Apple Event: May 7th at 7 am PT

Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

Pause/Resume

I'm exported a project from Motion 5 and Compressor is estimating 8hrs to render. I'm using a MacBook Pro and need to pause the render from time to time. If I pause the render, will I lose what's been rendered?

Posted on Aug 30, 2011 12:45 AM

Reply
7 replies

Aug 30, 2011 4:59 AM in response to Warwick Teale

****... We'll after hours debating I just tried the pause option and sure enough I lost 5hrs worth of rendering gggrrr... Kind of pointless having a pause button if it doesn't pause..!


Anyway just for the record, in sent my project from Motion 5 to compressor and chose the Apple ProRes 422 codec to encode too. I guess ProRes is one of those codecs that can't be paused :(

Aug 30, 2011 5:13 AM in response to blimpmedia

hmm, like most on these forums I use PRORES daily for working in Fcp7 /FCPX and Motion.app output etc. I use qmaster/compressor for both:

  • rendering .motn projects (care on the cores used) in compressor via qmaster and
  • transcoding material into distribution codecs (H.264 etc etc)


All in the qmaster/compressor paradym.. it was a buggy thing in V2 etc but it is very good now.


The trick is to MAKE SURE "job segmenting" is turned on fo rthe types of codecs you are doing with.


NOw this all works well if you have a great deal of CORES! Not much good in wifes 2010 macbook air.... works worse as the resources seem to "fight" with each other. Works better on 2011 macbook air i5 with 4 cortes.. u get the picture.


By definition this performs at hte very least a "segmented transcode".. assemebling the PARTS of the transcode job at the end. Works great!


If you look closer at the COMPRESSOR.app inspector for VIDEO you can se ethis here and also see the "multipass" tick as well... hence you have a vey powerful "segmented multipass transcoder" that can be distributed over many cores on many hosts.


I use a 16core Nehalem MAC PRO and a new i7 mac mini (8cores) with an ethernet connection...


super fast... ,eaning very good transcode through put for compressor. (not Nuke etc...)


Anyway each of these segments CAN BE PAUSED in the "SHARE MONITOR.app"! As you inferred if you can paue the job and depending on the CODEC for output (the target codec(s)) used, you shoud only lose that part not already cmpleted.


Unlike RENDERING, transcoding is very tricky stuff, especially when all the parts have to be ASSEMBLEd into the complex object at the end.


I've looked for many years for such a versatile transcode that DOES THIS (multipass segmented transcoding) ON LINUX and mr softee... (Kulabyte promised but could not do it).. Other like Anystream and the bunch do it ony for some codecs.. and so on. Most launch as many transodes in paralell.. NOT like compressor.app .


Anyway.. look at the way you job it set up.


Its really very good.


HTH

w

Aug 30, 2011 5:22 PM in response to David M Brewer

David there's one thing I learnt from your post, read pop up message and don't assume.


Warwick, I attempted Qmaster when I was using Compressor 3 and like you I found it buggy. To be honest when I purchased compressor 4, I didn't even bother setting it up, thinking it hasn't changed.


Anyway after your advise, I re-set Qmaster thanks to this tutorial: http://vimeo.com/1799266


I'm now re-encoding the same .motn file (as a test) and have noticed a MASSIVE increase in render time.


From the dreadful 8hrs, it's drop to 2-3hrs, massive difference. To be honest that's only sharing the available cores on my Quad MacBook Pro.


Just quietly, I'll be setting up my other two machine to share their cores here on 😉

Aug 30, 2011 6:57 PM in response to blimpmedia

Hi Blimpmedia, sounds like you have a good improvment in the REDUCTION of your transcode and motion render times.


I had teased along compressor.app 2,3 & 3.5 for years and knew and posted many experiences and workarounds on these forums. I knew it would work when it was stable so I perservered with it.


I found it so excellant for MOTION rendering when I tripped over a note in the user manual.. lol! Also confirmed my Mark Spencer.


MOTION RENDERING NOTES with COMPRESSOR: take extreme care NOT to get too over confident and GREEDY with the number of cores to use when your envorinment only has less than an optimal amount of ram.


THat is for example each INSTANCE of the compressord using Motion may yse over 2GB of RAM! This means on say a 12GB ramm machine 4-5 cores running 100% doing motion will "FLATTEN" (bring the host to its knees) due to the lack of real memory and the paging to disk. Frankly I learned the hard way on this. One needs to power-on restart to get over it.


So be vigilant and use the ACTIVITY.MONITOR.app and whatch the REAL memory usage and the paging and swapping too!


TRANSCODING: on the other hand it has been my experience to throw EVERTY available core at transcoding! Its very spiky numeric intensive and CPU matters! Especially for distribution codecs where I can use segmented multipass transcoding!.. compressor/qmaster is ripe for this!


Yeah just watch the REAL MEMOERY usage with MOTION and COMPRESSOR usage.


Infact, I'd suggest this simple adjustment to your workflow in Motion.app.


  1. Save the .MOTN projects and DONT render them inside the motion.app unless you really need too.
  2. Shutdown MOTION.app if you have nothing left to do..( to get some real memory back)
  3. Start COMPRESSOR.app with APPLE QMASTER active for the APPLE QMASTER cluster you want and MAKE SURE you only have the numbers of compressor instances you need wit hthe available MEMORY you have. (e.g a ROT is 2GB/core)
  4. Using compressor.app, Submit each .MOTN project as a separate BATCH to your favourite APPLE QMASTER cluster
  5. monitor the REAL memory usage on each host using ACTIVTY MONITOR.app
  6. stand back and watch with amazement or go for a long cup of tea and lunch with the wife or G/F


FWIW my next MACPRO (thunderbold one when it is available - rumours prevail!), will have at least 48-64GB of RAM on it for this exact reason above! I am constrained to 12GB on this Nehalem MAC PRO.. fine but can be troublesome.


I'm also using this imilar setup with FCPX "send to compressor".. reults are wide.. I'm trying to utilise this new i7 MAC MIN for a service node with FCP and APPLE QMASTER cluster.. more as I get it stable.


Post your results with Apple QMASTER cluster.. very interested.


HTH


Warwick

Hong Kong - where is it very hot today! 😀

Pause/Resume

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple ID.