HDMI: On Hiccups You Can Rely

To date, I’ve been routing my Xbox 360 and VUDU set-top box to the single HDMI input of the Hanns-G 28" LCD in my bedroom via a slick HDMI link-powered two-input switch made by Belkin. The Xbox 360 and VUDU both had dedicated optical S/PDIF inputs into my Pioneer sound system. When I recently came into possession of Roku’s Neflix Player, I temporarily tested it by swapping out the HDMI cable from the VUDU back panel. For sound, I employed the wimpy speakers built into the LCD, fed by the Netflix Player’s multiplexed A/V HDMI connection.

Today, I decided to formalize the Roku unit’s presence in my bedroom’s entertainment stack, and while I was mucking around with cabling, twist ties and power strips, I also resolved to move the PlayStation 3 upstairs from its prior location in the living room. This meant I needed to mate one optical S/PDIF input in the Pioneer sound system with three systems’ optical S/PDIF outputs:

• PS3
• Roku Netflix Player
• Roku SoundBridge

which a mechanical switch neatly solved.

It also meant that the LCD’s single HDMI input now needed to interface with four systems’ HDMI outputs:

• PS3
• Roku Netflix Player
• VUDU
• Xbox 360

I have several four-input HDMI switches on hand; I decided to try Radiient’s Select-4ce first, since it claimed not only HDMI compliance (all the way to 1080p) but also the Simplay Labs seal of approval. The Roku Netflix Player (at 480p on Select-4ce input port 2), VUDU (at 1080i on input port 3) and Xbox 360 (1080p on input port 4) all worked without a hitch.

And what of the PS3, on input port 1? Alas, that was a different matter. At its default 1080p output setting as well as at 1080i, the display would blank out at random occasions, for random durations. And on occasions when the image was stable, it was still marred by random noise. Switching to high-end cabling and/or shorter cabling reduced, but didn’t completely eliminate, the problems (leading me to conclude that there apparently is some difference between HDMI cables at different price points). But if I direct-connected the PS3 to the LCD, everything worked fine all the way up to 1080p. And if I ran the PS3 through the Radiient switch but throttled the console’s resolution down to 720p, the setup was stable, too.

I emailed Radiient’s President and CEO, Jano Banks, with my findings. His initial response was intriguing, to say the least:

We have had no reports of Playstation 3 failures with the Select-4ce, but we may not have a lot of people running v3.96 firmware. The Select-4ce’s inputs do have slightly different performance, all meeting specs to 1080p, but because they are slightly different, we have had some cabling/device setups work better or worse on different inputs. Have you tried putting the PS3 into input #2, 3, or 4? Input 1 is the most different (2, 3, and 4 are very close to each other).

I didn’t try port 4 per his suggestion, but I did try ports 2 and 3. And within my reply to Jano was the result:

Well, whaddya know…;-) Both inputs 2 and 3 worked fine with the PS3. I’ve now got Roku on 1 [editor note: unlike the PS3, it works fine here], PS3 on 2, VUDU on 3, and the Xbox 360 on 4. Can you give me some more technical background on what I ran into here? Impedence differences between the various inputs?

Here’s what Jano had to say in response:

Glad that solved your problem. Since you said that the problem occurred even at 1080i from the PS3 on Input 1, I can only guess that the issue may be impedance issues on the DDC [editor note: display data channel] lines, but that’s just a guess. Also could be Hot Plug/+5V signal levels. If the problem had only occurred at 1080p on Input 1 with the PS3, then it would point to TMDS - the high-speed communications lines. The Select-4ce is fully HDMI compliant, and the PS3 should be, too, but DDC is a gnarly bus, and cabling can really cause funny things to happen, which can screw up communication of the EDID (capabilities of the device) and, more importantly, HDCP handshaking. Perhaps the impedance and reflections of the DDC bus when going through our Input 1 is just slightly different than the other inputs, and that’s enough, combined with the cabling and PS3, to fail. If all components are HDMI compliant, it should work perfectly. Without testing the devices ourselves in our lab, I can’t say for sure what the ultimate culprit was.

I had thought that by evolving my setup beyond an older DVI/HDCP-input display (and corresponding HDMI-to-DVI adapter) to a newer, HDMI/HDCP-native LCD, I might free myself from a continuance of my past interface problems. Today’s lesson unfortunately teaches me not to be so naive, eh?