New lab scanner option: Auralab

All this fuss, we should have released our scanner almost a decade ago ahh well.

Steven Lee said:

n a separate thread I remember you posting how much extra work (years) it would take you to convert your software to accept commonly available inputs. Besides, nobody actually makes commonly available inputs dedicated to film scanning.

Click to expand...

Well, there's been some updates since then that I haven't really publicized, but I guess now is as good a time as any. My original code only worked with a couple Canon cameras, natively outputted Adobe DNG files, and was written specifically to run on my systems. Well, I've heard that many enthusiasts would prefer something that isn't tied to paying a monthly subscription model to Adobe just to use the thing, and I have had on my roadmap to support more hardware, and have heard that they'd like to run it on pretty much any OS they'd like.

Since then, I've generacised my code to be standard ANSI C17 for the core functionality, pushed all the camera specific stuff out to use the libraw library, so technically, I can support pretty much any camera Libraw supports as long as it's a bayer CFA. I may have to do specific tweaks for different CFA layouts, but the foundation for all that has been laid. Fuji X-trans stuff is weird and would require quite a bit more work to support well, so for the moment that's a TODO at some point. Also instead of natively outputting Adobe DNG files, it now just outputs generic 16 bit TIFF files with an embedded ICC profile. And, because the core functionality is generic ANSI C17 code, I can technically compile it to run on any platform that has an ANSI C17 compiler available with libraw and libtiff. It's been architected so that the core functionality is discrete from any UI I may want to put on top of it giving me the flexibility to go figure out how to support a UI on the various platforms I may want to officially support. Right now I just have a super basic interface for my lab operations driving it.

So what's left before I can actually sell this thing?

1. Color profiles for other camera makes and models. I have a generic Canon profile that I currently use. I imagine I'd need to do at least a generic Sony Profile, and generic Nikon profile.
2. Have a facility for specific custom profiles so if somebody wants to go to the trouble of making a profile for their specific camera, lens, light source, they can. This might be a service I provide for a fee.
3. A user interface that consumers won't hate.
4. I'm sure I'm forgetting something.

Steven Lee said:

My camera scanning rig was close to $7K and using it still sucks because every single component, except the Negative Supply holders, is not used as originally intended. And there's literally nothing I can spend more money on to improve my scanning experience. I can drop another $10K for a tiny marginal improvement but on the process/experience side I'll continue to suffer with a horrific contraption anyway.

Click to expand...

Ehh... my setup is orders of magnitude simpler, and in all honesty, it doesn't need to be that complex. I currently run a Canon RF100mm macro lens with a Canon R5 with a negative supply copy stand, and negative supply film holders. I splurged and got the motorized 35mm transport from them (life saver), and have the borderless 120 setup. I have slightly bastardized their stuff a little bit so I can quickly switch between film formats simply because their stock configuration assumes you're going be screwing stuff in and out (at least on the older versions, haven't looked at the newer versions), which is a huge time suck and doesn't work for a lab environment, and I don't have a vibration problem because I use a strobe for the light source. For 4x5 and 8x10, I just use an Epson V850Pro. It easily does 2400dpi which is absolutely massive files for sheet film.

Steven Lee said:

Creating a separate comment regarding another tangent you mentioned: business models. You raised good points there, the major one is the low LCV of one-time hardware purchase. But the elephant in the room is market size. Big enough market cures everything.

Let's say there 1M of people who're willing to spend $5K on a film scanner which offers meaningful improvement over their camera rig. Let's assume the scanner lasts 10 years on average and your market share is 25%. This means 25,000 units sold per year. If you take $3K from each sale, that leaves you about $1K per unit at ~65% gross margin, i.e. $25M/year for OPEX and profit. I may be off here and there for simpler math, but these numbers are not bad... assuming there's indeed one million people who'll want it.

Click to expand...

There maybe might be one million people worldwide, but not here in the U.S. I don't have any hard data to back any of this up, but my guess would be that here in the US, the number of people who are willing to pay $5000 for such a thing is down in the couple hundred thousand range, and that's being really generous. For software, I don't see the various software makers like NLP living it up large, so while they may be flush with cash, they don't appear to be that flush with cash, which says a lot about the potential market size. This is one of the reasons why I haven't just rushed to market as quickly as possible, the market doesn't look to be that big, and much to my own detriment, the last thing I want to do is damage other players by entering a market that isn't big enough to support having multiple competitors. Competition is good, but only when there's plenty of customers to go around.

@Adrian Bacon Looks like your setup is extremely similar to mine, except the focus of my splurging was a sturdier copy stand from Kaiser, since I found that vibrations was my bottleneck. The reason I pointed out at Jim's contraption as a marginal improvement is because he employs a software-controlled motorized focus rail with focus stacking to offset never-quite-flat focus field on macro lenses, and he can easily do 4x5, 120 and 135 without changing much.

I will be the first in line to buy your software, and I love that you're planning for the profile making capability. I have no idea how many people will buy it, or how many people will buy a $5K piece of scanning hardware. I was using 1M number hypothetically and it made the quick math simpler.

One observation I have is that there's a strong feeling of dissatisfaction among users of all existing color inversion software. For example, almost every Negmaster user I've chatted with, had tried NLP before and hated its heavy-handed approach and strange colors. I may be wrong, but feels like many people are in the waiting pattern, ready to switch to something better when/if it comes along.

Have you played with Filmomat Smart Convert?

Steven Lee said:

Have you played with Filmomat Smart Convert?

Click to expand...

Nope, and I'm not planning to. I also haven't purchased any other software licenses from other makers since the last thing I want to be accused of is stealing ideas. It'll be pretty hard to prove that if me or anybody close to me isn't even in their customer databases and I've never had a copy of their code on any of my systems. I've seen screen shots and the usual YouTube reviews, and mention of me and my code tends to generate link referrals from their respective forum sites that shows up in my website logs, so I know there's some level of awareness on their part that there's potentially some competition coming.

Steven Lee said:

One observation I have is that there's a strong feeling of dissatisfaction among users of all existing color inversion software. For example, almost every Negmaster user I've chatted with, had tried NLP before and hated its heavy-handed approach and strange colors. I may be wrong, but feels like many people are in the waiting pattern, ready to switch to something better when/if it comes along.

Click to expand...

My observation is that there's a bit of dissatisfaction due to what I perceive as the misguided attempts by the software to "be digital" and provide some amount of presets and/or digital controls. That's not the aim of any of my code. If you stop and think about it, color negative film was originally designed to be printed onto RA-4 paper. A paper that had a set contrast response and set color gamut. No matter what film you printed, it always printed against that set of unchanging constants and the result was the look you got. Your only real control was white balance, and maybe if you wanted to really get into it, some other more advanced darkroom techniques, but by and large, that was what you got and it was specific to that output medium. Well... A tiff file with an embedded ICC profile is also an output medium with a set gamut (defined and controlled by the ICC profile) and a set contrast response. Once you know what the average contrast response of correctly processed C-41 film is, and what the average spectral response is, it's actually not that many lines of code to adapt that to a given ICC colorspace. I actually have way more code just dealing with handling camera files and getting the converted files packaged up and ready for upload than I do actual conversion code. The specific version of code I use does have those knobs and dials exposed so that if I need to I can go and tweak things for specific frames or change something if I get a wonky roll of film in (looking right at you LOMO, and will be looking right at you Harman Phoenix), and I don't know that I'd actually expose those knobs and dials in an end user product just due to the fact that you really need to understand what it's going to do, so it's very much change at your own risk and super easy to get way out of whack. And as I've processed and scanned film over the years, the specific values have been refined to a point that it's rare that I actually need to change them, so it'd be more of an advanced use case, or maybe something somebody might use if they home develop and their C-41 process is nowhere near in-control. Either way, all the film is scanned against that constant, so the differences between them come right on through because I don't try to change it, or make it look like something else. The benchmark is a generic neutral rendering that's a mixture of most all the currently available emulsions, so it's generally relatively neutral, but something like Ektar is going to look different than something like Portra 400 in the same general way Ektar would look different than portra when printed on RA-4 paper and that's because the like RA-4 paper, the digital output has an unchanging set of responses that the film plays against. There's no shenanigans of trying to make it look like something else.

Adrian Bacon said:

There's no shenanigans of trying to make it look like something else.

Click to expand...

FWIW Adrian, it is that lack of "shenanigans" that I trend to yearn for and, I would suspect, that many potential customers would yearn for.

MattKing said:

FWIW Adrian, it is that lack of "shenanigans" that I trend to yearn for and, I would suspect, that many potential customers would yearn for.

Click to expand...

Yeah, I get it. I see the UIs of the other programs in screenshots and such, and I find myself asking "why is that even doing that? I'd never even want to do that, and yet that seems to be the default thing it does." Maybe it's their way of trying to value add? I don't know. I just find it to be very strange in general, because from where I sit, that's not a value add at all.