Log in

View Full Version : Better than DVD?


dragoman
24th March 2002, 10:32
Hi,

I begin this topic kind of open-ended. I was thinking about DVD, which uses MPEG-2 compression, and the fact that most people use it as the de-facto ideal in video quality. (Notice I said most people, to those who bristle at any mention of an all-inclusive idea)

Anyway, since I was drunk at the time, I'm kind of surprised I remembered it. But I did, and here is what I have found out:

DVD (Digital Versatile Disc, or Digital Video Disc as most people call it....officially it doesn't stand for anything :) ) uses MPEG-2 video compression, as most of you know.

First, some background on DVD:

Physcial specifications of a DVD (http://www.mpeg.org/MPEG/DVD/Book_A/Specs.html)
Very technical explanation of MPEG-2 workings (http://mpeg.telecomitalialab.com/standards/mpeg-2/mpeg-2.htm)

The question I am ultimately trying to answer is "Why is MPEG-2 regarded as the standard for video" and also "Is there anything better?"

Question 1: Why is MPEG-2 the standard?

From my research into the matter, MPEG-2 is a lossy compression method (like all MPEG formats) that uses a bitrate between 3.5 and 5 MB/sec. The DVD FAQ (http://www.thedigitalbits.com/officialfaq.html#1.3) states that better quality is achieved over 6 MB/sec (makes sense, it is a lossy format).

It seems to me that the reason that DVD has become the standard is a combination of three things: 1) MPEG-2 is the best way to compress the master video into a easily watchable format, 2) DVD provides a way to pack massive amounts of information (ie high bitrate video) into a package that is easily mass-produced, and 3) DVD offered a way to easily copy-protect the content, preventing (or at least reducing) movie piracy.

As it says on the DVD faq, video quality (especially video compressed using a lossy format such as MPEG) is subjective....one person's idea of good quality may not be another's. I base my assumption then that the decision to use the MPEG-2 format for DVD was based upon economic reasons rather than technical ones (this conclusion is also supported by the behavior of the MPAA in recent months).

One should also note that copy-protection was integrated into DVD from the start.

1) CSS (Content Scrambling System) prevents the data on the disc from being copied onto a computer's hard drive, unless the proper key was used to unlock the data. As we all know, this feature has been largely by-passed by now.

This feature was by-passed first by Linux users who wished to view DVD's on their computer system (something completely legal, btw) but were prevented by the fact that no software was yet-written to play DVD's on the Linux operating system. The Linux users by-passed the CSS encryption by developing De-CSS, which allowed them to copy the MPEG-2 files to their hard drives and thus watch movies. The resulting program, De-CSS was posted and distributed by 2600 Magazine, the mislabled (by the press) "hacker" website which has since stopped hosting the software (by federal decree). The MPAA has waged constant legal battles against anyone distributing this software on the internet (although we all know it's a moot point now).

Sidenote - How was CSS cracked?
The CSS system has a master list of 400 "keys" that are included on every CSS-encoded disc. The drive unit of a player is assigned it's own CSS decryption key, which must be licensed. This key is one of the 400 master keys, and is used to unlock the encryption on a disc which allows the player to decrypt and play the video. A "DVD-ripper" essentially does just that - it accesses the CSS keys on the disc and applies every one until the correct key is found, thus allowing the video to be copied onto the user's hard drive.

2) Another measure of copy protection was dubbed Macrovision. The correct term is APS, or "Analog Protection System", which is basically a circuit included in every DVD player ever sold. Macrovision adds a rapidly modulated colorburst signal ("Colorstripe") along with pulses in the vertical blanking signal ("AGC") to the composite video and s-video outputs (also component video). Macrovision may show up as stripes of color, distortion, rolling, black & white picture, and dark/light cycling. What this means is that it is virtually impossible to record a dvd onto a vhs tape without severe distortion occurring in the resulting video (using standard equipment).

The inclusion of this protection system was obviously intended to prevent the copying of dvd media onto vhs tapes (the so-called "Blockbuster pirates" - referring to those people who rent movies to tape them), and has been largely successful. Several computer programs have been developed to remove the copy protection from the digital video, and several standalone dvd players have appeared on the market with de-macrovision features (the most notable being the APEX 600A, which was later recalled by Apex Digital Inc. after protests from the MPAA were published). (Sidenote - the Apex 600A is no longer on the market, but future generations Apex players have the same function - my AD3210 has this feature to turn off the Macrovision).

3) Region codes. Region codes are not considered a form of copy protection by the MPAA, but the effects of this system cannot be considered anything else. One byte of data on a disc is given a certain value, one which corresponds to the values assigned DVD players sold in different regions around the world. The different region codes are as follows:

1: U.S., Canada, U.S. Territories
2: Japan, Europe, South Africa, and Middle East (including Egypt)
3: Southeast Asia and East Asia (including Hong Kong)
4: Australia, New Zealand, Pacific Islands, Central America, Mexico, South America, and the Caribbean
5: Eastern Europe (Former Soviet Union), Indian subcontinent, Africa, North Korea, and Mongolia
6: China
7: Reserved
8: Special international venues (airplanes, cruise ships, etc.)

A DVD player with a specific region code will play only DVD's with that region code. Software dvd players often offer the chance to change the region code up to a certain number of times, after which the last selected code becomes permanent. Hardware DVD players were region specific, until very recently. The aformentions Apex AD600A was a "multi-region" DVD player, which would play any region dvd (Sidenote - My Apex AD3210 plays all regions as well....to access this feature a hidden menu must be found).

The inclusion of the region code into the disc is entirely discretionary by the disc's maker, and is not required by law. However, most movie studios do so (after all, why not? They can charge higher prices in the US for Region 1 DVD's than they do in China for Region 6 dvd's, and vice versa...more money).

Content encoded without a region-code on the disc can be played on all players, regardless of region designation. However, these discs are hard to find and are normally made by independant film companies or are privately-generated.

*I know this seems to have become a dvd-piracy lecture, but bear with me....I prefer to be thorough)

One must understand that this point that the first four copy protection mechanisms explained here are not required by law to be used with DVD-discs.

Five copy-protection systems currently in development are mentioned below, but as they are not currently used they are not part of this discussion. Full explanation of them can be found at the DvD Faq (http://www.thedigitalbits.com/officialfaq.html#1.3).

These systems are:
Copy Generation Management System (CGMS)
Content Protection for Prerecorded Media (CPPM)
Content Protection for Recordable Media (CPRM)
Digital Copy Protection System (DCPS)
High-Bandwidth Digital Content Protection (HDCP)

This proves my third point above, that a reason for DVD to be used as a standard is the ability to copy-protect it's contents.


To move on to the second question, and the reason for this post:

"Is there anything better?"

Unfortunately, at the time a true lossless video compression technology does not exist in a consumer-applicable format (any developers reading this please correct me!).

Several formats that claim to have achieved lossless video compression include:

Autosophy compression (http://www.autosophy.com/videcomp.htm) - uses a "redefintion of information, based on object size and motion" - A few more technical details are on the site, but basically this is a stream-optimized video technology that only transmits information about things in a frame that have changed from one frame to the next. (Very similar to MPEG-4 video compression's use of keyframes). Promises to reduce bandwidth tremendously while preserving and even increasing quality of streaming video. Very interesting.

Matrox (http://www.matrox.com/videoweb/products/enduser/background/pdf/mathematically_lossless.pdf) has developed a "mathematically lossless compression" that delivers "better quality than uncompressed". Hard to believe, but the details are there. The term for the process is known as "entropy" encoding, which leaves the digital information unaffected, yet manages to compress the video at a rate of 1.6/1. (Note - requires Adobe acrobat reader)



There are many other "lossless" video compression formats out there (go to Google, you'll see for yourself), far too many to list or explain here.

But back to the original question. Is there something better than DVD?

Quality-wise, there must be. By defintion, MPEG-2 is a lossy format, which means that the video can be corrupted by the encoding process, and that data is lost. MPEG-2 can never (read, NEVER) be as good as the original master video.

A lossless compression system would have better results quality-wise than a DVD. However, such methods would probably result in a filesize greater than that of DVD, which requires more space to hold that information.

Advances in data storage are occurring at a rate that makes it conceivable that such massive amounts of data could be stored in a format easily marketed and used by the general public. That being said, it is conceivable that a lossless video format could replace DVD as the video "standard" for consumers in the future.

But I seriously doubt that this will happen in the near future. DVD is becoming entrenched in the consumer world, like CD did before it. After 22 years, CD-audio is still the standard audio format being sold today (although the Red-Book Audio Standard is far behind current audio formats in quality). DVD-Audio and Super Audio CD are two formats becoming more mainstream, but are not yet massively available.

DVD will be the same way, I predict. Certain video technologies will be created which are better in quality than DVD, but the DVD will remain the standard for years (and perhaps, decades) to come.

So the people out there who are constantly searching for better quality (I am one of them) will not be disappointed, but for the forseeable future we will continue to use DVD's as the source material for most of our encoding efforts, at least until the master source video for a movie is released to the public....

(wrapping up a very long post....whew!)

I invide comments, but please don't flame me....

dragoman

hyc
24th March 2002, 11:20
What was the point of this topic?

Media that's better than DVD should start coming out next year:

http://www.pcworld.com/news/article/0,aid,78324,00.asp
http://www.pcworld.com/news/article/0,aid,31062,00.asp
http://sharp-world.com/corporate/news/000323.html

This blue laser stuff has been in the labs for a couple years already and apparently the DVD industry is already signed on to it.

30GB per layer, more than 6 times the capacity of today's DVD. They're talking about encoding HDTV programs, which would just fit, using the same lossy compression algorithm that DVD uses today.

Lossless compression typically cannot do better than 2:1 on real-world data (on average). Lessee, 720x480x16 bits per pixel, 30 frames per second gives 20,736,000 bytes per second uncompressed. Let's say 20MB/sec. 30GB/20MB = 1500 seconds, still only 25 minutes uncompressed. With lossless entropy coding, maybe 50 minutes, the actual compression factor depends entirely on the source material.

So, even with these wonderful new discs that can hold 60GB, you can't encode a full length movie without lossy compression. Maybe DV encoding is good enough, 5:1 lossy compression. That lets you fit a 2 hour movie on a single layer at 30GB, not bad.

dragoman
24th March 2002, 11:37
Hi,

When you say "media that is better than dvd" I assume you are talking about storage space. Yes, media that can hold more info is coming out. The development of this high-capacity media is important to the development of a lossless video format and it's use by the public, for obvious reasons.

However, you cannot make any determinations as yet to the size of a "lossless" video technology that doesn't even exist yet and even in my post was only conjecture. The examples I provided were to illustrate the fact that there are current attempts to create a lossless video format.

The main point of my post was the issue of video quality, and what it would take for that new video format to become standard. I attempted to explain the process by showing the process dvd traveled to become the standard.

Sorry if any confusion resulted....

dragoman

hyc
24th March 2002, 11:56
I was talking about lossless data compression algorithms, like Lempel-Ziv and Huffman coding, that are already used today. The fact is that you can already do reasonable lossless encoding on video if you treat each signal channel as an independent stream. (If in RGB format, then R, G, and B must be compressed as 3 separate streams. Likewise for YUV.) And, on average, today's lossless algorithms will get 2:1 on complex imagery. Actually you can do better than that for typical broadcast-quality video, since it is produced in YUV color space and the UV resolution is 1/2 the Y resolution. Maybe 4:1 on average.

By the way, if you've ever seen a file archiver/format named ARC, you've probably seen my name before. I was the guy who discovered that so-called "uncompressible" files - files that were already optimally compressed using LZW - could be further reduced by applying Huffman coding to the LZW data. I know a thing or two about lossless encoding, in any context, not just data or video.

hyc
24th March 2002, 12:01
(the big problem with any lossless method is that its final compression ratio depends entirely on the specifics of the source data, so you cannot consistently assign a number to its efficiency. You cannot say conclusively "this 650MB disc can hold 75 minutes of lossless-compressed video" because you just can't know for sure, it may hold only 32 seconds of detailed, fast action video. It may hold 20 hours of still test-pattern. It all depends on the actual images.)

-h
24th March 2002, 12:03
I think 4:1 is entirely possible, probably more. Once I'm done adding interlacing to XviD, I'm planning on forking it into a lossless non-MPEG4-related codec (why put all that nice code to waste?). Motion estimation should bring ratios far better than huffyuv.

Not sure what I'll do about speed.. I doubt it'll be fast enough for capturing on most people's PCs, though it'll be faster than XviD currently is, due to no fDCT/quant/dequant/iDCT overhead. Probably just use PNG for the residual image data.

-h

dragoman
24th March 2002, 12:26
Hi,

Something that sparked my curiousity (my thanks to those who know far more about the subject than me explaining things)...

How do we know how massive (that is, how much space does it take up, digitally) lossless video is?

For that matter, how can it be determined how good a compression ratio can be achieved using "lossless" video? Against what are you comparing to when you say video is compressed at 4:1. What format is the "1" in?

dragoman

hyc
24th March 2002, 12:45
In the US, broadcast video is produced in 720x480, 16 bits per pixel. That is the standard used in all professional production houses, all TV studios, everything. The pixels are in YUV format with 8 bits for Y and 4 bits each for U and V. One frame of video is thus 691200 bytes. There's some additional overhead injected into the output analog stream to provide horizontal and vertical blanking, control signals, etc. but the digital image is a 720x480 frame. Professional production is usually done on streaming tape, which can typically hold several hours of raw uncompressed video...
As I recall, PAL base resolution is 720x576. Note that 720x480 is actually a 3:2 ratio, but TV doesn't use square pixels. The 4:3 aspect ratio quoted for television is correct based on physical dimensions, not on pixel count.

So that's your starting point, the "1" in "4:1". I already went thru the rest of the math above - 30 frames per second, so 30 * 691200 bytes gives 20,736,000 bytes per second. Multiply again by 3600 to get bytes per hour, etc. etc... For max rez High Def TV, use 1920x1080 resolution, but still 16 bits per pixel.

Also my 4:1 prediction was only based on 2-dimensional processing. If you throw motion estimation in there/3-D processing, you should certainly be able to do even better.

Interlacing is murder on compression ratios though. It destroys the natural sequence of the data stream, so you can't take as much advantage of similarity of neighboring pixels. It makes motion estimation much more difficult. To do it well you need to be able to process outside of realtime, so you can scan forwards and backwards from a particular frame. You can do this in realtime if your algorithm is fast enough, so that you can buffer a couple frames and manage it that way, but it's still tough.

-h
24th March 2002, 13:00
Interlacing is murder on compression ratios though. It destroys the natural sequence of the data stream, so you can't take as much advantage of similarity of neighboring pixels. It makes motion estimation much more difficult. To do it well you need to be able to process outside of realtime, so you can scan forwards and backwards from a particular frame. You can do this in realtime if your algorithm is fast enough, so that you can buffer a couple frames and manage it that way, but it's still tough.

Interlacing isn't too bad - MPEG4 has a motion estimation mode especially for interlaced frames - you can transmit two motion vectors, one for the top field and the other for the bottom, to find better temporal matches for artifacted areas. Also, the DCT can be applied to field-based blocks instead of frame-based to significantly improve compression.

And it can all be switched on/off on a per-macroblock basis.. fun fun.

-h

dragoman
24th March 2002, 13:11
Wow, thanks, I 'loined' something tonight....

So what you are telling me here is that this is the standard against which all video is measured? That there cannot be an improvement beyond this format?

Is YUV the best method to display digital video? Is there any work in progress to form something better?

dragoman

BTW, thank you for you answers so far...

ps. Sounds like it takes an enormous amount of computing power to play this.....what is the hardware requirements for this (just curious)?

Scuba
24th March 2002, 13:50
There are lot's of factors that come into all of this.

> For example the maximum quality used in brodcast when editing, not the final brodcasted signal is 32Bit uncompressd.
this is used when making effects in programs such as Adobe After Effects, Pinnacle Commotion and Discreet combustion.
> If you will look at the file size you get :
PAL > 720px X 576px X 32bit X 25FPS = 42MB/Sec
NTSC is the same. it's only 720X480 but 30FPS.

> I have such a file of 4min video and it's about 9gig in size.
Now I compressd the file using TmpgEnc using very high quality settings with average of 8Mbit to be DVD compatible but also I gave room to play with up to 10Mbit spicks Max.
> The file came out less then 240MB, and the quality is very good when comparing to the original.

It's important to explain that Mpg2 is not a compression standard. It's a DeCompression standard.

Saying that, I can write any compression algoritms that I like, and use what ever means and time that I have in order to compress the video, and as long as it confront to the Mpg2 DeCompression standard it supposed to work.

Why film look so much better on a DVD then home video ?
They have armies of geeks that can sit for day's looking at the video and optemise the compression.
While we are talking about using a faster CPU to compress the video faster, they are sitting and manually setting the location of the 'I' 'B' and 'P' frames.
Have any body here triad to do it with TmpgEnc :rolleyes:

Finally is a small equasion saying you can't get loosless compression Antrophosy wise because it will take infinit time to calculate.
So you must make compromises. Little less quality, but it will only tale 30Min to encode 1Min of video on my 1Ghz Athlon and I will be able to one day author my video onto a DVD and not in the end of times.

Are there tecnolagies in the market that can provide "better" quality then Mpg2 ? Probably.
Are this technolagies backward compatible with mpg2 ? Probably not !

The reason they used Mpg2 in the first place was deu lot's of reasons.
How complicated is to create the video ?
How complicated is to uncompress it ?
what is the final file size ?

> You can ask "Why mpg1" exactly like you can ask "Why mpg2"
for the time it was good and acheaved the needs.
In Mpg1 it was video on slow CD-Rom's. In Mpg2 there was more reasons.
MII and Beta SX are also using Mpg2 compression of sort's.
> Why didn't we started at Dvix 4 ? Why did we have DviX, then DviX 3 and 4 come only now ?
It takes time to develop a codec, and each and every coded was designd to acheave a goal.

I just hope you can undeststand my "Muxed" Ideas and writing :o

ulfschack
26th March 2002, 16:21
Do you know for sure that there's an army of geeks sitting tailoring each GOP? If so, in many masterings I've seen these soldiers were in dire need of thicker glasses (albeit with the piece of tape in the middle :) ).

I think that for most productions they're content with saying "Now on DVD ... bringing the audience closer ...bla" and shit like that with a booming voice, running it through any arbitrary encoder as soon as possible to get more money as soon as possible.

The biggest difference is due to the equipment used while aquiring the source. Compare this to a shaky, hand held, wildly panning DV camera at a party with nice punch and I think the prerequisites will speak for themselves.

theReal
27th March 2002, 07:17
Someone posted an article here recently that dvd developers have now agreed on the next generation of dvd. It said that this next generation will not be blue laser high-capacity disks (too expensive), but normal red-laser dvds, only the content will be in high quality mpeg4 (maybe to fit more video on a dvd? Or to get higher quality? I didn't quite get what the actual benefit was...)

What I want to say with this is underlining ulfshacks statement about the geeks who need better glasses - the companies are not aiming for the highest possible quality to satisfy some video freaks like us. They want to sell as much as possible with the highest profit possible. The quality must only be as good as that the standard customer thinks it is "really good, better than what we had before".
THIS is the reason why a lossy codec like mpeg2 is the standard for digital video right now...

movmasty
27th March 2002, 12:05
@hyc
>produced in YUV color space and the UV resolution is 1/2 the Y resolution

then rgb=1+1+1=3(8+8+8=24), YUY2=1+1/2+1/2=2(8+(8+8)/2=16)
this is a 2/3 compression, not 1/2
rgb isnt 32 bits, but 24, the other byte is the so called alpha channel, added from editing progs for editing purpose.

yuv is NOT a lossless compression, you loss some chroma info, just the eye doesnt notice,to cheat the eye is the base of lossy compression,
it is just different from dct of mjpeg, because if you recompress 100 times at 100% quality in mjpg you will get shit,
while if you recompress 100 times a movie in any yuv compression you will get EXACTLY the same movie as the first "loss"
That is.

YUY2 is the format used in broadcast, but you could do more,
i sometime use yuv9(8+(8+8/16)=9) like intermediate file,this is a compression of 24/9 ~2.66, and results are quite good,
but if you find this a too heavy chroma compression,you could always use the yuv12(8+(8+8)/4=12), this time a real 1/2 compression(24/12=2),with still better results
it works by using a unique chroma value every 2 pels horizontally AND every 2 pels vertically,
obviously broadcast cant apply yuv12 or yuv9 because interlaced format cant have a VERTICAL chrominance compression.

>In the US, broadcast video is produced in 720x480, 16 bits per pixel

In the US, broadcast video is produced...NOT in digital format,no pixels nor bytes there....
there are interlaced linees,that could be seen as EQUIVALENT to pixels,
and a HORIZONTAL BANDWIDTH that gives ABOUT 460 DOTS, that can be rendered with ABOUT 640 pixels.
the reasons why ITU choosed a 720 standard are others.
I am NOT saying that if you count the pixels on your ntsc screen you will count 640 instead of 720....
in fact very rarely the number will reach 600...on COLOUR 4:3 screens.
Look close at your tv, each pixels is composed of 3,smaller,R,G and B pixels,
then the real number of pixels is 3 times that of a b/w screen,
and, applying the yuy2 compression,3/2=1.5,
you will get the real res of your tv screen,~600x1.5=900 pixels....
that is why 16:9 screens looks ok with same bandwidt...
and also why PAL only added vertical linees to ntsc res....
and why movie films are 1.37/1 while screen is 2.35/1....

>The pixels are in YUV format with 8 bits for Y and 4 bits each for U and V

lets admit for a moment that there are bits on analogical tv,
4 bits? you mean 16 colours? this would be a very poor chrominance,
8 bits every PAIR of pixels isnt really the same of 4 bits for every pixel.

>I already went thru the rest of the math above

but a very imaginary math.....

Now....stabilished that with "lossless entropy coding" RGB coded files gets an average compression of 50%....we have to note that...
since yuv is a compression too(it puts ONE value in place of 2 or 4 values).....we will get less.

So,
YUY2 psycovisual lossy compression(2/3) + some true lossless algo ~38%
YUV12 psycovisual lossy compression(1/2) + some true lossless algo ~30%

these are the basis for further improvements,
and i think to the -h's lossless interframe compression, good luck!...

@theReal
to fit HDTV on a 4.7gigs DVD,
if you use mpg2 you will need like 30mbps that only very fast cpu and video boards could handle.

hyc
27th March 2002, 13:23
You've explained some things very well, but in fact the major US TV networks today do everything in digital. Conversion to analog is one of the last steps before the actual transmission.

auenf
28th March 2002, 03:26
Content encoded without a region-code on the disc can be played on all players, regardless of region designation. However, these discs are hard to find and are normally made by independant film companies or are privately-generated.

well, there is no such thing as a disc with no region coding, the discs that play in all player have every region selected when making a disc.


Why film look so much better on a DVD then home video ?
They have armies of geeks that can sit for day's looking at the video and optemise the compression.
While we are talking about using a faster CPU to compress the video faster, they are sitting and manually setting the location of the 'I' 'B' and 'P' frames.
Have any body here triad to do it with TmpgEnc

uhh, i dont think anyone would like to do that for a living, they would go crazy after a day.

basically all the big studios use a seriously expensive mpeg-2 hardware encoder, and make a few dozen copies, and get some people (usually employees or employee's family) to watch the disc, and check that it works fine and that there arent any encoding artifacts. if there is encoding artifacts, they can just re-encode that section with a higher bitrate.

Enf...

dragoman
28th March 2002, 07:28
Hi,

I think I was correct....there are EXTREMELY RARE DVD's with no region coding (remember, region coding is just a bit of data that corresponds to a region...) that play in all dvd players.

dragoman

auenf
28th March 2002, 14:09
if the disc doesnt have region coding, it would play in almost 0% of players.

when a disc is insterted into a, say, R4 player, the player checks the r4 on the disc, if its enabled, it will play, if its not enabled, it will not play, RCE works differently, which i wont go into right now. the fact that all the other regions on the disc are enabled doesnt matter to the R4 only player, but if the r4 wasnt enabled (the disc your referring to is one where no region is enabled, and hence would get rejected in this example).

region free players are just players that check all regions, which is how they get messed up with RCE, which i dont know exactly how works but i think its a all region enabled disc, but the disc uses some command sequences to check what regions are enabled, and if all regions are enabled, then it brings up custom error.

even if no region discs were possible (all authoring software shouldnt let you create a disc without region coding), i wonder if there is any player to actually play them anyway?

Enf...

movmasty
30th March 2002, 04:06
then better than DVD there is the HDTV, like 1080x1920
who built a HTPC captures from hdtv like hbo just copying the broadcasted mpg2(dtv is coded directly in mpg2)
and this is like 2.5mB/sec,that is 9G/hour.

then all true htpc owns a DVD writer,
and they span a hdtv film of 2 hours in 4/5 DVD....
or use RAID of 4/6 hd of 160g each
and divx compression is like shit in their mind,
also dvd quality is considered very low.....
naturally they dont see the 1920x1080 movies on small screens like us,
retroproiectors are a must.....

they are using exotic software like hipix....

all this make me feel like a.... papuasian :scared: