Responses to “I Almost Bought a Scanner” | Hacker News Reader

Responses to “I Almost Bought a Scanner”

leejo.github.io

92 pointsleejo49 comments

1. old software - For 5-15K hardware keeping dedicated $100 2007 iMac is a non issue. No need to upgrade when it does what its supposed to do.

2. "3,000 Euro repair bill" “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.” is BS from repair company. Where do they get the replacement boards from? answer is they dont, they fix it by replacing $1 component and charging for knowledge and rarity. You are paying for secret knowledge of where to hit the machine with hammer. Hasselblad neither manufactures chips nor ordered custom Firewire controller just for this very niche product. Someone like STS Telecom (https://www.ststele.com https://www.youtube.com/@ststele/videos) or iPad Rehab (https://www.ipadrehab.com/ https://www.youtube.com/@JessaJones/videos) will be able to do the repair at reasonable (compared to $3K) cost without overcharging for working on rare professional photography hardware. As a bonus they will happily publish reverse engineering results so others can also repair their scanners in the future. The issue might be as trivial and simple as bad capacitors in need of a recap, this is 15 year old hardware after all.

So, I missed the suggestion thread, but here's mine: hack together a DIY microscope scanning rig. It'll set you back less than 500 dollars. An open frame 3D printer like a Creality CR-10S can be had for a few hundred - that's your stage, complete with controller electronics. Replace the hotend with a bracket for a Pi Cam HQ - it's a decent 12MP sensor - and add a microscope lens [0]. Put a lightbox on the bed, then your film, then the thinnest piece of glass you can find to hold it down. Now you can easily raster-scan at over 17000dpi. You can even make minute depth adjustments using the vertical axis of the stage to correct for film flatness issues.

I'm sure there are many objections to this setup - I know resolution isn't everything and I'm unsure if a Sony IMX477 sensor is "good enough" for the author in terms of color and dynamic range - but you can't argue with the price!

[0]https://shop.pimoroni.com/products/microscope-lens-0-12-1-8x...

I assume the author knows their stuff about why consumer scanners aren’t good enough, particularly around dynamic range. However, for black and white film, I’m generally able to make out the crystal grain on 35mm with my consumer scanner and included tray, so I don’t know what more there is to resolve.

I used to own the scan he is talking about and various epson scanner, you cannot imagine the difference, Epson is fine and you are happy with it, then you see what an Hasselblad scanner can do for the same negative and for the rest of yout life you will think that Epsons are garbage.

I "used" to own it because it had some small hardware problems and i was unable to fix it or find someone who could repair it. I bought a digital back and never did film photography since.

Bouncing on that last bit: do you find digital to match/exceed your experience of film+hasselblad?

I have a Nikon Coolscan 4000 from days of old. I was able to scan a 35mm at about 20 megapixels of data... and it fixed scratches and identified dust (my father has used it for the past several years to scan all of the old family photos - including those from his parents) and it works very well.

The big difference was in the amount of data in the blacks of a good slide film that you could pull out with a scanner. You can get a really wide dynamic range on the film by scanning it and pushing more or less light through it depending on the film.

There's also something... tangible or natural that film grain captures that I just don't feel with for digital.

For example this https://www.deviantart.com/shagie/art/Provia-400-grain-no2-1... is a detail of https://www.deviantart.com/shagie/art/Black-Sea-Nettle-Provi...

https://www.deviantart.com/shagie/art/crop-of-golden-gate-ca... is the cable (that you can see) from https://www.deviantart.com/shagie/art/Golden-Gate-Tower-50mm...

For its time, 35mm + a good scanner was not something you could get near with digital cameras. The good DSLRs are there now for 35mm... and then I can argue with myself about the difference between the curves on film vs doing it in post in photoshop and the ease taking multiple photos vs getting one shot (or a bracket) and hoping and the cost of a good digital camera vs the cost of processing film... its arguments that you can see fought out in my camera bags.

And this is where I've scanned with an Epson scanner. I've got an older epson photo scanner. Its lid has a light in it (so it can transmit - and scan film). And it has a tray so that you can put 120 or 4x5 film in, secure it (without having it touch any of the glass and scratch) and scan a flat plane.

And it did fairly well. I used it to scan my xpan shots (since the Nikon scanner won't do that). https://www.deviantart.com/shagie/art/Sunset-Marin-Headlands... is one such photo.

The thing I liked about film, was I knew what I was getting out of the box, and the images would largely take on the look I wanted based on my choice of film.

Yes, the quality is better, you lose the film "feeling" but everything else is so much easier...

What is a "digital back"?

Edit: Wow! Thanks for all your replies! This sounds intriguing, but only tangentially related to scanning.

Now if you could turn a film camera into a high-quality scanner.. wouldn't that be something :)

I'm in the ignorant camp where I've been satisfied with Epson consumer grade scanners because I had no idea what I was missing.

Cheers.

I think in some film cameras, you can replace the part of the camera where film is exposed with a digital image sensor. Then you can use the same camera, but instead of exposing chemical-based film, you're recording digital images.

https://www.hasselblad.com/de-de/h-system/h6d-100c-digital-b...

You love the feel of your camera, its optics, and you don't want to throw away 20000 dollars? Pay a few thousand more and it's now a digital camera, everything else stays the same.

The large format approach is often a scanner - which works fairly well with a landscape. https://en.wikipedia.org/wiki/Digital_scan_back

It takes a bit of rethinking since the shutter speed should be thought of as exposure speed since it is the speed at which the scanner moves - not the duration the shutter is open.

Replaces a film magazine with a digital sensor

Has been made since the late 80s in various forms.

Phase One is the “premier” brand, with hasselblad closely following, then you have the defunct brands such as imacon, leaf, megavision (maybe around still), Sinar (around, part of Leica), Fuji made one, Rollei also made some. Dicomed made an actual 60mmx60mm one for a full frame 6x6 camera, still unique to this day.

Others have already explained, but many medium format camera systems were made to be modular long before the advent of digital. The viewfinder, lens, or the bit at the back which holds the film could all be hot swapped if you wanted to switch to eg a waist level finder or different film.

> I've been satisfied with Epson consumer grade scanners because I had no idea what I was missing.

Every time I meet someone who is expert in something I enjoy I have learned to tread carefully. It can often ruin perceptions - sure, it leads to a better result but it’s often very pricey.

Coffee, computers, networking, storage, backups, automation, tools, etc etc.

My life is a lot better and things a lot more reliable for the help I’ve had, but I’m poorer.

The silver particles in the film aren't very transparent and you can see this for yourself with a microscope. All tonal representations of B&W film, whether a scan or traditional enlargement, depend on limitations of resolving power to blur or alias the grain into a tone.

Correspondingly, you will see a difference in apparent grain from different scanners at different resolutions. If you think there is nothing to be gained by scanning on a drum or virtual drum, pick one of your best negatives and pay to have it professionally scanned!

I’ve definitely had plenty of medium format negatives drum scanned and I have found consumer scanners to not be great for color slide film. But, scanning a grainy/gritty film like HP5 or TriX doesn’t seem like quite the same kind of challenge.

https://epson.com/fastfoto-photo-document-scanner

If anyone needs to scan through bulk printed photos, I used this device to scan all my mom's stored photos for her wedding gift. I scanned over 12,000 photos over hte course of a few months so I couldn't bother with raising a lid. This thing makes short work of 20-30 photos per batch and it does a great job at scanning them. I host a small pikapod that has them tagged by a rudimentary AI for their faces and other objects. It was a fun little project and warmed my heart.

> I partially agree - but without the patent system where does the incentive for companies to invest millions (sometimes billions) into research and development come from? Altruism?

This presupposes that the patented invention is trivial to reproduce by mere knowledge of its existence - in which case, is that really the best use of a patent system?

Most worthwhile inventions can't be easily duplicated like that; the knowledge of how to manufacture it, continue iterating/improving on it, etc. is almost certainly concentrated in and around the inventor - giving the inventor a considerable first-mover's advantage. A better system, in that case, would be to ensure the inventor can capitalize on that advantage - in particular, that even individual inventors have access to capital such that they can start producing their inventions.

(A better system than that would be for everyone to have a safety net sufficiently strong such that they can collaborate on inventions solely motivated by intellectual curiosity rather than needing to worry about monetary gain, but I digress...)

> This presupposes that the patented invention is trivial to reproduce by mere knowledge of its existence - in which case, is that really the best use of a patent system?

This seems to be the central issue with many software patents, where the "invention" is either generic, obvious, or already widespread.

Take the whole set of patents covering touch screen interactions, for instance. Apple sued HTC for several patent infringements and lost, because the judge deemed things like scrolling through pictures "too obvious". The case covered four patents, as far as I remember. Apple also sued Samsung for the "slide to unlock" patent, and won, not because Samsung copied Apple, but because Apple held the patent, and it was considered valid.

Thus, there is no room for any sort of simultaneous inventions here. Samsung might have come up with the idea by themselves. Apple might have patented a concept that was used by others first. And the reason why a single company was awarded exclusive ownership over it, was timing and money, nothing else.

The upside is the patent system hasn’t been fucked up like the copyright system, so after 15 years all the really obvious patents start expiring.

Yeah, if the copyright system was brought back to the same state as the patent system, then that'd be a strict improvement - though that's less due to the current patent system being decent and more due to the copyright system being obnoxiously bad.

Often commercial software has tons of closed-source libraries that they don’t have the source code to. Even Doom had that when opensourced. It’s an annoying headache for a company to open source even for a project no longer in development.

They could still release what they do have. Depending on the nature of the missing bits, the community sometimes can replace or shim out chunks of proprietary code and still have an easier time than replacing the whole thing. Or, depending on the licenses involved, it can be possible to at least release object files so people can edit the core code then link against closed libraries to build the final library. That's what happened with OpenSolaris - Sun released most of the OS as open source, but couldn't relicense some chunks, so they released a bunch of blobs that had to be injected into the build to create the final product; AFAIK some of these persist as annoying blobs to this day, but some were replaced piecemeal by illumos with FOSS versions.

(Obviously, this is very dependent on how much lifting is done by what parts and how specialized they are; if you happen to use a closed zip library, that's easy for the community to work around, whereas if your app is a thin wrapper around another closed library then maybe opening it really is minimally useful)

Yeah, and it does happen (rarely) - the best way to get it to happen that I've seen is to have a major customer ask for it. So instead of normal users asking, have Ford ask for a guarantee that X years after maintenance ends, the portions of the code that can be GPL will be, or something.

Even then it's rare to work, but it sometimes can.

In this instance, though, the protocol is more valuable than the software.

If they released documentation on the Firewire protocol, likely somebody could write the software to talk to it fairly quickly. Hardware protocols are almost always the simplest thing that could possibly work--and generally simpler than that. As one myself, I assure you that hardware people are LAZY. :)

(In reality, the issue is that hardware is really difficult to change and expensive to develop. Consequently, you want your hardware to be simple and robust while pushing as much complexity as you can up to the software stack.)

Or, even more importantly, if they released some schematics and documentation, that Firewire port could be replaced by an Ethernet port.

The sticking point is likely the word "documentation". I suspect that "documentation" doesn't exist anymore--if it ever did.

> if you patent something and don’t act on it in X years then it enters the public domain. If you discontinue the thing that relies on the patent then it enters the public domain. And so on.

people have an extremely limited view of patents to suggest this.

so you're working a 9-5, spent all of your savings on getting a patent because you did perceive something everyone else neglected, and people on the internet are like "go form a startup and raise capital, or else"! that's not what they say, but they don't think about what they say. they imagine a person with a toolkit in a garage or writing software and that's it, and then imagine creating and monetizing that one thing is the best use of their time, otherwise the idea shouldn't have been contributed to society at all. funny, who living in an economic center and has a garage anyway?

until we can have a real discussion on patenting and patent holders, as well as the incentives in making the patent claims broader and more generic than the inventor really had in mind, this conversation will go nowhere.

Not exactly relevant for hi-def pro photography, but if you're struggling to scan paper photos, and you can't find a proper scanner, a hybrid solution that kind of works is to take a low-res scan (using any tech, even a phone) and enhance the result / augment the resolution using some kind of AI software suite or online SaaS.

I had always been impressed with the fidelity of the Kai’s Power Goo example facial images.

https://www.macworld.com/article/226834/an-ode-to-kais-power...

I believe a drum scanner was used.

In a parallel fashion, I haven't delved very far into the topic, but I have a feeling I wouldn't be able to get any of my trusty old SCSI devices working, both from a motherboard and a driver perspective.

Naive question incoming:

Does a good modern camera with a macro lens, shooting from a few inches away, have a higher effective DPI than this scanner?

If so, is the only problem trying to get a predictable, smooth grid of pictures from such a camera to stitch together?

And if so, couldn't you just stick the camera to... a CNC machine, say? Is that not a thing?

Weird colors, lens distortion, there is a lot of weird stuff when using a phone camera to 'scan' a paper document

All fixable with todays filters etc though I'm sure.

Sure? Maybe? Do you have a working setup?

When you get down to it - a scanner is "just" a really specific digital camera with a film holder. The problem is that quality is more about setting up the entire system consistently (film aligned on the scanning camera focal plane correctly, can shift from frame to frame, etc). As with all photographic setups, the chain is only as good as its weakest link.

It is probably possible for any single person to build, through sufficient time and effort, a "better" scanner than any model manufactured - but once you are spending that much time and effort on it you've "lost the plot" of the whole effort.

I guess what I would naively expect to exist here, is some kind of "kit" for turning a DSLR into a 2D flatbed raster scanner, in the vein of the Pi-based emulation portable "kits" — i.e. "here's a box with all the mechanical and structural stuff, flat-packed with foolproof assembly mechanisms that guarantee rigidity; just bring-your-own expensive electronic bit to slot into it, and watch it go." Where in this case the "expensive electronic bit" is the DSLR — probably with a Wiki-maintained list of known-compatible camera bodies + lenses.

If such a kit existed, the "sufficient time and effort" would amount to maybe five hours and $500. It'd basically be the same hardware + electronics as a cheap 3D printer — just without the need for one of the axes. (Mind you, that's presuming all the imaging is flowing directly from the camera to software on your computer, without any electronics in the device MITMing the signal to do computational photography stuff. If it needed to stitch the photos itself, it'd be a lot more expensive.)

Enough people seem to both have this exact problem, and have the ingenuity to be the one person who produces the kit to scratch their own itch but also the itch of everyone else in the community from then-on, that I would be somewhat surprised if at least an attempt at such a kit didn't already exist.

I mean, I agree that such a kit would be hugely popular and useful. You just don't understand how hard the problem is. There's a reason that people pay thousands of dollars for twenty year old hardware.

> I would be somewhat surprised if at least an attempt at such a kit didn't already exist.

Oh there are lots of attempts!

Do you expect that an existing mid-tier consumer 3D printer company could make such a kit (with sufficient quality to be useful) as a byproduct of their 3D-printer production line — the same way that iRobot makes "robotics platforms" that are just Roomba vacuums, minus the vacuum mechanism, plus an SDK?

If so, would it be a good idea to go bug one of the 3D printer companies — most likely, one of the ones that frequently pumps out new models to address the problems customers complained about in previous models — to try doing this?

I don't think so.

The problem is that you have a... thick 160μm plane (the 'target') that you need to orient exactly along another plane some distance away. You are going to need to mount things on the target and move them through with good speed and precision. Every link in the chain has the potential to throw off the alignment. Traditionally this is solved through a centralized manufacturing process where one group of people get really good at building (and maintaining) the devices such that they generally line things up reliably (i.e. a company making a scanner).

It's kind of like P=NP for manufacturing I guess? If you can really figure out how to instruct a random person to get an above-average scan out of their DSLR, you can instruct them to make almost anything.

Friend built a light box with some color calibration markers and DSLRed 20,000 slides in a few days and post processed the ones he liked. It's not professional grade but it worked fine.

I've been playing with this idea. Planning on making an attachment for my 3D printer, but haven't gotten around to it yet - for scanning medium format negatives

I use my DSLR on a tripod with a film holder and a light table. It seems to give much better results than an Epson.

I also have an old Mustek scanner with outdated drivers, with no hardware issues.

Luckily, my scanner is USB, and VMware Workstation allows to forward complete USB devices to a guest OS. When I need to scan something, I fire up a virtual machine with 32-bits Windows XP inside, with OG drivers, and scanning software.

Does Vuescan support it?

Never heard of that software. Indeed, it says “VueScan is compatible on Windows, macOS and Linux.”

Thanks for the tip!

Will give it a try next time I need to scan something.

VueScan is really really huge in the photog community for those who still have to handle (or restore) film negatives. Not only does it continue to support scanners that have long since been abandoned by their manufacturers, its just straight up better software.

I've used it for shits and giggles, and it even did quality work on a generic Epson flatbed better and faster (and lets face it, Epson flatbeds aren't good, getting anything good out of them is very difficult).

VueScan is amazing, and the author is a great guy who does most of his own support.

Vuescan is fantastic!

Aunt's computer had been upgraded to Windows 10 and HP printer/scanner had the scanner part stop working...

I was cursing Microsoft and HP for half an hour reading various failed workarounds. I was getting serious "shaving the yak" vibes until I found Vuescan.

Best $25 purchase for a software in the last 10 years.

Definitely true point about the Hasselblad. They make their money through legacy and brand these days. There are plenty of medium format competitors that outperform them at a far lower price.

If the goal is to scan xpan, why not just get an old CoolScan or something eqivalent, and stitch together the scans

Did you read the article? A cool scan is still far worse than an X1/X5 - these are considered to be the best scanners besides drum scans

And coolscans don’t work with xpan images according to another commentor below.

Reply on news.ycombinator.com