How To Edit 8K
Video - Threadripper + Ultrawide #Helpful Post
K
red code raw footage how do you
work
with files that are multiple
hundreds
of megabytes per second of
recording
with a complex and proprietary
compression
algorithm that allows them
to
retain effectively nearly the full
raw
quality of the footage while taking
up
a fraction of the space well
LG's
Monitor team seems to have an
interest
in this and the many
technological
challenges faced by
creative
professionals because they
sponsored
this video featuring their 49
WL
95 C otro wide monitor where I'll be
figuring
out just that so in the old
days
when anywhere from 4 K to 6 K was
cutting-edge
the solution was this $7000
red
rocket X card but in spite of it
using
either a custom ASIC or an FPGA so
both
highly specialized Hardware even
these
puppies were easily overwhelmed by
the
epic W and the red weapon featuring
reds
8k helium sensor so now this is
basically
an expensive paperweight ha
that's
the that's the joys of cutting
edge
technology right fortunately NVIDIA
has
been running around doing demos with
their
arty x-series graphics cards
showing
GPU accelerated real-time full
quality
8k playback so then is this the
new
way now $1300 is still a lot of
money
but it's a lot less than $7,000
could
this be progress there's only one
way
to find out
so
my first encounter with this monster
was
back at CES and while 49 inch ultra
woods
like this one so 32 by 9 aspect
ratio
look at this thing
have
existed before what makes this one
special
is that the pixel density since
its
effectively 5 K by 1440 is much
higher
than the ones that have existed
before
making it actually suitable for
creative
work so it has become apparent
that
I'm going to need to clear some
space
so it has become apparent that I'm
going
to need to clear even more space
wow
that's a lot of time line even with
my
task manager open here
okay
so first order of business is to
fire
up a plausible project so this is
an
episode of fast as possible and use
the
the normal workstations that we have
downstairs
so my CPU is a little better
this
is a 79 80 XE 18 core processor and
this
is a GTX Titan X Pascal so first
things
first we've got ourselves at
one-quarter
resolution you can see
scrubbing
around in the timeline nice
and
responsive playback basically
instantaneous
but whenever I'm not
paused
you can see it's quite fuzzy not
representative
of the finished product
so
let's crank it up to full so this is
a
K and this is working just fine no
problem
whatsoever and in fact while I'm
playing
it back we're sitting at a mere
10
to about 30% CPU usage and our GPUs
sittin
tight at 25 to 30% so at first I
was
kind of confused by all of this it's
been
a while since I've looked into a K
performance
on the timeline and I was
thinking
well hold on a second is there
nothing
special about these r-tx cards
that
makes them suitable for this why
does
Nvidia keep talking about this and
then
I
remembered something we shoot most of
our
projects at a much higher
compression
ratio than what they would
use
while shooting a VFX heavy Hollywood
production
for example so our data rates
are
much lower so let's go get some
seven-to-one
footage so let's start with
a
sample from red here this is a 7 to 1
compression
ratio 8k clip that was
recorded
at 24 frames per second so
let's
go ahead and fire this bad boy up
hit
that playback button it oh we are
looking
at spikes of up to 95 percent on
our
GPU and CPU usage that is pinned at
100%
while trying to playback that clip
that
is a stark contrast to the I think
we
use 2001 is that right David to the
20
to 1 footage that we capture and
that's
not even the worst of it here's
some
8k 7 to 1 footage at 30fps that we
captured
let's go ahead and drag that
onto
the timeline
yeah
whoof not quite
yeah
he moves like that sure why not so
what's
weird though is that even with
the
heaviest footage that we could throw
at
our workstation our GPU is actually
neither
sitting idle nor completely
maxed
out so it's being used to handle
the
debayering process which
reconstructs
the full color image that
we
see here from the samples that are
collected
by the image sensor but our
CPU
is still in charge of the entropy
decoding
which is a process that allows
the
most common input signals to be
substituted
for much shorter ones saving
space
and wavelet decoding which frankly
I'm
not familiar enough with to offer
even
a basic explanation and the
Wikipedia
article wasn't much help
anyway
though so it's maybe our problem
here
is we just need more
cpu
horsepower to keep our GPU fed AMD
to
the rescue then why don't we try
going
from 18 to 32 processor course
building
improvements Oh mission failed
okay
it worked didn't do anything and
interestingly
our GPU usage is higher
near
a hundred percent so this is kind
of
fascinating I've said it many times
before
that every system is effectively
bottlenecked
because
there's something that is
slowing
down the process or else it
would
be infinitely fast but this is
like
double bottleneck
it's
like they're perfectly matched to
each
other but still not enough for the
task
this is great to look what happens
to
Timeline scrolling when my CPUs at a
hundred
percent usage so while I was
troubleshooting
to prepare for this
video
I actually spent some of my time
with
the r-tx twenty eighty rather than
the
r-tx twenty atti
because
what I thought was that this
functionality
had something to do with
their
new and vank engine but actually
it
doesn't in fact there's nothing about
the
tensor cores or the ray-tracing
cores
or anything like that that makes
the
r-tx cards so good at this it's just
the
sheer frigging processing power of
all
those CUDA cores so we are gonna
jump
for 8k seven to one footage
straight
to the RT X xx atti and you
might
think to yourself well gee Linus
that's
a 1,200 to 1,300 US dollar
graphics
card of course it can handle it
but
that's not necessarily to be taken
for
granted I mean we installed CPUs
that
cost up to two thousand dollars and
they
couldn't do it it's all about
having
the right tool for the job okay
everything's
working normally here we go
this
is the moment
I
mean now that is kind of gotten its
act
together that's a lot better we are
now
at a hundred percent CPU 95% GPU but
we
are still dropping frames so that's
it
then no CPU on earth can handle this
30
frame per second 8k seven to one read
code
raw footage at least not as long as
it
has to handle the entropy decoding
and
wavelet decoding what if we could
offload
those to the GPU we are going to
need
a beta version of Reds decoder that
unfortunately
isn't supported by Adobe
Premiere
at this time fortunately it's
built
into the beta of redson DX pro and
we're
gonna start with the 24 FPS
footage
now it's a little flaky right
now
I'm actually not sure if this is
going
to work but let's go ahead and
give
it a shot so we're gonna enable GPU
decode
and it's flaking out on me so
it's
running it like eight frames per
second
fortunately the fix right now
seems
to be to just toggle between image
pipelines
and there it is smooth 8k
playback
that hitch by the way is the
clip
restarting but that's not even the
most
impressive part so our CPU is
sitting
at just 10% utilization and our
GPU
is up at 77 80% utilization so we
potentially
have some Headroom to spare
so
our overall usage for the system is
way
down like 10% of our 32 cores we
could
be doing this on a more human
workstation
you know one with six or
eight
cores by offloading this work to
the
GPU because the thing to remember as
well
is that when we were pinning both
our
CPU and GPU at a hundred percent
that
was just playback
we
weren't even applying any real-time
effects
like denoising the footage or or
anything
like that we were at the limit
now
we have Headroom to play with
let's
go ahead and try our 30fps footage
though
that's actually working a little
bit
better than I expected when I the
first
time around I actually had to
overclock
the GPU a little bit to get
this
quite so smooth but you know what
I'm
taking it so there it is our GPU
CUDA
usage jumps up near 90% but our CPU
stays
at 10% and we are smoothly playing
back
seven to one 8k red footage at
29.97
frames per second we get some
applause
sound effects in the video or
something
I don't know
this
is crazy so all that's left now
then
is to thank LG for sponsoring this
video
featuring their 49 WL 95 C W
ultrawide
monitor I mean I think it's
pretty
obvious why they wanted us to
feature
this in this video because it's
designed
for content creators like
photographers
filmmakers music producers
and
more who can really benefit from
this
kind of a canvas being able to work
on
more than one thing simultaneously or
just
be able to manage and see extremely
long
timelines without zooming out
giving
them a finer degree of control
and
it's not even just for creators but
office
work would also be a great option
for
someone like a programmer stock
trader
or really anyone who needs a ton
of
side-by-side screen real estate
because
it's basically exactly the same
as
having to 2560 by 1440 monitors side
by
side except that you have no bezel in
between
LG's also been hard at work on
software
that allows you to take two
inputs
select your desktop in your
laptop
and seamlessly share your mouse
and
keyboard between them just like
moving
between them so you can control
two
systems at once
it's
got full support for USB type-c
including
power delivery so you can use
it
as a single cable docking solution
for
your laptop and it's got support for
HDR
10 with
-
ten watt speakers built-in in fact it
was
didn't realize they were gonna be
that
loud that happened that happened a
couple
times during the video so check
it
out at the link in the video
description
and I guess that's pretty
much
it thanks for watching you guys if
you
disliked this video well you can hit
that
button but if you liked it hit like
get
subscribed or maybe consider
checking
out where to buy the stuff we
featured
at the link below also down
there
is our merch store which has cool
shirts
like this one and our community
forum
which you should totally join
No comments:
Post a Comment