Erwin Puts Reviews M8 IR filter

[bob cole]

Here's another report from the Leica Users Group:

At 8:51 PM +1000 4/3/07, G Hopkinson wrote:
>Folks, Erwin Puts has upoaded a new report on the use of IR filters on the M8.
>http://www.imx.nl/photosite/leica/M8_7/m8_7.html
>
>Some folks may like to look at it.
>
>cheers
>Hoppy

and some folks may note that the captions on the photos are reversed
as to which pictures and data are with filter and which without.

--
* Henning J. Wulff
/|\ Wulff Photography & Design
/###\ mailto:henningw@archiphoto.com
|[ ]| http://www.archiphoto.com

_______________________________________________
Leica Users Group.
See http://leica-users.org/mailman/listinfo/lug for more information

 

[Ben Z]

I've seen enough shots where foliage in broad daylight has been skewed by IR that if and when I buy myself an M8 I expect I will keep an IR filter on it for almost all my shooting. I don't consider that a problem because I've always kept a B+W MRC UV filter on my lenses and never noticed any ill-effects--though no doubt there are some people out there who have (or believe they have) greater discerning power than I and would argue that all filters degrade noticeably 😉

The only place I've ever been concerned about a filter is at night with point sources of light in the shot--like Erwin's first example. In those cases I have found reflections with my ubiquitous UV filters and would remove them for those type of shots. I am hopeful that with the M8 I wouldn't be caught between the proverbial rock and hard place, and could remove the IR filter in such conditions. From Erwin's one example it appears to be doable. However, an indoor shot with tungsten tracklighting or spotlights, and people dressed in black, that I don't know about. It seems like it might come down to having either to manually select the off-color clothing and correct the cast, or clone out reflections. Both would be tedious if there were a large number of shots.

Much as I respect Erwin's efforts (and I've met the man personally, he's a gentleman and a scholar) I feel like I might end up doing just the reverse of what he suggests, that is keep the filter on all the time like I do with UV currently, and only remove it if it looks like reflections will be a worse problem than IR.

 

[Richard Marks]

Ben I have to say even with tungsten and synthetic clothing in practice I have found its rarely a probelm. I suspect the actual situation for getting the magenta cast is rather more complex. It could even depend on the temperature of the garment relative to the suroundings and relative to the proportion of tungsten vs flash vs daylight in the exposure.

I would be very suprised if the filters made much of a difference to my images (which is just as well because they have yet to arrive in the UK!!)

I also along wiht a few other posters suspect that the 1.092 firmware may have adressed this issue in part although Leica themselves have made no such claim.

Best wishes


Richard

 

[willie_901]

Ben I have to say even with tungsten and synthetic clothing in practice I have found its rarely a probelm. I suspect the actual situation for getting the magenta cast is rather more complex. It could even depend on the temperature of the garment relative to the suroundings and relative to the proportion of tungsten vs flash vs daylight in the exposure.

Since the M8's release I have wondered why the IR effect does not show up in photographs where foliage is prominent. Many of the photos I've seen from IR film or digital cameras modified to enhance IR frequencies indicate plants reflect IR strongly. Yet no one ever discusses funky green colors in M8 photos where lots of foliage is present. Why would fabrics that reflect IR strongly be different from plant foliage?

I would be very surprised if the filters made much of a difference to my images (which is just as well because they have yet to arrive in the UK!!)

I would be surprised if you didn't see a difference. For instance blood vessels near the surface of the skin release IR energy. Of course if your favorite subjects are unlikely to reflect or generate IR energy, then you will certainly see no difference.

I also along wiht a few other posters suspect that the 1.092 firmware may have adressed this issue in part although Leica themselves have made no such claim.

It is impossible to know before the data is recorded where erroneous IR photon counts will occur, and how many counts will be from unwanted IR energy photons, and how many counts are due to wanted visible energy photons. Even a partial a post-data acquisition solution seems unlikely. Howeve, if IR contamination can be accurately and significantly reduced post-data acquisition, then Leica 's alleged 1.092 firmware algorithm will be a very valuable patent. IR contamination is a significant problem in digital video and still photography. A post-acquisition firmware solution would revolutionize digital imaging.

willie

 

[Richard Marks]

Hi Willie
I take plenty of portraits and am generally very pleased with the skin tones! I am well aware of the theories which would suggest improvement with an IR filter. I am simply saying that in practice as things stand it would be pretty difficult to make a substantial improvement to the vast majority of images.

As an M8 user yourself, are you experiencing a lot of magenta cast problems? If so then you would of course notice a big improvement.

With regards to the problem not arising when photographing plants, I tink the explanation is more that in the absence of IR enhancing filters, the total proportion of reflected infrared is small compared to the other reflected frequencies such that the effect is negligible. The cast issue usually seems to occur when overall daylight levels are reduced and with fabrics maximally IR radiating.

Maybe someone with more time on their hands than I should set up an experiment where the proportion of daylight to tungsten is adjusted with shots of dark synthetics.

Ultimately time will tell when we get our hands on some IR filters. Personally I am quite happy as things stand but any improvements however slight will of course be a bonus.

Regards


Richard Marks

 

[jaapv]

We cannot deny that the M8 has high IR sensitivity...

B&W 092 IR pass filter on the M8:

 

[Richard Marks]

Hi JAAP
Im not denying that at all!
I simply suggest that the mechanism for the magenta cast is probably when the proportion of infra red to other reflected light is relatively high. This would explain the issue with dark materials in artificial light. Its probably a matter that they reflect proportionally less light in the normal spectrum. If so then this may make predicting the need for an IR filter more reliable.

 

[Ben Z]

Since the M8's release I have wondered why the IR effect does not show up in photographs where foliage is prominent.

It does. I've seen some shots posted to specifically illustrate it, where green foliage has gone way over to the yellow side.

I also along wiht a few other posters suspect that the 1.092 firmware may have adressed this issue in part

It looks like they did an in-camera tweak similar to what Jamie Roberts' C1 profile does in post. I'm guessing that ultimately what Leica's going for is one type of color mapping when "IR Filter=Off" for those who don't use the filters, and another mapping with "IR Filter=On". If that's the case then a problem may arise if someone uses uncoded lenses if the IR Filter On/Off isn't a separate parameter from Lens Detection On/Off.

 

[Richard Marks]

Very interesting Ben
Does the yellow go away if an IR filter is added?
If Leica have as I suspect partially adressed this in the firmware upgrade they really should have mentioned it. I can however understand why they may chose to play it down a bit!
I guess we will have to await the arrival of 1,1o and of course the filters!

 

[jaapv]

Hi JAAP
Im not denying that at all!
I simply suggest that the mechanism for the magenta cast is probably when the proportion of infra red to other reflected light is relatively high. This would explain the issue with dark materials in artificial light. Its probably a matter that they reflect proportionally less light in the normal spectrum. If so then this may make predicting the need for an IR filter more reliable.

LOL Richard - I know 😉. Seriously, if you look at my - admittedly grotty- IR photograph above, you'll notice that it puts unnatural highlights and shadows in unpredictable places, making exact correction impossible. This is the overlay you get over your real shot at about -2 stops , and out of focus as well. The IR filter, imo, is mandatory to get rid of this interference.

 

[Ben Z]

Although I was and still am disappointed over the whole IR filter deal, from what I've seen and gathered in terms of trying to correct the IR casts without internal or external filters, I would much rather Leica did not try to minimize the IR cast through fiddling with the in-camera color mapping. That or profiles designed to get rid of "magenta blacks" is kind of like putting on perfume instead of taking a shower 😀 I would hope they either install new chips with better IR filtration, or if that's not possible without other dire consequences, then optimize the color mapping for use with an IR filter.

 

[Richard Marks]

Ben I very much agree with this. I think the least they could do is to say what they are doing with the upgrades. For all we know 1.092 may not have addressed the magenta issue and the rest could just be placebo effect! I do find the auto white ballance a lot better (and this they do admit to fiddling with) and this would have some effect on perceived colour rendition too.

 

[jaapv]

or if that's not possible without other dire consequences, then optimize the color mapping for use with an IR filter.

That is what 1.10 is supposed to do. And it is close,close,close.. (I hope)

 

[StuartR]

My experience is that the IR is present in almost every M8 picture without a filter, but you need to be aware of what you are looking at. It is not always a shift to purple in blacks...often it is a ruddiness in skin-tones, a drift in foliage color and so on. I notice this because I also use the DMR which has superb color, and it's the funniest thing...put an IR filter on the M8 and the photos match the DMR quite well. Take the IR filter off and weird shifts happen. I agree completely with Ben here...there are situations where you need to take the filter off, like with strong light sources at night etc, but in all other situations leave it on. It is annoying...particularly that they cannot fully multicoat the filters so there is more flare than normal, but getting accurate color is more important.