Nikon Coolscan 4000 - Vuescan/Silverfast Vs Nikon Scan 4.X

[Twotone]

Thanks both!

Well the iMac has arrived, it's currently installing 10.6 off a USB stick as we speak.

The RAM is a different story! Same brand, different packaging, one of the packages worked immediately, the other 2 sticks weren't recognised by the computer and it wouldn't boot, so it now has 2 X 2gb of original RAM + 2 X 4gb of new ram.

Amazon have told me to keep it (it'll cost them more to return it) so I'll consider what to do next with them! (Bin is likely)

Will probably roll with 12GB of ram total unless it becomes especially painful.

 

[brbo]

Try running only one old module, so 14GB. If I'm not mistaken, those Macs don't need matching modules.

 

[Twotone]

Try running only one old module, so 14GB. If I'm not mistaken, those Macs don't need matching modules.

Thanks, I'll try this later (checking the 'faulty' sticks individually) - just trying to get this OS X installed

 

[Twotone]

Thanks, I'll try this later (checking the 'faulty' sticks individually) - just trying to get this OS X installed

As I'm sure you expected, only one was faulty so she's now running 14GB, and currently receiving Nikon Scan 4.0.2 via bluetooth from my other machine!

 

[brbo]

To be honest, I expected none to be defective. I just thought that one original module needs to be present to act as the "slow" module. If you only use new modules maybe they communicate too optimistic timings to the memory controller and then the memory doesn't work. You could test this further to establish whether this really is just about one particular module that doesn't work... but I guess you'd now rather spend some quality time with Nikon Scan

 

[Twotone]

To be honest, I expected none to be defective. I just thought that one original module needs to be present to act as the "slow" module. If you only use new modules maybe they communicate too optimistic timings to the memory controller and then the memory doesn't work. You could test this further to establish whether this really is just about one particular module that doesn't work... but I guess you'd now rather spend some quality time with Nikon Scan

Certainly possible! I'll investigate in slower time Thanks!

 

[Twotone]

Good Morning Everyone,

Posting this for anyone who finds this thread as there have been some slight nuances which have taken time to resolve with installing Nikon Scan 4.0 on Snow Leopard 10.6.8.

1) In order for Nikon Scan to run (which originates from PowerPc days), Rosetta needs to be installed. This is a background app, not one you can find on the mac or download. This is installed during the OS X install by selecting it in the 'customise install' or something to that effect. The good news is that if you happen to miss this during installing OS X, you can put in the DVD/Bootable USB you used, and click the 'Optional Installs' file which loads up a wizard and you can install Rosetta.

2) The biggest issue that stumped me for a while was the Nikon Scan still not being able to open (it had the clear circle with line through it over the app, signifying it can't be opened), and gave the error "
You can’t open the application “Nikon Scan Installer” because the Classic environment is no longer supported.'

I had tried downloading the original software off the Nikon site that was 'zipped' in a proprietary bit of software called Stuffit (.sit) which unfortunately is no longer supported, even when downloaded off the app store. Thankfully, 'The Unarchiver' exists and covers this obscure compressed format. I thought I was on the home straight, but it still had the same issue! I had been uncompressing on my Mac Mini M1 and then copying over to the iMac on a USB stick and saw someone reference this issue on the FB CoolScan Users group in a one line comment but no resolution.

The fix was to download an updated browser that worked on the iMac (Legacy Firefox that had a different name to firefox), and use that to download the NS files in their .sit format, then download the 'Unarchiver'. By doing this and unarchiving, it was obviously picked up by the magical 'Rosetta' in the background and classed the NS software as perfectly accessible - it then opened.

I'm waiting on my CS 4000 to arrive, hopefully today, so In the mean time I've installed the NS software, however I may need to reinstall once I have the scanner - we'll see.

Either way, a success story, and I'm sure I won't be the only person to come across this issue!

PS. my initial confidence with the RAM was short lived - whilst it ran beautifully, I booted up the iMac last night several hours after shutting down and it was not booting again and doing the beeps. I'm going to investigate again today, but I'm probably going to buy some more 'recognised' RAM. In the mean time, she's back on the old 2 X 2GB sticks and working fine

 

[albireo]

While I felt the Nikonscan output looked great in terms of contrast and sharpness, there is a major caveat in terms of dynamic range, and colorspaces in particular. As much as I liked Nikonscan, it consistently clipped the highlights for me. I could play around with levels a bit, but then my shadows would get crushed. On challenging negatives I always seemed to lose recoverable detail in the extremes. This is apparently a known phenomenon, with plenty of threads on the topic here as well as other sites.

Possibly a dirty mirror. Follow Gleb Schlengtel's tutorial on how to clean your Coolscan's mirror. I did that, and noticed by rescanning the same images in Nikonscan 4.0.3 that the highlight clipping issue had disappeared.

If the above observation will survive more testing it will be amazing, because I haven't found anything that comes close to Nikonscan's C41 inversion algorithms so far. The colours are just -to my eye- perfect out of the box. No tinkering needed. That's right: in my workflow, and with common C41 film like Ektar, Colorplus, Ultramax, Nikonscan's inversion routines are just miles better than anything I've tried. This includes Vuescan' advanced workflow (which I still use for B/W), NLP (this was the worst of the lot), Grain2pixel, and Colorperfect (this is the only one that comes close).

 

[radialMelt]

Possibly a dirty mirror. Follow Gleb Schlengtel's tutorial on how to clean your Coolscan's mirror. I did that, and noticed by rescanning the same images in Nikonscan 4.0.3 that the highlight clipping issue had disappeared.

If the above observation will survive more testing it will be amazing, because I haven't found anything that comes close to Nikonscan's C41 inversion algorithms so far. The colours are just -to my eye- perfect out of the box. No tinkering needed. That's right: in my workflow, and with common C41 film like Ektar, Colorplus, Ultramax, Nikonscan's inversion are just miles better than anything I've tried. This includes Vuescan' advanced workflow (which I still use for B/W), NLP (this was the worst of the lot), Grain2pixel, and Colorperfect (this is the only one that comes close).

Thanks for sharing your experience. How could a dirty mirror affect the scans differently between different softwares? And besides that, I cannot see how a dirty mirror would create clipping in the digital realm. Weird contrast and ghosting artifacts yes. Clipping of the highlights, no.

In any case, mine was recently cleaned - along with the rest of the internals - and flocking added to the case to reduce internal reflections.

There are a number of threads here on Photrio as well as other message boards that go into detail how NikonScan does not handle colorspaces properly. I'll look some up later and share them here-- it is a well documented caveat that I personally have experienced.

Otherwise, I do agree with you, the output for most files is quite pleasing. I do notice more noise in some frames however, compared to my Vuescan workflow applied to the same frames.

 

[albireo]

Thanks for sharing your experience. How could a dirty mirror affect the scans differently between different softwares? And besides that, I cannot see how a dirty mirror would create clipping in the digital realm. Weird contrast and ghosting artifacts yes. Clipping of the highlights, no.

I'm not quite sure. I'd love to find a better explanation for this. My working hypothesis is that dust and traces of grease on the main mirror might create perhaps unwanted reflection patterns or glow that end up tricking the automatic exposure procedure performed by Nikonscan.

What I am observing is that, after a thorough cleaning of the mirror, using Nikonscan and Nikon auto exposure as done previously leads to a much better centred histogram than before. For almost all high contrast negatives, I used to see a histogram which was noticeably shifted to the right, suggesting the histogram was simply translated, rather than mis-shaped. After cleaning, the histogram retains its shape, but it's better centred in the available dynamic range of the 14bit sensor.

Interestingly, I wasn't experiencing major ghosting artefacts or weird contrast before cleaning the mirror. What I saw in Nikonscan was mild colour shifts (emphasized magenta) and occasionally the severe highlight clipping mentioned.

 

[warden]

I'm not quite sure. I'd love to find a better explanation for this. My working hypothesis is that dust and traces of grease on the main mirror might create perhaps unwanted reflection patterns or glow that ends up tricking the automatic exposure procedure performed by Nikonscan. What I am observing is, after a thorough, careful cleaning of the mirror, that using Nikonscan and Nikon auto exposure as previously leads to a much better centred histogram than before. For almost all high contrast negatives, I used to see a histogram which was noticeably shifted to the right, suggesting the histogram was translated towards the right, rather than mis-shapen.

Interestingly, I wasn't experiencing major ghosting artefacts or weird contrast before cleaning the mirror. What I saw in Nikonscan was mild colour shifts (emphasized magenta) and occasionally the severe highlight clipping mentioned.

I think cleaning the mirror is more important than which software we use to scan our negatives. Cleaning a dirty mirror makes a big difference in quality while scanning software makes a small one, if any, assuming we're not clipping highlights and shadows.

I liked the initial color from both Nikon Scan and Epson Scan back when I used them, but both seemed to think a bit too much about what they were scanning, and clipped highlights and shadows a bit. Not a deal breaker but I like to have a bit more room to make adjustments on my own. Epson Scan allowed the option to turn off color management entirely, which solves that problem, allowing the user to select their own black and white points and color adjustments after scanning. Does Nikon Scan allow this too? It's been many years since I've used it so I don't know.

Chris's approach at post #14 achieves a similar result to Epson Scan's "no color management" option, but with VueScan. His resulting scan is flat with no color balance, which is what it should look like, and then he has more to work with to make the final image to his liking without clipping anything.

 

[albireo]

I liked the initial color from both Nikon Scan and Epson Scan back when I used them, but both seemed to think a bit too much about what they were scanning, and clipped highlights and shadows a bit. Not a deal breaker but I like to have a bit more room to make adjustments on my own. Epson Scan allowed the option to turn off color management entirely, which solves that problem, allowing the user to select their own black and white points and color adjustments after scanning. Does Nikon Scan allow this too? It's been many years since I've used it so I don't know.

As regards your point regarding the 'strong' histogram decisions Nikon Scan makes. I agree. It would be good to be able to defeat the histogram clipping settings, while retaining Nikon's proprietary colour management. I haven't found a way to do it. As stated, however, it does seem that my histograms show much milder or inexistent histogram clipping following mirror cleaning.

In general, Nikon Scan does have the option to produce a raw linear positive which can be then color-managed and inverted externally. However, Vuescan allows that too, and I tend to prefer Vuescan's interface in general. So whenever I want to have complete control over the raw output of my scanners (eg when scanning black and white material) I'll go for Vuescan.

 

[Twotone]

Evening Folks,
Well there is some good news and bad news...

The good news is the CoolScan 4000 has arrived, and results look good, noticeably more detail when viewing close up.

The bad news - unfortunately, the scanner starts to malfunction as soon as Nikon Scan is opened, whereas it runs well in VueScan.

Obviously I've gone through the trouble of buying, and prepping a suitable Mac to host Nikon Scan, so I'm currently quite disappointed.

The scanner turns on fine, initialises and then green light is solid - like normal.

If you then add the SA-21 35mm feeder (converted to SA-30), it will do a short whirring/grinding noise, and then the light goes solid - assuming normal too.

If you then open Nikon Scan, the grinding noise starts again and stops, and just keeps repeating, but the green light never stops flashing.

If I use a Vuescan trial version I've downloaded, it works fine.

The Shtengel website suggests this issue (without the specifics of some software triggering it) would be due to a component failure, I have asked a repairer here in the UK what their impressions are, but I'm not especially impressed currently. I've also let the seller know (they are reputable, and have been very helpful so I'm hopeful if there is an issue it will be fixed).

The issue I have, is I'd like to take advantage of Nikon Scan, I'm not sure I have it in me to buy another computer to be able to test whether it's a Mac issue or a scanner issue.

Any words of encouragement/support/condolences appreciated!

Thanks

 

[Les Sarile]

Would it make a difference opening Nikonscan first then turning on the Coolscan?

 

[Twotone]

Would it make a difference opening Nikonscan first then turning on the Coolscan?

Sadly not!

 

[Les Sarile]

Sadly not!

Bummer. It does seem like a system issue given the hoops you had to jump through just to install it. But I don't know anything about the Mac as I've always Windows guy. I have a couple of Windows Vista PCs that I picked up for cheap just for this purpose.
Hope you get it all sorted!

 

[Twotone]

Just a little update for you all!

The old iMac I bought previously was sold as I wasn't able to use it as planned, but recently picked up a high spec Mac Pro 5,1 which is running the appropriate OS X for the Coolscan 4000.

I installed Nikon Scan 4 and updated to 4.0.2 and tried it, unfortunately the same strange behaviour was present again so it's obviously an issue with the scanner (most likely a power supply fault), as such it'll be getting the full works soon and this issue rectified.

Whilst Vuescan is working well for me (scanning as RAW DNG, inverting with NLP in Adobe LR), the ICE leaves a lot to be desired, so I'm very interested to see how Nikon Scan performs in my hands.

I'll report back in a few months

 

[Tom Kershaw]

I installed Nikon Scan 4 and updated to 4.0.2 and tried it, unfortunately the same strange behaviour was present again so it's obviously an issue with the scanner (most likely a power supply fault), as such it'll be getting the full works soon and this issue rectified.

Have you tried running NikonScan 4 on a PowerPC G4 machine? I have found this runs smoother than on an old intel mac mini.

 

[albireo]

Whilst Vuescan is working well for me (scanning as RAW DNG, inverting with NLP in Adobe LR), the ICE leaves a lot to be desired, so I'm very interested to see how Nikon Scan performs in my hands.

Yeah I've noticed the same. Nikonscan's ICE management and results are so much better than Vuescan's. For great results from old, faded, scratched negatives (a tiny minority of my output) Nikonscan and its ICE+ROC implementation is unrivalled. I do scan black and white in Vuescan though: it provides great resources to produce raw linear output positives.

 

[Tom Kershaw]

ROC implementation

With faded negatives and slides, I have white balanced with the colour picker tool within the adjustment curves in PhotoLine, and must admit I've not studied the Kodak digital ROC implementation in NikonScan.

 

[Tom Kershaw]

Here are some rather small samples supplied by Kodak:

Kodak's Austin Development Center - Photoshop Plugins & More!

Applied Science Fiction: The Future Is Clear

web.archive.org

To quote 'Using a complex set of proprietary algorithms, DIGITAL ROC Technology reads the dye signature in color negatives and slides and restores colors to the appropriate levels.' - implying the ROC tech is performing operations on the image data that may be difficult to substitute.

 

[Radost]

Try silverfast

 

[PhilBurton]

Try silverfast

What can Silverfast do better?

 

[Radost]

Negafix color control is very good. It edits the film in a way that … it’s very film print. And not photoshoppy. Great color cast tools.
ICE!!! For color and scratches. A lot of good tools for grain. .

The best thing is the workflow. Scans raw. Has a separate software for raw editing that lets you revisit old scans in the future.

Silverfast is still a working company. I talk to them about issues and ideas. They release updates all the time.

Software is ok a different level in 2024.

 

[PhilBurton]

Negafix color control is very good. It edits the film in a way that … it’s very film print.

Have you ever compared results with Negative Lab Pro?

And not photoshoppy.

What do you mean by "photoshoppy?"

Great color cast tools.
ICE!!! For color and scratches. A lot of good tools for grain. .

The best thing is the workflow. Scans raw.

What version of Silverfast? What settings?

(Do you also have settings for slides?)

Has a separate software for raw editing that lets you revisit old scans in the future.

The "add-on" to the base Silverfast product?

Silverfast is still a working company. I talk to them about issues and ideas.

As far as I am aware, their user forum hasn't worked for years. Are you able to use it? If not the user forum, how do you communicate with them?

They release updates all the time.

Software is ok a different level in 2024.

As you might infer, I tried to use Silverfast but got very frustrated after a while. This software is not cheap. I would be very happy to be able to use it, instead of having to purchase VueScan. Nikon's scanning software is not an option for me. If it matters, I have a Nikon 5000 scanner.