UHD - the questions to be answered... - Page 2 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #31 of 51 Old 09-19-2013, 10:40 AM
AVS Special Member
 
8mile13's Avatar
 
Join Date: Nov 2009
Posts: 3,679
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 71 Post(s)
Liked: 188
Quote:
Originally Posted by Glimmie View Post

Quote:
Originally Posted by 8mile13 View Post

Keep in mind that many manufacturers will not provide such an software upgrade on many of their products especially the older ones and i am pretty shure that a lot official cable provider, satellite provider etc.. receivers are not HEVC capable.

And that's simply because it's not technically possible in addition to all the other issues. Consumer products are built to a price and to get there they use custom designed ASICs. That's "hard" hardware. Sure you can download a software update to change the menu structure or possibly do some minor alterations to the decoder. But you aren't going to find an MPEG2 decoder in a consumer product that can be magically download to an MPEG4 or 5 decoder. We currently cannot build hardware like that at the current consumer price point.
http://mpeg.chiariglione.org/standards/mpeg-h/high-efficiency-video-coding
quote
''The coding layer of the current HEVC draft specification and the corresponding HEVC test model (HM) software and encoding algorithms is still based on the traditional hybrid coding approach as found in previous standard designs''
doesn't that mean that an update, at least theoretically, should be possible when owning a divice which includes a previous standard?
8mile13 is offline  
Sponsored Links
Advertisement
 
post #32 of 51 Old 09-19-2013, 10:54 AM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by 8mile13 View Post

http://mpeg.chiariglione.org/standards/mpeg-h/high-efficiency-video-coding
quote
''The coding layer of the current HEVC draft specification and the corresponding HEVC test model (HM) software and encoding algorithms is still based on the traditional hybrid coding approach as found in previous standard designs''
doesn't that mean that an update, at least theoretically, should be possible when owning a divice which includes a previous standard?

No, not at all. It means that someone can design a new MPEG encoder/decoder by using code blocks from an existing design. Also consider if a device even has a programmable architecture. Note my FPGA comment. Just because you can download a new menu skin or add some simple feature to a product does not mean you can completely change out the hardware function.

Can I load OS10 into a PC? Can I load Windows 8 into a MAC? It's a computer so it's just software right? And no an emulator program is not the same thing.

Why did Dish and DirecTV as well as many cable companies have to swap out hardware boxes when they went MPEG4 for HD? That was a huge expense. Believe me if there was any way around that they would have done it.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #33 of 51 Old 09-19-2013, 11:08 AM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Quote:Originally Posted by Glimmie View Post
 
Quote:
Aside from "where to put it" point you made below, I'm still not sure why this is a tough issue. Firmware updates are nearly everywhere. Even TVs that aren't connected to the internet have had SD card slots allowing updates. And asside from bandwidth, MPEGanything is a software issue. Unless the raw processing power required suddenly went up, which isn't out of the question.


Your are grossly over estimating the abilities of a firmware update with today's technology - I know as I am a hardware engineer. In order to have an MPEG decoder that is upgradable from MPEG 2 to MPEG 4/5/6/7/8 would require the hardware to be entirely FPGA based

 

I wasn't aware of that for sure---but I wasn't grossly over estimating anything.  I left room for that as a possibility with "unless the raw processing power required suddenly went up, which isn't out of the question."

 

Quote:
That's expensive. Can't do that in a $500 HDTV. And even if you were FPGA based, will that chip(s) have enough gates for the new decoder algorithm? Consumer video processing products are ASIC based. They have to be in order to meet the current price point. There is a limit to how much field programmable capabilities you put in a low cost chip.

 

Even though I'm a software engineer, I know what gates and ASIC are.  But what do you mean "have enough gates for the new decoder algorithm"?  You mean "an ASIC capable of supporting the bandwidth required"?  Why "gates"?


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #34 of 51 Old 09-19-2013, 11:19 AM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Originally Posted by Glimmie View Post
 
Quote:
Originally Posted by 8mile13 View Post

http://mpeg.chiariglione.org/standards/mpeg-h/high-efficiency-video-coding
quote
''The coding layer of the current HEVC draft specification and the corresponding HEVC test model (HM) software and encoding algorithms is still based on the traditional hybrid coding approach as found in previous standard designs''
doesn't that mean that an update, at least theoretically, should be possible when owning a divice which includes a previous standard?

No, not at all. It means that someone can design a new MPEG encoder/decoder by using code blocks from an existing design. Also consider if a device even has a programmable architecture. Note my FPGA comment. Just because you can download a new menu skin or add some simple feature to a product does not mean you can completely change out the hardware function.

Can I load OS10 into a PC? Can I load Windows 8 into a MAC? It's a computer so it's just software right? And no an emulator program is not the same thing.

 

Code blocks?

 

You have a point about there not being enough bandwidth in the existing CPU's to manage the new decoding algorithm without hardware support, but you've lost me on your OS10 and Windows 8 analogies.  To modify your tone to suit my needs: "Trust me, I'm a software engineer, I know".  Besides, for fun: http://www.makeuseof.com/tag/how-to-install-mac-os-x-on-a-pc-without-using-a-mac/, among many such links.

 

Something I need you to specifically clarify though: Are you claiming that the new encoding standards will be spec'd to run on specific ASICS and graphics cards?  If so, I find that alarming and pointless.  All software, unless doing "bottom level coding", use established API's below which can be anything including pure software on a fast CPU.


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #35 of 51 Old 09-19-2013, 11:23 AM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by tgm1024 View Post

Even though I'm a software engineer, I know what gates and ASIC are.  But what do you mean "have enough gates for the new decoder algorithm"?  You mean "an ASIC capable of supporting the bandwidth required"?  Why "gates"?

Because dedicated products like TVs DVDs, Bluray Players, are hardware based decoders. The MPEG algorithm is implemented in hardware for speed as well as cost. If using FPGAs, a particular MPEG2 decoder may need 8000 gates. The engineer chooses a part with 10,000 gates. New along come the MPEG5 decoder that needs 14,000 gates. See the problem. And it's a moot point anyway because no Wall Mart or Best Buy consumer product will use FPGAs. Just too expensive.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #36 of 51 Old 09-19-2013, 11:33 AM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Originally Posted by Glimmie View Post
 
Quote:
Originally Posted by tgm1024 View Post

Even though I'm a software engineer, I know what gates and ASIC are.  But what do you mean "have enough gates for the new decoder algorithm"?  You mean "an ASIC capable of supporting the bandwidth required"?  Why "gates"?

Because dedicated products like TVs DVDs, Bluray Players, are hardware based decoders. The MPEG algorithm is implemented in hardware for speed as well as cost. If using FPGAs, a particular MPEG2 decoder may need 8000 gates. The engineer chooses a part with 10,000 gates. New along come the MPEG5 decoder that needs 14,000 gates. See the problem. And it's a moot point anyway because no Wall Mart or Best Buy consumer product will use FPGAs. Just too expensive.

 

Yes, but I contend you're stating it incorrectly.  Decoding an image is ultimately a data issue, period end of story.  One form of data comes in, and a raster comes out.  If you need hardware assist to accomplish this in any particular box without adequate CPU, then so be it, but it's not written into any data specification any specific hardware design.  Or is this what you're actually saying: (again) that somehow a particular chunk of hardware is written into the decoding spec?

 

What's making me smile about this dialogue is that it sounds like 100 previous arguments between software and hardware teams in prior startups that I've tried to diffuse.


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #37 of 51 Old 09-19-2013, 11:35 AM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by tgm1024 View Post

Code blocks?

You have a point about there not being enough bandwidth in the existing CPU's to manage the new decoding algorithm without hardware support, but you've lost me on your OS10 and Windows 8 analogies.  To modify your tone to suit my needs: "Trust me, I'm a software engineer, I know".  Besides, for fun: http://www.makeuseof.com/tag/how-to-install-mac-os-x-on-a-pc-without-using-a-mac/, among many such links.

That looks like a lot of hardware and low level software hacking to me. I mean just slap a disk in or download some file and it's done. That's what some of you are expecting for an MPEG decoder upgrade. A TV that can decode any new compression standard, standards that have not even been developed yet.
Quote:
Something I need you to specifically clarify though: Are you claiming that the new encoding standards will be spec'd to run on specific ASICS and graphics cards?  If so, I find that alarming and pointless.  All software, unless doing "bottom level coding", use established API's below which can be anything including pure software on a fast CPU.

Encoding standards aren't spec'd to run on anything. An algorithm can be implemented in hardware or software. Show me a TV today that is based on a fast CPU architecture that can be upgraded to different video and audio codecs. Now to be specific I know you can build a media PC that will do all of this and connect it to a monitor or projector. But it's expensive, complex to use, and hardly an appliance type device.

The closest thing we have to a general media computer without being a MAC/PC is a BluRAy player. And look at all the limitations here in updatability. If there was an new MPEG5 BluRay standard, how many dedicated BluRay players could be upgraded. I'll bet none.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #38 of 51 Old 09-19-2013, 11:39 AM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Quote by Glimmie:
 
Encoding standards aren't spec'd to run on anything.

For algorithms like image compression, yes they are.  Again, unless there's something about MPEG4/5 that says "this hardware is required".

 

We're talking past each other in the timeline....I'm waiting for you to catch up :)


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #39 of 51 Old 09-19-2013, 11:54 AM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by tgm1024 View Post

For algorithms like image compression, yes they are.  Again, unless there's something about MPEG4/5 that says "this hardware is required".

We're talking past each other in the timeline....I'm waiting for you to catch up smile.gif

Algorithms are not specific to hardware or software. Take noise reduction for example. It was first done in hardware for speed. Today it can be done in software on general purpose CPU hardware in real time - within reason.

I never said MPEG4/5 requires specific hardware. In fact these algorithms are designed in software. What I am saying is that you cannot compete in the market place with a general purpose computer running a software decoder against a dedicated purpose hardware device.

You seem to be saying you could build consumer media products on general purpose computer hardware that is upgradable to new standards with simple software upgrades. I fully agree you can and in fact that's what the HTPC hobby is all about.

But that's not what is in the average living room. Consumer media products are largely hardware based. They are not capable of drastic design change via software upgrades. The initial point is that ATSC tuners today are MPEG2 based and there is no possibility for the millions of products out there to be upgraded to MPEG4 or 5 with software downloads. The hardware used greatly limits that.

My point is not about what is technologically possible, but rather what we have and what we are stuck with for some time to come. How many if any consumer STBs or TV's can be upgraded to MPEG5? I'm betting none at this time. Sure you could build one than can be, but what would it cost and would you be willing to guarantee the customer it's future proof for X years?

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #40 of 51 Old 09-19-2013, 12:45 PM
AVS Special Member
 
Chronoptimist's Avatar
 
Join Date: Sep 2009
Posts: 2,557
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 203
Quote:
Originally Posted by Glimmie View Post

Can I load OS10 into a PC? Can I load Windows 8 into a MAC?
Yes.
Chronoptimist is offline  
post #41 of 51 Old 09-19-2013, 01:26 PM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Originally Posted by Glimmie View Post
 
Quote:
Originally Posted by tgm1024 View Post

For algorithms like image compression, yes they are.  Again, unless there's something about MPEG4/5 that says "this hardware is required".

We're talking past each other in the timeline....I'm waiting for you to catch up smile.gif

Algorithms are not specific to hardware or software.

 

That's what I'm saying.  It sounds like that's what you're saying too, but with an emphasis away from the hardware thing (and OS10 on pc hardware, etc.)  Good enough for me. Same page.  :)

 

By the way, it's worth a few yucks to point out there's an inflection problem here too.  And this drives me crazy.  It's been the cause of many a virtual death threat nailed to many a virtual door in online conversations.  Usenet was a hoot for this.  Oye.  For example, one of your prior statements:

 

"Encoding standards aren't spec'd to run on anything."  Sounded at that moment to me as:

"Encoding standards aren't spec'd to be able to run on just anything", as opposed to the way it sounds now to me:

"Encoding standards aren't spec'd to run on anything in particular", which is of course what I've been trying to say.


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #42 of 51 Old 09-19-2013, 01:36 PM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by Chronoptimist View Post

Yes.

Ok bad example. But you know what I mean. It's not a simple novice proceedure like a firmware update.

And I'm still not sure that a universal thing either. It may work fine for Word and Excell but I question if some high end video edit or color correction software would run on a kludged system like that.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #43 of 51 Old 09-19-2013, 02:45 PM
AVS Special Member
 
8mile13's Avatar
 
Join Date: Nov 2009
Posts: 3,679
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 71 Post(s)
Liked: 188
Quote:
Originally Posted by Glimmie 

But that's not what is in the average living room. Consumer media products are largely hardware based. They are not capable of drastic design change via software upgrades.
It seems to me that PS3 update 1.82 is a drastic design change. No?

PS3™ System Update History
Video
. New for 1.82: Support for playback of AVC High Profile (H.264/MPEG-4) format video has been added.*
*High image quality encoding method used by Blu-ray Discs and other media.
. Super-White and x.v.Color output are now supported.
. Downscaled output of Blu-ray Discs is now supported. When 720p is selected as the video output setting of the PS3™ system, Blu-ray Discs that were recorded in 1080p or 1080i resolution can be played in 720p resolution.
8mile13 is offline  
post #44 of 51 Old 09-19-2013, 03:05 PM
AVS Special Member
 
Chronoptimist's Avatar
 
Join Date: Sep 2009
Posts: 2,557
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 203
Quote:
Originally Posted by 8mile13 View Post

It seems to me that PS3 update 1.82 is a drastic design change. No?
The PS3 is essentially a computer (albeit not one using an x86 processor) so it performs video decoding in software. Most televisions, players, and other devices have dedicated fixed-function decoding hardware in them.

Quote:
Originally Posted by Glimmie View Post

Ok bad example. But you know what I mean. It's not a simple novice proceedure like a firmware update.
Installing Windows on a Mac is - it's just like installing Windows on any other PC.
Getting OS X to work on a PC is somewhat more complex, but if you pick the right hardware it should be no different from running it on a real Mac - at least that's what I'm told. I don't really like OS X on a desktop computer, so I stick to Windows on my PCs.
Chronoptimist is offline  
post #45 of 51 Old 09-19-2013, 03:09 PM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Originally Posted by 8mile13 View Post
 
Quote:
Originally Posted by Glimmie 

But that's not what is in the average living room. Consumer media products are largely hardware based. They are not capable of drastic design change via software upgrades.
It seems to me that PS3 update 1.82 is a drastic design change. No?

PS3™ System Update History
Video
. New for 1.82: Support for playback of AVC High Profile (H.264/MPEG-4) format video has been added.*
*High image quality encoding method used by Blu-ray Discs and other media.
. Super-White and x.v.Color output are now supported.
. Downscaled output of Blu-ray Discs is now supported. When 720p is selected as the video output setting of the PS3™ system, Blu-ray Discs that were recorded in 1080p or 1080i resolution can be played in 720p resolution.

 

Sure, I would say so.  Even Sony introducing a new app on my TV that wasn't there before involves a large chunk of software coming down that embodies that provider's encryption encoding scheme and format.  That's all software making low level calls that the TV jobs off to whatever hardware assist it has.  I don't see anything stopping the TV from having it's own interpolation algorithms fixed or changed either so long as it's not entirely embodied in some non-reprogramable application specific IC.  Once you decide that you can download software, then you can have arbitrary levels of complexity show up so long as you have powerful enough hardware to support it at appropriate speeds.


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #46 of 51 Old 09-19-2013, 03:58 PM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by tgm1024 View Post

I don't see anything stopping the TV from having it's own interpolation algorithms fixed or changed either so long as it's not entirely embodied in some non-reprogramable application specific IC.  Once you decide that you can download software, then you can have arbitrary levels of complexity show up so long as you have powerful enough hardware to support it at appropriate speeds.

Somebody's got to pay for it. Yes it can be done. But not at average Best Buy prices.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #47 of 51 Old 09-20-2013, 08:27 AM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Originally Posted by Chronoptimist View Post
 
Quote:
Originally Posted by 8mile13 View Post

It seems to me that PS3 update 1.82 is a drastic design change. No?
The PS3 is essentially a computer (albeit not one using an x86 processor) so it performs video decoding in software. Most televisions, players, and other devices have dedicated fixed-function decoding hardware in them.

 

...which still might not matter the way some might think if it's not A-Z embodied in the hardware.  The existing firmware knows how to "hook" into that API (or device drivers) for that decoding hardware already.  As long as that decoding hardware isn't so specific that a firmware update can't modify it, then things are ok so long as the sum computing power is enough (as said over and over now).

 

And by the way, TV manufacturers are absolutely not so stupid as to put themselves into quite that degree of handcuffs.  It would require a service call for every decoding bug they discover.  There are certainly parts of the decoding support ICs that are applicable/useful to many algorithms.  Anything else is nuts.
 

Quote:

Originally by Chronoptimist:

 

Originally Posted by Glimmie View Post

Ok bad example. But you know what I mean. It's not a simple novice proceedure like a firmware update.
Installing Windows on a Mac is - it's just like installing Windows on any other PC.
Getting OS X to work on a PC is somewhat more complex, but if you pick the right hardware it should be no different from running it on a real Mac - at least that's what I'm told. I don't really like OS X on a desktop computer, so I stick to Windows on my PCs.

 

I think the point of the article I linked is that it's now no longer even a requirement to use another mac for this process.  Steve jobs moved away from other CPU's and adopted Intel probably at least partially for this keep-it-universal reason.  As an interesting aside, he also decided to stack is OS on top of his variant of Unix (a sort of Mach mix), which was interesting to me: he always decided surprisingly to not reinvent the wheel unnecessarily as often as people think.  The way (cough) NT did.  Note: I'm no Apple drone---and don't like OS-Xish things or iOS.

 

Quote:
Originally Posted by Glimmie View Post

Ok bad example. But you know what I mean. It's not a simple novice proceedure like a firmware update.

And I'm still not sure that a universal thing either. It may work fine for Word and Excell but I question if some high end video edit or color correction software would run on a kludged system like that.

 

Not true on the high-end software comment, and it's a bad example for a different reason.  By far, the most hardware dependent part of any software on a computer system is the operating system.  And that's by design.  Once that is in place, it will provide device drivers, and layers above those drivers, etc., for any application to use.  The very lowest that any application software would ever access are hooks into the device driver itself, which is wholly unusual and almost always disallowed because the OS and library layers above it are there for that reason---to make sure the OS is the only way to access anything in hardware.  Even the low level "open() and ioctl()" calls (if you're familiar with Most OS's, and perhaps C, C++), is an interface abstraction over a device driver, albeit often a thin one.


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
post #48 of 51 Old 09-20-2013, 10:10 AM
AVS Special Member
 
Chronoptimist's Avatar
 
Join Date: Sep 2009
Posts: 2,557
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 203
Quote:
Originally Posted by tgm1024 View Post

...which still might not matter the way some might think if it's not A-Z embodied in the hardware. The existing firmware knows how to "hook" into that API (or device drivers) for that decoding hardware already. As long as that decoding hardware isn't so specific that a firmware update can't modify it, then things are ok so long as the sum computing power is enough (as said over and over now).

And by the way, TV manufacturers are absolutely not so stupid as to put themselves into quite that degree of handcuffs. It would require a service call for every decoding bug they discover. There are certainly parts of the decoding support ICs that are applicable/useful to many algorithms. Anything else is nuts.
The demands of decoding 4K H.265 video, or even 1080p H.265 are significant when compared to 1080p H.264, just like H.264 is significantly more complex to decode than MPEG2 was. Unless the current generation of displays were designed with H.265 in mind (to a draft specification) nothing is going to be upgraded to H.265 or 4K decoding via a firmware update.

And "decoder errors" should not happen - it's a fixed spec.
Chronoptimist is offline  
post #49 of 51 Old 09-20-2013, 10:23 AM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
As for the Sony PS2,3,??? this is more a computer than a dedicated appliance. So yes, doing major upgrades to the decoding section is just software download.

However a BluRay player, DirecTV, Dish, or cable STB, uses an MPEG decoder chip. There is no API. There are hardware registers that must be setup by an embedded computer in the device but beyond turning on and off a few options, it is what it is. The core decoder function cannot be updated.

Then there is the in- between area of card based MPEG decoders. Here somebody puts a popular MPEG decoder chip, often the same chips the disk players and cable boxes use, on a PCI card. Here you have more flexibility which now includes an API but still you are largely limited to the chip set for the decoding algorithm. Most MPEG broadcast products use this approach inside an industrial PC due to the low market needs.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #50 of 51 Old 09-20-2013, 10:36 AM
AVS Special Member
 
Glimmie's Avatar
 
Join Date: Sep 2000
Location: Los Angeles, CA
Posts: 7,785
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 163
Quote:
Originally Posted by tgm1024 View Post

Not true on the high-end software comment, and it's a bad example for a different reason.  By far, the most hardware dependent part of any software on a computer system is the operating system.  And that's by design.  Once that is in place, it will provide device drivers, and layers above those drivers, etc., for any application to use.  The very lowest that any application software would ever access are hooks into the device driver itself, which is wholly unusual and almost always disallowed because the OS and library layers above it are there for that reason---to make sure the OS is the only way to access anything in hardware.  Even the low level "open() and ioctl()" calls (if you're familiar with Most OS's, and perhaps C, C++), is an interface abstraction over a device driver, albeit often a thin one.

This is the classic Windows and other higher level OS's rule. Thou shall not write to or read from hardware directly. But in my business, broadcast and film production, this is often ignored because of the speed requirements. In realtime image processing applications, the OS was often too slow for real time operation. And as these applications are usually dedicated function boxes, a "broken" OS was not really an issue.

I will say however that is largely a 1990s early 2000s decade problem. Today even commodity PCs are more than fast enough to let the OS handle hardware interaction even for uncompressed HD playout. Also the GPU has done a lot to make this possible too. But we have a couple of systems that do realtime 24 frame uncompressed 4K playout and they have a highly hacked Linux to make it all work. But that too will become obsolete as CPU and GPU speeds increase. In fact this device is over 5 years old and some newer competition uses 8-16 GPUs under Cuda to do the same thing.

Personally I still use DOS and C to write control software. There is in fact still a large industrial DOS user base for similar projects. WHY? It's fairly common to reboot Windows and even Linux PCs weekly if not daily due to problems. DOS programs can run for years with interrupted power. For simple to mid complexity control applications this is all you need.

Glimmie's HT Page
Being redone - comming soon!

Glimmie is offline  
post #51 of 51 Old 09-20-2013, 11:43 AM
AVS Special Member
 
tgm1024's Avatar
 
Join Date: Dec 2010
Posts: 5,815
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 90 Post(s)
Liked: 550
Quote:
Originally Posted by Glimmie View Post
 
Quote:
Originally Posted by tgm1024 View Post

Not true on the high-end software comment, and it's a bad example for a different reason.  By far, the most hardware dependent part of any software on a computer system is the operating system.  And that's by design.  Once that is in place, it will provide device drivers, and layers above those drivers, etc., for any application to use.  The very lowest that any application software would ever access are hooks into the device driver itself, which is wholly unusual and almost always disallowed because the OS and library layers above it are there for that reason---to make sure the OS is the only way to access anything in hardware.  Even the low level "open() and ioctl()" calls (if you're familiar with Most OS's, and perhaps C, C++), is an interface abstraction over a device driver, albeit often a thin one.

This is the classic Windows and other higher level OS's rule. Thou shall not write to or read from hardware directly. But in my business, broadcast and film production, this is often ignored because of the speed requirements. In realtime image processing applications, the OS was often too slow for real time operation. And as these applications are usually dedicated function boxes, a "broken" OS was not really an issue.

 

Sure.  I've done plenty of embedded work, much of it with DSP's <--- which by the way is always a design mistake---but that's for a different forum entirely.  One even for an aircraft display.  But I was responding to your statement about what would likely work once operating systems (in that case Windows & OS X) were ported.  Not embedded stuff.  I understand the desire for non-multitasking "mini" OS's too: DOS at least has the ability to have your application run as if the PC were in some dedicated standalone completely without memory segments and the like.   Modern OS's are dicey in that regard.  Getting a real-time OS to truly be a real-time OS  when other stuff wants to run is tricky as hell.

 

Anyway.....


Well Vinnie97, one of the kindest and most helpful and respected members here, got one of these. I wonder how much longer before I get such a message...
You have been banned for the following reason: No reason was specified.
Date the ban will be lifted: Never
tgm1024 is offline  
Reply Flat Panels General and OLED Technology

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off