Category Archives: Technical

“No glasses” 3D Holograms?

Light Field Labs has raised an additional US $28 million in funding to develop and produce free air holographic display technology. They are said to have a working prototype now and the additional funding will enable them to scale up to an actual product.

The aim is to create holographic objects that appear to be three dimensional and float in space without head-mounted gear such as augmented reality or virtual reality goggles.”

Source: Light Field Labs : 3D Holograms no glasses Deep Dive – fxguide

The principle people behind the technology had developed the Lytro camera technology. As best I can tell, it may be similar to a digital implementation of a conventional, analog, film-based hologram. In the original hologram technology, you look at a flat image that is, basically, like a window pane. As you move to the left or right, you see the true 3D image visible from that point in space. In the laser-based hologram, the window pane is a film that has recorded light interference patterns.

From the description down the page, here, my interpretation is they have created a currently small window pane that is replicating the light interference hologram concept, but in the digital domain. Obviously, it takes a tremendous amount of computational horsepower and for video, high bandwidth, both of which are becoming available as tech advances.

I presume, also, that this technology can be used to project objects in front of the viewing plane, as is done in stereoscopic 3D. In other words, actors or objects can be appear to be between you and the viewing screen – or behind the screen.

This tech creates true 3D that does not require glasses for viewing.

Common photography aspect ratios and print sizes are arbitrary

Still photography and motion pictures have, over history, used aspect ratios such as 4:3, 3:2, or for printing 4×5, 8×10 and what not.

These choices were arbitrary – based on practical design and implementation considerations of the time.

The popular 8×10 paper size came from how fine paper was originally manufactured and sliced down to size by hand, in Dutch paper mills and corresponded to the equipment size readily handled by the length of the arms of the mill workers. These cut 8×10 sheets were later cut to create 8×5 sheets, which in turn were sliced to 4×5 sheets. (I could not verify these claims independently but could not dispute them either. Of interest, the 8 1/2 by 11 inch sheet of paper we take for granted also seems to have come out of similar issues and stuck with us because of practical issues regarding manual typewriters, issues that no longer exist today.)

The 35mm standard came from early still photo film which happened to be 70mm wide, but was split down the middle by Thomas Edison to save money for making a movie film. After adding holes along side the film for pulling the film through their movie camera, the image area became 24mm wide measured across the film. Each image was limited to 18mm in the length direction – becoming a 24 x 18mm or 4:3 aspect ratio image.

This film was then adopted for new still cameras (Leica) which chose to double the 18mm to 36mm, hence 24mm by 36mm (the well known 35mm format) in a 2:3 (or 3:2) aspect ratio. The 1:1 ratio photo came from waist-level viewfinder cameras – since it was not easy to turn the camera sideways, they chose a 1:1 ratio.

The result is that today’s modern digital camera and print aspect ratios are arbitrary and based on design choices that occurred out of practical considerations in the 19th century and the early 20th century.

Source: history – What historic reasons are there for common aspect ratios? – Photography Stack Exchange

And then there is the 16:9 aspect ratio of HD, which is the compromise that came out of a committee that wanted to create a new TV standard to deal well with older 4:3 content and wide screen content which is wider than 16:9. Basically, an arbitrary compromise value.

There is also similar information on how did we end up with audio reel-to-reel tape recording at 7 1/2 inches per second? I was told it was because this was the speed at which 16mm film, with an optical soundtrack on the film, operated. I could not quickly verify if this was true though and could only work out that 16mm film seemed to go through at 7.1″ inches per second at 24 fps.

Using 2 cameras to create fake narrow depth of field images

Small sensor cameras – such as smart phones and point and shoot cameras – are unable to create significant blurring of the background or foreground. Narrow depth of field is mostly limited to large sensor cameras – or to long telephoto shots.

But, two camera sensors may be used to measure depth in the scene. One camera is used for the actual photo and the second for depth. Parallax, or the difference between the two camera images, varies by distance to the subject. This information is used to blur the original image based on distance to the subject. (Blurring is done by averaging local pixels together using a simple average or a weighted average.)

This means that software creates the narrow depth of field effect, rather than large sensors and expensive lenses.

The HTC One M8 smart phone has this feature today. The linked article gives examples of how this works, in practice. Take a look at their sample photos!

We compare the HTC One M8 camera with a Fuji X-M1 to see what its bokeh-style effects are really like.

Source: HTC One M8 Camera vs A Proper Camera: Fake Bokeh On Trial

Note that if the cameras are very close together, as is typical on a smart phone, the ability to accurate measure distance a long ways from the camera is greatly diminished. Image resolution and interaxial spacing both impact the capability of this feature.

Rumors are that the iPhone 7 will feature dual cameras for the same reason – to create narrow depth of field photos using tiny sensor cameras built in to the phone.

Currently, the best narrow depth of field comes from DSLR full frame cameras and expensive, large aperture lenses.

But post processing software is eliminating many advantages of the full size cameras. Modern post processing noise reduction enables many small sensor cameras to perform more like their big cousins in low light. And now, with dual cameras and depth processing, little cameras may soon deliver narrow depth of field at lower cost than the big guns.

This should be worrisome to the DSLR makers. Particularly as increasing numbers of shooters would prefer to travel light – and not have to carry big camera bodies and heavy lenses.

StereoStitch Introduces Real-Time 3D VR 360 for live 3D VR

Source: StereoStitch Launches a Real-Time 3D 360 Video Stitching Software for VR Live Streaming

Playing 3-D video games can boost memory

Playing 3-D video games can boost memory formation says a study from UC Irvine

Source: Playing 3-D video games can boost memory formation, UCI study finds

The study is published in the Journal of Neuroscience.

(Go Anteaters! My undergrad degree in computer science was earned at the University of California, Irvine 🙂 )

Canon announces 250-megapixel image sensor

Canon has today announced a new ultra-high resolution CMOS image sensor that packs approximately 250 megapixels into an area smaller than a United States postage stamp.

Source: Canon announces whopping 250-megapixel camera sensor – TechSpot

Low light image quality test: Nikon 1 V2 and J1, Lumix GH-2 and GH-4

I tested the low light image quality of the Nikon 1 J1 (electronically the same as the Nikon 1 J2 and the Nikon 1  V1), the Nikon 1 V2, the Lumix GH-2 and the Lumix GH-4.

The Nikon 1 is a 1″ sensor camera with a 10 MP image resolution and the V2 has just over 14 MP. The Nikon 1 is also an interchangeable lens camera.

The Lumix cameras are micro-four thirds sensors. In the 3:2 ratio in which these photos were taken, both have just over 14 MP of image resolution.

The Nikon 1 camera images were shot with the 1 Nikkor 18.5mm lens, at f/1.8 to f/2.5. The cameras were set to Program mode. White balance was set to Auto.

The Lumix cameras were shot with the Lumix 14-42mm kit lens, at 25mm (equivalent to the Nikon 18.5mm lens field of view) at about f/5.6. The cameras were set to Program mode and white balance was set to Auto. I realized afterwards the GH-4 was also set to a “Custom” photo setting that I use to reduce the highlights so it was not quite identical to the GH2.  The 1 Nikkor prime lens is also sharper than the Lumix kit lens.

These are 1:1 extractions of RAW images from Lightroom 5.x. All shots are at ISO 800 as my goal was to test low light situations.  Noise reduction and sharpness setting were what ever the default was at – thus, no attempt was made to clean up the noise. These tests are not laboratory quality – but the kind of tests that us hobbyists do to better understand our gear!

You can click on any photo to see the full size image.

LOWLIGHT EXAMPLES AT ISO 800

Nikon 1 J1

J1-800

Nikon 1 V2

The V2 shows a slightly higher noise grain than does the J1 (V1 equivalent).  However, the J1 is a 10 MP sensor and the image enlargement is not the same. When the J1 is resized to match the V2, the noise grain is closer in appearance to the J1.

Also note that in the smaller sensor J1 and V2, magenta chromatic aberration is apparent at lower left. This is easily fixable in Lightroom.  Also, those faint pink splotches captured in the V2 images were captured to some degree with all the cameras. There is a reflection from something going on – I just never noticed it before.

V2-800

Lumix GH-2

Interesting to see that the GH-2 had a lower exposure selection. Not shown, but an open window to the left of this section of the photo was bright for all of these photos.

There appears to be a very slight bit of chromatic aberration at lower left or at least some slight flaring at the high contrast points.

Lumix800

Lumix GH-4

The GH-4 did better on the auto exposure, white balance and the chromatic aberration. Even though the GH-4 is using the exact same lens as was used on the GH-2 photo above. That implies the GH-4 is doing some image processing on this RAW file that the GH-2 does not do. While there are differences in white balance making the grain harder to see in the GH-2, manually adjusting the exposure in Lightroom showed fairly similar levels of grain.

GH4-800

LOWLIGHT COMPARISON J1 VERSUS V2 AT ISO 1600

J1 ISO 1600 RAW

J1-ISO1600

J1 ISO 1600 RAW processed with LR noise reduction

J1-ISO1600-Processed

V2 ISO 1600 RAW

V2-ISO1600

V2 ISO 1600 RAW processed with LR noise reduction

V2-1600Processed

IMAGE RESOLUTION 1:1

The following is not a fair test. The Nikon 1 cameras used the 18.5mm prime lens while the Lumix used the 14-42mm kit lens. Unfortunately, I do not have a Lumix prime lens (25mm) which is needed to make this test fair.

What this does show is that the Nikon 1 with the prime lens is very sharp – perhaps sharper or at least on par with the 4/3ds camera with the kit lens. Each photo taken with the lowest ISO setting.

Nikon 1 J1 – ISO 100

J1-100

Nikon 1 V2 ISO 160

V2-160

Lumix GH-2 ISO 160

GH2-160

IMAGE RESOLUTION 3:1 ENLARGEMENT

Nikon 1 J1 – 10 MP sensor

J1-100-3

Nikon 1 V2 14+ MP Sensor

Because the J1 has fewer pixels, the 1:1 image above shows a wider area. Still, it is surprising how good the 10 MP image looks compared to the 14+ MP image.

V2-160-3

J1 ISO 100 Resized

With the images reset to roughly equal sizes, the greater resolution of the V2 becomes apparent. Here, the J1’s 10 MP image is enlarged to match that of the higher resolution V2. Bear in mind that these are really bad case/low light situations too.

J1-100-3-equal

V2 ISO 160 

If you look carefully, you can see slightly more detail on the ceramic cup, at left, and you can see more detail in the writing on the spice container in the middle.

V2-160-3-equal

J1 3:1 on the Creole seasoning container

J1-equal

V2 3:1 on the Creole seasoning container

While the resolution improvement in the V2 is visible, it does not have nearly as much visual impact as you would expect in going from 10 MP to 14 MP. Note the word “Original” in the yellow band area – the word “Original” is readable in the 14 MP version but not so well in the 10 MP version. Still … not a lot of difference, is there?

V2-equal

Lumix GH-4 ISO 400 3:1

Even at ISO 400, the GH-4 looks slightly better than the Nikon 1 V2 at ISO 160  (above). I did not shoot this test photo at ISO 200 (or using the expanded ISO setting on the GH-4, I could go as low as 100 – but there is apparently no improvement in image quality at that setting).

GH4-400-3

I did not shoot the test photo with the GH-2.

Conclusions

Small sensor cameras are doing better and better as each new product is announced.

The Nikon 1 J1 is already up to the J5 generation with 20+ MP and a better low light sensor. The V2 is presently up to the V3, with the V4 rumored to appear soon (and assumed to be similar to the J5 sensor).

The Lumix GH-2 is now up to the GH-4 generation.

In well lit areas, the Nikon 1 V2 appears to hold its own very well against the larger 4/3d sensor (the GH-2 and V2 are roughly comparable in that they were both sold around the same time). I have generally avoided shooting at ISO 800 and have probably never shot at ISO 1600! However, after these tests, I am comfortable that I can get decent results on any of these cameras at ISO 1600.

In low light, the larger sensor of the GH-2 and the GH-4 cuts the noise. You can probably shoot the GH4 at an ISO setting double that of the V2, for the same noise level. Not a big surprise.

The 4/3ds sensor also has noticeably wider dynamic range but none of these tests demonstrate that feature.

None of these cameras compete directly with full frame sensor cameras and their better low light capability. But then, they cost a tiny fraction of the cost of the full frame camera. If you are not shooting extreme low light situations, the smaller sensors may be entirely fine – which is the case for me.

I did these tests because I am thinking I may carry only the Nikon 1 V2 on some future trips. Not only is the camera small, but the lenses are small too and weigh very little compared to larger formats.

In the real world where most of us post photos to FB or Flickr, the image resolution we look at with this detailed pixel peeping just does not matter. These photos will all look fine on line.

As long as I am not shooting a lot of low light situations, there is not a big difference in image quality and I can usually control for the slightly narrower dynamic range of the 1″ sensor by shooting RAW and if necessary, using exposure compensation to control for bright highlights and dark shadows.

How important is lens sharpness?

Perhaps not as much as we’ve been led to believe: Lens Sharpness.

Ken Rockwell points out that in most real world scenarios, lens sharpness is just fine. Today we pixel peep to the absurd dimension, finding limitations that may not matter. And once we find the limits, there are often simple work arounds – such as stopping down the lens by a stop.

He also notes that sharpness tests are too often done by photographing a flat test chart, which bears little resemblance to our 3D world where everything is not always in perfect focus. There is a lot of interesting info in the above linked article.