B&W Negatives

Don wrote:

Photoshop: 6
Hardware: Nikon LS-50
Software: NikonScan 4.02

Taking a well deserved (!) break from Kodachromes I started playing with B&W negatives today (Ilford, if anyone is curious). I wonder what the panel thinks of the less "sensible" of the possible scanning permutations for B&W negatives:

1. Mono Neg/Grayscale – boooriiing… ;o)
2. Mono Neg/RGB
3. Color Neg/Grayscale
4. Color Neg/RGB
5. Color Pos/Grayscale
6. Color Pos/RGB

For example, scanning B&W negatives as Mono Neg/RGB has obvious advantages because some blemishes are much more easily corrected with access to 3 individual RGB channels, converting to grayscale only as the last step.

I think it stands to reason that scanning three channels, even of a B&W negative, might get more information than scanning only one channel. This is analogous to averaging several scans to attempt to get rid of noise. You can always convert to grayscale in PhotoShop. Whether the payback is worth it for scanning a large number of images is up to question, but I would certainly try this if I were attempting to get the most out of a single image.

I wonder if there are any similar advantages to some of the other permutations specifically in terms of, say, dynamic range or sharpness, etc.

For example, at first blush, taking an individual channel from an RGB scan seems sharper than a grayscale scan which, presumably, started life as a composite (all 3 LEDs on).

An example image would help illustrate what you’re saying. I suspect you may be making the almost universal error of mistaking noise for sharpness.

Another example is that fine scratches appear easier to spot if scanning as Positive and then inverting as the last step, etc.

Good point, scratches do probably show up easier as white on black.

Another fun thing to play with is duotones and tritones – these can give a very pleasant tint to an image, and if not overdone can be comparable to selenium and other toning operatons. Heck, Ansel Adams did it, so why not us? 🙂

You’re bringing up some interesting points that I haven’t heard mentioned before. I’d suggest posting also to comp.periphs.scanners, which specializes in the non Photoshop side of such discussions. —

Mike Russell
www.curvemeister.com
www.geigy.2y.net

"Don" wrote in message

Taking a well deserved (!) break from Kodachromes I started playing with B&W negatives today (Ilford, if anyone is curious). I wonder what the panel thinks of the less "sensible" of the possible scanning permutations for B&W negatives:

1. Mono Neg/Grayscale – boooriiing… ;o)

Why?

Bite the bullet and get CS then use 16-bit greyscale.

"jjs" …

"Don" wrote in message

Taking a well deserved (!) break from Kodachromes I started playing with B&W negatives today (Ilford, if anyone is curious). I wonder what the panel thinks of the less "sensible" of the possible scanning permutations for B&W negatives:

1. Mono Neg/Grayscale – boooriiing… ;o)

Why?

Bite the bullet and get CS then use 16-bit greyscale.

CS is not needed. I scan my B&W negs (also Ilford) in "16" bits* with version 5.5. The extra dynamic range is essential for normalising the tone range of the negative.

–Toby

* Photoshop’s "16"-bit mode is only 15 bits deep, even in CS.

"Toby Thain" wrote in message

CS is not needed. I scan my B&W negs (also Ilford) in "16" bits* with version 5.5. The extra dynamic range is essential for normalising the tone range of the negative.

So what’s the problem? Why is a monochrome scan "booooring" ? If your negatives have too much DMax, then it’s your problem. Expose/develop to accomodate. Want a pointer?

On Sat, 18 Sep 2004 10:43:12 GMT, "Mike Russell" wrote:

For example, scanning B&W negatives as Mono Neg/RGB has obvious advantages because some blemishes are much more easily corrected with access to 3 individual RGB channels, converting to grayscale only as the last step.

I think it stands to reason that scanning three channels, even of a B&W negative, might get more information than scanning only one channel. This is analogous to averaging several scans to attempt to get rid of noise. You can always convert to grayscale in PhotoShop. Whether the payback is worth it for scanning a large number of images is up to question, but I would certainly try this if I were attempting to get the most out of a single image.

I don’t really know what happens at the low level (may vary from scanner to scanner) but I’m fairly certain that the scanner uses all three LEDs to generate a white light source and then converts the acquired image to grayscale before passing it on. So in that sense I’m just postponing the gray conversion.

What I gain – and the main reason for scanning B&W negative as RGB – is that, for example, on poorly stored negatives there are sometimes, say, brown spots which are much more easily removed from an RGB image because they stand out so glaringly and are much more easily eliminated by going to individual color channels. On a grayscale image, that brown spot, is well… gray… but nevertheless quite noticeable as a blemish and yet comparatively more difficult to remove.

Another fun thing to play with is duotones and tritones – these can give a very pleasant tint to an image, and if not overdone can be comparable to selenium and other toning operatons. Heck, Ansel Adams did it, so why not us? 🙂

Oh, absolutely! 🙂 After converting the image to grayscale it looks… you guessed it… kind of gray… ;o) The sepia-like RGB looks much better and I’m seriously considering storing the final product (after all the touch-up and editing) as duotones.

You’re bringing up some interesting points that I haven’t heard mentioned before. I’d suggest posting also to comp.periphs.scanners, which specializes in the non Photoshop side of such discussions.

I did, but I thought I’d share it with you guys here as well.

Don.

On Sat, 18 Sep 2004 08:31:29 -0500, "jjs"
wrote:

1. Mono Neg/Grayscale – boooriiing… ;o)

Why?

That was tongue in cheek, hence the smiley…

But, as I just wrote to Mike (and hinted at in the original message) there is a very serious point here. For example, removing a brown spot from a B&W negative is much easier if scanning as RGB. The brown spot stands out much so more on an RGB image than on a grayscale image, where the brown spot is just as gray as the rest of the image, but nevertheless still very noticeable as a blemish.

Bite the bullet and get CS then use 16-bit greyscale.

PS 6 does 16-bit grayscale as well and scanning at maximum scanner bit depth goes without saying.

Don.

"Don" wrote in message

On Sat, 18 Sep 2004 08:31:29 -0500, "jjs"
wrote:

Bite the bullet and get CS then use 16-bit greyscale.

PS 6 does 16-bit grayscale as well and scanning at maximum scanner bit depth goes without saying.

But can V6 or V7 actually do anything with 16-bit greyscale? I have forgotten.

Back to an important point – just how does a scanner make a grey-scale image? I’d much prefer a scanner for monochrome that had no filters interfering with the optics. They can’t help performance.

Note that when you scan some modern B&W negatives you may get a blue or purple cast due to some of the anti-halation remaining on the film. I do not know if conventional scanners have the same mismatch of color filtering sensors, but in cameras the blue and red channels are most susceptible to noise – compared to the green channel.

"jjs" …

"Toby Thain" wrote in message

CS is not needed. I scan my B&W negs (also Ilford) in "16" bits* with version 5.5. The extra dynamic range is essential for normalising the tone range of the negative.

So what’s the problem? Why is a monochrome scan "booooring" ? If your negatives have too much DMax, then it’s your problem. Expose/develop to accomodate. Want a pointer?

I assume these questions are meant for the OP, they don’t relate to my posting. –Toby

"Toby Thain" wrote in message

"jjs" wrote in message
news:…

I assume these questions are meant for the OP, they don’t relate to my posting.

Note. Sorry for the misattribution, Toby.

On Sun, 19 Sep 2004 08:46:58 -0500, "jjs"
wrote:

PS 6 does 16-bit grayscale as well and scanning at maximum scanner bit depth goes without saying.

But can V6 or V7 actually do anything with 16-bit greyscale? I have forgotten.

You can’t do layers (although I found a workaround for some functions) but you can do levels, curves, clone, etc.

Back to an important point – just how does a scanner make a grey-scale image? I’d much prefer a scanner for monochrome that had no filters interfering with the optics. They can’t help performance.

In case of Nikons (which is what I have) the light source are 3 LEDs (RGB) and there is no white LED or need for filters because primary colors are generated by the LEDs directly. So the only way to nominally scan an image in "B&W" is to turn all three LEDs on (it actually happens in succession as each pixel is sampled with each LED on). So, in that sense it really shouldn’t make much difference where the grayscale conversion happens, in the scanner or in PS afterwards.

Don.

"Don" wrote in message

In case of Nikons (which is what I have) the light source are 3 LEDs (RGB) and there is no white LED or need for filters because primary colors are generated by the LEDs directly. […]

Excellent news. That’s the way it should be done. Thanks.

"jjs" …

"Don" wrote in message

On Sat, 18 Sep 2004 08:31:29 -0500, "jjs"
wrote:

Bite the bullet and get CS then use 16-bit greyscale.

PS 6 does 16-bit grayscale as well and scanning at maximum scanner bit depth goes without saying.

But can V6 or V7 actually do anything with 16-bit greyscale? I have forgotten.

Even versions 5 & 5.5 can do the most important adjustments in "16"-bit mode: Curves and Levels. That’s all I generally need to do to my neg scans (apart from Invert 🙂

Back to an important point – just how does a scanner make a grey-scale image? I’d much prefer a scanner for monochrome that had no filters interfering with the optics. They can’t help performance.

Use a drum scanner if you want the ultimate quality scan. Great for negs.

–Toby

Note that when you scan some modern B&W negatives you may get a blue or purple cast due to some of the anti-halation remaining on the film. I do not know if conventional scanners have the same mismatch of color filtering sensors, but in cameras the blue and red channels are most susceptible to noise – compared to the green channel.