ODEMAX in the fall?

[Jeff Kilgroe]

Rec2020 is a joke anyway. Right idea, failed final spec and an impotent execution thus far. Just like the whole HDMI 2.0 clusterfuck. I still think RED, Sony, Apple and maybe one or two others need to push ahead with some forward-thinking display products and define a new spec and standard. HDMI has become a scourge upon the industry and yet no one seems to want to fix it.

Right now, the industry is targeting HDMI 3.0 in 2020 to support "8K" (er.. 7680 actual horizontal pixels, so calling it 8K is more like a big fat lie) at up to 60Hz via dual-link with rec2020 space adapted to 8bpc color. And they are not in any hurry to get there. 4K is almost viewed as some bug up their ass that they don't want. At least as far as the HDMI standards committee is concerned. They're only tolerating it, because consumers want it at the moment and it's a somewhat necessary step to get to 8K.

IMO, there's no reason they don't have the HDMI spec already designed and established now to support 8K or 12K or whatever at up to 16bpc/48bit color and 240Hz with 64 audio channels and options for sequential or simultaneous 3D/stereo via packed frames. Throw the spec out there, CE makers will then have a better road map and the HDMI committee can go on vacation for the next 8 years or so, then come back and draft a new spec to last the next 25 years and retire. Apparently that's too much to ask.

 

[Wayne Morellini]

Rec2020 is a joke anyway. Right idea, failed final spec and an impotent execution thus far. Just like the whole HDMI 2.0 clusterfuck. I still think RED, Sony, Apple and maybe one or two others need to push ahead with some forward-thinking display products and define a new spec and standard. HDMI has become a scourge upon the industry and yet no one seems to want to fix it.

Right now, the industry is targeting HDMI 3.0 in 2020 to support "8K" (er.. 7680 actual horizontal pixels, so calling it 8K is more like a big fat lie) at up to 60Hz via dual-link with rec2020 space adapted to 8bpc color. And they are not in any hurry to get there. 4K is almost viewed as some bug up their ass that they don't want. At least as far as the HDMI standards committee is concerned. They're only tolerating it, because consumers want it at the moment and it's a somewhat necessary step to get to 8K.

IMO, there's no reason they don't have the HDMI spec already designed and established now to support 8K or 12K or whatever at up to 16bpc/48bit color and 240Hz with 64 audio channels and options for sequential or simultaneous 3D/stereo via packed frames. Throw the spec out there, CE makers will then have a better road map and the HDMI committee can go on vacation for the next 8 years or so, then come back and draft a new spec to last the next 25 years and retire. Apparently that's too much to ask.

Expecting them to go to 12k high bitrate etc now is a little unfair Jeff. The real word limitation, until we go to optical wires, is transmission speeds. We are at at least 10-20Gb/s currently, the prospect of doing beyond that non cheap cable I suspect night get thinner and thinner. Even thunder bolt is expecting to use multiple lines to get to 100gb/s on their optical cable convertor plugs (I think their might be ten pins I. the socket available).

The real solutions beyond HDMI for Red and others to go to formulate a proper pro/content media interface are thunderbolt/display link, that currently reaches up into 8k data rate territory and is to move to 100gb/s optical cables territory (I mention all the figures for other readers), the Silicon Image wireless HD, looking at 26Gb/s + , in future new WiFi looking at over 100gb/s. I'm coming up with a new technique to beat WiFi dara density as a solution for an application, WiFi direct and thunderbolt offer the advantage of io devicrs interface, and co-hosting WiFi over wireless HD would offer over 20gb/s boost. Thunderbolt 2 and WiFi are fairly data agnostic, so offer an easier way for such a group to bypass HDMI and offer the standard to manufacturers who only need existing hardware licenses plus the new data standard.

 

[Wayne Morellini]

Here's a question, does it have the latest display port or thunderbolt 2? HDMI, is a threat to the industry with such a mess of support and upgrade delays, 1.4b should have been 2160p60 at least. p30 was unacceptable, and now we have p30 only consumer sets as the mass affordable sets without rec2020, so pretty bad when people try to hook up 24/25/60p content from reports.

I don't even remember posting this, I was wondering why this thread turned up, boy I'm getting old.

 

[Jeff Kilgroe]

Expecting them to go to 12k high bitrate etc now is a little unfair Jeff. The real word limitation, until we go to optical wires, is transmission speeds. We are at at least 10-20Gb/s currently, the prospect of doing beyond that non cheap cable I suspect night get thinner and thinner. Even thunder bolt is expecting to use multiple lines to get to 100gb/s on their optical cable convertor plugs (I think their might be ten pins I. the socket available).

I wouldn't say defining 12K would be unfair. There is a difference between defining a specification and implementing it. The limitation on bandwidth currently is in the interconnects and switching hardware. I've seen 100Gbps pushed over standard CAT-5e cable, I know people working on bringing that tech to market and it is currently functional. Some HDBASE-T methods about to be finalized for HDMI 2.0 support are using stacked signal read-outs, dual and quad data rate techniques with signal rise/fall, etc.. These are going to be pushing 20Gbps over CAT-5e for full uncompressed video and all 32 audio channels, CEC and ARC and other HDMI features. That's neither here nor there in terms of defining a spec, as in what to do with the data once it's in the system, how to structure it, which formats to allow, etc.. I only threw 12K out there as a possibility, 8K is already working in what I would consider a beta form and there are as of yet only proposed specs for official standards. No reason these are not defined now to give hardware developers a more defined target to aim for.

The problem is the standards committees don't want the task of future thinking. They just want to manage what is here now. The bulk of the industry is fine with this approach because it hinders development speed and allows piece meal delivery of new products with annual upgrade cycles. If HDMI 2.0 were finalized with support for 72Gbps 12K @ 60Hz, you know good and well that some upstart display maker would try to build such a thing and disrupt the market. That is what I would like to see, and I'm thinking it's what you want as well. Unfortunately, the industry at large does not want that. The hardware makers, as well as the various standards committees, value their control over product evolution and the pace at which it happens.

Thunderbolt 3 will improve to at least 40Gbps, possibly up to 100Gbps. The latter being Intel's original target to hit by 2016. All will use the same mini-displayport connector and will work over both copper and optical cabling. The biggest limitation to the system currently is the active electronics within the cables.

The real solutions beyond HDMI for Red and others to go to formulate a proper pro/content media interface are thunderbolt/display link, that currently reaches up into 8k data rate territory and is to move to 100gb/s optical cables territory (I mention all the figures for other readers), the Silicon Image wireless HD, looking at 26Gb/s + , in future new WiFi looking at over 100gb/s. I'm coming up with a new technique to beat WiFi dara density as a solution for an application, WiFi direct and thunderbolt offer the advantage of io devicrs interface, and co-hosting WiFi over wireless HD would offer over 20gb/s boost. Thunderbolt 2 and WiFi are fairly data agnostic, so offer an easier way for such a group to bypass HDMI and offer the standard to manufacturers who only need existing hardware licenses plus the new data standard.

Yep. I think right now, our best hope is for Apple to release 4K+ displays that shed the requirement of HDMI or DisplayPort and use proprietary signaling over Thunderbolt. It would also really help to get GPU makers onboard for this, but that's a tough sell. The primary market for GPU makers is games and will be for the foreseeable future, even with all the advances in GPGPU computing, games are still the market hinge pin. And the focus there for now is still in the 1080p realm and to continue to increase detail and realism before compounding resolutions, which equate to an order of magnitude increase in texel processing power, texture sizes and more.

 

[Wayne Morellini]

I see, I was meaning that you can't really define the actual implementable interface until you know which physical system to use and its characteristics are defined, but it is here now from what you say.

So, this cat 5e, 100ghz, why do we need a multi wired HDMI port, HDMI 3 could be a file streaming cat 5e cable. If the tech could go into thunderbolt then problem solved, the active able might be simpler. I've been thinking a methods to compensate for noise and increase data throughput, what are they doing to achieve that data Raye Jeff.

You know, decades ago I considered making a digital rca plug version to do io as well. But I was young and unskilled so didn't do anything about it, failing to realises that there was a digital audio version already, for. what you are saying we could have upgraded that up to 12k lol! I would prefer a rca, or 3.5mm version of USB, rather than USB plugs. It was rather simple, you plugged in a dumb device it reverted to analogue and the daisy chain terminated. You plugged in a older spec digital rca, the daisy chain after that converted to the lower spec. It would have made tv, audio and home theater setups a breeze just keep plugging things in one after the other, with older spec equipment progressively down the end. Mark the ports and product with a clearly labeled version number so people know which to put next on the chain.

 

[Wayne Morellini]

Jeff, I'm interested in that 100gbs cat 5e cable thing, I'm always looking into products, and I know people that can do the interface chips for it low powered and really low cost (maybe even an RCA version).

How does the technique transfer to on board use and short haul? We really need terabit solutions for holo cameras?

 

[John Bellari]

I wouldn't say defining 12K would be unfair. There is a difference between defining a specification and implementing it. ...The problem is the standards committees don't want the task of future thinking. They just want to manage what is here now. The bulk of the industry is fine with this approach because it hinders development speed and allows piece meal delivery of new products with annual upgrade cycles... The hardware makers, as well as the various standards committees, value their control over product evolution and the pace at which it happens...

Amen!

 

[M Most]

The problem is the standards committees don't want the task of future thinking. They just want to manage what is here now. The bulk of the industry is fine with this approach because it hinders development speed and allows piece meal delivery of new products with annual upgrade cycles. If HDMI 2.0 were finalized with support for 72Gbps 12K @ 60Hz, you know good and well that some upstart display maker would try to build such a thing and disrupt the market. That is what I would like to see, and I'm thinking it's what you want as well. Unfortunately, the industry at large does not want that. The hardware makers, as well as the various standards committees, value their control over product evolution and the pace at which it happens.

Jeff, I'm surprised that you would guise this as some kind of conspiracy, which is essentially what you're saying. You know as well as I do that standards take time because they need to be completely vetted, approved by many different vendors and users with many different points of view and needs, and they need to stand up to scrutiny because they require manufacturers to spend a great deal of time, money, and effort to conform to them. You don't just come up with something and say "this is forward thinking so everyone should do it." You must have very specific descriptions, you must have completely vetted math behind any required transforms, and you must have an end result that multiple manufacturers can support and build to. You create roadmaps for things like LTO and USB because you anticipate certain advancements in both mechanical and computer engineering, but those roadmaps can and usually do change and evolve. Most standards have those roadmaps proposed, but the standard itself needs to be for the here and now because it is a complete statement of requirements that must be followed in order to assure proper conformity of multiple devices. It is not a conspiracy to rob the public of a future, it is rather a way of getting things built that can be manufactured in volume and work with other devices manufactured to the same standard. Today. Not 10 years from now. Or even 3 years from now. The fact that IT based technology moves very fast doesn't change the need, it just changes the roadmap.

 

[Jeff Kilgroe]

Jeff, I'm surprised that you would guise this as some kind of conspiracy, which is essentially what you're saying. You know as well as I do that standards take time because they need to be completely vetted, approved by many different vendors and users with many different points of view and needs, and they need to stand up to scrutiny because they require manufacturers to spend a great deal of time, money, and effort to conform to them. You don't just come up with something and say "this is forward thinking so everyone should do it." You must have very specific descriptions, you must have completely vetted math behind any required transforms, and you must have an end result that multiple manufacturers can support and build to.

I agree with your points, but I'm still standing my ground on this one. HDMI output for 8K via dual-link 36Gbps has been on the drawing board for years now -- So a proposed and possible goal or entry on a roadmap for the specification. It's already got backing from NHK, JVC, Sony and most likely others too. Everyone knows where the industry is headed with the target year being 2020. Manufacturers like SHARP have been demonstrating 8K displays for more than a couple years now, just as Samsung first demonstrated 4K displays at CES in 2006. While I'm not chalking this up to a conspiracy, I do feel that and will maintain that too many compromises have been made with HDMI 2.0 in order to pace the industry. Sony, Microsoft and others had to make concessions and back down from desired inclusions in the 2.0 spec in order to get it ratified. We are not getting 10/12bit x.v. or "deep" color with UHD+ resolutions in HDMI 2.0. Why is that? We are not getting timing support and marking for simultaneous left/right stereo image pairs. Why is that? We will have the 32 channel audio ability, but no requirements to implement, no defined requirements, only half-hearted guidelines for multiple audio profiles over these channels. You want to send 2-channel stereo, 5.1 DD, and 5.1 PCM simultaneously from your set-top box? That's awesome, but TV's and AV receivers may not reciprocate and support it, let alone understand your implementation and there is nothing in the spec allowing for channel marking or profiling to make it happen in a way that's plug-n-play for the consumer. Why???

You create roadmaps for things like LTO and USB because you anticipate certain advancements in both mechanical and computer engineering, but those roadmaps can and usually do change and evolve.

I would be happy if there was an HDMI roadmap… But there is no such thing. We have various specifications hanging out in a no-man's land of future support. 8K, dual / multi-link, locking connectors, rec2020, multi-streaming, etc.. Something you're saying manufacturers and standards committees are not going to do. We have the rec2020 spec, which almost no one is going to be able to use or bother trying to use with HDMI 2.0 due to the 8bit restrictions and the industry's current affinity to cloudy edge-lit displays. It's being shoved to the side, along with > 8bpc color support for UHD as most of the lower denominator hardware companies want to reserve that for "pro" equipment. The few who would like to implement it most likely will not because it will be out of spec for v2.0. And it seems that Panasonic, SHARP and JVC are in that camp as well, the don't want to support it camp, so they can charge a premium for it on their "pro" displays. OTOH, Panasonic at least implemented HDMI ports capable of v2.0 data rates and features ahead of Sony and others.

So I agree on the changing of standards or the proposed forms of such standards on roadmaps, as new technology arises. My point was, and still is, that HDMI has not only failed to create such a proposed road map. It has failed to address the here and now. It's a complete slap in the face to anyone in the industry who was looking forward to at least keeping the color support and functionality WE ALREADY HAVE WITH HDMI V1.X and moving forward to UltraHD resolutions with those same features.

 

[Blair S. Paulsen]

Go Jeff go!

The lack of support for greater than 8 bit color at UHD rez in HDMI 2.0 is, IMO, evidence of a mandate to shorten upgrade cycles.

Cheers - #19

 

[Wayne Morellini]

Jeff, I'm surprised that you would guise this as some kind of conspiracy, which is essentially what you're saying. You know as well as I do that standards take time because they need to be completely vetted, approved by many different vendors and users with many different points of view and needs, and they need to stand up to scrutiny because they require manufacturers to spend a great deal of time, money, and effort to conform to them.

You amaze me M, these are all elements used in conspiracies for collusion.

You have presented an ideal form of how things should work if everybody behaves themselves. and Jeff is merely persuaded that not everybody did so.

BTW I looked at the HDMI 2 feature list, and it does have greater than 8 bits at UHD in 4:2:2 (and 4:2:0) which is good enough to transmit a Bayer sensor pattern.

See here;

https://www.apertus.org/forums/viewtopic.php?f=4&t=1331

 

[M Most]

.... HDMI has not only failed to create such a proposed road map. It has failed to address the here and now. It's a complete slap in the face to anyone in the industry who was looking forward to at least keeping the color support and functionality WE ALREADY HAVE WITH HDMI V1.X and moving forward to UltraHD resolutions with those same features.

One of the fundamental differences between the way you seem to see this and the way I do is that I feel that HDMI was, is, and always will be a far too complex connection for continued evolution in a consumer based environment, even with things like HDBaseT as a long-run transport.. I feel it needs to be simplified, particularly in its physical connectors and attendant cables. In other words, I want to see a replacement for it, not an even more convoluted evolution of a flawed design. I would much rather see something much more open ended with high bandwidth and a more rugged but simpler connector replace the current form factor, something that CAN be evolved practically and economically in some of the ways you're hinting at. I'm just not convinced that HDMI is that technology.

 

[Wayne Morellini]

One of the fundamental differences between the way you seem to see this and the way I do is that I feel that HDMI was, is, and always will be a far too complex connection for continued evolution in a consumer based environment, even with things like HDBaseT as a long-run transport.. I feel it needs to be simplified, particularly in its physical connectors and attendant cables. In other words, I want to see a replacement for it, not an even more convoluted evolution of a flawed design. I would much rather see something much more open ended with high bandwidth and a more rugged but simpler connector replace the current form factor, something that CAN be evolved practically and economically in some of the ways you're hinting at. I'm just not convinced that HDMI is that technology.

M that is basically in agreement with me rather than in difference I've revealed the digital rca thing, but in reality the plug on it could have been a square/rectangle or round sort of plug the size of the cable. instead. Hdmi rather than a continued revolution is just a present practical reality to use to transmit HDMI out of spec on existing equipment. Display link has long overtaken it into new areas, especially with thunderbolt io, even with a wireless version under the wigig alliance. But I think it is vital to loom into cohosting on a wireless HD link, which is currently higher bandwidth at 60ghz. it may not be as useful. but as a high device bandwidth link it is before going to post ,100gb/s modes.

 

[Jeff Kilgroe]

One of the fundamental differences between the way you seem to see this and the way I do is that I feel that HDMI was, is, and always will be a far too complex connection for continued evolution in a consumer based environment, even with things like HDBaseT as a long-run transport.. I feel it needs to be simplified, particularly in its physical connectors and attendant cables. In other words, I want to see a replacement for it, not an even more convoluted evolution of a flawed design. I would much rather see something much more open ended with high bandwidth and a more rugged but simpler connector replace the current form factor, something that CAN be evolved practically and economically in some of the ways you're hinting at. I'm just not convinced that HDMI is that technology.

I actually agree with all of that.

I think that what really needs to happen here is to take the complexity out of the equation for the interface itself. An interface like Thunderbolt could offer a lot here -- it's simple, uses PCIe signaling, which is standard and completely agnostic to whatever data we want to send over it. Not to say that A/V equipment should dump HDMI for Thunderbolt or that Thunderbolt itself is the answer, it's a possibility to consider for the future. So instead of concerning ourselves over a future HDMI 3.0 or whatever and what features it can support, perhaps we could have Thunderbolt 3 at 40Gbps and that's as far as we need to read into the interface.

So what then? I suppose the industry could adopt a standard protocol for device communication. Output devices could broadcast resolutions, color modes, etc.. they are capable. Display devices could do the same for what they can accept. Connected devices could negotiate a best-case connection scenario based on these conditions as well as user input. And negotiate and establish a means to exchange the necessary data. Let the specification determine the bandwidth of the interface and how devices communicate, don't spell out in the specification what it is they communicate about or dictate what sort of modes, colors, etc.. the equipment is require or allowed to generate or display.

Considering that, I think this comes back to some of my concerns above about why the HDMI 2.0 spec misses the mark on so many key points. It's such a detailed specification, yet fails to address several aspects of A/V devices that all but shipping right now. There are extensions to the spec already defined and tiered delivery considerations. HDMI 2.0B / C can deliver that elusive 10/12 bits at 4K, for example. The whole thing just seems like a circus to me. On one hand it's easy to suspect an intentionally paced evolution of the industry. On the other hand, we regularly see the consumer electronics industry spending stupid amounts of time and money trying to solve the wrong problems.

 

[Wayne Morellini]

Hmm, what you say is right. What they can look at is a basic data protocols for a range of ability.

You can specify resolution in a 20 bit range with one bit as an extension to another (has scientific, forensic, data display and recording uses). You can specify ranges per pixel (primaries and data, their color spaces and data structure descriptions), bits per range, frame rate etc. What ever good they come up with can be accommodated by just plugging the numbers in and then negotiate the best fit with the device. Most of this big data capability will be hidden away by the use of the format description headers of course.

 

[Ron Jones]

Go Jeff go!

The lack of support for greater than 8 bit color at UHD rez in HDMI 2.0 is, IMO, evidence of a mandate to shorten upgrade cycles.

Cheers - #19

HDMI 2.0 does support greater than 8-bit depth with 4K/UHD @60Hz when using 4:2:2 and 4:2:0 chroma subsampling. Specifically 12-bit depth is supported for 4:2:2 and up to 16-bit depth with 4:2:0. It's just with 4:4:4/RGB that the bit depth is limited to 8-bits for 4K/UHD signals at 60 Hz. The above info is for HDMI 2.0 implementations supporting the full 18 Gbps bandwidth limit.

 

[Wayne Morellini]

HDMI 2.0 does support greater than 8-bit depth with 4K/UHD @60Hz when using 4:2:2 and 4:2:0 chroma subsampling. Specifically 12-bit depth is supported for 4:2:2 and up to 16-bit depth with 4:2:0. It's just with 4:4:4/RGB that the bit depth is limited to 8-bits for 4K/UHD signals at 60 Hz. The above info is for HDMI 2.0 implementations supporting the full 18 Gbps bandwidth limit.

Shhh, John... not when they are thinking they are all right

 

[Wayne Morellini]

At the risk of disagreeing with Jeff, I do expect a more affordable consumer targeted RedRay-ish device once HDMI 2.0 is readily available - IF - the marketplace for such a device catalyzes.

Cheers - #19

Yes, more like $200 or $100 is a good target for market penetration. Not going to see that soon, but we can only wish. The industry needs to establish film cooperatives to produce and market content into redray odemax. To establish 4k redray we need more than some low grade shoot 4k features along side high end features, we need a variety of content for people to come to and stay on service, like YouTube is trying to do. If Red is wondering what to do next (apart from 8k, cheap 4k video production/cinema camera, and holographic cameras) this is something to aim for.

 

[jon honeyball]

As an owner of a redray box, I have emailed odimax twice now. I get no reply. Shrugs.

 

[Al Lougher]

As an owner of a redray box, I have emailed odimax twice now. I get no reply. Shrugs.

Did Odemax die already?