Many different questions about scanning

[Alan Edward Klein]

Interesting, the 8200i review on the same site states it's resolution is 3250dpi as opposed to 3800dpi on the 8100? But you still have to scan at 7200dpi in order to get that.
Scan times per frame is 4 minutes without iSRD and almost 10 minutes with iSRD enabled.

That site also reviewed the Epson V800/850 and conclude that you need to scan at 4800dpi in order to achieve 2300dpi and that scanning at 6400dpi doesn't achieve any more detail.
Scan times for 35mm at 4800 is 1 minute 44 seconds without ICE and 3minutes 44 seconds with ICE.

So if as you said that your V850 is able to take absolutely all image quality the film has - except b&w, then why would you get the Plustek that takes considerably longer to make a scan and requires manual intervention to advance the filmholder for each frame? Couldn't you just apply more sharpening as needed?

Les, when I did a comparison of my V850 with a Howtek 8000 drum scanner with a 4x5 Tmax 100, I found that I was able to sharpen V850 scan of 2400 better than 4800. I don't know why that is so.
https://www.largeformatphotography....-Epson-V850-flatbed-scanners&highlight=howtek

 

[Les Sarile]

I truly don't get the 'speed' argument against dedicated film scanners. Probably because I'm a poor photographer, I rarely scan all 36 frames in a roll. I only scan and archive the 'keepers', which are, if I'm lucky, 5-6 photos on a 36 frame roll.

Nothing to get really as scan speed is only one performance factor to consider. Bad results from a fast scan may make that performance gain a moot point. However, slow scans with bad results might be worse . . .

I can see why in your workflow scan times and automatic film feed may not be big factors and maybe why you don't put much value in them. Conversely, I've put the time and effort in the shot and now all I want is for the scanner to give me the best representation from it consistently.

I think tha if you've found the proper tool to deliver the results you need - or want, then that's a good thing.

 

[Les Sarile]

Les, when I did a comparison of my V850 with a Howtek 8000 drum scanner with a 4x5 Tmax 100, I found that I was able to sharpen V850 scan of 2400 better than 4800. I don't know why that is so.

Sorry but i don't know much about the Howtek really.

 

[Alan Edward Klein]

Sorry but i don't know much about the Howtek really.

I shouldn't have mentioned the Howtek. That was coincidental to my point. I was responding to your earlier post about best setting for V850. I found I was able to sharpen V850 scan of a 4x5 BW at 2400 better than 4800. I don't know why that is so.

 

[Lachlan Young]

I shouldn't have mentioned the Howtek. That was coincidental to my point. I was responding to your earlier post about best setting for V850. I found I was able to sharpen V850 scan of a 4x5 BW at 2400 better than 4800. I don't know why that is so.

Probably because of lower noise.

 

[138S]

I shouldn't have mentioned the Howtek. That was coincidental to my point. I was responding to your earlier post about best setting for V850. I found I was able to sharpen V850 scan of a 4x5 BW at 2400 better than 4800. I don't know why that is so.

Alan, many times we require several consecutive sharpening operations of different radius and % intensity. If you go this way you will get same good job independently of your dpi scanning.

In the Epson V850 the sensor outresolves the scanner's lens by a x2.2 factor if scanning 6400dpi, optimizing this may require a particular sharpening and %. But we do not only sharpen the scanning, we also may be sharpening the taking lens work, that requires a different radius and another %.

Consider this analogy. In a "perfect" digital image you apply a blur filter taking 2pix radius and 70% intensity. Later you apply another blur filter of (say) 7pix and 20% raidius. If you want to recover the image you will have to apply two consecutive sharpening actions, no single sharpening will work perfectly.

When optimizing an image sharpness you may have to perform more than one sharpening. In the case the scanner resolution matches the lens resolution then with a single sharpening you may get a good result.

So yes... you may obtain a better result from a single sharpening operation at some scanning dpi, but if you perform several cascaded sharpening operations you overcome overcome that problem and you perform always an smart job.

Additionally, you need more sharpening operations in the workflow, after you resize to the final size (from the larger edition size) you may sharpen again. And additionally you may make another sharpening that is calculated from the intended viewing distance, the print size and the pixel density in the print, this is to maximize acutance perceived by the eye while totally avoiding overshot.

Because of the cascaded operations required it's essential to work 16bits/ch, to not degradate the image.

___

So the shorter answer is that the scanning work requires a radius for the sharpening and the taking lens work (+ film result) requires a different one. It can happen that depending on the scanning dpi a single sharpening works better because the required raidus match for both situations, but you can always cascade several sharpenings of different radius for a perfect job.

___

If you want, send me the 2400 and 4800 original crops and the sharpened results, and I'll give you the sharpening operations at 4800 that matches or improves the 2400 workflow.

 

[Alan Edward Klein]

Alan, many times we require several consecutive sharpening operations of different radius and % intensity. If you go this way you will get same good job independently of your dpi scanning.

In the Epson V850 the sensor outresolves the scanner's lens by a x2.2 factor if scanning 6400dpi, optimizing this may require a particular sharpening and %. But we do not only sharpen the scanning, we also may be sharpening the taking lens work, that requires a different radius and another %.

Consider this analogy. In a "perfect" digital image you apply a blur filter taking 2pix radius and 70% intensity. Later you apply another blur filter of (say) 7pix and 20% raidius. If you want to recover the image you will have to apply two consecutive sharpening actions, no single sharpening will work perfectly.

When optimizing an image sharpness you may have to perform more than one sharpening. In the case the scanner resolution matches the lens resolution then with a single sharpening you may get a good result.

So yes... you may obtain a better result from a single sharpening operation at some scanning dpi, but if you perform several cascaded sharpening operations you overcome overcome that problem and you perform always an smart job.

Additionally, you need more sharpening operations in the workflow, after you resize to the final size (from the larger edition size) you may sharpen again. And additionally you may make another sharpening that is calculated from the intended viewing distance, the print size and the pixel density in the print, this is to maximize acutance perceived by the eye while totally avoiding overshot.

Because of the cascaded operations required it's essential to work 16bits/ch, to not degradate the image.

___

So the shorter answer is that the scanning work requires a radius for the sharpening and the taking lens work (+ film result) requires a different one. It can happen that depending on the scanning dpi a single sharpening works better because the required raidus match for both situations, but you can always cascade several sharpenings of different radius for a perfect job.

___

If you want, send me the 2400 and 4800 original crops and the sharpened results, and I'll give you the sharpening operations at 4800 that matches or improves the 2400 workflow.

I appreciate your offer. But I'm going to keep it simple. SInce the 2400 V850 scans came close to the Howtek results, I don't want to complicate it and will continue to scan at 2400. The 4x5 BW's are already 265mb each in 16 bit grayscale Tiff. It will be three times larger when I start to scan Velvia color in 48 bit. Raising the scan to 4800 will make the files 4x larger, way over a GB, just is too much for probably marginal gain if at all. Plus I don't think Lightroom can handle them that large. Thanks again.

 

[albireo]

Conversely, I've put the time and effort in the shot and now all I want is for the scanner to give me the best representation from it consistently..

Oh, believe me - I always put the time and effort in the shot, too. And I want, and obtain, the best representation of the shots I like, consistently.

It's just that I'm extremely critical with my output and I like 5% of my shots.

Over the years, I've found that being extremely self-critical with regards to composition, lighting, exposure and framing was the best way to improve.

 

[Les Sarile]

It's just that I'm extremely critical with my output and I like 5% of my shots.

If I took the time to take the shot then I fully intend to keep the shot and therefore scan each and every frame. That I change my mind after seeing the scan happens. Sometimes I reverse my reversal even . . .

A final output is yet another matter altogether, but that is long detached from the scanning process because I don't have to redo the scan. Although I have rescanned some frames made from other scanners I owned before I got the Coolscan.

 

[138S]

But I'm going to keep it simple...

Yes... sure that for 4x5" you get an insane amount of image quality anyway, and from 2400 you get diminishing returns.

 

[Vincent Boman]

Some more questions about the SA-21 film holder. Does it crop the image in any way? Are there settings in Nikonscan or Vuescan that fix this?
I'd also like to know if anyone here has experienced any jamming with this holder?

 

[Les Sarile]

The SA-21 does in fact crop the top end of the film by about 80 pixels. It is physical in nature so there is no software solution to it.

I've scanned tens of thousands of scans later, not one jam.

 

[alanrockwood]

You may try it on your own, from the V750 serial you have a Silverfast license that includes iSRD dust removal, but you may need to upgrade the bundled version (around $50 if IIRC) if also multi-exposurewanting. The infrared dust detection is useful for color films because color dyes are transparent to the IR, so in the infrared image all dust is well detected. For BW the dust removal tool is iSRX in the Silverfast suite, the last version works pretty well.

See here min 20:32

The version of Silverfast that came with my V750 cannot run under Windows 10.

 

[138S]

The version of Silverfast that came with my V750 cannot run under Windows 10.

You may purchase an upgrade to the Version 8 SE Plus (https://www.silverfast.com/buyonline/en.html), not much ($50 ?), if you scan slides this is very interesting because you get Multi-Exposure feature that I guess it was not in your original version.

Reportedly the version 6 should run on W10, but it can have issues, I've ot tried it.

 

[Vincent Boman]

I’ve decided to not look for a Coolscan V/5000 anymore, because of the non-ideal performance from the SA-21 holder, and because in the end I’d much rather have a 9000 anyway. But now I’ve found a good deal on a Coolscan 8000.
With customs, it will cost me under 1500 dollars, which certainly is within my budget. How does it compare to the 9000? I’ve read about problems with banding, can they be worked around? Is there any difference in the ICE? Since the price is about half what I would expect for a 9000, I’d like to know just how much better a 9000 is.
It is also recently serviced, which is a bonus.

 

[Les Sarile]

I’ve decided to not look for a Coolscan V/5000 anymore, because of the non-ideal performance from the SA-21 holder, and because in the end I’d much rather have a 9000 anyway. But now I’ve found a good deal on a Coolscan 8000.

To be sure, it's your money and I am only providing information strictly based on my own experience and not something I read or heard. Comparisons are from my own film across many scanners - flatbed, film scanner, Imacon and dslr. More importantly, I am not trying to convince you - or anyone else, to purchase these scanners.

Just so you know that unless you have one of a very limited number of cameras with 100% viewfinder coverage, you would not have seen the cropped 80 pixels at the top anyway. Everyone who shoots slides and have them mounted will have much more then the 80 pixels cropped at the top.
Typical viewfinder coverage is about 94% which therefore hides 6% all the way around. A full res scan uncropped is about 5641 X 3729 so 80 pixels off the top is about 2%. Most of the time, the actual image on the perimeter is not even and you will have to crop anyway.

If you have many 35mm to scan, or you shoot primarilly more 35mm then MF - and workflow is important, then feeding strips of film is much less tedious then film holders. Of course the speed advantage of the 5000 makes the workflow much faster. The V scan is slightly faster then the 9000 but still no film holder.

Since I have no experience with models prior to the V, 5000 and 9000 all I can say is good luck!

 

[Deleted member 88956]

I’ve decided to not look for a Coolscan V/5000 anymore, because of the non-ideal performance from the SA-21 holder, and because in the end I’d much rather have a 9000 anyway. But now I’ve found a good deal on a Coolscan 8000.
With customs, it will cost me under 1500 dollars, which certainly is within my budget. How does it compare to the 9000? I’ve read about problems with banding, can they be worked around? Is there any difference in the ICE? Since the price is about half what I would expect for a 9000, I’d like to know just how much better a 9000 is.
It is also recently serviced, which is a bonus.

This is NOT coming from direct experience. But, $1,500 for an old old scanner? Are you sure this is a sound investment? Unless this is one that was hardly ever used, comes with all holders and has some sort of at least 6 month guarantee, you may be asking to be disappointed. Holders are hard to come buy, and crazy expensive when they do show up (same for 9000). Check on banding problems, which I understand 8000 developed at some point in its age on a rather universal scale. Those who had used one for long enough might be able to counter this, but banding was one issue that put off from even considering one some 6-7 years ago. 8000 holds its value because of general feedback placing it not far behind 9000 in results department, but I think just like 9000 is artificially way over the top in current pricing , 8000 is even more so.

 

[albireo]

Sounds like a plan. I'm also on the lookout for a 9000 and will buy one when the price is right.

You might want to join the facebook group 'Nikon Coolscan Users'. There are incredibly knowledgeable users on there who would be more than happy to assist you with any issues or questions. One of the members, Frank A. Phillips, actually professionally refurbishes Coolscan machines so that they operate back to original specs. He used to sell some of the refurbished scanners on ebay, though I think he has now set up his own website to do this. Worth checking out if your 9000 deal ends up not working out.

I’ve decided to not look for a Coolscan V/5000 anymore, because of the non-ideal performance from the SA-21 holder, and because in the end I’d much rather have a 9000 anyway. But now I’ve found a good deal on a Coolscan 8000.
With customs, it will cost me under 1500 dollars, which certainly is within my budget. How does it compare to the 9000? I’ve read about problems with banding, can they be worked around? Is there any difference in the ICE? Since the price is about half what I would expect for a 9000, I’d like to know just how much better a 9000 is.
It is also recently serviced, which is a bonus.

 

[Vincent Boman]

This is NOT coming from direct experience. But, $1,500 for an old old scanner? Are you sure this is a sound investment? but I think just like 9000 is artificially way over the top in current pricing , 8000 is even more so.

Yes, but what are the alternatives? In terms of quality? There's only one scanner, though not yet on the market, that does both 35mm and MF and that is the Plustek 120 Pro, which release date has been delayed several times. I can only imagine it will cost a fortune, and compared to the 9000 probably won't hold up. If it ever is released.
Considering the 9000 is still sold in a fully functional state by people who have had it in a dust cover, it's not like buying a pig in a poke.

You might want to join the facebook group 'Nikon Coolscan Users'.

I did yesterday, though I haven't been admitted yet. It sounds great, especially if you can find a 9000 at a good price.

 

[Deleted member 88956]

Yes, but what are the alternatives? In terms of quality? There's only one scanner, though not yet on the market, that does both 35mm and MF and that is the Plustek 120 Pro, which release date has been delayed several times. I can only imagine it will cost a fortune, and compared to the 9000 probably won't hold up. If it ever is released.
Considering the 9000 is still sold in a fully functional state by people who have had it in a dust cover, it's not like buying a pig in a poke.


I did yesterday, though I haven't been admitted yet. It sounds great, especially if you can find a 9000 at a good price.

Honestly, if I were in the market for MF dedicated scanner I'd start digging at Minolta Scan Multi Pro as that was a sure competitor to Nikon 8000/9000 in final scan results. I believe there are fewer offered than any Nikon at any point in time, but you never know if you can get luckier with Minolta (price wise), as all the talk in town is Nikon and nothing else. Minolta gave Nikon run for its money in scanners until they sadly left the field. I have Minolta's Scan Elite II (for 35 mm only). This was their top 35 mm scanner until 5400 showed up, although having seen results from both, I don't think I've missed much, in spite of evidenced difference in real resolution.

 

[Vincent Boman]

Honestly, if I were in the market for MF dedicated scanner I'd start digging at Minolta Scan Multi Pro as that was a sure competitor to Nikon 8000/9000 in final scan results. I believe there are fewer offered than any Nikon at any point in time, but you never know if you can get luckier with Minolta (price wise), as all the talk in town is Nikon and nothing else. Minolta gave Nikon run for its money in scanners until they sadly left the field. I have Minolta's Scan Elite II (for 35 mm only). This was their top 35 mm scanner until 5400 showed up, although having seen results from both, I don't think I've missed much, in spite of evidenced difference in real resolution.

For the moment, I will focus on a 9000, since it is arguably the best scanner that also includes ICE. Though on Monday I'll ask the people at my local lab exactly what they charge for Imacon scans, and what model Imacon they have. They have been on vacation for the past three weeks, so haven't been able to ask.

If the price isn't too outrageous, maybe I'll keep scanning with the Epson V700 and then go to the lab and have the good ones scanned in hi-res.
Depending on how many scans I would get for 3000 dollars, it might be worth it. Though I doubt it, I seem to remember the price being much too high.
Of course the Imacon lacks ICE as well. Now that I think about it, paying for Imacon scans isn't really an alternative, though I'll ask for the price, just to know.
Thanks for the suggestion of the Minolta scanner, though the price seems to be about half of a 9000, now I have to see how they compare.

 

[dwc13]

Interesting discussion.

OP - Have you decided on a Nikon 9000 because it is regarded by many as (one of) the best film scanner or because you truly need (one of) the best in order to complete the work you're doing? I think there have been a lot of good points raised in this thread for your consideration. However, in the end it's your time, money and projects/work flow, so you're in the best position to evaluate and make the decision.

I've had similar discussions with countless people over the years when it comes to buying/building a computer. Some people seemingly always try and do things on the cheap, even if budget is not the key consideration. Often -- but not always -- things don't go quite as planned and they have to expend more resources looking into and procuring another solution. On the other end of the spectrum, many tend to overestimate their needs, sometimes by a significant margin. This is especially true among those who consider themselves "prosumers". I understand why people desire to have "the best", especially when money is not a key consideration (must be nice!). However, clearly not everyone needs "the best" in order to get the job done well. But, if money is no object...

Right now I'm researching film scanners (35mm) myself. With advances in software tools, some of my old and less-than-stellar negatives -- there are more than I care to admit -- might be worth a second look. However, I don't have $3K to drop on a film scanner, mainly because it's just not worth that kind of money to me for what I want to accomplish. YMMV. Besides, I'm building myself a new computer and the (hopefully) soon-to-be-released Ryzen Zen 3 CPU, Nvidia Ampere graphics card w/water cooling, X570 motherboard and 32" 4K monitor all need funding. And yes, I truly will need these components for my, uh, work computer.

If you do end up investing $$$ in an older Nikon scanner, be aware Windows 10 Mainstream support ends October 13, 2020. Windows 10 Extended support ends 5 years after that (October 14, 2025). You might consider purchasing a retail (or OEM, but that license isn't transferrable) copy of Windows 10 before 2025 rolls around, just in case your computer dies. Or worse, you buy a new PC and the new MS OS (Windows 11? Core OS? Polaris? Tik Tok OS?) causes problems. Sometimes, it's better not having to re-invent the wheel (e.g., making things work under a new OS) when you already have a setup that has proven itself.

Good luck in your quest for a Nikon 9000 scanner.

 

[PhilBurton]

Interesting discussion.

Right now I'm researching film scanners (35mm) myself. With advances in software tools, some of my old and less-than-stellar negatives -- there are more than I care to admit -- might be worth a second look. However, I don't have $3K to drop on a film scanner, mainly because it's just not worth that kind of money to me for what I want to accomplish. YMMV. Besides, I'm building myself a new computer and the (hopefully) soon-to-be-released Ryzen Zen 3 CPU, Nvidia Ampere graphics card w/water cooling, X570 motherboard and 32" 4K monitor all need funding. And yes, I truly will need these components for my, uh, work computer.

A few months ago I rebuilt my desktop computer. I switched from an Intel P9/X79 ASUS motherboard to an AMD X570 motherboard with AMD 3900X CPU, new RAM and NMVe SSD. I'm holding off until nVidia announces their new 3000 cards before upgrading the vid card. Windows 10 did a seamless migration from the old Intel setup to the new AMD setup. No BSODs, not even one!

If you do end up investing $$$ in an older Nikon scanner, be aware Windows 10 Mainstream support ends October 13, 2020. Windows 10 Extended support ends 5 years after that (October 14, 2025). You might consider purchasing a retail (or OEM, but that license isn't transferrable) copy of Windows 10 before 2025 rolls around, just in case your computer dies. Or worse, you buy a new PC and the new MS OS (Windows 11? Core OS? Polaris? Tik Tok OS?) causes problems. Sometimes, it's better not having to re-invent the wheel (e.g., making things work under a new OS) when you already have a setup that has proven itself.

Are you sure about that Windows support date? MS releases major updates twice a year now, and those updates lose support on a fairly aggressive schedule, by historical standards.

Good luck in your quest for a Nikon 9000 scanner.

Yes. I have read that there are people who specialize in maintaining old Nikon scanners. However, I have not found anyone who can upgrade a Nikon scanner to USB 3. It's not critical but it would be nice.

Phil

 

[Deleted member 88956]

...

However, I have not found anyone who can upgrade a Nikon scanner to USB 3. It's not critical but it would be nice.

Phil

Wouldn't upgrading to USB3 require major hardware upgrade to benefit from faster interface? Seems like an impossible pipe dream, no? Hardware may have some margin in how much faster it can advance through the scanning process, but how much would one want to spend for a likely marginal true increase in processing speed.

 

[Vincent Boman]

Now, I've started to doubt even a Nikon Coolscan 9000. I've started to look into DSLR scanning, though I had previously decided that wasn't an alternative.
The results it can produce for the formats that I use, 35mm and MF, for the price seems to be the best alternative. Plus, it is the future of scanning.
No new dedicated film scanners that match the old great ones will be created, though new solutions for DSLR scanning will keep showing up.

The reason I had decided on a Coolscan 9000 is one, it does both formats that I use. Two, I've never seen the reason in spending less to get less, or rather to save some money now just to spend that money at some point in the future, when I would absolutely need higher quality. Maybe I'm alone in reasoning the way that I do, but I figured that I should be able to have the best version of my images possible so as not to spend time working on images that come from a lesser scanner, just to work on those images again later when they come from a better scanner.

Though now that I've started thinking about the possibility that a Coolscan 9000 might break in five years, just a number I made up, 3000 dollars seems to be a poor investment.
I thought it should be safe to buy one, considering that they've been kept alive for I guess fifteen years or so. But when will they die completely?
I don't know of any service here in Sweden, so the cost of sending it away, plus service... That's fine for professionals with too much money. I'm not there yet.

I do have reservations against DSLR scanning, because it feels ridiculous to use two cameras for one picture.
Why not then just switch entirely to DSLR? I personally haven't found any DSLR that could match ergonomically my Nikon FE with a 28mm 2,8 Ai-S. I'd also like to keep that lens, and to use that lens with some big adapter on a mirrorless Sony or Nikon is just uncomfortable. So maybe it is best to bite the bullet and start researching DSLR scanning.
Though of course there are just so many solutions for that, which will require a long while of researching. But when it's completely finished, I really do not have any more worries.
Buying a Coolscan 9000, there is the possibility that it goes far too early which my budget can't handle.

What I would like the most is to have a dedicated scanner, one purchase for one purpose. But the more I look for a scanner the less likelier it seems it is an alternative. Though I'd like to know just how long I could expect a 9000 to survive.