Wednesday
Aug112010
Apple verses Adobe, who compresses faster?
Wednesday, August 11, 2010 at 6:49PM I also want to mention that in this compressed video the difference between the two version look much more dramatic than it did as I made the screen grab. I sort of sound like an idiot saying they look comparable. It was odd that the Apple default bit rate was nearly TWICE what Adobe reommended for an iPhone video. Not sure why that is.
Reader Comments (12)
Could you do the test with the cluster enabled to see what that difference is? We have used the virtual cluster on all our machines and I've been pretty pleased with it. I did try to do a comparable compression for blu-ray with the Adobe Media Encoder and Compressor. I may have been doing something wrong but Media Encoder was about 5x realtime and compressor was like .75x realtime. I don't remember exactly, but it was significantly different. I need to try Media Encoder more. I'm just used to Compressor, so in a pinch I default to that. Thanks for posting this.
p.s. Computer specs were MacPro/octocore/2.26/16GB Ram
You have compared apples to well.....something else. The excuse for not optimizing Compressor is really lame. I am not a Compressor fanboy, but considering you optimized for Adobe, your test seems pretty biased.
In what way did I "optimize" for Adobe... I missed something.
Very interesting, Chris - thanks a lot. And I agree with your comment - I certainly couldn't see how you "optimized" Adobe. Just another example of an Apple apologist - even direct proof is not enough to convince those who drink the Steve Jobs Kool-Aid. Steve says the world is flat, so it just "is".
I really hope Grant G. comes back and explains how I "optimized for Adobe". I'd like to know... Maybe he thinks that because I DIDN"T optimize for Apple it was "as if" I had actually optimized for Adobe. I guess that COULD be it. The fact of the matter is that I just used both apps the way MOST people would use it, in the default settings. If I cared enough I'd do a reshoot with Compressor set up with a virtual cluster. But frankly, I DON'T care enough.
Chris,
I used this document - which you are right, took some finding but it's a keeper when it comes to cluster rendering in Compressor.
http://www.devia.be/news/article/setting-up-a-virtual-cluster-to-speed-up-compressor/
Thanks for your video, it's great!
I noticed myself that encoder vs compressor it's faster. I've never even thought it's that fast.
I am getting quite some errors with compressor lately, errors in the final product, so I think I will try and switch to encoder.
Any ideas how to integrate encoder with fcp?
any ideas why I ended up after using compressor, instead of a video, I got white background all 29min but the audio was fine? Doing it again worked, but the size of the file is the same as the one with white background. It's puzzling.
Really great article ,thanks for this useful information :)
As you can tell, I'm writing this now after the release of FCPX and although I guess you can call me an Apple fanboy and I'll most likely learn and use FCPX for projects where it's applicable, I am trying to get a handle on Adobe products because I think the market will lean more this way if everyone abandons Final Cut like they're saying. It's certainly cheaper than Avid and ties in with After Effects and Photoshop really nicely. I'm curious about the settings used in the test though. Did Compressor run 32 bit in this test while Media Encoder ran 64 bit? Will using the "cluster settings" that you discussed make Compressor run faster? I also noticed some banding artifacts in the Media Encoder result, which also ran a tad darker. Do you think this is because Compressor spit it out at almost twice the bitrate? Is adjusting that in Media Encoder as easy as it is in Compressor? And how much will that effect it's speed? If Media Encoder really is as fast as you've shown, I'll have to seriously consider adjusting my workflow. Thanks.
The BIG gotcha that people forget with Compressor when using a cluster is to turn Cluster Options to "Never copy source to cluster".
It's in the preferences for Compressor. If you set up a local cluster, this will slow down the time it takes to render a large file, as it wants to copy the source to the local hard drive. Often the cause of crashes if your boot drive doesn't have enough free space.
Using Compressor 4 with sharing turned on is much faster than Media Encoder CS5.5 and usually gives better results. I don't know why but all of my exports out of media encoder look terrible, artifacting and noise, just general low quality and it takes about twice as long as Compressor 4 using 3 of my 4 cores. Exporting h.264 10mbs @ 1280x720.
Given the recent events with Final Cut Pro X, I'm certainly looking to jump ship from FCS 2, even after literally learning the trade on Final Cut since version 1.2. Everything you say about Compressor is true of version 3 & earlier but there are 2 things that come to mind with this horse race.
First, I notice that the Compressor file is 1500 Kbps whereas the A.M.E. created file is 800 Kbps...HALF the data rate, right? That would presumably take half the time to create. It's not really "essentially the same thing" as the Compressor file. Also, I notice a very slight color variation between the two, with my eye favoring the Compressor file. Then again, it's also larger and took 4 times longer to compress. Even still, we're watching this in yet a third compressed video file.
Second, how does Compressor 4, which was released alongside the bastard iMovie Pro fair in this horse race? I know that chiefly, it is a multi-threaded app that can take advantage of all processor cores and avoid the Virtual Cluster(f#@%) that you had to setup with previous versions. It also integrates & streamlines Qmaster for using network resources, something which was a pain to setup before.
If Compressor 4 might offer a performance and/or quality advantage over Adobe Media Encoder, then it's certainly worth consideration on it's own merits. In spite of the fiasco that is Final Cut Pro X.