Red filter vs. Custom white balance for B&W photography

I've been told many times that red filters are no longer necessary to shoot B&W photography because my 5D Mark ii has an internal custom white balance feature where I can dial up my red channel's sensitivity (essentially duplicating the effect of a screw-in red filter).

I'm skeptical. Won't dialing up the red channel's sensitivity cause noise in that channel to also dial up? Aren't I still better off using a red filter on the lens?

I strongly suggest you take the image using the appropriate white balance for the conditions and then process it using Silver Efex Pro. It sometimes helps to create an HDR image before you do the conversion to monochrome. Here's a link to several I've done this way.

Think about it this way... Black and White film has certain sensitivity to each wavelength. You use filters to alter the amount of each wavelength that reaches the film.

You should be using an external editor to do the monochrome conversion (instead of using the in camera monochrome setting), so don't you want to provide the editor with as much digital data to work with? Also, by doIng it this way you can similulate all filters (light yellow through deep red, as well as blue and green), you never know what you're going to prefer during the conversion process...

Also, the program can "borrow" from other channels when trying to recover highlights and bring up shadows. If you only record on the red channel and blow the highlight, it's gone.

PFDarkside wrote in post #14653640
Also, the program can "borrow" from other channels when trying to recover highlights and bring up shadows. If you only record on the red channel and blow the highlight, it's gone.

I get your point about highlight borrowing from the green and blue channel, but here's the catch. The best B&W work I've seen (e.g., Nick Brandt), borders on infrared. I know many people don't think of infrared when shooting regular B&W, but shooting at the far end of the red frequency does give a superior result. And the only way I know how to seriously push the red is in camera. If I shot normally in camera and boosted the red in PS during the B&W conversion, I'd lose that infrared look. To get the infrared look (not the extreme look for that would be infrared B&W instead of "enhanced" B&W, and I'm going for "enhanced"), I need to pull out the minute amounts of red from the predominately green and blue scenes in the frame. Photoshop may be able stretch "borrow" from the B/G pixels but it can't pull it out to the extent that the infrared-enhanced B&W look requires.

Think of like this. If I wanted a true infrared look, nobody would recommend that I do it all in post. I would definitely need to use an Infrared filter (or convert my camera to infrared). The same logic applies when shooting "infrared enhanced" B&W. Same concept, the only difference is that I should use a deep red filter instead of an infrared filter. I don't want my B&W to be true to nature, I want them to pop with a touch of infrared without going too spooky (infrared portraits are spooky looking).

Typical "best practice" is to shoot raw and blend channels.

What a filter will do that cannot be done in post-processing is to allow only a narrow range of wavelengths to reach the sensor, instead of the relatively broad range that is able to pass through the filters on the sensor. A custom white balance is not going to help you a bit in this regard.

Whether using such a filter gives you a creative edge is something I can't even guess at.

Just as a bit of an aside, I faced a similar conundrum some years ago with some immunofluorescence microscopy images. Various bits of cells were tagged with dyes that fluoresced in different colours when stimulated by particular wavelengths of light.

That's all well and good if you use no more than three colours, red, green, and blue, and often a single exposure with all fluorophores lighting up is all you need as the separate signals can be extracted later since the colours emitted directly correspond to the sensitivities of the different camera photosite filters (by design of course). However, with quantum dot technology, just about any colour of fluorescence is now possible, and very useful. The problem is that with an RGB-based camera, if something is fluorescing yellow, for example, that cannot be distinguished from two colocated objects, one fluorescing red, and the other green, as both will be recorded the same way: some red and some green. Nothing can be done about this in post-processing.

The solution is narrow-band pass filters and multiple exposures. With a yellow filter, the light from the red and green signals simply won't pass, but the intermediate wavelength of yellow will, and then be recorded as a combination of red and green. The difference is that we know we had the filter on, so the signal must have been yellow, not red and green combined.

I assume you mean IR look to imply a very dark sky with bright clouds, something like this.

IMAGE LINK: http://www.flickr.com …s/noisejammer/5​101536596/

IMAGE LINK: http://www.flickr.com …s/noisejammer/7​475629748/

FYI, the Sigma SD1 camera is the first of its kind to have a removable IR blocking filter. one could remove it, then pop a mild IR passing filter on top the lens and shoot some nice B&W images. doing it this way will create information in the shot that photoshop is unable to pull out from the G/B channels, namely the infrared wavelength information. the only problem with this method is screwy focusing. if the sigma sd1 were a mirrorless camera, it would be an ideal B&W camera.