Scanning negatives - grain problems

[m_liddell]

I'm pretty sure the negs are fine since I have done testing to arrive at the dev time and acros in pyrocat has always produced very fine grain. Temperature is always 21C.

Scans are RGB with green channel used. The imacon scans are much improved and I'm beginning to think the problem may lie in the printing side rather than the scanning. After downsampling for output, if there is even a *tiny* amount of visible grain viewed at 100% it seems to appear in the print. Tonality is also quite poor. Printer is an HP B9180.

Next I'm going to try sending some files to a decent lab to be printed on a lambda or similar and see if the same problems arise. Until then, my process will remain entirely analog.

 

[timeUnit]

A fine art inkjet print is generally quite a bit sharper than a lambda print, and that may require a different approach when sharpening for output. At least in my experience. If the prints you are making are grainy, try adding a tiny bit of gaussian blur before printing. Of course, that won't save a very grainy scan.

Personally I've never gotten "grain free" results from silver based processes, nor would I want to. The grain adds sharpness and a sort of bite to a print, and personally it's part of the beauty of the analog process. If you want a more grain free rendering, I would definitely advice you to go digital. The Canon 5D is a great camera, and I'm sure it wouldn't take long before it earned it's cost compared to shooting color neg. If ou can afford it, keep the M7 for BW. There's no substituting the look of BW film, imo.

 

[timeUnit]

Oh, and the tonality from my scans in HP B9180-prints is equal to what I see on screen after softproofing etc. If the neg sucks, so does my scan, and of course, my print.

 

[sanking]

I agree with Ted. Green channel is nearly always the one with the least grain. That is true with every scanner I have used, and with every type of color negative film I use. However, if you are scanning color negative film with the intention of converting to B&W bear in mind that discarding one or more channels to minimize grain will most likely result in a change of the relationship in tonal values in the scene. This may or may not be desirable.

My experience with B&W scanning is primarily with stained negatives. With this type of negative there is very little difference in grain between the three channels. This should not be surprising since the high density areas of a stained negative is made up primarily of stain, which is a dye density with no grain. However, there is a considerable difference in contrast between the red, green and blue channels.

Sandy King

 

[lenny]

An update on this.
I sent some film out to be scanned on an Imacon which was recommened to me for neg film over 9000 and drum scanner due to the light source. The results were much improved. The grain is much finer and sharper than the nikon and is handlelled more easily with neatimage.

This is pure crap. First of all, an Imacon is a CCD scanner. By definition it is not as sharp as any drum scanner. Whoever recommended the Imacon over a drum doesn't know what they are talking about. Period. Further, the drum, with its aperture settings, will have much more capacity to deal with grain.

Inkjet prints from the scans can still not go anywhere near as big as traditional silver prints without grain becoming a problem. I think I will stick to traditional silver printing for my b&w unless I shoot slide and convert it.

I make huge prints all the time, most of which are grainless. You are simply being served by people who don't know what they are talking about. Darkroom prints don't have a fraction of the range of the inkjet. There are far more possibilities... mine look very much like platinum prints, but have a longer tonal range.

The project this week is to make a 17 foot long banana, in black and white... I am going to photograph it with an 8x10, scan it at 8,000 dpi, and the resulting 80,000 pixels should yield a print at 392 pixels. Its for a local interior designer for some million dollar house. Don't ask me why or what for, I am getting paid nicely, which is what I care about. I expect a beautiful print....

 

[Bob Carnie]

Sandy
are you ever blending two channels together and dumping the bad channel??
for colour printing I am now using the channels to build up luminosity in an image before changing over to Lab Colour to continue working on a file, and sometimes the combination of red and blue give me better results than keeping the green by applying them to the image.
From what I am reading you are using the green mostly for black and white , but do you not worry about losing some important information one of the other channels might bring to the table?

Are you printing in color or B&W? If in B&W I would recommend that you scan in RGB and drop the red and blue layers. The green layer is almost always the one with the least noise. I have made a number of prints from ASA 160 color negative fmedium format film this way and grain has not be been an issue in prints up to 11X14 in size.

I would also recommend Bruce Fraser's Image Sharpening with Adobe Photoshop CS2. He recommends specific steps for noise reduction prior to other image manipulations.

Sandy King

 

[sanking]

Sandy
are you ever blending two channels together and dumping the bad channel??
for colour printing I am now using the channels to build up luminosity in an image before changing over to Lab Colour to continue working on a file, and sometimes the combination of red and blue give me better results than keeping the green by applying them to the image.
From what I am reading you are using the green mostly for black and white , but do you not worry about losing some important information one of the other channels might bring to the table?

Bob,

No, I don't usually dump any of the channels. Since I work primarily in LF and ULF grain is never a problem with B&W negatives so there is no downside to keeping all three of the RGB records. However, when working with scans of MF color negatives which I plan to print as B&W, I would consider dropping the blue if it did not negatively affect tonal values. 

Sandy

 

[Bob Carnie]

Hi Sandy

sometimes the blue has good info in it , I am doing this a lot for luminosity effects in colour and the blue sometimes gets blended in with the other two for good effect.

Bob,

No, I don't usually dump any of the channels. Since I work primarily in LF and ULF grain is never a problem with B&W negatives so there is no downside to keeping all three of the RGB records. However, when working with scans of MF color negatives which I plan to print as B&W, I would consider dropping the blue if it did not negatively affect tonal values. 

Sandy

 

[donbga]

I also experience this effect with scanning Colour neg film, not B+W nor E6. My solution is relatively cheap and effective. Neat Image software works a treat, also Noise Ninja although that was dearer - hence my choice. It will work wonders with B+W if you want grain free HP5 for instance.

Plenty of sliders to twiddle with, personally I have found that Auto works very well indeed. Worth saving profiles, dependant upon neg size,/resolution etc. CAn then reuse these, it gives assessment of matching before carrying out the noise reduction.

Michael Almond has a good comparison of noise reduction software:

http://www.michaelalmond.com/Articles/noise.htm

Interesting reading.

Don Bryant