End of summer update
Happy end of summer, everybody! For my family here in Maine, the month of August always feels a little frantic as we try to cram in summer activities before school starts again. In September, we all slow down and settle back into a consistent routine. I'm looking forward to that!
On the surface, August was a slow month for FilmLab progress, as I didn't get to spend as much time writing code as I had expected. In addition to family activities, I spent a lot of time this month communicating with people in the FilmLab backer/user community. Some of this was technical support necessary to get everybody up and running with the beta programs for iOS and Android. But many of the conversations (and certainly the most valuable ones), were about your experiences using the first beta builds, and your feedback and ideas for making FilmLab better.
And let me say: your feedback and ideas have been awesome. To be honest, there was a part of me that thought I wasn't going to learn that much from user reports on the early betas: after all, I already had my own personal long list of bugs to fix and improvements I was planning to make. But in practice, getting your early feedback has been super useful and eye-opening. There have been a lot of a-ha moments for me as I saw things from a new perspective, or realized that some of my assumptions were wrong. So thanks, everybody: I'm grateful for all the feedback, and it's definitely going to make FilmLab 1.0 a much better product than if I'd developed it in private.
Priorities: Useful is better than smart
I'm currently making progress toward the next beta, which I hope to have ready to release next week. In doing so, I've had to think about priorities, since there's a lot to do and I can't do everything at once.
Here's what I've realized: You can think of FilmLab as being made up of a set of core tools, like perspective correction and negative-to-positive conversion, combined with a layer of artificial intelligence that automatically applies those tools when it recognizes film in a scene. My plan has always been to add full manual controls, so you can control the tools yourself, but I'd been planning to leave that part until the end, after the intelligent features were 100% working and reliable.
But seeing how people are using the current betas, I've changed my mind. Given the choice between useful and smart, useful is better. It can be annoying to have to manually do work that the computer should be able to do for you. But it's super frustrating when a supposedly smart feature fails, and gives you no manual fallback option: now you just have a useless product.
So for the next few betas, I'm going to prioritize manual features. Even if the smart features fail, there should be a way to get the job done. I hope that approaching the task list this way will make FilmLab more useful in the short term for people using the beta builds, while still getting all the smart features working reliably before version 1.0.
Real world sample images
The current beta builds of FilmLab don't have raw processing enabled yet, so the output has poor dynamic range and often has bad-looking noise reduction and smoothing introduced by the default JPEG processing on phones. Despite these flaws, some users have been getting some surprisingly good looking results, especially with medium format film. Here's a nice black-and-white portrait posted by @analogmoment on Instagram:
And here's a fun color image shared by @velviatic on the FilmLab Community site:
The best place to share feedback
Let me take a moment to plug the FilmLab community forum. There's a lot of useful discussion happening over there, and I think people have been finding it helpful to get a sense of how the FilmLab beta builds are working/not-working for other people. There are some good discussions around light tables and other complementary hardware as well. If you haven't already created an account over there, I encourage you to do so!
Thanks again, Kickstarter backers! FilmLab is a big project, and I definitely couldn't be spending this much time working on it without your support. I think September is going to be a really good month for FilmLab development, and I look forward to getting you new and improved betas soon.