ScanMate drum scanner DIY maintenance, troubleshooting, mods

I have cleaned the lam contacts: They where full of corrosion, as you said, and now there is no banding around. However, when I try to scan TIFF RGB RAW 16 bits i get this:

Screen Shot 2016-11-15 at 18.52.52 by david barreiro, on Flickr

When I scan at TIFF RGB RAW I get a normal file. i have noticed, in this particular scan, that i the grain is kind of long, like stretched in one direction. This is a capture of it at 200%. The scan is also not really sharp, but i guess that is because it didnt focus properly.

Screen Shot 2016-11-15 at 19.07.33 by david barreiro, on Flickr

Any thoughts? I am wondering if something went wrong when the drum got deatached two days ago...


Best regards: David
 
CQ has some bug, and You can't save raw 16bit file (from Scanmate 3000/4000/5000 scanners ), only raw 8bit or Tiff 16bit and 8bit.

Try ColorTrio but You must convert 8 bit scan to 16bit in PS before any editing.
 
CQ has some bug, and You can't save raw 16bit file (from Scanmate 3000/4000/5000 scanners ), only raw 8bit or Tiff 16bit and 8bit.

Try ColorTrio but You must convert 8 bit scan to 16bit in PS before any editing.
Click to expand...

Ok, now everything makes sense...
 
weppo said:
When I scan at TIFF RGB RAW I get a normal file. i have noticed, in this particular scan, that i the grain is kind of long, like stretched in one direction. This is a capture of it at 200%. The scan is also not really sharp, but i guess that is because it didnt focus properly.
Click to expand...

David, the stretched grain and softer image is due to some sort of blur filter applied by QC when scanning in anything but RAW. It has been suggested that the blur filter is added to hide potentiel jitter caused by vibrations in the scanner hardware. It's a rather ungraceful implementation to say the least. One that I hope someday will be fixed, however, given the scanners age, that's probably more wishful thinking than anything else.
 
MeloV8, here is what I see when I try. Click on the link to download the original TIFF scans from the scanner and you can confirm these results for yourself. If you scan as much as I do, profiling is important to get color right straight from the scanner. To me, calibrated scans in 16 Bits directly from the scanner is always much better.



TEST NEGATIVE A

8 BIT RAW (JPEG for reference only) - Click for original scan.

SM%20Neg%20Test%20-%208%20Bit%20Raw.jpg


16 BIT TIFF (Color Calibrated) (JPEG for reference only) - Click for original scan.

SM%20Neg%20Test%20-%2016%20Bit%20Tiff.jpg


8 BIT RAW -> 16 BIT Adobe RGB -> ColorPerfect with Portra 160 Settings (No Adjustments)

SM%20Neg%20Test%20-%208%20Bit%20Raw%20CP%20No%20Adjs.jpg


16 BIT TIFF -> Adobe RGB -> ColorPerfect with Portra 160 Settings (No Adjustments)

SM%20Neg%20Test%20-%2016%20Bit%20Tiff%20CP%20No%20Adjs.jpg


TEST NEGATIVE B - SAME ROLL

8 BIT RAW -> 16 BIT Adobe RGB -> ColorPerfect with Portra 160 Settings (No Adjustments)

SM%20Neg%20Test%202%20-%208%20Bit%20Raw%20CP%20No%20Adjs.jpg


16 BIT TIFF -> Adobe RGB -> ColorPerfect with Portra 160 Settings (No Adjustments)

SM%20Neg%20Test%202%20-%2016%20Bit%20Tiff%20CP%20No%20Adjs.jpg


​

I am sure the 8 BIT scan has all the information to get the desired color but my point is that you will need to spend more time playing with the scans in Photoshop to get it right. Each scan will require steps that will make color balance a manual step which will make getting consistent balance across multiple scans much harder.

Hope this helps or at least prove that profiling makes a difference. You can decide if it is a good difference or not 🙂

Pali
 
meloV8 said:
Color Calibration and profiling is only for scanning slide film, not for color negatives. Just read carefully instruction from ColorPerfect. This is the same as Silverfast HDR scanning mode and Vuescan Raw scan. If You work in different way, You simply doing this in wrong way.

http://www.colorperfect.com/scanning-slides-and-negatives/scans/Howtek/ScanMate/ColorQuartet/

Unfortunately Raw 16 bit not working with Scanmate 3000/4000/5000, so i use ColorTrio because CT give me raw file without any profile and possibility of gamma correction.

Just try 🙂
Click to expand...

It actually good to have a profile for the scanner, CQ actually should come with generic profiles for the hardware. The profile just tells that red is red, and so forth. It's good for negatives also.

The thing about negatives cant use color profiles just means you cant get the particular film calibrated. Like you can with positives, when you actually should have the exact film you use profiled if you want 100% perfect results.

If you just slap an AdobeRGB profile on a file that has no profile at all.. the colors will be a bit funky.

Quick way to test this is to take a screenshot, paste it into PS. As it has no color profile, the colors are wrong. If you put a AdobeRGB profile on it.. colors are still wrong. If you put the correct sRGB profile on it, colors will be good. You will notice a big difference working this way.

Even for a raw file, you can add a profile you have made for the scanner after it's scanned. Then convert that to profoto or adobergb for editing.

With a profile, you get a perfect representation of the negative you have scanned. Without it, it will be somewhat off. Like you can see on jzagajas examples.

Using 8bit raw is not smart anyhow, as you are skipping all the necessary adjustment you should do pre scanning when doing 8bit workflow. There is no room to waste in 8bit workflow, it's not suitable for raw scanning at all.

EDIT

Just to add to this, you can just use the standard profile that comes with CQ. Install the it8trsml.icc (for transmissive, if you scan reflektive also install it) on your system, and when you open a raw scan in PS assign that profile to the file. And then convert it to ProPhoto RGB or AdobeRGB.
 
weppo said:
Thanks for the answer!

-Another question: What is the best way so scan B/W with CQ? I had very cool results scanning color negative as slide (16 bit tiff) and then inverting in Color Perfect (i will post some soon), but i am not sure how to scan B/W properly.
Click to expand...

There is one trick you might wanna test, scan in RGB and then in PS just use the green channel. This will give you the sharpest results, as green is what is used to focus the scan. And B & R are slightly out of focus.

But this is mostly only noticeable for large scans (for prints), always scan for the size you are going to use.

There is no point in making 5000dpi scans if you are going to do 1000px images. You don't gain anything, but will loose while resampling the image to a smaller resolution (avoid resampling as much as you can, people don't realize that it's not like optical enlarging or reducing the image.. completely different in digital world). It will just introduce a sharpening like effect to the image.

And about the banding, i would suggest just getting new fresh bulbs. For quick solution, switch the bulbs around. As reflective bulb has probably had less use. The bulbs don't last that long, they have a lifespan of only 50 hours! so get few of them.

https://www.osram.com/osram_com/pro...-reflector-mr16/index.jsp?productId=ZMP_56075
 
So, I finally got round to having a look at changing the IR cut filter in my SM5000s. First, there are more bolts to loosen (4) to get the lamp bracket off, plus they are very difficult to manipulate with no room to get the Allen key in.

Then I've got a problem - the old filter is square but the new one is round. On it's own I don't think it will stay centred on the light pipe. And this part will get very hot, so I need to be careful what I use to hold it in place.

Pali, was this the same for you?

IR_cut.jpg
 
Adam, Yes it was the same but don't worry too much about the heat as I measured it and the filter area never gets hot. Just slightly warm but no where near as hot to burn anything including paper. I simple cut a small piece of thin cardboard to hold the filter centered and haven't had any issues. You can use tinfoil if heat is a concern and as long as it holds the lens in the cener, the rest will be OK once the filter is sandwiched between the two pieces.
 
I second the tinfoil as a safer path. Check for duct tape in your nearest hardware store, or part 1207792 in homedepot dot com
 
Hello!
I am planning to start wet mounting soon. What do i need to do that? I mean: what fluid do you recommend? Which sheets and what material do i need to get to do the sandwich mounting style? Any advice about how to mount with the integrated mounting station of the scanmate 5000? I dont have a wet mounting station, but would love to get one...
 
Pali, You just prove, that Raw file need little bit more tweaking in film profile in Color Perfect. As You now, every developed c41 film can have some differences in color mask and overall color balance, the developing procedure is hard to get every time the same colors. Whats why film profiles in Color Perfect or Silverfast are only starting point for more tweaking. From that point, there is no sens for it8 calibration for scanning color negatives, because You will correct color cast in film profile, even if scanner don't have perfect color reproduction. This is way how work scanning procedure whit Silverfast or ColorPerfect. Just read tutorial for this software.

Inverting color negatives are always subjective manner and that is why it is so difficult to achieve good results.

I may add that both of your examples, you can see a problem with the gamut of colors. Instead of red, both have something closer to pink. This can be seen both on the car in the upper left corner, as well as on the printing parameters exposure on the edge of the film. It is a mistake to assign the wrong profile colors at work and presentation on the Web.
 
weppo said:
Hello!
I am planning to start wet mounting soon. What do i need to do that? I mean: what fluid do you recommend? Which sheets and what material do i need to get to do the sandwich mounting style? Any advice about how to mount with the integrated mounting station of the scanmate 5000? I dont have a wet mounting station, but would love to get one...
Click to expand...

I use Kami fluid or "copy" of Lumina fluid, and polycarbonate foil.
 
There is no single magic way to scan and invert but telling someone that anything other than your way is wrong is too strong. For each his own and I know what I like. I am happy to know that what I like is different from what you like. I am sure we can still be friends 🙂

I do appreciate your thoughts though and agree that RAW is sharper.

Weppo: I use Kami and Gamsol (Lumina Copy) and prefer Kami. However, some say that Kami can cause internal crazing on the drums so Gamsol or Lumina may be a safer choice since you are starting fresh.
 
Simplest way is to scan in LOG domain - apply true log curve (without slope limit) or true Gamma 3.0 in Photoshop to linear scan) invert, do levels and that's all. This way you get film character. Everything else is additional tone mapping. In most cases you should get perfect colors this way otherwise scanner is doing something wrong.
 
Whatever I write here is just to help someone, not just to argue. Maybe it sounds strange, but scanning color negatives, it's my hobby for many years on various scanners. I checked a lot of scanning methods and some time just I returned to the recommendations of the programs SilverFast, VueScan, ColorPerfect.

For clarity, write how it looks once again.

- Scan negative as positive in 16-bit RAW (if allowed by the software). That means the file without profile colors, without IT8 calibration, and gamma 1.
- Transfer the file to Photoshop and create profile film in ColorPerfect, or reversing using SilverFast HDR.
- When positive is ready, save the file converting to sRGB if the image to be displayed on the network, or write to the Adobe RGB color space if it is to be printed. No need to write to the wider gamut for archiving, because this source file RAW has the widest gamut which has a scanner.

ColorQuartet problem. CQ does not want to save files 16bit Raw on scanners SM3000 / 4000/5000, which is why I use the option 8bit raw and before any treatment converts the raw 8bit to 16bit raw.
For me, even better option is a simple choice ColorTrio where I can choose the type of gamma before scanning. Scans with ColorTrio even better are suitable for the calibration IT8. That is why I am using IT8 calibration in ColorQuartet, because it is very limited, has no influence on the process of calibration.
Make you IT8 calibration in CoCa on Raw scans http://www.muscallidus.com/coca/ and assign IT8 when you need in photoshop or any other software.
 
You must log in or register to reply here.
Back
Top Bottom