IEEE Future Directions
Related Articles
I just read a review of L16, the digital camera that according to its producer would have been a game changer.
I have the L16 under my sight since it was announced, back in 2015. The producer, Light, claimed, and they still do, to have a game changer in the making. The camera is actually an optical platform, consisting of 16 lenses that depending on the type of shot can be simultaneously activated (up to 10 at a time) to create a set of images that the software will combine to create exceptional resolution, great bokeh or depth of field as required.
The basic idea is that you have many “cameras” taking the shot and then you make the best out of the harvested pixels. In a size format that is in between a DSRL (bulky) and a smartphone (thin) you get better performances. In the best situation you get a 70+Mpixel resolution (that is equivalent to a medium camera format costing 10,000$+) at a 2,000$ price.
The camera had a quite long “gestation” but it managed to ship in October 2017 to meet demand.
Does it deliver? Well, yes and no, according to the review I just read it is more No than Yes.
The reason for writing this post, however, is not to take side one way or the other but to look into the kind of products that are emerging and of whom L16 might be a forerunner.
In the review the reviewer points out the dissatisfaction in some of the shots taken with the L16, attributable to low light situations, not sufficient dynamic range, under use of the various lenses BUT he is also saying that it is likely that new software releases will improve the outcome significantly. Actually, a new software release in November showed some improvements.
The point is that what you get is an optical platform whose performances are tied to the software analysing the images taken by each lens.
As we have seen lately, the use of Artificial Intelligence in processing images (I played with Luminar this week end and I was impressed by what AI can do to a photo…) can be a game changer and applying it to the L16 it can transform it. The point is that the L16 should provide not a DNG (RAW format) resulting from its internal software but 10 DNGs resulting from the individual lenses and let an external software to make the most of it. Of course, it might be even better for the lay user to have AI inside the L16 doing the job. I guess with future release we are likely to see this happening (not sure if the current processing power in the L16 can sustain the AI processing).
At the moment we do not have a straightforward way to deal with the 10 DNGs that are generated by the 10 lenses shooting together, and this is because we do not have cameras creating an attractive market for developers to support it (we have stacking software but that is customised to focus stacking from the same optics or DHR stacking again from the same optics). Hence we are stuck with the software delivered by Light.
This is what made mo think about Industry 4.o. Why don’t extend the concept of an optical platform to the point that is becomes open to third parties, end user included, letting them to create features (software for image rendering) that can exploit the raw performances of the platform. I am not sure if the biz model of Light can stand in this new paradigm but I am pretty sure that this opening would stimulate many people to focus their efforts and skills on creating new ways of image rendering ad pretty soon we would have a continuous improvement that might lead to that game changing promised by Light.
At the core of Industry 4.0 is an open value chain able to attract a variety of players, including the end users in creating value. Actually, each image resulting from the internal processing from L16 if coupled with the original images taken by the individual lenses may become a test case for experimenting with new software and this may lead to continuous improvement. Would you be willing to become part of this improvement process by letting third parties to access your image and the source images? May be not all of your snapshots may be ok for sharing but many will and this can be the game changer.
