1 Reply Latest reply: Feb 12, 2014 12:55 PM by Studio X
pcalvin Level 1 (10 points)

With the previous version of Compressor, I created a cluster with all four cores in my processor (Processor Name:          Intel Core i7, Processor Speed:          2.5 GHz, Number of Processors:          1, Total Number of Cores:          4).  Does Compressor 4.1.1 take advantage of all 4 cores, or do I have to form some kind of "cluster" to do so?  Also,  if I have another Mac on the same wireless network, can I use it for distributive computing, or do they have to be connected another way?



  • Studio X Level 7 (27,030 points)

    Compressor 4.1 has eliminated all of this.


    Submit the job with the ADVANCED>ENABLE ADDITIONAL COMPRESSOR INSTANCES checked and you'll get the equivalent of the old creating instances. Be aware C4 now looks at your ram and processors and decides the limit of usefulness for you. Put another way, you can't add more instances than your computer will support.


    A couple of caveats here:


    1. C4.1 has a different rendering engine under the hood so the way you worked before (using instances to max processor accessibility) does not necessarily mean faster encodes.


    2. There was a quite a bit of overhead when using multiple instances in C3 which often added significant time to the overall render when C3 had to reassemble the pieces. We all had to spend a lot of time by trial and error finding the "sweet spot" for the settings.


    3. C4 has a more efficient rendering engine (see point #1) which means it doesn't have to work as hard for the same results. Think of it this way, if you and an professional cyclist both start riding up a hill, you'll be working at max effort very quickly while the pro cyclist (who is MUCH more efficient)jwill be way further along for the same effort (or can keep pace with you for significantly less effort).


    My advice, stop looking at the processor involvement meters and simply time the results. C4.1.1 is faster than  C3.5 (even with multiple instances) for similar encodes.