Back To Listings RSS Print

Hold on a moment

Shedding some darkness on sample-and-hold displays.

By The Sony Tech Guy | July 27, 2011

image

This is a beautiful picture.

Ironic but true: CRT phosphors flash briefly and then go dark for most of the frame duration. Film projectors also go dark when the film gets pulled down to the next frame. Isn't the whole point of these displays to produce light? After more than a century of flickering images, the present generation of monitors is built on sample-and-hold technology, such as LCD and OLED. Today's monitors can show pictures that are absolutely unblinking. Is constant illumination display nirvana? Actually, no. The purpose of video displays is not simply to produce light. It's to trick the human visual system into perceiving a sequence of still images as continuous motion. And that's a good reason to embrace the darkness.

Vision in the real world



To appreciate how darkness works, consider natural vision. In life, we see motion as a smooth, continuous flow. If we could chart the progress of a moving object, with time on the horizontal axis and the object's position on the vertical axis, we'd see a smooth line (for constant velocity) or a smooth curve (for acceleration or deceleration).

image

As we track object motion in the real world, it generally appears smooth.


Sample-and-hold displays



If we configure our new, sample-and-hold displays for a continuous, unblinking presentation, we actually create a new problem. As our visual system attempts to track moving objects based on a lifetime of experience with natural scenes, it expects continuous motion but the display shows step-wise motion.

image

When a sample-and-hold display generates an unblinking image, the motion is portrayed in discrete steps.



During the course of each step, the object's fixed screen location hardly ever matches the constantly moving location our brain expects to see. This difference between expected and actual locations confuses the human visual system. We perceive the result as motion blur.

image

The human visual system is tripped up by the differences between the expected and actual locations of objects on the screen. These differences, which we experience as motion blur, are shown as shaded areas in the diagram.



There are other types of motion blur. For example, we can create motion blur in the camera, where it may well be the deliberate choice of the director of photography. We can get motion blur in the display if the pixels are slow to transition from one frame to the next. But this specific motion blur is the result of the human visual system trying to make sense of a highly artificial presentation.

Page 1 of 2 pages 1 2 Next »

Share This

Back To Listings RSS Print

Get articles like this in your inbox: Sign Up

Disclosure, to comply with the FTC’s rules 16 CFR Part 255 This article was either written by Sony employees or for Sony by an outside contractor. It is intended for the Sony Channel on ProVideo Coalition, which Sony sponsors.

Comments

IEBA: | August, 03, 2011

Interesting concept.

Now tackle the issue with regard to scanning imagers versus progressive displays.

Tube to tube worked well because both were scanning and corrected the inerrant distortion.

CCD imager on tube displays was wrong because it’s a global shutter on a scanning display.

CMOS imager on LCD / OLED display is wrong because it’s once again a scanning imager on a global display.

Once we get global shutters on CMOS chips, then, not only will reality be imaged truthfully, but the images captured will match the displays they are seen on.

The Sony Tech Guy: | August, 18, 2011

IEBA:

Thank you for your thought-provoking comments, which I referred to some of our display technology experts.  I hope you find their response as interesting as I did. 

++++++++++++++++

“Tube to tube worked well because both were scanning and corrected the inherant distortion.”

Actually this is not true. Tubes were driven by analog saw waveforms that may or may not match the properties of the yoke. Both the camera tube and the deflection system of the CRT always had geometric distortion. In fact, the last BVM CRTs still had significant errors in geometry at the outer horizontal edges.
If you are talking about timing or rolling shutter artifacts, this is still not true. Even when a tube camera is used with a raster driven CRT, rolling artifacts are still easily seen due to velocity differences between the raster scan of the camera and CRT with relation to a specific detail in the image.
You can take a tube camera, feed a CRT monitor and still see a very similar effect with a quick pan of the camera.

“CCD imager on tube displays was wrong because it’s a global shutter on a scanning display.”

It depends on your point of view. In moving to a fixed matrix imager, you now removed all of the static geometric errors at acquisition. This made a significant improvement in image quality.
With regard to rolling shutter artifacts, these were no more exaggerated with the use of a CCD imager than they were with the Plumbicon or Saticon three-tube system and significantly better than any striped single tube system.

“CMOS imager on LCD / OLED display is wrong because it’s once again a scanning imager on a global display.”

No. You are now using fixed matrix imaging at both ends of the system. What you are seeing now is a much more accurate representation of your acquisition. The fact that rolling shutter artifacts are more visible is not a factor of the display technology rather than the fact you can view your content much more accurately. Should line 42 skew from line 1121 is simply a matter of physics, not an inherent flaw in the system.

“Once we get global shutters on CMOS chips, then, not only will reality be imaged truthfully, but the images captured will match the displays they are seen on.”

Maybe, but don’t count on it. If it were that simple, it would have been done a while ago.

Please login or register to comment