Infocomm Issues Draft Standard for Contrast on Direct View Displays – But is it Right?

InfoComm has issued (July 9, 2015) a new draft standard called “Direct View Display Image Systems Contrast Ratio”. It can be reviewed here.

But some believe the system contrast specification for watching movies is just plain wrong – and they have posted a video showing side-by-side examples to prove the point (see videohere). You should take a look and weigh in on this debate if you care about image quality.

The new standard from Infocomm is based in large part on the Projected Image System Contrast Ratio (PISCR) standard released about 5 years ago. This standard uses the ANSI 16-zone (4×4) checkerboard pattern and measures the contrast in the room to determine the system (room + screen + projector) contrast. Further, it then defines certain levels of system contrast for various tasks. The new direct view standard mimics these categories – and system contrast levels, but adds a 5th category – content creation for corporate communication. The activities and specs are noted below:

  • Passive Viewing requires a minimum contrast ratio of 7:1
  • Basic Decision Making requires a minimum contrast ratio of 15:1
  • Analytical Decision Making requires a minimum contrast ratio of 50:1
  • Full Motion Video requires a minimum contrast ratio of 80:1
  • Content Creation for Corporate Communications requires a minimum contrast ratio of 250:1 with a maximum black level of 1.25 cd/m²

The author of the video, Darrin Perrigo, shows two different projected images with very different ANSI and sequential (full white/full black) contrast ratios, as shown below.

Perrigo 1

The system on the left meets the PISCR specification for movie viewing while the one on the right does not. Yet most would agree that the image on the right is more desirable.

I asked Perrigo to provide more details on his testing and measurements, which he then supplied and which are summarized in the table below. A single projector was used with identical screens in two separate rooms for the above comparison.

Perrigo 2

These results seem very confusing to me as the black level in the living room is higher than in the home theater, which is counter intuitive. In addition, the ANSI white levels should be about the same for both rooms unless the projector has a very low lumen output. While Perrigo acknowledged the data above, he did not offer a clear explanation of the results, so measurement error is suspected and I leave it to you to interpret these results.

To get the Infocomm perspective on this issue, we contacted Joel Silver, the founder of the Imaging Science Foundation (ISF), a well known and respected expert on display metrology, and a member of the task force for the standard.

“Darin makes an excellent point about the sequential contrast and using it as part of the movie viewing criteria”, noted Silver. “But his results were in a dark room, and there, these two types of contrast do matter. For commercial AV, the lights need to be on for the most part and watching movies in a dark environment is not really done, which is why there is no real focus on sequential contrast”.

If movies are shown, they may be embedded in a PowerPoint slide and it is not practical to dim the lights for short clips, noted Silver. “It is the room and ambient lights that dominate the contrast in ProAV, so ANSI contrast is more relevant. For contrast improvement in projection, designers and integrators want to first try to fix the room to reduce reflections from walls and direct illumination on the display or projector screen. Next, they need to have a projector with a sufficient amount of light and third, you can spec a gray or light-rejecting screen to help improve contrast”, said Silver in our phone interview.

“For home theaters, Darin’s comments are right on the money as these are controlled light environments,” continued Silver. “Here, I recommend a minimum of 100:1 ANSI contrast and 1200:1 sequential contrast, which is the DCI theatrical spec. My aspirational spec is 150:1 ANSI contrast and 2000:1 sequential contrast”.

Silver also noted that the current CEA/CEDIA CEB23 Home Theater document recommends the high end 150:1 Intraframe (ANSI) and 2000:1 sequential spec. Over recent years I have learned that both a minimum performance spec and a high end spec will be better guidelines for the next revision of that document.

Perrigo suggests that an acceptable set of specifications for movie viewing should be a 20:1 ANSI contrast ratio and a 150:1 sequential contrast ratio – a far lower benchmark.

The new Infocomm Direct View doc also adds a fifth element called Content Creation for Corporate Communications. Silver explained that this was added to cover the scenario of a video editor for a corporate or church video production where the editor is working with a monitor that may have decent ANSI contrast but has elevated black levels. Doing the editing with this monitor could lead to using content that has a lot of dark noise and lack of dark details as this would not be apparent on the editor’s monitor. However, it could show up later on a better display system. That is why the maximum black level spec was added. Silver would like to see a minimum black level spec added to the “watching video” scenario too at some point.

If these issues are of interest to you or your organization, please consider joining the task forces at Infocomm, CEDIA and CEA to help develop the next standard. These are volunteer activities so consider contributing – if not for yourself, then for your company’s business interests. The CEA/CEDIA Home Theater committee will be meeting at the upcoming CEDIA show, if you are attending… – Chris Chinnock

Join our list

share this post

About the Author: Chris Chinnock

Chris Chinnock the founder and president of Insight Media. His areas of focus include the 4K ecosystem, laser displays, 3D/Light Field/Holographic displays, advanced imaging technology (HDR, HFR, WCG), VR/AR/MR and emerging technologies and products in the broadcast, consumer electronics, ProAV and the display industries.