I’m having a bit of trouble understanding how all the video editors and codecs work. Essentially what I want is to take my animation and render it into .png’s, then sequence the .png’s to make both a large uncompressed .avi and a much smaller compressed video (H.264 preferably since I know it works well). If you don’t want to give me a lesson on codecs, perhaps you could help me decide the quickest and easiest way to get these three sets of images & videos? Currently I’m able to use Blender to render the .png’s and Virtualdub to combine those images into an uncompressed .avi, but I don’t know how to make a smaller video based on H.264 codec. (I used avidemux’s mpeg 4 and noticed quite a bit of icky pixelation)
If you do want to help me understand codecs, read on.
As for the video compression, here’s how I thought it worked: You can download video editors, or other programs like blender which create/compress videos, and these programs compress the video according to a set of instructions known as the ‘codecs’ already installed on your computer. If you want different quality/size options, you can download different codecs. When I downloaded editors like virtual dub and avidemux, they only had access to the four default codecs (the Microsoft and Intel codecs), but Blender can compress to other formats including H.264, which is an excellent codec since it makes the video extremely small but still looks just like the uncompressed video. What I don’t understand is how Blender has access to this codec while the editors do not.
Also I’ve been able to render to uncompressed avi and h.264, but can I use an image sequencer built into Blender instead, allowing me to bypass any use of another editor?
One more thing. Why is the codec known as H.264/MPEG 4 if it renders to a .avi, not .mp4?
If I understand this correctly I might be able to help you here at least…I’m assuming you meant you’ve rendered frames from a 3D view straight to video, and what you want to do is render the PNGs instead…
Not sure which version of Blender you are using but I think the process is similar for all. This is for Blender 2.5x
In a new project, if you go the the video editing layout and hovering over the video sequence editor use:
shift+a -> image
Now navigate to the folder with all your pngs and box select them. Then select ‘Add Image Strip’ from the top right.
At this point it should drop your sequence into the VSE. Line it up so it starts at the first frame (with it selected use ‘G’), and set the last frame of the animation to match the number of images you loaded.
Now in the render panel make sure your output sizes reflect the size of your images (the aspect ratio should be correct at least). Finally, set your output file and codec you want, then make sure you tick ‘sequencer’ under ‘post processing’. This renders from the VSE instead.
Then just hit ctrl+f12 or the animation button in the render panel. Blender should fly through your image sequence now and output a video file as you specified.
Sansari, you’re doing everything right, rendering to image sequences then using any of the tools you list, VDub is a favourite for many here, so is Avidemux.
Your problem is lack of system codecs, two main types, Direct Show (newer) and Video for Windows (older) I think VDub uses VfW only, maybe wrong. FFmpeg is another option, make sure you have that installed system wide. Win32 binaries are available via Sourceforge. Avidemux bundles x264 with it.
So all you need to do is download and install some more system codecs. Suggest FFDShow Tryouts, download through Sourceforge and make sure it’s the tryouts version.
This will allow far more encoding choices. x264 is the best OSS codec for h264, you may need to play around with the quality settings / Profile in the encoder GUI in Avidemux to get better quality. h264 is pretty good.
if it were me & I couldn’t use my NLE, I’d render to TGA’s, then use virtual dub to make the uncompressed AVI, then use Super! to convert them to the codec of my choice.
Super! includes all the codecs it uses & has never messed with any of my system codec’s.
One of the filters is Resize. Use this to make your smaller version.
VirtualDUB can mke what you need.
VirtualDUB makes decisions about what codecs are “good” and what codecs are junk/malware for viewing illegally downloaded videos. Simply use the good codecs it detects. This is to your advantage because if you use those “junk” codecs your target audience may not be able to view them.
Basically install XVid and or DivX and just use those for AVI. Otherwise get Quicktime Pro ($29.00) and make MOVs.
Ok so I used the sequencer inside Blender and it’s perfect. It renders the pngs, then the sequencer saves both an uncompressed avi and an h.264 compressed avi. So if Blender can do all this, why do people use other programs? Is it just to access other codecs?
there’s stuff I can do in my NLE much much faster then I could do in blender’s NLE. Other NLE’s also support different plugins. IE Blender can’t do what After Effects can do. It can’t do what Vegas can do. It can’t do what Virtual Dub can do, etc.
I’d like Blender to spend more time being a 3d app vs a NLE, just because there’s so many good ones out there (unless a spin-off of Blender was built around NLE use & didn’t focus on 3D). Just like I’d rather see blender do less sculpting stuff & more modeling stuff, just because there’s some really really really good sculpting apps out there that I don’t see Blender being able to compete against.
A Chainsaw is more useful at cutting down trees vs a Swiss Army knife, even though you COULD do it with the knife.
Did you have audio?