B&W grain hell...

[boinkphoto]

<p>I have what I believe is a pretty good scanner outside of drum scanner, namely a DiMage 5400 II. When I scan slides, though having some telltale signs of being scanned in my eye, they are quite respectable.</p>

<p>However, when I scan B&W the grain is excessive to me. Far more than prints made in the (chemical) darkroom from the same images.</p>

<p>I talked to a friend in the digital business and he said this is a common problem with B&W scans and that even drum scanners have this. He said they even purposefully cause some blur to help fix this. Moreover he said the general technique is to actually print the film and then scan it - it yields a better scan than straight from negative.</p>

<p>The problem is printing first is a PIA (even though I still have my mothballed darkroom). I'd rather just find a way to make the scans work if possible. This is particularly true since I've have thousands I'd like to slowly bring into my portfolio as it were.</p>

<p>So, is there any way around this dilemma digitally speaking? Is there a magic bullet grain removal software?</p>

<p>Yes I know some exist, but do they really do the trick compared to say the print and scan technique above.</p>

<p>Just wondering - thanks in advance.</p>

[franklin_polk]

<p>Google "grain aliasing". Basically, the scanner makes the grain worse than it is on the film due to how the sensor works. A real drum scanner (not an Imacon, something that actually uses PMT tubes) should have far less of a problem with it as part of the problem is with the CCD sensor itself. On your scanner, try the grain dissolver function (it is integrated in ICE on the Minolta software, and independently controlled in Vuescan), as well as a noise reduction plugin. </p>

[janne_moren]

<p>Turn off _all_ sharpening. Most scanning software has it on by default for some reason. That seems to be the reason for excessive grain more often than not.</p>

 

[peterbcarter]

<p>If your scanner has infra red/ICE capabilities, you might want to check out Vuescan. It can use the infra red channel to do the BW conversion which gets rid of most of the grain. I discovered it by accident and have been loving it ever since.</p>

<p>Here are some examples I did while making some clip tests.....</p>

<p><a href="http://www.flickr.com/photos/peterbcarter/sets/72157621747994875/">http://www.flickr.com/photos/peterbcarter/sets/72157621747994875/</a></p>

 

[boinkphoto]

<p>A large thanks to all for your input here. I will give the Vuescan ICE controls, grain dissolver, and NR. I guess I can also try to drive it through Noise Ninja, though I'm not sure how well it works with B&W. I will also make sure that sharpening is disabled in the initial scan.</p>

<p>Thanks again!</p>

[boinkphoto]

<p>This actually looks like a pretty good solution as well now that I thought about NoiseNinja:</p>

<p>http://dougplummer.blogs.com/dispatches/2006/08/mimicking_the_l.html</p>

<p>Thanks again to all...</p>

[Craig_Cooper11664875449]

<p>Matt, although I have no idea about using the Minolta scanner, you should be able to scan B&W film without the grain becoming overbearing. I use Nikon scanners, and as others have mentioned, use Vuescan. Basically turn off all filters and corrections and just get a clean scan. ICE will not work at all with B&W emulsions so do ensure that is off. I scan a lot of B&W film and I can attest that any application of blur will destroy far more than correct; Ive tried NoiseNinja and that doesnt really work to remove grain either (not without destroying significant detail)</p>

<p>Also doesn't the Minolta have the Grain Dissolver http://www.scanhancer.com/index.php?art=22&men=3</p>

[boinkphoto]

<p>Actually I don't think it came with "Grain Dissolver" (I looked around to make sure I wasn't missing something), but rather Kodak's "GEM". I haven't tried it with B&W, and I suppose I should (kind of dumb that I haven't now that I think about it).</p>

<p>I'll see if, as you suggest, the cure is worse than the ailment. It is pretty ugly straight out of the box though.</p>

<p>Thanks!</p>

[jtk]

<p>Matt...I've used both Minolta 5400II and Nikon V, have not scanned B&W with the Minolta.</p>

<p>They're probably optically/electronically very similar... build quality/focus is different.</p>

<p><em>With Nikon V the trick is to a) use Vuescan per defaults or b) OR use Nikonscan, scanning using "slide" setting, then <strong>inverting in post processing</strong> (it's just one click). Both allow further grain reduction... you can do quite a lot of that without affecting sharpness.</em> <em><strong>Mild grain reduction moves the ultimate print to a condenser-enlarged look while holding more detail than an enlarger could.</strong></em><br /><strong><em></em></strong><br />Both Vuescan/default and this workaround eliminate the problem completely with Nikon.</p>

<p>I've scanned more B&W silver with it than color, and I've printed many to 12X18. Better than Durst enlarger with Leitz lens.</p>

<p>Minolta's scan software probably works like Nikon's (IMO it was as good as Nikon or Vuescan in color) ... try scanning the neg as if it was a slide, then invert in in Photoshop or Elements or whatever.</p>

<p>Don't buy anything until you've tried the inversion workaround.</p>

<p> </p>

[mattb1]

<p>I've wondered about trying fluid mounting for B&W films, I'd give that a try if you are really wanting to scan B&W. I'd think it would help with dust and scratches as well. But, I have no experience with it, only had been looking at it as an option.</p>

[mendel_leisk]

<p>The II version doesn't have the Grain Dissolver, and I believe also comes with a more directional light source, closer to the Nikon Coolscans.</p>

<p>FWIW, the Grain Dissolver (which the "I" version of the 5400 uses), is a small frosted plate that swings into the light path. It can be activated independantly (for scanning b/w without ICE), or if you select ICE through the OEM software you are forced to use it.</p>

<p>I posted some with/without GD examples here:</p>

<p><a href="http://www.photo.net/digital-darkroom-forum/006Uxf">http://www.photo.net/digital-darkroom-forum/006Uxf</a></p>

[jtk]

<p>fyi in my color neg tests the detail resolution of N V and KM II was visually identical, each at highest ppi (4000 and 5400). They looked equally good in other respects.</p>

<p>Minolta files were larger of same image at maximum setting, leading me to believe they both ran out of film/camera lens resolution at around 4000ppi.</p>

<p>There's no scratch/dust problem with B&W silver if one does even moderately proper processing and sleeving. That requires a last rinse using distilled or well-filtered water.</p>

<p>My glassine-sleeved 35mm silver from 60s/70s virtually never has dust/scratch problems nor does my recent polyester-sleeved silver...but the minilab work on color neg and c41 B&W I've suffered has virtually always revealed scratches later, when I've Nikon scanned it...one doesn't notice how bad minilab processing/handling characteristically is when one relies on minilab scans...the difference between minilab processing and dip/dunk processing is huge.</p>

[boinkphoto]

<p>Thank you all again for the responses. Photo.net really is a great resource and I appreciate that unlike some sites you don't get snarky responses for what might be considered dumb questions. I will try the inversion technique and look into the fluid mounts (though I doubt the DiMage could handle it not being a drum scanner - but still it's interesting).</p>

<p>I agree incidentally that 5400 DPI is beyond the resolution of the film, even Velvia 50, which is what I primarily use. I have no doubt 4000 DPI is more than sufficient. Also to note all of my B&W was hand tank/reel processing (the slides were dip and dunk). They were immediately sleeved in archival sleeves, but it's amazing how little it take to scratch/dust them up. Thank god for nose grease!</p>

<p>Thanks again to all!</p>

[mendel_leisk]

<blockquote>

<p>I agree incidentally that 5400 DPI is beyond the resolution of the film, even Velvia 50, which is what I primarily use</p>

</blockquote>

<p>When I was in the early days, scanning a mix of KC64 and Ektachrome with my 5400, I downsampled to 4000 dpi, thinking to save the drive space. And then I had a good long look at the results, and ended up starting again: saving the full 5400 dpi.</p>

<p>On most shots it's not a big difference, you're mostly rendering grain a bit better, but fine details do sometimes emerge between 4000 and 5400, in my experience.</p>

[boinkphoto]

<p>Thank you Mendel that's good to know. Frankly I'm struggling with the whole process with scanning because I'm trying to solve a few things:</p>

<ol>

<li>To give me access to my images for personal enjoyment and posting on the web.</li>

<li>For potentially selling stock.</li>

<li>Provide an archival storage point should my house burn or something.</li>

</ol>

<p>Doing full 16 bit 5400 DPI scans with some degree of oversampling is not only enormously space consuming but also time consuming. Thus I've kind broken it down to:</p>

<ol>

<li>I do roughly 10mp scans for personal/web use, then import into my Lightroom DAM setup. </li>

<li>I scan max everything for potential stock sale, clean, and then delete after conversion to JPEG.</li>

<li>For slides I really think are my best (which frankly aren't that good but...) I do max and actually save them.</li>

</ol>

<p>I'm not totally loving this workflow, but the realities of time and disk space don't offer much choice.</p>

<p>As far as B&W goes, well I'm trying to figure that out. I have far more of them, but mostly their candids of friends etc. For those that aren't, that's part of why I started this thread.</p>

<p>Thanks again Mendel!</p>

[mattb1]

<p>For fluid mounting I ordered the scan science kit, but have not used it yet. </p>

[jtk]

<p>In my experience with Astia (far more accurate/longer scale than Velvia, same resolution), 13X19 doesn't severely test the potential of 4000ppi. Incredibly fine detail with/without Ice4.</p>

<p>"They were immediately sleeved in archival sleeves, but it's amazing how little it take to scratch/dust them up. Thank god for nose grease!" </p>

<p>fwiw I think polypropylene sleeves (soft plastic) are inclined to scratch..I use polyester "side lock" type because the film doesn't have to travel over as much sleeve/dust as it does in polypropylene when inserted and removed. Also, I think dust tends to stick more to the softer poly. IMO the ancient glassines were better than polypropylene, though Savage and Agfa from the Sixties are now yellowing.</p>

<p>I was introduced to the magic of nose grease years ago, but it stopped being useful on my B&W after I had loaded a few thousand (!yes) Nikkor reels for E4 processing (30 rolls at a time). Practice makes perfect :-)</p>

<p>fwiw, I rely on Kodak Rapid Fix but don't use the hardener...laziness, mostly.</p>