IS Compressor 3.5 slower than 3.0?

[Ben Holmes]

Oh – Forgot to mention the franky scary Motion performance test figures as well.

Edit Out Ltd
—————————-
FCP Editor/Trainer/System Consultant
EVS/VT Supervisor for live broadcast
RED camera transfer/post
Independent Director/Producer

https://www.blackmagic-design.com/casestudies/detail.asp?case=therydercup

[Warren Eig]

Apparently.

https://www.barefeats.com/fcp7.html

Warren Eig
O 310-470-0905

email: warren@babyboompictures.com
website: https://www.babyboompictures.com

https://www.babyboompictures.com/BabyBoomPictures/AFX.html
https://www.babyboompictures.com/BabyBoomPictures/KnitWits_Movie.html
https://www.imdb.com/name/nm0251670/

EDITING REEL: https://www.babyboompictures.com/BabyBoomPictures/Editing_Reel.html
TITLE DESIGN: https://www.babyboompictures.com/BabyBoomPictures/Titles_Reel.html

[Jeremy Garchow]

I don’t really know about Motion, but what I notice with Compressor is that is takes a while for it to figure out what’s going on and ramp up to the encode. Once it finally gets encoding, it’s really fast.

The Share function is useless if you are using a virtual cluster.

I’m sure it will get ‘tuned up’ in a future release, at least I hope it will.

[Matt Lyon]

Did I understand this test correctly? They downloaded some h264 half res HD footage and re-compressed it to h264 HD DVD?

Also, they only did one trial for each test?

Sorry to be a grump, but these results aren’t really scientific. I’d be more interested on seeing some tests encoding pro res or uncompressed footage to h264. Also, there is lots of overhead that compressor has to do when encoding on a virtual cluster (splitting up clips and re-combining), so running tests on such a short clip (2min) is problematic. I’d be interested in seeing come comparisons of batches of longer materials (10 minutes and above).

I’m not saying these numbers aren’t alarming … but these tests are a little sloppy, imho.

Matt Lyon
Editor
Toronto

[Ben Holmes]

I’d hope so too Jeremy – Although, I fail to see what Apple could have done to mess up encoding in Compressor – as I said it’s just a front end on some encoding maths. Maybe it’s a quicktime issue…

Edit Out Ltd
—————————-
FCP Editor/Trainer/System Consultant
EVS/VT Supervisor for live broadcast
RED camera transfer/post
Independent Director/Producer

https://www.blackmagic-design.com/casestudies/detail.asp?case=therydercup

[Ben Holmes]

That’s the link I put in the first post Warren….

Edit Out Ltd
—————————-
FCP Editor/Trainer/System Consultant
EVS/VT Supervisor for live broadcast
RED camera transfer/post
Independent Director/Producer

https://www.blackmagic-design.com/casestudies/detail.asp?case=therydercup

[Jeremy Garchow]

[Ben Holmes] “Although, I fail to see what Apple could have done to mess up encoding in Compressor “

I don’t think they did, but with the new MacPros and Now Leopard there’s Hyper-threading going on so each core now essentially becomes two processors. It’s a lot more to think about, initially.

[Ben Holmes]

Barefeats are quick to add they need to do further tests. BUT a) they have a good track record on benchmarking and b) they were performing exactly the same test in each version – so whether or not it’s the right comparison, at least it’s a fair comparison.

I’ll be watching carefully for a follow-up, but anecdotally I think they’re right – Motion feels sluggish to me even on simple HD projects, and Compressor is slooooow…

Edit Out Ltd
—————————-
FCP Editor/Trainer/System Consultant
EVS/VT Supervisor for live broadcast
RED camera transfer/post
Independent Director/Producer

https://www.blackmagic-design.com/casestudies/detail.asp?case=therydercup

[Jeremy Garchow]

[Ben Holmes] “Compressor is slooooow… “

I wouldn’t go that far. It’s actually pretty fast if setup properly. It chews Episode up.