M9 and the dreaded magenta cast

[mgreernz]

Advertisement (gone after registration)

After 50 years experience with SLR's, I've taken a deep breath and enthusiastically launched myself back to a rangefinder. Ticking one more off my bucket-list, I've purchased a new Leica M9.

 

But I'm encountering a problem widely reported prior to the release of firmware 1.116, except that in my case, 1.116 was already factory installed. I'm attaching 3 examples, 2 of which were specifically taken against a grey sky to demonstrate the problem.

 

L1000299

Taken at ISO 160. There is an observable magenta cast to the left hand edge and is very similar to what appears in a number of the photographs I took in the first serious photo-shoot with this camera. Any post-editing of the photograph makes this magenta cast even more distinctive. The problem is occuring with any and all of the 3 lenses I am using (Summicron ASPH 2/28, 2/50 and 2/75). It appears at all ISO speeds, but is markedly more noticeable at higher ISO.

 

L1000300

Taken at the other extreme (ISO 2500), the same magenta cast to the left is clearly visible. In both photos, there are some traces of the cast to the right, but significantly less than to the left.

 

L1000108.jpg

This is more than just a magenta "cast". This is only the 108th photograph taken on this M9. It is not an accidental photograph: it was carefully composed and the photographs either side of the sequence of files are "normal". Clearly, something is seriously wrong with the way the camera has processed this file.

 

I would be grateful for some expert advice please, before making any decision to send the camera to Germany for assessment/repair.

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!

[Shade]

The last picture is seriously wrong - did you code the lens on your last picture? And which of the three lenses did you use?

 

And I am somewhat baffled that you have magenta casts in ALL your lenses... it shouldn't... I only have these magenta shift on my 18mm Super Elmar. And that is barely noticable after the newest update...

 

With my 50 Lux and 90 Cron, it's colour is just right...

[mgreernz]

Hi Shade - seriously wrong is also what I've assumed! All 3 lens are new, 6 bit coded versions. The file #108 was taken with the Summicron 2/50, the same lens used for the two grey cloud test shots.

[nryn]

I'm not entirely sure what's wrong, but people are going to ask: what software are you using to process these?

 

Something is definitely wrong with that last image. Every once in a great while I get a white balance jump when white balance is set to auto, but nothing like this.

[mgreernz]

A possible link to the RAW conversion process had also occurred to me. From uncompressed DNG, I therefore processed the original files using two different applications: Aperture 3.0.3 (which is my normal library management and workflow) and Photoshop CS5 with Adobe Raw 6.1.0.242. The results were pretty much the same for each.

[ho_co]

Welcome to the forum, and to the M9, Greer! But this is an ugly way to start, isn't it?

 

I'm unaware of any "dreaded" magenta cast. The only "dreaded" thing I know is the "dreaded green stripe" that can't be avoided in some lighting situations with the M8.

 

You've done some kind of correction in the second picture, making the center quite cyan.

 

To me, the first two shots look like the M9's "red edge." There are a lot of threads on the topic on the forum.

 

 

But besides the color, that third picture is badly blurred, possibly because of focus, possibly because of motion, more likely because of a camera problem.

 

Send the file to Leica and ask for their opinion.

 

That has happened exactly once in more than 108 photos, right? It shouldn't happen at all, but it could be that you pressed the release through without letting the meter wake up, or it could be a problem relating to storage to card. Your card is on the recommended list, right? Was the previous image still being written to card? Can you think of any particular thing that distinguished the taking of #108 from the others?

 

To me, the first two don't show a problem worth worrying about (but they do bother you, so forget about my opinion ). The third one is a serious messup, probably caused by the camera (my guess), but there are other things you can eliminate while awaiting Leica's response to your file submission.

 

You've got a strange one there. Do keep us posted as you get this taken care of. And good luck! This is a horrid disappointment on a camera just out of the box.

[adan]

Advertisement (gone after registration)

In the third shot, even the bottom edge of the image itself is blurring (black edge at bottom). That looks to me more like the kind of "melting" image one gets when a sensor is failing, than just camera shake: Sony DSC-F717 CCD sensor failure photo - Richard Allen photos at pbase.com

 

The strong color cast also looks like an M8 image shot with an IR-pass filter - i.e. infrared image.

 

The first two, while a bit more magenta-edged than I'd expect with 35/50 lenses, are within the realm of the possible with a functioning M9 (esp. at ISO 2500) - but that last one looks like a symptom of a major problem developing.

 

I would definitely e-mail Leica with that sample, and if possible figure out a way to host the original .dng file on the web for them to look at.

[mgreernz]

Thanks for the helpful comments Howard.

 

Yes I've probably used the term magenta carelessly - and it is probably better described as red.

 

Yes, the SD card is from the "approved" list (16GB Sandisk Ultra II, Class 4).

 

No, there's been no post-editing correction to either picture. They are "as-shot" conversions from RAW to email sized .jpeg (for posting to the forum). One has been taken at ISO 160, the other at ISO 2500.

 

As far as #108 is concerned - I can't now be sure of the exact circumstances. But it looks to be a carefully framed shot, suggesting that it wasn't taken in a hurry. Also the shots either side of this file are taken in the same area, are ok and spaced a minute or two apart. So that kind of eliminates the possibly of a write-to-card traffic jam.

 

But I'm very grateful for the comments that are coming in. From this end of the world (wet, cold and miserable this week), I don't want to ship a new M9 off to Solm for a 6-8 week absence unless it's likely to be necessary.

 

Michael Greer, Christchurch New Zealand

[mgreernz]

Thanks Andy.

 

You suggest emailing Leica - and I'd considered doing that, but ended up uncertain what the relevant email address would be. Anyone able to identify the address please? I can easily use DropBox to give access to the original DNG.

 

Sorry for such dumb questions from an old head flapping around a new camera.

___

 

Michael Greer

Christchurch, New Zealand

[zlatkob]

All 3 lens are new, 6 bit coded versions.

 

Was the camera set to automatically detect the coding? I presume you set it to automatically detect as all 3 lenses are coded, but it's one more thing to check to try to determine the problem in the first 2 photos.

[andybarton]

Just email Customer Services at Leica in Solms, with these examples and as much detail of the circumstances as possible

 

they will advise what to do next

 

their email address is on Leica's website.

[mgreernz]

Thanks.

 

Yes, Lens detection: auto

 

Michael Greer

Christchurch, New Zealand

[brill64]

this may sound like a really stupid question but was there a uv/ir filter fitted to your lens in the last shot?

[archi4]

The Leica Service addresses are at:

Leica Camera AG - Service - headquarters germany

 

maurice

[Shade]

Briir, a uvir filter has never brought up any problems for me during daylight, so I assume it shouldnt be a problem for him either.

 

Eitherway, the first two shots as many have mentioned seems to be still within reasonable range.. the last shot however will get me sending my camera back. Have you gotten more than one picture with exactly the same problem?

 

I had my sensor busted also or sensor board, shipped it to Leica Hongkong and I got it back within 10 days with a replaced sensor or sensor board..

 

I would hate if the picture was a keeper and turns out its a damaged file.. Id rather not use my m9 for a while rather than get damaged images..

[mgreernz]

...was there a uv/ir filter fitted to your lens in the last shot?

 

Nope. But thanks for the thought. The general discussion thus far, seems to be leaning in favour of the M9 needing a little overseas holiday...

______

 

Michael Greer

Christchurch, New Zealand

[250swb]

Welcome to the forum.

 

The top two seem to be shot at different f stops judging by the dust showing in #2, so maybe there isn't the consistencey needed to judge properly the red edge? For instance under exposure will accentuate 'red edge' if it is occuring. #3 does look weird, its a worry, but I'd shoot masses of pictures and try and reproduce it more times, to see if there's a common denominator involved through operation, or if its a fundamental recurring problem. A few people have had the odd glitch with the M9 that happens without warning or repetition, things like blank frames, so maybe it is just a writing to card error that won't come back. You could try another card maybe?

 

Steve

[jaapv]

The use of IR/UV filters can cause red shift towards the edges, so I would take some shots with the filters removed like Leica recommends. As for the last shot, it does indeed look like a sensor failing or glitch. If it doesn't occur again you might disregard it, but if you want to make sure send the file to Solms, as Andy recommends. I think I can predict the answer, though: "send the camera in":(

[vanhulsenbeek]

............. I think I can predict the answer, though: "send the camera in":(

 

Likely, but as you are in your first pictures (my camera started at about # 100) I would put in a firm request for an immediate replacement. Probably faster to have a new camera sent from Solms than this one repaired.

[Advertisement]

Hallo,

Look here: Leica M

[arthury]

Is the last shot repeatable or is it occurring at random?

Are you able to obtain good images after the last shot?

 

It would be nice if you grab the EXIF data and slap it into the next respond so we can glance thru it just in case there's something obvious.