V700 first scan, lots of questions...

[Wayne]

After years of whinging that I don’t have a scanner I’m now the owner of an Epson V700 refurb. Here’s my first scan from a recent roll through a Rollei:



Probably not a good test due to the extremely narrow DOF, but I just wanted to check that the thing worked after unpacking it.

I have a lot to learn and will no doubt be bugging you all with lots of questions. Here are a few to start with...

I assume it’s best to just get as raw a scan as possible and do all of the fiddling in Photoshop. Is there any advantage to using Silverfast which came with the scanner? I just used the Epson driver for the first scan. I tried it with and without digital ICE and to be honest the scan with ICE just took a hell of a lot longer and it didn’t seem very effective. Maybe I did something wrong.

To find the optimum height for the film holder I was going to pick a sharp, detailed shot and try a few different heights. Does that seem reasonable – any other ideas?

Is there any point in scanning 120 at higher than 2400 with the V700? Paul Headland gives a nice run through on his B&W workflow in this post: http://nelsonfoto.com/SMF/index.php/topic,14455.0.html   Any other hints or tips on workflow will be gratefully received

[LarryD]

I have the same scanner just learn and work on the curve... Looks damn great...

[Wayne]

Thanks Larry - for a quick first scan I think it went OK. LOTS to learn however...  :eek:

[jake]

As raw as possible. I think Silverfast is a kludge. Other people swear it is the best thing ever. I just want a clean scan and I'll do the rest in Photoshop. Epson scan works well enough for that. Digital ICE does take a longer time than a non-ICE scan. There isn't really a wrong way to do it, I don't think. My Epson scan software has both dust removal and digital ICE. I have never used the former but always use the latter on color (only film it works on) and it seems pretty accurate on dust and scratches. Some spotting is required occasionally though. If it is a refurb, and the digital ICE doesn't improve the dust aspect noticeably on your scans, then perhaps the refurb was not successful.

Adjusting height is a good idea. I should do this with my own scanner. I have noticed, however, that flatbed scans require sharpening in Photoshop pretty universally. They just scan soft to begin with.

I don't think there is any point to scanning above 2400dpi with 120mm. Often there isn't much point to scanning above 1200dpi, but scanning at 2400 will give you plenty of information for quite large prints with plenty of detail.

[xvvvz]

>>I have a lot to learn<<

Wayne Fulton's www.scantips.com tutorials are the place to start!

>>I assume it’s best to just get as raw a scan as possible and do all of the fiddling in Photoshop. Is there any advantage to using Silverfast which came with the scanner?<<

Silverfast and EpsonScan are both good.  It is a question of preference for workflow.  They are very different in terms of how you interact so people usually hate one and love the other.  One thing about silverfast, it is the 48 bit HDR mode which gets closer to raw than EpsonScan.  If you really want raw, this is the way to go in my opinion.  Be prepared for a very flat scan.

>>I tried it with and without digital ICE <<

Just an aside, remember that ICE cannot be used with all film types.

>>To find the optimum height for the film holder I was going to pick a sharp, detailed shot and try a few different heights. Does that seem reasonable – any other ideas?<<

Yes, that is all you need.

>>Is there any point in scanning 120 at higher than 2400 with the V700?<<

One reason is to scan at a higher resolution and then downsize to true optical resolution.  This does a bit of noise reduction during downsizing.  Also, some people are able to get a bit more than 2400 out of their scanners, so don't take 2400 as gospel. 

Do everything you can to get as much resolution in your initial scan.  All CCD flatbeds need sharpening but sharpening is easiest when you have a sharper scan from the get-go.

Doug

[Wayne]

Thanks for the tips and suggestions. I haven't had time to do any more scans yet, but will try out a few things (check optimum height, best resolution, Silverfast) soon. I'll try the ICE again Jorn (paying more attention this time) and see what happens. This thing still comes with 6 months of warranty, even though it's a refurb, so if it's not right I'll talk to Epson.

Doug, I'll probably spring for one of your holders when funds become available. When you talk about downsizing are you saying that I should try, say 3200 and then downsize the resolution to 2400 in Photoshop? I'm a bit of a digital dummkopf...

[sebastian toombs]

Doug, I'll probably spring for one of your holders when funds become available.

me too, i always have film flatness issues with my 4490, and my film ends up with fingerprints on it due to handling problems.

[LarryD]

I use lint less cotton gloves I get them at Sally Beauty Supply for a few cents in bulk....

[xvvvz]

>>When you talk about downsizing are you saying that I should try, say 3200 and then downsize the resolution to 2400 in Photoshop?<<

Yes.  Since it does something similar to averaging when it downsizes, it often averages out the bad stuff (noise) and keeps the good stuff.

Good tip about the source for gloves.  I have never seen that suggestion.

Thanks,

Doug

[martolod]

i tried downsizing in viewscan when scanning raw files(tiff) at large resolution and also dabbled with it when using jpgs, but i was not impressed.
 the most effective way i found to sharpen and end up with a crisp scan is the following work flow a friend of mine showed me
scan large (biggest practical resolution)
sharpen in PS (or equivalent)
reduce resolution by half in PS
               or
using existing image (jpg...tiff is to large for this)
increase image size by a factor of 2 - 4 (effectively doubling etc the final image size)
then sharpen and spot
the reduce image back to original size.

i can't tell why it works (i have not sat down and thought about it)it just does,
the downside is that you need a lot of computer power (and correspondingly it takes time)to process images in that way because of the large working file sizes, so i tend to use it only on a few few images that i want to show.

[Wayne]

So many things to try...

OK, Last night I scanned a portion of this picture taken with a Ricoh Diacord:



I tried 2400, 3200 & 4800 dpi and also 2400 with the neg holder 0.5mm higher to see if that made a difference to focus (see here).

I resized 3200 and 4800 to 2400 and looking at tiny crops of tiffs last night both myself and my (disinterested but) objective judge (wife) decided that the resized higher res scans looked better than 2400 and also that the higher setting for the holder was an improvement. There wasn't a lot in it, but the differences seemed pretty obvious. Looking at the same files today on my work 'puter I'm really struggling to see a difference. It is an LCD monitor though compared to the old CRT at home. I've uploaded the crops to photobucket as unsharpened jpegs (which may defeat the purpose of the exercise?), but see if you can detect any difference:

[jake]

Purely unscientific seat-of-the-pants observational (perhaps even) superstitious response to follow:

I am not unconvinced that there are geometries produced by scanning (which renders a curve as a series of squares) that interact with the geometries of monitors (which similarly render curves as a series of squares) and that depending on the interaction between the scanner's geometry and the monitor's geometry, the image can look either smoother or rougher than it actually is.

A similar but not same interaction happens between grain geometry and scanner geometry where a grainy negative that prints nicely in a darkroom will scan like a photo made with mud-pies.

And when I say "geometry" what I mean is the structure/arrangement of the receptors (or whatever they are called) on the scanning element that record the information from the image. Inevitably there are "decisions" made by the scanner as to what the regions of the image that fall in between those receptors look like, purely because an organic shape is being read & then rendered by an inorganic system.

I have images that don't look so hot on the monitor, but they print as very nicely detailed images on my inkjet.

[Wayne]

Yeah, I hear ya Jorn. It's all part & parcel of the digital darkroom I guess. I know nothing of wet printing (having only done a couple of prints when I was at school) and don't intend to get into home inkjet printing at this stage. I could get all carried away with testing this and that (which I don't really have time for anyhoo), but I think what I'll do at this stage is to take an eductated guess at a reasonable scanning technique and then get a few of the files printed up largish to see what they look like.

I want some reasonably good quality scans of some of my better pics so that I can put them together somewhere online & maybe print a few off now & then. I've also got all of those fantastic 100 year old negatives from my great uncle that I'm itching to digitise.

[P C Headland]

This site has some good tutorials on sharpening theory & techniques (and free PS actions) that should help get the most out of your scanner.

[Wayne]

Thanks for the link Paul. I'll try some of the TLR stuff at some stage and I'll probably pick your brains some more as well.

I've only had time for a bit more testing, but I think I've worked out that the highest setting (3.5mm) for the MF holder is the sharpest. I'll try to post some examples of that tomorrow. I did a quick job at scanning and processing one colour neg and it went OK, even if I didn't really know what I was doing. This jpg is 750 X 750 pixels and so is only about 15% of the area of a 2400 ppi scan. I need a faster 'puter...