UV/IR or not UV/IR for B&W with M8

[tofsla]

Advertisement (gone after registration)

I know its been asked 1000 times, but I can't find consensus. Do you use UV/IR filter when you shot B&W with M8? Thank you for reply

[Advertisement]

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

[simon_hughes]

I do, but then I'm shooting DNGs so I'm actually capturing in color and post-processing to B&W. So far I can't be bothered to remove the filters even if I know a particular shot or shots will end up being B&W.

[Shootist]

Since the only B&W setting in camera is for Jpg only I do the conversion in PS or LR after I take the shot. With that I'm not sure if the shot will look better B&W or color so I always use filters.

[gesper]

Since the only B&W setting in camera is for Jpg only I do the conversion in PS or LR after I take the shot. With that I'm not sure if the shot will look better B&W or color so I always use filters.

 

Same for me, except I use JFI profiles in C1 or Alien Skin Exposure in CS3.

[Paul_S]

A UV/IR filter impacts B&W images positively in at least two different ways (as it does color images BTW):

1. Sharpness: filter prevents IR radiation from projecting an out of focus image
   
2. Color translation: accurate color rendering in the sensor provides a better basis for B&W translation
   

Note that whether you use in-camera B&W jpeg or convert DNG to B&W, both use the same information from the sensor as their basis. It is this information that is effected by the filter.

[gmaurizio]

A UV/IR filter impacts B&W images positively in at least two different ways (as it does color images BTW):

1. Sharpness: filter prevents IR radiation from projecting an out of focus image
   
2. Color translation: accurate color rendering in the sensor provides a better basis for B&W translation
   

Note that whether you use in-camera B&W jpeg or convert DNG to B&W, both use the same information from the sensor as their basis. It is this information that is effected by the filter.

 

Paul is right on the mark. Sharpness and color fidelity are essential for an adequate gray conversion. Specially with really fast lenses, like the Nocti.

[sean_reid]

Advertisement (gone after registration)

There's no one answer to this. Using an IR-cut filter can reduce unsharpness in certain areas of the subject that are reflecting a lot of IR. But there's no reason at all that the color file a B&W conversion is based on need be "accurate". Working without an IR-cut filter can have interesting (positive) effects on tonality, lifting shadow values (in particular) in parts of the subject that reflect IR. And the slight unsharpness introduced, in certain passages, by the IR, can be desirable.

 

So, it all depends on what one wants the look of the BW file to be. I worked for several years as a professional B&W exhibition printer and its quite clear to me that there's certainly no *one best way* that a BW negative or print should look. The RAW file is, more or less, one's negative.

 

I suggest finding *your* answer to this empirically. Pick a subject that you often photograph and make pictures of it with and without the filters. Convert the RAW files and see what appeals to your eye.

 

Cheers,

[Guest Walt]

I'm basically in agreement with Sean, you have to look at the results you're getting. That said, I always use IR filters, for a number of reasons. I find that the results are much more consistent and predictable because you are not allowing non-visible light to be part of imaging-forming light. When IR is allowed to be a component of the imaging light, many odd things seem to happen, including the sharpness issues mentioned and also, in my opinion, blown hightlights from tonally light areas in the image that are also emitting a lot of IR. The IR in the image also changes the grayscale, usually unattractively in my opinion. The images from the M8 without filters (and without correction in the BW conversion) remind me of the grayscale I see from many direct Canon BW conversions--milky and flat, with shadow detail that I don't want. M8 BW JPEGS (the in-camera conversion) look like this too. I think this is the high red sensitivity of digital sensors as compared to BW film, and the IR sensitivity seems to exacerbate that.

 

The color blance problems (beyond those introduced by IR sensitivity) of the M8 are also significant in BW conversion. Each BW film has a characteristic chromatic curve (just as it has a "characteristic curve" or tonal curve). You can look at these curves on Kodak's website. Tri-X, as one example, is sensitive to blue, green and red in descending order. Its red sensitivity is quite low. Thus, I normally convert M8 images in ACR to black and white with a chromatic curve that depresses red, boosts green slightly and boosts blue significantly. This simply gives me a grayscale that I like and am familiar with. If the original DNG file color balance is "off," which is to say unnatural in color, I cannot use my standard chromatic curve for conversion (and standard conversion profiles like those from JFI or Alien Skin will not give expected results either). All BW conversion profiles start from an assumption of "normal" color balance in the DNG and if you don't have that, you are doing an experiment each time. For example, a tungsten scene shot at daylight settings, is very orange and my normal chromatic curve does not begin to correct this because it does not depress red enough. In this particular case, depressing the red adequately for my normal grayscale produces way too much noise and I am left with the choice of a lot of very coarse noise or a grayscale that over-represents red and is unattractive to me. So I avoid, as much as possible, this kind of DNG. Perhaps I should be doing a color correction on it before the BW conversion, but I haven't tried that yet. I look forward to Leica improving the auto color balance for BW conversion because I do not always have the time to manually adjust it.

 

As Sean says, the DNG file is one's negative, but the DNG file is a starting point with hugely more variability than a BW negative. I certainly never had to color correct a Tri-X negative before printing it.

 

Walt

[Jamie Roberts]

I'll just say that some folks that do exceptionally fine BW work with the M8 do NOT use filters, for precisely the reasons Sean mentions: better shadow tonality and increased red response.

 

BTW, in practical terms, I've never seen a printable focus or sharpness issue with the M8 attributable to IR that wasn't also attributable to something else wrong in the shot, with exception of the smudging I saw from Sean's christmas tree light demonstration. So be careful this holiday season with your BW pictures

[sean_reid]

Perhaps I should be doing a color correction on it before the BW conversion, but I haven't tried that yet.

 

Walt

 

Hi Walt,

 

I think that's a good idea whether or not one uses filters. Doing a quick grey sampling to set WB definitely gives one a more neutral starting point. Its what I do for the JFI BW conversions I do in C1.

 

Best,

 

Sean

[sean_reid]

I'll just say that some folks that do exceptionally fine BW work with the M8 do NOT use filters, for precisely the reasons Sean mentions: better shadow tonality and increased red response.

 

BTW, in practical terms, I've never seen a printable focus or sharpness issue with the M8 attributable to IR that wasn't also attributable to something else wrong in the shot, with exception of the smudging I saw from Sean's christmas tree light demonstration. So be careful this holiday season with your BW pictures

 

Hi Jamie,

 

As you know, I often do not use filters for BW work and your comments bring us right back to that idea about doing some empirical work. Photographers who work often in B&W would be wise to do experiments with and without IR-cut filters to see what pleases them.

 

BTW, I did, this summer, run into a subject where IR was causing a resolution drop in certain highlight areas, in particular. For that particular series of pictures, I put the filters on. But I like shooting BW with a naked lens.

 

Cheers,

 

Sean

[Photoskeptic]

I am not using the filters with my 35 cron ASPH and am getting absolutely gorgeous bw. I do use the filter with the CV 15, though.

[Val B]

Photoskeptic...I've sent you a private message. Check your mail...Thanks.

[tofsla]

Thank you everyone for reply. Main issue in using UV/IR filters are flare and ghosting i get about most of the time. I can't stand it. Practically every lens I have with UV/IR filter becomes flare monster! This why I am thinking to stop using it all together. And flare create much more unsharpnes then any IR light that gets to the sensor.

[sean_reid]

In that case, if you're working only in B&W, drop the filters.

 

Cheers,

 

Sean

[pklein]

Sean: At the risk of further betraying our general ages: "There are many stories before the Naked Lens. This is one of them."

 

So far I've been doing B&W with IR filters. Without them, I've seen unsharpness, overly pale faces with overly visible blood vessels, and noisy shadows. The best argument for not using a filter is the possibility of flare.

 

I use C1 with the JFI profiles some of the time. I also sometimes do a just a rough white balance and level adjustment in C1, then export to a 48-bit color TIFF and play with the conversion in Picture Window Pro (which I use instead of Photo-you-know-what). PWP has a really wonderful B&W conversion (Transformation-Color-Monochrome)where you can play with the filtration applied to each channel easily and see the results instantly.

 

For what it's worth, the green channel is the least noisy of the three by far. Blue is the noisiest, and red is still noisy. Unfortunately, IR light seems to affect red and blue channels the most, and green the least. So when IR boosts your shadows, it boosts the noise right along with the detail.

 

--Peter

[Jamie Roberts]

{snipped}For what it's worth, the green channel is the least noisy of the three by far. Blue is the noisiest, and red is still noisy. Unfortunately, IR light seems to affect red and blue channels the most, and green the least. So when IR boosts your shadows, it boosts the noise right along with the detail.

--Peter

 

Peter, usually with an M8, the red and green channels are equal for noise on a "normally exposed" shot; I haven't noticed the red with more on a good exposure.

 

I know in theory Bayer filtration means there is more information for the green channel and less for the other two, but in practice without filters the RED channel often has higher exposure relative to blue and green, which usually means less noise, not more.

 

However, at the margins of exposure, the blue channel in most digicams is often the noisiest (and the noisiest especially under tungsten, where it fights to colour balance), which is why a RAW workflow is so important for really great black and white (if you're going to channel blend, which if you mix RG & B you *are* doing).

[pklein]

Thanks, Jamie. I'll probably get more deeply into this soon. Bless me, for I have sinned: A Noctilux is out of my price range, so I bought a Canon 50/1.2, and will be using it to play in the dark, so I'll find some dark shadows to experiment with.

 

When I first got the M8, I did a few available light B&W conversions with a 35/1.4 ASPH, and found that yes, under tungsten, the red channel might have a higher exposure in bright spots than the green. But in the dark midtones and shadows it always seemed noisier than the green. And boosting those darker areas via no IR filter seemed to make them noisier.

 

Some of my B&W conversions have involved using no blue at all, or only up to 10%, plus 100% green and enough red to look more natural, but not enough to get ugly-noisy, especially in the shadows--usually 25-50%.

 

I guess doing this at the RAW stage would involve creating a profile, unless one can simulate it somehow with the white balance sliders.

 

Is it really better to do all this in RAW, rather than get the RAW reasonably right and export to a 16-bit TIFF for final mixdown? Some people say one way, some say the other, and they all have credentials. :-)

 

--Peter

[Advertisement]

Hallo,

Look here: Leica M

[neelin]

OFF TOPIC

 

Sean: At the risk of further betraying our general ages: "There are many stories before the Naked Lens. This is one of them."

--Peter

 

Speaking of which, a couple of the 6 million stories in the Naked City routinely had astounding cinematography for a 1960ish television production. Check some out on DVD for fabulous New York vista's/locals.

 

Robert

[Jamie Roberts]

{snipped}

I guess doing this at the RAW stage would involve creating a profile, unless one can simulate it somehow with the white balance sliders.

 

Is it really better to do all this in RAW, rather than get the RAW reasonably right and export to a 16-bit TIFF for final mixdown? Some people say one way, some say the other, and they all have credentials. :-)

 

--Peter

 

Peter--you will love the 50 1,2L BTW. It's a great lens, but very reminiscent of earlier R summilux design with massive vignetting wide open (well, you see it on a 5d, anyway!) and low contrast that sharpens up remarkably by f4. I like mine alot--it's a wonderful portrait lens, I think. But it's contrast and flare characteristics are more similar to the 85 1.2L and very different than the Nocti.

 

As far as BW from RAW goes, yes, the JFI profiles in C1 or some equivalent mixing in LR or something else will give you very nice images indeed. My personal BW workflow involves making a special colour conversion from RAW to a TIFF, then using AlienSkin Exposure 2 in photoshop, which I think is absolutely worth the money if you're interested in BW.

 

Having said that, there a million ways to make a good monochrome once you do the RAW-->TIFF thing. You don't want a JPEG because JPEG compression (in the blue channel particularly) actually usually increases the noise and prevents any blending.

 

So it's most important to me to limit lossy compression; that seems to give the best results if you're not doing the BW right in the RAW converter.