Panoramic pixels

Out of the blue, my friend Jonah sent me this picture taken from a kite.

1. It's a beautiful picture, taken on a beautiful island, on a beautiful day, at high tide.
2. I spent two summers in the home "De Wokkel" in this picture. Now a few years later a friend of mine who lives a continent away sends me a great aerial photo of the place. Does chance exist?
3. This very sharp picture is taken from a kite, apparently something very hard to do. Were any special techniques like super resolution used? Will we see image stabilization and super resolution techniques become standard features on our cameras soon?
4. There's an interactive panoramic version of this photo available from the author here. It takes a while to load but it's worth the wait.
5. Bonus points if you spot a car in this picture. Visitors can't bring their cars to the Schiermonnikoog island, and most of the locals don't have one either.

Mirrory pixels

IMEC just announced a new 11 megapixel CMOS-manufactured micro mirror array. Texas Instruments is quite successful with its DLP micro mirrors frequently used in projectors. IMEC now claims twice the pixel density of competing technologies.

Yet another pixel post-processing acquisition

Sigma Designs acquires Gennum, ST acquires Genesis, Zoran acquires Let It Wave and now IDT acquires Silicon Optix. Silicon Optix was another company that focused on developing the highest quality video format converters. This acquisition is a bit different from the others since IDT doesn't have any digital media application processors for set top boxes or Blu-ray players that need integrated post-processing functionality. Instead, IDT is likely to get synergy out of this acquisition by combining the post-processing products with their mixed signal consumer video products.

Parallel painty pixels

Every engineer loves Adam and Jamie from the Mythbusters, and now we video engineering guys have a reason to love them even more. Here's a clip showing a display Adam and Jamie built that uses 1100 paintball guns to draw a reproduction of Leonardo's Mona Lisa with a refresh rate of 80ms. They only refresh once, though.

Audio-focused pixels

Liquid lenses have been around for some time. Varioptic applies a voltage to a lens made of water and oil to change the optical properties and achieve focus or zoom. The Rensselaer Polytechnic Institute created a new optical system using a liquid lens and a small speaker. Instead of applying a voltage, they apply sound to move the water droplets to achieve a focusing effect.

Varioptic hasn't really made an impact on the camera module market yet. Will Rensselaer's technique finally bring liquid lenses into mass production?

Another pixel post-processing company acquired

Some time ago I wrote about post-processing companies being acquired by larger multimedia companies such that they can address the whole video pipeline, from image capture to display. It's a bit late, but for completeness sake I should really mention that another post-processing company was bought by a bigger multimedia company: Zoran buys Letitwave for $27M. Let it wave had Prof. Stephane Mallat on its staff, a world renowned expert on using wavelets for signal processing.

Watery pixels

In Canal City, a big mall in Japan, they built a cool new display that drops water in such a pattern that it displays images.

We've now seen wooden pixels, ping pong pixels and watery pixels here. What's next?

Big and expensive pixels

Today I was at the popular Mediamarkt electronics store in Eindhoven, the Netherlands, where I saw the "largest TV in the world"; a Panasonic 103 inch plasma display. That's well over 2.5 meters diagonal for us metric-centric people. It's got 1920x1080 spatial resolution with 12-bit per color channel resolution. The price tag was also quite hefty at 79,999 Euros ($120,000 dollars). For such an amount you can buy 200 22" flat screen TVs instead. If my math serves me right, that makes for a 580 inch (14.4 meter) diagonal TV. Quite a bit bigger and with a much higher spatial resolution.

I was surprised that the display didn't look more grainy though, even when standing relatively close by. In displays the Megapixel race doesn't follow the same pace as in the image sensor world. Still, with such large screen sizes, will we soon need to capture, store and transmit video with bigger than 1920x1080 resolutions?

Direct pixel manipulation

The below video shows a new intuitive and simple way of browsing through video material. Instead of browsing through time, by dragging the scrollbar or time bar, you can simply drag objects in the video. The video says it all. Funny side note: first author of the publication is Pierre Dragicevic. More videos and information here. You can even download their free "DimP" player.

Scalable pixel product

I've spotted the first real SVC product announcement. At NAB, MainConcept, a DivX daughter, presented their SVC implementation. SVC is the new video coding extension to H.264 that doesn't bring higher coding efficiency, but actually worsens it. I wrote about it earlier. The big benefit though is that you can decode parts of the bit stream in case you only need a smaller resolution picture. MainConcept writes "creating an SVC file only causes an approximate 10% file size increase compared to a regular H.264/AVC file".

10% is a lot in video compression.

Thin pixel machinery

In a recent post I briefly wrote about Sony's new OLED display. One of the main hindrances of market adoption of this technology is cost. But now DuPont and Dainippon have announced they're working together to develop equipment specifically to manufacture OLED displays. The plan is to make machines that basically print the display, using techniques similar to ink jet printers. The ultimate goal is for the OLED displays to achieve LCD price points. My prediction is that within 10 years OLED will have displaced LCD.

Shiny pixel fab being built

A little while ago I reported about the new IMOD display technology from Qualcomm, which should yield high quality displays that consume very low power. I saw the displays in action at the Mobile World Congress in Barcelona, but was unimpressed. The displays shown there were small, and perhaps similar to the LCDs that displayed the time on my watch about 30 years ago. Very few graphics and no moving pictures were shown. Still, LCDs have come a long way since the seventies, so perhaps the new IMOD displays will have a bright (no pun intended) future also.

What's the next step? High volume, low cost production facilities. Yesterday Qualcomm and Foxlink announced just that. They will jointly build a new dedicated IMOD fab in -- where else -- Taiwan. The fab is expected to be operational in 2009. There's no mention of how many units the fab can produce.

Can Qualcomm, the CDMA wireless communications company, be successful at entering such a new market? Will IMOD take off?

Thin pixels

Sony Mobile Display showed a 0.2mm-thick 3.5inch OLED display the other day in Tokyo. The resolution is 320x220, and since it is OLED I expect the picture quality to be quite striking. In contrast to LCD, OLED doesn't need a backlight, which means it looks like color-printed paper and is very readable in sunlight.

Lossy versus lossless pixels

The other day my dad -- who speaks English okay -- thought "lossless" meant "loose less". That's what prompted me to write this entry. There are two different ways of compressing video (and this holds for audio too). Lossless and lossy.

When compressing a video sequence with a lossy method, then decompressing it again, the decompressed images will be close to the original, but not quite the same. Even when using a low compression factor, where the decompressed images can't be distinguished from the original simply by looking at them, there is a difference, which can be computed. Almost all of the video compression methods we use today use such a lossy method, where the least important image data is thrown away.

In the lossless case there's no data lost. The compressed/decompressed images are exactly the same as the original. There's a big penalty though: lossless compression methods don't compress very well. The resulting files take up much more space or bandwidth. Lossless image or video compression methods are still used though, for instance in the medical field, and when storing fingerprint information.

Very very small pixels

Researchers at Stanford University are developing a multi-aperture image sensor which groups arrays of 16x16 pixels, then puts a tiny lens on each group. Their 3Mpixel image sensor in this way includes a total of 12,616 lenses, compared to a shabby single lens commonly found in cameras. The benefits are plentiful. The simpler electronic design means the pixels can be 0.7um, much smaller than Kodak's 1.4um pixels that I posted about earlier. Camera modules incorporating this technique can be made even smaller, cheaper, more robust, and, most importantly, grab better pictures. Instead of taking a single snapshot, the camera actually takes 12,616 pictures, which can be combined with digital image processing techniques to capture 3D image data, to accurately control depth of field, focus, etc. With enough image processing power available in the camera, this opens up a whole world of new possibilities.

A high level overview of the work can be found here and their technical ISSCC paper can be found here.

Pixels better than real life?

According to this survey by Motorola, Americans would rather watch the Superbowl on an HDTV than in person. “The survey results really speak to the popularity of high-definition programming,” said Doug Means from Motorola.

That's a lame study and a lame statement. The results of the survey don't say anything about the quality of HD video and how close it gets to being there. Yes, quite a few people would rather sit in their homes than take a plane and sit on a plastic seat for hours watching the game. Yes, a big screen TV presents a much better picture than an old Philco Predicta. But no, nothing compares to being there. And I can say that without having ever been to a superbowl game.

Pixel compression over time

Here's an interesting graph from Harmonic that I sometimes use in presentations. I often misplace it, so I figured I'd stick it here on this blog. That way I can always find it. The graph shows that video compressors are not all the same. They can be improved over time. This is an important fact for chip makers, since developing a complex chip these days often takes well over a year, and is then sold in the market for a year or so after. The longer you can keep your chip in the market, the more you will sell! If your chip includes a software programmable video subsystem, you can still take advantage of algorithmic improvements, just like Harmonic did, and deliver better video quality.

Very small pixels

A few weeks ago, Kodak announced their new 5Mpixel image sensor at the Mobile World Congress. The sensor has a 1.4 micron pixel size (1.4 by 1.4 micron) which means the sensor can fit in a 4x4mm camera module. That is about the size of a regular black ant. I am sure a bigger ant could carry such a camera. The Kodak sensor has some novelties. There is a new color filter pattern, which includes a "white" photocell receptor instead of just measuring the amount of red, green and blue. That will require quite some changes to the image processing algorithms. Another novelty is that the sensor measures darkness instead of light. Apparently that can be more accurately implemented in silicon. Like most new sensor introductions, Kodak promises higher quality images than anyone else.

Micron just announced that it spun its image sensor business out into a new company called Aptina. The business will be run by Micron's Bob Gove, who was previously at VLIW processor company Equator. Micron says they have already sampled an even smaller 1.2 micron pixel, which in the same 4x4mm tiny camera module would yield a 7Mpixel sensor.

Pixel etymology

Did you know the word pixel is derived from "picture element"? Here's a long video that details a search for the history of the pixel, by Richard Lyon. Lots of well known names in the field of video and graphics are mentioned. To skip over the introduction go to 2:20.

Perfect pixel patent

As early as 1929, Ray Davis Kell described a form of video compression and was granted a patent for it. He wrote, "It has been customary in the past to transmit successive complete images of the transmitted picture. [...] In accordance with this invention, this difficulty is avoided by transmitting only the difference between successive images of the object." Although it would be many years before this technique would actually be used in practice, it is still a cornerstone of many video compression standards today. It's the reason why video using MPEG can be compressed roughly a factor of 10 better than JPEG-compressed still images.

What technique can provide another magnitude of improvement in video compression?

My prediction is that we need to change focus from optimizing for best peak signal to noise performance to optimizing for psycho-visual perception. I.e. "how good do the compressed images look" instead of looking at minimizing the mathematical difference between the original and compressed imagery.