High ISO, Sharpness & Autofocus – Blackmagic Pocket Cinema Camera 4K – Q&A Part 1

High ISO, Sharpness & Autofocus – Blackmagic Pocket Cinema Camera 4K – Q&A Part 1


hi Carl here form proav TV and welcome to
the first one of our test videos for the Blackmagic pocket Cinema Camera 4k now
we shot with this and released a video recently up in London and thank you so
much for everyone who watched that the reception of something that video was
absolutely fantastic the amount of you commenting what you thought of the
camera down in the comment section down below it was great to see whether you
liked the camera or whether you don’t and a lot of this video and the
subsequent test we’re going to do have been based on what you’ve all been
saying down in the comment section and what you’ve been asking for us to test
out so if you’ve got more things that you want to test we’ve still got lots of
tests left to do make sure you leave a comment in this comment section down
below and let us know what you want to see for this video though what we’ve
chosen to focus on is the high ISO performance the raw versus the pro res
and all the much benefits to shooting in the raw formats things like the data
rates the different gamma profiles various little bits and bobs like that
like IR pollution for example because the original camera was quite prone to
that one so we wanted to test this one all little things like that so let’s
waste not waste any time and let’s jump straight into it so black magics cameras
have always had a bit of a reputation for not being fantastic in low light and
this is looking like it is the first camera to change that so what we did is
we did one of our high ISO tests the way we like to test high ISO is to have a
shot fairly darkly lit and maintain the lighting and the exposure as we ramp up
the ISO and the way that we do that is by dimming the light until we can’t do
it any further because the ISO is booting up so high and then we start
stopping down the aperture so we try and do as much of it as possible with the
actual lighting to maintain a constant the way we’re maintaining a constant
exposure is with waveform monitor and the color chart which you see in these
shots so we’re keeping the white color chart there at exactly the same level
for each of them so the main thing to pay attention to here is the color
chucks the exposure will go up and down a little
on the background so we’ve cropped here right into the color chart and we’re
comparing raw performance on the left and ProRes on the right 400 I can’t
really see much difference at 1250 the raw does seem a little bit noisier than
the pros this is what you’d expect war formats are normally slightly noisier
because it leaves all denoising until post 3200 I can definitely start to see
a bit of noise and of course it’s 6400 the cinema 10g is clearly noisier is
interesting to see just how well actually the pro res is holding up
though it’s noisy but it’s not too bad at all so this is the – Jun native ISO
so there’s 400 on the right and 3200 on the left the 400 is definitely cleaner
there is some noise to the 3200 now one thing I wanted to clear up a little bit
was your native ISO cameras there’s been a little bit of a misunderstanding I’ve
been seeing quite a lot online but what that actually means and the effects that
you’re going to see because a lot of people are saying well hang on the 3200
is also a journey a native ISO as well as the 400 and yet the 3200 is noisier
what’s going on there meant to be the same really all the dual native ISO
means is that there’s two sets of analog circuits behind each pixel on the sensor
and so without applying any digital gain you can take a reading at either 400 ISO
or 3200 ISO so for normal cameras with just 1 joule native ISO you take that
reading and then to get a high or lower ISOs you add gain to get higher or you
have negative gain to go lower and so as you get a negative gain it’s going to
make the signal slightly cleaner and less noise free noisy and as you bump it
up it’s going to add noise with the digital gain and as you go up or down
that’s going to limit the amount of dynamic range that you would normally
see so all the dual native ISO really means is that those points no gain is
being applied there’s no guarantee that that analog circuit from that ISO and
that unlocks okay put that ISO adjust as noise free as the other in fact nearly
always they weren’t there on the Jewish 5s that weren’t on the
VariCam and other cameras like that we’ve kill native ISOs and so I was
never really expecting them to be on this one so it’s not that much of a
surprise the 3200 is a little bit noisier than the 400 but what it is
going to mean is that you get fantastic dynamic range whether you’re in a
low-light scene at 3200 or whether you’re in broad daylight at ISO 400 one
thing we were asked a lot is how sharp 120 frames a second slow motion would be
so this is a 1080p shot windowed in a 120 frames a second we blown it up there
I was actually quite impressed by this there’s a reasonable amount of details
still there and I can’t see anything particularly nasty in fact this is 1080p
windowed on the left and 120 frames a second windowed on the right so that
both 1080p but ones at real-time and ones at 100 frames a second and actually
I can’t really tell much of a difference when you put it next to the 4k though
you definitely notice a difference the 4k is a lot sharper than the 1080p in
that crop now because it can crop in 1080p recording and it always cropped in
slow motion recording a lot of people have been asking how much that crops in
what does it do to your crop factor and what will it do to your lenses so if we
take a look at this shot here this is 1080p using the full size of the sensor
so it’s down sampling the full 4k sensor down to 1080p now this is 1080p with the
windowed option enabled so this is exactly the same field of view as you’d
expect when you’re shooting in slow-motion so and that works out to
about a two times crop so you can effectively double the focal length of
your lens when you want to shoot in 120 frames a second now we’ve done this
split-screen here to see if there’s much of a sharpness difference so we’ve
cropped into 150 times on a bush and we’ve matched the focal length in the
shot by moving physically further away when we’re windowing so we’ve got
exactly the same shot whether you’re in 4k 1080p
neppy with a window or 1080p 120 frames a second now what is interesting here is
just how sharp that 1080p is when you’re not using that windowed crop it really
is remarkably sharp it’s almost as sharp as the 4k so if you want don’t want to
shoot in 4k if you don’t want to deal with the 4k files you’re going to be
very happy with the performance of the 1080p mode and the 1080p using the full
sensor is significantly sharper than the 1080p using the windowed sensor
regardless of whether you’re shooting real-time or 120 frames a second so the original pocket camera suffered
quite badly from IR pollution when I our pollution is is it’s the infrared
pollution and how that is going to affect your image this particularly
shows itself when you’re using heavy levels of nd the reason for that is that
a normal ND filter will only cut visible light not infrared light so when you’re
shooting normally with that an ND filter all of the infrared light and all of the
visible light is going into the camera fine
then when you add nd you’re cutting down the amount of visible light that’s going
in and raising how sensitive or how light sensitive the camera is so that
you get the same results now that would be fine because you’re letting less
visible light in but you’re looking exactly the same amount of infrared
light in and so all of a sudden run there just being a little bit of
infrared light in your image there’s a whole load now this obviously different
sensors and different cameras are reacting different ways to this infrared
light and how this normally shows itself is a bit of a murky magenta tone in the
dark areas of your image particularly black clothing or as we see here on the
color chart the black parts the color chart you definitely see it now I don’t
have the right tools here right now to scientifically test this ideally you
want several sets of n DS you’ll want one that cut IR ones that don’t cut IR
you’d want fixed levels of nd so you know exactly how many stops you’re
putting in front of your lens I didn’t have access to any of that today when we
were doing the filming so what I did is I use the little variable ND filter from
tiffin which we actually use on our London shoot and so I did this little
split screen here outside very quickly went down and a color chart where on the
Left there’s no nd applied whatsoever and we were f-16 then working our way
along we’ve got a minimal amount of nd then about half way through though if nd
filters range and then all the way with the nd photo and white open at F at one
point to now actually I can’t see that many issues with the with IR here there
are definitely color shifts but this could have been down to so many factors
I’m not seeing that classic magenta tint in the shadow areas that you would
expect with IR pollution so I can’t scientifically say that this
doesn’t have any problems with the IR pollution but it does seem to be a lot
better than the original camera and for this with a variable ND filter from
tiffin doesn’t seem to be much of an issue we were getting a lot of questions on
data rates how long can you fit on a particular card when you’re in
particular formats things like that so what I thought I would do is I would
knock up this quick little spreadsheet here obviously this is a very boring
slide so I won’t show this here for too long but the full resolution of this
will be in our blog post and the link to that will be in the description the blog
post of course on our website now black magic don’t help the situation here
because they quote their data rates in megabytes a second
rather than normal megabits a second which might sound the same but they’re
very different things and normally in the video world we work in megabits a
second it might be 50 megabits the second broadcast quality or 300 megabits
a second as a codec something like that for example on the gh 5 which is coming
I’m sure many of you familiar with you can shoot it either 150 or 400 megabits
a second and you’ll need the faster SD cards to record the 400 megabits a
second so what I’ve done with this spreadsheet is I’ve taken each of their
code different codecs all of them at 30 frames seconds and I’ve converted the
megabytes a second value into a megabits a second value and then I’ve shown how
long you’re going to be able to record for on 128 gigabyte card 256 gigabyte
card or a 500 gigabyte SSD cuz I think those are going to be the really common
types of media that people are going to be using here now it’s really
interesting to note just how high some of these data rates are cinema DNG
lossless for example is crazy high on the data rate DCI 4k cinema DNG lossless
you’re looking at two thousand one hundred and eighty megabits a second
that is massive on the C 200 canons RAW format is 1,000 megabits a second and a
lot of people think that is far too much so I mean you’re if you want to shoot
lossless 4k law you really do need one very focused media indeed and two a lot
of storage space and a lot of processing power because those are very large very
heavy files you drop it down to 3 to 5 3 to 1 or 5 to 1 it becomes a lot more
manageable interestingly actually pro is 40 to HQ is larger
than five to one which I’m sure will actually to surprise some people for the
normal sort of levels I reckon most people are going to be shooting in
either the five to one law or a regular ProRes normal ordinary pro is 42 or 42
lights something like that where the data files become much more
manageable but they’re still high this is higher than the FS seven higher than
the c300 Mark to any of camp people that are used to cameras like that this is
high data rates of course if you already come from a black magic world with the
Ursa mini Pro you’re very familiar with this so if you want to delve into more
information on this data rate spreadsheet make sure you check out our
blog right so this is a big one obviously you’ve got war and you’ve got
pro res files and we’ve just looked at the difference in data rate between the
two of them so the big question really is is the rule worth that larger data
rate is it worth spending all that money on storage cards is it worth the extra
hassle in post-production now of course the answers yes they’re far better
quality files but for the majority of us is it going to make that much of a
difference so what we’ve got here is we’ve got raw compared to pro res
there’s one big shot you’re not really going to notice much of difference here
at all but what I’m going to do with these wide shots of London which I
filmed is I’m going to purposefully try and break the footage so we’ve got side
by side here we’ve got law on the right and pro res on the left this is pretty
much ungraded in resolve now I did the exposure right up to see into the shadow
areas and actually both are performing very very similar now I’ve done a crazy
grade where I’ve obviously changed all of the hues increase the contrast and
the saturation way beyond anything you would ever normally do and obviously
this is a horrendously ugly shot but this is purposefully to try and break it
and now here we’re starting to see a big difference when I crop in you can notice
lots of compression artifacts in the sky on the pro res and nothing on the Raw on
the right and also if you look at the buildings all the lines are much cleaner
on the war and it’s held all that information there
even though I’ve done such drastic drastic grading to it it’s holding up a
lot better and you can see it down here as well maybe to a slightly lesser
extent but all the lines and the hard refined edges on the buildings are
definitely cleaner and more accurate on the raw now this shot was quite
interesting because I’ve got some clipped clouds here I was quite pleased
with the dynamic range on this camera throughout all of my shooting and this
was one of the very few times I was actually able to clip the sky and clip a
cloud so I did it in raw and in pro res and here obviously I’ve horrendously
over contrast and saturated it to make it very obvious for you all on YouTube
but using the raw highlight recovery on the right here I was able to get all the
information back in that cloud was in the pro res file it is gone there no
matter what I do in resolve I cannot get that detail back on the right hand side
of that cloud whereas in the raw it looks absolutely great even though this
is a very extreme example you’re going to be able to take those benefits and
put them back into normal clips it’s just that you they’ll be much less
obvious to people looking online I wanted to make this as obvious as
possible for you all so that it’s very easy to see so the camera doesn’t have continuous
also focus it has partial to focus which you can activate with either a button on
the side or by touching what you want to focus on on the screen which is actually
quite useful I didn’t use it much on my shoe oh you’ve manually focused things
but of course because it uses a slower contrast based autofocus it’s not quite
as you would expect from some other cameras from Sony or Canon
so let’s look in more detail on what it looks like when you’re actually focusing
you can see as we tap on things it hunts in and out and as we show it larger here
and show you the actual footage you’re going to be able to see this even more
clearly it hunts backwards and forwards several times very quickly before
finally locking on to what is in focus it does seem to be very accurate and it
is getting it correct in these tests here but it is hunting like that so if
you want to actually include any of that footage
you definitely can’t use it while it’s focusing so there’s no chance of tapping
on things to pull focus between objects like we’re showing here it’s just not
gonna look good but there’s a one push get you in the ballpark for focusing
during you’re working very quickly it’s actually quite reliable and really quite
useful okay so there’s several gamma profiles
in the camera there’s film there’s extended video in this video so videos
on the Left extended video in the middle and film on the right film is
effectively their version of a log format and video is effectively their
rec.709 what extended video is is it’s it’s the normal video mode but with a
slightly more aggressive knee for the highlights and so it tends to compress
the highlights and give you the highlight performance or log but the
blacks and the saturation of rec.709 now here in this shot of me against the
window you’re really going to see the difference in film you can see all the
information you could color in the sky it’s not blown out in that tree in video
it’s completely gone it’s a white sky that tree looks pretty
ugly in the building doesn’t look great either
extended video it’s not that bright in the highlights but you’ve got very close
to the performance of the film but with the contrast and saturation of the video
profile now this outside one is actually very interesting because of what it’s
done to the sky if you’re going to use this as a gamma profile so that you
don’t have to grade your footage at all you might find some instances where it’s
a little bit less saturated than you would like although it looks a lot
better on the clouds and on my face there which is a little bit hot and
overexposed in this shot then the video wonders if you look at the saturation in
the sky and that blue it’s nowhere near as vivid and as vibrant as it is it
almost looks like a very subtly graded log shot so even though you shot in the
extended video you might still want to do a little bit of a saturation boost in
post I was asked many many times in the
comments about maría and aliasing and had we noticed any our shots and to be
honest know what we’ve done here though to test that is to do a very slow
panning shot of lots and lots of detail this is our Railway model here in the
show room lots of a lot of fine detail on the trees and lots of lines on the
lose of these model buildings and you can see here is I’ve cop in there’s
really not any issues here at all there’s no jagged ease no spinning
patterns nothing ugly here it’s looking pretty good and here’s a real world shot
from London and as I crop into that brick wall up in the top right-hand
corner I can’t see anything nasty there at all and again here with the bridge
once you crop in to that fine detail in the buildings in the distance I really
can’t see any issues here and if I was going to see those issues and aliasing
in any shot it probably would have been this one because water is notoriously
crop in here to these boats and the water rippling underneath them there’s
no purple fringing no dancing patterns this looks remarkably clean so that was
just the first batch of our tests like this we’ve got more ones planned we
definitely want to look at things like audio for example and there’s a whole
lot of other things in terms of the video quality that we also want to test
but if you want us to look at anything in particular let us know in the comment
section down below and I’ll do my best to look into it for you and I also try
to reply to as many of your questions down there as I possibly can
we’re also going to do a video comparing it to different cameras now obviously I
can’t include every single camera here I can only realistically include one what
I’ve got access to and two I’ve got to limit the physical amount otherwise it’s
going to take us forever to do the tests and it would be such a long unwatchable
video so I’m just going to choose the ones which I think I’m most relevant but
if you’ve got one you particularly want to see let me know in the comment
section down below so I’ve been really really impressed so far with
Blackmagic pocket Cinema Camera and it’s video quality let me know what you think
of the results of these tests down in the comment section down below and if
you want to pre-order yours and join the pre-order queue the links to our product
pages of course in the description thank you very much for watching and see in
the next video

Leave a Reply

Your email address will not be published. Required fields are marked *