I have a multi-part question regarding playback and rendering:

First, I'm wondering if there is any limit to the size/length of scribe being produced?  I am working on a fairly elaborate piece that will ultimately run approx. 10 minutes.  As I lay in my original SVG files (from AI), performance is great... but as more images are added, the playback is no longer fluid (as I would expect as the file size increases).

So...

I assume that rendering to Quicktime will solve this issue... Correct?

Alternatively, is it possible to output an image sequence for assembly in some other application like FCP or AE?

Because of the real-time playback issues described above - it is difficult to assess timing as I move forward.  Can I rely on accurate playback timing even thought the visuals are "chunky?"  In other words, is the program dropping frames to assure that 1 second=1 second during realtime playback?

Would be good to know before I get too deep or take the wrong turn...

Thanks for your help!!

David, there is no limit to file size or length in the application. However there is a limit to the RAM usage the application can handle as it uses Adobe Air so when it gets to about 1.2GB-1.5Gb of RAM it can start to misbehave. However if you are not using large image files (SVGs should be fairly small in comparison to JPGs and PNGs) you should be fine. Other things that soak up RAM are:

• Extreme zoom so try to stay as close to 100% as possible and resize your SVG instead 
• Lots of imported fonts as the character sets for all are held
  
• Large JPG, PNG or GIF files

Thanks Barry!

The tips and link is really helpful.

Onward.... 

Further on the note above...

As file size grows and RAM is depleted, playback does indeed start to misbehave - chunky playback (skipped frames, I presume), strange masking behavior, etc.  Good to know the reason behind it as more SVG's are added to the file...

With that in mind, I'm still wondering if I can rely on the TIMING of the real-time  playback (even if it looks odd with skipped frames, etc)?  I am trying to adjust scribe timing to sync with narration, and still wondering if I can rely on timing of the working copy (skipped frames are fine if it keeps it all in sync - without SLOWING DOWN real-time playback).  If that is not the case, is there a workflow that you recommend for long scribe files like this (other than SVG guidelines mentioned above and in link) to confirm everything is in sync before committing to the final render?

With the preview if there are glitches it should keep the sync between video and time. 

I would run test renders to Sho.co through the 'share video online' button as this renders on our servers and allows you to carry on working. Can set these to private (you'll need to log-into Sho.co to view) so nobody else can see them.

Also there is a preview play button on each element of on the timeline and this will play from that point and that will have the voice over as well so can use that to test certain points.