Scanner with 8K resolution

[Alan Edward Klein]

OK thank you very much for your answers.
One question again: How do you judge the quality of a scanner besides resolution and bit depth?
Is there any parameter that matters besides that when it is about telling how much quality loss you will have?
Or can you generally say that the more resolution the negative is scanned with the lower the quality loss is?
I know there is also the MTF of the scanners that would be important but you rarely find those online, you would have to figure that out yourself.

The minimum dMax is also very important. The higher the number, the better the scanner can "see" through the denser parts of the film. That's important to see shadow details better.

 

[MattKing]

I 'invested' I an Epson V850 Pro a few years ago..It came with Viewscan software..

Are you sure it wasn't Epson ScanSmart or Epson Scan software?
Viewscan software is a product for microfilm.
To the best of my knowledge, Epson scanners have never been packaged with Vuescan (not Viewscan) software.

 

[mshchem]

Boy now I'm really confused. I've got a little Nikon Coolscan V, I picked it up used. I use Vuescan no troubles. I was using to scan my Dad's early Kodachrome work. I made some prints on a Canon inkjet, maybe a bit bigger than 11x14, came out beautiful.

I wish I could take a course on how to do all the computer Voodoo. I use Lightroom for my Digital cameras, Photoshop is too much.

 

[Wallendo]

Much of the confusion in this thread is related to the terms "4K" and "8K". These are generally terms related to the number of pixels in the horizontal plane only of a digital image (the vertical resolution varies with aspect ratio). These terms have little applicability to still film use. Even a low quality flatbed can produce a 8K scan if the original is large enough. My Plustek OpticFilm can produce a scan from 35mm film at its 3600 dpi setting of ( 3600 dpi * 36 mm / 25.4 mm/inch) or 5K. At its 7200 dpi setting, I can get a 10K file, but it doesn't look a lot different from the 3600 dpi file.

I have always found it interesting how the movie and still film industries have always described things by different incompatible definitions.

 

[Les Sarile]

Thank you so much for all your interesting posts!
What do you think about using the Nikon LS-5000 to scan 35 mm negatives? I know it's not 8K, but it is still over 4K.
Do you think it is enough to scan 35mm negatives without quality loss? Do you think 16 Bit is enough?

OK thank you very much for your answers.
One question again: How do you judge the quality of a scanner besides resolution and bit depth?
Is there any parameter that matters besides that when it is about telling how much quality loss you will have?
Or can you generally say that the more resolution the negative is scanned with the lower the quality loss is?
I know there is also the MTF of the scanners that would be important but you rarely find those online, you would have to figure that out yourself.

Ultimate detail capture is not only dependent on the scanner but on the film, the lens and settings, the scene and the conditions. For instance if you are shooting handheld with wide open aperture on a long poor quality lens in foggy conditions at a scene that requires huge DOF with a poor quality film then all the reach of a Heidelberg Tango 10,000dpi will not do it a bit of good. OTOH if you have your camera on a tripod, mirror lockup, ideal lighting enough for sufficient shutter speed and aperture setting shooting a flat test target on a wall on Kodak Techpan @ ISO25 developed in Kodak Technidol then neither a Coolscan's 4000dpi nor will a 36MP Nikon D800 be sufficient to capture the detail achieved on the film . . . see below.

Full res version -> http://www.fototime.com/8372250EA44CB06/orig.jpg
Bottom left shows the full arranged 4 X 4 target with the center area in red 100% crop.
Above it is a 14MP Pentax K20D scan of the film.
Above that is the Coolscan 4000dpi scan.
Above it is the 36MP scan with the Nikon D800. Although the D800 applies more pixels to the area, it is almost the same as the Coolscan if not slightly less.
Large crop on the right is an 4.5X optical magnification of the center area that clearly shows detail not resolved by these methods.

However, if you shoot good quality high resolving color negatives like Kodak Ektar 100 with some care, then perhaps the Coolscan 5000 may be sufficient. Not to mention that it will render accurate color and contrast results in about 30 seconds or clean dust and scratches in about 50 seconds per frame.

Full res version -> http://www.fototime.com/EEA4F124C726025/orig.jpg

BTW, if you shoot more color negatives then any other type of film then the Coolscan LS5000 is as ideal a scanner as you can get. With dust and scratch removal if scans a frame much faster then any other desktop model/brand. This is also far faster then any DSLR capture which may capture each frame in seconds - with the right setup, but post work will be far longer just to do color inversion let alone dust and scratch removal.

 

[grat]

Boy now I'm really confused. I've got a little Nikon Coolscan V, I picked it up used. I use Vuescan no troubles. I was using to scan my Dad's early Kodachrome work. I made some prints on a Canon inkjet, maybe a bit bigger than 11x14, came out beautiful.

I wish I could take a course on how to do all the computer Voodoo. I use Lightroom for my Digital cameras, Photoshop is too much.

Sometimes, not knowing makes it easier. Techies (I include myself) sometimes get hung up on the details. I haven't used Photoshop / Lightroom in some time, but as memory serves, Lightroom is fine for tweaking colors, cropping, resizing-- anything that isn't pixel-level editing, which is where Photoshop comes in.

I've switched to Affinity, which has most of the features of Photoshop, but has simplified the interface by breaking it into "personas". It's photo-healing tools are good enough I've given up on ICE.

If you really want to learn how to do more with Lightroom and other programs, there's a youtube channel by "Robin Whalley" in the UK-- he has a bunch of books available as well, but he demonstrates how to do common tasks in various packages such as Lightroom, Photoshop, Affinity, Capture One, etc.. The accent's a bit heavy, but clear, and the information is solid.

 

[Les Sarile]

I've switched to Affinity, which has most of the features of Photoshop, but has simplified the interface by breaking it into "personas". It's photo-healing tools are good enough I've given up on ICE.

What scanner/software do you use?

 

[Alan Edward Klein]

Much of the confusion in this thread is related to the terms "4K" and "8K". These are generally terms related to the number of pixels in the horizontal plane only of a digital image (the vertical resolution varies with aspect ratio). These terms have little applicability to still film use. Even a low quality flatbed can produce a 8K scan if the original is large enough. My Plustek OpticFilm can produce a scan from 35mm film at its 3600 dpi setting of ( 3600 dpi * 36 mm / 25.4 mm/inch) or 5K. At its 7200 dpi setting, I can get a 10K file, but it doesn't look a lot different from the 3600 dpi file.

I have always found it interesting how the movie and still film industries have always described things by different incompatible definitions.

You forgot the 3600 for the other direction. It's 1.41" x 3600 x 1" x 3600 = 18.3MB (Note: 25.4mm = 1" and 36mm = 1.41")

If you use 2400dpi, which is what I use to scan 35mm on my V600 and V850, you get 1.4" x 2400 x1" x 2400 = 8.1mb which may be what the teacher wants. The student should ask.

 

[Henning Serger]

Thanks for the analysis and suggestions. At this time I don't do any printing and planned to have it professionally scanned when I do for wall-mounted prints. So I'm pretty satisfied with the V850. It seems a little better than my V600. The only printing I might also do is for a personal photobook that will use medium format and 4x5's. So the print size won't be larger than maybe a double spread in the book. Most probably just on one page. I think the V850 should suffice for medium format film scans for the book. What do you think?

Alan, you are welcome.
Concerning your last question: I don't know which quality level you personally consider as "good enough" for you. The quality requirements / standards of different photographers vary very much due to their individual preferences. Therefore I would suggest: Just try it. Maybe make at first a smaller book to test it (and your lab / book supplier).
For me personally the quality of an Epson flatbed scanner is not good enough for most of my personal applications. I own an Epson flatbed, too, but I am using it very rarely for negatives and transparencies, mostly for written documents and office applications, and sometimes for prints.
Everytime when I look at a flatbed scan of one of my medium format transparencies and compare them to the original on a lighttable, viewed through my Schneider-Kreuznach 3x medium format loupe, I am so disappointed by the the flatbed scan, and so excited by the outstanding quality of the orginal under this excellent loupe (and of course when projected with my Rollei medium format slide projector with the AV-Xenotar lens).

Best regards,
Henning

 

[Henning Serger]

Thank you so much for all your interesting posts!
What do you think about using the Nikon LS-5000 to scan 35 mm negatives? I know it's not 8K, but it is still over 4K.
Do you think it is enough to scan 35mm negatives without quality loss?

No, definitely not. Not at all.
To make it clear clear just one example from my numerous test series: I've got a system resolution (lens+film) with Provia 100F of 120-135 Lp/mm, with my standard test lens Zeiss ZF 2/50 at f5.6 and an object contrast of the testchart of 1:4 (two stops).
I have scanned it with a Nikon Coolscan 5000 and got a resolution of only 60-65 Lp/mm. So just half of the original resolution!
But when I project this transparency with my Leica, Zett or Kindermann 35mm projectors and the best projection lenses for them (Leica Super-Colorplan, Zeiss P-Sonnar, Kindermann/Doctor-Optics MC-B) I get about 115-120 Lp/mm resolution.So I can almost fully exploit the extremely high colour reversal film resolution with projection.

Another example: With ADOX CMS 20 II - the by far highest resolving material on the market - I've got under the same test conditions 240-260 Lp/mm (the diffraction limit of white light for f5.6). Even the best drumscanners have only delivered up to 135 Lp/mm in my tests. But with optical enlargements and projection I've got results in the 200-230 Lp/mm range.
So with scanning you will always have a quality loss. You can only choose how big that will be.

We are meanwhile so 'brainwashed' by digital marketing that all "has to be digitalized" that we too often overlook the better original solutions / alternatives.
And too often the limitations of the complete imaging chain are ignored: Customers buy very expensive 24 MP, 36 MP, 45 MP or even 60 MP digital cameras, and then in 99.99% of the time they only view their shots on a 2k (2MP) or 4k (8MP) computer monitor. So they pay a huge amount of money for a resolution which they afterwards destroy by their monitors.
The same is of course valid for film photographers, who are using with film a medium which is capable of capturing huge amounts of detail, then scan it and also view the results only on their computer monitors.
And it is unfortunately also often overlooked that by using a complete analogue workflow there are other advantages, too:
- by using BW or colour reversal film you already have a finshed picture in perfect quality after processing
- no further steps like scanning needed, so you avoid both the time consuming step of scanning (if you do it by yourself), or the additional costs of scanning (if you use a scan service)
- so you have an extremely easy, convenient and cost-efficient workflow, which also offers perfect quality.

Best regards,
Henning

 

[Alan Edward Klein]

Alan, you are welcome.
Concerning your last question: I don't know which quality level you personally consider as "good enough" for you. The quality requirements / standards of different photographers vary very much due to their individual preferences. Therefore I would suggest: Just try it. Maybe make at first a smaller book to test it (and your lab / book supplier).
For me personally the quality of an Epson flatbed scanner is not good enough for most of my personal applications. I own an Epson flatbed, too, but I am using it very rarely for negatives and transparencies, mostly for written documents and office applications, and sometimes for prints.
Everytime when I look at a flatbed scan of one of my medium format transparencies and compare them to the original on a lighttable, viewed through my Schneider-Kreuznach 3x medium format loupe, I am so disappointed by the the flatbed scan, and so excited by the outstanding quality of the orginal under this excellent loupe (and of course when projected with my Rollei medium format slide projector with the AV-Xenotar lens).

Best regards,
Henning

Thanks for your input, that's a good suggestion.

 

[Wallendo]

You forgot the 3600 for the other direction. It's 1.41" x 3600 x 1" x 3600 = 18.3MB (Note: 25.4mm = 1" and 36mm = 1.41")

If you use 2400dpi, which is what I use to scan 35mm on my V600 and V850, you get 1.4" x 2400 x1" x 2400 = 8.1mb which may be what the teacher wants. The student should ask.

It is a terminology issue. The generally accepted use of resolution when measured as "K" is horizontal resolution only. Most motion pictures are currently displayed at 2K or 4K resolution, meaning there are 2000 or 4000 horizontal resolution and variable vertical resolution. Current televisions and projectors use the label "4K" for a 3840x2160 resolution. And Apple uses the term 5K for its 5120x2888 monitors, I guess because it sound cool. Digital cameras usually use Megapixels, but some labs describe the results of their scans in Megabytes, which actually provides very little information about resolution.

We really don't know what the instructor is requesting. Is is high resolution detailed scanning? 4 or 8 Megapixel scans (not particularly high resolution by modern standards)? Files of 4-8 megabyte size to save disk space? Or images to display on a 4K or 8K monitor?

 

[NortheastPhotographic]

I think we're getting well beyond information actually useful to the OP, and worse it's stuff that has been stated again and again on this forum. If you've been itching to show off everything you know about image resolution in the form of math formulas and graphs, maybe now is the time to ease off the trigger there, cowboy.

 

[Oren Grad]

We still don't know the teaching purpose of this exercise.

Regardless, the OP is evidently determined to find the highest-fidelity digital capture of a 35mm original. I don't think you can achieve that by reading spec sheets, which at best are descriptions of the scanner hardware and not of the output files. My guess is that the best that can be achieved will end up coming from either a high-end drum scanner or possibly a Hasselblad 400c-MS copy stand setup, plus a skilled, experienced operator either way. Not sure which one would win out - I haven't seen controlled comparative tests. The cost will be monstrously large files, the benefit will be faithful capture of image character down to the grain level.

 

[grat]

What scanner/software do you use?

Epson V800 with Silverfast 8 SE Plus (The scanner came with SE, but included a discount on the upgrade to SE Plus).

 

[Jonno85uk]

I think we're getting well beyond information actually useful to the OP, and worse it's stuff that has been stated again and again on this forum. If you've been itching to show off everything you know about image resolution in the form of math formulas and graphs, maybe now is the time to ease off the trigger there, cowboy.

Oh boy. I'm glad somebody said it.
Did we ever get an answer as to what the OP was going to use the scans for

 

[Chan Tran]

Oh boy. I'm glad somebody said it.
Did we ever get an answer as to what the OP was going to use the scans for

I think the OP stated his purpose here
"Thank you so much!
I need the scans to compare analog and digital photography.
I am not sure yet whether to scan on my own or using a service. I would rather want to use a service but my teacher wants me to scan on my own.
I'll see what I can get."
However, I think it's difficult because you digitize the film you must lose something so it's not a fair comparison.

 

[Jonno85uk]

to compare analog and digital photography.

That's vague do you not think?

 

[Chan Tran]

That's vague do you not think?

I don't think it's that vague. I think I know that the OP wants to get the best of film (in digital form, it's kind of oxymoron) and see how it fares with the digital. They are evaluated in the digital domain so it's kind of not fair at all.
I would suggest that the OP have a 35mm color negative print made optically in the small 4x6" size and then trying to print a digital file the best he can in the same 4x6" size and evaluate them under a loupe. I bet the film print would beat the digital at least in term of resolution in this case.

 

[Jud23]

Thank you all for your answers and your interesting input!
To be more specific about my project, it is about figuring out if shooting analogue is still viable.
Actually it is about literally films (moving pictures) but I can not compare analogue and digital moving pictures since I have no chance to get the equipment for that so I compare photos.
If moving picutres are shot analogue, it is common to digitalize, for postproduction but also for distribution. This is why I compare them in digital form.
The question of my project so far is how I can compare them on a fair level since it would be unfair to compare a great digital Sensors quality with poor quality scan to come to the conclusion that digital is as good as analogue. So I have to find a comparision basis. I still have to think about that. When I posted the question I thought it would be best to compare the sensor to the analogue quality so that I need a scanner which is capturing this quality on the best level. But now I know that this is not possible so I need to take a step down and include the problem with quality loss in digitalization in my project somehow.
You have given me a lot of input and actually I don't have further questions at this point. Maybe I will have some in time again. So there is no need in giving me further answers but of course I am happy and thankful if you have further advices, but as mentioned, no need for it at this point.
Have a nice day and thanks again.

 

[Alan Edward Klein]

Movie film works differently than stills. Because you're looking at let's say 24 frames per second, all the grain and out-of-focus individual frames blend in. This gives the viewer an illusion of better quality than they may actually be. So comparing an individual shot from a movie film to a single shot taken within a still camera, are not comparable.

 

[ParkerSmithPhoto]

Hello
Does anyone know Filmscanners with a resolution of 8K?
If not, which scanners have a resolution of 4K?
I need the best resolution for a project.

Best regards and thanks in advance

Resolution and detail are often two different things. When I fluid-mount scan medium format negatives, I choose 2400 PPI. My experiments show there is no increase in DETAIL beyond this point, only a larger file. Yet I can still see the threads in a shirt when I print these files.

 

[Les Sarile]

Thank you all for your answers and your interesting input!
To be more specific about my project, it is about figuring out if shooting analogue is still viable.
Actually it is about literally films (moving pictures) but I can not compare analogue and digital moving pictures since I have no chance to get the equipment for that so I compare photos.
If moving picutres are shot analogue, it is common to digitalize, for postproduction but also for distribution. This is why I compare them in digital form.
The question of my project so far is how I can compare them on a fair level since it would be unfair to compare a great digital Sensors quality with poor quality scan to come to the conclusion that digital is as good as analogue. So I have to find a comparision basis. I still have to think about that. When I posted the question I thought it would be best to compare the sensor to the analogue quality so that I need a scanner which is capturing this quality on the best level. But now I know that this is not possible so I need to take a step down and include the problem with quality loss in digitalization in my project somehow.
You have given me a lot of input and actually I don't have further questions at this point. Maybe I will have some in time again. So there is no need in giving me further answers but of course I am happy and thankful if you have further advices, but as mentioned, no need for it at this point.
Have a nice day and thanks again.

In your other thread "Questions regarding analog photography" I responded with what I think is a significant difference between the two.

Resolution is certainly one far easier to quantify as Henning points out digitization is a big compromise to serve a convenience but color fidelity is yet another.

 

[runswithsizzers]

You forgot the 3600 for the other direction. It's 1.41" x 3600 x 1" x 3600 = 18.3MB (Note: 25.4mm = 1" and 36mm = 1.41")

If you use 2400dpi, which is what I use to scan 35mm on my V600 and V850, you get 1.4" x 2400 x1" x 2400 = 8.1mb which may be what the teacher wants. The student should ask.

If you are multiplying pixels times pixels, I believe the correct units would be "MP" (MegaPixels), and not MB (MegaBytes), right?

Knowing the megabyte file size tells us how much space a file will take up in storage, but I believe megabytes are much less useful for estimating the resolution of an image.

 

[Les Sarile]

If you are multiplying pixels times pixels, I believe the correct units would be "MP" (MegaPixels), and not MB (MegaBytes), right?

Knowing the megabyte file size tells us how much space a file will take up in storage, but I believe megabytes are much less useful for estimating the resolution of an image.

Filesize is also useful as it shows how much - or how little, compression is applied in the case of lossy filetypes.