We get tons of questions on how to output After Effects project renders, mainly which format and codec to use. Topher decided to do some research and wrote this long article that aims to try and answer some questions. It is worth noting that the After Effects built-in help has improved dramatically in recent versions and it is constantly updated online. We urge anyone reading this article to also read the AE help for additional reference.
I will be the first to tell you that this is going to be a boring article. Possibly mind numbing. But I have been wanting to write this article for the longest time because I can’t count how many times I have been asked what Output Module to use, what it does, and a thousand other questions that go along with it. There is a great deal of information at help.adobe.com’s After Effects section. So go grab an energy drink or a cup of coffee, and gear up… this is a big one.
Adobe Clip Notes
Adobe Clip Notes is a pretty neat little trick when you are working with clients who might not have certain codecs enabled on their system. What this format does is output the video in either QuickTime or Windows Media, the two most standard formats on all computers… and wraps the video in a PDF file that you can choose to password protect or not. There is also an area for you to make notes about your video and save it, and your client can make notes back, save it and send through email. This is a great way to work with all the proofs you might be sending back and forth repeatedly. Check out Aharon Rabinowitz’s tutorial on this over at Creative COW.
This is a pretty obvious output format, but basically it lets you set up a piece of video as an animating GIF image that you can specify for it to have transparency, background color, dithering, or looping enabled or not.
Cineon is the format used to output the frames of your composition to be later put onto film. You can choose the correct color profile for the film stock that will be later printed to. You can assign color profiles in the Project Settings dialog box. For a very in depth look at Cineon formats, check out Pete O’Connell’s article at Creative Cow.
This is the native file format for Electricimage and unless you are working with this 3D program you probably won’t use it.
Essentially this is the same thing as an FLV. It is a MainConcept H.264 video codec wrapped in a Flash video container. The good thing about F4V is that it gives you many more high resolution output settings (up to 1080p) than the regular FLV output. It also looks better and produces smaller files than .flv. There is also a trick with .f4v files, if you literally change the file extension to .flv, it will make it more compatible with online flash video players.
This is the main Flash Video format (F4V is just the latest version of this format). Like F4V is lets you customize bitrate, frame rate, field order, and the audio settings. The main difference with this output module, is the video codec that is uses to compress: Sorenson Spark and On2 VP6. This is a very lightweight format, and with over 98% of all internet users have flash installed on their computers, it makes this file format the most popular video format on the web. If you want to reach the most people, this is probably what you want to use as your distribution format.
This is a format that is made to be brought into Photoshop as a filmstrip with frame number, reel, and timecode, displaying every frame right next to the other, so that you can edit it by comparing frames side by side. This is generally used as a rotoscoping technique if you choose not to do your roto in After Effects itself.
This is the most commonly used Quicktime compression for video, because it produces really clear, very small file sizes. The format settings for this output format are very similar to the F4V settings (since it is essentially the same codec under the hood), and it gives you a bunch of settings preformatted so you can produce the videos immediately for iPod, iPhones, or PSP. This format is kind of a pain to work with when you are trying to render out HD video. You can see my tutorial on blog.videohive.net for tips on rendering HD H.264 video.
There are two different Blu-ray codecs to render out to, but this new H.264 Blu-ray output format is a little newer and compresses a lot more than the MPEG2 version. The only catch to use this is that it takes a lot more processing power and more time. The recent new Nvidia QuadroCX card greatly increases render times by utilizing the CUDA engine built into the card. This output format creates a file with a .m4v file extension. For a comprehensive article on the different Blu-ray formats, check this out.
The .IFF Sequence is the native format of Autodesk’s Maya. This file format embeds only RGBA information, and not Z-depth.
A lot of the image sequence output formats are the same deal, in that they can only encode RGBA information. This choice, however, can’t encode alpha information into it. The file size is a lot smaller than some of the other choices, and should only be used for final output since it is highly compressed.
Kind of obvious. This renders an .mp3 audio file for the length of your work area of your composition. Options available in Format Settings are quality, sample rate, channels, and bit rate. If you want to export a .wav or .aiff file format, you can do so by going to File>Export>WAVE or AIFF, and then choose the settings that you want.
The different MPEG2 output formats actually gave me quite the runaround on figuring out how they work, and why they are different from each other. There are two formats to output for DVD, this one, the MPEG2 format renders to a .mpg file extension and gives you the choice in the format options, under Multiplexing to use a MPEG2, SVCD, DVD, or TS format. This seems like the output format you would want to use if you were to be making a Super Video CD.
Like I wrote earlier, the H.264 Blu-ray codec compresses better, but other than that, the settings are all virtually exactly the same layout. I guess it depends on what file format your Blu-ray authoring program is calling for, because this Output Module outputs the files in a .m2v file extension rather than the .m4v that they H.264 Blu-ray module uses.
This is the standard format that you are going to want to render to for taking your standard definition video files into a DVD authoring program such as Adobe Encore or Apple DVD Studio Pro.
Finally! You can render files out to be used in Avid directly out After Effects… except there is only one catch, it’s only available in Windows. There are a couple little features in it once you choose the output module and hit Format Options. First off, you can choose between NTSC and PAL (obvious), choose your resolution between 1:1 or DV, and then choose a version of the file between OMFI 1.0 or 2.0. The only difference between the two, is 1.0 is supported for backward compatibility with older Avid Systems. One more little catch, this output module doesn’t render audio with it, so you are going to have to output some audio tracks to sync up later.
OpenEXR was developed by Industrial Light and Magic for use in their feature films. This is a high dynamic-range image file format that allows you to recover highlights, luminance and other features of photos that you generally wouldn’t be able to without using floating point. Unlike the Radiance format, with .exr you can choose to use an array of different lossless and lossy compressions including Piz, RLE, Zip, Zip16, PXR24, B44, or B44a. There are also options to encode Luminance and Chroma information. This is becoming the de-facto HDR workflow format.
If you are a Windows user like me and had no idea what a PICT Sequence was all these years… you aren’t alone. I actually had no idea, and had to Google it myself, and found out that it is a standard graphics format on Macs. This output format actually just creates an image sequence like any of the other image sequence modules, and can contain RGBA information.
One of the more popular image sequence output formats, PNG sequences are usually preferred because they encode color and transparency really well. This format can render out RGBA information as well as progressive or interlaced.
Same as all the other image sequences, this renders out to .psd file extensions and has the ability to render RGBA color information.
Alright, the big one. QuickTime is probably one of the output containers you will use most often in your rendering for almost any client. There are a lot of options for quicktime so I will describe every codec. Each one of these sub-codec compression types encode with the codec type, but keep the file extension as a .mov. Keep this in mind when you are trying to do something like get a .mp4 file from the H.264 sub-codec compression type. Depending on your platform some of these codecs might not be available. You can always install additional codecs on your system and they will show up when you hit the Format Options button.
The 3ivx MPEG-4 5.0.3 is a version that claims to be better than Apple’s native MPEG-4 compression. According to their website, 3ivx can produce the same quality video in half the size, is five times faster encoding, and is fully compatible with the Apple MPEG-4 decoder built into QuickTime. This of course is in conjunction with QuickTime alone, and I haven’t been able to actually test these facts within After Effects itself. This codec also has its own options available when you are in the Compression Settings dialog box; when you hit options, there will be Basic as well as Advanced options, specific to this codec.
Quicktime Animation is next, and is quite possibly one of the most used codecs. It renders out a lossless, visually uncompressed QuickTime movie. This means it usually produces a huge file. Supports RGBA color information encoding. Until recently, this was probably the most popular mastering codec.
CineForm HD/2K/4K is the codec you are going to want to use if you are working with RED files or any other gigantic file formats. This codec also has options specific to itself, in that you can check to enable RGB 4:4:4 encoding and if your source is an interlaced video source.
DV-PAL is kind of self explanatory. You can set your frame rate, compressor quality, scan mode and aspect ratio. Additionally, the DV/DVCPRO – NTSC has all of the same options as DV-PAL, just it encodes to NTSC instead. Finally there is the separate DVCPRO-PAL format, which, yeah you guessed it, has all the same options again, but renders to DVCPRO-PAL format.
Now, the infamous H.264 compression format. This is probably the best formats for compression that you can use. When you watch those amazing looking hollywood movie trailers on Quicktime.com, they are all compressed with h.264. It very customizable, you can set the KBytes per second to that you want it to be with the checkbox for "Limit data rate to". You can do constant or variable bit encoding which can make better looking, smaller files. This is quickly becoming the most popular distribution codec since it began being supported by Flash in it’s latest version: F4V.
JPEG 2000 essentially renders the video as an image sequence, but then kind of "wraps" that sequence in a .mov container and gives it the ability to encode audio along with it. Unlike the JPEG Sequence output format, this codec can also encode RGBA information.
The MPEG-4 Video compression type, doesn’t use the same compression as the H.264 codec, but it does give you the same amount of options including the "Limit data rate to" specification box.
There is a major difference between the Quicktime Animation compression setting and the NONE setting. Animation does Run Length Encoding (RLE) to maintain a visually lossless image but a smaller file size. This means you get a visually uncompressed file which is going to be smaller in file size than NONE. But like animation it also lets you set keyframes in your video every however many specified frames you want. Unless your client specifically asked for the NONE codec, I would recommend you use Animation when you want an uncompressed movie.
Photo JPEG is the same as JPEG2000 except for the fact that it doesn’t support encoding of alpha information. You can still render an audio track along with it though.
PNG is another image sequence wrapped in a .mov type of format. This compression lets you choose if your video is to be interlaced, and then also provides a few filters for you to choose from, although you should probably keep it on "Best". This codec also support RGBA information and encoding with audio.
The last codec in the Quicktime output Container is the TechSmith EnSharpen compression. Techsmith makes the Camtasia screen recording software, so this is meant to replicate their type of compressions they use in their software. It gives you the option of using RGBA information, but also gives you a numbered chart on how much compression to use having 1 be the fastest (most) compression and 9 being the slowest (least) compression.
The Radiance Sequence output format is a 32bit image sequence format for use with high dynamic range workflows. Its extention is .hdr. These image sequences can later be brought into .hdr image editors to manipulate as a whole or as individual frames. ILM’s .exr format is probably a better format as it also a 32bit format but produces smaller file sizes.
SGI stands for Silicon Graphics Image, and was originally the native image format for Silicon Graphics workstations. It has since become a popular format for high-end 3D workflows. This format has the ability to encode RGBA information and within the Format Options dialog box there is a checkbox to switch on RLE compression.
TIFF is another image sequence output format, but a very is very versatile and high quality. Along with being able to encode RGBA like pretty much everything else, it can also render with floating point. In the option dialog box, there are options to check on or off LZW Compression, which is a universal lossless data compression, as well as IBM PC Byte Order.
Right off the bat, when you choose the Targa Sequence output format, the format options make you choose to render to 16, 24, or 32 bits per pixel. From there you can choose to render with RGBA and also if you want to use RLE Compression.
Video for Windows
This is the output format that you are going to want if you are looking to output a .avi file format. There are almost 20 different subcodecs to choose from, and each one has a different type of compression, and different features available to it. Keep in mind that I use the K-Lite Media Codec Pack, so if I am talking about a few options you don’t have that could be the reason.
Microsoft DV is the codec that FireWire-captured DV footage is encoded in coming from miniDV tape. If you are editing together effects shots in After Effects shot on DV, and putting them together in Premiere Pro later on… this is the codec you will want.
The 8-bit and 10-bit YUV (4:2:2 YUV)‘s are the codecs recommended for capturing, processing, and displaying video in the Microsoft Windows operating system.
The TechSmith Screen Capture Codec is the codec that Camtasia encodes to when you do screen recordings in the program. This codec gives you options of keyframing however many frames as well as a slider for Faster Compression or Better Compression.
The Cinepak codec by Radius gives you the options of keyframing every certain amount of frames, as well as limiting your data rate to a specified Kb/sec. There are also options to render color or black and white.
The Heliz YV12 YUV Codec is one of the basic codecs used by Video for Windows for encoding. It is almost identical to the I420 codec. The only restriction with the Helix YUV codecs is width must be a multiple of 4 or the render will fail or the file won’t work correctly.
The Xvid MPEG-4 Codec has a ton of options that come along with it. For starters there are a bunch of Profile @ Levels to pick from, different encoding types, and Quality presets. Remember even though this is an MPEG-4 Codec, this will still give you a .avi file.
The ffdshow Video Codec has a massive menu for you to customize how your video will be rendered out. It’s almost like a second render queue. You can choose from over 20 different encoders, bitrates, how many different passes and types of passes, Masking, Quantization, Motion Estimation and a bunch more. I am not sure, but I am not sure if this comes standard with After Effects.
There is also a PICVideo M-JPEG 3 VfW Codec available which is a lot like the JPEG2000 options that Quicktime gives you. It basically creates a JPEG image sequence and wraps it in an .avi giving you the ability to render audio with it. There are also options for effecting the Luminance, Chrominance and Compression Qualities.
The DivX 6.8.5 Codec basically creates your video so that it is playable back for different profiles of DivX video players. You can choose from mobile to full HD profiles, and even go deeper into the settings to further customize your output.
Microsoft Video 1 is very basic, it has a slider to set the compression quality, and options to keyframe every certain amount of frames, and also set the data rate to conform to a specific KB/sec.
3ivx MPEG-4 3.0.5 is essentially the same options as above inside the Quicktime Movie output module.
Microsoft also has an RLE compression codec to encode video to .avi. This one is also very basic with the same features as Microsoft Video 1
Lastly, there is the No Compression option, which outputs a completely lossless .avi video file from your composition. This is optimal if you are working on shots and need to have the utmost detail possible.
Like I noted above in the MP3 output format, this one only renders an audio track. All that is able to be customized is the kHz, bitrate, and channels.
Yes, some people still use Windows Media. If you go into the options for the Windows Media Output Module, you will find a ton of presets for NTSC and PAL source download options catered to different speed connections as well as presets for Zune, Creative Zen and Palm devices. Everything else is completely customizeable as well as legacy .WMV video and audio codecs, bit rates, frame rates, and other settings.