New lab scanner option: Auralab

[Jon Buffington]

Having used both a Kodak Pakon scanner and also a Noritsu lab scanner, I much prefered the color output of the Noritsu. Scans off the Pakon using Kodak software was the worst color I've ever seen off a professional lab scanner. It really made film look like crap. The rez may have been ok, but if the color isn't good, what's the point? Also the Pakon needed longer strips of cut film to put into it without jamming. The Noritsu allowed shorter strips. So I wouldnt have high hopes this new Kodak scanner would be much better. I also dont know why 120 keeps getting forgotten with Kodak. Last time I saw a Kodak Pro scanner be able to use 120 is with the HR500, and that scanner is well over 20 years old now. At least the Noritsu scanners (some of them) had the option for 120 trays.

As the previous poster stated, this has nothing to do with Kodak. I just picked the info up of the pakon fb group (which is a user group, not affiliated with any manufacturer/company).

As for color output of the Kodak Pakon f135+ (using PSI), I find Kodak color algorithms far superior and accurate to competition. I have owned one since 2012 so have 11 years and thousands of rolls scanned. Of course preference is subjective so horses for courses. We have options and anyone is able to use the device of their choice. This still in development scanner offers another option to the dwindling options we have and why I originally posted it.

 

[braxus]

I think it was the software our Kodak print stations were using with the Pakon. It did not look good to me really. But yes color is subjective.

 

[foc]

I have used both Frontier and Pakon scanners professionally and while I preferred the Frontier colour, the Pakon produced its own unique colour.
As has been said colour preference is subjective.

 

[Adrian Bacon]

Ran across this today on the kodak pakon group on FB. Looks like a commercial lab option, brand new product on the market. Pricing is still TBD but from comments from the advertiser/developer, it will be spendy. Anyways, I know there are a few on here that have labs so this may be an option vs using older/ancient lab scanners.

https://auralab.photo/en/?fbclid=IwAR0p-nTcWKd5b0PH0x2n7i4HPHebsSIjWgKj533R35MyrKimdI4IY38XW5Q

Heh.

I’ll believe it when I see it.

from a hardware perspective, sure, no problem.

as someone who is intimately familiar with developing scanning software for a lab environment, I’m seeing a lot of promises that are years away unless they have stacks of cash and a hoard of software guys that are also familiar with high volume scanning workflows (I,e, guys like me).

the software side of that link has a lot of buzzword that sound great, but actually realizing it in any reasonable and cost effective time frame is a tall order.

 

[MattKing]

I was hoping to see Adrian's thoughts about this.

 

[Adrian Bacon]

I was hoping to see Adrian's thoughts about this.

If they can actually pull it off, wonderful, however, I suspect that they're going to discover that it's actually a lot harder to do than they think it is. I've labored away on my scanning software for years at this point, eating my own dog food, and I still don't have something that I would consider releasable to the public. It works, for me, but needs a lot more work to handle the general public throwing files from pretty much every camera available (and scanners too, I suspect), and a much more polished user interface before I'd ever consider selling my code.

To be fair, they do have the advantage of only having to support their own hardware, but knowing what I went through to get to where I am now with my own code... Me thinks they're going to hit a number of harsh realities.

 

[Jon Buffington]

@Adrian Bacon
Long delay in my response to this. I appreciate your insight and thoughts on this.

 

[Adrian Bacon]

@Adrian Bacon
Long delay in my response to this. I appreciate your insight and thoughts on this.

No problem. I hope they can pull it off, because there just aren’t enough solutions out there, but if you stop and think about it, film has been on the incline for a good while now. If a good scanning solution was really that easy, you’d think somebody would have done it already, but has it really happened? I’ve been working towards that, but I’m not there, and won’t be for a while, partially because I also have a film processing lab to run, so the time spent making my code better suited for the masses is somewhat limited. But even if I could spend full time working on it, I’m still a ways off.

 

[Jon Buffington]

No problem. I hope they can pull it off, because there just aren’t enough solutions out there, but if you stop and think about it, film has been on the incline for a good while now. If a good scanning solution was really that easy, you’d think somebody would have done it already, but has it really happened? I’ve been working towards that, but I’m not there, and won’t be for a while, partially because I also have a film processing lab to run, so the time spent making my code better suited for the masses is somewhat limited. But even if I could spend full time working on it, I’m still a ways off.

Most definitely! The legacy stuff is just getting so long in the tooth. I use a Kodak Pakon f135+ scanner but software only plays nice on a dedicated xp machine, even vitual xp craps out. Thankfully, some reverse engineering has happened recently which makes it work on Win 10 and 11 64bit machines, albeit somewhat crippled to TLX client demo....the PSI software (only good on the plus models of the 135) makes scanning a joy and super simple. TLX is clunky but works. Picked up a epson 4990 recently for MF and large format, doesn't work on my win10 x64 machine. Off to vuescan I went since my last purchase from 10-13 years ago is gone....maybe I should just dslr scan but don't want to deal with another headache. Really, I should have bought an hr500 when they were cheap a decade ago. Ditto on the fuji sp3000...had one offered to me for 2k a decade ago. I should have bought it.

Anyways, good luck on your software development. Lack of a modern dedicated film scanner is the crux for many in the film world today.

 

[OrientPoint]

Thankfully, some reverse engineering has happened recently which makes it work on Win 10 and 11 64bit machines, albeit somewhat crippled to TLX client demo....the PSI software (only good on the plus models of the 135) makes scanning a joy and super simple. TLX is clunky but works.

Actually PSI does work on Windows 10/11 with the reverse engineered SKM Pakon drivers. The problem is that the Pakon installer (which is obviously still Windows XP vintage) gets confused and sticks the program files in "C:\Program Files x86" but configures the ODBC connection for the database PSI uses to track jobs to expect files in "C:\Program Files". To resolve this issue you can either adjust the ODBC connection parameters, or just copy the Pakon directory in "C:\Program Files x86" to "C:\Program Files"

Then change the properties on the PSI application to run as administrator and with Windows XP SP2 compatibility mode, and you should be ready to scan.

I've let the driver author know about it and he'll probably issue instructions on how to do it at some point, but he's a pretty busy guy.

Also, PSI does work with all models of Pakon (135, 135+, 235, 235+, 335). The main difference with PSI on the 135 vs 135+ is maximum resolution. PSI with the 135 is limited to 1500x2250. With the 135+ (or 235/335) it's the full 2000x3000 resolution. Using TLX it is possible to get the full 2000x3000 resolution on a 135, but it's slower than on the other models.

 

[Jon Buffington]

Actually PSI does work on Windows 10/11 with the reverse engineered SKM Pakon drivers. The problem is that the Pakon installer (which is obviously still Windows XP vintage) gets confused and sticks the program files in "C:\Program Files x86" but configures the ODBC connection for the database PSI uses to track jobs to expect files in "C:\Program Files". To resolve this issue you can either adjust the ODBC connection parameters, or just copy the Pakon directory in "C:\Program Files x86" to "C:\Program Files"

Then change the properties on the PSI application to run as administrator and with Windows XP SP2 compatibility mode, and you should be ready to scan.

I've let the driver author know about it and he'll probably issue instructions on how to do it at some point, but he's a pretty busy guy.

Also, PSI does work with all models of Pakon (135, 135+, 235, 235+, 335). The main difference with PSI on the 135 vs 135+ is maximum resolution. PSI with the 135 is limited to 1500x2250. With the 135+ (or 235/335) it's the full 2000x3000 resolution. Using TLX it is possible to get the full 2000x3000 resolution on a 135, but it's slower than on the other models.

would be nice to use PSI on win 10...I am not tech savy enough to know how to implement what you just typed I knew there was a reason why folks with the non plus used tlx, thanks for the reminder.

 

[OrientPoint]

Well hopefully we'll come up with a script to automate this soon. Or maybe I should write it up myself with pictures.

 

[henryyjjames]

Pricing is still TBD but from comments from the advertiser/developer, it will be spendy.

It looks like it will be priced around 20-25k USD. yikes.

 

[Jon Buffington]

It looks like it will be priced around 20-25k USD. yikes.

For a mini lab/boutique lab, this is not too outrageous. The older sp3k's and other long in the tooth minilab scanners are getting pricier. Heck, my ancient kodak pakon f135+ is selling for well over 1k these days which is to me crazy. Then again, they were north of 10k when new back 20 years ago. I am just glad more new options are coming out. We need solutions.

 

[gswdh]

20-25k for this is insanely expensive considering the manufacturing cost.

I would like to know if others would be interested in a product like this if it cost around 1k. I designed all the electronics and software for a scanner like this a few years ago. It's all working but I haven't assembled the mechanical parts yet. I could finish this off and get it on kickstarter with not a lot of effort. What do you lot think?

 

[OrientPoint]

20-25k for this is insanely expensive considering the manufacturing cost.

I would like to know if others would be interested in a product like this if it cost around 1k. I designed all the electronics and software for a scanner like this a few years ago. It's all working but I haven't assembled the mechanical parts yet. I could finish this off and get it on kickstarter with not a lot of effort. What do you lot think?

People are paying $1,000+ for 20 year-old Pakon scanners. There's demand at that price point. Whether it's enough demand to make it worth your while is something to take up in the business plan.

As a Pakon user, I'd be interested in a replacement so long as it was sub-$1,500.

 

[gswdh]

Thank you for your reply. This is the tricky part - gauging the demand. It's the reason I didn't take it any further because I started speaking to people to try and get an understanding of interest but didn't get anything reliable.

The engineering side is easy for me, the marketing and getting the word out there is the expensive part. You can't just put it on kickstarter and expect it to sell... you need a few hundred thousand in ad campaign budget.

 

[NortheastPhotographic]

Yeah, and Noritsus are... well, they're crap. Color's bad, only output relatively low res JPEG, they don't handle a lot of situations well, like adding banding to blue skies. They take a lot of babysitting and even then you can't fix crap like the banding issue. If I could get a decent 24MP RAW or TIFF without all kinds of crap being wrong I'd happily give the many hundreds of dollars I spent on my home scanning rig to the lab to do it for me.

The Noritsus haven't been updated in a generation, there is obviously no active development or improvement going on, software or hardware. But especially software.

At the very least, if this new upstart hits the market it'll spur some competition. Though it's a big "if" there.

I love confidently incorrect posts. I'm sitting next to my HS1800 right now and know a lot about its output, and how to get the best out of it. I also have 2 SP3000s and a Creo Eversmart Supreme II. The HS1800 is capable of excellent results with a competent operator. It's very color accurate to the film, some people just prefer the Fuji Image Intelligence, which is less accurate but often pleasing. Unless your scanner has a faulty part, there is no banding if you run calibration.

I've been speaking to the Aura people for a few months and they have a working prototype. Still early days, lots to do and work out, but it's promising.

I am also in talks with another shop who are looking into making a new Lab-Scanner. Promising times! (No I won't talk about it).

 

[QuisAmet]

Thank you for your reply. This is the tricky part - gauging the demand. It's the reason I didn't take it any further because I started speaking to people to try and get an understanding of interest but didn't get anything reliable.

The engineering side is easy for me, the marketing and getting the word out there is the expensive part. You can't just put it on kickstarter and expect it to sell... you need a few hundred thousand in ad campaign budget.

The home scanning community is fairly close knit. I'd say if you could bring a relatively affordable ($1k) scanner to market that competed with legacy scanners for quality, whilst having relatively high throughput and being up to date, the marketing would do itself.

 

[OrientPoint]

It's not just about price and hardware. The software is (arguably) the hardest part. With the Pakon, for example, the hardware is quite good. Pretty robust, compact (at least the 135 model), very fast, and excellent scan quality. The software is pretty awful though, and most significant obstacle to keeping those units usable.

If your scanner is a Pakon-style feed-a-roll type (which I hope it is) then it's probably not going to be something you can simply interface with Vuescan. Scanner-specific software is a huge investment to create and support.

 

[gswdh]

The home scanning community is fairly close knit. I'd say if you could bring a relatively affordable ($1k) scanner to market that competed with legacy scanners for quality, whilst having relatively high throughput and being up to date, the marketing would do itself.

That's good to hear! I think I'll finish the prototype in the next few months and post something on the forums.

 

[gswdh]

It's not just about price and hardware. The software is (arguably) the hardest part. With the Pakon, for example, the hardware is quite good. Pretty robust, compact (at least the 135 model), very fast, and excellent scan quality. The software is pretty awful though, and most significant obstacle to keeping those units usable.

If your scanner is a Pakon-style feed-a-roll type (which I hope it is) then it's probably not going to be something you can simply interface with Vuescan. Scanner-specific software is a huge investment to create and support.

Yes, totally agree, this was half the reason I started making a scanner. I have finished the basic software for the scanner with a good, fast and stable graphics interface. It has basic features to control the scanner, set the number of frames to scan and load film presets to optimise for signal to noise. Indeed it is a roll feed style. Currently, the software is written in Python but can easily be ported to C++ as I wrote everything around the Qt framework, multithreading and all. I would make all the software and technical details of the scanner USB interface fully open source so the scanner HW would likely have life after many many years and not be held back like what's happened to older scanners now. If fact, all the work so far is open here:

Electronic Design
FPGA Project
PC Software GUI

For me, I think the hardest part of the design would be the mechanical. Getting a reliable film transport that will standup to curly film etc. and not have any chance of destroying people's film will be tricky. Also, inevitably injection moulding will be needed which is pricey to setup but is amazingly low cost and high quality once done.

 

[PhilBurton]

That's good to hear! I think I'll finish the prototype in the next few months and post something on the forums.

Seriously???

 

[OrientPoint]

Seriously???

How hard could it be ? ;-)

 

[PhilBurton]

How hard could it be ? ;-)

Doing a circuit board isn't child's play. Neither is a mechanical design with motors and a power supply. And manufacturing to tight tolerances is no easy feat. Probably needs a team of people.