How to add a new dimension to your print
Lenticular imaging for 3D and animation effects goes in and out of fashion, but it’s proving particularly suited to UV flatbed printers that can print directly onto the plastic lens material. There’s even a system that does away with lenses altogether now. Simon Eccles takes a stereoscopic look.
Printing is an inherently two-dimensional process, but we live in a world of three dimensions (or four if you count time). Wouldn’t it be nice if we could reproduce depth, the third dimension, in print? Well, it is possible, but it’s fiddly. Lenticular imaging offers the only commercially viable route to 3D print today, and this month sees the announcement of a new system that promises to be cheaper, better and tailor-made for flatbed UV inkjets.
Adobe is also wondering out loud about what features it should be introducing to Photoshop over the coming years. 3D image capture and processing is one possibility, though not necessarily for 3D print.
Let’s fill in the background first. Those of us blessed with two eyes experience the world in stereoscopic vision. Thanks to instantaneous ‘post-processing’ by our brains, we experience a sense of depth when we look forwards, so we constantly know how far away an object is, and if it’s moving we can predict its trajectory. The calculations involved make you proud to be human until you remember that when your dog catches a Frisbee, it’s doing it too.
Given that depth perception is so important, it’s disappointing that stereoscopic photography and ‘3D’ printing has never been good enough or compelling enough to capture a wide market.
The original stereo photographs that require special viewing frames or coloured spectacles are obviously a non-starter for general print and display work. The best route to single-image 3D print so far has been lenticular printing, which gives a full-colour single image that can be viewed without special equipment.
With lenticular imaging a series of images from different viewpoints is split into narrow stripes that are interlaced and placed behind parallel clear plastic lenses, which have a ribbed appearance. The refraction of the lenses coupled with the different viewing angles seen by each eye, mean that different interlaces appear as the viewpoint shifts – this can be used for either stereoscopic 3D or simple animation and change effects.
The introduction of UV-cured inks for offset presses, and more recently flatbed UV inkjets, allows images to be printed directly onto the rear surface of the plastic lens sheet material, instead of printing on paper and then laminating it to the plastic. For the past few years flatbed suppliers such as Inca Digital and Zünd have often printed lenticular images as demonstrations at shows. HP Indigo, Presstek and Karat digital presses can also be adapted to print onto the lenses for smaller formats.
The stereo effect actually works best with thicker lens materials with ‘coarse’ lens pitches, which even lower-resolution inkjets can cope with. It’s much better than the thin lenses needed for hand-held viewing in applications such as novelty lenticular greetings cards and book covers.
Whatever the size and process used, lenticular’s commercial impact has remained limited by the high cost of the lens material, while its aesthetic appeal is limited by the material’s ribbed outer surface.
Now HumanEyes has introduced a new ‘Lens-free’ method of creating lenticular images that doesn’t use the expensive moulded plastic lenses. Instead it uses a normal flat glass or clear plastic, with parallel lines printed on the front surface, and the interlaces printed on the back. The lines coupled with the refraction of the clear material act as a cross between a mask and a virtual lens.
Removing the need for moulded lenses means the cost of production will plummet, making lenticular much more attractive for signage. The effect only works when it’s backlit, but this means there’s a lot of potential for indoor and outdoor signs, vending machines and so on.
Anyone with a flatbed UV inkjet could print this – HumanEyes’ UK distributor NCS (about to be renamed Turning Point) has already been running demonstrations on a Durst Rho in its Newcastle showroom. Product specialist Andy Metcalfe says that a flatbed user could be up and running for about £9,000 including training. The software itself (before training and set-up) costs £7,500 for an ‘unlimited’ size version suitable for wide format flatbeds.
Lens price apart, HumanEye’s software price has been pretty high until recently. The original entry level Studio3D edition costs £2,500. This can output proofs up to A2 size on desktop inkjets and paper, with a manually aligned lens sheet provided for checking.
Last year HumanEyes announced an entry level package called Creative3D, which can be bought on-line for just $299 (about £150). This may open up the concept to a much wider audience – previously the ease of use of PrintPro has been rather cancelled out by its off-puttingly high price. The original Creative3D could not output test prints and relied on previewing the effect on-screen through with red-green anaglyph lenses, but in Janary HumanEyes announced a version 2.0 that will output test prints that can be viewed through manually positioned plastic lens sheets.
The idea is that designers set up the images in Creative3D or Studio3D and transfer the files to a user equipped with one of the output-enabled PrintPro editions for final output and printing.
The high price of both print software and lens material explains why so far there are only half a dozen or so UK printers equipped to output this way. However, illuminated signage is one of the most common uses for lenticular, so the new Lens-free method may increase its adoption among flatbed-owning printers.
The HumanEyes lenticular output process was originally secondary to what the company was created to do. Boffins at Jerusalem Hebrew University invented a method of capturing stereoscopic panoramas by moving a single camera through an arc (typically on a boom arm), taking a lot of overlapping shots and using software to calculate a partial 3D image. HumanEyes was set up to market this commercially and lenticular was chosen as a printable output method.
Attractive though the idea of stereo/3D printing may be, it’s not actually what’s driving lenticular today, whether from HumanEyes or rival lenticular software. Ironically most HumanEyes users find the boom arm stuff too fiddly and have preferred to use the software’s secondary abilities to create flips (where completely different images suddenly appear) and morphs, which is basically a smooth multi-frame transition or animation). These effects aren’t new at all, though the HumanEyes software makes them particularly easy to achieve. HumanEyes has also recently added the ability to convert conventional ‘flat’ 2D photographs into pseudo-3D by cutting out objects in the images and arranging them on layers at different simulated distances.
What’s the Adobe connection? Nothing you can buy yet. But Adobe Labs in California has developed a prototype lens and digital camera combination that can capture depth information within a single photograph. The real purpose is to allow the image to be sharply focused by processing in the computer, rather than depending on the lens. A short QuickTime demonstration sequence shows the focus moving all over the place. Right at the end, it reveals an intriguing side effect: the whole image rotates slightly in 3D. The camera captures a ‘2.5D’ (partial 3D) image just like lenticular can reproduce.
‘We did tons of research and found the number one reason for bad photographs is the focus going wrong,’ explains Dave Story, vp digital imaging product development at Adobe Labs. ‘The camera has not progressed much since its invention, except to go from film to digital. Essentially we have an automated pinhole. But what really has changed is the software. I think the future of photography will be in computational photography – both the CPU inside the camera and in post-processing. Lens makers are already reaching the limits of what optical glass can do, but we have not scratched the surface of what software could do.
The Adobe Labs solution is nowhere near production yet, but here’s how it works. Adobe has sourced a huge 200 million pixel image sensor from somewhere (it won’t say where, so it’s probably military). It’s teamed up with a lens manufacturer (again anonymous) to create a honeycomb optical array of 20 prisms and lenses that fits in front of a more or less standard camera lens. The prisms are all angled differently, so in effect each one sees the image from a slightly different angle and focal point. ‘It’s like an old stereogram camera on steroids,’ says Story. Technically it’s a variation of a largely experimental process called plenoptics, which in earlier forms dating back to 1908 has put an array of hundreds of microlenses just in front of the film or sensor.
Fitted with this array, the 200 mp camera shoots 20 x 10 mp image segments in a single exposure. This is a 600 MB image, but Adobe is looking to a future when processors and storage will be able to cope. The really clever bit comes when you import that mega-image into Photoshop. Adobe Labs has created a prototype plug-in it calls a ‘Focus Brush.’ Using that, you can set the focus point and depth of field anywhere between close-up and infinity. You can have several sharp points at different distances, or angled planes of focus.
Story says that it will take years to go from prototype to a commercial camera, and for the processing to make it into a future Photoshop. He predicts that the cost and complexity means it will probably be pitched at professional applications before consumers. ‘We’re trying to help people do the impossible with Photoshop in future,’ he says. ‘If we don’t let them do things they never could before, they won’t keep buying it!’
Back to top
|