I would reach out to one of these 3rd party companies with one of these scanners and just talk to somebody. Even better if you can find one locally to talk to someone personally. If they are a large corporation with evilCorpOverlords, move to the next one to see if you can find a smaller company. Just talk about your project with them. I would be shocked if you didn't find some like minded person that would be willing to help get the scans done for you.
I worked for large post facilities that would most definitely have told you to pound sand. I've also worked for smaller facilities that used the same equipment. We frequently would take on projects like this, especially if we were slow. It made those boring times in between projects much more interesting. Sometimes there was also something we learned in the process that made us even more experience for future projects. I've pushed to accept some of these jobs personally.
At the end of the day, what's the worst that could happen from the conversation? They say no? Volunteer to come in and work with it yourself during off times or other types of ideas. Show them your passion for it, and get them excited for your project. You'll be amazed at what the community will do for others.
I reached out to a few people over a couple of months while testing the scanner, including a couple of third party service/repair shops. There are also dedicated user groups around these scanners that are all happy to help and give tips. They were extremely helpful but the conclusion was indeed that the Firewire port or main board is on the way out.
It is possible to fix it, and the third party service/repair shops do have the parts and expertise (some of them are former Hasselblad service employees). The problem is the cost. To quote: "The problem is when the FireWire port dies the only repair that is ever proven to be effective has been to replace the entire main board. I currently have sufficient parts to replace this should you wish. The price to completely refurbish the scanner and replace the main board is 2800+ VAT "
So 5,000 Euro for the scanner then 3,500 Euro for the service/repair. That's a little bit more than I'm willing to drop on this project, especially given the other factors - it might need repair/service again in 12months time, the discontinued software and needing old OS/computers to run it.
> when the FireWire port dies the only repair that is ever proven to be effective has been to replace the entire main board
You aren't talking to the right person... A circuit board is a set of components and wires... With enough expertise, it's always possible to find and replace the faulty component/wire. Someone who says the whole board needs replacing simply doesn't have the expertise.
It would be like hiring a builder to fix a window frame, and them saying "the only fix is to just knock the house down and buy a new one".
Sometimes the expert time required to find the fault isn't worth it when a whole new board is cheap... But that isn't the case here.
I agree with the thrust of what you're saying and OP should definitely take the board to someone who will do board rework and see what they say.
But I don't agree with:
>With enough expertise, it's always possible to find and replace the faulty component/wire.
That depends on what's happened. If it's an ASIC or something else complicated and application specific that's died, then the thing is probably toast unless you can find another board that died of a different cause.
It really depends on what exactly is broken. If it's a BGA chip, it's much easier to just replace the whole board, because almost no one has the facilities or expertise to replace BGA packages on a board correctly. If it's a connector or a capacitor, you can do that with a soldering iron if you have some skill. If it's some little 14-pin SOIC package (with 0.5mm pin spacing) and you have a hot-air rework tool, it's trivial to replace it. Of course, most hobbyists don't have hot-air rework tools (they're not expensive though).
I think parent poster meant asking them to scan images for you, rather than repairing a scanner. You mention the average cost for that in your post, but maybe going for someone who's doing it as a one-off rather than providing it as regular service could shrink down the price.
A fair guess is mechanical strain on the port. The first one should do is to implement some kind of strain relief, like putting a donor Firewire port on a separate board or plate, like in desktop PCs.
you'd be surprised at how accomodating and chill people are in Switzerland (the writer mentions living there).
i needed a couple prints for a birthday and my local shop was closed, so i went to a business that printed ads, banners, fliers, etc with incredibly expensive professional xerox printers. they printed what i needed, had me pay a fair price (less than the shop), and let me use their facilities for cutting/finishes.
I often got the sense, when living in Switzerland, that if there was a designated process for something it would nearly always be followed to the letter, which at least provided a great deal of certainty (although rarely the process didn't benefit either party). If there wasn't one in place, people seemed to operate using a healthy dose of understanding and common sense. This is true virtually everywhere but to varying degrees, with the two seemingly melded together more in some societies.
Through various driver discs and archived files online I was able to compile some resources and threw them up on Archive.org (available on the FlexColor page of my site).
My biggest gripe is when people don't document fixes. This whole rabbit hole started because Hasselblad removed the software downloads from their site and the rest of the community doesn't really have a clue when it comes to software preservation. Flat CMS's are important to the process as well. It looks like Archive.org picked up on the page because of this post and made a couple snapshots:
https://web.archive.org/web/20230126040120/https://pathar.tl...
> That’s 2500 to 7500.- CHF worth of scans if I get a third party to do it.
This is how I ended up with my first laser printer.
My aunt was asked to publish a limited edition of a book she translated to German. The printer wanted a ~1000 dpi film and we got quotes on how much it'd cost to get the protolith done. We then concluded a good laser printer that could print on transparencies would be far cheaper and fit our budget, so we asked if that would work for the printer. When they said it would, we got the printer, I tweaked the halftoning a bit and off we went.
In the end, the German translation we did was perceivably better quality than the professionally typeset Portuguese version. And I got a nice laser printer.
I know a few students who bought those big business lasers and 3rd party toner to print pirated textbooks. This was before good tablets with stylus support were commonplace. Several books would already pay for the machine.
I came very close to buying one of those large car-sized print systems. I thought about it because a graphics shop was upgrading and the printer's brains were very close to a Dandelion or Daybreak Xerox workstation (same UI elements, similar boot sequence, same keyboard and mouse), but, after a closer look, it seemed to be impossible to extricate the brain from the rest of the body and, so, the whole machine went to someone who'd love it, cherish it, and print with it (instead of me, who'd kill it, sell the organs and keep the brain alive in a jar).
I would advise the author to just buy the scanner, use it to scan everything, then sell it on. If you make a loss on resale, just consider it the cost of renting a very high quality scanner, but of course there's a significant chance you'll make a profit.
this is good advice. I've done it with everything from weird machine tools (tool and cutter grinder) to woodworking tools to snowboards to guitars. if you don't treat big purchases as permanent, you'd be amazed at what money you can make and what items you can use and enjoy for just the opportunity cost of your capital.
I have a friend who buys lots of used stuff, and sells it the moment he stops using it.
He's extremely patient and waits for a good deal, and usually makes money on the resale.
It lets him go quite deep into a niche hobby without breaking the bank, then when he inevitably gets bored and picks up another hobby it's not a waste.
It's honestly hard for me to sometimes wire my brain to think of total cost of ownership as taking into account the fact that durable goods can really maintain their value. I feel like I've been so conditioned to either hoard things, or worse, just think that they have no resale value from having been purchased once, and that's simply not true.
When you remember that there are decent ways to resell things, and that plenty of classes of hobbyist goods and professional tools maintain their value extremely well, it really changes the entire equation.
Getting into amateur radio had had me do this re-think. The Icom IC-7300 has apparently been the best-selling HF transceiver for several years now (think Toyota Corolla for relative cost and performance), but even so, numbers in the low tens of thousands of units sold since it was introduced. They're 1200-1300 EUR new, and 800-900 EUR several years used. the other models are even more niche.
It is so common for amateur radio enthusiasts to re-sell (and re-re-sell) gear that swap meets are still a central activity of their meetings and conventions.
In most cases yes, but the author said this one was showing signs of the firewire port failing. If it dies in your hands that really sucks financially.
Hardware is repairable, difficulty depends on the mode of failure but if the port itself is faulty that's a pretty easy fix most of the time as long as you can get your hands on a suitable replacement which are (usually) pretty cheap
> The software that drives these scanners is compiled for 32bit architecture and hasn’t been updated in well over 10 years
Might be worth it to try using Wine to run the windows software if they have windows drivers. Wine should run on mac if that's your poison.
> It seems the Firewire ports on these things start to go bad after about a decade. If that happens you’re looking at a 3,000 Euro repair bill
A soldering iron is cheap, and so are firewire ports on digikey. I doubt they invented their own proprietary firewire connection. Ferrari's use volvo parts, you can probably fix it for like $3.
Does userland driver software, which probably wants to make all sorts of weird syscalls to communicate with hardware, tend to actually work under Wine?
I'll second this. I bought some ridiculous behemoth of a Fujitsu scanner that could do 60ppm 20 years ago when it was manufactured and got it working under SANE in Linux. I reported a bug that they fixed to make it work (it didn't initially) I think because it had really old firmware they'd never seen before.
I have so far successfully: used eeprom programmers, flashed a multi of android software, interfaced with a PLC from under wine with no problems or workarounds whatsoever
People have done it with vmware ESXi and passthrough. But you can't do it with say, Virtualbox. So with Firewire, real OS on the metal is usually the answer.
The state of dedicated film scanning is so bad these days, especially for larger than 35mm formats that it seems most people still shooting film are resorting to using a digital camera and a macro lens with something to hold the film.
You're correct. Most people who do their own scans either use a macro lens or a telephoto with extension tubes to create a macro lens of sorts. Then you need a negative holder such as this one: https://www.negative.supply/shop-all/basic-film-carrier-35. You need a light table or LED hot light, and a way to hold the camera and you're in business.
Personally, I've been scanning my own 35mm negatives for the past few years with both a Full-frame DSLR and an APS-C Mirrorless camera, and get much much better results than I'd been getting via my local film processing lab and their scanning.
Scanning negatives with a modern digital camera still retains a lot of the filmic quality that we've come to expect via shooting film. Obviously you lose the "resolution" and ability to blow those images up once scanned, but you can 100% still see the grain, color reproduction, and analog anomalies that we're used to by shooting film.
Yeah, I have recently switched to "scanning" my 6x6cm and 35mm negatives with a macro lens and digital camera. It's probably not as flat out perfect as a really good scanner, but the end results look amazing and fit my workflow for digital images as well.
I recently got a camera with "pixel shift" technology and using that the files have 96mb and really show off the grain. Plus it's just faster than scanning ever was. I am really very happy with the setup.
You don't "need" much tbh. My setup is a cardboard tube, a lens I found in a broken scanner and some tape. I'm using an old kindle as backlight since it doesn't have pixels = no need for diffusion.
If you don't use a planar lens it will never really look good.
What I do for my part is just use an enlarger to which I attach my digital camera with various tubes and adapters to get the distances right, and an enlarger lens. These lenses are infinitely cheaper than "regular" planar lenses and they're just made exactly for the job, and the whole assembly is rigid and aligned so it's easy to get the whole negative right.
Scanning medium format can be a pain but I've had decent results with an Epson V700 for 4x5 film. A lot more cost effective than a Flextight and I maintain some VMs with the firmware on various operating systems that have traveled with me through various desktop builds.
I'd like to see some better FOSS options but images are a lot better quality than they were using cheap desktop film scanners in 2005.
I shoot lots of 120mm on a Mamiya 7 and am currently stuck with an Epson V600 as my best option. Unfortunately, the scanning experience sucks for 6x7 photos because of how not-long the bed is. You can only scan two 6x7 photos at a time, and then have to turn the film around. It's a tedious process that means most of my film remains unscanned.
As others have mentioned here I am also looking into trying to fauxscan with a DSLR and a macro lens, but I'm also wondering if there is enough of a market/interest in a more DIY/FOSS scanning project.
I've considered trying to build something more standard that could be printed or manufactured, that combines some small pointgrey (CV) cameras and macro lens, a light, and some film rolling mechanism to keep the film tight - the idea being that you could basically "scan" the film as a moving roll, scanning a whole 120mm roll in seconds. Pointgrey cameras also have a great api, so you could actually build software on top of it.
I've long thought about putting time to this, but the intersection of pro film shooters and programmers is so small that I'm not sure it would be worth it. If anyone else is interested in the idea, I'd be happy to collaborate on it!
Alternatively, the state of digital cameras is so good these days that they can be used to digitize film. https://www.youtube.com/watch?v=6pnqF8G_wiM highlights a setup with a Panasonic G9 that features an 80MP in-camera high resolution mode. A used body can be found as low as $500 in the USA, and it's a camera too!
I wet scan on a flatbed, though I will admit the process is tedious and holding a negative up to a lamp and taking a snapshot with my phone is 80% as good ;)
Wet scanning was the default for drum scanners, the gold standard in the 90s. Essentially, you lay a drop of mineral oil on the glass drum, create a bond with the negative and then tape it down for good measure since the drum spun at a pretty good clip.
Wet scanning is worthwhile if you have a higher-end flatbed scanner like the Epson V700, 750 and 900(?) because they use higher-quality sensors and real lenses.
I almost bought a drum scanner a decade ago but would have to maintain a Mac Quadra for it to run. For large format film, a wet mount scan on an Epson V750 with VueScan software is great.
This is a good point. Speaking of points, in the way a laser turntable pickup is noisier than a needle, would an electron microscope scan of a frame of film be best, or is there a sweet spot where the grain noise is at an optimal minimum?
Another one is that nobody made a feeder for movie film, for ANY of the consumer or "prosumer" scanners at any price.
If you want film scanned today to individual files per frame, it's still a bunch of money if you can find someone to do it... and good luck getting it done right.
The movie film scanners that are top notch are easily in the six figure range, and go up depending on the features. In a past life, I was researching a scanner to buy for film restoration purposes, and I can't remember the name of the company or unit, but it would move each frame into place and then "gently" press down to flatten the film, and then triple flash it to monochrome CCD with the appropriate RGB filter. It was a big machine and with the flattening and triple flashing was not real time capable, but not as slow as I would have expected. Somewhere between 12-18fps for 2K and slowed down for 4K+ scanning. The project was never funded, so it was all for naught, but I enjoy getting to research new toys like that while on the clock so to speak.
The Northlight is definitely one of the units we looked at, and was a top contender. Ultimately, it was between the Northlight and a second vendor, but funding was pulled before we ever got to make a decision. So, no new toys for me and the team =(
I understand (I work in the industry), but the point still stands that these companies could have added greatly to the appeal of their scanners by offering a motion-picture attachment, no matter how slow.
I was considering trying to build one out of an inspection microscope, but gave it up and just had my most-important reels scanned to TIFFs.
>I was considering trying to build one out of an inspection microscope, but gave it up
justifying the cost of the people that did not give up and made an actual product. "if this shit was easy, everyone would do it!" is something i remind myself all the time
Sure. I don't begrudge the cost of the devices. In this case I'm calling out the missed opportunity of having built most of such a device already and not following through to expand its appeal to more potential customers by offering a particular accessory.
Also the opposing viewpoint to remind yourself of is that hey, SOMEBODY was the first to make one of these things and started a company to do it! I've given up on ideas before starting because of the manufacturing technology required, only to see a company like GoPro come out of nowhere and own a market segment.
Cool! Thanks for the link. I checked out a few DIY projects years ago, and even bought a few broken Super-8 projectors to scrounge their film-advance mechanisms. But I haven't revisited the state of the DIY art for a long time.
> It seems the Firewire ports on these things start to go bad after about a decade. If that happens you’re looking at a 3,000 Euro repair bill if you can find someone with the parts capable of replacing them. Hasselblad will still repair them but it’s a pain to get the scanner to them and they charge almost twice as much as third part service shops.
...
> This thing is remarkable. The bigger brother (the X5) even more so. Imacon/Hasselblad had a load of patents on the technology that means no other manufacturer can replicate it. There are other high spec scanning solutions, of course, but none that come close to this form factor.
...
> 15 years passed, at which point Hasselblad discontinued the scanners. The cost of modernising the interfaces was not worth it. In fact, only 7 years passed before Hasselblad effectively discontinued them as that was when they stopped updating the software.
This is the dark side of patents.
A company that has used the patent system to run everyone else out of business despite not being particularly innovative (the author describes a "simple" system for assuring the film is perfectly flat), refuses to keep their product line up-to-date or properly support it, charges a fortune for a service that from the sounds of it doesn't actually fix the problem with the interfaces, which may have been purposefully designed to fail anyway...and an entire market segment just dies.
It's sad that those $3000 bills for repair will probably be going to organizations like museums trying to preserve their collections or make them more accessible. Or be unaffordable to such organizations.
Yep, I deal with this all the time in industrial automation. It's even worse than the consumer space there, because the cost of breaking stuff is so high that the default is to invent something and then coast on your laurels for a few decades. Infrastructure and tooling that somebody's dad bought 30 years ago is now too expensive to update, so you keep it limping along well past its normal service life...
In particular, the author laments the high cost of scans:
> Here in Switzerland I am looking at a cost of around 30.- CHF a frame minimum if I get a third party to scan them for me.
And worries that you have to have something like the 2007 iMac to run it:
> So you need a dedicated old rig to run the proprietary software to drive the scanner.
But doesn't seem to quite connect 2 and 2 together: Those third party shops all have ancient dedicated rigs that they bought a decade or two ago. The use of that equipment is what your $30 per scan is buying! There are label printers and laser markers and time clocks and press brakes and CNCs and inventory management systems at shops all over the world, running some PLC RTOS from the 90s, or DOS or Windows 98 with PXI cards or printer parallel ports or even completely proprietary logic boards, that are decades old and you can't get parts or service or updates for them.
It's ubiquitous in any less profitable industries that aren't on the cutting edge.
It's hard to maintain and gradually update this equipment. I fear it's going to get harder as IIoT services die off and engineered lifetimes shrink. It's hard to debug a machine built before the Internet, with coffee-stained schematics and only enough IDE hard drive space for 8-character variable names, but I worry it will be harder still to debug a machine built well after the Internet with ubiquitous documentation available only behind a login to a server that's no longer online and enough hard drive space for gigabytes of third-party libraries...
I mentioned this in another comment, but sometimes people who are dealing with these systems just don't dive in far enough either. I was able to get my FlexTight scanner working on Windows 11 64-bit by modifying the INF for some existing 64-bit drivers. For some reason nobody had tried this before, or even just documented it. Even if there were no 64-bit drivers I've gotten it to work on a Windows 10 32-bit install.
Especially in the photo world there's so much tech being used, but not a lot of people that understand the implementation/engineering of it. Not faulting them, but I can't tell you how many people I know that just work off external drives. No NAS, no backups, just originals spread across fragile external hard drives. Hell, I knew some people that up until ~2011 still stored originals on ZIP disks and never managed to move them to something a bit more stable.
Going back and working on older Macs is an absolute pain. Now there's four separate architectures that may or may not have the right software/driver at varying levels of implementation. It's amazing how often the answer is "just use Windows". Not that it's perfect, but at least you weren't dumped like last week's spaghetti.
I dove in pretty far - to the point of running dtrace on the flexcolor software, and it seemed to get stuck in a thread that was receiving data from the scanner. The Firewire port is dying (well, now it seems to be dead) so this is not a software issue and more a "i'm not prepared to drop 5,000 Euro on something that may not be fixable".
They might have a patent on this particular setup, but the FlexTight system produces worse results compared to traditional drum scanning. Due to the way it uses a lens between the CCD sensor and the drum also means that you get some quality issues around the edges of larger frames. It also means you can't scan large formats like 8x10. It's cool tech and it is a shame that it's stuck behind patents, but it's not like there's not other options out there.
Also the hardware is funny. IIRC they didn't iterate on the actual components much, and the move to FireWire basically just integrated a SCSI->FireWire adapter into the case. I haven't seen the inside of an X1/X5, but I'd be interested if these could be repaired by using a different SCSI adapter of some type.
I don't know anything about scanning film, but perhaps a microscope scanner could be used to do that? 8000ppi is nothing for those, there you're looking at ~10 pixels per micron (see
https://cancer.digitalslidearchive.org/#!/CDSA/acc/TCGA-OR-A...
). If you can fix the film somehow to a pathology glass slide (75x26mm) you could load 100s in the machine and have it scanned overnight. It will be multiple GBs per slide though.
They do cost about 2x-20x as much but they are fairly ubiquitous in research hospitals or universities. Some (like Philips) also have annoying software so people are getting rid of them for cheap
which is odd, as I've been playing with home LED lighting searching for non-Hue lights. not one of them and their accompanying software can hold a candle to what the Philips Hue app and control of the lights can do. Hue may be more expensive, but the hassle free and ease of use goes a long way.
The relatively smaller no. of people would put up with the quirks of speciality wonky software as long as it get the job done somehow, but the much larger no. of average consumers won't put up with a buggy software that fails to make the bulb do what they want. So better UI, better testing and so on.
Not familiar with Philips, but other microscope scanners I used, Aperio, Leica, Zeiss, all have dedicated computer attach to them, with special, sometimes strange looking interface card. They have precision stage, which holds glass slides, move under lens. USB is becoming common especially for transferring image. The software and driver is part of the scanner. Resolution of these scanner is quite high, but generally their lens (microscope objective) have very short work distance to the slide, usually less than 2mm, sometimes less than 1mm. So need to find a way to keep the film flat on the supporting glass slide while the distance between film and objective is still short. The output is limited to the vendor format. Most of the time is 8bits per RGB channel. An annoying character of the microscope scanners is their color is not accurate. One might get notable different color from the same slide with different scanner model. Therefor they may not suitable for scanning film.
I spent months scanning about 2500 family negatives and slides on an Epson Perfection V600 photo scanner.
While it's no FlexTight, I am happy with the results, especially because I had no plans to crop.
In hindsight, I wished I had used SilverFast rather than the Epson scanning software. SilverFast offers Multi-Exposure which does two scans for maximum dynamic range and then merges them into one.
Also, the Epson default film holders have no ability to flatten the film strips so I probably ended up with softer images in many cases. I believe there are 3rd party adapters that address this.
The V600 is better than most and has a CCD sensor instead of an inferior CIS, but the limiting factor is the optics. The V700/V750/V800/V850 have proper lenses.
The old-school way to scan film is to lay the film directly on glass with a thin layer of mineral oil. Works terrific for medium and large format. We used to do this for drum scanners.
I also use an Epson Perfection (V750 Photo). These machines produce very good results if you are careful with film positioning. I still haven't figured out a way to scan really old negatives in rolls: despite buying several magic holders, I've yet to find one that can oppose the force of a nearly 80-year old film roll.
I would not recommend third-party software, though. The problem with scanner software is that every developer seems to think that I have unlimited time to tweak the settings for every scan, and that scanning those 5 negatives is my only job for the next month. That might be the case for some people, but trust me, if you're looking at several thousand scans, you do not want to tweak each one individually. You want software that works with you. And so far every third party program I tried did not have this approach.
I have the same scanner. Would you mind sharing your setup? I've only used mine for flat-bed document scanning, but do have a bunch of old family slides.
Do you have the illuminated back? If so, your cover is maybe 4" (10cm) thick.
I have the v750 and scan medium and large format film with excellent results.
You can buy Viewscan software from https://www.hamrick.com, made by a NASA JPL engineer, if I remember correctly.
You get a perpetual license (I reinstalled after 12 years and the updated software ran on my old license code). The software can pull detail out of even the worst negative. I cannot recommend this highly enough! I do have SilverFast as well, and have tried the Mac-native scanning, but Viewscan is the best.
If you have slides and you haven't kept the original film adapters, you can find them on ebay or even Amazon. They are simple plastic holders, nothing special.
You can usually select a particular film type if you scan color negatives that will automatically color correct for the film quirks. All in all, a very easy process once you get set up.
I'm flabbergasted, yeah, it does seem to have the illuminated back. I never looked far enough to see that the mat covering the inside of the lid was removable. I don't recall film holders at all, but will definitely have to look into getting some.
I bought this thing when we moved, to get set up ASAP for a remote home office, so documents and some photos have always been the purpose. I had taken the slides in a box to a photo place, where they used some setup with a DSLR to get JPEGs, but now I'll need to dig the box back out. Thanks!
+1 for a cool title, and great choice of a camera. I once used a friend's Hasselblad in the Scottish highlands, and the photo looked almost nicer than reality.
A refusal to share the source code of a driver in the 1970s (on the side of Xerox Corp.) angered RMS enough back then to start the Free Software Foundation, and the rest is history, after all. So let's see what your solution will be...
What you could do is buy the scanner and after your project offer others to scan their slides to get some of the money back. Or team up with others and split the cost of the scanner upfront (this latter scheme requires someone to hold the physical device, I think donating it to a library after the project would be a fair mechanism, so each party - and others - can still use it later).
I wonder if the company would be willing to grant a patent license to an open hardware version of the scanner. They’re not manufacturing them anymore, nobody else seems interested in making the damn things… let it go.
Sounds like maintaining these high end medium format cameras and scanners can be a real hassel.
I have a Nikon Coolscan V and nothing is wrong with it but the switch, but it’s hardly worth repairing or trying to sell on eBay. Working, it might be worth more than I paid for in c2005.
Maintaining high end medium format cameras is really not bad at all -- most are very simple and many have no electronics, so as long as the shutter mechanism continues to work, everything else is a piece of cake.
A C>oolScan V is worth a fair amount of money already, and will only increase in price. Send it to Frank Philips, who repairs CoolScans more or less full time. You can find him on the FaceBook CoolScan group
I had a Nikon Coolscan 9000 that was the best scanner I ever had for 35mm and 120mm film. I worked professionally and would submit perfect scans to clients and retouchers that were far better than anything the labs were producing.
I bought a Pakon (Dogbowl variant) and modded an old Macbook Air to run it. It was great.
But, again and again, I found myself running up against entropy of time and software. It was always going to fail. I miss my Rolleiflex, I miss my Leica, I miss my Mamiya RZ and my Mamiya 7II. I miss film and writing shot notes in my notebook.
But the manufacturers and the labs are all going in one direction, digital, and the price of film and development is rising.
Since I've moved to Amsterdam from NYC I've looked at darkrooms/dokas but it's all too hobbyist and caught in an older era. Reluctantly, I've gone all digital and adapted my workflow to make it less painful: profiles, curves, mobile cloud syncing, batch treatments, decent historical archiving.
I've found a great digital printer and a great print shop, and now the only real issue if finding a good framer.
I used to work in a photo studio in Germany, we had 2 of these and the predecessor to the Hasselblad/Imacon X1. The difference between them was amazing. I'd still love to have one of those Coolscan 9000s, they are great machines. Even back then (around 2004) we used Vuescan for the Nikons. It was already an issue to get up to date drivers for our Macs.
The Flextight is definitely better than the CoolScan, but not approaching drum scan quality. A very good CCD scanner with a great enlarger-quality lens, and the transport is very good at keeping the film flat along the line scanned by the CCD, by the expedient of bending it cylindrically, but it's not the same dynamic range as the photomultipler tube on a drum scanner, or even what the old $50K prepress X-Y flatbed scanners like a Creo Scitex or Fuji Lanovia could do.
Oh I know/agree. It is literally described as a virtual drum scanner by B&H for example though, as inaccurate as that is. To me the main difference is CCD vs. photomultipler.
The X1 was probably better, as it was a [almost] drum scanner.
But scanner tech seems to have basically "hit a wall," in the last decade, or so. Not many advancements in the imaging. I think digicams and pure digital images/documents have made it difficult to justify the cost of developing them.
VueScan did not support Digital ICE4, which was the only way to scan Kodachromes with automatic dust removal using the infrared channel (Kodachrome is opaque to the infrared LED used on the Coolscan V ED or 5000 ED). Only Nikon Scan did, and possibly SilverFast (Digital ICE was invented by Applied Science Fiction, a subsidiary of Kodak).
I have a bunch of 120 film that I'd like to get better scans of someday. The old Flextight scanners were on my radar, as well as the Coolscan 9000. I still wonder what the best way to scan these old negatives is. I have some scans done with one of the Epson flatbeds, but they're limited in what you can get out of them. I've also taken some shots to get drum scanned, which gives fantastic results, but isn't justifiable given the quantity of film I want to scan.
I've considered trying to get an old drum scanner and learning to do it myself but it would require some dedicated space which I don't have right now.
A lot of these shots were taken carefully, with good exposure, on a tripod, good focus, correct aperture, and slow film (by modern standards). There's an enormous amount of detail in some of these negatives which just doesn't show up in most scans. On optical prints I can even count the stitches in people's clothing from full-body portraits, if I look at the print with a loupe.
I had hoped that film scanners would get cheaper as time went on, and some day I would just be able to buy a nice scanner and just plow through my film. Seems like my best hope is for somebody to make a jig where I can connect a digital camera and use that as a makeshift scanner--I know these jigs exist, but MF film is still a bit of a beast.
Yes, but the used scanners are at least 10 years old at this point, have precision mechanical parts like steppers and gear trains that will wear out eventually. If it was (hard) used by a service bureaum as is quite possible, it probably has few cycles left.
I’m helping my dad with a SciTex scanner – I think it’s a SciTex EverSmart Supreme? It’s a similar proposition in many ways as far as I know. High-end professional equipment from the golden age of digital prepress. Firewire is involved. Will essentially only run on old Macs. With the SciTex scanners I believe a PowerPC Mac is a requirement in practice.
From reading the post, the biggest issue with these Hasselblad scanners is of course the scanner mainboard, and I assume the same is true of the Scitexes.
A smaller problem is needing to have old Macs around. Regarding that, I am now curious if SCSI/Firewire controller passthrough into a VM can help. In the same way that recent PC hardware can pass a PCI-Express GPU into a virtualized macOS guest. MacOS VMs in qemu-kvm on recent Linux kernels running on IOMMU-enabled hardware affords a lot of control and compatibility – the guest OS gets direct control of a physical PCI-Express device, and it works.
Doesn’t solve the mainboard problem on the Hasselblad side of course. And doesn’t completely solve the old-mac-hardware problem on the other side either. But it might reduce the hardware dependency on the Mac side into just a SCSI or Firewire controller card instead of a whole old Mac.
Have you had the chance to look into this side of it enough to give any hints about it?
And/or can I be useful in any way? Have set up macOS VMs with physical GPU passthrough and have working prepress experience with Macs ranging back to a Macintosh Plus :)
The post as it is is already immensely valuable insight into the whole ancient-prehistoric-digital-prepress world my father and I sometimes burrow a little into – many thanks!
You are actually on the money. We do this at my company. With much more challenging scanners than the Flextight. We've virtualised the Fujifilm SP3000 this was as well the the Noritsu S-1700 (which is a rarer precursor to the HS-1800 with very similar specs).
> Hasselblad refuse to update it to modern architectures, and refuse to open source it to allow others to do that.
Hasselblad are idiots then.
You could try to find more owners of that hardware and chip up together to hire a reverse-engineer to make an open source driver. Compared with the cost of hardware and complexity of using old computers to interact with it, it should be negligible.
While this is not one of the 7200 scanners claimed by VueScan[1], it might be worth talking to those folks. With their experience they might be able to reverse-engineer the protocol.
If not (and I'm sure this is not an acceptable idea to a purist like OP) I would consider using a good consumer-grade scanner and post-process with Topaz AI[2], which produces absolutely astonishing results in my experience. Yes, maybe it is inventing the pixels it adds, but they look like the right pixels nonetheless.
I had a Nikon Super Coolscan 9000ED that I ended up selling off and still have my CoolScan 5000ED with the slide and strip film autofeeder when I finally get around to scanning my father's old slides and negatives in my copious spare time™. The 9000 could scan a X-Pan slide in a single pass, the 5000ED in two scans with stitching. I had dedicated a first-generation MacBook Pro to run the software as they had a FireWire port, using BootCamp to run Windows on it as the Windows version of Nikon Scan was more stable.
Almost all the proper film scanners were discontinued around 2010, that the Flextight series survived another decade is amazing, but Imacon bought Hasselblad (not the other way around) and now DJI has bought the combined entity, and obviously has no interest in the legacy of film.
So someone needs to get the idea into their heads that lugging medium-format cameras around would be the perfect excuse why people need to buy larger and more expensive drones
Well, the post-war Hasselblad medium format system started as an aerial reconnaissance photo system, as did lenses like the legendary 38mm Zeiss Biogon in the SWC.
Brought back “Fond” memories of my Polaroid Sprintscan 120 (4000dpi) [0] which could handle negs of up to 6x9 size. It was a slow, noisy beast, but gave great scans.
Around the turn of the century I used it thousands of 120 film (6x6cm) negatives shot from my beautiful 1980s Plaubel Makina 67.
I’ve often thought of resuscitating the camera from its pelican box sarcophagus but the idea of scanning the negatives gives me chills.
I would love for the hacker crowd to start looking into repurposing old scanner technology.
I am waiting EAGERLY for the first open source scanner made for negatives. It's so sorely missed, and so far i'm on my third flatbed and not happy at all. It's silly when the line-scanner is the same for all of them more or less.
Someone should really do a reverse engineering on some of it.
Please consider not calling it "free software". I've tried to explain to RMS that it's terrible communication, but I think he likes the wordplay.
The term "free" product already meant something to people, and continues to mean that.
There is generally no opportunity to (as RMS might imagine) say "I'm glad you asked about the name. We mean free as in freedom. Let me assemble my lecture podium and explain..."
Instead, people who hear the term have been hearing the opposite of the intent.
Which leads us back ot the current problem, of people thinking it's all random hookups with open source, rather than meaningful relationships.
I don't think it's a naming problem, I think libre software has been under direct attack on a similar angle to "embrace, extend and extinguish".
But yeah I agree naming doesn't help. I'm biased because in my language libre and gratis are distinct. I'll keep that in mind, even though I'm a bit pessimistic and don't think it will help, I guess it won't hurt either.
Agreed that naming isn't the only problem. It's a complex environment of different understandings and goals, including adversarial dynamics (and sometimes very underhanded behavior, including EEE).
I think the foot-dragging by some on the name, after it's been pointed out to them, is kinda symptomatic of their difficulty operating in this environment. And they often end up discrediting action that could be viable, but needs to win buy-in from others.
Stallman is an odd character. He's deeply passionate about the FOSS movement, and its success means more to him than anything except his jargon choices. He'd rather see the whole thing burn than change the words he demands it be described with.
That's a bit unfair. He passionately advocates for his choice of language because he thinks it's important to the future of free software. You might disagree with him, but he's not exactly shy about justifying his opinions.
Software freedom didn't really disappear from public discourse, it never appeared in the first place, unless you count amongst a bunch of nerds nobody in the real world cares about or listens to. It may sound harsh, but that's reality.
Open source appeared in the public discourse because it had tangible benefits and people could understand what it meant.
I don't think open source killed the concept of software freedom, I think the concept of software freedom as a concept literally never caught on to begin with.
Honestly, I'm a bit tired of the Stallman worship at this point. A lot of people were pessimistic about the future of software and the internet, and they're not all prophets. He observed what was happening when other people chose to naively ignore it and hope for the best, and that's not nothing, but frankly, if he was so smart, then you'd think he'd have a solution. Nonetheless, it seems like everything Stallman actually proposes is impractical and out-of-touch with the current state-of-affairs, which makes me wonder what we're supposed to do other than marvel at the fact that someone was able to reason that this would become an issue some day.
Open source (OSI definition) is exactly the same thing as Free Software (FSF definition) so your comment makes no sense without clarifying further. Are you perhaps referring to Source Available software vs OS/FS?
The problem is in messaging. The whole point of the OSI was that the FSF was too ideological to be appealing to most businesses. The problem, is that the ideological angle is the point. The OSI would have never pitched the GPL, and the FSF would have never pitched the MIT license.
When you release software under a license that makes approximately zero demands of the other party, the other party won't respect your software. When you release software under a license that makes even low effort demands of the user, like reciprocation, they will.
The MIT license is a FSF certified "Free Software" license, it provides all of the "Essential Freedoms". The MIT is not "copyleft" like GPL is but copyleft is not the same thing as free software.
Once again, that's not the point. The ideology required to promote copyleft licenses at all is the point. The OSI's standpoint is that you can distribute software that provides the user with their freedom. The FSF's standpoint is that you are ethically obligated to.
Stallman made up his own definition of open source which is different than the common definition. The OSI best sums it up here https://opensource.org/osd their definition is essentially the same as the FSF essential freedoms. Without classifying "Stallman's definition of Open Source" people will assume something much like the OSI one.
What Stallman calls Open Source is what the rest of the world calls Source Available. Stallman shows his lack of communication skills again here by endlessly fighting to redefine words for no purpose.
Stallman appears to be considering "source available" licenses which forbid modification (and sometimes deployment) as "open source." This is not the definition of open source that everyone else uses. Certainly the OSI would not agree. There is more information here: https://opensource.org/osd
RMS was around in the 1980s when "open source" was still being defined. So maybe he thinks he can define it however he likes. But to the rest of us it doesn't make sense, like saying "software that has a free software license, like GNU Emacs and Microsoft Word." One of those two doesn't belong in that sentence, and saying that I'm using my own custom definition of "free software" is a pretty shitty way to argue.
In this case I don't think it would really matter. An MIT-licensed driver/app to run this scanner would be perfectly acceptable and useful to the community.
Copyleft just ensures that other parties can't take someone's open-source work and close it up and profit off of it. I don't think copyleft is really necessary for when a corporation open-sources some software essential to the use of one of their products.
The reason is probably that there are way way less devices in use now. Graphics cards are made by 2+1 companies (nvidia, amd and somewhat intel), webcams are uvc compatible, mice and keyboards are "driverless", there are less and less printers and scanners in use, and even those are wireless, noone uses digital cameras with need for gphoto2 anymore (either mobile phones or pro equipment and a card reader), flash drivers are all the same, soundcards are mostly onboard now, screens are all the same, etc... so no critical mass to bother companies anymore.
Back in the time you had a bunch of soundcars, palm pilots with activesync, lpt printers, usb printers, scsi printers, zip, jazz drives, etc.
The only pain currently are unstable wifi drivers, because there are still quite a few chipset makers) and well.. custom equipment used by a handful of people, such as op here.
There are less external devices you’d plug in, but internal hardware is still a huge battle. Internal laptop webcams usually don’t work on day one, finger print scanners usually don’t work, suspend and sleep is big problem for most new laptops.
This is nice-sounding, but ultimately there is almost no incentive for hardware manufacturers to open source their drivers. Linux isn't a large market and macOS and Windows users generally aren't concerned with it. On top of that, many of these products (though maybe not the one in particular) is basically disposable.
The way this is handled in industry is to have 20-year-old computer running Windows XP that is just for a specific product that only has drivers for one version of Windows.
If more people committed to open source, there would be more pressure on hardware vendors.
I suppose all that's needed is for a good chunk of Windows and Mac users to stop thinking of open source as only Vegas flings of opportunity, and more like a life partner.
Linux is even worse at maintaining compatibility with old drivers. Opening your source doesn't help when the OS doesn't offer a stable API; you can try to get your driver into the kernel tree, but that takes forever, has high code quality standards, and can still lead to you getting rejected essentially arbitrarily. So you can publish a driver that will work on linux version xyz, but that requires udev and systemd and glibc and everything else to be of versions that match.
"Should" or not, the reality is it's rare. Rejecting low-quality hardware drivers tends to mean ending up with no drivers rather than with high-quality ones.
but in linux it does not matter as much, you have the source, it is unpleasant to update it to a different api but not actually that hard, much easier than when all you have is a binary blob that you want to get running on your system.
Has this ever worked? Even now it seems that hardly any vendors put out linux software and when they do it's garbage that tries to import shared libraries that existed on ubuntu 13.04 or something. Almost all the hardware support on linux came from people in the linux space reverse engineering and developing those drivers which are then in a state that can be upstreamed.
The wine/proton project has had far far more impact than decades of posting on game forums asking for a linux build.
> Almost all the hardware support on linux came from people in the linux space reverse engineering and developing those drivers which are then in a state that can be upstreamed.
Almost all sounds like exaggeration, Intel&AMD have done a lot, and even ARM vendors have contributed lot of drivers. See for example LWN author stats for employers, plenty of HW vendors there: https://lwn.net/Articles/902854/
Yep, it really makes me curious what the best DPI you can get on a consumer negative scanner is. It's pretty obvious from that comparison that it's a low-dpi scan upscaled to the same size for the sake of comparing them. But surely there are a few high end models that can do reasonable looking 3200 or 6400 DPI scans?
I wonder what kind of results modern upscaling software could achieve with that less detailed scan. I've seen much more impressive improvements than what it would take in this case, and I reckon the end result could look even better than the Hasselblad scan.
I'm sure film purists would scoff at this, but there's a large market that would prefer cheaper solutions that do this in software. We're already used to these enhancements in modern cameras, so why not scanners?
That's just making up stuff, though. I don't think only "film purists" scoff at this. Anything that even remotely has the chance to be used for archival purposes must not be upscaled that way. Not only do you effectively fill in details that simply aren't there, you also don't even have any foolproof way to know which details are made up by upscaling. Worse, some super fancy upscaling might falsify parts of the image that are still above the threshold of where you would need upscaling.
Imagine a historian going through old photos (potentially ones that were previously deemed uninteresting, even) and is led astray by made up details. Imagine a criminal investigation doing the same.
In short, unless you have very specific purposes in mind, and always be careful to clearly mark upscaled images as such and to keep the originals, upscaling film photos that way is a terrible idea.
Before anyone asks, I do think there is some value to upscaling of old photos and videos, like the popular HD/4k 60fps upscaling of videos from the 1800s: It gives even casual viewers a better sense of what the actual scene, and the life depicted within, might have been at that period. It's easier to accept this as "more real" (even though it's ironically less real).
But I really wish any such upscaling was very clearly marked as such, and pointed out to even casual viewers with disclaimers: Unlike the old choppy, blurry, low-res video, this is not a sample of reality back then anymore, but a speculative reconstruction.
To be fair upscaling with videos isn't necessarily making information up, but depending on the method used may use information from other frames, wich I think isn't the same as speculative reconstruction for photos.
For increasing the frame rate for videos this depends on the model used, but in general most of them don't add any new information, but instead calculate how something looked in-between two frames. https://medium.com/axinc-ai/flavr-a-machine-learning-model-t...
Definitely! I was explicitly talking about things like "AI upscaling". Not simple interpolation (which keeps the image spectrum the same and is effectively just anti-aliasing), for example.
To be honest though, the method you linked does seem to fall in the bucket I am warning against: It does make information up, it does add that made up information. Those frames in between are not simply interpolated, or you would not need an ML model.
Case in point, individual droplets in that example animation may not have existed in reality, or moved the way they did here. But hopefully that's fixed by just dropping the interpolated frames, at least.
It's taking the next step in line with what all digital cameras already do. They all do post-processing and make approximations to what the eye actually sees. Just because upscaling technology is still in its infancy and it's not entirely foolproof yet, is not an indication that it won't become seamless and undetectable in the future.
I agree that such features should be optional and all processing should be clearly marked, but to dismiss the technology altogether is a mistake.
Those upscalers never looks remotely correct to the trained eye beyond minor scaling increases, even with state of the art (I work in post production).
I worked for years for a professional photographer with a vast commercial archive.
After every finished project we would copy the files to a thumb drive and print the top 10-20 images at around 14" long edge >300dpi and place everything in a simple archival box.
The logic being that even if the digital copy becomes unsustainable because of interface change or degrading, you could still scan or photograph the prints.
Most analog prints you see 'digitized' on Instagram are iPhone photographs of prints laid flat. It's all a bit ridiculous.
Paper will last, a 100 or so years is not that hard to achieve, honestly. A bank I worked for had documents dating 140 years and they were just in a box most of the time. They handled it carefully, kept proper moisture in the room, but that was mostly it.
What about the original photo film? Isn't that the ultimate backup in such situation? There will be an option of potentially scanning it with better equipment or skill in future. Like it's done nowadays firm classic analogue movies.
I've recently read a great story of a son of a local artist who found a box of photographic film left behind his relative 80 years ago and it was "relatively well preserved, just sitting there in a box".
This reminds me: please let me know if you found a tape backup solution that is feasible for a small homelab!
afaik even the best color papers (for wet prints) will last 20-50 years before starting to show color shift, that's in a darkbox with optimal humidity. b&w obviously is much better
Modern pigment prints seem to perform a bit better, 65-120 years according to some studies
Color film and photo paper is made with dyes that fade and shift over time. An inkjet print with pigments will last longer. Only monochrome silver metal film or paper will last indefinitely, and it's the gelatin layer that will last, some substrates including consumer-grade acetate film and most papers will degrade. The best archival format is silver on polyester film. Color can still be achieved by way of three exposures in RGB, similar to Technicolor.
Ok, so the old Hasselblad scanners are breaking down and his big box store scanner isn't good enough. The unanswered question I have is why he doesn't buy a modern professional scanner. Modern pro scanners blow past his 8000 dpi benchmark at less than a quarter the cost of the old Hasselblad.
1) optical resolution is not the same as DPI (which I think you know, but its worth pointing out for those people chatting at the back of the class)
2) the way they film is placed on the scanner makes a huge difference to the quality.
the hassblad scanner effectively sucks the film down onto some sort of curved plate, which means that the optics can assume that the film is going to be at position x +- 0.05mm. This is critical because the closer the sensor is to the film, the more the placement affects the quality.
This means that in most highend cases, the film holder has a greater bearing on quality than the sensor/optics.
I have the advantage that at work I have access to a medium format digital camera. This means that I can take really good macro pictures of film negatives. but because the film isn't flat, its a challenge to accurately capture the entire frame. I need to get a proper film gate that slightly tensions the film to make it straight. then I need to worry about getting the camera at 90 degrees to the film.
For example, the Epson Perfection V850 Pro scanner has an optical resolution of 6400dpi for $1300 while doing the full bed, greatly exceeding the X1's 3200dpi in that mode. If you want to match the film scanning performance, the Plustek OpticFilm 8200i SE film scanner has a resolution of 7200dpi for just $400.
PlusTek and Epson are absolutely nowhere near the quality of discontinued scanners like FlexTight, Nikon CoolScan and Fuji Frontier. Modern scanners claim high DPIs, but the real resolution is far below. What good is scanning at 6400 dpi if the lens only resolves at 3000 dpi. I've scanned multiple tests on an Epson 700 myself, and found that there was no visible difference between scanning at full and half resolution.
I have that scanner you're talking about and a Nikon V ED. The Nikon with 'only' 4000dpi massively out-resolves the V850. It's true resolution is somewhere just over 2000dpi and that's after adjusting the trays height.
Because of that I only use the V850 for 4x5 and 120. It's a waste of time for 35mm even if it's quicker to scan multiple frames.
That’s a lot of images for a fine art photography book. Particularly if the images are 60cm on the short side…and at full bleed and through the gutter a book would be about 80cm wide.
It is also a lot for a gallery show at that size — at about 1600mm x 600mm, 75 images would cover the floor of a one bedroom apartment…two hundred would cover a the floor of a spacious three bedroom house.
An editor might be a good way to reduce cost of the goal is a book. And no agent would want an artist to release dozens of images at once.
I wonder if there is an alternate optical solution using a macro lens. Use an even backlight and a dedicated macro lens to take a digital picture of the negative and postprocess. This is essentially what the Hasselblad scanner seems to be doing.
Kinda wish there were more details on this page to understand or interpret what the author is talking about or compared it to.
I have a Canon 9000F (consumer grade) that is getting very old now and often difficult to get running when I pull it out, but it scans film at 2400dpi which has been adequate for my home archival purposes. I don't know if this is in the ballpark of what the author used as the comparison for the consumer scanner (nor is the DPI cited). About 8 years ago I was going to start a new film scanning project and thought "Hey, maybe now's the time to buy an updated film scanner" but I learned there is very little on the consumer market and it's gotten very expensive.
I think the 9000f hasn't been made in a while but I still find it to be a great scanner when I need this.
There are a couple of sample zooms at the very end of the page comparing an X1 (6300dpi) to the best the author could get on a consumer scanner (unspecified dpi, but fuzzy). The range given is 6300dpi-8000dpi for the scanners the author's talking about.
I just bought a Fujitsu ScanSnap S1500 off of eBay recently, and upgrading to macOS Monterey has made it unreadable. Time to downgrade back to Mojave by way of Time Machine, because Apple hates legacy hardware of third party OEMs, never mind its own.
ExactScan Pro supports the S1500M on modern versions of macOS, and is superior to ScanSnap's own software in most ways. Mojave won't help, I had to switch to ExactScan when I "upgraded" to Mojave (although I believe Fujitsu brought back S1500M support in the original ScanSnap software on Mojave by popular demand a few years later).
Treasure it, the S1500M uses a CCD sensor, unlike all subsequent ScanSnaps that use inferior CIS sensors with much worse color fidelity, and even photo-oriented models like the Epson FastFoto series:
It was previously scanning on Mojave for me, but good to know, maybe I can use ExactScan Pro to get it to recognize my S1500M without me having to downgrade. I had stayed on that macOS release for quite some time for 32-bit app support, not to mention updating to Monterey also borked my MBP's ability to recognize NTFS drives, needing me to get Paragon NTFS.
!! I have a S1300, was going to scan stuff soon. Yikes, I didn't think it wouldn't work. Have you tried VueScan software? I have a license for VueScan and can try it out later tonight.
(I also have a S1100 which Fujitsu sent me because a won an online draw!)
At this point I’ve got devices with basically every single usb type and I’ll have to keep cables around for them all. The only one that actually seems to have dropped out of existence is micro usb 3.0 which had the little wing bit sticking out the side you used to get on phones and portable hard drives.
> A Hasselblad Flextight X1. Google the thing for the price if you’re curious, and no I wasn’t going to pay that much, I was going to pay less than half of that (a bargain to be fair).
I'm surprised FireWire port repair is that difficult/expensive. Can't be that much to it, you just need to source the same/similar controller from another (much cheaper!) device, no?
I doubt you'd even need to get a new controller. If the wording in the article about what breaks -- just the port itself -- is accurate, the ports themselves cost a few bucks on a site like DigiKey, and even if you have no experience soldering, you could fix it, working very slowly, carefully, and methodically, over a few hours. (Someone with experience could do it much faster, of course).
Sure, I get that you might be squeamish about opening up a $16,000 device yourself and taking a soldering iron to it, but a port/connector is about as low risk as it gets when you need to repair electronics.
If you don't have a decent soldering iron or any of the related accessories, this "$3k" repair will cost under $100. If you do have one, it will cost you under $10.
I wouldn't suggest a simple fix on an expensive device as a first soldering project. It is not that you need much experience, but it is risky and a little experience removes virtually all of the risk. Either practice on a few throw away boards with a similar part first, or ask someone you know who has relevant experience to do it for you.
Another fun thing (that I or someone I know might have done) is when you do a great job replacing the port with an exact replica but in the process somehow manage to desolder a resistor that's about the size of a small shard of a grain of rice from the board AND simultaneously knock the board so that the resistor goes flying into your carpet so now even though everything else is good the system won't work because you don't have the tools or know-how to locate a replacement.
Heh heh been there, done that. Was reflashing the eeprom on a $800 raid card and sheared off two surface mount resistors while pushing the card back in the slot. That horrible sinking feeling when you already know without checking that you’ve screwed up big time!
So, it's probably not the phyiscal port, but the controller and/or the interfacing. If you read the article fully, it mentions that the only known way to really fix the issue is to completely replace the main board... which would tell me that the "firewire port breaks" is a red-herring and something deeper in the electronics fails over time.
I'd imagine more of it has to do with trying to beat every penny of those desperate to get more use of their $15k+ scanner, and less to do with it being a technically complex task. That's assuming it truly just being a port issue is correct.
It's a niche market, and they're just taking advantage of such niche market, however morally gray it may be.
I made the big mistake of purchasing a "CZUR Book Scanning solution" the year before last. The camera on it is sooo shit that I ended up just resting my phone on top of the thing and getting better results just using it as a lamp.. Sure there is some laser contour detection of pages and then flattening, if you use their windows only software to with it, but the resolution is so low its only good for black and white text only books.
I thought it was interesting how in the crop comparison, the effective low pass filtering on the worse scan makes it look better to my eyes.
Don't get me wrong, the better scan is still much better. The correct way if you want to go for that effect is to get the better scan and then explicitly apply whatever low pass filtering you want, not to low pass filter with a less detailed scan. But I thought it was a neat demonstration. Like squinting your eyes.
Yeah, and I also have to say that my "looks better" likely only applies to seeing a small part of the image magnified to an intense amount, not when viewing the image the part comes from itself.
Still, I thought it was a neat demonstration, that's all. I see "more" on the right side: The weird patterns in the windows become actual reflections (as they are), and the building gains more 3-dimensionality (a part of it is obviously recessed).
I don't have to create 160cm wide prints but I have scanned medium format frames on an Epson v600 flatbed and the results with the right software is more than serviceable. In fact, Nick Carver uses a v800 to scan fairly high quality images he creates big prints from. Although he does go to a drum scanner shop for the mammoth 6x17 prints but it simply goes to show that the return on value is very high on the right consumer grade scanner.
My dad used to produce prints in the 90's with a hybrid digital/film workflow. Scan with a FlexTight -> post in Photoshop -> write back to 4x5 with a LaserGraphics film recorder -> darkroom printing. Using this process he did up to I want to say 40x48 prints.
someone shooting on a vintage hasselblad that still costs used near as much as many modern high tech cameras and printing 5'x2' finished pictures isn't interested in serviceable. serviceable isn't the result they're looking for.
I mean, yes, I did qualify that I don't want the same fidelity from my scans but I also shoot on a Bronica SQ, which is no Hasselblad but was still professional equipment when it was released and is still extremely well respected and gives me great results.
Prints are also extremely forgiving, especially viewed from a normal viewing distance, from personal experience of having made high quality prints of my photos.
But again, you can choose to spend however much you wish to to get the last bit of detail out of things.
The camera is an optical device that along with hasselblad lenses has (many) detectable differences in the pictures taken, along with film choice. Someone shooting with a particular camera aiming for large prints is aiming for high excellence not serviceability.
there are bad cables (shielding etc) but the camera+lens+film is literally a sensor/recording device and not really comparable. Whether you like the thing they're doing is a separate consideration, but the person in the article is not looking for ok or even good results. They want everything the best it can be.
Scanning is almost always the weakest link when it comes to film photography these days and sometimes it's really bad, it's not a "copper vs gold plated cable" scenario, more like a "20$ logitech speakers vs "$2000 semi pro setup" to listen to .flac scenario
You can easily run most Flextight scanners on windows 10. I think you just need to install two version of Flexcolor. I have extensive experience in dealing with old scanners as this is actually my line of work. In our company we have almost over 20 scanners of certain types from various brands: Flextight, Noritsu, and primarily Fuji. We use a lot of KVM virtualisation using linux to passthrough PCI cards to VMs.
I would be interested in hearing more about your experiences running these and possibly some notes how they compare regarding scan quality etc. Do you primarily scan 35 mm?
Yes we primarily deal with 35mm color rolls, which I'd say accounts for 90% of our business. Though at our scale I wouldn't call the amount of Black and White or Slide we do as insignificant.
Generally I'd say the Noritsu Scanners yield the best results for time spent for 35mm/120. I'd say results wise they are 90% there compared to Flextights with the added benefit of having some ICE capabilities (the infrared readout for dust masking). The Noritsu 1800 is probably the best scanner on the market right now though it's price point certainly reflects it. We have about 4 in production.
That being said our favourite machine is the Fujifilm SP500 which is a dedicated 35mm Scanner, with superior ICE capabilities to that of the Noritsu though not as sharp but I'd say has about 80% as good as the noritsu though the Fuji do have a nicer colour science. We have about 16 of these guys.
The Flextight honestly doesn't get much use because it is far too slow of a workflow. We've had about 3/4 different models throughout the years all SCSI. I haven't used any of the new ones. Though honestly I don't think they are worth the time. The need to cut the negatives to use them alone is prohibitive and makes it more of a tool for special circumstances in our opinion.
I have yet to use many drum scanners or the fuji highend flat bed scanner but if someone has any to play around with one I'd like to give em a go.
Thanks for the details! The Fujifilm SP500 doesn't seem as utterly unattainable as an $18,000 Noritsu, but it's still a bit of a stretch for anything but commercial work. I have a few of the latest and greatest Nikons which still yield pretty great results.
Article author here - can you let me know where you're based (details in my profile)? I'll be looking for a solution to keep scanning my Xpan negs, and possibly my F1 project at some point in the near future.
I wonder whether making multiple scans on a decent and more accessible scanner with a small physical shake after each one and stacking the images can get you acceptable higher resolution results. After all, all smartphones do it today for each photo, and the astronomers have been doing it for decades.
> If that happens you’re looking at a 3,000 Euro repair bill if you can find someone with the parts capable of replacing them. Hasselblad will still repair them but it’s a pain to get the scanner to them and they charge almost twice as much as third part service shops. .... I went through several weeks of debugging to eventually conclude the port or main board was bad and was this not prepared to drop 5,000 Euros on it
So I don't know the author's own personal financial situation, but for me given the options 'E5000 plus weeks of effort and uncertain results' or 'E8000 and some legwork to find someone to do the work' or 'E11000 and some annoyances to have the manufacturer do it' I don't think I would have made the same choice.
Do modern equivalents of these scanners exist, but they're too expensive for the author? Or has the market for such things declined to the point where you just can't buy them anymore? (If so, why? Everything's just end to end digital now, so nothing to scan, I guess?)
Feels like the kind of thing you’d be better off renting than buying. I wonder if there are any facilities you can buy a day pass or something and use expensive gear like this. I know there are for woodworking and related hobby gear.
Ten years ago I paid for some proprietary software to help me collate my photographic archive. The project was very successful but the software is also 32bit Intel only. Since Catalina onwards I have not been able to access the archive at all. It causes me a lot of anxiety.
Thanks Adobe / Apple! Hopefully Lightroom 6 runs on Windows, but I just never seem to have gotten around to trying. I think it’s mostly out of fear that it won’t work and I will realize I am shut out of my collection metadata forever.
It does run on Windows, and on the Mac you have the option av installing the current versions from Adobe. Even without an active subscription, you'll still have access to the library module (i.e. all your metadata) and can do various exports etc.
"Absolutely no contest, and there’s no way I can interpolate the consumer scan up to achieve files good enough to print 160x60cm. I need high quality scans of these negatives."
I have a strong sense that the original author has not tried any of these upscaling techniques:
Even for consumer scanners, it seems I’m helping my parents fight unmaintained proprietary firmware on Windows every few years. Eventually I just set them up with SANE.
Sadly it seems SANE has no Hasselblad drivers at all, which probably makes sense given that these are professional devices with a very small install base. Even worse, given the high end features I suspect the firmware isn't just a clone of some other device like it is so often with scanners.
Probably not too hard to reverse engineer with some skills and tools, but well beyond what a random photographer is likely to be able to do. The real shame is the company squatting on the driver source code and refusing to update it. It is supremely frustrating when companies treat the driver software--useless without the hardware for it to run--as top company secrets. The software does nothing without the hardware your company sells, so who cares if someone could "steal" it?
I recently got a (35mm/APS) Canon scanner working really well with http://www.sane-project.org/ - not the same as the one mentioned in the article, but a really good collection of open source backends for old film scanners.
>Nobody demands you to give out the stuff that you don't use for free, why should we demand it from a company?
The two aren't comparable. IP is not the equivalent of physical objects.
A better way to reframe it is that the company is effectively keeping ideas hostage.
They aren’t doing anything with them, but are preventing anyone else from making something as good to do the same job, for no good reason.
>Nobody demands you to give out the stuff that you don't use for free
People are demanded to give up land for the public good. It's "eminent domain" in the US, but called different things elsewhere. Compulsory Purchase Order in the UK.
They paid, say, $10 million to make it, and now it's theirs. If you want it so badly – buy it. If public is not willing to pay a small fraction of this, like $100, then public doesn't need it.
I don't know but I suspect the consumer grade scanner can actually be useful. Scan several times, stack the images, and then deconvolute them. I think you could probably get great quality. Now how long it would take I don't know, probably way too much time though.
What’s the best option for standard 35mm film? I have a bunch of negatives sat in a box that I’d love to get digital copies of before they end up getting damaged. Should I buy a scanner? Or am I best off using a a scanning service?
IMO as a person who did a lot of analog photography going to a decent lab and having them scan it will be the best option unless you plan to develope and scan a film every month.
Buying a scanner of a quality that matches the lab is a not going to be cheap and unless you have a steady output of negatives (or a huge collection of material and the the time to scan it) it is not going to pay off.
I'd say go to the lab unless you wanna have the scanner for other things (e.g. artwork).
I have a Nikon Coolscan that I keep running. On Windows there's a "compatibility mode" that I think means "pretend you have a 32-bit address space" that I need to use to run the drivers.
Could you use a 35mm condenser enlarger to project the image directly onto a scanner bed? You would probably need to disable the light in the scanner, remove the glass, and upgrade the light in the enlarger.
I still don't understand why 32bit can't be executed in a special process mode or in Rosetta. Just seems like a ridiculous hill to die on. Few apps use more than 4gb of memory.
Doesn’t really feel like Apple died on this hill. It’s been years now and other than old obscure hardware, there hasn’t been any real loss. Even looking at my steam library, it seems like basically everything made it over in the end.
this is one of the reasons I would highly recommend digitizing any analog film media you might have lying around that has any value whatsoever. Technology like this, that greatly assists in “crossing over“ the old media to the forever-digital media, is disappearing in accessibility due to lack of utility – once everyone goes digital-native (which has already occurred), there is less and less of a reason to have high quality analog to digital converters.
I have a stash of old black & white photos from my grandparents' youth. They are nice but I would love to digitize them and was wondering if there is a better way than a flatbed scanner built into one of those cheapo Canon MGX inkjet all-in-one printers. Ultimately I want to run the scans through some colorizing filters and reprint them into an album to make a fun xmas present.
No, your top priorty should be washing the film thoroughly to get all the fix out after developing. Then store it properly in a dark and not to warm and humid area. That way, it will last for centuries (at least for black and white) and can be rescanned every time the digital files are lost
Ah yes. The "hope a flood or fire or theft or loss or baby/dog got to it and tried to eat it or simple human forgetfulness never happens" plan. That deserves not only a "no thanks," but a "that is a fucking ridiculous and quaint but also very sad (in the sense that this person has not realized that all analog media is already dead) notion"
Also, you missed the entire point of my comment. "Rescanning" will one day NOT be possible, at least not cheaply or without great effort. The market for people scanning analog to digital will dwindle to nothing, and thus the products that try to satisfy that market will also, unless you will be willing to fork over a ton of money for something custom.
The best you'll get is a bad reproduction taken with a cell camera of the original media that is then perhaps "enhanced" (read: pixels are invented out of thin air that were never there to begin with) by some AI.
What do you mean by “get”? I have an Epson V700. I have done plenty of side by side comparisons. The sharpness of the grain in the scan is what matters, not merely “getting” the grain. To speak of the resolution of a film exposure is incoherent. Grain != pixels. The V700 is also inconsistent because of the floppy plastic mounts and because they lack a mechanism to really keep the film flat.
If you can see individual grains you are already limited by the film itself, anything beyond that is noise. Sharper grain is just a higher spatial frequency noise.
Because Epson scanners don't actually resolve the grain. Low resolution scans tend to exaggerate grain... which can make you think you're resolving the grain.
Also, photos aren't only about resolution and sharpness. Colour and tone reproduction is also important. And consumer scanners are acceptable, but nowhere near as good as drum scanners. Even minilab scanners (Fuji Frontier, Noritsu) are no match for drum scanners.
It can't. It really can't. Even with betterscanning.com holders you can't beat dedicated film scanners or dlsr scans. Drum-scan obviously being the best. I've never wet mounted on flatbeds so can't comment.
My almost 2decades old Nikon V-ED, which I bought when it came out, still outperforms the most expensive flatbed you can buy now (V850-pro), which I also have.
Again, I will write. The job of an OS is to provide an abstracted shim between HW customers have, and the software they have. When this doesn't happen it is a giant fail. There are untold thousands of devices just like this scanner for which the choices to drop support (or change) the OS driver ABI puts people in a bad position because frequently this low volume specialty hardware costs more than the OS vendors product and a new PC/etc to run it because the HW vendor uses MS/Apple/etc's decisions as a chance to force their customers to upgrade to the latest HW. Frequently when the old stuff continues to work just fine.
So IMHO this is a giant middle finger from the OS vendors to their customers, because (as an OS developer myself) dropping support for these kinds of things are rarely done because its costing developer time to have modules that are mostly untouched for years in the tree, or infrequently provide a small shim from the new driver model to the old one. And even when it turns out to be real effort, the OS vendors show us they can provide very transparent shims as long as it benefits them and not a second longer (ex most recently doing transparent x86 emulation on arm based machines).
On the other hand, there is little that necessitates this sort of thing to be part of the OS instead of just being separate piece of software like VueScan
This isn't the MS-DOS/PC-DOS era where it was necessary for software to have full and direct access to communicate directly with the (ISA 16-bit bus attached, whatever) hardware and peripherals on a desktop workstation, because that can crash the whole multitasking OS.
In the case of the scanner in the original post here which appears to be from the era of MacOS 10.4, 10.5 and 10.6 , the operating system clearly doesn't let the scanning software talk DIRECTLY to it, there's an abtraction layer of some sort for the SCSI or (Firewire 400?) bus.
This paragraph from the blog post stuck out to me:
> The software that drives these scanners is compiled for 32bit architecture and hasn’t been updated in well over 10 years. Hasselblad refuse to update it to modern architectures, and refuse to open source it to allow others to do that, so you are stuck using an OS that can run it.
It's fine if the scanner company doesn't think it's worth putting in the work to add 64-bit support to their software themselves, but they're the ones artificially making it harder to support indefinitely by keeping the code private despite seemingly having no plans to ever sell anything that uses it again. It seems weird to me to blame OS vendors for operating in bad faith when the people who own the actual code that no longer works aren't even willing to let anyone else attempt to do the work to update it for them.
In business terms, open-sourcing a piece of software they haven't touched for ten years means spending a number of hours on doing that for no obvious gain¹. Legal has to clear this for approval (risks? liability?), someone has to decide if opening up the code won't give away company IP, and someone has to actually find the source code (not a given at all), put a licence on it (going through legal again), upload it, and provide a minimum level of documentation.
Doing nothing on the other hand, costs nothing.
1: There is the positive marketing you gain from such a move, but someone has to quantify if doing this is worth it.
That's fair, but I still don't see why that somehow means that the OS vendor should therefore be expected to keep that code working in perpetuity. I have no love for Apple or MacOS, but I just don't see how they're ultimately the ones who should be blamed for this situation.
Apparently the Apple Silicon macs don't have the capability of running 32bit binaries. Since this software is abandoned, it's not like a longer support window would have given them time to recompile the drivers.
Intel Macs do, though; until Apple removed it. Worked fine in Mojave, but me (and many others) one day rebooted our Macs to find dozens of programs, libraries and plugins that no longer worked.
Whole thing sorta smells like Apple trying to further assert the dominance of AArch64, what with their insane stake in the ISA and it's eventual dominance. Not a terrible thing inherently, but the push to depreciate 32-bit hardware and software alike was pointless. 32-bit ARM can (and should) live alongside AArch64 as an ISA. From what I understand, the M1 even has secure 32-bit emulation mode built-in, without any OS functionality to leverage it.
Whatever Apple does, it's hard for developers to treat the Mac like a regular development target now. Besides the ISA change, the push for Metal-agnostic GPU pipelines has left developers scrambling. Massive open projects like Blender didn't even have an Apple Silicon strategy until Apple sponsored one. It's a bit of a mess, and kinda tragic that Apple is pointlessly complicating cross-platform development.
For all the hate Microsoft regularly gets, they are pretty amazing about that: I have ancient software that still runs fine on Windows 11 through layers upon layers of compatibility libraries baked into the OS.
I'm sure it's costing them dearly, but there's plenty of niche and industrial hardware like this that lasts much longer than a generation of operating systems, and there's no alternative. Linux pays lip service to compatibility, but ABIs change constantly and it's easy to get stuck in dependency hell. MacOS X is... MacOS X.
Speaking of smelly: My scanner just broke down last week. The powersupply started smelling burned plastic and then gave up the ghost completely. Can't find a powersuppy fitting the special contact. It has been working since windows 95 or so, was now running windows 10 without problems and without having to install special drivers from the producer. Old was a canon, early USB model.
Given the vast difference in power, it’d probably have been enough to just cascade all the various layers they already have. Arm > x86 > ppc > 68k
Even if each layer loses, say, 50% performance, the fastest 68k was only 75mhz.
(Although now that I think about it I think the handled the 68k/ppc transistion by just making everyone compile for both and slap em into the same binary. Memory is hazy.. it’s been at least 20 years since I’ve touched a classic mac)
You're confusing the "fat" binary format that allowed developers to ship the same application file for both 68k and PowerPC with the emulator that allowed for compatibility with existing 68k code. Much of the OS and apps remained 68k to the end of classic Mac OS's life.
Yes, Apple only cares about allowing supported software to easily keep running. This means they can implement more modern and secure systems than Microsoft is able to achieve. It's also the reason Windows is plagued with 32 bit software currently in development as MS has no way to push for the removal of obsolete features.
I worked for large post facilities that would most definitely have told you to pound sand. I've also worked for smaller facilities that used the same equipment. We frequently would take on projects like this, especially if we were slow. It made those boring times in between projects much more interesting. Sometimes there was also something we learned in the process that made us even more experience for future projects. I've pushed to accept some of these jobs personally.
At the end of the day, what's the worst that could happen from the conversation? They say no? Volunteer to come in and work with it yourself during off times or other types of ideas. Show them your passion for it, and get them excited for your project. You'll be amazed at what the community will do for others.