Scanner + printer = interference pattern?

[ola_tuvesson]

<p>I have been lurking here in the forums for some time and have found them to be one of the best resources for technical info, tips & tricks and answers to a wide range of questions. It so happens I have a few of my own which I have been unable to find an answer to, here or elsewhere, and now that I've finally plucked up the courage to expose my ignorance I figured this would be the best place to post. <br /><br />For a number of years I've been using a Nikon D200 and a couple of lovely Nikkors (20-35mm f2.8 AF-D and 85mm f1.4 AF-D) but I've always longed (perhaps irrationally) to go back to film, black and white in particular. I was lucky to get hold of an FM3a body a couple of months ago and have been shooting a few rolls of Ilford Delta 100 (rated 80 ISO) with it and the aforementioned Nikkors. While I still shoot mainly digital I have really enjoyed the more focused style and manual steps required to produce images from film - and I have been very pleased with the results too!<br>

Now I don't have a darkroom, instead I'm trying to follow a hybrid workflow with scanning and digital printing which which seems to have become fairly popular of late. For this I use a Pacific Image PrimeFilm 3650 Pro scanner (lets me scan uncut rolls which is great) and an Epson Stylus Pro 5500 printer from which I've had some surprisingly good results, particularly using the "black only" method. After spending some time setting up the scanner the ~17Mp 16-bit TIFFs it produces need very little (if any) adjustments and print beautifully up to 13x19" (which is the max size for my printer), provided the neg is well exposed. Life couldn't be better. <br /><br />But there's a fly in my ointment and it's driving me nuts. Depending on the size I print at I get varying density banding pattern which must be caused by an interpolation error between the scanner and the printer. No banding whatsoever is present in the scanned neg, even if I hunt for it with extreme curves adjustments - and importantly it changes frequency non-linearly with the size of the print. I have tried printing at 1440dpi but I just get a different frequency of bands. I suspect that the grain in the film as captured by the scanner causes an interference pattern with the printer driver's "raster". <br /><br />Does this sound like a reasonable assumption or have I got things completely wrong? And more importantly, does anyone have any suggestions what I could to to get rid of this effect?<br /><br />Any input much appreciated, <br /><br />Ola<br /><br /></p><div></div>

[ola_tuvesson]

<p>Here is a 100% crop from the previous image.</p>

<p>Ola</p><div></div>

[Ed_Ingold]

<p>Grain is random, and cannot cause a "beat" or Moire pattern. However, the scan has uniform pixels which may cause the banding you notice due to the way it is interpolated to the pixel pattern of the printer.</p>

<p>Printers lay down discrete dots in 8x8 dot matrices which translate to pixels. The number of dots in that matrix determines the density of that pixel - 0 to 64. The ink dots are randomized to mask this regularity.</p>

<p>Banding is usually caused by clogged ports in the print head. You should try printing test patterns and cleaning if necessary.</p>

<p>If that doesn't solve the problem, try cropping your images to print size and resampling to 180 or 360 ppi and printing to that exact size. Those numbers are multiples of the basic pixel resolution of a 1440 printer - 180 ppi.</p>

<p>Another thing to try is to print in "fine" mode, which half-steps the head and paper advance, possibly eliminating banding through overlapping passes.</p>

[jtk]

<p>Black-only exaggerates banding if there's even the faintest print-head clogging. </p>

<p>You might try Quadtone Printer Driver (QTR). It does a better job simulating silver than does black-only, IMO...though black-only can create interesting grainy effects, especially obvious in small prints. Black-only was popular before QTR but that seemed to decline when people learned QTR was just as "sharp" and, evidently, stable. I've seen zero reports that neutral or cool black QTR settings make less stable results than black-only...the yellow in warmer settings is allegedly less stable but I've seen no controlled research reports. If a warm-toned print changes in a few decades due to fading yellow, it'll get cooler but nothing else should be evident.</p>

<p>If you're using non-Epson ink/pigment it may (did often in reports) result in print head clogging. Most Epson-clog lore (ancient history) had to do with non-Epson pigments, CIS systems etc (you can confirm by browsing). CIS systems were/are often bought to to save money by using non-Epson pigments (even non-pigment "inks").</p>

<p>Both Epson's black and alleged "pure" carbon black are inherently a little red (maybe 3 CC-R), making a little Cyan (as with QTR) or extremely cold-toned paper (eg Moab Kayenta)necessary if neutral black is the #1 goal.</p>

<p> </p>

[ola_tuvesson]

<p>Ok, so, after a few months of procrastination, and the gradual realisation that good B&W printing will never be possible to achieve with my SP 5500, I have now become the owner of an Epson Stylus Pro 4000 instead. Unlike the 5500 this machine is supported by QuadTone RIP and there are several alternative ink options available, including dedicated black & white ones. It also does A2 and handles roll paper so quite a nice upgrade for my tiny budget. I am pretty sure the banding issues I had with the 5500 were caused by poor rasterisation algorithms in the Epson print driver* and I'm betting on QTR being the solution.<br /><br />The replies from John and Edward were really helpful in making this decision - many thanks to both of you!<br /><br />Since the printer only just arrived I'm currently happy to use the ink it came with and print some colour pictures with the standard Epson driver - but I will eventually start to run out of ink and at that point I would like to invest in a refillable cartridge system. Now I have heard horror stories about clogged up heads from third party inks but I'm prepared to take that risk in the interest of saving a fair chunk of $$$. Besides, I'm fairly confident I can unclog the head by reverse flushing should this happen (I did this for the 5500 a couple times which was pretty easy - and from what I've seen removing the head is even easier on the 4000). But since I do all this on a budget which would make most of you laugh I have to try to keep costs under control. I've done a fair bit of searching on the web, reading in various forums and browsing web shops to get an idea of prices. A plan is formulating and it goes something like this: </p>

<ul>

<li>Get a full set of transparent, spongeless, refillable cartridges </li>

<li>A set of 500ml or 1L bottles of the standard colour ink set</li>

<li>Add an additional set of cartridges of the same type for printing with MIS carbon based inks</li>

<li>And of course some bottles of B&W inks</li>

</ul>

<p>But how big should these sets be? Do I really need to swap out all carts every time or can I get a B&W inkset which shares one or more "blacks" with the colour set? And does the B&W inkset really need to be 8 carts in size? Won't I get sufficient quality from QTR with just four? I don't worry about tone at this point, decent neutral tone B&W is more than enough for now. <br /><br />I haven't been able to find a clear answer to any of these questions, can anyone here perhaps shed some light? <br /><br />Many thanks, <br /><br />Ola<br /><br />*) I forgot to mention in my original post: the banding was perpendicular to the print head travel direction; so very unlikely to have been caused by a clogged head.<br /><br /><br /></p>

[ola_tuvesson]

<p>This has been moved to a new thread: http://www.photo.net/digital-darkroom-forum/00XhTi Please post any replies there!</p>

<p>Ola</p>