Monochrome Spectral Response

[Stephen]

Hello Everyone,

I have a basic question that I can't seem to find an answer to.

I own a Nikon D300 and it has several Picture Controls such as Standard and Landscape. Each of these color Picture Controls has a specific spectral response and tone curve.

Does the Monochrome Picture Control have a specific spectral response like the color picture controls? Does it have its own particular look like TRI-x does?

Thanks, Stephen

[Dave Humphries]

Hi Stephen,

I am guessing here, but I suspect the spectral response will be neutral, but determined in a large part by the colour temperature you have set (or the camera sets if on Auto-WB).

If you shoot RAW, as you really should without a reason not to, then it is academic because the RAW file will contain the colour information and only your Nikon PP program will be re-applying the monochrome selection you set on the camera. I expect, if you open the RAW in a third party editor, it would be in colour.

I'm sure there's someone here who can be a bit more authoratative than I.

(btw - after replying here, I found the exact same question in an old thread, nwhich you posted before this new (and better titled) question, so I have removed that other one to keep all the answers in one place)

This is a good question - thinking a little more upon it, I might surmise that Tri-X obviously has a certain spectral response - i.e. ignoring contrast ranges and gamma effects, for a given amount of noon sunlight, the Tri-X emulsion must produce a certain set of levels when viewing a colourchecker chart. This is presumably, at a specific kelvin figure for WB, reproducible on your D300. Dunno what that answer is though

Welcome to the CiC forums from ...

[Stephen]

I know that if I shoot bw in raw then open it up on cs5 it will be a color image. Since the color images are not neutral I can't understand why a bw image would be. This issue has always vexed me.

[PaulMiller]

If I shoot in B&W, why is the raw file still in colour?
The very short answer to this is that for most digital cameras, the raw data from the sensor always represents a colour image. Each pixel in the raw file represents the brightness of a point in the image viewed through either a red, green or blue filter. The filters are fixed to the sensor chip and cannot be removed.

A Black and White image is produced by combining a mixture of the R,G,B channels. By changing the relative proportions of R,G,B in the result you can mimic the effects of various colour filters used in black and white film photography. The combination can either happen in the camera (when you shoot in JPEG format), or in software on the computer when you shoot in raw.

It is much more flexible to shoot in raw and do the conversion on the computer afterwards, because you can adjust the response to suit the image.


Does the Monochrome Picture Control have a specific spectral response like the color picture controls?
Yes. It is possible to change the spectral response by changing the colour settings of the picture control to simulate the effect of e.g. an orange filter.

Does it have its own particular look like TRI-x does?
Yes, and the different settings give different looks to the result. The main difference in look between digital B&W images and film is that the digital image will lack grain - they look much too clean compared to film. Specialised software exists that can mimic the appearance of grain to give a more film like appearance to black and white conversions.

[Dave Humphries]

Since the color images are not neutral, I can't understand why a bw image would be.

Define (to yourself) your meaning of "not neutral" then think about what that means for a B&W image - how would you know?

~

A digital B&W image isn't necessarily neutral, it must be dependent upon the WB being used - which determines how much output, for a given colour subject, the R, G and B sensor produces (in a jpg file).

Both WB and iso are preset (at manufacture) with Tri-X, or any other film - unless you 'monkey around' with the processing (e.g. to 'push' it to pretend to be a higher iso). Since you have no control, you take it for granted, but as I recall, the spectral response was usually on a little graph on the bit of paper in the box.

So there's no means of comparing film and digital beyond the 'back to basics' idea I outlined above; use a known colour temperature light source, a known spectral response subject, then vary the digital parameters to get the same shades of grey for each given colour square as Tri-X produces - then you'll have what you want.

Perhaps a simpler way is to say it is; shoot a grey card, use that as a custom WB in camera, then take your B&W images with that same colour temperature setting - so now a monochrome shot of the grey card, even in a mono in-camera jpg must be neutral, yes?
But if you change the colour temperature of the light source, put something brightly coloured near the grey card that reflects some light onto it - then your 'neutral' goes out the window, but in mono you just don't notice it. In colour, the eye+brain is sensitive to it if it knows what colour something should be.

We can certainly see that a colour image taken in mixed lighting that doesn't work, e.g. inside a cathedral with sodium + daylight + candle light, can look perfectly fine in monochrome -leading one to a (technically false) view that mono is neutral.

Hope that helps,

[Stephen]

Thinking out loud: I always thought that camera makers would have to apply a specific spectral response to the monochrome setting, just like they do for the standard or vivid color settings (ignoring other things like wb and filters). It seem they would have to do this to get any kind of a picture since the raw data coming off the sensor must be manipulated in some way to get an image, either processed into a jpg in-camera or tagged as in a raw file for processing later. So, it may be fair to say that Canon, Nikon, Pentax, etc each has its own black and white "look" just like different types of film. If what I'm guessing at is true then the camera makers might want to play up their bw "look" for marketing sake.

You know, back in the film days one could get the spectral response from the film maker for either bw or color film. Why are digital camera makers so secretive?

Steve

ps This is a great web site!

[revi]

(...snip...)
You know, back in the film days one could get the spectral response from the film maker for either bw or color film. Why are digital camera makers so secretive?

Perhaps because there is no specific spectral response for B/W (compared to colour), the camera will have a spectral response, which then can be adapted (up to a point) by the post-processing. And because the public at large isn't interested all that much by such technical points (or by B/W in itself).

The few that might be interested mostly shoot in RAW anyway, and thus (can) ignore the factory settings for B/W and determine their own conversion (channel mixer). A bit like in the later days of film: most camera owners used colour film, B/W was already something for the more serious amateurs and professionals.
I'm heavily generalising here, ofc.

Also, processing from RAW not only gives you the possibility to determine the spectral response, but also what tinting (if any) you want to apply.

Remco

[Dave Humphries]

Hi Steve,

OK, ya got me biting

Looked through my D5000 manual and, as you say, it isn't mentioned.

So I just took four images of my garage door (dirty white) and green wheelie bin.

But first I determined a suitable (ish) exposure for the prevailing light, which is dull overcast (and raining); 1/90s f/5.6 iso 800, so I set the camera on manual exposure so these didn't change. I also set the camera up to take RAW + Fine jpg.

Image 1, normal colour image with white balance set to Cloudy
Image 2, Monochrome image with white balance set to Cloudy
Image 3, normal colour image with white balance set to Tungsten
Image 4, Monochrome image with white balance set to Tungsten

Comparing the two monochrome images, the second, with the incorrect (red biassed) Tungsten WB is noticeably darker than the more correctly WB shot. But if I didn't know the cause, I could just assume it was underexposed and open up a bit more.

Looking at the the histogram of the jpgs in Elements 8;
The brightest white is about 231 vs 202 (of the 256 levels) 29 apart.
The wheelie bin peaks are at 116 and 85 respectively, 31 apart.

In colour, the garage door looks very blue in the Tungsten WB shot, as you'd expect.

So basically I think this confirms my initial suppositions, in monochrome the spectral response is dependent upon the WB set. If I had a proper gretag colour checker, we could see whether the reletive brightness ratios of a veritable rainbow of different colours, varies with colour temp set, but I don't - so we'll have to make do

You know, back in the film days one could get the spectral response from the film maker for either bw or color film.

That's because it was set during in manufacture and only they knew it.

Why are digital camera makers so secretive?

It is no secret, it just isn't relevant because it varies so much based on the WB set - an option not available in film days.

Ain't dig-it-all wonderful?

At the risk of you asking another difficult question it's a bit like saying "why don't camera manufacturers say what speed the camera is?" - they sorta do; but it can be set over such a wide range that compared to Tri-X at 400 ASA (if memory serves me) it isn't so relevant in the digital age.
As to whether Nikon's 200 iso = Canon's 200 iso, well, I don't care, they're close enough for me

[Stephen]

Digital is indeed wonderful! Sometimes I shoot in-camera bw (I kinda like the look...which got me thinking about the question) and sometimes I use Silver Efex. It's good to have options!

Steve

[rob marshall]

Digital is indeed wonderful! Sometimes I shoot in-camera bw (I kinda like the look...which got me thinking about the question) and sometimes I use Silver Efex. It's good to have options!

Steve

If I shoot for BW I always switch the picture mode to Mono and boost the contrast in the settings for that style. The RAWs come out in colour anyway, but the in-camera generated JPEG shows quite clearly in the viewscreen what it might look like. It's impossible to tell if you use a colour picture mode.