[theora] NHW Image codec

Raphael Canut nhwcodec at gmail.com
Fri Feb 6 10:00:04 PST 2015


Yes, the next step for this compression scheme is to make it adaptive, more
rle for plain regions, more differential coding for high contrast.I could
start for example by decomposing the wavelet DC image in 16x16 blocks and
giving 4 modes to each block: 1) plain (blank) block, more RLE, 2) high
contrast, edges, more differential coding, 3) mid, current scheme, 4)
texture, grain.As in the current state this scheme gives pretty good
results on wavelet DC parts, this could be even better.

Else, I'm slowly working again on the NHW codec.I would like to code the
lower quality settings (-l3 and below, high compression) but it seems hard
with my algorithm and HEVC intra is doing very well in this domain.On the
contrary, for mid/normal compression (-l2 to -h3 setting), I prefer the NHW
codec to HEVC because it has more neatness, I also prefer in this range
Daala codec because of its grain and precision.

Would it be possible to create a codec only for mid/normal compression (and
upper quality)? Is there a niche for this kind of codec? My codec will be
still royalty-free and very fast.


2015-01-23 18:31 GMT+01:00 Raphael Canut <nhwcodec at gmail.com>:

> Hello,
> So here is the compression scheme that I don't know if is covered by
> patent.
> I use it on the wavelet DC parts.It outputs 1 byte (8 bits) words.There is
> 5 modes:
> - 1(MSB) | 7 bits to store value (0->127 range)
> - 01 | 6 bits to store val[n+1]-val[n] | val[n]-val[n-1] if each in the
> range [-4,4]
> - 001 | 5 bits to store val[n+2]-val[n+1] | val[n+1]-val[n] |
> val[n]-val[n-1] if each in range
> - 0001 | special case
> - 0000 | 4 bits for RLE+prediction
> It is a simple compression scheme, it is not as good as context modeling +
> arithmetic coding, but it gives good results on my data (wavelet DC parts)
> and is very fast.
> My question is: Does someone know if this compression scheme is patented?
> Also, as it is quite simple, maybe the patent has expired (more than 20
> year old)?
> Also, if someone knows a better compression scheme for DC parts (not too
> complex, I don't want to use context modeling + arithmetic coding for now,
> even if I know that patents on arithmetic coding have nearly all expired),
> do not hesitate to let me know.
> Cheers,
> Raphael
> 2015-01-18 10:02 GMT+01:00 Raphael Canut <nhwcodec at gmail.com>:
>> Hello,
>> Ok, and that's too many work to review a source code.In my codec, I have
>> 3 compression schemes, I think 2 are not patented, but the third... I don't
>> know.For the rest, I think my codec is patent-free (I don't use
>> SPIHT,EZW,zerotree methods), even the wavelet transform is new and don't
>> use the lifting scheme nor the convolution product.
>> Else, if you found time to review the codec, do not hesitate to let me
>> know what you think of it and if you would have remarks.Would be much
>> appreciated.
>> Cheers,
>> Raphael
>> 2015-01-17 23:01 GMT+01:00 Basil Mohamed Gohar <basilgohar at librevideo.org
>> >:
>>> On 01/15/2015 12:39 PM, Raphael Canut wrote:
>>> > Hello,
>>> >
>>> > Yes, that's right.I still think that it would have more weight if I
>>> > could claim that Xiph has reviewed the source code and find that the
>>> NHW
>>> > codec is royalty/patent-free (better than me alone claiming
>>> > this...).Maybe it was in this sense that Ralph Giles answered me?
>>> >
>>> > Cheers,
>>> > Raphael
>>> That's not a guarantee that anyone can make, not even a patent
>>> attorney.  The current landscape is such that "being clear of patent
>>> concerns" is impossible.  Someone can be destroyed by the legal process
>>> that results in them being exonerated, but then left with no will nor
>>> resources to continue development.
>>> I don't say this to scare you away, just understand that asking Xiph to
>>> "clear" the patent issues doesn't make sense.  Xiph is a non-profit
>>> organization that advances the cause of free, open, and royalty-free
>>> technologies, but cannot offer a technically legal service as requested.
>>> --
>>> Libre Video
>>> http://librevideo.org
>>> _______________________________________________
>>> theora mailing list
>>> theora at xiph.org
>>> http://lists.xiph.org/mailman/listinfo/theora
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.xiph.org/pipermail/theora/attachments/20150206/15f8fa7b/attachment.htm 

More information about the theora mailing list