Afrikaans
Akan
Albanian
Amharic
Armenian
Azerbaijani
Basque
Belarusian
Bemba
Bengali
Bihari
Bosnian
Breton
Bulgarian
Cambodian
Catalan
Cebuano
Cherokee
Chichewa
Chinese (Simplified)
Chinese (Traditional)
Corsican
Croatian
Czech
Danish
Dutch
English
Esperanto
Estonian
Ewe
Faroese
Filipino
Finnish
French
Frisian
Ga
Galician
Georgian
German
Greek
Guarani
Gujarati
Haitian Creole
Hausa
Hawaiian
Hebrew
Hindi
Hmong
Hungarian
Icelandic
Igbo
Indonesian
Interlingua
Irish
Italian
Japanese
Javanese
Kannada
Kazakh
Kinyarwanda
Kirundi
Kongo
Korean
Krio (Sierra Leone)
Kurdish
Kurdish (Soranรฎ)
Kyrgyz
Laothian
Latin
Latvian
Lingala
Lithuanian
Lozi
Luganda
Luo
Luxembourgish
Macedonian
Malagasy
Malay
Malayalam
Maltese
Maori
Marathi
Mauritian Creole
Moldavian
Mongolian
Myanmar (Burmese)
Montenegrin
Nepali
Nigerian Pidgin
Northern Sotho
Norwegian
Norwegian (Nynorsk)
Occitan
Oriya
Oromo
Pashto
Persian
Polish
Portuguese (Brazil)
Portuguese (Portugal)
Punjabi
Quechua
Romanian
Romansh
Runyakitara
Russian
Samoan
Scots Gaelic
Serbian
Serbo-Croatian
Sesotho
Setswana
Seychellois Creole
Shona
Sindhi
Sinhalese
Slovak
Slovenian
Somali
Spanish
Spanish (Latin American)
Sundanese
Swahili
Swedish
Tajik
Tamil
Tatar
Telugu
Thai
Tigrinya
Tonga
Tshiluba
Tumbuka
Turkish
Turkmen
Twi
Uighur
Ukrainian
Urdu
Uzbek
Vietnamese
Welsh
Wolof
Xhosa
Yiddish
Yoruba
Zulu
In this lesson, we're going to learn how to do image stabilization with a track.
OK, so I'm just starting off here in 07_begin,
which is just going to be a blank Nuke script.
So I want to go ahead and drop in the footage we're going to be using.
So let's go ahead and grab your reference files folder and just grab that
image stabilize folder and drop that onto your node graph.
And that's going to give you the footage we're going to work with for
the remainder of this course and for this lesson.
So go ahead and plug your viewer in.
And we'll go ahead and hit Play just to cache those frames
and take a look at what we've got here.
So we're going to be doing a lot with this clip.
In some later lessons, we're going to learn the last type of tracking,
which is going to be a four point track,
and we're going to learn how we can use that with a corner
pin to replace the screen on this phone.
But the first thing we need to do, before we get to that point,
is to stabilize this shot.
So you can see that the shot was shot with a hand held camera because we've
got quite a lot of movement around here on the edges.
So we're going to get a cleaner looking shot and also a
better track eventually for that four point track on the
phone by stabilizing the image first.
So let's go ahead and get ready to do that.
And I want to show you an additional technique before we hook up our tracker.
And that is using color correction to help bump up the
contrast in your shot and therefore make your tracker have a
little bit more accurate of a track.
So let's go ahead and just come down here to our node graph,
hit the Tab key, and we'll type in color.
You just want to choose the Color Correct node.
And you can just drop that between your read and your viewer node.
And then let's go ahead and just come over here to Contrast and pull that up.
Now,
you don't want to do it too high because if you
blow these colors completely out,
then Nuke will actually have a worse time tracking.
So you just want to bump up the contrast a little bit,
just so kind of in this area over here we've got just kind
of a better area between the chair leg or the coffee table
leg and that black area.
And then also let's track, let's say, this part right here,
maybe this grout line.
This is a great place of contrast right there.
And because we've got this nice intersecting line,
that's going to give us a really good pattern to track.
Any time you have intersecting lines like that,
it really gives you a great reference point.
So for a image stabilization like this,
you're definitely going to want to use two points.
You can do a one point track,
but because it was handheld you want to get that rotation
information in there to help correct.
So I'm going to go ahead and close out the color
correct now and we'll add our tracker.
So you can just hit the Tab key down here in your
node graph and type in tracker.
And that's actually going to drop the tracker on after the color correction,
which is what we want because we want that color
correction to be affecting the tracker,
affecting the pixels that the tracker's looking at.
So let's go ahead and zoom in up here on our leg of our
coffee table and we'll add a track.
And I'm just going to position this.
Let's just kind of have it right here in the middle where
this shadow kind of turns into a V.
So kind of just right about there.
Or you could maybe even come up here to kind of the tip of that.
That would work as well.
Any time you just have a feature that has a very stark contrast right around it,
you can kind of see this area might here not be as good because
the pixels around it might be a little closer.
Anytime there's more contrast from pixel to pixel,
you're going to need the best information.
So we'll use this for our first track.
And you want to pull this over here to the side so
you can see your transformation, rotation,
and scale.
We'll just do the translation.
And then let's move this over here and zoom in on this grout line.
Notice if you zoom in first before adding a track,
if I just hit a track, it's going to add that right in the middle of the viewer.
So it's a lot easier to add or to move that closer to where you want it
if you're zoomed in that position in the first place.
OK, so I'm going to pull this down.
We don't need that pattern area to be quite so large.
And let's go ahead and turn on the translation,
rotation, and scale for this tracker.
Then let's zoom back out.
And I have a feeling this one's going to be pretty easy
compared to some of the other ones we've done,
especially since we added this color correction.
So let's go ahead and we'll just Shift-Select to both of those tracks.
And it looks like this time since we've started on the 38th frame,
we probably should have gone back to the first frame.
But that's OK.
We can just track backwards and then come back to frame 38 and track forwards.
So this time let's just hit this button right here that says Track To Start.
And that's just going to kind of automate that for us.
That looks like, for the most part,
that's really staying nicely in place because,
honestly, it's not that bad of a shaky image.
So in fact, I think with both of these tracks selected, we can just watch this.
And just kind of looking right here,
this really looks like it's staying in place very nicely.
So let's just come back over here to frame 38 and we'll hit Track To End
and finish up that track all the way through to the end.
OK,
so right here we've run into a little bit of an issue where our
timeline is only showing 44 frames but clearly that tracker was
trying to continue to go past that 44.
So we're going to come over here to where we see
global and change this to Input.
And basically that's going to give us the full amount of
frames that we have for this shot.
So let's continue to track the rest of this shot.
I'm just going to come back here to 74,
which is the final frame that the track stopped on.
And then we'll just tell it to continue to track to the end.
And it'll track all the way to 297.
So that's quite a few frames.
That's a really long track.
But again, because we don't have a lot of movement in our shot,
that's going to be just fine to deal with even if it is a little bit
longer than some of the other clips we've worked with.
So while this is finishing up,
let's go ahead and talk about what we're going to be doing next.
So just like before, whenever we get our tracking data,
we need to apply it to something to move the image.
So in this case, we're not really going to have a second thing to move.
So we've used our tracking data so far to move something around the shot.
But basically, in this scenario,
we want to take the tracking data and use it on the image
itself in order that it makes the image basically move
backwards of those movements.
So it ends up just kind of counteracting that and what you
end up with is a still looking shot.
So the way that we start to apply this data is slightly
different from what we've done in the past.
So now that this is finished tracking, let's go ahead and apply this.
So come over here to your Transform tab.
And what we're going to do is to kind of look at this transformation.
So right now the transformation is set to None.
So this is usually going to be when you're calculating your track,
you want this to be set to None.
And then we can come in here and change this to the stabilize function.
So go ahead and click that.
And because we did a stabilization with two points,
we'll just choose this first option, not stabilize one point.
So just click stabilize.
And believe it or not, that's all you have to do to apply that data to itself.
So now let's hit the Play button again.
And you're going to see two things happening.
Now our image looks much more still.
We don't have that crazy movement that we had earlier.
But we also get this black edge here.
So that is basically what happens whenever you move an image in this way and
basically what's happening is it's rotating down or moving down to counteract
that rotation and transformation information.
And then the black outside of the image is being shown.
So that's not what we want.
What we'll have to do to counteract those black areas is to scale our image up.
And one other thing that you can do before you change that is to make
sure that the reference frame is set to the right frame.
So right now it's still fixing this because it's very close.
But you'll notice that our reference frame is 0 and
this reference frame is set to one.
So you want it to be from the first frame.
So I'm just going to go ahead and change that to 0.
And that's actually going to help it be a little bit closer to what it
really should be doing to counteract that frame by frame.
And if your sequence that you're working with,
if you're not using our project files,
starts on frame one then you would leave that reference frame to frame one.
It's basically whatever first frame you have there.
And it's usually going to be a 0 or a one.
OK,
so now let's go ahead and fix the scale of this so that we
don't get those black bars on the edge.
Now,
what I'm going to do is kind of scrub forward to a
place where I see the most black.
Looks like that's mostly happening here at the top.
So we also get a little bit down at the bottom here towards the end.
So maybe right about there, actually.
That's quite a bit.
So let's go ahead and have our transform node to scale this up.
So what I'll do is just kind of zoom out a little bit.
And also, at this point we don't need that color correct node,
so you can just select that and delete it,
because we already have all our tracking data.
That's just going to make it look a little bit less blown out.
And you can go ahead and select that tracker because we want the
Transform node to appear between that tracker and the viewer.
So we'll just come over here to our Transform button,
choose the Transform node, and that'll drop that in there for us.
And you can see that's right there in the middle.
Perfect.
And let's go ahead and bump up our scale a little bit.
So looks like right in there on 106 pushes that black right outside the edge.
And let's just go ahead and play through.
Just make sure that those dotted lines you see don't go on the
inside of this solid gray line that's your frame.
So one of the drawbacks of image stabilization is that
you're going to lose some resolution.
So it's not going to be actually a different size.
It's still going to be your 1280 by 720.
But the pixels that are there to make that up are going to be stretched a
little bit further than they originally were intended.
Now again, the scale is only 1.
06, so it's not a huge change.
You're probably not going to notice a huge difference.
But it just kind of depends on the shot.
So if you end up shooting things bigger than they need to
be and then doing your stabilization,
then you can scale them down and you'll preserve more data.
That's actually going to be better to do it that way.
So if you shot it at 1080 and your final output is going to be 720,
you would want to do your transformation in between that change.
So this is starting to look pretty good.
So we've got our image stabilized and the tracker was set up.
We've got that transform node in there just to bring the scale up.
Really, really looking nice.
So now that our footage has been stabilized,
we're ready to learn how to do a four point track to
replace this screen in the phone.
And we're going to learn how to do that in our next lesson.
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.