Update on Leica and 486 IR Cut Filters

[carcam]

Advertisement (gone after registration)

Thanks Scott

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jrc]

I just went out and shot a white garage door with a 21 f2.8 with a B+W UV IR filter in place and the corners seemed definitely cyan to me (opened in Photoshop with no adjustments.) Actually, not just the corners, there was a whole vignetting effect. However: I did this fairly late in the evening without good light. I'll go out tomorrow morning in better light and try again. But I was not especially happy with what I saw tonight.

 

Edit: Whoops; moderate brain damage alert. I rechecked my camera, and the lens selection was set to On, rather than On+UV/IR. The lens is coded all right, but it wasn't expecting the filter...so...I'll go get my pointy hat and sit in the corner for a while.

 

JC

[Guest guy_mancuso]

That damn user error comes around once in awhile. LOL

[grober]

I did some experimentation tonight with my CV15/4.5, a white wall, and tungsten illumination.

 

scott

 

Scott,

 

You did all this testing with an UNCODED Heliar, yes?

 

-g

[sean_reid]

I just went out and shot a white garage door with a 21 f2.8 with a B+W UV IR filter in place and the corners seemed definitely cyan to me (opened in Photoshop with no adjustments.) Actually, not just the corners, there was a whole vignetting effect. However: I did this fairly late in the evening without good light. I'll go out tomorrow morning in better light and try again. But I was not especially happy with what I saw tonight.

 

Edit: Whoops; moderate brain damage alert. I rechecked my camera, and the lens selection was set to On, rather than On+UV/IR. The lens is coded all right, but it wasn't expecting the filter...so...I'll go get my pointy hat and sit in the corner for a while.

 

JC

 

Even leaving that mistake aside, it won't be just the corners that are affected with an IR-cut filter. The reason I keep calling this a cyan drift (or red vignetting) is that its a gradual drift to cyan that begins in the outer zones and increases in intensity as one moves further and further off-center. The pattern looks a lot like some kinds of luminance vignetting.

 

Cheers,

 

Sean

[sean_reid]

I did some experimentation tonight with my CV15/4.5, a white wall, and tungsten illumination.

This is a worst-case example, because of the higher red content and the greater amount of long wavelength reds waiting to be cut out by the vignetting effect of the IR-cut filter.

 

I shoot with the lens focussed at infinity, two feet from the wall to blur things. First, the natural color temperature, determined by white balancing at the center of the image in C1,

is 2500K, with a color shift of -6 (in the green direction, correcting for warmth). With an E39 Leica UV/IR filter the color temperature is 2550 K and the shift needed is -1. That's a very slightly colder temperature, but 5 C1 units cooler (greener) than without the filter. The B&W 486 filter also white balances to 2550 K with a shift of +1, so it is 2 units cooler than the Leica filter.

 

The cutoff effects of the B&W 486 are also stronger than the Leica filter. Just measuring the depression of the red channel compared to the green channel, and using intensity (the square of the RGB numbers, to undo the gamma correction), the 15/4.5 even with no filter shows a 10% depression of the reds in the extreme corners of the image (due to the existing IR filter). The Leica filter (with no corrections applied in firmware) depresses the red intensities by 30% and the B&W 486 depresses reds by 40% in the corners. It is a small but noticeable difference in this case. The difference should be smaller in daylight.

 

The filters are definitely not the same, but I suspect many of us can live with the B&W filters.

 

scott

 

Thanks Scott. That's interesting. Down the road, we may discover that the B+W actually cuts more IR than the Leica filter. I'll find out in testing at some point.

 

Cheers,

 

Sean

[sean_reid]

Advertisement (gone after registration)

OK, I just finished my first round of controlled tests with a coded Elm 21 Asph. As always, these were done at night so that only one type of lighting illuminated the test board. It's lit by two Lowel Tota Quartz lights both bounced into umbrellas. The lighting across the board varies by less than 1/3 stop and the board is a uniform color. Quartz (tungsten) lighting is much richer in infrared content than daylight so this test represents (more or less) a worst-case scenario.

 

I tested the three 55 mm filters I had on hand: Schneider 486 MRC (which was loaned by Leica as what used to be the "official" filter), B+W 486 and Heliopan Digital. The good news is that with all three filters, the new firmware makes the files much more neutral in the outer zones. With the two 486 filters, the corners actually can tend to drift a bit towards red-green rather than blue green, but its pretty mild. There's just a hint of it in the corners but my guess is that with normal subjects (as opposed to my white test board) it might be hard to detect. For example, with the lens at F/2.8 and the B+W filter mounted, the center RGB is 181,181,181 and the upper left corner is 110,112,102. The Heliopan filter, by contrast, still drifts slightly towards blue-green. For example, at F/2.8, the Heliopan center is 181,181, 181 and the bottom left corner is 103,112,111.

 

I have more tests to do with lenses of other focal lengths but my hunch, again, is that B&W 486 filters and new firmware are going to be OK with the Elmarit 21 Asph. for most work, especially under daylight which is not as rich in IR. This should apply, in increasing degree, to the 24, 28s, 35s, etc.

 

I'll do some more "real world" testing with the Elm 21 Asph and the 486 this weekend. I think the lenses where it's really going to be important to use the official (current) Leica IR-cut filter will be the WATE and (for those who hand code) the CV and Zeiss 15s. Otherwise, I think the longer lenses may be OK with 486s.

 

This is all, of course, to the best of my knowledge and experience so far.

 

Cheers,

 

Sean

[Artichoke]

this pretty much confirms my experience

I appreciate your posting this so promptly

[sean_reid]

My pleasure. It's good news for those who want to use the 21 now, given that the Leica 55s may not be available (at least in the USA) for awhile.

 

Cheers,

 

Sean

[waterlenz]

The official statement given to me this morning simply reads:

 

""Leica Camera can only guarantee the correct function for the Leica Filters. Leica has no control about how, e.g., the coating of the B+W filters change in the series."

 

Cheers,

 

Sean

__

 

Sean Reid

http://www.reidreviews.com

 

Thanks for all the info. I suspect it is this 'control' issue that prevents Leica from offering the generic lens settings via menu that folks are asking for ad nauseum.

 

Interesting is that they apparently selected a slightly less strong IR cut filter than 486, this should result in less oblique reflections, possibly done to enable better color correction in corners of extreme wide-angle. I am sure we will hear eventually.

Tom

[sean_reid]

Thanks for all the info. I suspect it is this 'control' issue that prevents Leica from offering the generic lens settings via menu that folks are asking for ad nauseum.

 

Interesting is that they apparently selected a slightly less strong IR cut filter than 486, this should result in less oblique reflections, possibly done to enable better color correction in corners of extreme wide-angle. I am sure we will hear eventually.

Tom

 

Hi Tom,

My pleasure. I haven't seen any requests for generic lens settings by menu, which thread was that? I do support having menu options to manually select the same Leica lenses that can be read by codes.

 

Cheers,

 

Sean

[waterlenz]

Hi Tom,

My pleasure. I haven't seen any requests for generic lens settings by menu, which thread was that? I do support having menu options to manually select the same Leica lenses that can be read by codes.

 

Cheers,

 

Sean

 

As you may have surmised I am referring to the 'coding' issue, (mainly to enable one to use generic lenses, i.e., non-Leica) many folks seem to want a something ala Nikon D2 series, which I use too, often at the wrong FL and aperture. It is a good joke - I have exif data for 800mm lens shots that readout 35mm f/1.4, 28mm f/2, etc !!! No effect on camera function, pix look OK. Unlike Leica, I suspect the user set data does not affect lens corrections during raw processing - CA, vignetting (done by Nikon capture NX, and some earlier versions too, not in-camera).

Tom

[scott kirkpatrick]

Scott,

 

You did all this testing with an UNCODED Heliar, yes?

 

-g

 

Yup. Have to take an older-style LTM to M adapter off an old 19mm wide angle lens before I can code my CV15, which came with the modern Voigtlaender adapter, narrowed over the code sensors.

 

And the camera was set to "lens detection OFF" just to make sure. That submenu is in a confusing order. It goes ON, ON + UV/IR, OFF, with the strongest setting in the middle. It would make more sense to me if it read "OFF, ON, ON + UV/IR" so that as you go down the list you are turning on more features.

 

scott

[LJL]

LJ, what size did you get for the CV 15? In the little info packet which came with the lens, they advise the 15 cannot take a filter, though there are threads cut into the shade.

 

Sorry I missed this question in the thread....was working

 

For the CV 15/4.5 lens, I am using a Lecia 39mm UV/IR cut filter. Based on other posts, I use the o-ring technique for mounting. I picked up a 1-5/16" ID, 1-7/16" OD, 1/16" wall thickness o-ring at a hardware store....$0.50-0.60, I recall. I carefully stretched the o-ring over the threads on the Leica filter. It is a bit tight, but once on, it holds well. I then carefully pressed the Leica filter with the o-ring into the lens hood petals of the CV 15/4.5 lens. This is also a fairly snug fit, but with patience and evenly applied pressure it goes in. Once the o-ring hits the lower petals section, you need to ease it past where it tends to bulge out a bit. I used my fingernail as a wedge to help it in. Once in, the fit is snug, the filter is very close, but not touching the front lens element, and it will not dislodge. Very strong, clean and comptletely removable installation for the filter. Folks have use silicon gell, tape and paper wedges, but this o-ring method seems the best and most secure without doing any damage to the lens or filter.

 

Those really are not filter threads in the lens hood petals on the CV 15. I think they are ridges for both strength and light control from reflection, not threads.

 

I ordered my Lecia 39mm UV/IR filter from Tony Rose at Popflash Photo. He shipped it in a day or so. I think they still had some in stock last time I checked the other day.

 

Hope this helps.

 

LJ

[MP3]

Thanks for the info, Sean.

 

Real nice to know, since Leica filters are nowhere to be found, won't ship from Leica until God knows when, and cost 50% more than any other brand.

 

Did they really have to alter the original specs of the B+W IR-cut filter instead of just tweaking software to its specs? Give me a break.

 

I had to buy four filters (two B+W, two Heliopan) because Leica didn't ship their promised filters, and now they're telling me I should trash them?

 

Same feeling here, hammam! I got 4 B+W 486 already. Will Leica re-imburse us with more Leica IR-UV filters?

[jlm]

i'll post this in the 1.10 thread as well.

 

coded my 15mm Heliar with 39mm B+W 486 as the WATE; selected 16mm in the drop down mene, then ON+IR filter for lens selection. it works! (firmware 1.102, obviously)

 

when the replay image comes up, at first it has the cyan corners, then in a second or two, the correction kicks in and the display changes w/o cyan.

 

same thing on re-playing an image: it first comes up w/o correction, then after a second or two, the correction shows up.

[Guest guy_mancuso]

As mentioned several times by Sean and myself there are only 4 or so lenses that you want the leica filter for sure. The WATE 49mm with adapter, Zeiss 15mm although needs a test, Cv15mm is extremely clean with the Leica and very good with the B+w and the Zeiss 21 and Leica 21mm and those two are still being tested but we are leaning at the leica filter. The rest work fine with the B+W the only difference you MAY see is a slight color shift that would be global which is a easy fix and will be probably be perfect with a good WB. That's it I have been using the B+W for weeks with 1.102 and they work fine. So no need to trash anything. Just these lenses mentioned will probably need the leica filter and certainly the WATE which i tested a lot wih the new adapter John made and a 49mm filter. The only filter that is tough to get is the leica 55mm , the 49mm was on the shelves at a couple dealers and my bet is Monday there will be more in stores, since a container came in last week.

[Guest guy_mancuso]

Great John the CV15mm i have not personally tested but seen results, does anyone have both the B+W and leica to run a side by side on the CV 15mm . I know the leica filter works great the question was does the B+W clean up the cyan cast completely.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[LJL]

Great John the CV15mm i have not personally tested but seen results, does anyone have both the B+W and leica to run a side by side on the CV 15mm . I know the leica filter works great the question was does the B+W clean up the cyan cast completely.

 

Agree that it is great news. My only caveat is to look at the files on the computer, and not just the camera LCD. I do think that many may find both filters acceptable, and no doubt that the firmware does a good job at both vignetting and cyan contamination correction. My tests with only the Leica filter on the CV 15 are very clean, as you mentioned several times Guy, and have seen yourself. That other color balance shift you mention, Guy, is the only other thing I have been looking at also. Like you, I do think that it is easily corrected in processing, but for the cleanest straight from the camera shots (e.g., JPEGs for some), the Leica corrections using Leica filters is very good, and much, much better than a lot of what we were seeing earlier, I think. Great stuff. Thanks for all the continued efforts.

 

LJ

 

P.S. I do not think that things beyond 35mm are going to matter much at all, with respect to any differences in color among the various filters (Leica or B+W). On the widest stuff, Leica filters may provide a bit less intense cyan that cleans nicely with the new firmware. Just my thoughts and tests so far.

[Guest tummydoc]

I agree Sean the 21mm is the one we are just on the fence about and for users of the 21mm and a B+W please try and give some feedback on this lens so we know where we stand on the 21mm. On the WATE it seems okay either way at the 21mm setting but the Elamrit 2.8 we need to know about , I don't have one to test. Thanks

 

I made that test yesterday and reported on it on another thread. I haven't the Leica 55mm filter (nobody has TTBOMK) but I used a 486 and a Heliopan. My test was white wall, AWB, MWB and custom-WB, fluorescent (lowest IR) and incandescent (higest IR). What I found was there was neglectable difference between the filter brands, but in both cases there was very slight residual cyan in the corners under incandescent light and more noticeable red corners under fluorescent light. Assuming the Leica filter corrects more, and there will be no cyan in high-IR light, I would expect there might be possibly more red corners in low-IR situations. That's just a guess based on logic; however it is a proven fact to me, that the firmware is invaribaly set at an arbitrary level for each lens, and does not take into account different amounts of IR. That is to me a more serious issue than the cyan to begin with. Since I will need to use Panotools to remove the residual drift the firmware leaves behind, I'm going to skip the expense of coding (and the bother of self-coding my uncodable lenses), and also stick with Heliopan filters which I have found are the best coated to avoid scratches and the reflections Erwin and others have seen with the Leica filters.