ScanMate drum scanner DIY maintenance, troubleshooting, mods

[monkeyfist]

Looks like you're doing something wrong. No matter what scanner you use, you need to scan at the highest (real) resolution of the scanner if you want to achieve the highest scan quality and print quality.
People are learning their whole life, so everything is in front of You...

Yes, if you want to achieve the highest resolution of the scanner. But as i want to achieve a quality print, i scan at the size i print. I'm not into scanning lines, or proving my scanners max resolution.

You don't gain anything from scanning at highest resolution, and then scaling the image down. Only thing you are doing is degrading the image quality.

The whole, "scan at highest resolution" would work only if the scaling process that happens in computers would be the same as optical. But it's not, digital is nothing more than what's currently shown in the photo. The extra resolution you had, is gone when you scale the image down. It did not mysteriously add resolution to the end image. It's just gone.

Needless to say, that scanning at 11000dpi always would be idiotic anyhow. Or do you really scan 8x10" @ 11000dpi to get that mysterious extra resolution to you scaled images?

Digital does not enlarge or reduce the size of the image, it creates a new image based on the old. And this causes issues that lead to loss of image quality.

Personally, i think inkjet prints usually look pretty bad. And its really hard to make quality inkjets from film. So i do my best to retain as natural look as i can. Digital just don't like grain.

 

[brbo]

You don't gain anything from scanning at highest resolution, and then scaling the image down. Only thing you are doing is degrading the image quality.

Provia 100F, pushed 2 stops. Which one is "degraded"?

(sorry, not a drum scan as I can't be bothered to mount the drum to prove to myself what I already know)

 

[monkeyfist]

Provia 100F, pushed 2 stops. Which one is "degraded"?

(sorry, not a drum scan as I can't be bothered to mount the drum to prove to myself what I already know)

Comp 2 is scaled down, comp 1 is unscaled. A tiny bit of sharpening on the comp 1 will make it look the same.

What scanner did you use? Also, transparencies will always have noticeably softer grain because of the diffusion.

It would be better to post pictures on actual file server, so they would not show up scaled like in flickr.

(you need to click this image to see it unscaled, it's about 1/3rd of a 645 negative)

The difference is not major, but it's there. And as a theory, this clearly shows the difference of digital vs analog. As people seem to have hard time understanding, that more scanner resolution has no effect on anything else except print size.

The only way to actually increase the resolution of the final products, is to increase the PPI of your monitor or your printer. Currently the large scale printers do 600ppi (canon) and 720ppi (epson). And highest screen resolutions on some mobile devices are around 400-500ppi?

Seems we soon have display tech that can produce higher resolutions than what we can currently achieve with inkjet. Maybe i'll switch from inkjet to OLED, at some point. lol.

Here is another example if found from my archives:

This one shows how using unsharp mask will result in the same look as the downscaled image. It's not a perfect example, as i was not interested in dialing in the unsharp mask to match it perfectly. But works as a reference.

The labels are a bit confusing, but the first one is scanned at 11000dpi and then scaled down to 5000dpi, the second is scanned at 5000dpi and then an unsharp mask was applied to it to match the look of the first image.

 

[brbo]

Comp 2 is scaled down, comp 1 is unscaled. A tiny bit of sharpening on the comp 1 will make it look the same.

What scanner did you use? Also, transparencies will always have noticeably softer grain because of the diffusion.

It would be better to post pictures on actual file server, so they would not show up scaled like in flickr.

The scanner is Minolta 5400. You can view the files in full size on flickr. The comp1 is scanned at 5400dpi scaled down to 2700dpi in PS (bicubic (smooth gradients)), comp2 is scanned at 2700dpi. So much for "degraded"...

BTW, I get a bit more resolution (measured with USAF1951 target) when scanning at full resolution and downsizing than scanning at lower resolution (you do need to pick different resizing algorithm). It's called oversampling and it's a real thing. This is especially true with CCD scanners and really obvious with consumer desktop scanners.

Point is, you are NOT degrading image quality with scanning at the highest scanner resolution.

 

[monkeyfist]

Point is, you are NOT degrading image quality with scanning at the highest scanner resolution.

Actually, you can. As shown in my example that compares sharpening to the scaling.

And i thought it was Minolta 5400, did you scan as raw? As its the only way to turn the software sharpening off from the minolta scan.

 

[brbo]

Actually, you can. As shown in my example that compares sharpening to the scaling.

Well, of course you can. The question is why would you use a wrong scaling method?!

Your example is riding on an assumption that sampling at lower frequency is more "true" and that the fact that you can (somehow) make a lower resolution file look similar to a (somehow) downsized higher resolution scan proves that assumption. I don't agree with this logic.

Anyway, if you are doing prints at standard sizes virtually every print will go through some sort of digital scaling. It's better to have control over the scaling than to trust the dumb machine (scanner or printer) will do it better.

And i thought it was Minolta 5400, did you scan as raw? As its the only way to turn the software sharpening off from the minolta scan.

I scanned as raw and then applied color profile made from IT8 target. I used Vuescan, so there is no sharpening in any mode (unless you explicitly enable sharpening).

 

[help!]

I'm curious, how does the plug-in work?

The plugin can be used as a standalone application. Unfortunately, I failed to run it on Windows XP ('ScanMate not responding...'). As it was designed for Windows 3.1, I suppose this is a 16/32 bit issue.

 

[Kamph]

The plugin can be used as a standalone application. Unfortunately, I failed to run it on Windows XP ('ScanMate not responding...'). As it was designed for Windows 3.1, I suppose this is a 16/32 bit issue.

Damn, I was hoping that the plug-in would be a way to circumvent some of the bugs in QC. What version of PS did you use?

 

[apbruce]

Sm11000

Sm11000

Hi All

I run a SM11000 on a Mac G4.
I use ColorQuartet 4.0.2 on OS9.2 (Scanner came with a CQ4 dongle)
I seems to have got lucky and my machine (from 1995) doesn’t seem to exhibit the problem with banding in the shadows. I did have to replace the power supply a few months into owning it though, but I have a spare non-working SM11000 from 1998 that I can use for parts.

By far the majority of the scans I make are colour negative. Scanned in RGB TIFF 16-bit mode. Always wet-mounted.
I then invert using Colorperfect changing just emulsion type, the black and BP tails.
Mostly my scans are medium format, 5x4 and 10x8”.
Having said that I scan medium format and 5x4” at 4000ppi which means that I need to scan 5x4” film in two strips and stitch (to avoid the ‘Error no 1-6 ReadData error stripsize’ issue with the internal memory buffer) and with medium format i’m right up against the limits of the memory buffer. As a general rule when scanning 16-bit I try to keep under 10,000 pixels per rotation, however recently I have had scans crash with the ‘Error no 1-6 ReadData error stripsize’ issue at resolutions as low as 7500px per rotation. I power down the scanner and restart the mac and repeat the scan and it works just fine - this is incredibly frustrating and can be very time consuming.

I also find that scanning medium and large format at 4000ppi requires the film to be so perfectly flat to the drum that I normally only mount one sheet of 5x4 or one strip of medium format at a time to each of my drums, just so it means I can get the film really tight to the drum. I do think the small size of the scanmate drums is an issue here, you really are forcing the film to curve round a pretty small diameter piece of acrylic.
However I haven’t found mounting just one sheet or one strip of film causes too much imbalance of the drum and my scanner doesn’t seem to suffer from any really notable wow and flutter. I have placed several KG's of metal weights on top the scanner though, and have it placed on a very stable low desk that has yet more weights placed on it to try and keep things as stable as possible.

Today I was reading a few posts here about 8-bit scans being sharper than 16-bit scans and I did a few tests (scanning 6x6 colour neg at 4000ppi) and couldn’t see any noticeable difference in noise, colour and artefacts. I was however very surprised that the 16-bit scan took over twice the length of time as the 8-bit scan. I would have thought that the scan time was only effected by the resolution and not the output format. Does it really take the ADC’s so much longer to output 16-bit data?
Maybe two years ago I replaced the old IDE hard disk drives in my G4 with a Sonnet IDE-SATA adapter and fitted a new SSD to the machine. I was surprised to find this actually sped up the scan time a fair bit.
I also noticed that a few people mentioned the ‘Raw scan’ setting - is this in a later version of CQ (version 5?) or in ColorTrio?
My understanding was that effectively, unless you are adjusting the levels/curves or USM yourself in ColorQuartet then all scans made in this software are by default linear and raw?
I have never been satisfied with the default scanner input ICC profiles (it8trsml.icc and it8rfsml.icc) though. So in my workflow I delete these default ICC profiles so the scanner makes its scans with no ICC profile, and then when I open the image in photoshop for the first time I assign an ICC profile that I have created using a silverfast it8 target and BasICColor. This seems to me a much more accurate method rather than using the default profiles which i find clip and skews the colour for both slide and negative film.
I would like to get the it8 scan feature in CQ to work, but haven’t yet been able to. It took me a while to figure out that you need to scan the it8 target in L*A*B* mode, but then it seems the .txt reference file for my it8 target isn’t read by CQ. I believe this is because CQ needs XYZ, LAB and hue and chroma measurements, and my reference file doesn’t have hue and chroma and i’m certainly not keen on doing the maths for this myself.

Has anyone ever found out what the actual micron measurements are for the scanmate apertures 1-6?
My scanner is generally kept on aperture 4 most of the time.

Thanks very much for all of the tips and thoughts

Andrew

 

[meloV8]

If your scanner is capable of working in raw tiff 16 bit, then stick to this option. Aperture 1 gives you the sharpest scans, so if you use aperture 4, you probably will not see the difference in sharpness between RAW Tiff and Tiff.
You do not need to calibrate the scanner in CQ, it8 target scans in RAW TIFF 16bit and then calibrate in the external program is the most correct.

 

[DKimg]

If someone is interested in the following ScanMate Plugins please leave a comment or PM:

- ScanView ScanMate PC Plug-In For Photoshop and PhotoStyler 1.3.0.
- ScanView MultiMate for PC 1.1.0.
- ScanView MultiMate for Mac 2.0.

Unfortunately, the 'ScanView ScanMate Mac Plug-In For PhotoShop 1.6.' floppy disk isn't working.

I'm interested of the plug-ins. Maybe I'll start drum scanning again this weekend. At least its something to play around with. 🙂

 

[baronkatz]

I finally got the power supply and blowing issues resolved. Looks like it was a bad transformer. The bad news is that suddenly, now, the scanner makes an initialization noise and doesn't actually initialize. The noise just continues and continues.

It happens when the scanner is plugged into the computer. It will seem like it initialized after a while and then go back to making the noise. It goes away when I unplug it from the computer.

Here's a video:

https://www.youtube.com/watch?v=wT8L3bwLXlY

 

[sheel]

Hello,

Is anyone using ScanMate (5000) with an Intel Mac and Ratoc Firewire-to-SCSI adapter? I'm having some stability issues and can't tell if it's because of PPC emulation on OSX, the Ratoc settings, or possibly the scanner itself (e.g. SCSI card).

Gear:
* Late 2008 MacBook Pro (Intel), Rosetta compatible for PPC emulation
* OS X 10.5.8
* ColorQuartet 5.3.1
* Ratoc Firewire to SCSI Adapter, with dedicated power supply, Firmware 1.33
* Scanmate 5000


Symptoms:
* I have strange symptoms where sometimes the ScanMate will "reset" itself as if comm has dropped
* For example, in the middle of doing a preview scan the scanner will stop, reset itself, go into ready state, and then CQ hangs on the preview progress bar
* I end up having to reboot the entire setup to get CQ to connect

 

[sheel]

Hello,

Is anyone using ScanMate (5000) with an Intel Mac and Ratoc Firewire-to-SCSI adapter? I'm having some stability issues and can't tell if it's because of PPC emulation on OSX, the Ratoc settings, or possibly the scanner itself (e.g. SCSI card).

Gear:
* Late 2008 MacBook Pro (Intel), Rosetta compatible for PPC emulation
* OS X 10.5.8
* ColorQuartet 5.3.1
* Ratoc Firewire to SCSI Adapter, with dedicated power supply, Firmware 1.33
* Scanmate 5000


Symptoms:
* I have strange symptoms where sometimes the ScanMate will "reset" itself as if comm has dropped
* For example, in the middle of doing a preview scan the scanner will stop, reset itself, go into ready state, and then CQ hangs on the preview progress bar
* I end up having to reboot the entire setup to get CQ to connect

Correction, I was at firmware 1.23. I must've downgraded at some point. I just upgraded it back to 1.33.

Anyways, I'm trying some new configurations on the RATOC, will see if that helps with anything.

Data Transfer Speed - Async
Delay time between power on and SCSI bus reset - 15s

All the other parameters are default, except "SCSI SIM driver" which I set to off

 

[baronkatz]

Urgent Scanmate 5000 Parts Needed

Urgent Scanmate 5000 Parts Needed

I finally figured out that the problem is most likely within my power supply - as I tested the stepper motor and other components.

 

[ndcowboy]

Hi, I'm new here. I just acquired a Scanmate 3000, but unfortunately nothing included with it.

Excited to get it going and see if it works. I have a computer with a DB25 port. Will that work? If so, I think I'd have to order a DB25 Male to CN50 Male SCSI cable.

I also need supplies for mounting. What is the process for mounting 35mm negatives only. Any suggested supplies? Do I need Mylar overlay? I have never mounted a drum. I see a KAMI starter kit on Aztek.com but it $215... more than i paid for the scanner.

 

[Netsoft2k]

Hi, I'm new here. I just acquired a Scanmate 3000, but unfortunately nothing included with it.

Excited to get it going and see if it works. I have a computer with a DB25 port. Will that work? If so, I think I'd have to order a DB25 Male to CN50 Male SCSI cable.

I also need supplies for mounting. What is the process for mounting 35mm negatives only. Any suggested supplies? Do I need Mylar overlay? I have never mounted a drum. I see a KAMI starter kit on Aztek.com but it $215... more than i paid for the scanner.

Congrats on your purchase. SM 3000 is a great scanner and I am sure you'll enjoy it when you get it up and running. You will need a SCSI connection for the scanner and may need to get a SCSI card for it. The DB25 connection is not a SCSI interface so unfortunately, you cannot use it to communicate with this scanner. Before you get into mounting supplies, I would suggest that you get the scanner working, cleaned (inside optics), and tuned before perfecting the mounting. Do make sure to review this entire thread from beginning to end as most of what you need to get it working and troubleshoot is already well documented.

Goodluck and let us know if you have any specific questions.

Pali

 

[meloV8]

Hi, I'm new here. I just acquired a Scanmate 3000, but unfortunately nothing included with it.

Excited to get it going and see if it works. I have a computer with a DB25 port. Will that work? If so, I think I'd have to order a DB25 Male to CN50 Male SCSI cable.

I also need supplies for mounting. What is the process for mounting 35mm negatives only. Any suggested supplies? Do I need Mylar overlay? I have never mounted a drum. I see a KAMI starter kit on Aztek.com but it $215... more than i paid for the scanner.

I have a SM3000 and it's a nice machine. The quality of the slide scan is good. To scan color negatives is not very good (like most Scanmates) but it can be used ... For wet scanning I use a fluid similar to Gamsol. For this scanner, you do not need to buy expensive ( but great ) Kami foil. All you need is a high quality polycarbonate foil, or possibly a cheap Mylar from Amazon.

 

[Kamph]

To scan color negatives is not very good (like most Scanmates) but it can be used .

This statement is very puzzling to me. I've never had a scanner that gave me as good scans from negatives as my Scanmate 5000. With my SM I scan as a positive, invert in PS, and set manual black and white point for each color channel. In 99% of the cases I end up with a picture that's incredible close to what I actually saw (with the characteristics of the specific film emulision of course). None of the Flextights, Epsons, and Plusteks I've used comes close to the colour depth of my SM.

 

[meloV8]

I am a huge fan of color negative films, I have bought various scanners for many years to achieve the best possible color reproductions. At the moment I have not found anything that has better colors from the negative than the Pakon / Kodak scanners. I have Scanmate 4000 and 3000, these scanners, although they offer a higher scan resolution, unfortunately they do not achieve what Pakon gives in the automatic settings. I also compare to the Primescan D8200 drum scanner, where Scanview scanners are poor in color reproduction, noise levels, etc.

If you have time and willingness, check for yourself if you can get closer to Pakon ...

https://www.dropbox.com/sh/9edn3277bj6aipc/AACEnYhHiQw3EYGf9GCP9K4wa?dl=0