Skip to main content

Compression Hardware, Software and Codecs

Lesson 1 from: Faster Video Compression

Larry Jordan

Compression Hardware, Software and Codecs

Lesson 1 from: Faster Video Compression

Larry Jordan

buy this class

$00

$00
Sale Ends Soon!

starting under

$13/month*

Unlock this classplus 2200+ more >

Lesson Info

1. Compression Hardware, Software and Codecs

Lesson Info

Compression Hardware, Software and Codecs

video compression is unnecessary. Evil takes a long time, and you don't know if your settings are any good until compression is complete. So today I want to answer these questions. Which hardware works the fastest? Which software works the fastest. Which Kodak works the fastest? How do you optimize settings for YouTube video and Facebook? And how should you optimize settings when you're creating a downloadable video? Over the last month, I've been testing various combinations of hardware, codex source files, compression software and settings and written about them in a series of four articles. Here's a really good article to start with that gets you access to the entire Siri's goto larry Jordan dot com slash articles slash speed test Compare video compression software. Notice all the hyphens. This is one that I where I compared six different video compression software running on a Mac Pro in an I Mac and a variety of different settings, and we're gonna talk about the results of this te...

sting and other testing that we've done in today's Webinar First questions. Which hardware should you use if you're compressing video for the Web? While a Mac book Pro was marginally faster, the Mac is the best choice for compressing Web video at best. A new Mac Pro was roughly half assed. Fast is an I Mac when you're picking the I Mac. What matters is CPU speed, GPU, speed and throughput. Storing files on an internal SST versus an external hard drive. And the source Kodak of your video influences. Compression speed will talk more about that in just a minute. The Mac Book Pro was probably 5 to 10% faster than the Eye Mac, but it's also multiple times more expensive. So the best price performance isn't I. Mac 21 or inch. Both EI Mac and the Mac book Pro Use Hardware and GPU acceleration when compressing a video for the Web using the h 0.2 64. Kodak, The new Mac Pro does not support hardware acceleration, which is a significant detriment when it comes to video compression. And as I mentioned I max or less expensive than Mac book pros, Mac Book Pro is faster, slightly, but the I Mac is much more affordable, especially if you're getting it for video compression. If you're trance coding from camera native formats to pro reds for two to this does not require hardware acceleration. So I would expect a new Mac pro to be faster than the Eye Mac due to its architecture and GPU speed. When we are trans coding for editing and say Final Cut Pro 10. When you're trying to side which software, Here's the results of what was fastest. If you're running a Mac Pro handbrake with a compression setting of V B R. One passed h 10.2 64 it's absolutely the fastest, followed by compressor variable bit rate single pass, followed by EMP. Extreme clip, variable bit rate, single pass, followed by hand brake, variable bit, right to pass and Adobe Media Encoder VI Br one pass before on the night back Compressor vi br one passes the fastest, followed by hand brake, followed by compressor with multiple instances turned on, followed by compressor to pass, followed by hand brake to pass. There's a big difference in how the two computers compress. We'll talk more about that in just a second Adobe media encoder, and this is a really huge note. Adobe Media Encoder has a new version coming out very soon, which I havent tested yet cause I only test the released versions of the software I'm looking forward to seeing when the new version does arrive, whether it's any faster, which could change the results of these tests. MPEG Stream clip is fast, but it failed three out of three times when it was running in to pass. Modi could not reliably get it to compress when it was in two pass mode. For this reason, I'm having a hard time recommending stream clip just on general principles. I also tested Tell the Stream episode of Sorenson Squeeze, and those were the two slowest at compressing video for the Web. Now there's lots of different video formats. There's lots of different reasons to compress from, say, an M XF file. Toe a pro reds for two to or from H 20.2 64 ABC HD to progress because you want a trance code convert to file from a camera native format to an editing format. I didn't test any of that because that's going to be a whole different round of parameters and testing results. I was looking specifically at creating a video for the Web using the h 0.2 64 Kodak. Another question was asked about whether the ex dot to 64 Kodak would work. This is the open source version of H 640.2 64. We didn't have any software that ran the X stop to 64 Kodak that was installed, and I wanted to stay with this much of the installed default settings as possible so we could get a better look at the architecture of the software without introducing the vagaries of a different Kodak. I would expect ex dot to 64 to be similar in speed toe h 640.2 64 when you running Adobe Media Encoder. The new Mac Pro was roughly 50% faster, then the same Mac Pro running the latest version of compressor, which is 4.2 compressing the h 0. 64 from a pro rez for two to source file was about twice as fast as compressing from X T. Cam noticed that when the source file format changed, compression speed changed source files and programs four by four Compressed about this fast is pro rez for two to H Q. And I would expect all pro rez versions to compress equally quickly, though it must be said that in most situations progress four by four compressed oh, about 5% faster, then progres for two to H Q. When you're working with apple compressor compressor was roughly twice as fast running on an I Mac compared to running on a new Mac pro. There was no compression software or scenario when compressing video for the Web, where the Mac pro was faster than the Eye Mac and the screen size of the I Mac 21 inch or 27 inch did not make a difference, but did make a difference was CPU and GPU. Speed compressor has an option called multiple instances. Multiple instances allows multiple iterations of apple compressor to run at the same time, If you are running on a Mac pro tern, multiple instances on it compressed the same file between 11 and 47% faster. And when you're running multiple instances, X'd cam showed better results than progress. However, you want to leave multiple instances off when you're using an I Mac or a Mac book pro multiple instances compress the same file roughly three times slower. When multiple instances air turned on pro rez compressed faster the next Zicam by about 20%. Again, changing Codex of the Source file has an influence on how fast you can compress. I only tested extra cam Progress four by four and pro rez for two to H Q. And of the three extra cam compressed more slowly than either version of progress, probably due to its got its group of pictures structure. Extra cam was extremely slow when compressed, using multiple instances taking almost 10 times longer to compress than when multiple instances were not enabled. Pro Rez four by four compression speeds were always the fastest, except when using Tell a stream episode and Adobe Media encoder in to pass mode. So the general recommendation that I have is, if you have the option, master your projects for either adobe media encoder or compressor master using pro rez rather than other Codex. Trans coating not only saves time during the editing process, but it will save time and sometimes significant time during the compression process. Which settings should you use if you're compressing videos for YouTube video or Facebook, all of which re compress your video into multiple different formats? Always use V p r. One pass its variable bit rate. One pass. The default rate that set in the different compression software is between 10,000 and 16, K bits per second, and that default setting is fine. And there's no significant difference in image quality between a 10,000 K bps compression rate on 16,000. So just leave the default alone, and there was absolutely no advantage to to pass when you're bit rates are set 10-K or above. For downloadable files, though, the situation is much more complex because of needing to keep the files small so they download quickly and you keep the quality high, and here settings will vary by software.

RELATED ARTICLES

RELATED ARTICLES