tim parkin

<p>Quite often the inbuilt sharpening (that some drum scanner operators don't know how to switch off and some drum scanner software doesn't let you) can cause huge problems. Also, inbuilt software inversion is typically 'shite' to use a British term. The best auto-neg inversions I have seen come from a Fuji Lanovia which uses Colorkit (the same software in Fuji Frontiers). <br> Also, after lots and lost of testing, setting the black and white point in the scanner software doesn't produce better results than making a raw scan and setting the black and white point in Photoshop (tested on ICG, Fuji Celsis, Aztek Premier, Screen 8060 and Heidelberg scanners). The high bit depth of the scans and the 'dither' inherent in film means no visible data loss using this method. <br> On top of this, Photoshop allows you to set the black and white point perfectly and then you can 'compress' the highlights and shadows, instead of clipping some of the 'grain' (which can make things noisier). <br> Finally, the other test we ran on all of these scanners was to see how effective the aperture was at reducing grain/noise. The answer is 'fairly' but on further testing we realised that you get a better result using a film grain friendly noise reduction plugin (in our research Imagnomic Noiseware was the best). Larger apertures reduced detail at the same time as grain, noise reduction could be targeted to keep detail but reduce the prominence of grain in a way visually similar to aperture changes. <br> we (http://www.drumscanning.co.uk/about/colour-negatives/) recently won a tender to scan a major fashion photographers images which was judged on visual quality alone. These were done using Photoshop inversion techniques we developed in-house (just ask if you want a quick guide) and noise reduction software instead of aperture control. </p>

<p>Hi - it's Tim Parkin here who runs 'cheap drum scanning'. <br> I get a lot of that "if it's so cheap it must be crap" comments. However, I've been chosen to scan work for various professional photographers for book publishing and fine art (Thames and Hudson, Victoria and Albert museum, Cicerone, Ahmed Ertug, Paul Wakefield, David Chow, etc) with many of these sending test scans out to various labs around the world. <br> There are a couple of reasons I'm so cheap and also a couple of reasons why the scans I produce are still better than 99% of the people who run drum scanners worldwide. <br> Why I'm cheap? Drum scanners used to be expensive ($300k for a new Heidelberg Primescan). So anybody that bought at that price needs to pay it back. In the last few years the prices have plummeted and a good drum scanner can be had for $10k. <br> People who can operate drum scanners are rare. Most companies that have a drum scanner need a professional and then they have to work out what that professional is going to do whilst they're waiting for a drum to finish (a drum full of 35mm at high resolution can take 24hrs+). You're paying for that person, that machine and also for the rent, rates and maintenance on the place they work. <br> Now I bought my drum scanners at a reasonable price so the purchase cost is pretty irrelevant in a years operating costs. Maintenance is essential, Karl Hudson does as good a job for me as for anybody who he maintains Heidelberg scanners for. <br> I have a PhD in engineering and a background that includes design and print. My PhD included computer science and so the mechanical, optical and computing side of scanning is something that I'm happy with. There are few people out there who create there own custom profiles from multiple sources using ArgyllCMS. There are also few people who have worked out how to get non-sharpened transparency scans from Newcolor (myself and Derin Korman investigated this and found a solution which we're happily sharing). <br> So for the costs involved, I only charge for the time taken to clean and mount your film to the drum, set the scans up, unmount and process. <br> I don't charge for cleaning - there are many people who can do this cheaper than me and I spent the time instead getting clean originals (I've been told my scans are the cleanest most people have seen despite me not spotting them). <br> So ... which scans are the best? Well that's a tough one. I tested all of the available drum scanners (including Lenny Eiger's services) and my conclusion was that the best balance of shadow recovery, colour rendition, sharpness and consistency came from either the ICG or Heidelberg scanners. The Heidelbergs weren't quite as sharp as the ICG but they were clearer in the shadows. The ICG could get sharper results but not quite as 'rounded' in look. <br> The bottom line for me was that the ICG's charge a ridiculous price for support - a cost I would have to pass on to customers. <br> Bottom line is that I'm happy to do a couple of test scans for people if they wish - drumscanning.co.uk</p> <p>p.s. As for how much detail you can get from a drum scanner. We did a massive comparison of film vs digital and we used high end microscopes to check the results. No scanner available gets all of the detail out of a well exposed film. If you have access to a scanner that gets 5500dpi or higher you can get pretty close though (this seems to be about the break even point for most 'normal' films). If you're shooting T-Max or Adox CMS20 with a Mamiya 7 or good 35mm camera/lens all bets are off though!<br> For your normal (f/22 or smaller) 4x5 transparencies, 3000dpi is usually enough, 4000dpi is the maximum you'll need. <br> https://www.onlandscape.co.uk/wp-content/uploads/2011/12/miroscope-700.jpg<br> https://www.onlandscape.co.uk/2011/12/big-camera-comparison/</p> <p>p.p.s. If you've any questions I'm happy to answer them. Oh and one final spanner in the works. The sharpest scans I've ever seen are from an Imacon and a Screen Cezanne flatbed (and also from a macro lens on a 5DSr) these have there own issues though. Here's an example - including an Aztek Premier sample. This one to show that two scans at 4000dpi aren't necessarily the same (and demonstrate the clean scans from the Heidelberg at 4000dpi). <br> http://www.drumscanning.co.uk/wp-content/uploads/2015/09/portra400-mamiya7-differentscanners.jpg</p>

<p>As the last comment pointed out - any noisy file rotated by a small amount will cause these issues. If you want to avoid it you need to scale the image up by a reasonable amount 50% or more, and then rotate and scale back again. </p>

This is what you get when you rotate a noisy file by a small angle. The only way round it is to upscale, rotate, downscale...

<p>Hi Rodeo Joe, <br> You might be interested to see a comparison between the D800E + Zeiss 25mm with a Mamiya 7 medium format scanned on a flatbed..<br> <img src="http://static.timparkin.co.uk/static/tmp/D800E-vs-Mamiya7-including-colour.jpg" alt="" width="1130" height="838" /></p>