MPEG 4 Information

[DragonLips]

As posted on Streaming Media.com
http://www.streamingmedia.com/article.asp?id=8569&page=3&c=7

MPEG4 – Internet Streaming and Synchronized Multimedia
Where MPEG2 was designed to scale up to broadcast and high-definition quality and operating requirements, MPEG4 goes the other way. It's designed to scale down – down to dial-up internet bandwidths and to tiny devices like cell phones and PDAs; as well as still remain viable for high-quality desktop streaming up to 1Mbps. MPEG4's AAC audio codec is the root of the MP4 file type, recently popularized by Apple's iTunes, among others.

But MPEG4 is much more than just an audio and video compression/decompression scheme. It's a container for all kinds of media objects (images, text, video, animation, interactive elements like buttons and imagemaps, etc) and a way to choreograph them into a synchronized, interactive presentation. (see Interactive Authoring for MPEG4 for more info). MPEG4 also has standard interfaces to allow plugging in a DRM scheme called Intellectual Property Management and Protection (IPMP).

MPEG4 is still at the frontier of media technologies. The specification is extensive, and each vendor implements it in their own way. Try a variety of MPEG4 tools and you'll find lots of incompatibilities. But some are working to smooth the landscape. The Internet Streaming Media Association (ISMA) is an industry consortium dedicated to interoperability among MPEG4 products and services. Essentially, any implementation that's ISMA-compliant will work with any other.




Personally, I need to read up on this a bit more. Apparently MPEG 4 is very scaleable for vendors. They can shape the codec to fit their needs. I am concerned as it appears to be what MP3 is to wav. It's more compression should allow more channels, but I have to think that is going to hurt the product. .wav format is better than MP3, the file sizes are too high. Mp3 is better than WMA format, but WMA has more manageable file sizes. I think there is some real truth to the fact that the larger the size of the image/sound, the better it is. Personally I'd rather see 50 channels through voom pumping out the most prestine images, than seeing 450 channels of slop. I think the reality we're facing is MORE compression with MPEG4, which should dampen image quality. The file sizes will smaller for easier transmission, but I'm skeptical at this point. Each change in compression has taken us further away from the original presentation. We'll have to see.

 

[Sean Mota]

In a way you are right, it will compress the channel more but the compression scheme will give you as good quality as MPEG2 is right now.

 

[rang1995]

Even with competition from proprietary video compression formats from RealNetworks and Microsoft Corp., the clear advantages of MPEG-4 in regards to quality and scalability will ensure its eventual dominance in the consumer electronics and streaming industries. MPEG-4 includes state-of-the-art compression techniques, which allow it to provide greater compression and higher quality than MPEG-2 and earlier standards. In the case of MPEG-4, scalability refers to MPEG-4's specific provision to allow it to work well over a range of bandwidths and screen sizes.

These scalability and quality advantages unique to MPEG-4 are driving the standard's acceptance within the streaming, electronics and even security/surveillance markets. The ISMA, for example, is promoting the use of MPEG-4 as the encoding standard for video delivered over the Internet.

Manufacturers of set-top boxes and DVD players are also increasingly providing support for MPEG-4. This will allow home users to access the wealth of video available on the Internet and will enable distribution of high-quality, low-cost video content on CD-ROM as well has the more expensive DVD disks. Camcorder manufacturers view the scalability and quality of MPEG-4 as integral to new applications, such as sending video clips by e-mail. The 3G mobile phone industry consortia have also adopted MPEG-4, and phones with MPEG-4 capability are now coming onto the market in Japan.

 

[rang1995]

or try this

http://www.altera.com/solutions/consumer/industry/csm-industry.html

 

[squicken]

Whoosh. Right over my head. Where does the flux capacitor fit into all this?

 

[vission]

Personally...

I've used various encoding technologies, and I can't say that MPEG-4 is terribly better than other current codecs. In fact I have found it to be worse than most current codecs using the same bandwidth.

If MPEG-4 had been released when it was supposed to be instead of being held up by licensing issues it may have been the best at the time. The best thing I see that it has going for it is compatibility, not quality.

Here is a link to a comparison in which I agree with:

http://www.extremetech.com/print_article/0,1583,a=121163,00.asp

I must admit that I have had great success with the WMV9 codec, the picture quality is superb at the same bandwidth. Also you may find this interesting:


http://www.eetimes.com/sys/news/OEG20040302S0013

The biggest issue that I see is the lack of true WMV9 (real-time)hardware encoders. As this will be a necessity for any sort of video transmission, however it will not be long before such hardware exists. Even your next video card might support it:

http://www.extremetech.com/article2/0,1558,1567092,00.asp

Enjoy!
Vission

 

[DragonLips]

Vission,
Those links really scare me! I'm seeing MPEG 4 as a terrible way to go with this. I know it's early, but cost cutting by saving bandwidth costs will lead to a terrible product. If VOOM want to carve it's own nitche into the sat world, it needs to do so by offering STELLAR PQ at uncompromised value! I don't see that when I look at MPEG 4. I agree with you on the WMedia format, but I think we won't see hardware encoders because it's got Microsoft's name on it. HD encoding there is STELLAR!! Hell, I'd take Divx at this point.

 

[ZeitGeist]

Hmmmm...

I think most of you guys that want to sit around and wring hands about the quality of WMV vs. MPEG4 should take a valium. I'm not going to get into the religious debate over whether the Windows Media codec is better or worse than MPEG4 but I would suggest that Ziff-Davis ExtremeTech is an unreliable source for information on which codec is better for the following reasons:

1. Divx is MPEG4
For some reason, the monkey who wrote the comparison doesn't seem to know that Divx utilizes a version of MPEG4

2. Quicktime/Sorenson and Quicktime/MPEG4 are not professional products
These products have not been developed as a professional encoding scheme to be used by video professionals or to be embedded into professional (or even consumer) level equipment. Basically, they exist so that DV video from a camcorder, or movie trailers, can be made into a small file for incorporation onto a web page where the video can be streamed by a viewer in a manner that is cheap and easy and with very little overhead on the server side. No movie, television broadcast, or satellite/cable video stream is going to be encoded with these codecs. They were never designed to do that.

3. Dubious tests
Taking previously encoded material from DVD or videotape and then re-encoding with another codec is inherently flawed and completely worthless as a test. Talking about MPEG artifacts and the differences between various codecs without actual empirical data is no more scientific than saying one picture just looked better than another.

I could keep going on but the ultimate fact is that there is an excessive amount of opinion and precious little fact in most of the opinions about how Windows Media is better than MPEG4. One thing is certain, we're getting MPEG 4 so if you want another codec you better look for a satellite or cable provider that is going to use a set top box with Windows Media - I don't think one exists or is going to exist anytime soon.

 

[soledade]

I think most of you guys that want to sit around and wring hands about the quality of WMV vs. MPEG4 should take a valium. I'm not going to get into the religious debate over whether the Windows Media codec is better or worse than MPEG4 but I would suggest that Ziff-Davis ExtremeTech is an unreliable source for information on which codec is better for the following reasons:

1. Divx is MPEG4
For some reason, the monkey who wrote the comparison doesn't seem to know that Divx utilizes a version of MPEG4

2. Quicktime/Sorenson and Quicktime/MPEG4 are not professional products
These products have not been developed as a professional encoding scheme to be used by video professionals or to be embedded into professional (or even consumer) level equipment. Basically, they exist so that DV video from a camcorder, or movie trailers, can be made into a small file for incorporation onto a web page where the video can be streamed by a viewer in a manner that is cheap and easy and with very little overhead on the server side. No movie, television broadcast, or satellite/cable video stream is going to be encoded with these codecs. They were never designed to do that.

3. Dubious tests
Taking previously encoded material from DVD or videotape and then re-encoding with another codec is inherently flawed and completely worthless as a test. Talking about MPEG artifacts and the differences between various codecs without actual empirical data is no more scientific than saying one picture just looked better than another.

I could keep going on but the ultimate fact is that there is an excessive amount of opinion and precious little fact in most of the opinions about how Windows Media is better than MPEG4. One thing is certain, we're getting MPEG 4 so if you want another codec you better look for a satellite or cable provider that is going to use a set top box with Windows Media - I don't think one exists or is going to exist anytime soon.

Well said. I could not agree more. One more thing to add. WM is not a real time codec. Broadcasting requires a real time codec. Even with a custom ASIC WM is not designed for real time encoding.

In my experience in our lab at work MPEG4 is the same quality as MPEG2 in less than half the streaming throughput.

 

[dledeaux]

I would also like to point out that they did not include MPEG2 at all, so we really have no frame of reference for how it would look with some of those same scenes. It could look worse...

 

[Sean Mota]

Well, this should bring up more questions/answers to the table now. Voom to use WM9

 

[rang1995]

http://www.chiariglione.org/mpeg/standards/mpeg-4/mpeg-4.htm

 

[vission]

I think most of you guys that want to sit around and wring hands about the quality of WMV vs. MPEG4 should take a valium. I'm not going to get into the religious debate over whether the Windows Media codec is better or worse than MPEG4 but I would suggest that Ziff-Davis ExtremeTech is an unreliable source for information on which codec is better for the following reasons:

1. Divx is MPEG4
For some reason, the monkey who wrote the comparison doesn't seem to know that Divx utilizes a version of MPEG4

2. Quicktime/Sorenson and Quicktime/MPEG4 are not professional products
These products have not been developed as a professional encoding scheme to be used by video professionals or to be embedded into professional (or even consumer) level equipment. Basically, they exist so that DV video from a camcorder, or movie trailers, can be made into a small file for incorporation onto a web page where the video can be streamed by a viewer in a manner that is cheap and easy and with very little overhead on the server side. No movie, television broadcast, or satellite/cable video stream is going to be encoded with these codecs. They were never designed to do that.

3. Dubious tests
Taking previously encoded material from DVD or videotape and then re-encoding with another codec is inherently flawed and completely worthless as a test. Talking about MPEG artifacts and the differences between various codecs without actual empirical data is no more scientific than saying one picture just looked better than another.

I could keep going on but the ultimate fact is that there is an excessive amount of opinion and precious little fact in most of the opinions about how Windows Media is better than MPEG4. One thing is certain, we're getting MPEG 4 so if you want another codec you better look for a satellite or cable provider that is going to use a set top box with Windows Media - I don't think one exists or is going to exist anytime soon.

1. He might not have said it but it is a known fact(to those it concerns) that DIVX is a hack of Microsoft's out-dated non-ISO version of MPEG-4. As such I would have expected the quality to be worse than that of a ISO-MPEG-4 encoder. I honestly want MPEG-4 to succeed and blow away all others I'm not trying to be on Microsoft's side, I just want the best codec for a given bandwidth (at least till Blu-Ray).

2. Is there a recommended ISO-MPEG-4 codec available then, so that I can compare it's quality to other codecs? Which vendors ISO-MPEG-4 codec do you use?

3. I will agree that using a compressed source is not an ideal reference however, all 4 codecs tested were given that same source and the results were from that source, I'm not sure in the real-world encoders are always thrown pristine reference material. Also, I don't remember the author ever mentioning he was a scientist or that he intended on pulling out the scopes and giving us bit by bit comparisons. It was a real world test, when people on here are complaining about PQ it isn't using hardware, it's what they perceive.

I am actually looking forward to MPEG-4 on Voom, it'll "hopefully" mean more HD channels without PQ loss. I do not expect WMV to take over the world it just looks pretty, heck it's uselless for the DVD work I do. If you have some HI-Quality MPEG-4 material I'd really like to see it especially in 1080i format so I could compare it to the WMV 9 HD samples, I'd really like to see them.

Enjoy!
Vission

 

[Ken F]

Well said. I could not agree more. One more thing to add. WM is not a real time codec. Broadcasting requires a real time codec. Even with a custom ASIC WM is not designed for real time encoding.

There is no such thing as a "real time codec." Equipment is either available for real-time encoding with a codec, or it is not. Real-time HD encoding solutions are not currently sold for WM9 and H.264. However, real time encoding solutions for WM9 are in development. For example, Harmonic and TANDBERG are both demonstrating their new real-time encoding solutions for WM9 SD, and both are developing products for WM9 HD.

See this Microsoft announcement from NAB2004.

 

[rang1995]

found this om google any help???
http://www.updatexp.com/codec-for-windows-media-player-9.html

 

[vission]

There is no such thing as a "real time codec." Equipment is either available for real-time encoding with a codec, or it is not. Real-time HD encoding solutions are not currently sold for WM9 and H.264. However, real time encoding solutions for WM9 are in development. For example, Harmonic and TANDBERG are both demonstrating new real-time encoding solutions for WM9 SD, and both are developing products for WM9 HD.

See this Microsoft announcement from NAB2004.

Thanx, Ken!

Interesting read so we just may see WMV9 in our next Voom box. I wonder if the current box can be upgraded to support it or if that will mean a box swap?

I would guess we may need a box swap since WMV9 HD does take a little umph to decode, even in dedicated ASICS.

Should be an interesting year.

Vission

 

[Ken F]

Microsoft has some WM9 HDTV demos material at the link below.

http://www.microsoft.com/windows/windowsmedia/content_provider/film/ContentShowcase.aspx

Try the Scooby Doo2 or Coral Reef adventure clips in 720p, or 1080p if you have a fast computer. Keep in mind that VOOM will probably use higher bit rates than seen in most of these demos, so quality should be even higher. Also note that most of the material on this demo page is sourced from film (like what you see on HBO-HD), rather than video.

 

[DragonLips]

I wonder If we'd all have to then run out a buy the latest/greatest sets that support 1080p. I've read that several of the newer DLP sets slated for production this year will support 1080p.

 

[vission]

Oh my

I hope not I don't think I could control my bodily functions if I saw 1080p picture quality.

 

[madpoet]

1080p is oustanding... even when it's just being output on a device that only does 1080i .

It's a SERIOUS strain to decode though.