Changing RGB to Black and white

[Annibale G.]

I've scanned an image from a negative BW in Rgb . Which is the best method to convert it in BW in photoshop? Do I have to do this as first thing or after I use curves, levels and other stuff in photoshop?
Thanks
Annibale

[jacksparrow]

if the negative is B&W I think you just change the image mode from RGB to grayscale, no need to change levels or anything as part of this conversion process. I would say you don't even need to do it, since the image on the screen will be exactly the same. I'm not sure about that if there's printing involved though, since then calibration would be involved.

[strick]

First step: I'd save the RGB-file as a "masterfile". Correct its tonal values and contrast in general. Remove dust spots.

Second step: My favourite method is to convert a copy of the masterfile into the grayscale mode and then to use curves, levels and so on.

There are people who like more to use the channel mixer ("Kanalmixer" in german - I don't know if this is the right translation). For this method you need a RGB file.

Using desaturation to get b/w-files results in my opinion in weak outcomes.

Cheers,
Greg

[DTD]

For the best control the Channel Mixer method is the best way.
Do it as an adjustment layer, then you can come back and still work on the colour image. It's good practice to always work in ways that do not 'destroy pixels'.

Go Layer>Adjustment Layer>Channel Mixer. Tick the monochrome box, then adjust the sliders. As a rule of thumb the totals should add up to 100.


Apply curves and levels etc. after this as you need to judge how things look in monochrome.

[abrewer]

Okay, you've scanned from B&W in RGB mode, and you want to convert to B&W?

I'm not sure I follow your question completely as you don't have color information to begin with, unless the scanning has introduced some kind of cast into the image.

My preference, when working with color negs, is to convert first. I really like the channel mixer as pointed out above. However, given that you're already B&W from the scan, I'd just make a Hue/Saturation adjustment layer and shove the Saturation slider full left, then back it off just a bit while adjusting the Hue to tint the image slightly to my preference.

I usually do the Curves adjustment after the conversion so I can evaluate highlights and shadows as they will appear in the final product. One trick I've learned is to set your preferences in Curves to slightly less than the maximum and minimum. That is, set your preferences for max highlight to 245 (from 255 default) and shadows to 10 (from 0). It prevents blowing the highlights and plugging up the shadows when you use the eyedropper to set the black and white points in the image. Then all you have to do to vary exposure is simply move the middle part of the curve back and forth, up and down, to get the tones like you want them.

My workflow is probably different from yours in that I sharpen first thing out of the scanner, then do retouching etc. at the end. The reasoning is I don't want to sharpen my corrections and make them any more obvious than they already are. If I do any sharpening at the end at all it is minimal, and to highlight lips/eyes etc or chrome parts on a car...the things I want to really leap out and bring the image to life.

Hope this helps.

Thanks.

Allan

[DTD]

True you don't have colour information – but you do have three channels of greyscale, each of which is different.

Remember computers are greyscale, it's only output devices such as monitors that are colour.

Found your tips about curves useful.

[strick]

Originally Posted by DTD

True you don't have colour information – but you do have three channels of greyscale, each of which is different.

In other words, scanning b/w-photos in RGB mode results in more information / better files. Just try it.

Greg