All posts tagged display

Motorola Battery Indicator- Really a Color Error?

While I do not have the time to respond to everything that is written about PenTile technology, occasionally I will see something that is not quite right which is then picked up by other blogs and gets a life of its own. This can happen from a single photograph. Perhaps you have seen one or more recent blogs that claim that PenTile cannot render color correctly and then go on to illustrate this using the same badly focused photo.

Here is one that appeared today from

As indicated by the logo, this photo seems to have originated with XDA.CN.

Several bloggers have shown this photo of a Motorola smartphone display stating that this proves that PenTile RGBW displays in Motorola’s phones exhibit this type of color error. I would be the first to agree that photographing high resolution displays is very challenging, especially at such high magnification. It is easy to overexpose the image, as was done here. Overexposure can cause color shifts, because the elements in the camera’s CCD of the proper color will become saturated and can no longer respond. Only at the right hand edge, where there is less light due to blur, did the camera capture the correct color.

Battery indicator on Motorola Atrix

With the proper exposure, from the equivalent display in a Motorola Atrix, you can see above that the highly magnified battery icon is rendered with the correct color over the entire area. You will see that there are more red subpixels on the top of this, since it was the intention to shade the green toward yellow at the top. And, here is the same photo, but zoomed in to show just the battery indicator:This photo matches exactly with what I see with my own eyes when looking through an eye loop.

Some of you may say it is different in a Droid X2 or another Motorola phone, but it is the same technology and the same PenTile firmware.

Of course, it was never intended that viewers examine these displays at this magnification, so you will see pattern visibility in these photos that is not apparent to people with normal vision for normal use. Fully saturated green on PenTile RGBW will have a bit of a checkerboard look that can be seen at this magnification. This does not occur with less saturated colors.

Finally, I tried to reproduce the color error by overexposing my photo of this same display and was able to reproduce the effect below. Of course, this will be slightly different for different cameras with different camera CCDs.

Battery Indicator Overexposed - note the yellow center with the green edge.

As you can see for yourself, PenTile RGBW is rendering the colors in this battery icon correctly in Motorola phone displays.

Fonts – comparing RGB Stripe to PenTile

As promised, this blog will illustrate small fonts compared between RGB stripe LCD, PenTile RGBW LCD and PenTile RGBG .

In the comment from Khalid H. he said that Arabic fonts cannot be read on his HTC phone equipped with PenTile, implying that PenTile cannot render Arabic fonts.  Unfortunately, Khalid did not mention what font size or style of Arabic font, and there are so many. I borrowed a clipped section of an Arabic font from Middle East International Services ( I displayed these fonts on my PenTile RGBW LCD and tried to photograph these incredibly small fonts.  Even with the best lens on a Canon SLR, I went beyond the resolution limits of this lens to get a sufficient focus.  So instead, display simulator tools were used to render these characters in RGB stripe, PenTile RGBW and PenTile RGBG as is used for OLED.

I have to agree that I could not read these on PenTile, but then I also could not read these on RGB Stripe, since I do not read Arabic.  Just the same, I am posting the comparison between RGB Stripe, PenTile RGBW and PenTile RGBG .  I compared  these using an eye loop to the actual fonts displayed on our PenTile and RGB stripe screens and the rendering is matched exactly.  As you can see below, the comparison speaks for itself.  PenTile RGBW and PenTile RGBG can render these tiny single stroke fonts equivalently to RGB stripe.  Even the white dot in in the small square is rendered perfectly, or at least it seems that way to a non Arabic reader, but in any case it too compares well to RGB stripe.

The relative brightness is a true representation of how PenTile RGBW will appear on a screen at comparable power.  In each case there are 600 pixels of width, but with PenTile we eliminate one-third of the subpixels, so 600 pixels includes more information than for RGB stripe.

RGB Stripe LCD



For good measure, I did the same exercise with an Arial, English letter font.  Ariel was chosen because there were good options for small single stroke fonts.  I cannot see any problem with how this is being rendered in PenTile.  While some may say that fewer subpixels should lead to fuzzy fonts, it simply is not true.

RGB Stripe LCD



These fonts are zoomed in to a level that one should never look at fonts.  Both PenTile and RGB stripe are looking rather jaggy.  That is true for both RGB stripe and PenTile displays.  Of course, large fonts that are not single stroke, especially when displayed on larger diagonal displays can look better than these small fonts, except that they too will also look more jaggy at the edges.  PenTile is never recommended for low dpi, not because it renders incorrectly, but because the pattern will become visible at lower dpi.

Note that I did not try to test these on phones or with specific types of software, because I don’t have access to the wide variety of phones that some bloggers seem to have.  There are variations in SW and other factors unrelated to PenTile that can make these images look different than the rendering engine is capable of producing.



Update in response to the comment from Khalid H. last weekend.  He suggested that such disconnected fonts were not as difficult and suggested that we look at Arabic Droid fonts which are more connected.  With his help we downloaded this font and generated some text in similar size small fonts to what we used last time.  Then we  rendered these as RGB Stripe, PenTile RGBW and PenTile RGBG to enable a comparison between RGB stripe and PenTile LCD and PenTile OLED versions.  You can see this result below.  Khalid, I hope this is more convincing.

RGB Stripe LCD - Arabic Droid Font

PenTile RGBW LCD - Arabic Droid Font

PenTile RGBG OLED- Arabic Droid Font

Samsung Experience Media Tour

300 dpi PenTile tablet panel

On June 6-7, Samsung hosted a media event in their amazing facility on the third floor of the Time Warner Bldg at Columbus Circle in NYC.

This gave us a chance to show off our 10.1″ WQXGA to editors and bloggers who didn’t have the chance to see this at SID DisplayWeek in LA. Scott Birnbaum VP of New Business Development for Samsung Electronics presented both PenTile and Transparent Display technology that Samsung is now developing. It is always very useful to collect live feedback on our newest displays.

Joel Pollack demonstrating PenTile RGBW display.

Award Winning PenTile Display

It seems that one of the displays using PenTile technology has received an award, the Silver Display of the Year Award (see below).  Actually the award was given to Samsung SMD who created the display for the Samsung Galaxy S that used innovative methods for combining an OLED display with an optically bonded capacitive touch panel that was a big part of the Galaxy S success.  Many bloggers have already noted that this display was PenTile RGBG.  For some time now Samsung SMD has been demonstrating  and selling displays using PenTile OLED in sizes of 3.1-inch, 3.3-inch, 3.5-inch, 3.7-inch,  4.0-inch and 4.1-inch.  It is great to know that the prestigious Society of Information Display has recognized the “exceptional display quality” of this design.

The press release on this year’s awards can be found here:


Silver Award: Samsung Mobile Display On-cell Touch AMOLED

Samsung Mobile Display developed its OCTA – on-cell touch active-matrix light-emitting diode (AMOLED) – display to meet the ever-increasing performance and industrial design demands of today’s modern mobile devices. Unlike prior-generation touch-capable mobile displays, the OCTA display integrates the touch sensor onto the display itself, rather than fabricating the sensor on a separate glass substrate and then laminating it onto the display. This innovation, embodied in the OCTA’s projected capacitive touch detection technology, greatly reduces product weight and thickness, while increasing touch performance and response via the multi-sensor input capability that enables gesture recognition for the most advanced mobile devices. Moreover, because AMOLEDs are self-emissive, no backlight is required, allowing the Samsung OCTA display to deliver very high performance and exceptional display quality, with highly accurate and sensitivity-optimized touch input, in a module that is less than 2 mm thick and provides nearly 100-percent light transmission and exceptional outdoor visibility.

Interface to a PenTile RGBW Display

We noticed some PenTile related comments in the Russian publication at HABRAHABR.RU.  Since so many readers of that site went to our blog, I thought it would be appropriate to answer a question raised there ( shown below for those who can read Russian).

Это понятно: попасть в устройство от Джобса лестно каждому. Но этому дисплею такая честь вряд ли светит: во-первых iPad использует другое соотношение сторон, во-вторых пикселы PenTile сильно отличаются от нормальных. Чтобы работать с таким дисплеем, программа должна понимать, что экран состоит не из пикселов, а из полупикселов двух разных типов. В основном это решается на уровне драйвера, но многие приложения придётся адаптировать. В-третьих такое разрешение даст слишком большую нагрузку для мобильных процессоров, что выльется в лаги и повышенное энергопотребление.
Но успехи в увеличении разрешения дисплеев радуют.

If we understand this correctly, there is some concern about connecting an existing system to a PenTile RGBW display panel.  It was correctly understood at this site that we have, to date, put the electronics for driving PenTile into the driver chip, but it was not clear how the signal had to be prepared to drive this display.

The answer is simple.  One feeds the display input a conventional RGB signal since the display module is plug and play.  Internally to the driver chip, we perform  a gamut mapping from RGB to RGBW.  The driver can handle 24-bit color, but if one puts in less bits of color the module will handle that as well.

Day 1: SID recap

Visitors to the PenTile SID demonstrations at SID

Day 1 at SID is behind us and we’re pleased to say we had an incredibly busy day. The Samsung/Nouvoyance demonstrations were mobbed, almost to the point that it was difficult to walk through virtually all day. The demos of the new WQXGA 10.1″ tablet display, as well as those for the multiprimary PenTile displays, performed stunningly.  We were glad to see some tech industry experts, including some folks from Engadget, come through and take a tour of the new tablet display.

But we were also pleased to see that the PenTile multiprimary display demonstration got a bigger crowd throughout the day. We are encouraged that this means good things to come for our entire line of PenTile displays.

Demonstration of PenTile Multiprimary Display

Thanks to all who stopped by yesterday. If you’re in the neighborhood of SID, stop by our booth #707 and see the displays for yourself!

Regarding WQXGA Pixel Density

I’ve come across a number of comments made online today in regard to pixel density in the WQXGA display, and I’d like to address that. One post in particular, in response to an article on, said the following:

Saying that this screen has a resolution of 2560×1600 is pure marketing BS, and I’m kind of surprised that you fell for it. As you noted, “Or, put another way, a panel with the same number of pixels as a traditional screen will have higher resolution.”

That’s because there are only two subpixels per pixel: half of the pixels are red-green, the other half are blue-white. This means that no single pixel can ever produce the color assigned to it without help from its neighbors. If the pixel at x123,y234 is asked to produce color #ABCDEF, it can’t do that because it has no blue if it’s a red-green pixel, and it has no red or green if it’s a blue pixel. These displays fudge a 2560×1600 resolution by using the missing colors from neighboring pixels to fool the eye into thinking that the other pixel is actually displaying a color composed of all three primary colors. If the pixels are small enough, you may actually be fooled, but the image will be subtly inferior to a true 2560×1600 display.

I appreciate the passion that DOSGuy and others show, but I must disagree with assertions about resolution. They seem to be confusing pixels with subpixels or dots. And DOSGuy, in particular, sounds like he’s of the camp that believes that resolution is based upon counting dots.

That assertion is, however, in contradiction to how the world’s leading metrologists for the display industry, who have written the accepted standard for measuring display resolution. Display resolution is based upon measurement of modulation contrast ratio—more specifically Michelson contrast ratio. There is currently only one display metrology standard organization, Video Electronics Standards Association (VESA), who’s standard is now being combined with the ISDN standard for a new and comprehensive standard to replace the VESA Ver 2.0  standard by an SID subcommittee. Nothing in this standard talks about counting dots—nothing.

So why not? The reason is partially historical. At one time the display of choice was a CRT. It had a Gaussian shaped spot. Two adjacent spots overlapped and compromised resolvability when the overlap reduced  the modulation contrast ratio. The lower limit was then set  50%.

Later passive matrix LCDs came out that had well defined dots in an RGB stripe pattern and people could count dots. A white pixel had a combination of one red, one green and one blue dot. Still there was crosstalk, so modulation contrast ratio could be compromised.

Display metrologists realized that modulation contrast ratio was where the rubber met the road. It was what the eye really saw and what could easily be measured. What good did it do if there were lots of dots that bled into one another for any reason? For many years this definition using modulation contrast ration has stood the test of time with display experts.

PenTile displays meet and exceed the definition of modulation contrast ratio for the resolution as defined. In this case for 2560 x 1600, meaning that one can write a series of 1280 black and white line pairs in one direction and 800 line pairs in the other direction and can write diagonal line pairs in any other azimuth of the same pitch and always meet or exceed 50% modulation contrast ratio. While the spec on speaks of black and white lines the PenTile display can do the same with all other colors.

So can the PenTile display write any color at any pixel? Yes it can. Remember that we are dealing with logical pixels that are comprised of varying numbers of subpixels. What the instrument measures and what the eye sees is the center of luminance energy at each and every one of the corresponding 2560 x 1600 pixels and it is possible to write every color to each of these that one can write to an RGB stripe display. The instruments can measure this and the human eye can see all of these dots of all colors.

For more detail on Michelson contrast modulation measurements and PenTile, please read this white paper on the topic.

It’s definitely not marketing BS — it’s perfectly within the specifications defined by the display industry’s metrology experts.