While I find that a warming filter can be nice, particularly with outdoor images and with certain film emulsions, I don't always like the extra warmth. Especially with extra saturated films like Ektachrome VS.
Moose is shooting pretty much all digital with D1's these days, so the warming filters he still uses are essentially unnecessary and may actually get in the way of workflow for the best digital image output.
Whatever the physical qualities of today's CCDs may be, the image is processed in the camera to deliver a nicely "Nikon" balanced image (at least in my CoolPix 990). On top of that, you can do a custom white balance to compensate for an unwanted color cast from ambient light. With the D1, most pros use various software to extract a usable image from the camera's native RAW image file. From what I've seen of these, there's a considerable difference in the color balance of extracted images from, say Bibble compared to other the results from extraction utilities. Alan will know much more about this than I.
Since every digital image should be tweaked for output and/or profiled for the destination colorspace, it makes more sense to me to do any warming electronically than by using a single glass filter for every image. For anyone but Moose, it's easier and cheaper to manage your filters without having to stack in a warming filter as your standard. Sure, there's a cheapo Hoya "Moose" Polarizer, but there's no Moose ND or ND gradient (at least not yet). By the same logic, it makes little sense for someone shooting digtitally to build a big collection of color effects filters since you can easily (and with much more control) create the same effects in Photoshop.
Bit crazy, this: the NEF to JPEG image came from a converter of my own...
I got interested in the whole business after reading Phil Askey's (http://www.dpreview.com) review of the D1, which went into quite a bit of detail about the RAW mode and compared the various converters. He showed up a flaw in QImage that I'd noticed on occasion: while it retains a lot of shadow detail, it may exaggerate the background noise from the CCD. http://www.dpreview.com/reviews/nikond1/page15.asp
I wondered whether I could do any better.
What I've got at the moment is a missing colour interpolation algorithm that does a good job of averaging out noise while retaining maximum detail: it's the output of that - literally what the CCD 'sees' - you're seeing in the previous post.
All I've got to contend with now (Ha, Ha) is the colour space conversion process: I'm currently struggling my way through the workings of CIE XYZ and Lab colour space converters, gamut mapping algorithms and so forth.
Just read his latest D1 update and it surprised me quite a bit: firstly that he's using the native JPEG mode which can have colour problems (it doesn't use any specific colour space, but output is very similar to NTSC) and secondly that the resolution is good enough for his customers - although I guess if you're Moose Peterson then you get to call the shots...
Wow, this has been an eye opener. Thanks for the discussion; I'm certainly learning a lot.
I revisited his site and see that he uses digital exclusively now so perhaps the use of the warming filter is a moot point at least that's what I'm getting from this discussion. (Please forgive me if I'm wrong, but I have another stack of door hardware submittals to wade throught - again.)