Goodbye HDMI

[Malibu Rapper]

Hello HDbaseT. A bit old news for some if you're following the industry. The good news is I already have a TON of cables for the next standard. I just started wiring up the house with some more keystone jacks thanks to Monoprice.com. I love the comment "it'll only be a matter of time before Monster finds a way to throw gold or other precious metals into a Cat 6 cable and release it as a "premium" HDBaseT cable." It's about time!

Story from here.

New standard offers both lower costs, the potential to deliver more information and convenience

LG Electronics, Samsung Electronics, Sony Pictures Entertainment and Valens Semiconductor have been secretly conspiring to kill HDMI. And today they set their plans into motion, introducing a brand new audiovisual standard, HDBaseT. While that name may sound confusing, before you slap your head, this will not likely introduce a new kind of cable to your house.

Past A/V standards from industry groups have introduced a plethora of connectors (S-Video, HDMI, DisplayPort for example), so that makes this release all the more unusual. Instead of a new connector, it is based on the Cat 5e/6 network cables, commonly referred to as "ethernet cables".

It supports cable lengths up to 328 feet. The cable can pass HD and 3-D video signals, as well as data through an integrated 100MBit Ethernet connection. That data feed should allow for new internet-connected TV services, such as Google TV which delivers advertising-funded services to TV sets.

The standard also has many other advantages. For one, it will help declutter the growing mess of cables in the average household. By repurposing ethernet cables, it should also dramatically lower costs, both for the manufacturer and the consumer.

About the only loser in the situation may be "premium" cable makers like Monster Cable Products. However, it'll only be a matter of time before Monster finds a way to throw gold or other precious metals into a Cat 6 cable and release it as a "premium" HDBaseT cable.

About the only loose end is what mini-connectors will arise out of this new standard. Currently many smart phones, such as the HTC EVO 4G offer mini-HDMI connections. Perhaps more advanced mini-USB connections will answer that dilemma, though.

The standard will begin its rollout later this year. The majority of its volume will hit in 2011.

 

[AndyH]

I am thinking about buying a new 3D TV. Should I wait?

 

[Malibu Rapper]

I am thinking about buying a new 3D TV. Should I wait?

I'll let the #1 Laker fan Dave answer that one...

 

[TURBO2GO]

LOL.... Just wait til next year bradley. Just signed Ray Allen back, the coach decided to come back, everyone took a paycut to keep the team together. Looks like Jermaine Oneil may be coming here, as may shaq.


Well I personally despise HDMI. In fact to us in the custom installation industry, it is almost an unusable standard full of problems. Starting with the connector itself that tends to fall off and come out too easily to its transmission, to HDCP issues, it is nothing but a nightmare. It's also very low in length of transmission cables allowed (about 50 feet) and using adaptors is never reliable. I wish the whole thing would go away. I love the idea of HDbaseT, but it is competing with other formats. AndyH I hate to tell you to wait, because you may be waiting for nothing. This may not make it at all. Even if it did, I am not sure it would concern you unless you are planning on distributing HD content throughout a house. Then it too may have problems, and there may be several itterations of the standard. No, I would not wait if it was me.

 

[Ben]

Through my past and current work, I've worked closely with HDMI since almost the beginning, including on a joint project with Silicon Image, and with Intel on the HDCP front.

In my opinion, practically speaking, at a consumer pricepoint, HDMI reached it's technical limits a few years ago. Subsequent revisions drive up the cost to get full functionality, and don't fix some of the root issues with an essentially outdated means of transmitting high speed/bandwidth data. HDMI was originally designed for 4Gbps transmission, (for a 1080p display at 60Hz) through a relatively short cable, which was fine and good for that purpose. Version 1.3 increased that data rate to 10Gbps, which was good, but what they don't tell you is that to get the higher data rate or to use long cables, expensive, heavy, thick cables are needed, which are priced more than the average consumer is willing to pay. For full throughput longer length cables, cables can be almost 1/2" thick, weigh several pounds, and cost $50 or more. Sure, you can always buy a cheaper/thinner/lighter one, but there's no guarantee it will work 100% if you have a high bandwidth application. Some of the eye diagrams using inexpensive and/or long cables don't look very pretty. If you don't keep the cable length relatively short (1-3 meters), buy an expensive cable, or use expensive active repeaters, it's quite likely that you will experience problems at the higher data rates. For even higher speed iterations in the future, it's likely the standard will move to a dual-link setup. The standard itself is also difficult to work with and implement properly with HDCP, such that, twice a year, manufacturers of HDMI enable equipment have to get together in one place for a full week to test interoperability with each others devices using a plethora of different cables. Suffice to say, no device is always 100% interoperable with all the others, but it's a good testing ground to see what you need to fix.

Coupled with the fact that Silicon Image has a monopoly on licensing, high license fees, poor technical support, and are difficult to work with in general, it's no surprise that the industry is starting to define new standards and push other means of transmitting high bandwidth data.

That said, viable successors to HDMI are HDBase-T and optical links, such as Intel's upcoming Light Peak. Another format war could be coming. In my opinion, optical is the superior technology, because it's much more scalable than an electrical transmission system. HDBase-T is attractive because it uses the ubiquitous low cost Ethernet cable. But, how scalable and how long of a cable you can use at full bandwidth with HDBase-T remains to be seen. I think the initial quoted length was 100m? Optical links for consumer applications will initially be 10G, scalable to 100G, with cable lengths of up to 500 meters. Optical cables are also much thinner and lighter than Ethernet for comparable data rates, and have no EMI/RFI issues. Both are better than HDMI.

 

[flaminio]

I love the comment "it'll only be a matter of time before Monster finds a way to throw gold or other precious metals into a Cat 6 cable and release it as a "premium" HDBaseT cable."

They'll be playing catchup to Denon and their $500 ethernet cable.

 

[joboy77]

They'll be playing catchup to Denon and their $500 ethernet cable.

Just read some of the Amazon reviews for that cable. Freaking hilarious.

 

[clr1024]

Just read some of the Amazon reviews for that cable. Freaking hilarious.

I was reading some of them too. This one cracked me up

364 of 366 people found the following review helpful:
2.0 out of 5 stars Great cable, but too fast., June 23, 2008
By Matthew Sidor "seadour" (Boulder, CO USA) - See all my reviews
(REAL NAME)

This review is from: Denon AKDL1 Dedicated Link Cable (Electronics)
Transmission of music data at rates faster than the speed of light seemed convenient, until I realized I was hearing the music before I actually wanted to play it. Apparently Denon forgot how accustomed most of us are to unidirectional time and the general laws of physics. I tried to get used to this effect but hearing songs play before I even realized I was in the mood for them just really screwed up my preconceptions of choice and free will. I'm still having a major existential hangover.

Would not purchase again.

 

[Midnight_Raven]

Through my past and current work, I've worked closely with HDMI since almost the beginning, including on a joint project with Silicon Image, and with Intel on the HDCP front.

In my opinion, practically speaking, at a consumer pricepoint, HDMI reached it's technical limits a few years ago. Subsequent revisions drive up the cost to get full functionality, and don't fix some of the root issues with an essentially outdated means of transmitting high speed/bandwidth data. HDMI was originally designed for 4Gbps transmission, (for a 1080p display at 60Hz) through a relatively short cable, which was fine and good for that purpose. Version 1.3 increased that data rate to 10Gbps, which was good, but what they don't tell you is that to get the higher data rate or to use long cables, expensive, heavy, thick cables are needed, which are priced more than the average consumer is willing to pay. For full throughput longer length cables, cables can be almost 1/2" thick, weigh several pounds, and cost $50 or more. Sure, you can always buy a cheaper/thinner/lighter one, but there's no guarantee it will work 100% if you have a high bandwidth application. Some of the eye diagrams using inexpensive and/or long cables don't look very pretty. If you don't keep the cable length relatively short (1-3 meters), buy an expensive cable, or use expensive active repeaters, it's quite likely that you will experience problems at the higher data rates. For even higher speed iterations in the future, it's likely the standard will move to a dual-link setup. The standard itself is also difficult to work with and implement properly with HDCP, such that, twice a year, manufacturers of HDMI enable equipment have to get together in one place for a full week to test interoperability with each others devices using a plethora of different cables. Suffice to say, no device is always 100% interoperable with all the others, but it's a good testing ground to see what you need to fix.

Coupled with the fact that Silicon Image has a monopoly on licensing, high license fees, poor technical support, and are difficult to work with in general, it's no surprise that the industry is starting to define new standards and push other means of transmitting high bandwidth data.

That said, viable successors to HDMI are HDBase-T and optical links, such as Intel's upcoming Light Peak. Another format war could be coming. In my opinion, optical is the superior technology, because it's much more scalable than an electrical transmission system. HDBase-T is attractive because it uses the ubiquitous low cost Ethernet cable. But, how scalable and how long of a cable you can use at full bandwidth with HDBase-T remains to be seen. I think the initial quoted length was 100m? Optical links for consumer applications will initially be 10G, scalable to 100G, with cable lengths of up to 500 meters. Optical cables are also much thinner and lighter than Ethernet for comparable data rates, and have no EMI/RFI issues. Both are better than HDMI.

Only thing I would like to add into this is that although optical cable may seem better due to it's longer length and bandwidth they are also very fragile. We work with Fiber Optics all day and those are the most troublesome because you have to be very delicate with them. Any kink makes it a useless cable with no method to repair while a ethernet cables is definitely more robust and can handle a beating relatively speaking.

 

[HIJACKER]

I am thinking about buying a new 3D TV. Should I wait?

I'm waiting until you don't need to wear the stupid glasses and it's integrated into the TV itself.

 

[NSXrebel]

I'm waiting until you don't need to wear the stupid glasses and it's integrated into the TV itself.

Is that even possible?

 

[Ben]

Only thing I would like to add into this is that although optical cable may seem better due to it's longer length and bandwidth they are also very fragile. We work with Fiber Optics all day and those are the most troublesome because you have to be very delicate with them. Any kink makes it a useless cable with no method to repair while a ethernet cables is definitely more robust and can handle a beating relatively speaking.

I think optical cables are getting better, and for consumer use, it's less of an issue than for enterprise/datacenter/telecommunications use. Optical TOSLink cables using plastic/polymer fibers have been around for a long time, and proven themselves for use in a consumer/prosumer setting. Optical links for something like HDMI or home networking should have the same or better reliability - good enough for home use. Newer designs even allow smaller radius bends using newer "bendable" fibers. Using ethernet cable is definitely a huge plus, but it will be much harder to push it to higher bandwidth or longer cable length without signal degradation as bandwidth requirements go up as time goes on. That's why we don't see any claims of HDBaseT being able to scale up significantly as bandwidth requirements increase over the next few years.

 

[TURBO2GO]

I think optical cables are getting better, and for consumer use, it's less of an issue than for enterprise/datacenter/telecommunications use. Optical TOSLink cables using plastic/polymer fibers have been around for a long time, and proven themselves for use in a consumer/prosumer setting. Optical links for something like HDMI or home networking should have the same or better reliability - good enough for home use. Newer designs even allow smaller radius bends using newer "bendable" fibers. Using ethernet cable is definitely a huge plus, but it will be much harder to push it to higher bandwidth or longer cable length without signal degradation as bandwidth requirements go up as time goes on. That's why we don't see any claims of HDBaseT being able to scale up significantly as bandwidth requirements increase over the next few years.

Very true but in my work, we need to send signals across a house, sometimes these homes are 15-20,000 sqaure feet. Toslink cables are not an option. Fiber is extremely hard and time comsuming tro terminate. We've done homes with fiber at the customer's request, but I have yet to see one of those fiber lines be put to actual use. HDMI is a pain because its length is limited (and it sucks in many other ways). The appeal of using Cat5e is great because it is cheap, tough, easy to terminate. If you need more bandwidth I don't know why you couldn't run two. Or three. Or four. I much prefer to pull 4-5 runs of cat5 than to mess with fiber in a residential/business commercial setting. Even Cat6 sucks to pull. Ity won't bend and if you bend it you kill it. IMO until they address these other issues, cat5 will always be the winner. The problem with all this is there are always too many guys sitting in a lab or a boardroom making decisions and none of them actually RUN the stuff in any sort of real setting.

 

[ashfixit]

HDMI good riddance - bad rubbish

HDMI is terrifying and a technically flawed creature. Think of ISDN; the technolgy (I Still Don't Need). With piles of good/fair/bad transports available; why pick this terrible thing.

- DVI-D - what was wrong with that?
- SDI - that's what the pro cameras/mixers use!
- displayport
- standard cat5e will support the data rates with 350mhz per pair;
- 10Gb Ethernet on CX4 cabling - or cheap multimode fiber
- Fibre Channel 4gb/10gb
- STM for multiplexed synchronus support (If the 100ns latency of ethernet framing is too much to bear)
- DWDM - it works
-lightpeak is cool, but unObtanium

HDMI was rushed in to create an intentionally broken transport for the consumer market to favor 'stakeholders rights', and sell landfill real estate.
- HDCP is unspeakably bad for consumer rights (ever see an HDMI recorder?)
- ridiculous mechanical connectors (4 of them!!!1)
- backwards compatibility record abysmal so far
- canned formats obsolete (480p, + 32khz audio, really?)
- color space conversions are implemented wrong.
- TMDS is a laughably narrow channel descrption for the TDM buss
- differential electrical spec is entirely too weak to drive any reactive cables.
- i2c, spi, ethernet, power and tdm on same cable?? really??? reeks of personality disorder
- sillicon image - unclean, need shower

 

[Ben]

Very true but in my work, we need to send signals across a house, sometimes these homes are 15-20,000 sqaure feet. Toslink cables are not an option. Fiber is extremely hard and time comsuming tro terminate. We've done homes with fiber at the customer's request, but I have yet to see one of those fiber lines be put to actual use. HDMI is a pain because its length is limited (and it sucks in many other ways). The appeal of using Cat5e is great because it is cheap, tough, easy to terminate. If you need more bandwidth I don't know why you couldn't run two. Or three. Or four. I much prefer to pull 4-5 runs of cat5 than to mess with fiber in a residential/business commercial setting. Even Cat6 sucks to pull. Ity won't bend and if you bend it you kill it. IMO until they address these other issues, cat5 will always be the winner. The problem with all this is there are always too many guys sitting in a lab or a boardroom making decisions and none of them actually RUN the stuff in any sort of real setting.

You're right - I'm the guy in the lab and boardroom.

I actually just attended a Silicon Image HDMI 1.4a update seminar today - they are pressing forward with this, but it looks like they are also looking to the future, and optical is one of the things they are studying for the next generation. They are well aware that they are hitting the limit with the current implementation.

I would say your customers are a subset with special needs that most people don't have. My statements were aimed at the average consumer, most of who don't live in 15000 square feet homes. I was talking about off-the-shelf cabling that the customer can buy cheaply, in precut lengths (1.5, 3, 6, 10, 20, 50 meters) and plug and play, not a custom pull for a killer system in a mansion. Such cable could be used for a source to TV link, or, a next generation link from PC to external storage, etc. The possibilities are many. HDMI currently can't do this easily or cheaply at any length above 3-6 meters.

 

[Ben]

Re: HDMI good riddance - bad rubbish

-lightpeak is cool, but unObtanium

It's on the verge. Once the standard is fully defined, I think it will be full speed ahead with Light Peak. Intel is effectively trying to kill USB, and if Intel has their way, USB 3.0 will be dead out of the gate.

And, once it is introduced, it will be low cost enough to be a consumer standard.

 

[TURBO2GO]

Ben although I work in those homes, I and many custom installers also work on smaller projects, normal homes, normal sizes, etc. A lot more people than you think want to distribute video or not have a stack of equipment in their family room. A basic home theater install often requires the equipment go elsewhere, in a basement, a closet, whatever. All these are problems and the custom installation industry is HUGE, much bigger than you think. When Best Buy has a custom install division, there is a reason for it. We are all screwed with HDMI which is extremely unstable, and now they are talking about killing the component outputs of all video devices outputing anything other than 480. No HD. All HD will have to run ONLY through HDMI. If you knew the number of service calls I get... because HDCP didn't sync and now there is no video.... it is so frustrating. It is almost an unusable format. Fiber is just not at a place where it can be a solution for residential. I am not surprised that the 3 manufacturers behind HDbaseT are TV companies.

 

[ilya]

Another point to consider. Assuming HDMI does bite the dust in the future, and a more stable, reliable signal carrier becomes the standard (HDBaseT or otherwise), will it still eliminate all of the HDCP handshake issues that exist today.

The handshake issues due to cable integrity/speed may be eliminated, but what about all the HDCP sync issues due to incompatiblity or poor communication across different brands and devices. The cable will improve, but all the copy protection and security schemes will remain.

I'm curious to see if a new, more stable security standard will emerge or if the same handshaking/sync will continue on the new platform.

 

[TURBO2GO]

Another point to consider. Assuming HDMI does bite the dust in the future, and a more stable, reliable signal carrier becomes the standard (HDBaseT or otherwise), will it still eliminate all of the HDCP handshake issues that exist today.

The handshake issues due to cable integrity/speed may be eliminated, but what about all the HDCP sync issues due to incompatiblity or poor communication across different brands and devices. The cable will improve, but all the copy protection and security schemes will remain.

I'm curious to see if a new, more stable security standard will emerge or if the same handshaking/sync will continue on the new platform.

Please don't remind me. This subject ruins my day. HDCP belongs in the toilet.

 

[Malibu Rapper]

Ben, so is displayport's future the same as HDMI?

 

[Ben]

Ben, so is displayport's future the same as HDMI?

It's anyone's guess at this point. Displayport really never gained much traction outside the PC space.

And, you can bet that Silicon Image and Intel aren't going to let HDMI and HDCP go without a fight. It's too much of a cash cow right now. Without HDMI, there is no Silicon Image. HDCP, is Intel's, and thus backed by all the money and resources of Intel. Whatever new standard emerges, you can bet Intel will try to get HDCP or some form of Intel controlled security inside.

Since Silicon Image and Intel have a monopoly on HDMI and HDCP, they are both very difficult to deal with, and both have very poor technical support. There isn't even a phone number you can call for HDCP support, only email and fax, which, of course, takes several days to get returned. And, when you ask for a representative's phone number, they say they don't have phones. Yeah, right.

 

[Ben]

Ben although I work in those homes, I and many custom installers also work on smaller projects, normal homes, normal sizes, etc. A lot more people than you think want to distribute video or not have a stack of equipment in their family room. A basic home theater install often requires the equipment go elsewhere, in a basement, a closet, whatever. All these are problems and the custom installation industry is HUGE, much bigger than you think. When Best Buy has a custom install division, there is a reason for it. We are all screwed with HDMI which is extremely unstable, and now they are talking about killing the component outputs of all video devices outputing anything other than 480. No HD. All HD will have to run ONLY through HDMI. If you knew the number of service calls I get... because HDCP didn't sync and now there is no video.... it is so frustrating. It is almost an unusable format. Fiber is just not at a place where it can be a solution for residential. I am not surprised that the 3 manufacturers behind HDbaseT are TV companies.

It will be interesting to see what happens. I had a meeting with Samsung a few days before the announcement of HDBase-T. They dropped some hints about HDBase-T, but didn't provide much detail. A day earlier, I also met with 2 groups within LG, and they said nothing about HDBase-T. From what I gathered from the Samsung meeting, the lowest cost solution will be the one that will ultimately win. Samsung is about cost, cost, cost. Regarding optical, you are right, as of today, it's not there yet. But, you will see some huge improvements/breakthroughs/announcements in the next year or 2, I think. For say 3000 or even 5000 square foot "normal" homes, 100 meter cables should be sufficient, so you can use off the precut shelf cabling without having to cut/terminate.

 

[Ben]

Another point to consider. Assuming HDMI does bite the dust in the future, and a more stable, reliable signal carrier becomes the standard (HDBaseT or otherwise), will it still eliminate all of the HDCP handshake issues that exist today.

The handshake issues due to cable integrity/speed may be eliminated, but what about all the HDCP sync issues due to incompatiblity or poor communication across different brands and devices. The cable will improve, but all the copy protection and security schemes will remain.

I'm curious to see if a new, more stable security standard will emerge or if the same handshaking/sync will continue on the new platform.

As I referenced in a previous post, twice a year, all manufactures of HDMI enabled devices spend a week together in one place to hash out all the incompatibilties, HDCP sync, and other issues. I remember back in 2005, it was as 3 day thing, and now it's grown to a full week. There are always dozens if not hundreds of issues with each device, then the 2 manufacturers have to hash out whether the problem is with one device, or the other, etc. It's a long convoluted process, and it's absolutely impossible to fix everything. That is what you see on the market with HDMI/HDCP problems.

As I referenced in another previous post, whatever new standard comes out in the future, you can bet Intel won't give up HDCP or some HDCP successor without a fight, and it's such a cash cow, you can bet they will fight.

 

[TURBO2GO]

For say 3000 or even 5000 square foot "normal" homes, 100 meter cables should be sufficient, so you can use off the precut shelf cabling without having to cut/terminate.

Don't forget that first of all any sort of termination will probably not work as we have to run this wire through holes in studs, sheetrock, ceiling joists, conduit, etc. A fat end is always a problem not to mention it just gets destroyed.

We pull lots of HDMI through walls and ceilings, the connector is always too fat, and it gets damaged very easily.

Third thing is, none of the HDMI wires I have seen or that most installers use is actually in-wall rated wire. It does not have a CL rating. It is technically ILLEGAL to be run behind walls and ceilings. The jacket does not meet code.

If I was to pull a pre-terminated fiber wire, it has all these issues, plus the fact that the fiber will just not survive going through a hole in a stud then taking a bend.

Last thing is, pre-made wire by its nature is nothing but a pain in the ass to us, because it is never the correct length. I am dealing with bundles of wire that are a foot thick. On large projects we can fill a room with wire. Imagine if none of this wire was the correct length and we had to coil up excess. Imagine the spagetti. I can't remember the last time we used ANY pre-made cables. We custom make all wiring to proper length. Sometimes we need to terminate at a junction box. A wall plate, etc. It is CRITICAL that we be able to determine exact length. If you handed me a bunch of 100M pre-made optical cables at a new construction site (or worse in a retro-fit install) and said "run these", you would not get a friendly look from me. You'd get this: and not this: :smile: Maybe you'd get this:

Cat5e solves all these issues I have now. HDCP free HDbaseT all the way baby!!!

 

[Arshad]

I'm in the industry too. HDCP isn't going anywhere, regardless of what type of wire the signal is on. We should all be glad it's not something even more cumbersome... everything considered, it's pretty lax.