t be much in it for either C or N bodies to migrate...

The M lenses have the edge over R among the wide-angle lenses but for 50mm and up it's a toss-up ... and many people are finding the R wides adapted to a Canon body produce significantly better results than Canon's lenses do.  As for the 50mm and longer lenses there are no 'adequate' lenses: all of the current lenses are either best-in-class or essentialy equal to a well-calibrated cherry-picked N or C lens.  No woof-woof, zooms included.  A list of the really stellar R lenses would include everything from 50mm to 800mm, as well as the current wide zooms.

Trouble seems to lie in having to have a back focus longer than ideal, which isn´t required in RF cameras due to the happy lack of mirrors! Sensor construction might make this problem even more acute.

Actually the longer flange-to-sensor distance works in the R's favor.  The M8 incorporates several work-arounds to compensate for the short back-focus and consequent angle of incidence.

[Rob C]

But I understand that retro-focus designs, there to artificially give a greater space between rear element and film, is require ONLY because of the difficulty of the mirror box and the clearance required for it.

Yes, I also see that the less the angle angle of incidence to a sensor and its caves, the better for the results, but that´s just another form of problem, which I suppose can equate to the SLR problem just mentioned - either way, the RF (M) bodies (for film) allow for the best uncompromised design parameters, no fiddling about required; so, in that context, the M8 shares some of the problems of a reflex.

Ciao - Rob C

[telyt]

But I understand that retro-focus designs, there to artificially give a greater space between rear element and film, is require ONLY because of the difficulty of the mirror box and the clearance required for it.

Yes and no.  The M cameras with light meters require retrofocus wide-angle lenses in order to meter properly.  The primary disadvantage of retrofocus wide-angle lenses is weight and bulk.  Advantages include being able to tweak the design to reduce light fall-off toward the edges, and on digital bodies the angle of incidence is more favorable.

There are several excellent retrofocus designs among the R lenses including:

15mm f/2.8 (made by Schneider for Leica)
19mm Elmarit-R
28mm Elmarit-R

[Huib]

I keep on holding bad luck with Canon lenses. Especially with the 50mm F1.2. This lens was chosen as the best from 5 new lenses. But I discover later that one corner wasn’t sharp. I went to the CPS service and thought that I could wait on it. After 1 hour they told me that they will need much more time to fix it. It was to difficult to repair it immediately. They will send it to me when it is fixed.

[Slough]

If Sigma and Tamron can do it!
[a href=\"index.php?act=findpost&pid=148582\"][{POST_SNAPBACK}][/a]

They cannot do it, at least not without compatibility issues, Sigma lenses are well known for needing rechips when Canon release new cameras, I'm not sure about Tamron.

I suspect Leica will not release Canon mount lenses because they cannot licence the electronic coupling from Canon, and hence cannot ensure compatibility. This means that they would have to make manual aperture and manual focus lenses with a Canon mount. Which is not much different from using a Leica lens with an adaptor ring. So why bother.

And in any case, they might not want to encourage people to use their lenses on another makers cameras.

[Slough]

Nikon lenses do have sample variation. My 12-24 F4 AFS lens is useless between 12mm and 14mm as the corners are smeary at all apertures. Some are perfect though. Unfortunately I bought a used sample, and I suspect this was the reason the original owner sold it. I should have tested it thoroughly and returned it.

Rorslett reported problems with many samples of the 17-35mm F2.8 AF lens until he found a good one. Apparently Nikon claimed to have tightened up the production QC, but that might just be marketing bull.

Sample variation is a fact of life unless you pay big money. About 25 years ago, I heard that a semi-pro photographer (friend of the family) had a deal with a local camera shop. He would take home a bag full of new lenses, test them all, and keep the best one. He was trained as an optical technician, hence his ability to do the tests. So this is nothing new.

I think what has changed is that digital has made it much easier to pixel peep, and find faults. And of course zoom lenses are now dominant, and since they are more complex, both optically and mechanically, there is more likelihood of misalignment.