My previous articles in this series have mentioned a few factors by which the 3D implementation effort is different from HD a decade ago, rather than similar, as some industry experts have expressed.
In this fourth installment, I analyze how consumers could embrace the 3D effort, and I provide some ideas to help them evaluate the adoption of 3D for their particular application: Either as a replacement of HDTV, or as a value added feature for 3D content, whether it is sourced from satellite, cable, or Blu-ray.
[url=http://www.hdtvmagazine.com/articles/2010/02/3d_tv_at_ces_2010_was_it_actually_like_hd_a_decade_ago_part_4.php]Read Article[/url]
3D TV at CES 2010 - Was it Actually Like HD a Decade Ago? (Part 4)
-
Rodolfo
- Author
- Posts: 755
- Joined: Wed Sep 01, 2004 8:46 pm
- Location: Lansdowne VA
-
jerfilm
- Major Contributor
- Posts: 82
- Joined: Fri Oct 29, 2004 12:46 pm
How about using DVRs?
Understand I'm non technical, but as you were discussing the issues and mentioned that BluRay will be able to do the necessary bandwidth for full res 3D (I did get that right, didn't I?) it occurred to me that if we are talking about occasional 3D content, how about constructing a DVR that can take two HD broadcasts done separately, one for each channel, put them together and then show them on your TV? Of course broadcasters may not want to spend 6 hours of air time to broaccast a 3 hour movie? Or perhaps it could be done on two separate channels at the same time and then combined......
Jerry
Jerry
-
Rodolfo
- Author
- Posts: 755
- Joined: Wed Sep 01, 2004 8:46 pm
- Location: Lansdowne VA
Jerry,
Actually broadcasters are facing the possibility of losing their current bandwidth so I do not believe two channels for the two eyes in real time would be a possibility:
http://www.hdtvmagazine.com/articles/20 ... asting.php
Considering the constrains many broadcasters will face under the pressure of the FCC to maximize whatever bandwidth is left, I do not believe your proposal can happen, including the time delay style you suggest, doubling up the hrs for a DVR to store for a later viewing, putting both images together in the STB or the 3DTV.
Here is an article published today that confirms the risk I anticipated on my article above:
http://voices.washingtonpost.com/postte ... s=posttech
As I said the frame compatible format has been preferred as general distribution method (including satellite and cable) and that loses half of the 3D resolution, but even then I believe delivering 3D content would be low in the list to the problem broadcasters are facing.
However, satellite and cable have the power to assign extra bandwidth for a quality 3D VOD business model if the price is right for the subscriber and the service provider, and I suggested that to DirecTV and Bell Canada at CES, but that is not happening either.
Additionally, new STBs would have to be produced to handle the dual image in full resolution if that ever happens, as opposed to just apply a firmware upgrade to existing STBs to handle the 3D protocol of the frame-compatible format, as it is happening with satellite and cable, which uses about the same bandwidth of an HD channel (because it loses half of the resolution to serve the two eyes with the same number of available pixels per frame)
Best Regards,
Rodolfo La Maestra
Actually broadcasters are facing the possibility of losing their current bandwidth so I do not believe two channels for the two eyes in real time would be a possibility:
http://www.hdtvmagazine.com/articles/20 ... asting.php
Considering the constrains many broadcasters will face under the pressure of the FCC to maximize whatever bandwidth is left, I do not believe your proposal can happen, including the time delay style you suggest, doubling up the hrs for a DVR to store for a later viewing, putting both images together in the STB or the 3DTV.
Here is an article published today that confirms the risk I anticipated on my article above:
http://voices.washingtonpost.com/postte ... s=posttech
As I said the frame compatible format has been preferred as general distribution method (including satellite and cable) and that loses half of the 3D resolution, but even then I believe delivering 3D content would be low in the list to the problem broadcasters are facing.
However, satellite and cable have the power to assign extra bandwidth for a quality 3D VOD business model if the price is right for the subscriber and the service provider, and I suggested that to DirecTV and Bell Canada at CES, but that is not happening either.
Additionally, new STBs would have to be produced to handle the dual image in full resolution if that ever happens, as opposed to just apply a firmware upgrade to existing STBs to handle the 3D protocol of the frame-compatible format, as it is happening with satellite and cable, which uses about the same bandwidth of an HD channel (because it loses half of the resolution to serve the two eyes with the same number of available pixels per frame)
Best Regards,
Rodolfo La Maestra
-
videograbber
- Major Contributor
- Posts: 146
- Joined: Fri Sep 24, 2004 7:10 am
Since it's a given that any delivery mechanism other than Blu-ray will be at half-res...
One thing I find puzzling is that anyone is considering using the over-under encoding at all. I thought it was pretty well understood that visually, humans are more sensitive to resolution vertically than we are horizontally? So cutting 1080 in half, to 540, makes far less sense to me than a side-by-side encoding, with 1920 cut to 960.
I.e., 960x1080 to each eye seems far more reasonable than 1920x540 to each eye. This is especially true since very little HD content really has the the full 1920 horizontal resolution to start with (ATSC requires filtering that limits it to 1770 max), and thus would suffer less from compressing what IS there (e.g. from HDCams at 1440) down to 960. I.e., less than a 50% net reduction.
So is there something desirable about over-under that I'm missing, because I don't see the appeal.
- Tim
P.S. No discussion of the pros and cons of the checkerboard method?
One thing I find puzzling is that anyone is considering using the over-under encoding at all. I thought it was pretty well understood that visually, humans are more sensitive to resolution vertically than we are horizontally? So cutting 1080 in half, to 540, makes far less sense to me than a side-by-side encoding, with 1920 cut to 960.
I.e., 960x1080 to each eye seems far more reasonable than 1920x540 to each eye. This is especially true since very little HD content really has the the full 1920 horizontal resolution to start with (ATSC requires filtering that limits it to 1770 max), and thus would suffer less from compressing what IS there (e.g. from HDCams at 1440) down to 960. I.e., less than a 50% net reduction.
So is there something desirable about over-under that I'm missing, because I don't see the appeal.
- Tim
P.S. No discussion of the pros and cons of the checkerboard method?
-
Rodolfo
- Author
- Posts: 755
- Joined: Wed Sep 01, 2004 8:46 pm
- Location: Lansdowne VA
Thanks for bringing this up.
Your question maybe enough material for at least two articles, and I will keep it in mind. However I covered some of the subject of formats on the article about 3D from the HD World Conference in NY in Nov 2009, linked form part 1 I believe.
At my meetings there were several discussions of the methods, who prefers what and why, but frankly it ended up as a matter of choice when considering the pro/cons of each, and not even HDMI LLC anticipated the possible choice when they released 1.4 without the format because the market was still undecided, Steve Venuti, HDMI president said.
Now that frame-compatible format will be put into 1.4a. But it still a manner of adding the 3D structure protocol to a chip/device that was designed for that even without 1.4, all depends how the manufacturers implemented the device/chip.
Regarding checkerboard, on TIs site there is a document that highlights how that works. I suggest paying attention to their use of language about "full resolution" and "sampling" which is a contradiction: one could include all the pixels or one could sample pixels to choose some from each image, but not both.
As you may know Samsung and Mitsubishi are doing this for a few years on their RPTVs using the "wobulated chip" which has made itself famous by the way it uses the mirrors to make your eyes believe they are seeing an image created at once by a full resolution chip when is actually having half of the mirrors on the DMD and is using speed to complete the pixels of an image.
I think I commented about my viewing experience, maybe on a thread. I never liked the technique, nor did I like the 3D demo at CES with a 3D Blu-ray player/w/adapter. Perhaps some of the "dislike" is sourced in knowing that is not using a full chip like front projectors.
There are a lot of Mitsu fans that would defend this technique, and that is fine; I have very good relationship with their execs, especially David Naranjo, but am not a fan of the approach since the "wobulated" chip was introduced for HD several years ago, rather than using the full chip, allegedly because manufacturers preferred to keep the cost down on their RPTVs.
Best Regards,
Rodolfo La Maestra
Your question maybe enough material for at least two articles, and I will keep it in mind. However I covered some of the subject of formats on the article about 3D from the HD World Conference in NY in Nov 2009, linked form part 1 I believe.
At my meetings there were several discussions of the methods, who prefers what and why, but frankly it ended up as a matter of choice when considering the pro/cons of each, and not even HDMI LLC anticipated the possible choice when they released 1.4 without the format because the market was still undecided, Steve Venuti, HDMI president said.
Now that frame-compatible format will be put into 1.4a. But it still a manner of adding the 3D structure protocol to a chip/device that was designed for that even without 1.4, all depends how the manufacturers implemented the device/chip.
Regarding checkerboard, on TIs site there is a document that highlights how that works. I suggest paying attention to their use of language about "full resolution" and "sampling" which is a contradiction: one could include all the pixels or one could sample pixels to choose some from each image, but not both.
As you may know Samsung and Mitsubishi are doing this for a few years on their RPTVs using the "wobulated chip" which has made itself famous by the way it uses the mirrors to make your eyes believe they are seeing an image created at once by a full resolution chip when is actually having half of the mirrors on the DMD and is using speed to complete the pixels of an image.
I think I commented about my viewing experience, maybe on a thread. I never liked the technique, nor did I like the 3D demo at CES with a 3D Blu-ray player/w/adapter. Perhaps some of the "dislike" is sourced in knowing that is not using a full chip like front projectors.
There are a lot of Mitsu fans that would defend this technique, and that is fine; I have very good relationship with their execs, especially David Naranjo, but am not a fan of the approach since the "wobulated" chip was introduced for HD several years ago, rather than using the full chip, allegedly because manufacturers preferred to keep the cost down on their RPTVs.
Best Regards,
Rodolfo La Maestra