HDMI, DVI And HDCP: A Hands-On Update

To say that my earlier writeup (discussing the difficulties interfacing a Toshiba HD-A1 HD DVD player to a Samsung HP-P4261 plasma display via a Belkin AV24502 PureAV 3-into-1 HDMI switch intermediary) generated industry interest would perhaps be an understatement! Analog Devices (who may be Belkin’s silicon supplier, although given the switch’s age, I suspect Silicon Image) offered to run the switch through the company’s internal validation lab. Belkin volunteered to do its own testing of the AV24502, using in-house Toshiba HD DVD players and other equipment.

Michael Schaller, the Director of Compliance for HDMI Licensing, LLC, has been in contact via email to better understand my system configuration. And more recently, Toshiba’s HD DVD partner, Microsoft, connected me with Radiient’s CEO, President and founder, Jano Banks. Microsoft’s Kevin Collins identifies Banks as ‘the co-inventor of HDMI’, and Banks claims to have influence with Toshiba.

First off, here’s a brief summary of my quandry.

• The Samsung display has a HDCP-cognizant DVI input, thereby requiring a HDMI-to-DVI translation cable.
• The Toshiba HD-A1 (running latest firmware v2.3) works fine when direct-connected to the display via such a cable, as does a Sony PlayStation 3 (also firmware-current).
• The PS3 also works fine when the Belkin HDMI switch is in-between the console and display.
• The HD-A1, conversely, refuses to proceed with the Belkin switch ‘in the picture’, citing a HDMI ERROR (which may reflect either a HDCP ‘handshake’ problem or, as Collins suggested, an inability to parse the display’s EDID).
• The HD-A1’s problems also interrupt playback of other switch-connected gear, regardless of which video peripheral is switch-selected at the time.

I shipped the Belkin switch to Analog Devices on the 11^th. Although I haven’t yet gotten a formal report from the company, initial feedback I received a week ago from compliance engineer Chris Vaughn indicates no problems discovered with the Belkin switch so far.

ADI’s exoneration of Belkin jives with Belkin’s own findings. Below are the contents of an email I received from company PR representative Jackie Romulo last Thursday:

Here is what my team found with their tests:

We tested the Toshiba HD-XA1 (very similar to Toshiba HD-A1) HD-DVD player with Belkin 3×1 HDMI switch and a Flat Panel with DVI input. It does not work.

We tested the Toshiba HD-XA1 with Belkin 3×1 HDMI switch and a Flat Panel with HDMI input. It works.

We tested the Toshiba HD-XA2 HD-DVD player (more recent model) with Belkin 3×1 HDMI switch and a Flat Panel with DVI input. It works.

We tested the Toshiba HD-XA2 HD-DVD player (more recent model) with Belkin 3×1 HDMI switch and a Flat Panel with HDMI input. It works.

We tested the Sony Playstation 3 with Belkin 3×1 HDMI switch and a Flat Panel with DVI input. It works.

We tested the Sony Playstation 3 with Belkin 3×1 HDMI switch and a Flat Panel with HDMI input. It works.

We also conducted the same exact testing with a Terk 3×1 HDMI switch and faced the same issues.

Conclusion:

The Toshiba HD-XA1 HD-DVD player demonstrates a poor implementation of HDMI and HDCP; as such, it does not work with an HDMI switch connected to a flat panel with a DVI input. It also stop playing a movie every time you switch inputs, with any HDMI switch on the market. The newer Toshiba HD-XA2 HD-DVD player demonstrates a much better handling of the HDMI/HDCP protocols.

I’ve learned a few key lessons so far from this exercise:

• Don’t automatically assume that the product which at first glance seems to be the Achilles Heel really is. In this case, I had admittedly been leaning towards fingering Belkin as the culprit, since everything worked fine when the switch wasn’t in-line. But, as my successful PS3-plus-Belkin experiment had suggested, the HD-A1 was also a potential weak link, and in fact it now appears to be the root problem source.
• I’d always believed that, at a base level, the differences between DVI-plus-HDCP and HDMI-plus-HDCP were only physical (i.e. connector-level). Granted, HDMI supports many incremental features beyond DVI, such as multiplexed audio-plus-video and optional component video transfer (versus RBG-only in the DVI case). But I’d always thought that a HDMI video source in conjunction with a connector translation cable source would ‘down-throttle’ and communicate with any DVI video destination, as long as (and only if necessary given the content requirements) they both ’spoke’ HDCP. This experiment shows that my prior understanding was incomplete.
• Sometimes the simplest solution is the best one. Fortunately, the DVI connector on the back of the plasma display is easily accessible. I’ve ordered a second HDMI-to-DVI adapter cable, and for now I’ll just swap out cable connections at the display whenever I want to transition between using the HD-A1 and the PS3. Longer term, I hope that a HD-A1 firmware upgrade will improve its ability to work with DVI/HDCP destinations and switch intermediaries. If not, there’s also always HDfury, in combination with the Samsung plasma display’s VGA input…