"Deep Color" - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 
Thread Tools
post #1 of 30 Old 10-27-2006, 10:44 PM - Thread Starter
Advanced Member
 
Timwit's Avatar
 
Join Date: Feb 2001
Location: Summit County, CO
Posts: 564
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
From:

http://hdmi.org/press/pr/pr_20061026.asp

Quote:
“During the first half of 2007 we expect to see HDTVs with HDMI 1.3 functionality, allowing them to display Deep ColorTM content. We also expect the introduction during 2007 of HDMI 1.3 technology for PCs, audio-visual receivers and a range of other source and display devices.â€
Quote:
HDMI 1.3 more than doubles HDMI’s bandwidth and adds support for Deep Color technology, a broader color space, new digital audio formats, automatic audio/video synching capability (“lip syncâ€)...
Quote:
Silicon Image Demonstrates Deep Color
The HDMI 1.3 World Tour includes one of the first opportunities to see Deep Color. HDMI 1.3 supports 10-bit, 12-bit and 16-bit (RGB or YCbCr) color depths, up from the 8-bit depths in previous versions of the HDMI specification....

(Note: HDMI Licensing previously used an alternative naming scheme referring to 10-bit, 12-bit and 16-bit color as 30-bit, 36-bit and 48-bit color, reflecting the bit depth of all three colors (RGB or YCbCr) combined.)

Many filmmakers today digitally record and process motion pictures at greater color depths than consumer home theater equipment has been able to reproduce. Movie studios have had to reduce the color depth of their films for home distribution in order for them to play on consumer equipment. However, the advent of 10-bit digital displays and HDMI 1.3 paves the way for player devices and media that can deliver digital movie and game content in nearly lossless visual form, providing consumers with a level of visual acuity and realism never before available in their homes.

Benefits of Deep Color support include:

- Allows HDTVs and other displays go from millions of colors to billions of colors
- Eliminates on-screen color banding, for smooth tonal transitions and subtle gradations between colors
- Enables increased contrast ratio
- Can represent many times more shades between any two colors, and many times more shades of gray between black and white. At 10-bit color depth, four times more shades would be the minimum, and the typical improvement would be eight times or more.
Will Blu-Ray machines, with HDMI 1.3 from the player to the display, take full advantage of Deep Color?

Deep Color is sounding like something I should wait for. No? Especially at current Blu-Ray prices.

Opinions?______________________________

Regards
-Tim
Timwit is offline  
Sponsored Links
Advertisement
 
post #2 of 30 Old 10-27-2006, 10:48 PM
AVS Forum Addicted Member
 
bfdtv's Avatar
 
Join Date: Nov 2002
Posts: 13,531
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Will Blu-Ray machines, with HDMI 1.3 from the player to the display, take full advantage of Deep Color?
Blu-ray and HD-DVD specifications currently limit content to 8-bit, 4:2:0. Hence, no Blu-ray disk will take advantage of "deep color."

Future players could interpolate to use "deep color," but it's unclear what benefit, if any, this would have for 8-bit 4:2:0 Blu-ray and HD-DVD disks.

The only probable source of "deep color" content in the next five years is...games. If you do a lot of gaming, there may be some benefit to a "deep color" display in a few years, but otherwise, I wouldn't worry about it. Keep in mind you must have a HDMI 1.3 display that actually takes advantage of "deep color" to see any benefit from it for such future content.
bfdtv is offline  
post #3 of 30 Old 10-28-2006, 04:54 AM
AVS Forum Special Member
 
Neo1965's Avatar
 
Join Date: Jun 2006
Location: CANADA!
Posts: 3,722
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by bfdtv
Blu-ray and HD-DVD specifications currently limit content to 8-bit, 4:2:0. Hence, no Blu-ray disk will take advantage of "deep color."

Future players could interpolate to use "deep color," but it's unclear what benefit, if any, this would have for 8-bit 4:2:0 Blu-ray and HD-DVD disks.

The only probable source of "deep color" content in the next five years is...games. If you do a lot of gaming, there may be some benefit to a "deep color" display in a few years, but otherwise, I wouldn't worry about it. Keep in mind you must have a HDMI 1.3 display that actually takes advantage of "deep color" to see any benefit from it for such future content.
When you have enough pixels, I wonder if anyone really can tell 8bit/channel from 12bit? That is assuming of course the display processor itself actually has a frame buffer that would store the 36bits/pixel and keep all intermediate processing values high enough to preserve the 12bit/channel.

I doubt it, but like 96kHz or 192kHz 24/48bit audio, people will likely go for this stuff even 99.99999% of the population can't tell them apart. Such is the curse of addiction to new toys.
Neo1965 is offline  
Sponsored Links
Advertisement
 
post #4 of 30 Old 10-28-2006, 02:26 PM
AVS Forum Special Member
 
gooki's Avatar
 
Join Date: Apr 2002
Location: Queenstown, New Zealand
Posts: 3,799
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
I wonder if anyone really can tell 8bit/channel from 12bit?
I know i can't tell the difference between 8bit and 16bit/channel photographs at my work. Tested on CRT and LCD.

International HDDVD Screenshot Archive (Full 1080p Images): www.hdmovies.co.nz
gooki is offline  
post #5 of 30 Old 10-29-2006, 08:41 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
My own policy for the time being, is that as soon as some piece of literature speaks of "deep color" as a feature, my BS shields go up for whatever is about to follow in print. ;)

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #6 of 30 Old 10-30-2006, 06:53 AM
AVS Forum Addicted Member
 
bfdtv's Avatar
 
Join Date: Nov 2002
Posts: 13,531
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
It's worth noting that existing HD-DVD and BD chipsets already do their video processing at 10-bits or higher, so extra "head room" for video processing isn't really needed.
bfdtv is offline  
post #7 of 30 Old 10-30-2006, 07:51 AM
Advanced Member
 
welwynnick's Avatar
 
Join Date: Jul 2005
Location: Welwyn, Herts, UK
Posts: 958
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Quote:
Originally Posted by bfdtv
Blu-ray and HD-DVD specifications currently limit content to 8-bit, 4:2:0. Hence, no Blu-ray disk will take advantage of "deep color."
I read somewhere that while HDDVD specs do limit colour to 8 bit 4:2:0, this was not the case for Blu-ray, with it's potentiallly greater capacity. I beleive all current releases only use 8 bits/colour, but there is actually provision in the BD spec for greater depth - 10 or 12 bits, I'm not sure. I can't find the liks any more. Does anyone know any better?

Cheers, Nick
welwynnick is offline  
post #8 of 30 Old 10-30-2006, 09:11 AM
AVS Forum Addicted Member
 
bfdtv's Avatar
 
Join Date: Nov 2002
Posts: 13,531
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
I read somewhere that while HDDVD specs do limit colour to 8 bit 4:2:0, this was not the case for Blu-ray,
Both formats are the same with regard to 8-bit 4:2:0. In discussions elsewhere on this forum, it was noted that AVC supports 10-bit with the Hi10P and Hi422P profiles, and VC-1 could support more in the future with an extension, but neither of these expanded color profiles is in the HD-DVD or Blu-ray specifications at this time. Both HD-DVD and Blu-ray are identical in their video codec support.

Note if AVC Hi422P profile were added to either specification in the future, HDMI 1.1 would be sufficient to support it.
bfdtv is offline  
post #9 of 30 Old 11-03-2006, 05:23 PM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
It's possible that some players - like the Toshiba XA2, which has been announced to support "Deep Color" - may convert the colorspace from the 8 bit on the discs, ie "upsampling" to provide 12 bit output.

No different in concept, to how the Chord upsamples 16 bit CDs to 24 bit.
rdjam is offline  
post #10 of 30 Old 11-03-2006, 06:50 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
You're not suspicious in the slightest of the generation of 24-bit data from a 16-bit source?

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #11 of 30 Old 11-03-2006, 07:55 PM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
rdjam is offline  
post #12 of 30 Old 11-03-2006, 08:04 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
Can you explain how it is done, in your own words?

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #13 of 30 Old 11-03-2006, 08:08 PM
AVS Forum Special Member
 
FilmMixer's Avatar
 
Join Date: Oct 1999
Location: Los Angeles Area, CA. USA
Posts: 8,783
Mentioned: 172 Post(s)
Tagged: 1 Thread(s)
Quoted: 2383 Post(s)
Liked: 3186
Mr. Hanky and I have gone around before on this subject. He is of the opinion that 24 bit audio cannot sound any better than 16 bit audio in the home environment due to the laws of physics and accoustics...

rdjam... prepare for his response... He is very smart and is never wrong... just ask him :)
FilmMixer is offline  
post #14 of 30 Old 11-03-2006, 08:36 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
Actually, the question I posed is entirely different than the topic you bring up. Purposely, I did not bring up anything to do with the "wonders of 24-bit sound", because it is a clear detour road to off-topic land. What I DID ask, if you care to actually read my post, was if he can explain how some product can spontaneously generate 24-bit level audio from a 16-bit audio source. Is this clear?

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #15 of 30 Old 11-03-2006, 08:41 PM
AVS Forum Special Member
 
Neo1965's Avatar
 
Join Date: Jun 2006
Location: CANADA!
Posts: 3,722
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Actually, I doubt how the 8bit to 12 bit YPrPb can be done to create new information when there is no scaling involved. I think you'd have better quality if there was a way to recreate the 4:2:2 or better still 4:4:4 from the 4:2:0 colorspace.

But since they don't exist anymore, I'd say neither the 4:4:4 recovery or the 12bit upsampling of color makes too much sense if the codec is 8bit 4:2:0. Perhaps we need a new HD format using h.264 FRExt --- might be nice --- except I don't see how today's display can handle that level of color depth, but they'd have no difficulty taking the 4:4:4 space as HDMI does allow that --- which panasonic is using, even the YPrPb 4:2:2 used in the Toshiba HDMI out is not bad...
Neo1965 is offline  
post #16 of 30 Old 11-04-2006, 12:58 AM
AVS Forum Special Member
 
FilmMixer's Avatar
 
Join Date: Oct 1999
Location: Los Angeles Area, CA. USA
Posts: 8,783
Mentioned: 172 Post(s)
Tagged: 1 Thread(s)
Quoted: 2383 Post(s)
Liked: 3186
Quote:
Originally Posted by Mr. Hanky
Actually, the question I posed is entirely different than the topic you bring up. Purposely, I did not bring up anything to do with the "wonders of 24-bit sound", because it is a clear detour road to off-topic land. What I DID ask, if you care to actually read my post, was if he can explain how some product can spontaneously generate 24-bit level audio from a 16-bit audio source. Is this clear?
I can read... your point was clear... I just know where your questions usually lead when talking about 24bit audio.... You didn't request of anybody else in this thread to explain how 8 bit color could be interpolated into 10 or 12 bit 'spontaneously'..

Adding 8 LSB zeros onto a 16 bit word is simple, and it's not going to change the level... You can debate the benefits doing that forever.... I was just giving rdjam a heads up...

But every time we have a discussion about 24 bit you turn it into a conversation about level and SPL's... just because you have a dynamic range of 144db vs. 96db it doesn't translate into 'level'.....

I do have a great respect for your knowledge in the science and theory behind all of this stuff.. I really do. It greatly overshadows my more practical experience to be sure.

As they say.. those who can do, and those who can't........... :)
FilmMixer is offline  
post #17 of 30 Old 11-04-2006, 01:37 AM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
You are quite obnoxious for the professional you profess to be. I'm sorry you have been so wounded in past discussions that you have to resort to platitudes.

For the benefit of this topic, just put a sock in it if you are not going to contribute anything useful. I am no stranger to rdjam, anyway. He's a big boy and can handle himself. The question remains if he can actually explain in his own words how something works, after challenging that I must not understand how it works. THIS HAS NOTHING TO DO WITH THE VIRTUES OF 24-BIT SOUND, NOR HAVE I INTENDED TO BRING THAT TOPIC UP. Until then, you have nothing to complain about.

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #18 of 30 Old 11-04-2006, 06:30 AM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
Wow - sounds like I'm walking into a death-trap :p

Well, I'll start with the sound analogy first, since it's one that everyone is familiar with - even though it's a slightly different situation (because you are also changing the sampling rate, not just the bit-planes).

With the audio, going from 16 bit, 44 khz to (say) 24 bit, 88/96 khz can make a difference. Having more bits to work with allows finer control when creating the "tweens" - ie setting the value of the extra sample bits created by the upscaling.

In this case, the audio upsampling can be compared to good upscaling of video. If you take standard SD video and send it to a 10 foot screen, its limitations become quite apparent. But if you run the SD video through a Video Processor scaler, it can look fantastically better. Although it could not create detail that wasn't already there, it created new values from the existing data to better take advantage of the capabilities of the output device.

The audio upscaling is the same thing. Some folks have audio equipment that exceeds the limitations of 16 bit 44 khz audio and they can hear it - upscaling the audio gives what they consider cleaner and smoother sound on their system. It could be said that it didn't create additional data, as above, but it made the audio sound better on their system.

Did I avoid the death-trap-of-death? :p
rdjam is offline  
post #19 of 30 Old 11-04-2006, 01:53 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
You didn't answer the question- how does it work? What you gave is what a salesman would say to sell the product (what it may or may not do for you, if you use it). The question is- how does it create additional information, that wasn't there in the source?

Quote:
In this case, the audio upsampling can be compared to good upscaling of video. If you take standard SD video and send it to a 10 foot screen, its limitations become quite apparent. But if you run the SD video through a Video Processor scaler, it can look fantastically better.
Not really- video upscaling is more similar to audio oversampling (which has been done since the early days of CD). It is not the same as creating accuracy bits that were not there to begin with. In either case, you are not creating extra detail, but you are making what detail is present more amenable for a larger format (the former dealing with larger positional bandwidth and the latter dealing with higher time domain).

It's not a good example, at all, of transcoding 16 bit audio to 24 bit audio and magically obtaining the benefits of 24-bit sound.

Also bear in mind that the "Video Processor scaler" does what it does by changing/manipulating the signal. It's not exactly a good example for moving towards the level of greater transparency to the source, because it is effectively changing the image from its condition in the source. Certainly, it makes it "look better" on a big screen, but not more accurate to the source. It's not much different than using an EQ to make something "sound better". It most certainly can be done in certain scenarios, but it isn't exactly becoming more transparent to the source. The whole point of going from 16-bit to 24-bit is move in the direction of transparency to the source, not away from it.

I'm suspecting that wherever you get this information from, has sold you some serious snake oil. By all means, you have the perception that it looks/sounds "better", but you have gone beyond the realm of science/information theory as to how it can achieve more accuracy than the source contained in the first place.

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #20 of 30 Old 11-04-2006, 02:03 PM
AVS Forum Special Member
 
FilmMixer's Avatar
 
Join Date: Oct 1999
Location: Los Angeles Area, CA. USA
Posts: 8,783
Mentioned: 172 Post(s)
Tagged: 1 Thread(s)
Quoted: 2383 Post(s)
Liked: 3186
Quote:
Originally Posted by rdjam
Did I avoid the death-trap-of-death? :p

I would say not :cool:
FilmMixer is offline  
post #21 of 30 Old 11-04-2006, 02:26 PM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
Quote:
Originally Posted by Mr. Hanky
You didn't answer the question- how does it work? What you gave is what a salesman would say to sell the product (what it may or may not do for you, if you use it). The question is- how does it create additional information, that wasn't there in the source?
Oh well, the deadly-death-trap-of-fatal-death strikes again :p

As you know, I am not a mathematical-genius-slash-engineer, but I am pretty sure that when doing even the most basic video processing, like contrast, brightness and color adjustments, the higher the precision the data is manipulated in, then the less risk of horrible things like banding and artifacting being introduced.

This is why many video processors use 10-bit processing instead of 8-bit, even though the original source data may have been 8 bit.

I'll quote Greg Rogers from Accupel, from another thread, as he probably knows more about this than both you and I.

Quote:
Originally Posted by gregr
The HDMI interface is just an interconnect between components (video processing engines). The 12-bit interface is currently adequate to connect the results of one processing engine to another when the original consumer source is only 8-bits. The individual processing engines can use as many bits as they want for processing headroom. It is advantageous to maintain 10 bits after scaling, and arguably perhaps even 12 bits, although I don't know of any video processing chips (other than those specialized for gamma expansion as display drivers) that currently output or input more than 10 bits. If the scaler performs 4:2:2 processing the current 12-bit HDMI video interface seems more than adequate for practical consumer video applications. Even professional SDI interfaces are only 4:2:2 10-bit.

But the Deep Color pixel packing is to provide greater than 8-bit RGB 4:4:4 and YCbCr 4:4:4 video transmission. Therefore, if an external scaler converts the 8-bit 4:2:2 YCbCr video to 4:4:4 prior to scaling, and then does 4:4:4 scaling, it becomes advantageous to transmit the results of that scaling as 4:4:4 10-bit or 12 bit video if the display will accept and maintain the video in the higher bit resolution 4:4:4 format and not convert it back to 4:2:2 for internal processing. If all of that takes place, then the HDMI 1.3 spec will eventually provide an advantage over the current spec.
rdjam is offline  
post #22 of 30 Old 11-04-2006, 02:56 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
Quote:
Originally Posted by rdjam
Oh well, the deadly-death-trap-of-fatal-death strikes again :p
That should indicate to you that you were way off base to accuse me of "not understanding" and in turn being suspicious, when in reality, you have no understanding of how it works, either. You just assume it does something good when transcoding from 16 to 24 bits, because that is what you have been told. That is poor science. That is the way a salesman cons you.

The reason I ask how it works, rather than what it does for you is because when you determine the "how", it will then become clear if the "what" is actually possible, given how it works. If it isn't, then you know someone is trying to pull a fast one (or really has no idea, themself).

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #23 of 30 Old 11-04-2006, 03:00 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
Quote:
Originally Posted by rdjam
This is why many video processors use 10-bit processing instead of 8-bit, even though the original source data may have been 8 bit.
This I can agree with because at least it is a tacit admission that the signal IS being manipulated. There is no claim of fidelity to the source, nor is there a requirement, by virtue of using such processing.

The fact remains, by doing so, you are simply using the extra 2-bits as overflow or buffer for the processing (which may or may not have needed it, depending on the nature of the processing). The result may not even need 10-bits to be expressed. It was just the processing to achieve the 8-bit result that benefits from the extra accuracy (sort of like using the full accuracy of a calculator to perform a series of operations, but the end result is still rounded to the accuracy of the original data, because we know mathematically the result cannot contain more precision than the data even if the result happens to contain extra decimal places). It is NOT restoring 10-bit information that may or may not have been present in the studio master.

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #24 of 30 Old 11-05-2006, 09:39 AM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
Quote:
Originally Posted by Mr. Hanky
That should indicate to you that you were way off base to accuse me of "not understanding" and in turn being suspicious, when in reality, you have no understanding of how it works, either. You just assume it does something good when transcoding from 16 to 24 bits, because that is what you have been told. That is poor science. That is the way a salesman cons you.

The reason I ask how it works, rather than what it does for you is because when you determine the "how", it will then become clear if the "what" is actually possible, given how it works. If it isn't, then you know someone is trying to pull a fast one (or really has no idea, themself).
There's no doubt in my mind that deep color has benefits, certainly considering the fact that color and contrast get manipulated in most systems.

So I've outlined some of why these benefits exist, and yet you continue to claim that it is all nonsense?

I agree that lots of claims are made of different systems and approaches, sometimes they make sense, and sometimes they don't add up (ie Bluray). With Deep Color, there is good logic as to why it is a positive.

Sorry, I'm giving up on you, as there will be no constructive outcome here ;)
rdjam is offline  
post #25 of 30 Old 11-05-2006, 04:08 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
Quote:
Originally Posted by rdjam
So I've outlined some of why these benefits exist, and yet you continue to claim that it is all nonsense?
You failed to explain in your own words how to extract 24 bit precision data from 16 bit precision data. The same applies to getting 10-bit from 8-bit in video. The fact remains that the figurative "perpetual motion machine" remains a myth.

No one is arguing that having a bit of headroom precision is good to have when you are processing something. The fact remains that the spontaneous generation of higher precision data than the source is more than a bit dubious. You extract from the source at its native bit precision, do your processing at greater precision, then pass on the result in the native precision of the source. Anything more and you open yourself up to all sorts of spurious artifacts and noise that occurred from the processing, itself. This is a basic tenet of mathematical calculation. Perhaps you will learn this once you get to a few college level chemistry lab courses- you don't report "10" as "9.999875" even if the calculation device supports that precision, if the source data was only 1 decimal place to begin with. If you are past the college level already and still don't see the problem with your point, then you cannot be helped, I'm afraid.

"Deep color" sent over a data line will be relevant when it actually becomes encoded as such on the disc and from the source. Anything less is just parlor tricks- not genuine utilization of the increased precision.

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #26 of 30 Old 11-05-2006, 04:23 PM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
Quote:
Originally Posted by Mr. Hanky
The fact remains, by doing so, you are simply using the extra 2-bits as overflow or buffer for the processing (which may or may not have needed it, depending on the nature of the processing). The result may not even need 10-bits to be expressed. It was just the processing to achieve the 8-bit result that benefits from the extra accuracy (sort of like using the full accuracy of a calculator to perform a series of operations, but the end result is still rounded to the accuracy of the original data, because we know mathematically the result cannot contain more precision than the data even if the result happens to contain extra decimal places). It is NOT restoring 10-bit information that may or may not have been present in the studio master.
By the way, you are just plain wrong here, and it indicates the lack of understanding that I mentioned above.

The "extra 2 bits", as you call them, are not used as "overflow or buffer for the processing" - in fact, it's hard for anyone to even understand what you could mean by that.

In fact, using 10 bits allows the Video Processor to select from a far higher number of finer increments for pixel values when making adjustments to the picture, even if the original source data is 8 bit. When changing things like gamma, brightness, contrast and color, some pixel changes will be large, and some very very minute. Having to make bigger jumps than necessary is what causes the quantizing effects like banding.

The VP can select from over a billion values at 10-bit, and over 68 billion at 12 bit.

This is compared with only 16 million values at 8 bit.

One can argue that the human eye cannot see the difference between the 16 million colors we already get at 8 bit - but the simple truth is that we can very easily see the banding artifacts caused by the quantization when image manipulation goes wrong at 8 bit.

Using 10 bit maipulation is very positive, and maintaining a minimum of 10 bits throughout the chain is a must for the new home theater that I'm currently spec'ing.

No one has suggested that any box will create "new" 10 or 12 bit values out of thin air - that was merely your attempt to re-direct the "issue" at the beginning of the discussion.

Sorry to disappoint you by sticking to what I have said.

That having been said, I will note that if you are outputting 720 from one of these new Deep Color HD players, the player could easily create 10 or 12 bit data from the original 1080p 8 bit data, as it will be combining data from the extra pixels during the conversion.
rdjam is offline  
post #27 of 30 Old 11-05-2006, 04:36 PM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
As a further note to my above post, I will note that the new Toshiba XA2 has an on-board 12-bit video processor, which allows you to make your calibration settings in the player.

This speaks exactly to the situation I have described above.

Quote:
Originally Posted by Toshiba
The new high end Toshiba HD-XA2 HD DVD player is designed to output 1920 x 1080p, the highest HD signal currently available, via High Definition Multimedia Interface (HDMI™) 1.3. HDMI version 1.3 is the most current, major enhancement of the HDMI specification, which adds support for Deep Color technology. The HD-XA2 also incorporates a 297MHz / 12 bit Video DAC with high-quality, 4x oversampling for increased bandwidth for true playback of an HD picture to a video source. The HD-XA2 comes with a picture setting function allowing customers to optimize picture quality with user adjustable settings for color, contrast, brightness, edge enhancement and block noise, among others. The new HD-XA2 HD DVD player’s 1080p output capability complements Toshiba’s new Cinema Series® Pro LCD TVs which provide true HD 1080p input capability via HDMI. The HD-XA2 will be one of the first next generation players to support the video portion of HDMI version 1.3. HDMI 1.3 supports the demands of future high definition display devices, such as Deep Color enabling 36-bit (RGB or YCbCr) color depths, up from the 24-bit depths in previous versions of the HDMI specification. This allows Deep Color compatible display devices to deliver outstanding video quality.
Another situation where the Deep Color 12 bit color depth has the advantage is during color conversion from YCbCr to RGB, etc. This is where MANY problems occur today at only 8 bits.
rdjam is offline  
post #28 of 30 Old 11-05-2006, 04:42 PM
AVS Forum Special Member
 
Mr. Hanky's Avatar
 
Join Date: Mar 2006
Posts: 6,234
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 124 Post(s)
Liked: 99
Quote:
Originally Posted by rdjam
By the way, you are just plain wrong here, and it indicates the lack of understanding that I mentioned above.

The "extra 2 bits", as you call them, are not used as "overflow or buffer for the processing" - in fact, it's hard for anyone to even understand what you could mean by that.
Yes, I mean the cumulative effect of 10-bits. By referring to "2-bits", I was inferring extra precision. This changes nothing described earlier.

Quote:
One can argue that the human eye cannot see the difference between the 16 million colors we already get at 8 bit - but the simple truth is that we can very easily see the banding artifacts caused by the quantization when image manipulation goes wrong at 8 bit.
If you can see the banding, it is because the data compression is giving you less than 8-bit performance, in the first place. Did you think of that? If you compress things enough you might be lucky to get "4-bits of precision", even though it is still considered "8-bit video". That is the whole essence of data compression- to truncate data that it thinks you wouldn't notice otherwise.

Quote:
Using 10 bit maipulation is very positive, and maintaining a minimum of 10 bits throughout the chain is a must for the new home theater that I'm currently spec'ing.
Within processing domain, yes. Across a data line where the source is 8-bit, not so much.

Quote:
No one has suggested that any box will create "new" 10 or 12 bit values out of thin air - that was merely your attempt to re-direct the "issue" at the beginning of the discussion.
Actually you were. You are attempting to slide away from that fallacious stance now.

Quote:
That having been said, I will note that if you are outputting 720 from one of these new Deep Color HD players, the player could easily create 10 or 12 bit data from the original 1080p 8 bit data, as it will be combining data from the extra pixels during the conversion.
More funky blending concepts? That's all we need. [sarcasm] How about just retain the fidelity to the source and demand premium content in the source? A bit of processing is probably ok, but relying on it to generate "a billion extra colors" that were not there in the first place is really heading in a bad direction. At that point, you aren't viewing "deep color", but doing some severe hack'n'slashing on some material that isn't even up to 8-bit standards, just to cover up the fact. That's why I refer to it as "parlor tricks".

Margot Robbie is THE most exquisite human creature on the face of the planet, in this moment in time.
Mr. Hanky is offline  
post #29 of 30 Old 11-05-2006, 07:07 PM
AVS Forum Club Gold
 
rdjam's Avatar
 
Join Date: Mar 2005
Location: Miami, FL
Posts: 9,847
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 18
Quote:
Originally Posted by Mr. Hanky
Yes, I mean the cumulative effect of 10-bits. By referring to "2-bits", I was inferring extra precision. This changes nothing described earlier.
"Yes" you realize you were incorrect? Or "Yes" you'd like to try and restate it?


Quote:
If you can see the banding, it is because the data compression is giving you less than 8-bit performance, in the first place. Did you think of that?
Wrong again - while encoding with less bit planes can also cause banding that has nothing to do with the discussion here regarding bluray or HD DVD.

Quote:
Within processing domain, yes. Across a data line where the source is 8-bit, not so much.
I'm afraid you still don't understand what is being discussed here.


Quote:
Actually you were. You are attempting to slide away from that fallacious stance now.
Actually, I wasn't - which anyone can see by reading the start of our "discussion".


Quote:
More funky blending concepts? That's all we need. [sarcasm] How about just retain the fidelity to the source and demand premium content in the source? A bit of processing is probably ok, but relying on it to generate "a billion extra colors" that were not there in the first place is really heading in a bad direction. At that point, you aren't viewing "deep color", but doing some severe hack'n'slashing on some material that isn't even up to 8-bit standards, just to cover up the fact. That's why I refer to it as "parlor tricks".
I don't think you even WANT to understand what is being discussed here.

If you've ever used Photoshop to scale images, the first trick is to change the bitplanes of the image to be scaled down to the maximum number. That way when it is scaled down, the image retains the maximum data by having the derivative values calculated with the higher accuracy.

If you don't understand any of what has been discussed and illustrated here, then you were given too much credit by filmmixer, above.

I've been quite tolerant of your gap in understanding but can't be bothered to put up with further slurs from you. This conversation will only get more absurd at this rate, so I'm out.

Cheers.
rdjam is offline  
post #30 of 30 Old 11-05-2006, 07:13 PM
AVS Forum Club Gold
 
kevinca1's Avatar
 
Join Date: Jul 2002
Location: california
Posts: 3,493
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
This is going no where
kevinca1 is offline  
Sponsored Links
Advertisement
 
Closed Thread Blu-ray Players

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