subscribe to rss feed
subscribe by email

Puppet Kaos -

Case Study: Putting a Character into a Car with Green Screen and Garbage Matte Effects

We now know how to do the green screen effect and garbage matte effect; it’s time to combine them. Here’s how I put Moostifer into a car in Episode 12: Daylight Saving. Again, I used Adobe Premiere Pro but you don’t have to for this to work. This is the final result we want:
cow_car_final.jpg

First we need the raw footages. The first picture is the car. It’s a still picture, not a video. The second picture is a video that we filmed in front of a green background for green screen effect.
cow_car_car.jpgcow_car_cow.jpg

Now, resize both images so the sizes match up. And then remove the green background using green screen effect. Notice that there are some green parts on the corner that we didn’t do a good job removing. Also Moostifer is sort of hanging in the air because the bottom of the video is lower than the driver’s window frame. That’s fine because we’ll soon use a garbage matte to fix it.
cow_car_green.jpg

Now, apply the garbage matte. When you draw the matte, make sure this shape is approximately the shape of the car window. This will create an illusion that Moostifer is actually sitting in the car, because he is being blocked by the windshield, the car door, etc.
cow_car_garbage.jpg

And that’s pretty much it. Moostifer is now sitting in the car! Now just a little bit of fine-tuning: I adjusted the colors a little bit to add a red tint to match our background graphic. I also adjusted the brightness of the picture to make the character darker since this is supposed to be a night shot. And here’s, again, the final result:
cow_car_final.jpg

And there you have it!

(Of course, you can always get the puppet into a real car and then film it. I didn’t feel like going through the trouble of finding a location for just one shot though.

Related Posts:
Editing Trick: Garbage Matte
How to do a green screen effect: the theory
How to do a green screen effect: a case study
Episode 12: Daylight Saving

Editing Trick: Garbage Matte

What is a “garbage matte”? According to the all-knowing Wikipedia:

A “garbage matte” is often hand-drawn, sometimes very quickly made, and can be used to exclude parts of an image that another process, such as bluescreen, would not remove. The name stems from the fact that the matte removes “garbage” from the procedurally produced image. This “garbage” might include a rig that is holding a model or the lighting grid above the top edge of the bluescreen. Garbage mattes can also be used to include parts of the image that might otherwise have been removed by the bluescreen, such as too much blue reflecting on a shiny model (”blue spill”).

In this case, we start out with this raw footage, designed to be a green screen shot:
garbage_matte_raw.jpg

It is filmed in front of a green background so we can apply the green screen effect to replace the background:
garbage_matte_green.jpg

Notice that on the left edge, there’s a little bit of green that wasn’t removed very well. And then there’s that black thing on the lower righthand corner…. that would be my head (more about my head later… or not). Those are the two things that we do not want to see in the video.

So what do we do? Garbage matte comes to rescue!

With garbage matte, we can define an area that we want to keep in the final picture and leave out the rest. In Adobe Premiere Pro, this is done by first finding the garbage matte effect under Video Effects => Keying, and drag it into the Effect Controls Window:
garbage_matte_eff.jpg

And then in the Effects Controls Window, you can start changing the coordinates of each of the four corners that defines that garbage matte region:
garbage_matte_efc.jpg

And if you had “Garbage Matte” selected (highlighted), you can just drag the four corners to define the desired region. Now we see Bobby, who’s within the trapezoid-shape region we defined, and we no longer see the two things we were trying to remove!
garbage_matte_matte.jpg

Garbage mattes have other uses too. For example, in Episode 12: Daylight Saving, I used the trick to put Moostifer into a car. More about that later…

Related Posts:
How to do a green screen effect: the theory
How to do a green screen effect: a case study
Episode 12: Daylight Saving

Editing Trick: Alpha Glow

alpha_glow_result.jpg
Ever wonder how the characters were able to toss that ball of “daylight” around in Episode 12: Daylight Saving? Okay, maybe you didn’t, but here’s how to do it just for the record and for the people that care. :-P

I used Adobe Premiere Pro, and for good results, your best bet is probably to look for third-party plug-ins that specifically do glowing effects. In this video, however, I chose to use the built-in Alpha Glow effect. This effect adds a glow to specific objects in a video, be it some kind of text, shape, or a picture.

Note that for the ball of “daylight” effect, you really only need to draw an oval and make it glow. In this example, I’m going to have other shapes like rectangle, triangle, and text glow as well. They are just there to serve as examples.

alpha_glow_compare.jpg

First create a caption file and put things that you want to glow in it, like the left half of the above picture. (The right half is the final result after applying the Alpha Glow effect.)

And then go to the Effects window, in the effects window, under Video Effects => Stylize, find Alpha Glow and drag it into the Effect Controls window of the caption/title file:
alpha_glow_eff.jpg

And then in the Effect Controls Window, you will see the different variables. The glow is currently the default value: gray. So although the quarter circle is yellow and the text is green, they are still glowing the same gray. For best effect, you might want to tweak the color of the object and the glow to match each other. You can also change the “Glow” option to make the emitting rays longer or shorter.
alpha_glow_efc.jpg

And that’s how the glowing effect is done. And to have the characters hold and toss the glowing objects, it’s just like making them hold and toss non-glowing objects. Just carefully adjust the positions and sizes of the caption file to match the characters’ hand movements. The glowing effect is associated with the caption file so you don’t need to worry much about it.

And that’s how you throw daylight around. :-)

Related posts:
Episode 12: Daylight Saving

Editing Trick: How to Turn Day into Night (by adjusting brightness and contrast)

I’m kinda behind on writing up these editing tricks, aren’t I? Anyway, here is one trick that’s useful when you want to film footage in daylight and then pretend it’s filmed at night or in a dark area. I’ve used this trick in both Obscure UCLA Facts You Didn’t Know, and Episdoe 12: Daylight Saving.

Let’s look at the before and after pictures. The one on the left was the raw footage. It was filmed in front of a green background because it was originally planned out to be a green screen shot. I later abandoned the idea. The one on the right is the finished product, as seen in the video.

day2night_compare.jpg

The effect is done by adjusting the video’s brightness and contrast settings. I did mine in Adobe Premiere Pro, but most editing software should have this option.

First, in the Effects window, find Video Effects => Adjust => Brightness and Contrast. Drag it into the Effects Controls window.

day2night_effwin.jpg

And then in the Effects Controls window, adjust the brightness and contrast settings. I changed the brightness to -100 and the contrast to 13. So basically I dialed down the brightness and upped the contrast.

day2night_efcwin.jpg

We are trying to make the image much darker, so we dial down the brightness. This will give you the picture on the left. Now, why did I turn up the contrast? This is actually not that important of a step, but I’d like to get more contrast between the dark background and relatively bright character to emphasize the idea that my character is in a dark room. When we do that, we get this image on the right. The difference is not huge but you can see the difference when they are side by side:

day2night_compare2.jpg

Now you can turn day into night (just like Mr. Burns)!

Related Posts:
Filming in the Dark or Filming in the Light
Obscure UCLA Facts You Didn’t Know
Episdoe 12: Daylight Saving.

Filming in the dark or filming in the light

Some scripts call for things that happen at night. Sometimes I film them in the dark, and sometimes I don’t film them in the dark, but instead make the images darker in editing. So what’s better? Let’s check out my little experiments.

In Episode 12 - Daylight Saving, I experimented with filming in the dark for the power outage scene in the beginning. This is the outcome.
filmed_in_dark.jpg

And in Obscure Facts about UCLA You Didn’t Know, we filmed the tunnel scene which takes place at night, but we filmed it in broad day light because we were there in the morning. So I had to take a bright and light picture and make it dark in editing. This is the result.
filmed_in_light.jpg

Now, which one is better? I’d say they each has their own pros and cons.

Pros for filming in the dark
In the first picture, the darkness is authentic because it was indeed filmed in the dark. When we are in the dark, we don’t see colors as well as we do in daylight. This is because we have two kinds of photoreceptors in our eyes, rods and cones. Cones are the ones in charge of gathering color information but cones don’t work well in darkness. Rods are the ones that work better in low light situations but they don’t gather color information. So in the dark, we see shapes and movements, but not so much colors. When this translates to video, it means you’d want to get a dark picture without so much color information. What the camera captured was similar to what our eyes saw in this case.

Pros for filming in the light
In the second picture, this image is done by taking a video in broad daylight and then digitally decreasing the brightness level in a video editor. Since it’s not filmed in the dark, there are more richness in the colors. Now, the richness in colors is good for a better contrast of the foreground characters and the background. This is good if you want to see more clearly what the characters are doing, and it makes the foreground characters stand out more. Sometimes, having some light in the foreground can be used as a trick to emphasize how dark the background (and hence the overall picture) is. This shot below is also filmed in daylight but darkened in the editor. You can see a pretty cool contrast between the foreground and background.
filmed_in_light_etc.jpg

Cons for filming in the dark
If you watch the Daylight Saving video that the first picture is from, you might have noticed that Bottle Monster seems to come in and out of focus a lot. This is because I was filming in the dark and I manually dialed the exposure setting on the camera all the way down to get an even darker picture. This gave the camera’s auto focus mechanism a hard time to focus on Bottle Monster, because his color is similar to the dark background. This doesn’t happen to Mac and Cheese as much because he is yellow and bounces off more light. This problem can probably be solved by using the manual focus setting on the camera, but really, when you are filming in the dark, you simply don’t have as much control over the image because the camera has a harder time sensing lights, and you have a harder time looking at the camera’s LCD screen for what’s being filmed as well.

Cons for filming in the light
Like I mentioned earlier, it’s good for contrast between the foreground and background. However, if you don’t want that much contrast and color, and instead just want everything dark, then filming in the light isn’t a good idea. Also, any shots that’s manipulated in the video editor might come out having an artificial feel. It might not look as natural as a shot that’s actually filmed in the dark.

So which one would be a better choice?
This probably depends on what kind of shot we want, but in general, I’m not going to go for either extremes. In the future, I’m going to film something in dimmed lights where the brightness level is higher than that dark picture and lower than the daylight. This will give me a more natural feel to the final product while also giving me some control in terms of brightness and contrast with colors. Since that’s probably enough lights for the camera’s auto focus mechanism, I’m probably not going to use the manual focus. I’m probably not going to be adjusting the exposure either, but instead I’ll adjust the light level as best as I can. Filming in the dark can certainly be tricky, but the ones done well are certainly interesting to look at. :-)

How to do a split-screen effect with a matte: a case study

We are now doing a case study on how to do a split-screen effect with a matte. This effect was first experimented in Episode 3: Crazy Names and I have been using the effect for every episode ever since. The video footages I’m using for this tutorial was from Episode 7: Big Dipper but it just might as well be any other episode. The finished product looks like this:
matte_practice_composite.jpg

Again, the editing software I use is Adobe Premiere Pro. If you use other software, the procedure should be pretty similar still. Also, I’m not going to talk about what mattes are and why they work this way over again since I already talked about them in the previous post, How to do a split-screen effect with a matte: the theory. If you haven’t read it, go read it now because it has cute stick figures and a couch. Boy, couches are hard to draw than I originally imagined. :-P

First we need to film the left side of the screen. We’ll call this Clip 1.
matte_practice_left.jpg
You can see my head in the lower righthand corner, but that’s alright. We are not using that half anyway.

Second, we need to film the right side of the screen. We’ll call this Clip 2.
matte_practice_right.jpg

And the last item we need is a matte that’s black on the left, white on the right, and a gray gradient in the middle. The red frame is not part of the matte. It’s just there in case you are reading this in an email or RSS reader that has a white background.
matte_practice_matte.jpg

And now it’s time to edit.

1) Drag Clip 1 into video 1 (bottom layer) and Clip 2 into video 2 (layer above video 1): If you’ve read the previous post, you know why it’s in this order. ;-)

2) Create a matte track: Create a new video track and rename it to “matte”. Well, you don’t have to rename it but it helps prevent confusions. Drag the matte into this track. Make the track invisible by clicking the eye to the left of the name of the video track. This is because while you want to use this matte, you don’t want the viewer to see all these black and white colors occupying the screen. They are there for a purpose, but not for viewing. So turn it invisible.

Now your timeline should look similar to this:
matte_practice_tracks.jpg

3) Apply the matte: To do this, go to Effects window. Under Video Effects => Keying, you will find Track Matte Key. Drag it into Clip 2’s Effect Control Window. And then under “Matte”, choose the video track that the matte is sitting on (in this case, “matte”). And then for the composite option, choose “matte luma” (the default is probably “matte alpha”). We are deciding the transparency of the pixels based on how bright a pixel in the matte is. That’s why we use the matte luma option.
matte_practice_options.jpg

And when you have done all these steps, the black on the left of the matte will turn the left half of Clip 2 transparent, allowing the left half of Clip 1 to show through (Mac and Cheese). The white on the right side of the matte will turn the right of Clip 1 opaque, thus showing the right half of Clip 2 (Bottle Monster) while blocking the right half of Clip 1 (background and my head). So you see the final result:
matte_practice_composite.jpg

Now you can go nuts with the awesome split-screen effect! :-D

PS. In case you are wondering, the “reverse” option reverses the matte. That is, black is now opaque and white is now transparent. If you have the two clips in this current order, clicking on “reverse” will make the final result just the green background, since we are now showing the halves without the puppets. However, if you had put Clip 2 in video 1 and Clip 1 in video 2, this option will come in handy.

Related posts:

How to do a split-screen effect with a matte: the theory
Editing Trick: How to have two puppets on screen at the same time with only one puppeteer
How to do a green screen effect: the theory

How to do a split-screen effect with a matte: the theory

This post is about how to do a split screen effect, where there’s a background image across both parts of the screen, but you film the left and right halves seperately to be put together later. Here’s some ideas on how it works. For a more detailed step-by-step tutorial on how to actually accomplish this effect in an editing software, come back in a few days!

Let’s say you want to film you sitting on one end of the couch talking to someone on the other end of the couch, who is also played by yourself… something like this:
matte_theory_composite.gif

You’ll need three items to accomplish this effect:

1. The left half: Film yourself on the left half of the screen. We’ll call this Clip 1:
matte_theory_left.gif

2. The right half: Film yourself on the right half of the screen. We’ll call this Clip 2:
matte_theory_right.gif

3. Make a matte that defines how to composite these two halves together:
matte.gif
(Note: The outside red frame isn’t part of the matte. I only put it there in case you are reading this post in an RSS reader with white background.)

Now, what’s a matte and how does it work? A matte is something that you apply to a layer of video to decide how opaque or transparent the final result is going to be. Here we are using a grayscale picture, which is a picture with different shades of gray, with full black being the darkest gray, and full white being the lightest gray. When you apply this matte to a layer of video, the part that gets the white will be entirely opaque, and the part that gets the black will be entirely transparent (meaning, you won’t see anything). The gray parts are the semi-transparent parts. How opaque or how transparent the gray part is depends on how far it is from full black and full white.

For those that want to know why it’s that way (which is not essential for using this effect but might help your understanding), think about your monitor or TV when it’s off. When it’s off, it looks black. This is because your monitor displays colors by adding three beams of light (red, green and blue) together to form one pixel. When there’s no light, you get black. When they are all on full power, you get the sum, which is white. Different mixes of light beam / color strengths would add up to different color. So in a matte, white means that you want all colors to pass through, making that part fully opaque. On the other hand, black means the absence of colors, or not letting lights through. In that case, you get an image that’s transparent. A mathematical way to look at it is, for white, you are multiplying every pixel value by 1 and for black, you are multiplying every pixel value by 0. For gray, it’d be something in between like 0.2, 0.5, 0.8, hence making the pixel semi-transparent.

Now you want to put the two images together. First you want to put Clip 1 in the bottom layer (background) and Clip 2 in the top layer (foreground), and then you apply the matte to Clip 2. What this is going to do is:

1. The black part of the matte, which is the left half, will make the left half of Clip 2 transparent. When that part is transparent, it will show what’s in the background. This is how the left half of Clip 1 shows through.
2. The white part of the matte, which is the right half, will make the right half of Clip 2 fully opaque. So what we are seeing in the final composite is the left half of Clip 1 and the right half of Clip 2.
3. The gray gradient part of the matte, which is down the middle line, will make the middle region of Clip 2 semi-transparent. This is to accomplish a natural blending between the two videos. You don’t want to see a harsh dividing edge between the two halves. That’s why you want the two videos to blend down in the middle.

And then feed these three elements into your editing software (I’ll show you how in the next post), and you shall get the final result:
matte_theory_composite.gif

So that’s how mattes work. A detailed tutorial with actual video footages in a real editing software will be posted up in a few days. Stay tuned. :-)

PS. This effect was first experimented in Episode 3: Crazy Names, but you can see it in pretty much every single episode after that.

Related posts:
Editing Trick: How to have two puppets on screen at the same time with only one puppeteer
How to do a green screen effect: the theory

How to do a green screen effect: a case study

In Episode 3: Crazy Names, I attempted the green screen effect for the first time. But initially, I wasn’t trying to use a green screen effect. I was just going to film one puppet on the left half of the screen, and then another on the right half of the screen, and then composite the two together. Here are what I started with:

green_practice_raw.jpg

When I used the cropping effect to take out the half that I didn’t want, what I got was this result on the left. Hm… that didn’t work so well. It was obvious that the two halves were shot seperately. That is when I decided to try a green screen effect (I happened to have filmed this in front of a green backdrop, which helped). What I got was this result on the right… and it was so much better.

green_practice_goal.jpg

To do this, first you need to decide what’s the foreground and what’s the background. In this case, we are going to use Mac and Cheese as the foreground and Bottle Monster as the background. This is because Mac and Cheese is yellow/orange and less similar to the green background than Bottle Monster is. In Adobe Premiere Pro, this means you want to put the Bottle Monster footage in the bottom video layer and the Mac and Cheese footage on top of it.

Now select the Mac and Cheese video. In the Effects window, find Video Effects => Keying => Chroma Key and drag it into the video’s Effect Controls window. This effect allows you to key out a color you select, in this case, green.

green_practice_chroma.jpg

Now it’s time to set the variables associated with it. First, use the color picker / dropper to pick the shade of green to key out (make disappear). You might want to pick something that’s in the middle of the darkest and lightest green you want to key out. And then set the other variables. The most important variable here is Similarity. Our footage is not ideal, especially since it’s not intended to be a green screen shot initially. The Similarity setting allows you to set how much error you want to give in comparison to the green you picked. If this value is very small, only the pixels that’s very similar to the shade of green you picked will be keyed out. If this value is big, you will get rid of more different shades of green. Note that, when this number is too big, it will start to include colors like black and make Mac and Cheese’s mouth transparent. We don’t want that. The Blend is a variable that decides how much the two images blend together.

green_practice_finished.jpg

A lot of this is trial and error. Here I included the results of a few different sets of numbers. Some of these look better than the others. Some of them have background that’s not completely keyed out. Some of them have parts that became transparent although it shouldn’t have. So play around with the numbers until you find the best result! :-)

green_practice_variations_80.jpg

Related posts:
How to do a green screen effect: the theory
Editing Trick: How to have two puppets on screen at the same time with only one puppeteer
Editing Trick: Cropping
Episode 3: Crazy Names

How to do a green screen effect: the theory

In my post Editing Trick: How to have two puppets on screen at the same time with only one puppeteer, I said that I’ll talk about the green screen effect. The green screen effect is something that’s very widely used in film and TV productions. Huge portions of movies like Star Wars and 300 are filmed in front of a green screen so they can use a lot of computer generated images in the background. It’s a very useful effect. So, let’s talk about how green screens work.

When you use a green screen, you first film the foreground object (usually some kind of character) in front of a screen that’s green (first picture). And then, you insert your background (second picture) in the back (where backgrounds belong). And you’ll have a composite like the third picture:
green_theory_stick.gifgreen_theory_back.gifgreen_theory_composite.gif

What the software will do for you, is look through every pixel in the first picture for green. When it sees a green pixel, the pixel becomes transparent so the background layer can show through. So whatever was green in the foreground picture is replaced by what’s in the background in the final result.

When you use this method, you should be very careful because if the character is wearing anything that has green on it (like the question mark on the shirt in the following picture), it will disappear too! And then you get a shirt that’s partially transparent and you can see through the guy’s body and see the background. So be very careful.

green_theory_stick2.gifgreen_theory_composite2.gif

What if you want to have a foreground character that wears green? In that case, you might want to use a blue screen effect instead. That is, film the character in front of a screen that’s blue, and in editing, let the computer pick out all the blue pixels instead of the green ones. In many movies, both are used depending on what’s in the foreground.

Why don’t we use red screens? It’s because that our human skin has red in it and you might end up with strange effects that you don’t want. Of course, this might or might not be an issue for puppets. I haven’t tried it out yet.

In the next post I’ll talk about how to do it with actual footages.

Related posts:
Editing Trick: How to have two puppets on screen at the same time with only one puppeteer

Editing Trick: How to have two puppets on screen at the same time with only one puppeteer

In Episode 3 - Crazy Names I needed to put two characters on the screen at the same time but all I have was one puppeteer (me). Now how do we accomplish that? Before I go into that, let me refresh your memory with these three screen shots, each representing one way of putting two characters on screen.

Three ways of putting two puppets on screen with only one puppeteer

There are three methods, each with its own advantages and disadvantages.

1) One puppet in each hand:
How it works: By… putting a puppet in each hand? Simple enough. In fact, that’s how Episode 1 - Making a Podcast is done.
Pros: No tricky editing required, yay!
Cons: Talking heads only. There are absolutely no arm movements. Also, the character operated by my left hand (in this case, Bottle Monster) has worse lip-sync because I am right-handed.
How to recognize it in this video: Neither character’s arms are moving.

2) Using a green screen
How it works: Film one (or both) characters in front of a solid green background. In editing, all pixels that’s green will be removed and replaced with something else.
Pros: You can do this as many time as you want (meaning, you can have 100 characters on the screen at the same time if you want) provided that you can get the lighting right and bother to do many passes on shooting.
Cons: For something that’s home-made, it’s hard to have a green screen that’s perfectly lit (meaning, all the green background pixels are lit to be the same shade of green).
How to recognize it in this video: Mac and Cheese cast no shadow on the background but Bottle Monster does. Mac and Cheese didn’t blend with the background entirely due to bad lighting.

3) Do a split screen effect using a matte
How it works: On the first pass, we shoot what’s supposed to appear on the left half of the screen. On the second pass, we film the right half. In editing, we put these two halves together using a matte (mask) where we show the left side of one video, the right side of another video, and then blend whatever that’s in the middle.
Pros: This allows you to film the two characters on different passes. Perfect for two characters that occupy opposite sides of the screen.
Cons: Neither of the characters can cross the middle line, or the illusion will be lost. (If they tried to cross the middle line, whatever crossed the line would vanish.) Sometimes the background colors of the two footages don’t blend very well and require extra corrections.
How to recognize it in this video: Characters always stayed away from the middle line. In some shots the left background didn’t blend perfectly with the right background (which I paid more attention to in subsequent episodes but not this one). Bottle Monster’s dad’s fishing pole started to vanish a lil bit when it’s closed to the middle blending section.

While the first method is easiest to do, the second and third methods give the characters the freedom to have arm movements. Both the green screen and the split screen method take more planning and require two passes to film each shot. But hey, when you only have one puppeteer (or actor), you can still use these editing tricks to make two appear on screen.

Are you confused yet? :-P

If you are confused by the first method, hm… I don’t know what to say.
If you are confused by the second and third method, don’t worry. This is just an overview. More detailed step-by-step explanations will be posted in the following weeks. :-)

Next Page »

Based on FluidityTheme Redesigned by Kaushal Sheth Sponsored by Web Hosting Bluebook