Pre-Conversion Processing for B&W

[gerryp123]

Any thoughts on how a color image should be processed before converting to B&W? I once thought that this processing should be to achieve the best color image possible, but I now see that consideration must be given to tones that are well separated in color, but merge when converted.

Your opinions on this?

[Donald]

Gerry - A discussion on this took place in respect of one of my images, in which I was invited to walk through what I do.

You'll find it from post #16 onwards here. It may or may not go some way to answering your question.

[James G]

Gerry,
I've seen Donald's thread and would definitely recommend you read it. I'm getting back into B&W myself (kind of stopped doing it when I went digital).
I picked up a reference to a book only a few days back: -
From Oz to Kansas: Almost Every Black and White Conversion Technique Known to Man: , by Vincent Versace

It was on this forum, but I've lost the original reference so must apologise for not properly crediting my source, anyway, I purchased it, and it is proving educational and quite a good read on B&W techniques etc.

cheers, James

[Manfred M]

As with all things in photography, there are no rules, only guidelines, and they are meant to be broken. The books that write that you should get things right in colour first do have a point. There are more tools in Photoshop to manipulate colour and of course once you have done the conversion to B&W you are throwing away pixels, so that is really the limit to the theoretical reason why this work flow makes sense. There is a way of working in Photoshop where I avoid throwing away the pixels. I work in Photoshop, never in Elements and rarely in Lightroom (and I never do B&W conversions in Lightroom).

I personally find I prefer early conversion to B&W; if using a RAW image, I will do some minor tweaking in ACR or one of the other RAW converters and will do an import sharpen and will then proceed directly to working in B&W, mostly because I find all of the manipulation of a colour image a waste of time, In B&W I am looking at tonal ranges and contrast, not colour. In colour work I will to get the colours right and the two often have you moving in opposite directions in terms of how I process the image.

My B&W conversions are always done using a B&W layer mask (and I do a lot of manipulation with the colour sliders to tweak the image). That way I have not "damaged" the colour information in any way and can still use all of the colour tools. I will use curves and the other adjustment layers to tweak the image. At that point I will either output the image or drop it into Nik Silver Efex Pro 2. At other times, I will jump directly into Silver Efex Pro 2 and work the image there.

I find that this way, I am only worried about shades of gray and much of the PP work is actually a lot easier.


I've recently posted a few images I handled using this workflow:


This is the original colour image (looks pretty terrible):








This is what the image looks like after I did the conversion that I described above:










This is an image where I dropped the colour image straight into Silver Efex Pro and then Photoshopped the output. This image has had really heavy duty retouching:







I find that my work flow works gives me the results I want regardless of what the "experts" say. I find that colour and B&W require a totally different approach, which I why I tend to turn my images into B&W as right away.

[gerryp123]

Thanks folks.
Some good information here.

[Mike Buckley]

I now see that consideration must be given to tones that are well separated in color, but merge when converted.

That's because different colors can have the same luminosity. Another way of saying it is that two different colors when displayed in so-called color can appear as the same tonality when displayed in grayscale.

One way to deal with that is to select part of the image and convert to grayscale using one color filter and to select the rest of the image and convert using a different color filter. Another way to deal with that is to simply select part of the image (perhaps a person's hat) and change the color so that once it is converted, it appears as you want.

The stuff that Manfred discusses about losing data depending on how one converts is very important and is discussed in the book that James mentioned. It's so important that it's in Chapter 1 of the book.

The theory about editing the color version before converting to grayscale has to do with having the full range of colors available before converting. If the image is lacking the full range of colors at the time of conversion, it will also lack the full range of grayscale.

[Mike Buckley]

I've lost the original reference so must apologise for not properly crediting my source

No apologies needed, James. I'm glad your purchase seems to be working out as well for you as it is for me. You might want to add a post to that thread.

[arronlee]

I am a newbie here and almost a green hand on image processing and conversion work. But I have to admit that choosing a fine tool is so important because only the one whose way of processing is simple and fast can make the work much more convenient to operate. Checking its free trial page according to its tutorials about how to to get the related image program is also very neccesary. I hope you good luck all the time.



Best regards,
Arron

[revi]

I am a newbie here and almost a green hand on image processing and conversion work. But I have to admit that choosing a fine tool is so important because only the one whose way of processing is simple and fast can make the work much more convenient to operate. Checking its free trial page according to its tutorials about how to to get the related image program is also very neccesary. I hope you good luck all the time.

Best regards,
Arron

Except that your link refers to an SDK (aka Software Development Kit), which is absolutely useless to anyone but a programmer, it's not a standalone program like photoshop.

I have reported this post and your post in the "TIFF, when to use" thread as spam (link to the same SDK site )

[Donald]

Thank you, Remco.

This account has now been closed. All previous posts by the member were promoting the same product line.

[Brev00]

I don't follow any one rule here. Once in a while, I jump right into a black and white conversion with little editing beyond the initial raw processing. I can use Silver Efex, BW Effects, or the conversion tool in Elements (gotta switch to 8 bits for that). Other times, I process the image all the way through for color and then convert. There have been times I prep an image in raw specifically for the conversion (adding more blacks than usual, for instance), but I am seldom clever enough for that. I enjoy black and white but not as a primary form of photographic expression. I don't think tones necessarily merge when converted. That depends on the color channels which can be adjusted for contrast or conversion or somewhere in between. Which is why I still do a lot of my b&w work in Elements: the color channel choices are straightforward and easy to adjust to get the right flavor. I can then take it into Silver Efex for further enhancements or not. There are so many choices I will occasionally convert a file two different ways in two different programs. And, like both!

[Mike Buckley]

Other times, I process the image all the way through for color and then convert.

Vincent Versace explains in his book, From Oz to Kansas, why taking the image all the way through the color-post processing before converting to monochrome is important if we want to make the most of an image.

[chauncey]

Let me preface this by saying that I am not a B@W guru and use no plug-ins, but...
do a lot of work with selections in Photoshop. I'm more than comfortable making numerous
selections in an image and converting each selection differently. Consider the times you may
want the yellows lighter/darker in one part and the opposite in another part...problem solved.

I don't limit myself to doing this only in B@W but, do so in a lot of what I cough out.