Will RR (original) benefit from TB2 Enclosure?

[Adam Howden]

Great info in this thread. I'm about to receive my RRX and am wanting to run it in a Sonnet Echo Express III-D as a mobile solution on a latest macbook Pro. I'm curious about the 16x lane (8 electrical) specs, can anyone here explain that briefly?

Sorry not meaning to hijack.

 

[Vico Martin]

In my experience - scaling on the Rocket is faster than software scale - which is why I always use hardware scale on dailies. I can run some more tests on a tricked out machine - but I have never encountered faster scale via software vs. hardware - so ... I agree on the ability to tweek the sharpness and scale algorithm - Mitchell better for UP - Lanczos better for down, etc. - but speed .... I have not seen software scale beat the hardware scale in terms of speed.

Mark, can you choose hardware scaling outside Redcine-X??

 

[JASPER GLAVANICS]

I have been testing my RRX in a Sonnet Echo Express III-D, and on my new Mac Pro, I didn't see much use in having it. The Mac Pro performs amazingly well, and the RRX actually caused issues when exporting from Premiere. I think the issue was when the export sequence went from 5K footage to 1920 HD footage (used from 5D aerials). The export stalled and froze my whole system. Did this a number of times until I unplugged the RRX and then it exported the sequence in the same time as with the RRX. So when mixing different file formats as an export .. this equals difficulties.

I haven't however tried it on the new Mac Book Pro.... can someone post how their experience with that setup was?

 

[Shiblon Wixom]

The Rocket-X has some scaling issues right at the moment, hopefully it's fixed soon, had issues going from 6k or 5k FF to 1080p, but 6k HD and 5k HD to 1080p seemed to work fine.

 

[Shiblon Wixom]

Shiblon,

Something is off then, you should see big speed difference between all of that. Thunderbolt 1 (and 2) will not be able to full res 6K debayer at 38+ fps. Math will easily proof that. Sounds like you are scaling on the Rocket-X and exporting, so your export codec could be the limit (encode speed limit). Won't know until you provide more details.

Rob

I ran a few more Transcode timed tests with Redcine-X and a Magma TB2 Chassis and Rocket-X, again, using 6kHD Dragon 30 sec and 1 min clips, timed tests in Redcine-X were aprox. 15 seconds for the 30 second clip and aprox. 30 seconds for the 1 min clip. I also timed 5k Epic and 4k Red One clips at close to those same results, in every case the system is using around 55% to 65% of system resources. This is transcoding to ProRes 4444 1080p RedLogFilm.

When I transcode the same clips in Redcine-X to ProRes 4444 4k HD output, then the transcode time dropped to a 1:1 time basis and the system seemed to be maxxing out all available system resources.

I did remove all the extraneous PCI-e cards in the TB2 Magma Chassis and in the quad-core 4GHz 4770k hackintosh, which seemed to improve the transcode speeds a little vs my first round of timed transcode speed tests. I'm also transcoding to/from an internal Areca HBA Controller and Raid5 x8-Drive 15k rpm SAS drive array.

 

[Blair S. Paulsen]

In my experience - scaling on the Rocket is faster than software scale - which is why I always use hardware scale on dailies. I can run some more tests on a tricked out machine - but I have never encountered faster scale via software vs. hardware - so ... I agree on the ability to tweek the sharpness and scale algorithm - Mitchell better for UP - Lanczos better for down, etc. - but speed .... I have not seen software scale beat the hardware scale in terms of speed.

My point was in reference to the delta between doing scaling AND sharpening as separate nodes/processes, VS, using a scaling filter that by the very nature of its inherent properties adjusts the perceived sharpness to my intended level. Sort of a 2 for 1 strategy. You mention Lanczos for downscaling and its always the first one I audition for that, though certain material is better served by other "flavors". IMO, when you can achieve the right amount of "sharpness" via the downscaling process vs just adding sharpness digitally you are more likely to retain a more natural rendition.

YMMV, some material will not be well suited to any particular formula, be creative.

Cheers - #19