Streaming DVD Video over the Internet
Sexy Commando writes "According to this article on ZDNet, the new codec, H.264, is able to stream DVD quality video using bandwidth as little as less than 1Mbps. The new codec requires 3 to 4 times as much CPU power than MPEG-2 to process the video. Now we can have two movies on 1 CD. Cheers."
Well, I've got no phd in DVD technology, but the AC3 sound alone would take up far more than 1mbit all by itself right?
One of the reasons im not into watching movies on my PC is that I cannot take advantage of my DTS gizmos.
If this is just for video quality - Count me out.....
True ravers don't need drugs
From another article:
"To date, LSI Logic has not outlined the company's plan for how and when to introduce silicon capable of handling H.264. Umesh Padval, LSI Logic's senior vice president of broadband entertainment division, acknowledged that Bob Saffari's group - responsible for professional video market - has seen a growing demand for H.264. But as far as the volume consumer H.264 market is concerned, he said: "The actual deployment for H.264 is not solidified at all."
Padval predicted that the volume market for H.264 won't emerge before early 2005."
if it takes 3 to 4 times more cpu power just to decompress it, how long does it take to actually make these files? I've done some DivX-ing and 16 hour compression sessions are too long.
MABASPLOOM!
It's just software. One good content "trigger" and distribution becomes a non-issue. For example, imagine that Apple puts h.264 into QuickTime 4.1, and puts out the next big trailor or streamed event out in h.264. This worked for MPEG4 quite nicely -- millions of user installs in a week.
If the question is whether consumer computers are fast enough to support h.264, well, I haven't heard anything yet to indicate that would be an issue. The only devices so far discussed as not having the muscle to support h.264 were PDA's and wireless devices.
What fun!
I think someone was a little bit overoptimistic :-).
Yours, Martin
It claims this new codec can get the same quality at 33% lower filesizes than other MPEG-4 codecs, but it doesn't say WHAT MPEG-4 codec. There is more than a 33% difference between existing MPEG-4 codecs alone! Are they comparing this to DivX 5.x, arguably the current leader in quality? Or are they comparing it to Microsoft's ISO MPEG-4 encoder, with it's horrid quality?
Regards, Guspaz.
Many DVDs are already coded at a fairly minimal 2Mbps or so*, so this really isn't as big a deal as people might think. If someone developed a truely intelligent encoder (for starters, one that didn't follow a static frame type pattern such as I-B-P-B-I) and fed it a really clean signal then we could really make progress.
* ZThe bitrate is according to an industry insider who gave a talk at UC Berekeley. The bitrate is low so that they can fit all the extras on a DVD, which most consumers value more than movie quality.
It says 3-4 times to process the video? Does that mean just to encode or encode and decode? If it takes 3-4 times as much to both, then the CPU will be an issue. An 800MHz PIII will not be able to handle it well and chances are only dual systems of any kind will be able to handle it very well. There's a reason why Appple went to dual processor configurations. DVD encoding/decoding takes a lot of CPU usage as is.
I disagree. Now people are encoding to divx in better than real time on some machines so i dont think its a problem to take 8 hours to encode a movie. I remember when i did 20 hour encodes back in the day. Its really not a big deal.
unzip; strip; touch; finger; mount; fsck; more; yes; unmount; sleep
The little problem is when you get more storage or more bandwidth, you tend to find ways to use it to the max...
I set up a file server for my brothers' MP3 archive (he is a DJ and a few use mp3's on an external drive and laptop for gigs, just because you don't have to lug hundreds of CD's around), and it has 240GB of RAIDed storage, that lasted 6 weeks... it's been teetering around the 10GB mark for a while)... plus a 5Mb broadband doesn't help in keeping the HDD free! We're just saving up for another storage upgrade.
Are you local? There's nothing for you here!
A couple of guys from Hardforums and I played with this codec back in early July. There is a writeup here about how to create a file. At the time we were using it, it was a pain to compress and decompress. At the time the encoder required a YUV file and the decoder produced a YUV file. On a 1.2 TBird it took about 14 hours to encode the Final Fantasy - The Spirits Within Trailer. It was about 2:45 I think. The file produced was 11.2MB. A comparable (quality as best we could tell) Divx 4 encoding was about 35MB, both started from DVD and contained no audio. Decoding was about 2fps on my machine. Remember that these times are using files that were written to be correct, with no efficency added in. In fact, one of the guys on the JVT team told us if we were able to improve the compression at all to let them know.
;)
/.ed to be able to still see it, just at a lower quality.
Btw, JVT stands for Joint Video Team, which is the group resposnible for developing the standard. It used to be H.26L, and looks now to be called H.264. The ftp below is the once that is used by the people developing the standard, so don't hit it too hard
And here's what you all have been waiting for. the Source Code to it. I dunno how it's changed since I used it last, but the newest version we had available was 3.2 and they are now on 4.2. Version 3.7 came out shortly after we finished our tests, but there were no compression speed changes from the few quick tests we ran on it, as well as no file size changes.
Also, one intereting thing that I didn't see when glancing over the linked article was that the server's software will monitor the connection and playback and if there are too many dropped frames it will decrease the quality. The opposite is true as well, the quality will increase based on the connection and playback. Of course the server would be able to disable this as well, but would be nice if a video stream got
From occasional obvservation, I have come to the conclusion that Divx requires AT LEAST %50 more CPU power than MPEG2 software decoding.
Now, mind you, these number are not at all easy to come by because most people don't use the same quality levels as a standard MPEG2 DVD.
Most people use a slightly lower resolution, this will definitely save you some CPU cycles decoding the video.
Another thing most people don't think about is that the actual video data used to describe the scene is directly related to the processing power required to decode the scene. Finer detail reflective of a true MPEG2 stream will use a lot more processing power than your standard "movie on a disk" Divx encode.
If you use dual-pass, you can approach MPEG2 quality @ 720x480 at approximately 2mbit. Accomplishing the same at 1Mbit is a great acheivement.
You people seemed to have missed the major point of this codec, and that is we finally have a code capable of delivering TV over DSL, or other low-bandwidth LANs like 802.11b. There are movements in place to create real-time encoders from MPEG2, and this is where the REAL market acceptance will be.
Man is the animal that laughs.
And occasionally whores for Karma.