It is extremely unlikely that much will actually happen. Remember that there are only 4 more years on the MP3 patent...

The problem isn't in video or picture formats, the problem is in software patents. If a country permits this stupidity, it should expect stupid behavior from lawyers and judges. (and decrease of innovativeness in the country)

So far this road to xxx is used only by US, and Japan. It looks like something is in the air in US, and territories under its influence.

You might look at the letter of Donald Knuth to a patent office.

BTW if you need an image codec for lossy compression, you might use my image codec. It has quite efficient hierarchic image decomposition, and work quite nicely for natural images. Of course you might review possition of IBM on range coding, and if they'd think they have patented such a trivial algorithm. Or you might sue patent office for allowing a trivial algoritm, or mathematical equation being patented. ~_^

Wow. This answer brings many questions.

1. Is IBM claiming patented status of this range coding?
2. Is this range coding also at the heart of PNG, VP3 and the oggs?
3. What codec of yours? (Link?)
4. What range coding patents (USPTO numbers?)

Many questions have I, I know.
Stop I now before used to talking like Yoda I am.

Thanks!

Quote:

Original post by medevilenemy It is extremely unlikely that much will actually happen. Remember that there are only 4 more years on the MP3 patent...

In this case it is a very good time to milk the cow. The current trends in management keep antisocial greedy behaviour and shortsighted lack of common sense and honour in great esteem.

I agree. It is normative to squeeze the last possible revenue out of a dying patent. Furthermore it is also possible to "glean" after the patent has expired by suing for "losses" incurred during the life of the patent. I do not know what the statute of limitations is for patent infringement but it probably comprises the span since they changed their policy to include decoding MP3, as well as including the time for them to eventually get around to the little guys.

Normally patent lawsuits are intended to convince a potential licensee to sign on the dotted line for future licensing coverage, but in the case of an expired patent they have no future revenue to lose, and if they can show that you probably would have signed then they can probably sue you for what you would have paid as a "loss" that they had incurred. In a civil case they need to prove a loss and that is a fairly straightforward way of doing so.

There is the mitigating factor that if they come off as complete jerks regarding MP3 they could scare everyone away from MPEG4 into the arms of ogg theora. But if they figure most people won't bother then they could consider that as acceptable collateral damage and force those who don't know any better to sign for MPEG4 by their "making an example" of the last MP3 users, somewhat like the media moguls made an example of the file sharers. Today most would not particularly desire to hear the sound files that were once the subject of those severe penalties but nobody will soon forget the SS tactics they used in making an example of them. Likewise it is not inconceivable for them to make an example of past users of expired MP3 patents to encourage MPEG4 licensing.

For those who have included MP3 decoding or encoding in their projects in the past a good way to prepare for potential action might be to collect and keep in one place the proof that MP3 usage did not have more than 100,000 dollars of revenue associated with it in any one year. Now they might stretch the definition of "associated" to include revenue from all projects (because people may have heard of your other projects from using the MP3 ones), in which case you might show that you never made more than that total income from all your projects during each year in question. Once having done that, as far as I can figure out, there might be no need to go any further. That is as far as I can tell. I may suggest that you to go check things out for yourself but I do NOT give legal advice.

Quote:

Original post by inspired Wow. This answer brings many questions. 1. Is IBM claiming patented status of this range coding? 2. Is this range coding also at the heart of PNG, VP3 and the oggs? 3. What codec of yours? (Link?) 4. What range coding patents (USPTO numbers?) Many questions have I, I know. Stop I now before used to talking like Yoda I am. Thanks!

1. They might have patented arithmetic coding. However arithmetic coding is a prime example of a mathematical type of algorithm (that is also somewhat trivial). So it shouldn't be patented in first place. It's just a two mathematical equations. Range coding is closely related, and I suspect that some optimizing compilers might compile them into the same binary code.

2. PNG uses zlib and image decomposition. VP3 uses (possibly DWT) and DCT with standard way to encode video files. Ogg vorbis uses MDCT. (modified discrete cosine transform, that stuff that uses two inputs to create one output)

All are standard algorithms, taught on universities around 30 years. PNG was created to remove problems with GIF files. However, patent for LZW expired few years ago.

3. A little video codec that I'm sometime doing in a free time, it supports full screen transformation, it's able to transfer codec (digitally signed) inside a video file. While it's still unfinished, and I don't know if it would be better than h264, it could be easily used as a image compressor at somewhat JPEG2000 level.

As for link
As Xellos said it's a secret.

I would need to move my butt and modify it into workable level. It would be somewhat easy, that two parts are well tested.

4. Range coding is supposedly patent free. (Of course it depends on acombination of a lawyer, judge, and a technical consultant.) ^_^

Quote:

Original post by inspired even if you provide software to decode an MP3 into ogg

That would be a horrible idea anyway; both are lossy formats, and each discards slightly different information, so converting from one to the other will cause a loss in quality. Always encode your mp3s or vorbis files from a lossless source format. Note that converting mp3->wav->ogg (I've seen it suggested before) still wouldn't help - once that information has been discarded you do not get it back.

Quote:

Original post by Raghar As for link As Xellos said it's a secret. I would need to move my butt and modify it into workable level. It would be somewhat easy, that two parts are well tested. 4. Range coding is supposedly patent free. (Of course it depends on acombination of a lawyer, judge, and a technical consultant.) ^_^

That consultant could be a concern. I watched in horror as a technical consultant decieved a whole courtroom. I was the only juror to recognize the victem's ongoing need for medical care. I am wondering if VP3 is a better bet in court than range coding, if for no other reason than that it has been patented then given away; thus the USPTO has blessed it.

Quote:

Original post by Kazgoroth Quote: Original post by inspired even if you provide software to decode an MP3 into ogg That would be a horrible idea anyway; both are lossy formats, and each discards slightly different information, so converting from one to the other will cause a loss in quality. Always encode your mp3s or vorbis files from a lossless source format. Note that converting mp3->wav->ogg (I've seen it suggested before) still wouldn't help - once that information has been discarded you do not get it back.

Right on. Just thinking of general compatibility with all users. Perhaps best to simply provide a link to CDEX. They also discourage compressed to compressed encoding.