<s>
SMPTE	B-Algorithm
ST	I-Algorithm
2117-1	I-Algorithm
,	O
informally	O
known	O
as	O
VC-6	B-Algorithm
,	O
is	O
a	O
video	B-Algorithm
coding	I-Algorithm
format	I-Algorithm
.	O
</s>
<s>
The	O
VC-6	B-Algorithm
codec	B-General_Concept
is	O
optimized	O
for	O
intermediate	O
,	O
mezzanine	O
or	O
contribution	O
coding	O
applications	O
.	O
</s>
<s>
VC-6	B-Algorithm
,	O
like	O
other	O
codecs	B-General_Concept
in	O
this	O
category	O
uses	O
only	O
intra-frame	B-General_Concept
compressions	O
,	O
where	O
each	O
frame	O
is	O
stored	O
independently	O
and	O
can	O
be	O
decoded	O
with	O
no	O
dependencies	O
on	O
any	O
other	O
frame	O
.	O
</s>
<s>
The	O
codec	B-General_Concept
implements	O
lossless	B-Algorithm
and	O
lossy	B-Algorithm
compression	I-Algorithm
,	O
depending	O
on	O
the	O
encoding	O
parameters	O
that	O
have	O
been	O
selected	O
.	O
</s>
<s>
Earlier	O
variants	O
of	O
the	O
codec	B-General_Concept
have	O
been	O
deployed	O
by	O
V-Nova	O
since	O
2015	O
under	O
the	O
trade	O
name	O
Perseus	O
.	O
</s>
<s>
The	O
codec	B-General_Concept
is	O
based	O
on	O
hierarchical	O
data	O
structures	O
called	O
s-trees	O
,	O
and	O
does	O
not	O
involve	O
DCT	B-General_Concept
or	O
wavelet	B-Algorithm
transform	I-Algorithm
compression	O
.	O
</s>
<s>
The	O
compression	O
mechanism	O
is	O
independent	O
of	O
the	O
data	O
being	O
compressed	O
,	O
and	O
can	O
be	O
applied	O
to	O
pixels	B-Algorithm
as	O
well	O
as	O
other	O
non-image	O
data	O
.	O
</s>
<s>
Unlike	O
DCT	B-General_Concept
based	O
codecs	B-General_Concept
,	O
VC-6	B-Algorithm
is	O
based	O
on	O
hierarchical	O
,	O
repeatable	O
s-tree	O
structures	O
that	O
are	O
similar	O
to	O
modified	O
quadtrees	B-Data_Structure
.	O
</s>
<s>
In	O
the	O
VC-6	B-Algorithm
standard	O
an	O
up-sampler	O
developed	O
with	O
an	O
in-loop	O
Convolutional	B-Architecture
Neural	I-Architecture
Network	I-Architecture
is	O
provided	O
to	O
optimize	O
the	O
detail	O
in	O
the	O
reconstructed	O
image	O
,	O
without	O
requiring	O
a	O
large	O
computational	O
overhead	O
.	O
</s>
<s>
The	O
ability	O
to	O
navigate	O
spatially	O
within	O
the	O
VC-6	B-Algorithm
bitstream	O
at	O
multiple	O
levels	O
also	O
provides	O
the	O
ability	O
for	O
decoding	O
devices	O
to	O
apply	O
more	O
resources	O
to	O
different	O
regions	O
of	O
the	O
image	O
allowing	O
for	O
Region-of-Interest	B-Algorithm
applications	O
to	O
operate	O
on	O
compressed	O
bitstreams	O
without	O
requiring	O
a	O
decode	O
of	O
the	O
full-resolution	O
image	O
.	O
</s>
<s>
At	O
the	O
NAB	O
Show	O
in	O
2015	O
,	O
V-Nova	O
claimed	O
"	O
2x	O
–	O
3x	O
average	O
compression	O
gains	O
,	O
at	O
all	O
quality	O
levels	O
,	O
under	O
practical	O
real-time	O
operating	O
scenarios	O
versus	O
H.264	B-Application
,	O
HEVC	B-Algorithm
and	O
JPEG2000.	O
"	O
.	O
</s>
<s>
A	O
variant	O
of	O
the	O
technology	O
optimized	O
for	O
enhancing	O
distribution	O
codec	B-General_Concept
will	O
soon	O
be	O
standardized	O
as	O
MPEG-5	B-Algorithm
Part-2	I-Algorithm
LCEVC	I-Algorithm
.	O
</s>
<s>
These	O
planes	O
might	O
be	O
RGB	O
or	O
RGBA	O
pixels	B-Algorithm
originating	O
in	O
a	O
camera	O
,	O
YCbCr	O
pixels	B-Algorithm
from	O
a	O
conventional	O
TV-centric	O
video	O
source	O
or	O
some	O
other	O
planes	O
of	O
data	O
.	O
</s>
<s>
The	O
standard	O
focuses	O
on	O
compressing	O
planes	O
of	O
data	O
values	O
,	O
typically	O
pixels	B-Algorithm
.	O
</s>
<s>
To	O
compress	O
and	O
decompress	O
the	O
data	O
in	O
each	O
plane	O
,	O
VC-6	B-Algorithm
uses	O
hierarchical	O
representations	O
of	O
small	O
tree-like	O
structure	O
that	O
carry	O
metadata	O
used	O
to	O
predict	O
other	O
trees	O
.	O
</s>
<s>
The	O
core	O
compression	O
structure	O
in	O
VC-6	B-Algorithm
is	O
the	O
s-tree	O
.	O
</s>
<s>
It	O
is	O
similar	O
to	O
the	O
quadtree	B-Data_Structure
structure	O
common	O
in	O
other	O
schemes	O
.	O
</s>
<s>
The	O
more	O
data	O
that	O
can	O
be	O
predicted	O
,	O
the	O
less	O
information	O
that	O
is	O
sent	O
,	O
and	O
the	O
better	O
the	O
compression	B-General_Concept
ratio	I-General_Concept
.	O
</s>
<s>
pixels	B-Algorithm
)	O
can	O
be	O
dequantized	O
at	O
different	O
resolutions	O
and	O
used	O
as	O
predictors	O
for	O
higher	O
resolutions	O
.	O
</s>
<s>
VC-6	B-Algorithm
is	O
an	O
example	O
of	O
intra-frame	B-General_Concept
coding	I-General_Concept
,	O
where	O
each	O
picture	O
is	O
coded	O
without	O
referencing	O
other	O
pictures	O
.	O
</s>
<s>
As	O
a	O
result	O
,	O
the	O
VC-6	B-Algorithm
bitstream	O
contains	O
all	O
of	O
the	O
information	O
for	O
all	O
of	O
the	O
planes	O
of	O
a	O
single	O
image	O
.	O
</s>
<s>
An	O
image	O
sequence	O
is	O
created	O
by	O
concatenating	O
the	O
bitstreams	O
for	O
multiple	O
images	O
,	O
or	O
by	O
packaging	O
them	O
in	O
a	O
container	O
such	O
as	O
MXF	O
or	O
Quicktime	B-Application
or	O
Matroska	O
.	O
</s>
<s>
The	O
VC-6	B-Algorithm
bitstream	O
is	O
defined	O
in	O
the	O
standard	O
.	O
</s>
<s>
Six	O
different	O
non-linear	O
upsamplers	O
are	O
defined	O
by	O
a	O
set	O
of	O
processes	O
and	O
coefficients	O
that	O
are	O
provided	O
in	O
JSON	B-General_Concept
format	O
.	O
</s>
<s>
These	O
coefficients	O
were	O
generated	O
using	O
Convolutional	B-Architecture
Neural	I-Architecture
Network	I-Architecture
techniques	O
.	O
</s>
