Minola Scan Dual IV users- Your Process?

[eIII]

as i have mentioned in previous posts, i am jumping into the world of rangefinder photography/scanning/photoshop/computers at one time. it is all very exciting/overwhelming/confusing as well! fortunately many hours on this forum has been a tremendous help. the amount of knowledge here is AMAZING!


i now have my film developed only at walmart and just started scanning the negatives with my recently purchased Scan Dual IV. i went with this model after reading about all the success many on this forum has had with it and viewing their galleries/blogs and asking questions.

i am having fun messing around with it and the photoshop elements 2.0 it came with, but i feel like i am throwing punches in the air. i am not very efficient and the results are so-so. how many adjustments do i make prescan? which features are beneficial/ essential and which are a waste of time? what file type (jpeg? tiff?) do i manipulate the picture with and how do i save it so i am not wasting hard disk space yet still be able to post the image in a gallery on this forum or a photoblog and still be able to email it to br printed out at a decent size (8x10)?

these are a few of the many questions i have. i know i will learn more as i keep working, and i will, but sometimes the frustration gets the best of me. 🙂

with so many here using the SD IV scanner/ photoshop elements i was hoping a few might be so kind to give a brief overview of your scanning process before this noobie/rookie runs out into the middle of traffic!! maybe we can all learn something new or do something slightly different for faster/ better results.

thanks SO much!!

bill

 

[sockeyed]

I've got the same scanner but I won't pretent to be an expert with it. In fact, I too, end up wanting to run into traffic when it comes to scanning colour negs (I can never seem to match the colours).

But, I can tell you about how I do B&W: I scan my negatives as COLOR negatives in batches with no tweaking in the scanner software. I believe that autoexposure is 'on', but that's about it.

I import the tifs into Photoshop, crop, invert and desaturate, then I adjust the levels to get the contrast range that I want. After that - dust removal, if necessary, and saving as the highest quality jpg. That's about it!

I'm generally quite happy with the results I get from the Minolta.

 

[dmr]

Here's what I'm currently doing. I've kinda come about this process which works for me. What works for you might be different.

First be sure the negative is as clean as you can get it. **This is one of the things that really makes a difference.** I use canned air, sometimes a dust brush, and I was using filtered H2O in the worst cases but I recently bought some commercial film cleaner for the real grody negatives.

I almost always scan at 3200dpi, 16 bit, and 2 or 4x sampling.

After the first prescan I'll almost always do an auto-focus on a part of the negative or slide that's not dead center, but off center and shows a lot of fine detail. **Again, this is one thing that makes quite a difference.**

Then I'll usually auto-crop, I don't really have any reason, but I'm in the habit of doing it. 🙂

Then I'll tweak the levels. This isn't really super critical (I think?) when doing 16 bit scans, but it sets the range of the machine more in line with what's on the negative or slide.

Then I'll scan and save to a 16 bit .tif file. This will be monstrodious, like 80 megs or so for a full frame. I figure it's better to have too much detail and information than not enough. After I collect a bunch of these, I'll write them to a CD ROM and nuke them to free up the space.

I do everything else in Photoshop, and usually saving the original .tif in case I wanna go back to the original for some reason.

For an almost 8x10 print or so, I'll just size it down a bit at 300 dpi, then do any post processing. If I use unsharp mask I do it in post, as opposed to during the scan. I'll also do a final level adjust and such. I'll often times save such a thing as a Photoshop .psd file. This will be about 18 megs at 8 bit and 36 meg or so at 16 bit.

For web, I'll size it down to 72 dpi and lately (ahem!) I've been making the long size 800 pixels. 🙂

I've found the scanner is consistently reliable for good scans for reasonably well exposed color negatives, color slides, and B&W negatives.

Hope this helos. 🙂

 

[anselwannab]

I posted this before and it comes with the disclaimer that I came upon this reading others posts and scanning the same negatives over and over, and this is the work flow that seems to make the most sense to me and gives me something I can work with.

On the SD-IV B&W scanning, I was having real problems until this week when I started a new work flow. B&W negatives always seemed too dense to scan with blown highlights and no detail in the shadows. I use the Minolta software and Photoshop Elements 3.

1. Scan as Color Positive
2. In Exposure control, move main slider until you get as much of the exposure in with out clipping the high and low.
3. Go to Image correction, the first button on that tab is for "Tone curve and histogram". Move the sliders for the black and white up to the exposure, and then move the gray point back and forth keeping an eye on highlight and shadow control (remember, everything is reversed B&W) I have more of a problem with blowing high lights so get curves with humps rather than dips in them.
4. I have the settings under "Open preferences dialog box" at the top of the screen (little hand pointing towards a checklist. I have:
- Exposure control for negatives set to manual.
- Of the next four check boxes I only have Autofocus at scan checked.
-4x sampling
- 16 bit color (I would think this give you more levels to play with in PSE)
-adobe color space
I then scan it at 3200dpi (why screw around with having to come back?)
5. Scan and save as a .tiff

I then open it in PSE3, use:
-Filter- adjustments -Invert
-Image -Mode -grayscale
-Enhance - Adjust Lighting -Levels (you can adjust the black, white and midpoint)
Maybe not the best way, and I've heard that the Invert and grayscale can be done in diferent ways. This method lets me level it in 16bitmode all the way through. After the last step you can convert to 8 bit and use layers to adjust different parts of the picture. I always have to burn an dodge parts to get the highlights and the shadows right.
Don't forget removing dust!

Like I said, just the way I've been doing it, it seems better to me. None of the scans in my gallery were done this way, but I might even redo some of them since I think it might address some of the highlight/shadow issues.

I do some work in the scan software just so that I I use all the dymanic range of the pic and don't have empty levels at the high and low side. Iwould think these empty levels would leave less to work with to tease out details in the shadows and highlights.

As to size, a full frame 35 negative at 3200dpi gives about 3000x4400 scan when you get done cropping out frame and edge funkyness. That's good for almost an 11x14 at 300dpi.

One thing I was thinking about doing was taking a picutre of a gray card and using that as "job"? when scanning other frames. That way the scanner should know about midtone gray?

I'll try to post some that I did this new way. All the ones in my album are just straight scans, mutitaled in PSE3.

Mark

 

[peter_n]

I am no expert either but I do exactly the same as sockeyed. Exactly. I only scan the ones that look good on the index scan and then scan the best looking negs at 3200 dpi. I am happy with what I get but I believe I have low standards as I'm only scanning for the web, not for printing. I only use B&W film BTW.

I use the Minolta software and have been happy with it, but the insistent drumbeat for Vuescan on this forum and others has me thinking that I might want to invest in it.

 

[dmr]

but the insistent drumbeat for Vuescan on this forum and others has me thinking that I might want to invest in it.

I've looked at the web site for Vuescan, mainly because of the fans it has here, but I really don't see anything it has that I can't do with the bundled software plus the full (legal copy) of Photoshop. I know I'm not using all of the features in the Minolta software at alll, such as the tweaking of various things and USM, not to mention batch scanning. I'm kinda like Peter in that I only scan those that look promising.

 

[peter_n]

I understand it "pulls" a lot more detail out of a negative than the bundled software can. I think I am only happy with my own scans because I'm ignorant and don't know any better. I look at many scans in the galleries here and they are a heck of a lot better than mine.

 

[sockeyed]

Anselwannab suggests that you switch the mode to greyscale. My lab, however, suggested that it was better to always use an RGB image and desaturate to get a greyscale image. They print on a Fuji Frontier. I found that desaturated RGB images print better than greyscale ones. I know not why, but it seems to be the case.

 

[eIII]

thanks all- this stuff is better than the maual! i'll be working through the suggestions today and i can already see this is going to be a BIG help! thanks for taking the time to write this stuff out.

billy.

 

[JohnL]

I do much the same as DMR, but I would point out that however clean you get the negative with canned air, etc, there always seem to be quite a few residual spots which need to be cloned out. This can be quite a lot of work. Like DMR, I use 3200 dpi, 16-bit color and 4x sampling. For fixing contrast and tone I use curves in PictureWindow Pro. I also adjust color balance and satruation "to taste". Then I convert to 8 bits per channel.
One thing I find useful is to uprez the scanned image by a factor of 2x (giving an effective 6400dpi) - for this I use S-spline pro. Then do the cloning and then down-rez to whatever you need for printing etc.
The example below was taken with an SLR (still waiting to collect my RF) on Superia 200, and is about the best I have achieved so far. In the full-sized scan the colours in the water are very subtle and look like water-colour painting at high magnification.
Egyptian Goose

 

[Nikon Bob]

This is a site that I found very useful http://www.scantips.com/ . There are helpful sections on how to set up (calibrate) your monitor, without using something like a colour syder, and colur management. These are good to get you close to printing the colours that you see on your monitor. It is very frustrating to get all the components working together but with a little experimenting and lots of time you will get the work flow and results the you want. Good luck as there is a really big learning curve, at least there was for me.

Bob

 

[eIII]

ok. i've spent the day scanning and playing in PS and i'm making good progress. i think. mostly i've raised a few more questions.

i started scanning a few negatives at 3200 dpi/ 16bit/ 4x sample. a huge 80 mb file takes about 8 minutes to scan and another 1-2 to open n photoshop. too long. is that average or is it my g3 500 mac 128k? i know more ram will help. the scans look great but i have much to learn in order to get the raw image looking good.

in the meantime in order to save time i want to focus on scanning at a good quality for web work- posting in the gallery here and on pbase/flicker. that way i still have time to work on my photog skills, which is what i'm really after! what settings (dpi/bit/sample rate) do you web users scan at for acceptable results?

dmr, after you finish working on that massive 80 mb file, what size do you drop it to to save on your hard drive? i know you make a copy on cd rom to come back to later if necessary, so do you just save it at a web sized jpeg? (100k-300k)?

what size do others save? sockeyed and peter_n? when i get a cd back from wally world the images are around 2mb each. that seems like it could start adding up fast. what is a good size to balance efficiency and quality?

thanks!

 

[dmr]

i started scanning a few negatives at 3200 dpi/ 16bit/ 4x sample. a huge 80 mb file takes about 8 minutes to scan and another 1-2 to open n photoshop. too long. is that average or is it my g3 500 mac 128k?

Ouch! That's much much longer than I've ever had. I've used it on 3 machines, an IBM laptop, maybe about a year old, a Compaq laptop about 2-3 years old, and my very generic no-name PC at home. My guess is a couple minutes max to scan. I'll time it next time I do some, but it seems to vary a bit from negative to negative. I haven't figured that out yet. On the old PC (don't ask me what speed) it maybe takes 30 seconds to pull up in the old Photoshop 5, and I thought that was long. The IBM takes maybe 10 seconds.

what settings (dpi/bit/sample rate) do you web users scan at for acceptable results?

If I am scanning something ONLY for the web and not for printing, I would use something like 1200dpi, 8 bit, 1x sample. Then size it to 72 dpi and maybe 12 inches on the long side at most. Most of the stuff I've ever posted on the web has been sized down from the full res scans.

dmr, after you finish working on that massive 80 mb file, what size do you drop it to to save on your hard drive? i know you make a copy on cd rom to come back to later if necessary, so do you just save it at a web sized jpeg? (100k-300k)?

I'll often times keep my "good" working file, which is almost always an 8 bit 300dpi .psd file of maybe 18 megs or so, on the hard drive. I've experimented saving them as .jpg files at max quality, and they seem to come out 8 megs or so for the same size image, but I'm in the habit of using .psd. I'll probably have to start moving these to CD eventually.

My "workflow" lately has been to crop and size first, then tweak levels and such and then clean up the dust and stuff. If I do make a web size file, I'll do that last. I guess I save these by virtue of the fact that I have not been deleting them. 🙂

thanks!

You're very welcome, hope this helps. 🙂

 

[eIII]

thanks again, dmr. that helps a bunch!

yes, my scans are unbelievably slow. i haven't ruled out operator error and there is a good chance i may be doing something goofy.

my goal is to learn to learn to take good pictures and share them on a budget. wal mart develops for $1.76 a roll and now i can scan my own. i can use this site and others for constructive criticism. i want to do minimal PS work and keep the process simple.

eventually i'll do my own developing, but i don't want to be overwhelmed and thus to frustrated to keep going. one thing at a time!

the help from this site has been GREAT! what did people do before internet forums?? 🙂

billy

 

[Nikon Bob]

eIII

If you are using PS Elements 2.0 then do not scan at 16 bit as I believe to open the file PS Elements 2.0 converts it to 8 bit , 3.0 works with 16 bit. Also do not use multiple passes unless you need to. This may help speed up your scan times. I do not have a Mac, but on my PC PS Elements 2.0 opens a 108 mb Tiff file in 3 seconds. I suspect the long scan times and times to open are RAM related as you have guessed. I am using 2 gig of RAM having moved up from 512. Using a Min 5400 at max rez with no ICE the scan takes 1 minute for a 110 mb file size. You can easily resize the original large file down for use in posting here in PS. I usually save a Jpeg copy compressed to under 170 kb for RFF.

Bob

 

[eIII]

nikon bob, what do you do with the 110 mb file once you've made your adjustments? save to disk? i assume you print your photos, sometimes nice and large?

thanks for the tips, and i am sure it is the ram holding me back. 128k is puny.
billy

 

[Nikon Bob]

eIII

You can save the original to a CD/DVD, an external HD or other storage device. That is your choice. I should also add that I do my work on a copy of the original scan so if I screw it all up the original is still there as a start point. I print small at home, only 8X10 in., which is about a 20 mb Tiff file at 300 dpi for the printer. Check the Scan Tips site out, it explains it a lot better than me. You have to get the PC, the Monitor, the scanner and the printer all working and talking together in the same language at the same speed. It takes awhile.

Bob

 

[peter_n]

[snip] what size do others save? sockeyed and peter_n? when i get a cd back from wally world the images are around 2mb each. that seems like it could start adding up fast. what is a good size to balance efficiency and quality?

I also get a large 80MB TIFF file from the scan and save it to disk (I have a fairly powerful computer with 800GB of disk space and 1 or 1.5GB of RAM (I forget which). After working on it in Photoshop I save the TIFF file then also save it as a high quality JPG file of between 150 and 180K in size. The JPG file goes to the gallery here.

 

[eIII]

NB, i checked and bookmarked scantips. good stuff, thanks.

my 'puter power is definitely lacking in comparison. i think ram is fairly inexpensive and will be a huge improvement. i'll stay simple everywhere else so i don't lose focus on taking pictures. i want to be a phographer, not a consumer/ computer geek! not that there is anything wrong with that. 🙂🙂

billy3

 

[Nikon Bob]

eIII

That is the trouble with home scanning/printing, you have to get semi geekish. You are right that Ram is cheap these days and no need to go overboard like I did, 1 gig might do the trick nicely. Let us know how you make out, especially if the extra RAM does the trick.

Bob