Yellow Filter On An M9, Good Idea?

[Lenshacker]

Advertisement (gone after registration)

I bought the Monochrom, and use a set of filters for the effects that I want. It's the second monochrome digital camera that I've bought.

 

I've suggested that the OP try both methods himself and draw his own conclusions. There are good points and bad points for each method. Oddly enough, Silver Efex 2 came with the Monochrom. I've compared converted images to images generated with the Monochrom. No regrets buying the Monochrom.

[Advertisement]

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

[AlanG]

I bought the Monochrom, and use a set of filters for the effects that I want. It's the second monochrome digital camera that I've bought.

 

I've suggested that the OP try both methods himself and draw his own conclusions. There are good points and bad points for each method. Oddly enough, Silver Efex 2 came with the Monochrom. I've compared converted images to images generated with the Monochrom. No regrets buying the Monochrom.

 

I'm glad you are happy with this approach but I don't see how the MM applies to this discussion.

 

If there are "good points" to using filters over an M9 or other color sensor for b/w photos vs. post processing, do you have examples you can post that illustrate this?

 

As far as I can see, all a filter can do is reduce the amount of information captured in one or more color channels. It seems to me it would always make sense and be easier and more flexible to capture as much data and then later remove whatever you want to change.

 

Remember that we can adjust each color individually in post... we are not limited to simply lowering part of one or more color curves as a filter does. You can select by color and mask off certain regions and work on parts of an image if you like. E.g. darken sky without darkening other blue areas. If you use a red filter over the camera to darken a blue sky, red lips will be lighter as will anything else red. But if you just darken blue in post, red will remain unaffected. There are countless examples like this.

 

BTW I just remembered that I used C1 not DXO in my posted examples.

[sblitz]

Alan -- you are absolutely correct. it is, however, fun to put a filter on the M9 (yellow, orange, blue, even IR) and shoot bw jpegs. The camera brings you high class output, some of which is difficult to replicate in software. Aside from doing this for kicks, you are correct -- shoot color, convert in pp, and adjust color curves to taste.

 

Steve

[jaapv]

Compensating exposure for the filter in front of the lens (filter factor) won't make the non-filtered colors lighter as those have to be exposed correctly to not clip the highlights.

Why not clip in one or two channels if you are going B&W anyway? The remaining channel(s) will still retain the detail.

[AlanG]

Why not clip in one or two channels if you are going B&W anyway? The remaining channel(s) will still retain the detail.

 

Maybe they won't have the detail you need.

 

Just think for a minute. If you are shooting b/w film, a strong red filter will block gteen and blue light. If you place a yellow filter in front of the red filter, you will just be reducing the overall expisure. You will not be reducing the blue light because the red filter will cut it out.

 

What I'm trying to make clear is that converting to b/w in post gives photographers options they never had with film. Why not learn how to use these rather than using the very crude method of colored filters and limiting what you can do?

[jaapv]

I don’t agree. You can do the same corrections you want to do but the only difference is that the colour channels are balanced differently, reducing the amount of manipulation needed per channel, reducing the degradation caused by the shifting done in postprocessing.

So the filter will not be limiting, but facilitating the postprocessing.

[AlanG]

Advertisement (gone after registration)

I don’t agree. You can do the same corrections you want to do but the only difference is that the colour channels are balanced differently, reducing the amount of manipulation needed per channel, reducing the degradation caused by the shifting done in postprocessing.

So the filter will not be limiting, but facilitating the postprocessing.

 

I'm frustrated that I keep explaining that I work on individual colors not by manipulating the color chanels to simulate the constrained effects of a filter.

 

And in your case you are not balancing the color channels. You are just reducing what is recorded in one or two channels. Try to keep in mind that a filter just blocks light. You can't add light to the unblocked color channel without overexposing highlights.

[jaapv]

:confused:In an RGB environment how can you work on individual colours without manipulating the colour channels?

I fear we will have to agree to disagree on this.

[Lenshacker]

Color filters have finer spectral resolution than the color mosaic filter of the sensor. If you know in advance that you are going to convert the output to the equivalent of a Y52 filter, best to use a Y52 filter. Half of the green response of the sensor is below 520nm. With a Y48 filter- about 20% of the Blue channel response is from light above 480nm. The 90% cutoff for Blue is at 540nm, which is above the Y52.

 

You can make the RGB channels lighter and darker relative to each other, but once the light is collected you cannot "filter out" specific wavelengths. There is too much overlap.

[pgk]

Try it! I have. Using colour filters doesn't work for me - personally I can get 'better' results from converting my M9 (and M8) colour files to B&W rather than using filters. But that said and technicalities aside, whatever you are more comfortable with has to be more effective for you. Absolute perfection does not exist and there are always different routes to achieve a similar result. So if you are happier using a yellow filter and converting to a preconceived B&W result then its a good idea. If you aren't, its not!

[AlanG]

Color filters have finer spectral resolution than the color mosaic filter of the sensor. If you know in advance that you are going to convert the output to the equivalent of a Y52 filter, best to use a Y52 filter. Half of the green response of the sensor is below 520nm. With a Y48 filter- about 20% of the Blue channel response is from light above 480nm. The 90% cutoff for Blue is at 540nm, which is above the Y52.

 

You can make the RGB channels lighter and darker relative to each other, but once the light is collected you cannot "filter out" specific wavelengths. There is too much overlap.

 

We are not doing a science project here and why do you and Jaapv keep looking at it as trying to cut out a range of colors before the image is captured? Trying to simulate a filter seems really stupid to me. Filters were only used in the first place as a crude approach because nothing better was available with film.

 

In a color photograph we can select any color to make it darker before converting to b/w. How that color was formed in the first place is not relevant and why you would need to filter out specific wavelengths for any practical use is lost on me. (Beyond IR photography and the like.)

 

I keep asking for you to present examples that show why using a filter is a superior approach and nobody has shown any. Since some believe this is so I assume some have shot extensive tests. I have posted examples demonstrating how to use pp and can show more if you like. Photography is about making images not theoretical discussions. So if a filter is so useful why not spend some time proving that the filter can produce results that can't be achieved in post?

[Lenshacker]

If you want to know the difference between post-processing a color image versus using a cut-off filter, as to which gives the best results- there is a science to it.

 

If you have no interest in understanding what your post-processing is doing to the image, then move on.

[A miller]

I'm not sure I see the utility of a yellow filter on a digital camera. I do, however, think the a skylight filter could make sense for a classic look

[AlanG]

If you want to know the difference between post-processing a color image versus using a cut-off filter, as to which gives the best results- there is a science to it.

 

If you have no interest in understanding what your post-processing is doing to the image, then move on.

 

This is a visual medium so it should be easy to demonstrate under what situations one approach or another might be better. I have demonstrated a few and you have demonstrated none as far as I am aware. You only speak in hypothetical terms that do not necessary apply to any useful "real world" applications. Yet you ignore the samples I have shown that demonstrate actual variations that could never be accomplished via a filter. I am not trying to get a science lesson from you on this. I already have had many.

 

I would never tell anyone to move on but perhaps you should consider thinking of applying your approach and showing it rather than telling me to get lost if I disagree with you. Why not show the OP examples of why using a yellow filter on his M9 is a good idea?

[jaapv]

I would suggest that everybody chooses his own approach and does his own experiments. I would use a filter on my colour camera, or rather I wouldn't as I would grab the Monochrom , for me a conversion is an escape hatch.

[AlanG]

I would suggest that everybody chooses his own approach and does his own experiments. I would use a filter on my colour camera, or rather I wouldn't as I would grab the Monochrom , for me a conversion is an escape hatch.

 

So you haven't done any that you can share with us?

[250swb]

I can see the point of using a yellow or other colour filter on the camera.

 

Adjusting the colour channels in post processing starts to stretch or compress the information available, so if you want a really dark sky, as in replicating a dark red filter, you are starting to make that colour channel scream, the next thing that happens is posterisation where the transition of tones breaks down completely. But do it with a proper red filter and the information is already adjusted closer to the tones in the sky you want, and any further adjustments in contrast etc have not already been compromised by stretching the red channel every which way. It is a mistake to think you lose information by using a filter on the lens, you get the same amount of information, but arranged in a way of your liking. Stretching the channels in post processing doesn't then add information to fill the gaps created it can only work with what you started with.

 

Steve

[jaapv]

So you haven't done any that you can share with us?

Of course I have for myself- and I have colour experiments with a yellow filter to check the effects on colour balance and the relationship between the channels, years ago - to my satisfaction. No need to repeat nor bothered to search for them.

As I said, agree to disagree. Bless

[Advertisement]

Hallo,

Look here: Leica M

[AlanG]

OK I shot some quick tests.

 

With a red filter (Nikon R60,) there was only detail in the red channel... the others were totally blank. This led to quite a loss of detail as shown in the last photo. What do you expect? When using a red filter you are only using 1/4 of the pixels. It was on a heavy tripod.

 

How do you like them apples?

 

Regarding tones. I used the color to b/w sliders in C1 to darken blue and cyan to just above where noise started to show.

 

On the red filter image, there of course was nothing to adjust, but I tweaked the contrast and tones a little to try to get the sky as dark as in the converted image.

 

I was preparing to do some more tests using color charts, but after doing these I don't see the point. A filter requires an increase in exposure, reduces the number of pixels used in your image, puts another layer of glass in front of the lens, costs money, and adds time and inconvenience.

 

In the end you don't get as strong an effect (you can use a weaker effect of course) or even a fraction of the control you get from working with a color image in post.

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!

[AlanG]

:confused:In an RGB environment how can you work on individual colours without manipulating the colour channels?

I fear we will have to agree to disagree on this.

 

I guess that is something for you to learn. There is a big difference between what C-1 does via color sliders to b/w conversion than what is accomplished by gross changes to color channels. Likewise you could take any image into almost any pp software, select blue sky, darken it, and then make a b/w conversion... all without altering the color channel... just that specific color or color selection region.