[theora] NHW Project - lower quality settings

[Raphael Canut]

Hello,

I have re-tested -l4 high compression setting and it's clear that it lacks
of precision on degraded, rather blurred images. So I don't know if it is a
good idea to base the other lower quality settings (-l5,-l6,...) on -l4
setting.I have tested the NHW codec against x265, x264, Daala, WebP,
Rududu, DLI and it's clear that at high compression these very good codecs
have more precision.But they also tend to blur/decrease neatness-sharpness
of image where the NHW codec would enhance image neatness/sharpness.But
that's right that on degraded, blurred images where there is no neatness to
enhance, the results start to be not good and lack of precision.So for now,
I recommend to use -l4 setting on well-defined, good quality images.

I have also "simply" increased the wavelet coeff threshold of the dead-zone
for -l5 setting (to save -5Ko), but the quality drop is too big for now.The
other codecs (above) don't have that drop in quality from -l4 to -l5
compression.So apparently, there needs to be more smartness to pass from
-l4 to -l5 with the NHW codec.I'll try to work on this, but it's extremely
complex, for me...For example, Rududu doesn't have this problem, the codec
source code has been made available recently on GitHub, I have started to
study it and it is very powerful: advanced SPIHT tree-like coding, advanced
variance-based context modeling, advanced range coding, and other things...

So to sum up, all the very good codecs quoted above have a very good
precision at high compression but they tend to decrease neatness/sharpness,
on the contrary of the NHW codec, but on degraded images where there is no
neatness/sharpness, the NHW codec then lacks of precision...

Last but not least, for now the big advantage of the NHW codec I see, is
that it is very fast.The NHW codec is way faster than the codecs above, for
example it is still now H.264 technology that is used in mobile devices,
and the NHW codec compared to ultra-optimized x264 is x3 times faster to
encode and x1.2 times faster to decode, that is to say that with good
optimization (SIMD, multithreading) the NHW codec will be x12 times faster
to encode and x5 times faster to decode than x264...

Again working on high compression is very complex, and any help would be
very welcome.And I know there are experts out there, more skilled than me,
that could have very good ideas for the NHW Project, if they could find the
time...

Do not hesitate to contact me!

Many thanks!
Cheers,
Raphael

[Raphael Canut]

Hi David!

Many thanks for your very encouraging and kind feedback!!! (Hope you don't
mind that I made your post public on the Theora channel).

I try to be the more sincere as I can in my posts.Some people suggested me
to make demo pages to explain how the codec works step by step, and I
realize that it would be good, because certainly it will show that the NHW
codec is not that complex and it will be more easy then to contribute for
people who want and have time.

For now, I focus on high compression settings, for me it's very complex,
but currently high compression settings are mandatory for a new image/video
codec.

Many thanks again for your feedback!
Cheers,
Raphael

I enjoy reading your posts and always find them educational. I don't
have anything to contribute other than my appreciation. Thank you!

Post by Raphael Canut
Hello,
I have re-tested -l4 high compression setting and it's clear that it

lacks

Post by Raphael Canut
of precision on degraded, rather blurred images. So I don't know if it

is a

Post by Raphael Canut
good idea to base the other lower quality settings (-l5,-l6,...) on -l4
setting.I have tested the NHW codec against x265, x264, Daala, WebP,

Rududu,

Post by Raphael Canut
DLI and it's clear that at high compression these very good codecs have

more

Post by Raphael Canut
precision.But they also tend to blur/decrease neatness-sharpness of image
where the NHW codec would enhance image neatness/sharpness.But that's

right

Post by Raphael Canut
that on degraded, blurred images where there is no neatness to enhance,

the

Post by Raphael Canut
results start to be not good and lack of precision.So for now, I

recommend

Post by Raphael Canut
to use -l4 setting on well-defined, good quality images.
I have also "simply" increased the wavelet coeff threshold of the

dead-zone

Post by Raphael Canut
for -l5 setting (to save -5Ko), but the quality drop is too big for

now.The

Post by Raphael Canut
other codecs (above) don't have that drop in quality from -l4 to -l5
compression.So apparently, there needs to be more smartness to pass from

-l4

Post by Raphael Canut
to -l5 with the NHW codec.I'll try to work on this, but it's extremely
complex, for me...For example, Rududu doesn't have this problem, the

codec

Post by Raphael Canut
source code has been made available recently on GitHub, I have started to
study it and it is very powerful: advanced SPIHT tree-like coding,

advanced

Post by Raphael Canut
variance-based context modeling, advanced range coding, and other

things...

Post by Raphael Canut
So to sum up, all the very good codecs quoted above have a very good
precision at high compression but they tend to decrease

neatness/sharpness,

Post by Raphael Canut
on the contrary of the NHW codec, but on degraded images where there is

no

Post by Raphael Canut
neatness/sharpness, the NHW codec then lacks of precision...
Last but not least, for now the big advantage of the NHW codec I see, is
that it is very fast.The NHW codec is way faster than the codecs above,

for

Post by Raphael Canut
example it is still now H.264 technology that is used in mobile devices,

and

Post by Raphael Canut
the NHW codec compared to ultra-optimized x264 is x3 times faster to

encode

Post by Raphael Canut
and x1.2 times faster to decode, that is to say that with good

optimization

Post by Raphael Canut
(SIMD, multithreading) the NHW codec will be x12 times faster to encode

and

Post by Raphael Canut
x5 times faster to decode than x264...
Again working on high compression is very complex, and any help would be
very welcome.And I know there are experts out there, more skilled than

me,

Post by Raphael Canut
that could have very good ideas for the NHW Project, if they could find

the

Post by Raphael Canut
time...
Do not hesitate to contact me!
Many thanks!
Cheers,
Raphael
_______________________________________________
theora mailing list
http://lists.xiph.org/mailman/listinfo/theora

[David Willmore]

I don't mind at all. I imagine there are more people like me out
there who have been following your work, but haven't said anything.

I agree that a series of pages explaining how things work may lead to
more contributions. At the very least it would lead to more people
being better educated on how wavelets work in image compression. But,
this is your project, if you don't have the time or interest, don't
feel compelled to do it. Clearly, NHW is a labor of love for you and
I wouldn't want that to change!

Post by Raphael Canut
Hi David!
Many thanks for your very encouraging and kind feedback!!! (Hope you don't
mind that I made your post public on the Theora channel).
I try to be the more sincere as I can in my posts.Some people suggested me
to make demo pages to explain how the codec works step by step, and I
realize that it would be good, because certainly it will show that the NHW
codec is not that complex and it will be more easy then to contribute for
people who want and have time.
For now, I focus on high compression settings, for me it's very complex, but
currently high compression settings are mandatory for a new image/video
codec.
Many thanks again for your feedback!
Cheers,
Raphael

I enjoy reading your posts and always find them educational. I don't
have anything to contribute other than my appreciation. Thank you!

Post by Raphael Canut
Hello,
I have re-tested -l4 high compression setting and it's clear that it lacks
of precision on degraded, rather blurred images. So I don't know if it is a
good idea to base the other lower quality settings (-l5,-l6,...) on -l4
setting.I have tested the NHW codec against x265, x264, Daala, WebP,
Rududu,
DLI and it's clear that at high compression these very good codecs have
more
precision.But they also tend to blur/decrease neatness-sharpness of image
where the NHW codec would enhance image neatness/sharpness.But that's
right
that on degraded, blurred images where there is no neatness to enhance,
the
results start to be not good and lack of precision.So for now, I
recommend
to use -l4 setting on well-defined, good quality images.
I have also "simply" increased the wavelet coeff threshold of the dead-zone
for -l5 setting (to save -5Ko), but the quality drop is too big for now.The
other codecs (above) don't have that drop in quality from -l4 to -l5
compression.So apparently, there needs to be more smartness to pass from
-l4
to -l5 with the NHW codec.I'll try to work on this, but it's extremely
complex, for me...For example, Rududu doesn't have this problem, the codec
source code has been made available recently on GitHub, I have started to
study it and it is very powerful: advanced SPIHT tree-like coding, advanced
variance-based context modeling, advanced range coding, and other things...
So to sum up, all the very good codecs quoted above have a very good
precision at high compression but they tend to decrease
neatness/sharpness,
on the contrary of the NHW codec, but on degraded images where there is no
neatness/sharpness, the NHW codec then lacks of precision...
Last but not least, for now the big advantage of the NHW codec I see, is
that it is very fast.The NHW codec is way faster than the codecs above, for
example it is still now H.264 technology that is used in mobile devices,
and
the NHW codec compared to ultra-optimized x264 is x3 times faster to
encode
and x1.2 times faster to decode, that is to say that with good
optimization
(SIMD, multithreading) the NHW codec will be x12 times faster to encode
and
x5 times faster to decode than x264...
Again working on high compression is very complex, and any help would be
very welcome.And I know there are experts out there, more skilled than me,
that could have very good ideas for the NHW Project, if they could find the
time...
Do not hesitate to contact me!
Many thanks!
Cheers,
Raphael
_______________________________________________
theora mailing list
http://lists.xiph.org/mailman/listinfo/theora

[Raphael Canut]

Hi David,

Many thanks again for your answer!

Yes a proper website and demo pages are very needed, not to say mandatory
to attract attention and contribution.But I don't have the skill and
software to do graphical design, and as I want to do professional looking
website and demo pages, I have then to put it on pause and search for a
solution...

I am also currently very involved in quality improvement and high
compression, it takes a lot of time, so I am focusing on it for now.

Yes the NHW Project is a labor of love for more than 11 years now, and I
had to have a lot of motivation all these past years because the industry
was really not interested in the NHW Project... Let's hope that with the
recent quality improvements, it will change, it will start to interest some
people and maybe we'll find a niche for the NHW Project, would be so
great... but very not sure...

Many thanks again for your support, it is much appreciated!

Cheers,
Raphael

Post by David Willmore
I don't mind at all. I imagine there are more people like me out
there who have been following your work, but haven't said anything.
I agree that a series of pages explaining how things work may lead to
more contributions. At the very least it would lead to more people
being better educated on how wavelets work in image compression. But,
this is your project, if you don't have the time or interest, don't
feel compelled to do it. Clearly, NHW is a labor of love for you and
I wouldn't want that to change!

Post by Raphael Canut
Hi David!
Many thanks for your very encouraging and kind feedback!!! (Hope you

don't

Post by Raphael Canut
mind that I made your post public on the Theora channel).
I try to be the more sincere as I can in my posts.Some people suggested

me

Post by Raphael Canut
to make demo pages to explain how the codec works step by step, and I
realize that it would be good, because certainly it will show that the

NHW

Post by Raphael Canut
codec is not that complex and it will be more easy then to contribute for
people who want and have time.
For now, I focus on high compression settings, for me it's very complex,

but

Post by Raphael Canut
currently high compression settings are mandatory for a new image/video
codec.
Many thanks again for your feedback!
Cheers,
Raphael

I enjoy reading your posts and always find them educational. I don't
have anything to contribute other than my appreciation. Thank you!

Post by Raphael Canut
Hello,
I have re-tested -l4 high compression setting and it's clear that it lacks
of precision on degraded, rather blurred images. So I don't know if it is a
good idea to base the other lower quality settings (-l5,-l6,...) on

-l4

Post by Raphael Canut

Post by Raphael Canut
setting.I have tested the NHW codec against x265, x264, Daala, WebP,
Rududu,
DLI and it's clear that at high compression these very good codecs

have

Post by Raphael Canut

Post by Raphael Canut
more
precision.But they also tend to blur/decrease neatness-sharpness of image
where the NHW codec would enhance image neatness/sharpness.But that's
right
that on degraded, blurred images where there is no neatness to

enhance,

Post by Raphael Canut

Post by Raphael Canut
the
results start to be not good and lack of precision.So for now, I
recommend
to use -l4 setting on well-defined, good quality images.
I have also "simply" increased the wavelet coeff threshold of the dead-zone
for -l5 setting (to save -5Ko), but the quality drop is too big for now.The
other codecs (above) don't have that drop in quality from -l4 to -l5
compression.So apparently, there needs to be more smartness to pass

from

Post by Raphael Canut

Post by Raphael Canut
-l4
to -l5 with the NHW codec.I'll try to work on this, but it's extremely
complex, for me...For example, Rududu doesn't have this problem, the codec
source code has been made available recently on GitHub, I have started to
study it and it is very powerful: advanced SPIHT tree-like coding, advanced
variance-based context modeling, advanced range coding, and other things...
So to sum up, all the very good codecs quoted above have a very good
precision at high compression but they tend to decrease
neatness/sharpness,
on the contrary of the NHW codec, but on degraded images where there

is

Post by Raphael Canut

Post by Raphael Canut
no
neatness/sharpness, the NHW codec then lacks of precision...
Last but not least, for now the big advantage of the NHW codec I see,

is

Post by Raphael Canut

Post by Raphael Canut
that it is very fast.The NHW codec is way faster than the codecs

above,

Post by Raphael Canut

Post by Raphael Canut
for
example it is still now H.264 technology that is used in mobile

devices,

Post by Raphael Canut

Post by Raphael Canut
and
the NHW codec compared to ultra-optimized x264 is x3 times faster to
encode
and x1.2 times faster to decode, that is to say that with good
optimization
(SIMD, multithreading) the NHW codec will be x12 times faster to

encode

Post by Raphael Canut

Post by Raphael Canut
and
x5 times faster to decode than x264...
Again working on high compression is very complex, and any help would

be

Post by Raphael Canut

Post by Raphael Canut
very welcome.And I know there are experts out there, more skilled than me,
that could have very good ideas for the NHW Project, if they could

find

Post by Raphael Canut

Post by Raphael Canut
the
time...
Do not hesitate to contact me!
Many thanks!
Cheers,
Raphael
_______________________________________________
theora mailing list
http://lists.xiph.org/mailman/listinfo/theora