It's difficult to say because we don't know a few crucial things:
First, we don't know about your workflow: do you edit, then do AE work and incorporate the work into the final edit? What editing application do you use?
Second, we don't know the nature of your AE work. Do you intend to use features like Rotobrush, Warp Stabilizer or one of the many effects that force AE to look backwards & forwards in time to create any given frame? They can slow down AE drastically.
Third, do you intend to push AE to its current limit in the 3D realm by using features such as ray tracing? They also can hinder AE speed, particularly without the proper third-party card; the number of such cards is very low for a Mac.
In general, H.264 is considered to be an ACQUISITION codec; it's good enough, but just barely; for example, you can pull adequate chroma keys -- adequate, not outstanding. For serious, big-money projects, you may be better served by getting out of H.264 as soon a possible -- as the footage is ingested, for example -- rather than working in that codec throughout. So-called production codecs won't compromise image quality.
Thanks for your response, I edit with Avid MC6, then once my frame ranges are confirmed I import the appropriate source QT that was shot on the 5D mkii into AE & trim my work area to the required frame range. I carry out any number & variety of VFX tasks but the actual processing speed of the application is not a major worry for me. It's when I then come to export the finished product that I am unsure about.
When you say "you may be better served by getting out of H.264 as soon a possible" I had been given some advice about creating image sequences (.exr or .dpx) from the H.264 QTs & then to use them to work with in AE to avoid further compression but I didn't really understand that at the time. So in layman's terms it sounds like AE can struggle when working with footage compressed with codecs syuch as H.264 & if the footage is converted & then imported as an image sequence instead these compression issues are irradicated as in theory there is no longer any codec in the workflow at this point, correct?
When it comes to exporting the finished shot then, to avoid an unwanted compression am I restricted to using codecs such as animation or uncompressed YUV or is there someway that I can somehow I can use a codec to try & match the true data rate of the information contained in the shot? Or once the original H.264 QT has been converted to the image sequence have I in effect reset the compression status of the shot with any previous compression artefacts now basically burnt in to the image?
These may well be basic compression questions but it's only recently that I have been required to produce shots that are to be used to online with, I've been more in the offline relm so these issues aren't as critical.
I hope I'm making sense with this.
Tell you what: if you can transcode that H.264 footage to DNxHD as it goes into the machine, you'd be good to go throughout your entire workflow. DNxHD makes high-quality files in a remarkably compact form. It's extremely good at preserving image quality.
No matter what you do, any problems with the H.264 footage you acquired will remain in the images throughout the workflow: it was recorded that way. AE and other applications can do some minor fixup if you have image quality issues, but huge problems require a reshoot.