Leica M9 and IR - should we keep our filters?

[holgerf]

Advertisement (gone after registration)

…

I still think a "dual" correction could be achieved via menu option with or without IR filter, specially because Leica engeneer know how to deal with it and second for people shooting both M8 and M9, they could leave the filters on all the time.

 

Second that especially to easily use the lenses on M9 and M8 as well.

 

I expect that many will keep their M8 body as backup instead of trading in and buying a second M9 as backup. In this case there is no practical way to switch the bodies without great inconvenience of getting the lousy filters on and off.

 

For those the option Lens Detection On + UV/IR would be the solution. The algorithm should be the same just reciprocal–or did I miss something?

 

Best

Holger

[Kittyphoto]

Well said Holger. Exactly my thought.

 

kitty

[reddawn]

That is not consistent with the results I've seen when testing this. There is indeed a difference between the IR sensitivity of the M9 and a Canon 5D Mk II. It is especially apparent under tungsten.

 

Leica's own testing results are in line with what my tests showed as well so I'm not sure what DPReview did differently. It looks like they may not yet have tested under tungsten.

 

I'm guessing you haven't seen the tests that were part of my initial review. There is a difference but it primarily might be of consequence to people doing color critical work under tungsten.

 

Cheers,

 

Sean

 

hi sean,

 

phil did do the test under tungsten lights. from dpreview....

 

"With the M9, Leica has addressed the IR problem with a redesigned sensor cover filter, which is now thicker and made from a different material. To demonstrate the difference this makes compared to the M8 we created a simple test scene lit by 800W tungsten lamps (which of course produce lots of infrared, so this is a pretty tough test). The contrast is stark - the M9 renders the scene correctly where the M8 fails."

 

I've seen the samples in your review too and in the end, my own personal conclusion is I would not bother with the little contamination that exists, and will use M lenses with no IR cut filters, the way they were intended!

[sean_reid]

hi sean,

 

phil did do the test under tungsten lights. from dpreview....

 

"With the M9, Leica has addressed the IR problem with a redesigned sensor cover filter, which is now thicker and made from a different material. To demonstrate the difference this makes compared to the M8 we created a simple test scene lit by 800W tungsten lamps (which of course produce lots of infrared, so this is a pretty tough test). The contrast is stark - the M9 renders the scene correctly where the M8 fails."

 

I've seen the samples in your review too and in the end, my own personal conclusion is I would not bother with the little contamination that exists, and will use M lenses with no IR cut filters, the way they were intended!

 

Thanks for the info.

 

I don't plan to use filters on the M9 either but I think its important that people see the differences for themselves and then make an informed choice about whether or not to bother. One would want to for catalog work, some fashion, maybe certain weddings, etc.

 

Cheers,

 

Sean

[pascal_meheut]

I do not plan to use them either as getting rid of them is one of my main reasons to switch to the M9.

But I would like to see if they make a difference with skin tones under certain lights or if the redish/pinkish faces we've seen sometimes are just a calibration/WB problem. In general photography, this is something I encounter more often and which is more painful to post-process than some black material slightly becoming purple.

 

And as Leica seems to deliver M9 in Paris rather slowly, I cannot test it myself yet

[adan]

Pink reds are common if one tries to use a calibration created under daylight for shots made in tungsten or other low-blue light. One really needs to create separate calibrations from a ColorChecker or similar standard under all different kinds of lighting. Ideally even different calibrations for different ISOs (or at least ranges, now that M9 has so many ISOs). I get pink skin under white-balanced tungsten light even with UV/IR filters in place on the M8 if I use my "sunlight" calibration.

[pascal_meheut]

Advertisement (gone after registration)

et pink skin under white-balanced tungsten light even with UV/IR filters in place on the M8 if I use my "sunlight" calibration.

 

So do I but I always considered it was because IR filters were not 100% efficient. The DMR gave me the same skin tones under tungsten light and with flash as did other cameras. The M8, not, even with the filters.

[sandymc]

That is not consistent with the results I've seen when testing this. There is indeed a difference between the IR sensitivity of the M9 and a Canon 5D Mk II. It is especially apparent under tungsten.

 

Leica's own testing results are in line with what my tests showed as well so I'm not sure what DPReview did differently. It looks like they may not yet have tested under tungsten.

 

There is a rational (and probable!) explanation for the discrepancies in test results that we're seeing - there are two parameters to a filter, the frequency that it starts cutting at, and the amount it cuts (I simplify). Leica have stated that they are using a different material - it is likely that the new material attenuates a lot, but only begins cutting IR relatively far into the IR spectrum.

 

So the results that you would see when comparing would depend on the spectrum of the lighting - how much "deep IR" it had relative to how much "near-red" IR.

 

Pity we don't have a full sensor data sheet - that would tell us for sure.

 

Sandy

[mweiner]

@Sean

@Pascal

 

You say you will not be using UV/IR filters most of the time.

 

For people who do prefer to keep a filter on for lens protection,

would you recommend the UV/IR or just UV?

 

Thanks

[sean_reid]

@Sean

@Pascal

 

You say you will not be using UV/IR filters most of the time.

 

For people who do prefer to keep a filter on for lens protection,

would you recommend the UV/IR or just UV?

 

Thanks

 

If the M9's slight sensitivity to IR doesn't bother you (and it is mild under most lighting) then I'd go with UV filters through 50 mm and then IR/UV for 75 mm and up. I say the latter because there shouldn't be a cyan drift penalty for the extra IR filtering on the lenses with narrow fields of view.

 

Cheers,

 

Sean

[pascal_meheut]

I second Sean. When I want to protect a lens with a filter, I buy a good UV one, like the a B+W. I may not use an IR filter at all on the M9 even with long lenses because I do not like outside and inside reflections.

[suden]

Per Leica's website from their FAQ on the M9:

 

I use both the M8 and M9. Can I leave the UV/IR filters on the lenses?

 

For focal lengths of 35mm and longer, yes. For focal lengths wider than 35mm this is not recommended, because through the UV/IR filter a color shift will appear in the corners.

 

I think I will just keep my UV/IR on my 35mm LUX and see how it goes.

[Guest EarlBurrellPhoto]

Leica could add a correction mode into the M9 which allows us to use IR filters.

Would also help those people who like switching lenses back and forth between M8 and M9

 

DINGDINGDINGDINGDING! Give that man a cee-gar! That's exactly what Leica should do, and I hope they man up and do it in a firmware update, not listen to the suits in marketing who're afraid it'll stop some people from buying a second M9 instead.

[jonoslack]

Just two pictures I shot today with and without IR filter (ISO 1000, 1/8th second, so pretty dark, -1 stop, halogen light). RAW processed in Apple Aperture, but no post-processing, except white balance adjustment by clicking at the grey blanket. Just look at the liner of the jacket. I did the same test three years ago and the entire jacket was magenta, same for the carpet in the background. This is quite a torture test, so Leica did a pretty good job, but not an absolutely perfect one (just as they say in the Q+A document). I would continue to use IR filters at black-tie events, etc. (no problem for me, actually).

 

Regards

 

Wolfgang

 

Hi Wolfgang

I did lots of similar torture tests with the M9.

 

The difference that I found from the M8 issue was that (on the rare occasions it cropped up) it was easily corrected by dropping the magenta luminance in Aperture.

 

The conclusion I came to was that I needed to keep my eyes open in weddings etc. but that I would be able to deal with it if it occurred.

 

Indeed, I've now done 3 weddings with the M9, and in no case have I needed to do any adjusting. The only time it's cropped up in 'real' life situations was a late evening shoot in natural dusk, and a lady's black t-shirt had a magenta tinge - which was easily fixed.

 

The danger with using an IR filter, is that you may find you get colour shifts at the corners using wide angle lenses.

 

 

For those the option Lens Detection On + UV/IR would be the solution. The algorithm should be the same just reciprocal–or did I miss something?

 

Best

Holger

 

There was much discussion about this during testing, and whether they should have an IR on setting in the menu, but they decided that wider than 28mm it wasn't really possible, and that for 35 and above it didn't really matter! The difficulty was due to the larger sensor and much more oblique lighting. (as I understood it).

[jonoslack]

So do I but I always considered it was because IR filters were not 100% efficient. The DMR gave me the same skin tones under tungsten light and with flash as did other cameras. The M8, not, even with the filters.

 

Hi There Pascal

The M9 still seems to give too much pink in skin tones at high ISO, however, unlike the M8 it seems to be easily fixable with a small hue and luminance adjustment,

 

What's more important is that the M9 doesn't make all my middle aged mates look like they've just drunk a couple of litres of red wine!

[ho_co]

These real-world results are nice and on-target, of course.

 

But the first picture I saw on the forum that illustrated the M8's IR sensitivity was one that someone made of the top of his R-D1. It really made the camera look laughable because Epson used so many different materials for the top.

 

So may I suggest that someone photograph the top of an R-D1 under tungsten light with the M9? That would bring us full circle.

[Artichoke]

skin tones & artificial fabrics, particularly in artificial light are well known problems for cameras with weak IR filters

foliage is another problem & one much more difficult to correct well in post

I wonder how well the M9 handles foliage sans IR filter

[Guest EarlBurrellPhoto]

skin tones & artificial fabrics, particularly in artificial light are well known problems for cameras with weak IR filters

foliage is another problem & one much more difficult to correct well in post

I wonder how well the M9 handles foliage sans IR filter

 

I doubt if it would be an issue. Even black synthetics under tungsten light show only a fraction of the IR sensitivity of an un-filtered M8. It's enough to require IR filters only if you or your clients' standards are picky.