Nikon Coolscan 8000 vs Sony A7RII

[plummerl]

I also highly recommend NLP. I'm using it with my Sony A7RIII. My first use was with a Schneider Componon-S, 80mm f4, with my Olympus Auto Bellows. I've since switched to a Sigma 105mm f/2.8 DG DN Macro Art, a very sharp lens. I had been using my Canon US4000 for 35mm. Not bad, but takes forever. The speed that arises using the Sony (with the EFH) is astounding.

I've been digital scanning 35mm (negative and slide), 645, 6x6 and 4x5, close to 30,000 images. Here are four examples, 2 6x6 (4900x4900 resolution) and 2 4x5 (6356x5085 resolution). The files are the result of simply cropping to the image and truly one click processing with NLP.

4x5 #1
4x5 #2
6x6 #1
6x6 #2

 

[xkaes]

Since I don't ADOBE, that's just another hoop to jump through. I'll stick with my scanner that does it automatically -- for free.

 

[Patrick Robert James]

I've stuck with scanning because it is more convenient for me with a Nikon 4000. I can stick a roll in there make a few adjustments then come back later and it is done. The time I spend engaged is less than with a camera scan doing an entire roll, and the quality is about the same (Sony A7Rii), with the edge for color going to the Nikon. Plus there is very little to do after the scan. For medium format there isn't going to be much of a comparison. The Nikon 8000/9000 will be leaps and bounds better unless you want to stitch. Not worth it in my opinion. Of course you may not need that quality. Just depends on you. If you want the best quality but don't want to buy a drum scanner, stitching is the way to go. It has a huge time penalty though.

 

[runswithsizzers]

I also highly recommend NLP. I'm using it with my Sony A7RIII. My first use was with a Schneider Componon-S, 80mm f4, with my Olympus Auto Bellows. I've since switched to a Sigma 105mm f/2.8 DG DN Macro Art, a very sharp lens. I had been using my Canon US4000 for 35mm. Not bad, but takes forever. The speed that arises using the Sony (with the EFH) is astounding.

I've been digital scanning 35mm (negative and slide), 645, 6x6 and 4x5, close to 30,000 images. Here are four examples, 2 6x6 (4900x4900 resolution) and 2 4x5 (6356x5085 resolution). The files are the result of simply cropping to the image and truly one click processing with NLP.

4x5 #1
4x5 #2
6x6 #1
6x6 #2

Apparently, a Google account is needed to view those links?

 

[plummerl]

Apparently, a Google account is needed to view those links?

Sorry, forgot to setup the sharing settings!!!

 

[John Wiegerink]

I also highly recommend NLP. I'm using it with my Sony A7RIII. My first use was with a Schneider Componon-S, 80mm f4, with my Olympus Auto Bellows. I've since switched to a Sigma 105mm f/2.8 DG DN Macro Art, a very sharp lens. I had been using my Canon US4000 for 35mm. Not bad, but takes forever. The speed that arises using the Sony (with the EFH) is astounding.

I've been digital scanning 35mm (negative and slide), 645, 6x6 and 4x5, close to 30,000 images. Here are four examples, 2 6x6 (4900x4900 resolution) and 2 4x5 (6356x5085 resolution). The files are the result of simply cropping to the image and truly one click processing with NLP.

4x5 #1
4x5 #2
6x6 #1
6x6 #2

Those files are really very impressive, to me at least.

 

[Romanko]

For medium format there isn't going to be much of a comparison. The Nikon 8000/9000 will be leaps and bounds better unless you want to stitch.

Technically, you can achieve the same pixel count of ~130Megapixels with camera scanning if you use pixel shift. What is the optical resolution of Nikon 8000/9000 and how does it compare to that of a good macro lens? Nikon uses a 16 bit AD converter while the resolution of the camera is 14 bit. Does it make any noticeable difference in the quality of the image?

 

[albireo]

Actually, I do recommend the NLP plugin to anyone who is copying color negative film with a digital camera, and who is also using Adobe Lightroom. I think NLP is a powerful plugin offered at a fair price. But I would hate for someone to spend their $100 thinking they are going to be able to click one button and get instant results with beautiful, natural looking color, everytime. That did happen for me once or twice, but in my experience, it is rare.

Don't get me wrong - I agree with you completely. I much prefer the colours I'm getting straight out of a Nikonscan all auto inversion than those I get from a string of third party tools I've tried - I was actually sharing some thoughts about this in another thread. I find this impressive given Nikonscan is a really old, unsupported, free bit of software.

However - more in general, my impression is, based on discussions I've had on social media, that for the average user who chooses to scan with a camera, NLP can give results that are as good as, or better than, lengthier sessions involving trial and error manual curve tweaking.

 

[Archiloque]

Technically, you can achieve the same pixel count of ~130Megapixels with camera scanning if you use pixel shift. What is the optical resolution of Nikon 8000/9000 and how does it compare to that of a good macro lens? Nikon uses a 16 bit AD converter while the resolution of the camera is 14 bit. Does it make any noticeable difference in the quality of the image?

On a Nikon 9000, a 6x9 picture scanned at 4000dpi will give you a 150mpx file. Without bayer color interpolation. The Coolscan 8000/9000 lens is also absolutely fantastic, supposedly not that far from a printing Nikkor 105. It also has IR dust removal. I think it still may have the upper hand in MF against camera scan unless you're going full stitching.

 

[John Wiegerink]

Don't get me wrong - I agree with you completely. I much prefer the colours I'm getting straight out of a Nikonscan all auto inversion than those I get from a string of third party tools I've tried - I was actually sharing some thoughts about this in another thread. I find this impressive given Nikonscan is a really old, unsupported, free bit of software.

However - more in general, my impression is, based on discussions I've had on social media, that for the average user who chooses to scan with a camera, NLP can give results that are as good as, or better than, lengthier sessions involving trial and error manual curve tweaking.

On a Nikon 9000, a 6x9 picture scanned at 4000dpi will give you a 150mpx file. Without bayer color interpolation. The Coolscan 8000/9000 lens is also absolutely fantastic, supposedly not that far from a printing Nikkor 105. It also has IR dust removal. I think it still may have the upper hand in MF against camera scan unless you're going full stitching.

The biggest problem I had with NikonScan was not with the performance of the software, but was the fact that no matter how hard I tried. Believe me I tried hard and long, I could not get NikonScan to run on a newer Windows 10 PC. Yes, I tried all the work-a-rounds on YouTube. I preferred NikonScan over Vuescan. Still, the main question for me is will I see a difference in a 13X19" print off my Canon inkjet printer using the Sony camera or the Nikon LS8000? I'm beginning to think not.

 

[xkaes]

You're having trouble/problems with Windows 10 and older, fully-functional, hardware & software???

No kidding!!!

 

[Archiloque]

The biggest problem I had with NikonScan was not with the performance of the software, but was the fact that no matter how hard I tried. Believe me I tried hard and long, I could not get NikonScan to run on a newer Windows 10 PC. Yes, I tried all the work-a-rounds on YouTube. I preferred NikonScan over Vuescan. Still, the main question for me is will I see a difference in a 13X19" print off my Canon inkjet printer using the Sony camera or the Nikon LS8000? I'm beginning to think not.

Test and report? I mean you're right. You don't necessarily need the best hardware to produce acceptable results. Especially when you don't print the size of a wall. I find my 9000 very convenient but it's definitely an overkill. I don't NEED a 100mpx scan of my pictures. Almost nobody does.

As for Nikon Scan, did you try to run it in a virtual machine with windows Xp? It can be quite straightforward.

 

[sojournermike]

On a Nikon 9000, a 6x9 picture scanned at 4000dpi will give you a 150mpx file. Without bayer color interpolation. The Coolscan 8000/9000 lens is also absolutely fantastic, supposedly not that far from a printing Nikkor 105. It also has IR dust removal. I think it still may have the upper hand in MF against camera scan unless you're going full stitching.

I would broadly agree with this. Pixel shift on an S1r claims 180Mp, but that’s a bit made up. Better to think of it as 47Mp true colour with ‘interstitial samples’ to reduce aliasing.

With stitching you can pretty well achieve what you want. Pixelshift at 1:1 would be lots of detail and a massive file…

 

[sojournermike]

I would broadly agree with this. Pixel shift on an S1r claims 180Mp, but that’s a bit made up. Better to think of it as 47Mp true colour with ‘interstitial samples’ to reduce aliasing.

With stitching you can pretty well achieve what you want. Pixelshift at 1:1 would be lots of detail and a massive file…

Adding to this, that scanning resolution can matter if you’re working with a highly resolving system - grain aliasing can be pretty significant and worth experimenting a bit to get results you like.

I’ve not tried it, but it might be that doing a single shot pixelshift image of a 6x9!frame might actually benefit from a smaller aperture, for example, as a form of anti-alias filter. One to play with when I’ve a day to scan and make big prints just for research purposes.

 

[John Wiegerink]

As for Nikon Scan, did you try to run it in a virtual machine with windows Xp? It can be quite straightforward.

No, I didn't try that and just used Vuescan. I'm not really going to say NikonScan gave me better results for B&W, but since I was so used to using NikonScan it seemed/seems easier to use than Vuescan.
Yes, I'm not blowing up wall size poster or prints. Also, I still wet print in the darkroom and if I have a negative worth its salt it will get wet printed.

 

[sojournermike]

No, I didn't try that and just used Vuescan. I'm not really going to say NikonScan gave me better results for B&W, but since I was so used to using NikonScan it seemed/seems easier to use than Vuescan.
Yes, I'm not blowing up wall size poster or prints. Also, I still wet print in the darkroom and if I have a negative worth its salt it will get wet printed.

The lack of support for newer operating systems was one of the big reasons I sold mine. I want this part of the process to not be full of hurdles.

 

[Archiloque]

It really isn't that difficult.

It took me less than an hour to install Nikon Scan and the Coolscan 9000 in windows 11. The biggest hurdle was finding a Firewire adapter compatible with my laptop. Otherwise I found it to be straightforward once you get pass the unsigned drivers installation problem.

 

[xkaes]

The lack of support for newer operating systems was one of the big reasons I sold mine. I want this part of the process to not be full of hurdles.

I avoided the hurdle by keeping my XP machine -- with its old software and old hardware. Works great.

 

[Chan Tran]

The OP already has the Coolscan 8000 I think he should use it. Using the scanner can be slower than a camera but it's much easier. The 8000 is a high resolution scanner so unlikely a camera can yield better quality than it.

 

[John Wiegerink]

The OP already has the Coolscan 8000 I think he should use it. Using the scanner can be slower than a camera but it's much easier. The 8000 is a high resolution scanner so unlikely a camera can yield better quality than it.

I'm certainly not complaining about the Coolscan 8000's output. In fact, I'm very happy with it. I was just wondering since having the Sony A7RII if I really need both and could maybe get by adequately with just the Sony A7RII and a good flatbed scanner. The Nikon Ls80000ed and LS9000sd are selling for a high dollar at the moment, so it might be a good time to part with it before it breaks down.

 

[Steven Lee]

Adding to this, that scanning resolution can matter if you’re working with a highly resolving system - grain aliasing can be pretty significant and worth experimenting a bit to get results you like.

The grain appearance is a big and complex topic. Personally, I find it more important to discuss than scanning resolution because grain jumps at you regardless of the level of detail in a scene. Badly scanned grain can ruin an image even at smaller magnifications.

I can't even generalize here, because scanning can ruin grain in different ways, both in hardware and software. I never owned a Flextight, but my local lab claims that its aggressive sharpening cannot be turned off.