Goodbye HDMI hello HDbase T and hello 4k monitoring for Epic?

[Michael Brennan]

LG, Samsung and Sony have agreed a new AV standard called HDbaseT that will supercede HDMI.
Panasonic hasnt signed up yet.

This new standard for consmer tv and audio kit is designed to accomidate 10Gb/s uisng ethernet connectors and cable.
The standard has been created to allow daisy chaining between internet, tv, recorders and audio equipment. (one cable linking each piece of kit)
The standard is aimed at delivering everything electronic on the planet down one cable, including 2 x full 1080 (3d) and according to the specs, 4k!

There is no question TV manufacturers are gunning for 4k sets up to 150 cm in screen size becasue they do not need to significantly change or re-tool the existing substrate production line.

Up to 100 meter runs at 10Gb/s is this the way to go for monitoring for Epic?

Apparently hardware is about to be included in top end sets currently on the production line and will be adopted on all models in early 2011.

What are RED's thoughts on delivering a 4k monitoring output to HDbaseT?


Mike Brennan

 

[Ryan Sims]

This is really some of the best news I have heard in a long time. I have to deal with RGBHV, DVI and HDMI extenders over cat5 on every job now. Nothing but problems. This will solve so many problems in the A/V and digital signage industries.

 

[Stuart English]

HDbaseT

HDbaseT

As a proposed alternative to HDMI 1.4 - in the sense that both can carry 1080p 3D, 4K 24 fps and Ethernet control signals - four key questions for any manufacturer would be:

- cost (in parts and license fees)

- copy protection scheme

- availability of chip sets / cores

- adoption. i.e. will Panasonic and others adopt it or stay with HDMI.

 

[tillHavis]

Can anybody answer why Light Peak would not do exactly what HDbaseT is proposing and more ? It can carry bus power. Ok maybe not enough to run AV equipment, but apart from this why another connection standard ?

One thing is clear however; we are on the cusp of a new standards war. Intel is dragging its feet (on purpose) by not supporting USB3. My guess is they are buying time until Light Peak is finalized and tested. The first Intel Light Peak motherboards will prob be seen by the end of the year or early 2011 and Apple, Intel, Sony are on board so again why not go with Light Peak ?

Light Peak can carry power and be used with all current standard ports including Hdmi and USB so in theory it will be backward compatible and has higher data transfer rates.
I understand it will be more expensive on launch but this will drop over time like all new standards. Also it can carry information further as it is an optical standard, and can be made as rugged and heavy or as thin and light as demanded.
Is their something I'm missing or is this just a competition, control and cost thing ?


My dream for the future would be a 30" RED 3-4K monitor with existing sdi connectors plus Light Peak with at least 10 bit 4:4:4 color space. This would make it backward compatible and cutting edge at the same time. Maybe its not a dream. How about it RED ???

 

[Dane Brehm]

I like the forethought of moving into a single link 10gig-e protocol but this is at least 3-5yrs away. HDMI 1.4 will be 1.5 and achieve the same speed.

I've been using 1gige for Phantom since 2006 and was thankful when 10gige Cinestations came out but I have yet to see a rental or post house set up a 10gige Desktop for Rental. I'm curious to see what SMPTE thinks of it.

Still waiting for a affordable 10gige Laptop first

 

[Wayne Morellini]

SHD at 50 fps at 16 bits per pixel in 4:2:2 = 12.8mb/s
3D double. Is not professional, a step backwards from HDMI. If they could have cheaply produced higher speeds that would be different. Is this an attempt to avoid the American licensed HDMI standard? I'd rather they used next generation USB3+ (designed to eventually go to 100mb/s) than this, at least that would greatly simplify interoperability and interfacing on PC's/Audiovisual hardware (and potentially lead to a simplified sub-standard for hard drive use).

Dare we mention UHD 3D.

 

[Stephen Gentle]

Can anybody answer why Light Peak would not do exactly what HDbaseT is proposing and more ? It can carry bus power. Ok maybe not enough to run AV equipment, but apart from this why another connection standard ?

Theoretically you could eventually be able to carry Ethernet, DisplayPort and this HDBaseT format all at the same time on a Light Peak cable.

 

[Wayne Morellini]

When they announced USB 3.0 they announced that it was designed to carry an optical connector up to 100gb/s. I have just been reading some comments from them and yes, light peak could be plugged into a usb socket.

http://www.wired.com/gadgetlab/2010/04/intels-light-peak-technology-could-kill-usb-30/

They have been working on this for sometime before USB3 was announced. Why, I think they intend to make a inter chip/inter or intra board optical bus too. They made an advancement a few years back in the first viably practical silicon lasers (using an obvious technique, amazing it was missed for so long), which makes this very possible. I don't know if anybody else has practical IP to compete with this (which mans a potential licensing bonanza). If they have cracked on chip processing it is an whole new ball game. So maybe there was shift between announcing USB3 and announcing this.

At least it looks like they are listening, making it protocol independent. I have not read the USB3 spec yet, hopefully it has the features I was pushing for that, including simplified stratified modes on top of a protocol independent (+channels) mode allowing simple straight point to point links, to allow simpler 'targeted' device controllers and roil your own inter device protocols. If they can make this simpler and cheaper then 100gb ethernet, then bring it on, I'd much prefer this to 10gbe, or HDMI. I have been waiting for this sort of bus for nearly twenty years. The virtual hardware for my OS design was designed around a optical bus concept like light peak with innovative multiple device capability.

If they play their cards right, you would be able to send multiple streams at different speeds, allowing slower devices like mice and keyboards along side UHD graphics streams without having to share and interpolate with the UHD hardware, allowing their controllers to be made slower, cheaper and simpler. If they could only archive this level.

This sort of stuff is like gravy and candy all rolled up in one, it can take a good cook to make it work.

 

[Kevin Olsen]

I think LightPeak is really interesting and very handy, but I think the cost of twisted pair cable and the ease of termination make HDbaseT really attractive. This is still a consumer driven market so anything that makes things easier to install and run wire for in a home is a huge advantage.

What is LightPeak going to cost per foot? Can it be terminated in the field? Is it stable in adverse climates (attics, crawlspaces, etc...)?

HDMI is really cool, but try running an HDMI cable in the wall to a projector in a home that's already built. These kinds of problems prevent lots of people from getting the products they want, in the locations with the aesthetic they desire. There are twisted pair baluns but they can become cost prohibitive.

I am all in favor of this movement. How come trillions of dollars travel securely over Cat5e every day and we need a stupid cable three times the thickness with a crappy connector that must be factory terminated (HDMI) just to get 1080p to our tvs?

KO

 

[Wayne Morellini]

I don't know any details but from their comments, it is very fine, flexible (but how much of a curve can FO use) and probably cheap. Because they can do on chip silicon laser, the cost of chips could, if not at the start, be very cheap. Originally when they talked about this, I was worried about all these issues you mentioned, and considering the 20mb/s link speeds achieved in lab electrically, but they seem to be talking the right talk. The cost of the last end of optical has been coming down for decades, they used to talk about the last mile, then the last 30 feet or so, now they can talk about cracking the last inch. I don't see this HDbaseT as valid unless they are looking at moving to the 100gb/s Ethernet standard and that can be done as cheap or cheaper then light peak.

If I were them I would compare Ethernet to Light peak in performance and price to the consumer, and if it was a clear winner, drop Ethernet at once and get a binding 'Open' agreement with Intel as soon as practical.

 

[Wayne Morellini]

This really puts plans to buy a Macintosh when they get USB3 (and preferably UWB ha, ha ) DX11 level GPU, multi-touch, and any of SHD, 3D, OLED, on hold. I might be waiting till they get a Terascale or Larrabee version) if I have to wait for light peak. Well, at least I might be able to get to play around with a 144 embedded core prototyping board I have been waiting (non-intel).