Red edges and new firmware......

[ho_co]

Advertisement (gone after registration)

Sandy, thanks for the continued careful consideration you're giving the issue.

 

Last October I jokingly suggested (http://www.l-camera-forum.com/leica-forum/leica-m9-forum/102760-18mm-m9.html#post1083786) that the problem might be some kind of misalignment of the microlenses. At the time, it seemed to me the only possible explanation, with unfortunately difficult correction possibilities. The serious consequences of that are the reason I phrased it humorously.

 

 

 

Assuming that the problem is indeed a microlens misalignment, consider this:

 

What about an individually-calibrated camera? By whatever means, Leica analyzes the M9 before shipping, and the analysis gets stored in on-board ROM. That way the universal software could call up the individual camera's correction data transparently and automatically.

 

Cameras in the field would have to go in for a ROM update, but it would likely be a quick turnaround. The upgrade would be necessary only for those customers who were experiencing the red edges, so it wouldn't be a universal recall.

 

The problem could thus be fixed quickly. The improved Kodak sensor you predict could be installed beginning with a certain production run; with its introduction, the individual calibration would be discontinued. There would be no need for an M9.2.

 

Short-term, the increased cost would be covered by Leica's normal price increases. When cameras with the older sensor and ROM correction go in for after-warranty repair, Leica's charges would include the replacement of the earlier sensor, and the customer would as usual receive a camera updated to latest standard.

 

 

 

 

 

Or, of course, there's the re-think workaround I suggested yesterday at http://www.l-camera-forum.com/leica-forum/customer-forum/116238-la-izquierda-roja-homage.html#post1233090.

[k-hawinkler]

Sandy, thanks for the continued careful consideration you're giving the issue.

 

Last October I jokingly suggested (http://www.l-camera-forum.com/leica-forum/leica-m9-forum/102760-18mm-m9.html#post1083786) that the problem might be some kind of misalignment of the microlenses. At the time, it seemed to me the only possible explanation, with unfortunately difficult correction possibilities. The serious consequences of that are the reason I phrased it humorously.

 

Assuming that the problem is indeed a microlens misalignment, consider this:

 

What about an individually-calibrated camera? By whatever means, Leica analyzes the M9 before shipping, and the analysis gets stored in on-board ROM. That way the universal software could call up the individual camera's correction data transparently and automatically.

 

Cameras in the field would have to go in for a ROM update, but it would likely be a quick turnaround. The upgrade would be necessary only for those customers who were experiencing the red edges, so it wouldn't be a universal recall.

 

The problem could thus be fixed quickly. The improved Kodak sensor you predict could be installed beginning with a certain production run; with its introduction, the individual calibration would be discontinued. There would be no need for an M9.2.

 

Short-term, the increased cost would be covered by Leica's normal price increases. When cameras with the older sensor and ROM correction go in for after-warranty repair, Leica's charges would include the replacement of the earlier sensor, and the customer would as usual receive a camera updated to latest standard.

 

Or, of course, there's the re-think workaround I suggested yesterday at http://www.l-camera-forum.com/leica-forum/customer-forum/116238-la-izquierda-roja-homage.html#post1233090.

 

How much time & effort & cost would that individual calibration require?

How much would a sensor replacement cost?

[sandymc]

What about an individually-calibrated camera? By whatever means, Leica analyzes the M9 before shipping, and the analysis gets stored in on-board ROM. That way the universal software could call up the individual camera's correction data transparently and automatically.

 

Howard,

 

It's possible. What I don't know is how much the characteristics of the individual lens impact on the required calibration, although we do know that it does make some difference. However, given that per Peter's post above there does now appear to be at least one 18mm that doesn't show red edges, while most 18mm's do, I suspect that you would need to calibrate for each individual lens as well.

 

Regards,

 

Sandy

[Alnitak]

Markowich, I'm not aware of any of those lenses (24lux 18SE) who doesn't show the red edges problem. It will be nice to see 2 of them behaving differently with an M9.

Probably in most situation it's hardly noticeable, but may try and shoot upwards against the sky on a sunny (or well light) day, I bet you may find the red edges with both lenses @ full aperture.

I've seen it with my 28cron @ f2 as well, and that's supposed to vignette far less than the 24lux at FA (if the red edges corner was related to vignetting).

 

I've said it before, and I'll say it again: I have the 24 'lux and it DOES NOT show a red edge, no matter what the aperture is. The only Leica lens I have that shows it is the 21/2.8 Elmarit, and even with that lens, it's very faint on my camera. Unless I am shooting a very light subject like snow (not something I get a lot of here in southern CA), its just not a problem in real-world shooting. My WATE has never shown a hint of red on any subject.

 

I have had significant red edges with my CV 12, 15 and 21, and on a decentered copy of the CV 28/2. On a decentered CV21, it was much worse than a good copy of the CV 21--to that end, I really feel strongly that lens issues do come into play, although they are clearly not the root cause.

 

Jeff

[ho_co]

Good point, Sandy. I don't deny that there is certainly some lens variability, but I think it is probably very little.

 

Without knowing more, it's possible that Peter's body shows minimal susceptibility to the red edges. It may be that any 18 SE he tried would not show the problem on his camera, though it might on another body. (As for the 15, Leica isn't likely very concerned about that. )

 

Actually, I'm just arguing for a hypothesis. The pursuit is what interests me.

 

And from what Daniel said above, Leica is already on the trail, so we'll know soon enough.

[ho_co]

Jeff--

There are several ways to branch out from your experience. Let's try just one:

 

First, Leica probably isn't particularly worried about our Voigtländer lenses. The 15/8 Hologon probably also behaves badly on the M9, and that won't worry Leica either.

 

Second, it's reasonable that improperly assembled lenses will make the problem worse.

 

Down to brass tacks:

 

It would be interesting if you or Peter or anyone else who has one of the "problematic" Leica lenses which doesn't exhibit a problem on his camera could try either:

a) the same lens on several other M9s, particularly on an M9 that has shown a problem with another sample of that focal length; or

other samples of that focal length on his non-problematic M9.

 

I'm only laying out a strategy to try to pin down a bit more detail for the problem. I'm glad that your 24 Summilux is fine on your camera. What that doesn't tell us is whether your lens is equally clean on another M9, or whether your M9 fares as well with another 24 Summilux.

 

From your description, it sounds as if your sample of the M9 may be almost free of the problem, as I speculate Peter's is as well.

 

As I said above, I think the problem will probably be primarily with the camera. I think it goes without saying that certain lens designs will be more prone to the red edges, and that certain samples of lenses will be worse than others.

 

You guys with wide-angle lenses and M9s should all get together for a round-robin swap meet.

 

That would answer a lot of questions about which we can presently only speculate.

 

 

 

 

BTW, all of this is just a summary of what Maurizio already said in two sentences:

... I'm not aware of any of those lenses (24lux 18SE) who doesn't show the red edges problem. It will be nice to see 2 of them behaving differently with an M9. ...

[jonoslack]

Advertisement (gone after registration)

You guys with wide-angle lenses and M9s should all get together for a round-robin swap meet.

 

That would answer a lot of questions about which we can presently only speculate.

 

 

HI There Howard

at the risk of being boring and repeating myself (nobody seemed to take any notice last time). The magnitude of the seriousness of the shift on different lenses is not really only speculation:

 

So, I'll repeat myself

 

Most obvious:

18 SE

21 Elmarit

24 Elmarit

28 Elmarit Asph

 

Slight

WATE

28 'cron

 

Negligible

24 'lux

21 'lux

24 Elmar (I think)

 

One thing which might be more relevant than different cameras is the way people process their photos - do an auto-levels and this sort of thing can suddenly appear in a picture where it wasn't visible before.

 

I've looked at some of these lenses on 3 different camera bodies, and the results were pretty much identical. (Of course, this is hardly a statistical sample!)

 

As far as what Maurizio said, I've yet to see anyone reporting it on the 24 and 21 'lux lenses . . . but of course I could have missed it.

[jonoslack]

Howard, Sandy

Surely the microlenses are 'misaligned' intentionally - i.e. shifted towards the outside of the sensor to allow for the increased incidence of light.

 

What isn't much mentioned about the 'red shift' is that it's often connected with a 'cyan shift' in the other corner - it's just that this is much less obvious in most natural situations (like skies and snow, where it simply looks like vignetting).

 

Do we need to even postulate about a misalignment - can't the effect be explained by the fact of the shifted microlenses, light spillage from extremely wide angled light and the lopsided Bayer pattern?

 

p.s. unlike the above post this is entirely speculation

[wlaidlaw]

(nobody seemed to take any notice last time)

 

Please Miss; please Miss - I noticed!

 

Wilson

[jonoslack]

Please Miss; please Miss - I noticed!

 

Wilson

 

oops . . .so you did! . . . what's worse is that I was befuddled by your reply, and only rescued from embarrassment by Andy!

[sandymc]

Do we need to even postulate about a misalignment - can't the effect be explained by the fact of the shifted microlenses, light spillage from extremely wide angled light and the lopsided Bayer pattern?

 

Well, yes, we are making the assumption that the red edges are not simply inherent in the design of the sensor and thus essentially inescapable given current technology.

 

But I think that there are two reasons to believe that this isn't just the price you pay for microlenses on the Kodak sensor:

 

1. The feedback I've received is that the effect does vary from camera to camera, although I would have to admit that I don't have one M9, let alone two to compare.

 

2. Also, if this is just inherent in the sensor, then Leica has surely been just a bit dishonest? That would mean that all the sensors on all the test cameras that Leica built do this. Surely they should have put their hands up by now?

 

Regards,

 

Sandy

[SJP]

HI There Howard

at the risk of being boring and repeating myself (nobody seemed to take any notice last time). The magnitude of the seriousness of the shift on different lenses is not really only speculation:

 

So, I'll repeat myself

 

Most obvious:

18 SE

21 Elmarit

24 Elmarit

28 Elmarit Asph

 

Slight

WATE

28 'cron

 

Negligible

24 'lux

21 'lux

24 Elmar (I think)

 

One thing which might be more relevant than different cameras is the way people process their photos - do an auto-levels and this sort of thing can suddenly appear in a picture where it wasn't visible before.

 

I've looked at some of these lenses on 3 different camera bodies, and the results were pretty much identical. (Of course, this is hardly a statistical sample!)

 

As far as what Maurizio said, I've yet to see anyone reporting it on the 24 and 21 'lux lenses . . . but of course I could have missed it.

If you check the lens designs at Leica Camera AG - Photography - Lenses it is clear that the most problematic lenses protrude into the body the farthest, i.e. the rear element is relatively close to the sensor. So my belief is that this is the cause of the problem combined with a slight asymmetry of the microlens positioning on the sensor which only becomes noticable at oblique angles of incidence. This is exactly in line with Leica's opinion that a FF digital M is (was) impossible, it is precisely the reason they took so long in developing the M9 as you are pushing the boundaries of what is possible.

 

Note that the 21 and 24 lux are designed such that the rear element is located at the bayonet position, this prevents any problems as the angle of incidence on the sensor is similar to 35 mm and higher focal lengths.

[noah_addis]

...can't the effect be explained by the fact of the shifted microlenses, light spillage from extremely wide angled light and the lopsided Bayer pattern? ...

 

 

This makes the most sense in my opinion, though I'm not an engineer and it's just speculation.

 

I do have two M9s (or rather I did, one is at Leica NJ for the line problem). My 21 Elmarit Asph performed identically on both bodies, which were delivered a few months apart.

 

I decided to drop my camera off in NJ so I could talk to the technician. The repair staff had not heard of the red edge problem but he was very interested. And no one else was around to answer my questions of if/how/when will the problem be fixed.

 

I did my initial tests in the snow but later tried a more consistent test by shooting a large white piece of paper in diffuse open shade, camera focussed at infinity. I converted in ACR with as-shot WB and in PS I set a neutral point in the center of the frame.

 

While the red edge (at left and bottom) was definitely there at all apertures, it may be easier to see as you stop down. The vignetting of the lens wide-open tends to mask it a bit.

 

I think some of the variability does have to do with testing methods, processing methods and maybe even monitors. The problem is less obvious on my 13-inch macbook with a profiled screen than it is on my 24-inch Imac (profiled with the same equipment).

[xtoph]

Has this been stated publicly somewhere?

 

Jeff

 

i haven't seen a public posting yet, but i did receive today an email from an official leica service rep in germany (m.s.) who tacitly acknowledged the issue (i had asked regarding my problems with the 28 cron on the m9) and stated, quote:

 

"Especially the correction profiles for the wide angle lenses will be reworked with the next Firmware-Update, that will be available in ca. 4 weeks on our homepage."

 

so this is encouraging. assuming that they really can do the fix in fw.

 

btw, the problem i am seeing with the 28 cron is not subtle; it is plainly visible in real-world pics.

[jonoslack]

Well, yes, we are making the assumption that the red edges are not simply inherent in the design of the sensor and thus essentially inescapable given current technology.

 

But I think that there are two reasons to believe that this isn't just the price you pay for microlenses on the Kodak sensor:

 

1. The feedback I've received is that the effect does vary from camera to camera, although I would have to admit that I don't have one M9, let alone two to compare.

 

2. Also, if this is just inherent in the sensor, then Leica has surely been just a bit dishonest? That would mean that all the sensors on all the test cameras that Leica built do this. Surely they should have put their hands up by now?

 

Regards,

 

Sandy

 

HI Sandy

one thing I'm 99.99% certain of is that Leica were not aware of this as an issue - let's face it, it wasn't really observed for some time, added to the fact that if they were aware they would have been less aggressive with their vignetting corrections.

 

I certainly realised that it was an issue with the CV15 when I was testing, but it wasn't really part of the remit to work out what was happening with other than Leica lenses - and I never noticed it on any of the Leica lenses I was using.

 

Do you have feedback from anyone with two cameras who says that it varies? Because it's very much a function of post processing, and of course of the lens.

[Mauribix]

Ok, so now I'm posting some samples of the 24lux on an M9 showing the threaded "red edges issue".

Ok, these were caught in Oslo, so the snow is more than an "alley" in this case, but that's just to say that even the snowy landscapes are often a "real world" environments for those who live not in the sunny south.

 

Jeff, I'm glad to know that your copy of the 24lux doesn't show the issue, I'd like to know why it behaves differently from those two samples I've tested.

One was kindly borrowed to my by my mate Jason in Oslo.

 

Here we are:

24lux @ f1.4 M9

DNG are available under request.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Just Alex]

Wasn't the M8 debacle enough of an embarrassment for Leica? Isn't anyone testing designs or performing quality control at Leica before a product is released?

 

It's unacceptable that professional equipment costing this much has such a fundamental flaw with its operation. What's worse is that Leica has said nothing while its customers are left to speculate.

 

Doesn't someone have access to the necessary equipment/engineering knowledge to come up with a more definitive answer?

[jonoslack]

Ok, so now I'm posting some samples of the 24lux on an M9 showing the threaded "red edges issue".

Ok, these were caught in Oslo, so the snow is more than an "alley" in this case, but that's just to say that even the snowy landscapes are often a "real world" environments for those who live not in the sunny south.

 

Jeff, I'm glad to know that your copy of the 24lux doesn't show the issue, I'd like to know why it behaves differently from those two samples I've tested.

One was kindly borrowed to my by my mate Jason in Oslo.

 

Here we are:

24lux @ f1.4 M9

DNG are available under request.

HI Maurizio

Now I HAVE seen the red edge on the 24 'lux!

I'm currently eating my words mngg mngg mngg

all the best

[thrice]

Has this been stated publicly somewhere?

 

Jeff

 

No, I have an e-mail from Stefan saying this. If you don't believe that just wait 4 weeks I feel no need to post a private e-mail on the forums.

 

If people would rather complain instead of take pictures for the next 4 weeks far be it from me to question them. But don't say Leica haven't acknowledged or aren't working on the issue. Why not e-mail Leica or your Leica dealer if you need affirmation.

 

In the meantime if anyone wants to sell me a 35 cron asph, 18 Super Elmar or 21/2.8 asph because they don't think Leica will ever fix it, I am offering premium prices from the perspective of these being redundant lenses

 

I find a lot of the hypothesising as to the cause very interesting though, and that seems to be the way the conversation is going which is encouraging. How many of you have set the lens detection to off and shot your wides? Non-uniform cyan drift would more indicate a physical difference rather than a correction asymmetry.

 

HI Sandy

one thing I'm 99.99% certain of is that Leica were not aware of this as an issue - let's face it, it wasn't really observed for some time, added to the fact that if they were aware they would have been less aggressive with their vignetting corrections.

 

Hey Jono, why not send an e-mail to Leica to ask? Hard to be certain based on pure speculation of their intentions.

[Advertisement]

Hallo,

Look here: Leica M

[Mauribix]

HI Maurizio

Now I HAVE seen the red edge on the 24 'lux!

I'm currently eating my words mngg mngg mngg

all the best

 

lol Jonathan,

 

BTW, I'm confident Leica's gonna fix the issue soon.