Export / render to mp4 keeps stopping

[coinman]

Working on my forth project, which is close to ten minutes long. The first three rendered to mp4 OK (with some issues), but this one will not. It gets through maybe 15% of the project, and then the render window just disappears. The mp4 file is there, but not complete and does not work at all.

No screen saver on, using Mcafee anitivirus, if that could be the culprit. Windows 10.

Trying to render to image sequence right now...and same problem. I can re-try and get further along, but it's going to take me 40 failed renders to get to the end.

If anyone has an idea, please let me know.

[Greenlaw]

I wouldn't bother rendering the animation directly to MP4, especially for such a long animation. It's more efficient to render to a PNG image sequence first and then compile the frames after rendering. You can compile the frames in a new Moho project with only the frames imported or through a compositing or video editing program.

The reason for this is that it takes a lot of extra computer resources to render the animation while also holding all the frames in memory to create the movie file, increasing the chance that something will error. It's easier, and maybe faster even, for the computer to render one frame to a file at a time. Once you have the entire image sequence rendered, compiling the frames into a movie file is quick and easy because your computer no longer has to deal with processing bones deformation, lighting effects, warping images, anti-aliasing, etc.

Other benefits when rendering to an image sequence: if the render fails, you can just pick up where it left off and continue. By contrast, if it fails while rendering to a movie file, your data is pretty much hosed. You will need to start over and hope it doesn't crash again. Finally, if you decided to change your animation, rendering to an image sequence allows you to replace only the frame range that changed...so need to render the whole thing again.

As mentioned, you can compile the frames in another Moho Project file, but I recommend using a compositing or video editing program to create the video file. Moho can do it but you may find more codec and quality options available in another program.

Rendering to an Image Sequence first is common practice in animation production and not specific to Moho. Many professional animators do this when working with Moho and other animation programs, 2D and 3D.

BTW, 10 minutes is unusually long for a single animation file. Normally, scenes are broken out into individual projects and then edited together in a video editor. If you really are working in a 10 minute long project file, I suggest breaking out sections to make it easier to manage. And if you're already doing that, never mind.

Hope this helps.

[coinman]

Thank you again!

I am sold on what you suggest. Fortunately, I am an expert at Vegas, so working on it now. Even though you mentioned Vegas earlier, it did not cross my mind until minutes ago that I could use it for compiling, which is an exciting plot twist. It seemed like you were suggesting that some studios actually use Vegas, which is also fascinating to me.

A little bummed that I have to re-edit my audio for this project, but it will clearly be much easier to do future audio edits in Vegas.

[hayasidist]

rather than editing the audio into separate files using something like audacity you can import (all 10 mins) into moho and have as the visual element of that file just a set of storyboard sketches or a "clapperboard" stills to show the scene changes (i.e. when the actual moho animation files start and end) -- then export that in chunks (give yourself a few frames for pre- and post-roll) and import the resulting a/v as references for the animation files. The "original" audio stays intact and becomes a layer in the compositor.

[coinman]

Thank you!

It occurred to me the second I woke up today to just render this project with the images turned off to get my completed audio track on a mp4. When you say export, do you mean mp4, or something even better?

[hayasidist]

Ah! you've used Moho to assemble the soundtrack elements?!
well ... are you on 13.5? if so:
File / Export / Export soundtrack

[coinman]

For this project already, yes. Sadly, I do not have that option in my version of Moho. }:^)

[hayasidist]

For this project already, yes. Sadly, I do not have that option for this one. }:^)

sadly?? you mean about using moho for soundtrack assembly? well -- no biggie -- are you on 13.5? if so you have the export soundtrack option. (might be windows only as it exports a .wav)??

[Greenlaw]

If you have .AIF as an option, that should work too. Mac should at least have that format.

Note: normally, I work the other way around.

First I'll make an animatic in my video editor (Vegas in my case,) and do my audio assembly there. Then I export each scene and audio track as separate files. Next I'll animate in Moho using the imported footage as reference. When I'm done animating the scene, I export the footage as an image sequence, composite in Ae or Fusion Studio, and export a movie file from there for editing in the video editor. I don't need the audio file from Moho because it already exists in the video editor.

Back in the video editor, I place the composited footage in a track above the animatic track, and the existing audio should remain in sync. After the animation stage is completed for the entire production, I'll go in and improve the audio and mix in the video editor for the final movie render.

This describes my personal workflow but it's more or less the same at my workplace but with some different software.

One more tip:

When exporting my scene clips from Vegas for Moho, I export each scene as a JPEG sequence and the audio as an uncompressed .wav or .aif. The reason for this is a JPG image sequence streams well over even a slow network. That, along with the uncompressed audio insures that I'm seeing my reference with frame-by-frame precision...this can be critical for lip sync animations.

To do this in Vegas, if you mark the head/tails for all your scenes on the timeline, there's a way to get it to export each marked scene automatically rendered as a separate JPG sequence and audio file. I like to use a Vegasaur script designed for this but you might be able do this natively in Vegas with some extra effort. If you use a different video editing program, there is probably an equivalent way to do this.

Hope this helps.

[coinman]

Thank you. A lot to think about - not sure if I want to go 100% this route yet, but might filter that way in time.

With this current project, if you export an mp4, would the mp4 audio still be the same quality as the original wav files? Just in case I ever do it this way again.


Todd

[Greenlaw]

if you export an mp4, would the mp4 audio still be the same quality as the original wav files? Just in case I ever do it this way again.

MP4 is just a container and the quality really depends on the audio and video codecs you apply but, generally speaking, movie formats like MP4 are reserved for the final stages because they will be compressed for streaming performance. If you use MP4 with a lossy codec throughout the entire process, the image and audio is likely to degrade (i.e., like when you resave a jpeg image over and over again.)

Ideally, you would use a lossless codec (for image sequences/movie files) throughout the process, which is still compressed but with no loss in quality. Some lossless codecs are optimized for editorial (better performance) use and others for archival (slower but more compact).

So, in my personal workflow, I usually render to image sequence first because images are easier and more efficient to work with in compositing, then export lossless movie files from compositing to editorial because movie files work better in a video editor. Finally, from the video editor, I'll export a final lossy file for distribution and a lossless version for archival. The workflow at most professional animation studios is similar.

For 2D animation, I like to use PNG image sequence because it's a lossless format, meaning it's compressed but there is no visual loss in quality, and because almost any program can open and edit it. Generally speaking, you want no image degradation at all when compositing. In higher end production, I use EXR, which is also lossless and has many more features useful for compositing. For editorial, I had been to using .avi with the lossless MagicYUV codec for years but lately I'm also using Apple ProRes in Windows programs that support it. Technically speaking, most ProRes choices are not lossless but it's very high quality and there is little difference visually.

For final output to streaming services like Vimeo and YouTube, I like to use the H.265 codec for its support for 16-bit color, high quality, and compactness. For me, black levels also look more correct with H.265 in streaming without applying additional levels correction like I do with H.264.

Hope this helps.

[coinman]

Great to know. Thank you.

[Greenlaw]

Sorry, I focused a lot more on video even though you asked about audio, but the info is generally the same for audio.

It's normal to do your audio editing when you're in the animatic phase before you start animating anything. It doesn't need to be final quality but the timing should be pretty close to what you intend for final.

As mentioned earlier, what you export from there for animation reference should be uncompressed audio to insure frame accuracy when you're animation. This is critical for lip sync animation. If you export a video file, you don't need to export it with the audio because the original audio already exists in your editing program. You can export your fully editing movie with audio from there.

Technically, you can do all this in Moho but if you do that you're probably making the task harder than it needs to be because Moho is not a video or audio editing program.