Thank you for the vote of confidence!
I don’t know that I’d ever sell a kit or components, I’m very much interested in community development for this niche/sector of the scanning market. I’d be curious to know what RetroScan users want (if anything) differently from their machines. If anyone has a spare $13k lying around I’d be interested in getting one just to learn how they approach certain challenges!
John,
If you want the brutal truth, I think they would love a 4K, 10-bit camera. I love how the machine works, but the 2K camera is a weak link, in my opinion. Now that the Mark II is out, it’s capable of transferring 35mm, and I wouldn’t waste my time transferring 35mm on this particular model. I want more than 8-bit images to work with, because a lot of my films have dramatic color fade.
The community would also like a film cleaning option built on to the unit. It saves time for the home movie transfer businesses who use it, because there’s less setup time. (The PTR roller setup would also be a step in the right direction.)
Stabilization is another factor. I’ve read your post about Resolve, but if the film jumps for a frame or two, then it can throw Resolve out of whack. (I have experienced the same issue with Apple’s Motion.) And, in my limited experience, it requires a metric ton of processing power for anything longer than a 10-second clip. Maybe a standalone stabilization app would be a better solution. Many film transfer machines do downstream stabilization, but they cost thousands and thousands of dollars more.
Those are just a few thoughts on what I would love to see in the RetroScan Mark II.
Keep up your good work! -----Todd
Good to know!
A trick I’ve found with software stabilization in the meantime is to go straight to Blackmagic’s Fusion software (they’ve started merging it with Resolve and it’s not great in that form, but if you can find the standalone version it’s AWESOME). Fusion has been rock solid stable for me and it’s faster at stabilizing footage. I use that to produce the frame sequence that I then bring into Resolve for color and framing work.
Here’s the last standalone verion of Fusion 9.0.2 from July 2018:
https://www.blackmagicdesign.com/support/download/54fc7e36d6fe466d95bc2e583c359582/Windows
Dear John
I am using laser detection as well. Challenge I have is that the detection is not stable enough. Hence, the film is not transported at the same distance. A sprocket hole has a length of 1.143 mm. Based on what I see on my images I estimate a variation range of 0,2mm. This is approx 20% and is too high. Also after software correction the movie is shaking too much. Any suggestion. I tried using a bit of black paper with a small pinhole but that did not do the trick.
Mike
I understand the issue, my machine still suffers from the same problem - the solution that got me close and allowed it work was using the pinhole filter on the laser output. It effectively increased the spatial resolution of the detector down to a small fraction of the sprocket hole height. I tried paper at first but found that printing a thin but solid cap for the laser and then carefully melting a pinhole into it with a small needle worked okay (this took a few tries to get right - some luck might have been involved but there were many attempts). It’s not perfect and the image still requires stabilization in post (albeit much less), but there are other physical modifications that could be made to my machine before immediately investing further in the laser detector block that would yield greater overall image stability. For starters, I’d really like to have my gate machined from a block of delrin or PTFE rather than 3D printed, and my roller situation could be improved.
In the short term I’ve made a test stand to modify my sprocket detector from a “through the film” design to a “reflected” design. The pinhole attenuated the laser enough that the “through” design worked reliably well, but I can’t shake the suspicion that the reflected design might work better or differently. The next iteration of the sprocket detector would include optics that put the sprocket hole edge at the focus of a “spatial filter” type setup to really get the effective resolution up. I’m hoping to evaluate that soon - the whole pandemic situation paradoxically put a dent into the time I was able to spend on my machine. I was hoping for more progress in the last year after the lighting changes but with everything else going on it’s nearly stalled.