ODEMAX in the fall?

[Elsie N]

Its not a question to me of how they ultimately deliver. Its their course of conduct in the past towards consumer end users. Nothing more, nothing less. Elsie. Didn't mean to seriously wound you and I know I didn't, just an all too easy shot at a huge target with a perfect set up pitch from Justin. He can be my pitcher in the home run derby anytime.

Heh, heh...

 

[Corey_O]

Mark I'm sure when you gave Sony constructive criticism, about a product which they had not released yet, they responded toward you in a prompt and favorable way because you are the all mighty "consumer end user"!

Why didn't sony listen to you about 84" 4k dumbo ears? ; )

 

[mark haflich]

When I email Sony, obviously a much much bigger company, they at least respond. Remember the old Sony motto? Sony. No baloney. Sony didn't blame everything on them and the world catching the flu. The Sony 4K server launch was a complete cluster. But they have been at least forth coming. And yes. I and others are the all mighty consumer end user. And if you don't get it you will perish.

 

[Corey_O]

Can't say I do remember that classic motto."Sony. No baloney". Why would they ever have discontinued being associated with the word baloney.

Sony baloney or Sony no baloney! Did they actually put a sandwich in a commercial?

I love it. I own plenty of sony stuff no hater here. ; )

 

[Justin O'Neill]

Sony didn't blame everything on them and the world catching the flu.

I occasionally run into people who don't like RED's fly by the seat of your pants culture and I realize it's not for everyone. Personally I find it refreshing and exciting but I'm sure that partly has to do with being with them from the beginning when you really could know everyone who worked there and see it grow from a scam to an industry leader. I talk to people who come up with all kinds of theories about why RED misses deadlines, or make big changes to long held plans but I can tell you the explanation is always the simplest. They are a small, passionate group pushing the boundaries of technology and they make mistakes and can be overly optimistic about the amount of time it takes to develop products. It is unclear exactly how closely related RED and Odemax are, but from my dealings with them they exhibit the exact same ethos as RED. It sucked we didn't get the planned announcement at Sundance but there is no conspiracy. It is a very small group and they all did get sick. These aren't big multi-natonal companies that can fly in an executive to cover in a situation like that. Everyone who knew enough to make some kind of announcement was taken out.

And yes, I am aware I have drank the cool-aid.

 

[mark haflich]

Red was never ever a scam and I am sure you didn't mean literally that. Their initial show appearance appeared to be most unsubstantiated hype but they pretty much delivered what they were hyping. Odemax could have responded with a blast email to inquiring consumers and their delay even until now is relatively harmless except for consumer relations because Red has not really progressed very far in fulfilling orders. The first production rune and now a long long running coming soon and a price increase. Because the core of Odemax doesn't appear to give the tiniest nod to its potential end user, no matter how it may grow, I will have nothing to do with them. Time for a bourbon.

 

[Jeff Kilgroe]

http://www.streamingmedia.com/Artic...ins-the-Significance-of-HEVC-H.265-90599.aspx

Great video regarding the hurdles and implementation of H.265.
Interesting points- 3-5x the encode time over 264, 1.5-2.5x overhead on decode, Promised 50% bandwidth saving hasn't been achieved yet.

Not going to get the 50% bandwidth or data savings unless comparing 4K. The larger the frame size, the greater the potential to squeeze it down. Even then, H.265 is falling way short of a lot of the hype. Takes 3 to 5 X longer to encode (or 3 to 5 X as much compute power to encode at the same rate as H.264). Same for decode, it requires several X more compute power. Decode on H.265 seems to be more demanding that REDCODE by a good bit. Not going to be editing away with X.265 footage anytime soon without transcoding and/ or using accelerator hardware. The one advantage there is that it's open and the industry is dead set on making it the next standard, so upcoming GPUs and CPUs will have X.265 acceleration integrated.

In my professional opinion, H/X.265 sucks balls. But it's where we're headed. Sure there are some superior options out there, but all those superior options are proprietary.

 

[Blair S. Paulsen]

(snip)
In my professional opinion, H/X.265 sucks balls. But it's where we're headed. Sure there are some superior options out there, but all those superior options are proprietary.

VP9? ;-)

IAC, IMO, the one crucial part of this is the need for a widely supported codec that can be decoded on the fly by an ASIC. General purpose computers are designed to handle a wide variety of tasks and that flexibility requirement is anathema to maximum efficiency for any single task. An ASIC, in simple terms, is a custom hardware based solution that has essentially no flexibility at all. Spinning an ASIC capable of a complex task like decoding HEVC/H.265 is neither simple or cheap, but once accomplished it offers huge economies of scale in mass quantities.

AFAIK the HEVC/H.265 roadmap posits a marketplace for billions of decoding ASICs embedded in virtually every media device from smart phones to the TV in your living room at a wholesale cost to the device makers of less than $10/ea. Among other upsides of this topology, if the ASIC is doing all the decoding it takes the load off other computing/processing resources. On top of everything else, a dedicated ASIC if extremely power and space efficient vs FPGAs/CPUs/GPUs.

Having a solid codec with a realistic potential to last at least a decade is key to an ASIC based decoding scenario. If HEVC/H.265 is not that codec, then the core value proposition of ASIC based decoding is severely handicapped.

Cheers - #19

 

[Jeff Kilgroe]

Wile I somewhat share your enthusiasm for dedicated HEVC/ x.265 processors, I don't foresee these road maps coming to pass. The only ones out there making official announcements are companies like NVidia and PowerVR, who intend to roll support into other upcoming products. Ultimately I think this will play out just as the whole x.264 timeline did. So I'm thinking slow adoption at first, hardware acceleration available in desktop and mobile GPU platforms by mid 2014. Broadcom is at the forefront of having a dedicated ASIC for x.265 / HEVC, but they may or may not push forward. Intel's HD5000 GPU in the next iteration - 5th generation Core CPUs should have support integrated.

As it stands, there are already a number of common CPU and GPU combinations on the market that can decode h.265 at 8bit 4K in real time without straining too hard. By the time it's mainstream, 4K monitors and UHDTV's will be in proliferation, 5th gen Core CPUs as well as next gen GPUs like the GeForce 800 series will be shipping as well.

a huge strike against commodity ASICs to serve here is that while the new x.265 specs are open, everyone is going to have their own flavor, just as we see with x.264 and AVC-Intra variants.

 

[Stacey Spears]

VP9 is inferior to H.265 regardless of what Google tells you. The industry will move quickly to H.265. Encode times will improve a lot over the next couple of years. How it compares to REDRAY is unknown at this point.

The compression specs / standards are for decode only. Every 264 decoder must be capable of decoding every 264 encoded bitstream, even when those bitstreams are not complaint. Producing a complaint bitstream seems to allude many encoder vendors. Poor decoder vendors have to deal with the mess. This applies to MPEG-2 as well. It will most likely apply to 265 in the future. As bad as this is, its nothing compared to the interop nightmare that is HDMI. 265 does have more parallism built in, but its optional, so GPUs will continue to suck at decoding without dedicated decoding chips like they have today.

HDMI 2.0 standard is now approved and its sad. This is where Sony, and other large CE companies, really do deserved to be badmouthed. Buying up smaller companies in order to obtain voting rights to block the incorporation of DisplayPort into the HDMI 2 standard was dishonest. You would be amazed at how many display vendors did not want 10-bit+ support. The only thing more depressing than HDMI 2.0 is how the game console fanboys are behaving right now over unreleased products.

 

[Chad Lancaster]

ODEMAX has been tweeting "8113" "0801" & "VIII I"

 

[Neil W. Smith]

HEVC, VP9 and HDMI 2.0

HEVC, VP9 and HDMI 2.0

VP9 is inferior to H.265 regardless of what Google tells you. The industry will move quickly to H.265. Encode times will improve a lot over the next couple of years. How it compares to REDRAY is unknown at this point.

The compression specs / standards are for decode only. Every 264 decoder must be capable of decoding every 264 encoded bitstream, even when those bitstreams are not complaint. Producing a complaint bitstream seems to allude many encoder vendors. Poor decoder vendors have to deal with the mess. This applies to MPEG-2 as well. It will most likely apply to 265 in the future. As bad as this is, its nothing compared to the interop nightmare that is HDMI. 265 does have more parallism built in, but its optional, so GPUs will continue to suck at decoding without dedicated decoding chips like they have today.

HDMI 2.0 standard is now approved and its sad. This is where Sony, and other large CE companies, really do deserved to be badmouthed. Buying up smaller companies in order to obtain voting rights to block the incorporation of DisplayPort into the HDMI 2 standard was dishonest. You would be amazed at how many display vendors did not want 10-bit+ support. The only thing more depressing than HDMI 2.0 is how the game console fanboys are behaving right now over unreleased products.

Very interesting perspective, Stacy .... a couple of follow up questions if I may:

1) apart from Google's inability to market certain of their product lines in an effective way, what from a technical point of view are the key differences between VP9 and HEVC algorithms? ... what's the basis of the inferiority complex of VP9?

2) will HEVC decoders also have to decode all the H.264 bitstream variations? ... as well as all the H.265 variations?

3) what do you see as the main shortcomings of HDMI 2.0 spec? ... why do the panel vendors want to be limited to 8bit color gamut?

I thought that the Sony Triluminous (read xvYCC) wide gamut UHDTV feature was based on a 10 bit color space ... no hard evidence for that, just an assumption.

4) lastly, do you see a certain company from Cupertino throwing a major wrench into the mix and shifting the goalposts for the UHDTV content creation and delivery ecosystem in a way that the ITU has no idea of? ... think what they did to the music delivery and consumption industry a few years back.

Neil

 

[M Most]

4) lastly, do you see a certain company from Cupertino throwing a major wrench into the mix and shifting the goalposts for the UHDTV content creation and delivery ecosystem in a way that the ITU has no idea of? ... think what they did to the music delivery and consumption industry a few years back.

Except that Apple has not achieved any kind of dominance in video downloads or streaming as they did in the music business. The video pie is much more diverse, split between not only Netflix and Amazon, but Google, Hulu, Vudu, Sony, Microsoft, and others. I don't see where Apple or any other single player is going to be able to put that genie back in a bottle, no matter how much they might want to.

 

[Elsie N]

ODEMAX has been tweeting "8113" "0801" & "VIII I"

That suggests a dual release (a month apart '-) for Dragon upgrades and some sort of opening up of their services. Wonder if this also means the beginning of .RED encoders.

 

[Jeff Kilgroe]

VP9 is inferior to H.265 regardless of what Google tells you.

Yep. Not only that, but much like the whole Android vs. iOS ordeal, there are some out there who also consider it "stolen technology".

HDMI 2.0 standard is now approved and its sad. This is where Sony, and other large CE companies, really do deserved to be badmouthed. Buying up smaller companies in order to obtain voting rights to block the incorporation of DisplayPort into the HDMI 2 standard was dishonest. You would be amazed at how many display vendors did not want 10-bit+ support. The only thing more depressing than HDMI 2.0 is how the game console fanboys are behaving right now over unreleased products.

I've been following the whole HDMI 2.0 ordeal and it's so frustrating. I just don't even have any way of comprehending how this could have been botched so badly. The big guys, namely Sony, have done a huge disservice to everyone. Including themselves! They have literally strong-armed into existence an interface standard that isn't even adequate for their existing display technology, let alone all their new display technology to come over the next few years. Makes no sense and I often wonder if their interface people have any clue what the rest of the company is cooking up.

I can understand the motivation to push out DisplayPort. Typical industry politics for sure and It wasn't the answer either, IMO. But they could have at least made an effort to replace it with something better. The one good thing that may come of all of this, if Apple and Intel play their hands right, is Thunderbolt could push in as the next dominant interface for computer monitors (among other peripherals). Thunderbolt doesn't have to limit display signaling to DisplayPort and other lesser supported formats. On its PCIe topology, they can create any signaling interface they wish. I've seen some stuff here that looks very promising, but can't say more than that.

That suggests a dual release (a month apart '-) for Dragon upgrades and some sort of opening up of their services. Wonder if this also means the beginning of .RED encoders.

I think it just means the ODEMAX service goes live on August 1. There have been numerous tweets...

Remember, ODEMAX is independent of RED and has nothing to do with .RED encoding or encoders. .RED encoder software is already in the hands of every REDRAY owner. It's definitely in beta form as it doesn't yet support several desired frame rates and a few other options.

 

[Elsie N]

...
Remember, ODEMAX is independent of RED and has nothing to do with .RED encoding or encoders. .RED encoder software is already in the hands of every REDRAY owner. It's definitely in beta form as it doesn't yet support several desired frame rates and a few other options.

Jeff, are you saying ODEMAX services will be open to all (or at least some) codecs other than .RED? I've been operating under the impression only .RED codec would be used in their service.

 

[M Most]

Remember, ODEMAX is independent of RED and has nothing to do with .RED encoding or encoders. .

Directly from the Odemax web page:

Designed to exclusively support the REDRAY Player and CRIMSON Projection systems from RED Digital Cinema.....

I've always been under the impression that not only is Odemax using the .Red codec exclusively, but that Red has a significant investment in it. Perhaps Jarred could clarify whether or not that's the case, of if he's at liberty to say.

 

[Elsie N]

Directly from the Odemax web page:



I've always been under the impression that not only is Odemax using the .Red codec exclusively, but that Red has a significant investment in it. Perhaps Jarred could clarify whether or not that's the case, of if he's at liberty to say.

Whew! I was beginning to think I was the Aaron Rodgers to the ODEMAX Ryan Braun, and was going to have to apoligize to Mark Haflich for my part in some of our past disagreements. That would have been one awful tasting plate of crow to eat. '-)

 

[Stacey Spears]

1) apart from Google's inability to market certain of their product lines in an effective way, what from a technical point of view are the key differences between VP9 and HEVC algorithms? ... what's the basis of the inferiority complex of VP9?

VP8 is an evolution of VP9. 265 is an evolution of 264. 265 has the benefit, and detriment, of several companies providing IP. A lot mostly to ensure their piece of the royalty pie in the future. That means not everything submitted is worth having. VP8 was on par with VC-1 and both were technically inferior to H.264, though VC-1 beat H.264 MP in several blind tests done by Hollywood studios in 2004/2005. Sadly Microsoft's boasting lead to specific tools being added to H.264 high profile, which fixed the shortcomings main profile had. As Jeff pointed out earlier, its all about the encoder implementation. Not all 264 encoders are created equal. (Same for any codec) There is only one VP8 encoder that I am aware of. The VP8 spec has pages copy and pasted straight from H.264, which makes their claims about IP free laughable. 265 has the worlds leading experts on video compression contributing to it. They are not necessarily video people and I use the term video to mean digital images, not 60Hz interlaced content. They get math and algorithms but sometimes miss the image quality mark. 265 has a superior tool set to VP9. It has wide industry adoption. They did target mobile over UHD during its development because that is where the money is. I wish 4:2:2 10-bit would have been part of it from day 1 instead of a year later. If we get UHD Blu-ray, it will probably go with H.254 10-bit 4:2:0. We should know by years end.

The current open issue with 265 is the patent pool and royalties. Until this is worked out, supporting 265 is a risk. MPEG-4 ASP failed because it had no caps. 264 did not originally have caps, but VC-1 forced a royalty cap. Anyone selling an H.264 decoder must pay $0.20 to MPEG LA until they reach the cap. MPEG-2 has no such cap and everyone must pay $2.00 to MPEG LA forever. If 265 does not have a cap, it will fail. Hopefully VP9 will motivate a cap for 265.

2) will HEVC decoders also have to decode all the H.264 bitstream variations? ... as well as all the H.265 variations?

No, a 265 decoder only needs to decode 265. With that said, no one could sell a product (ASIC) that only supports 265. Chips will continue to support all codecs. (sans VPn) MPEG LA was looking at forming a patent pool for VP8, not sure what is currently happening. They would probably do the same for VP9, which throws a wrench into IP free.

3) what do you see as the main shortcomings of HDMI 2.0 spec? ... why do the panel vendors want to be limited to 8bit color gamut?

Until publically available, I am limited on what I can say. They added 4:2:0 support specifically for 10+bit UHD. They should have supported 10/12-bit 4:4:4 UHD @ 60 Hz. (or 120 Hz) Decisions were made around todays technology, not tomorrows.

I thought that the Sony Triluminous (read xvYCC) wide gamut UHDTV feature was based on a 10 bit color space ... no hard evidence for that, just an assumption.

The "mastered in 4k" Blu-rays are 8-bit and claimed to support xvYCC. You don't need more bitdepth to support the wider gamut. There is a "fight" going on now about wide gamut in general. One of the concerns against it is displays not properly displaying existing 709 content. The fear is existing content not will out of whack with the wide gamut. Display manufacturers get so much wrong today. HD sets still suffer the same artifacts as SD set, just more resolution. We are now seeing the same thing with UHD sets.

4) lastly, do you see a certain company from Cupertino throwing a major wrench into the mix and shifting the goalposts for the UHDTV content creation and delivery ecosystem in a way that the ITU has no idea of? ... think what they did to the music delivery and consumption industry a few years back.

If they come out with a display, of any resolution, I hope they meet system standards and not try and push their own. Google is pushing VP9 for UHD on YouTube. Others, like Netflix, are moving to H.265. I suspect Apple would follow H.265 as well. While a lot of content is shot in UHD+, not a lot is being finished.

Right now a lot of the HT magazines are kind of dismissing UHD. Most of the reviews of been done without actual UHD content.

 

[Jeff Kilgroe]

Jeff, are you saying ODEMAX services will be open to all (or at least some) codecs other than .RED? I've been operating under the impression only .RED codec would be used in their service.

Perhaps I didn't word that correctly, but what I was saying is that ODEMAX is an independent company from RED. And ODEMAX does not have any bearing or any say as to when .RED encoder software or REDAY features are released. ODEMAX is a REDRAY channel partner and will be the first, of hopefully many, to launch services on the REDRAY platform.

Directly from the Odemax web page:

I've always been under the impression that not only is Odemax using the .Red codec exclusively, but that Red has a significant investment in it. Perhaps Jarred could clarify whether or not that's the case, of if he's at liberty to say.

ODEMAX is using the .RED codec exclusively. For now their services are only on REDRAY or will only be on REDRAY for the foreseeable future and that is .RED only, obviously. I was just saying that ODEMAX is an independent company. Something both Jim and Jarred have stated. Now as to whether or not RED has invested in them is a different matter and who knows whether or not they're at liberty to say.

RED has invested on some level I guess. ODEMAX has maintained a presence at RED Studios and I think they have had to work closely with RED to get this off the ground. While they may be independently owned, they are not fully independent. There is probably a lot of mutual sharing between the two in developing the REDRAY distribution pipeline.