HDMI 1.4 and Through the Ages

[admin]

Key Digital had an interesting newsletter that got sent out this week. In it they had a table which went through the various iterations of HDMI, from version 1.0 through the newest 1.4. We thought it was a good chart, though we disagreed (slightly) with some of their assessments on the importance of each of the versions. As I hinted at in my article on The Twelve Biggest Industry Mistakes of the Digital Age I believe that HDMI is really more of a stop-gap interface whereas it could have been a new standard. Instead, we have a largely "stupid" new cable system that merely carries data. HDMI 1.4 does little to change this reality.


Discuss "HDMI 1.4 and Through the Ages" here. Read the article.

[davidtwotrees]

Good article, Clint. It was interesting to note that v1.0 hdmi did everything 99% of consumers need to do in their home systems. The progression of versions in hdmi is really a snapshot of the consumer electronics industry, imho. They are making everything WAY too complicated for the end user.

[BoredSysAdmin]

I'm not a grammar Nazi, but Clint - your spellcheck missed this one:

"It is not clear if HDMI will comp7ete with DVI or Display Port in this market"

[Biggiesized]

It will take 4 times the encoded bandwidth data rate to transmit 4Kx2K/24 compare to 1080p/24. That means to transmit one 4Kx2K it will take the space of four current HDTV channels or 24 current SDTV channels.

Where is your evidence for this? Sure, the frame size is four times as large, but who says the encoded stream must be four times as large? Ben Waggoner believes 4K/24p can be done at around 60 Mb/s with the latest codecs. Let's not forget to mention that future codec design will migrate towards wavelets, which can be way more efficient than DCT.

[dgreenstein55]

"Where is your evidence for this? Sure, the frame size is four times as large, but who says the encoded stream must be four times as large? Ben Waggoner believes 4K/24p can be done at around 60 Mb/s with the latest codecs. Let's not forget to mention that future codec design will migrate towards wavelets, which can be way more efficient than DCT."


Biggie - The point of stating that 4K resolution will take 4 times the amount of data space is that using the current standards of encoding/compression MPEG4, it would take 4 times the bandwidth to display 4 times the picture! All we need is more compression schemes to further disrupt the video we currently look at, 1080p while fantastic is always compressed and for anyone who has ever seen a clip of completely uncompressed video, like myself, you will learn that uncompressed 1080p is far more breathtaking that a compressed 4K stream.

Ultimately HDMI will not have enough bandwidth to carry what will become available in the future of video. DVI has always been the better solution, for one it LOCKS and two it is Video only, and any experienced Home Integrator knows that a high end Disc player will always do a better job of audio decoding and sending it out via analog so it stands to reason that a medium to transfer data should always serve one purpose and in the case of video we need to quit cramming audio and control and so on down the pipe!

[Anagoge]

I think the description of the Audio Return Channel misses one big use case that has immediate benefit for me. Right now, I have to run two cables between my TV and receiver. One is the normal HDMI for input of DVD type signals. The other is the optical audio from the TV to the receiver so the receiver can play the audio for over-the-air (OTA/"antenna") HDTV signals. The Audio Return Channel would remove the need for this second optical cable and allow the TV's audio to travel to the receiver on the same cable that supplies the DVD input to the TV. I know many people ignore them, but OTA HD local stations are less compressed than the cable/satellite equivalents here.

[buzzy]

The industry viewed HDMI first and foremost (by far) as a vehicle to implement copy protection. That everything else (like ease of use, reliability, and capabilities, and long term migration path) was an afterthought, is a natural result of that mindset.

And this is the same kind of thinking that led to the music industry imploding. Rather than focus on customer needs, they kept trying to bend consumers over to meet their needs.

When they start thinking about consumer needs and experience first, they may be able to turn the tide in their favor. Otherwise, people have other things to do.

[KurtBJC]

Where is your evidence for this? Sure, the frame size is four times as large, but who says the encoded stream must be four times as large? Ben Waggoner believes 4K/24p can be done at around 60 Mb/s with the latest codecs. Let's not forget to mention that future codec design will migrate towards wavelets, which can be way more efficient than DCT.

HDMI isn't sent compressed. If you've got four times as many pixels to send, it takes four times as much bandwidth.

Kurt
Blue Jeans Cable

[Biggiesized]

Okay, there seems to be a misunderstanding between us.

Are you stating it takes 4 times the bandwidth to deliver OVER HDMI? Or are you stating it takes 4 times the encoded bandwidth over a cable system?

I'm arguing the latter.

[Biggiesized]

Biggie - The point of stating that 4K resolution will take 4 times the amount of data space is that using the current standards of encoding/compression MPEG4, it would take 4 times the bandwidth to display 4 times the picture!

Do you have any 4K source material? Would you like to run a test between 2160p50 and 1080p50? We can set the test parameters as follows:

1080p50 content is allowed a peak bitrate of 40 Mb/s.
2160p50 content is allowed a peak bitrate of 80 Mb/s. (that's only DOUBLE the encoded bit rate)

Are you saying that this cannot be done or that it requires a bit rate of 160 Mb/s (since 2160p is four times the frame size)?

If you're game, download the Crowd Run source files from Xiph:

http://media.xiph.org/svt/2160p50_Cg...ter_SVTdec05_/