M Monochrom, filters and Silver Efex Pro 2

[dante]

Advertisement (gone after registration)

Let me give you these insights as a person who uses b/w contrast filters on a very regular basis.

 

1. Recommended filter factors are not very helpful - because they are computed using "average" scenes (or, more likely, grey card). The entire purpose of filters is to selectively emphasize (or de-emphasize) the brightness of colored objects that may be small or big in your scene. TTL filters don't know what you are trying to do with a scene, and and Leica M's sensor is unable to tell how predominant colors are in the scene - and the silicon exposure meter and the CCD imager have their own independent color responses. None of this is a new issue. Take a look at "recommended" filter factors for film and then compare then to the recommendations for (a) Tri-X and ( T-Max. Big differences.

 

2. Yes, with a couple years of practice, you can develop an intuitive sense of what filter you want to use for a given situation. Which seems kind of pointless when a monkey with Channel Mixer does not need all eternity to return the results of Ansel Adams.

 

3. Changing filters on the fly is a great way to drop filters or hoods, cross-thread filter rings, and nick the front element on your lens. Unlike changing channel mixing at your desk under controlled circumstances, bracketing filtration (or even changing filters) in the field can be challenging when you are in unsafe places, you have just hiked to the top of a mountain, or you find yourself in any other circumstance where you can't set everything down on a flat, clean surface.

 

4. Some types of condition, like quick-changing variably cloudy landscapes, can require multiple changes in a short amount of time.

 

5. Just as with film, fiddling with filters on an MM will lose you some shots.

 

And, of course, if you are paranoid that UV or skylight filters will degrade the performance of your lens, you're not going to want to live with the permanent addition of filters for contrast control.

 

Filters really only exist to bridge the gap between our own perception and what our imaging materials (digital or analog) do. I don't think they have any independent charm.

 

In some ways, the MM is attractive to me - but in a lot of ways, a 28MP color bayer sensor (which would match the MM in resolving power, beat it in dynamic range, and at least equal it in speed) could be a better black and white platform. It would also cut out using two cameras.

 

Dante

[chris_tribble]

Dante - I find this illuminating and useful. Thanks.

[Advertisement]

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

[jaapv]

However, the 18Mp resolution of the MM, albeit with 24 Mp acuity, avoids some of the drawbacks that high-resolving sensors like the D800E have shown up i.e. demanding technique and insufficient lens resolution.

Just one remark on emulating filters in Photoshop by tweaking colour channels - it will increase noise.

Edited September 2, 2012 by jaapv

[atufte]

Leica Monochrom + Silver Efex Pro II - (no filter)

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!

[lct]

...Just one remark on emulating filters in Photoshop by tweaking colour channels - it will increase noise.

Less so than by increasing ISOs due to the loss of 2 or 3 stops caused by optical filters.

[brianv]

A yellow filter brings the Monochrom down 1 Stop to 160, the same base ISO as the M9. A RED filter- the QE of the M8 and M9 is ~17%, down two stops and you get a 4.5MPixel camera. Will be interesting to compare when my Monochrom comes in. Arrays with higher resolution means smaller pixels, usually means lower saturation count. I've read users of the D800 complain about blown highlights. Hard to get real numbers as most vendors do not publish the sensor data sheets, but I've seen some numbers that suggest the D800 sensor is 1/2 the saturation count of the KAF-18500. Truesense did not publish the KAF-18500 data sheet, but did so on detectors in the same 6.8um family that were produced at the same time.

Edited September 3, 2012 by brianv

[IkarusJohn]

Advertisement (gone after registration)

A yellow filter brings the Monochrom down 1 Stop to 160, the same base ISO as the M9. A RED filter- the QE of the M8 and M9 is ~17%, down two stops and you get a 4.5MPixel camera.

 

I don't understand this - why would reducing EV by 2 stops also reduce the pixel count from 18 MP to 4.5MP? Surely, it makes no difference, save that you either have to drop the shutter speed or open the aperture.

[AlanG]

Let me give you these insights as a person who uses b/w contrast filters on a very regular basis.

 

1. Recommended filter factors are not very helpful - because they are computed using "average" scenes (or, more likely, grey card). The entire purpose of filters is to selectively emphasize (or de-emphasize) the brightness of colored objects that may be small or big in your scene. TTL filters don't know what you are trying to do with a scene, and and Leica M's sensor is unable to tell how predominant colors are in the scene - and the silicon exposure meter and the CCD imager have their own independent color responses. None of this is a new issue. Take a look at "recommended" filter factors for film and then compare then to the recommendations for (a) Tri-X and ( T-Max. Big differences.

 

2. Yes, with a couple years of practice, you can develop an intuitive sense of what filter you want to use for a given situation. Which seems kind of pointless when a monkey with Channel Mixer does not need all eternity to return the results of Ansel Adams.

 

3. Changing filters on the fly is a great way to drop filters or hoods, cross-thread filter rings, and nick the front element on your lens. Unlike changing channel mixing at your desk under controlled circumstances, bracketing filtration (or even changing filters) in the field can be challenging when you are in unsafe places, you have just hiked to the top of a mountain, or you find yourself in any other circumstance where you can't set everything down on a flat, clean surface.

 

4. Some types of condition, like quick-changing variably cloudy landscapes, can require multiple changes in a short amount of time.

 

5. Just as with film, fiddling with filters on an MM will lose you some shots.

 

And, of course, if you are paranoid that UV or skylight filters will degrade the performance of your lens, you're not going to want to live with the permanent addition of filters for contrast control.

 

Filters really only exist to bridge the gap between our own perception and what our imaging materials (digital or analog) do. I don't think they have any independent charm.

 

In some ways, the MM is attractive to me - but in a lot of ways, a 28MP color bayer sensor (which would match the MM in resolving power, beat it in dynamic range, and at least equal it in speed) could be a better black and white platform. It would also cut out using two cameras.

 

Dante

 

All of this makes sense to me although I found filters pretty easy to use as long as I had the one I needed with me. But I'd like to point out that there are also other more selective ways to adjust the colors in a scene captured in RGB prior to b/w conversion than by using the color channel controls. This can be accomplished in the raw stage or on tiffs.

Edited September 3, 2012 by AlanG

[AlanG]

I don't understand this - why would reducing EV by 2 stops also reduce the pixel count from 18 MP to 4.5MP? Surely, it makes no difference, save that you either have to drop the shutter speed or open the aperture.

 

I presume he means that the red channel of the M9 is made up of only 4.5MP whereas a two stop reduction red filter on the MM will still have 18MP but will only be 80 ISO equivalent.

 

One would have to do a test on the MM to see if a red filter that only requires 2 stops of compensation really has the same effect as using just the red channel of the M9.

[AlanG]

Just one remark on emulating filters in Photoshop by tweaking colour channels - it will increase noise.

 

Maybe you would be darkening the color and suppressing any noise. (E.g. a typical blue sky.) And if you lighten a color via the raw conversion process there may be enough "headroom" that you won't see additional noise.

 

The pros and cons of all of this have been discussed many times and as far as I can see the MM will always be at a disadvantage to the MM except for times where the MM will produced more detail and less noise (better grain look?) in a given situation at a large enough size for this to matter to you. I don't think the answer is clear cut as to when this happens. I think this will only be answered if someone does a really comprehensive good series of tests... locked down and in typical handheld shooting.

 

One peculiar thing about b/w photography that muddies the water a bit is that grain (and grain-like noise) is often seen as an positive aesthetic attribute whereas this is often looked at as a defect in color.

Edited September 3, 2012 by AlanG

[IkarusJohn]

I presume he means that the red channel of the M9 is made up of only 4.5MP whereas a two stop reduction red filter on the MM will still have 18MP but will only be 80 ISO equivalent.

 

One would have to do a test on the MM to see if a red filter that only requires 2 stops of compensation really has the same effect as using just the red channel of the M9.

 

I was wondering about that, but surely the sensor will still simply record 18 MP of the image that it is exposed to. It will have two stops lower EV, and that image will have significantly reduced red bandwidth, but MP and resolution will not be reduced.

[AlanG]

"The pros and cons of all of this have been discussed many times and as far as I can see the MM will always be at a disadvantage to the MM" - I meant to write disadvantage to the M9

 

BTW typical red filters that most photographers use have a factor of 3 stops and it will probably require a filter that deep to get the same effect as using just the M9's red channel.

 

Also - sorry for so many grammatical errors but I hastily posted late at night and did not re-read after editing.

Edited September 3, 2012 by AlanG

[brianv]

I don't understand this - why would reducing EV by 2 stops also reduce the pixel count from 18 MP to 4.5MP? Surely, it makes no difference, save that you either have to drop the shutter speed or open the aperture.

 

If you want to emulate the use of a RED filter with an M9 conversion to monochrome, only 1/4 of the pixels are used: it becomes a 4.5MPixel camera. The Blue and Green pixels are excluded from the final image. You get an 18MPixel "interpolated" image made from 4.5million actual pixels.

[jaapv]

We can discuss this until we are blue in the face, but in the end the M9 (10?) will be chosen by those of us that like the color images and color channel tweaking of the bayer camera and the MM by those of us that err... prefer the MM, ie the look the sensor imparts to the images.including filters.

The real battle ought to be fought in the photoforums

[htimsdj]

This is not correct. All color information is used - red items are made a lighter shade of gray than they would otherwise and blue items a darker shade.

 

If you want to emulate the use of a RED filter with an M9 conversion to monochrome, only 1/4 of the pixels are used: it becomes a 4.5MPixel camera. The Blue and Green pixels are excluded from the final image. You get an 18MPixel "interpolated" image made from 4.5million actual pixels.

[AlanG]

What is not being considered here is that when one is using the M9 color image to make a conversion, one can use the entire 18MP of detail and only lighten or darken specific colors. I don't see why one would typically resort to the drastic effect of only using a single color channel before converting to b/w... that would be discarding an advantage of shooting with RGB in the first place. Looking at the M9 file as if it were a b/w file shot through a deep filter would be a mistake in my opinion.

Edited September 3, 2012 by AlanG

[Ivan Muller]

Judging by the questions re colour filters it would seem that many a MM will be purchased by camera collectors rather than 'photographers', sad but true....

 

But what about the aesthetics??? an M with a yellow/red/green/blue disk in front of the lens just doesn't look right does it..? unless of course Solms can get Hermes to make matching leather covers for ones favourite filter...just think of it, a beautiful black with red leather body to match the red filter in front...and just think of the aftermarket possibilities of easily interchangeable leather covers, shutter releases and grips to match the filters..the mind boggles...

[jaapv]

I hope you do not put me in that category. Although I tend to buy small fry collectibles from time to time I normally use my stuff....

Edited September 3, 2012 by jaapv

[c.poulton]

But what about the aesthetics??? an M with a yellow/red/green/blue disk in front of the lens just doesn't look right does it..?

 

Eh...?

 

Never was (is) a problem with the film Leica's - I think that my M2 looks gorgeous with her silver cron and yellow filter on!

 

 

Christian

[Advertisement]

Hallo,

Look here: Leica M

[brianv]

This is not correct. All color information is used - red items are made a lighter shade of gray than they would otherwise and blue items a darker shade.

 

 

You are simply wrong. If you take a file from the M9 and want to emulate the effects of a R60 filter, you do not use any of the green or blue channels. They do not have any response in that portion of the spectrum. If they are a shade of gray, and get used- then the image does not emulate a Red filter. The only pixels that are used are those under the Red portion of the mosaic filter. You get information from 4.5million pixels. With the M Monochrom, you get output from all 18MPixels.