Sean Oneil
Forum Replies Created
-
I’m predicting a free update to FCP 6.5, etc. that will work with Snow Leopard’s new features like Grand Central and OpenCL. Just like when 4.5 was a free update to officially support HD formats.
Just looking at it from a business perspective. If I’m Apple I’d want to encourage pro users to upgrade to Snow Leopard and replace any remaining G5s with Intels. A free update to FCS is a great way to do that. Then next April they can come out with a totally revamped version.
Sean
-
Sean Oneil
March 9, 2009 at 7:18 pm in reply to: OT A RAID made out of 24 256G SSD’s – Thats nuts! (and fast)It’s a shame those things are still really expensive. The cheapest 256GB SSD is $500.
Sean
-
[Bruce Feagle] “Doesn’t ProRes use YUV color space and Animation RGB?”
I don’t think that’s a serious problem. I’d be more concerned with all the generations of lossy compression going on. Animation is lossless but ProRes and RED RAW are not.
Sean
-
Check out freenas.org – that is without question the best method for a DIY NAS.
Sean
-
[Darren D’Agostino] “I could buy almost 20 SATA drives for $2,000 + add the cost of controller cards and chassis to house them, + the actual server + hardware firewall + wires and maybe a new power backup. I don’t see how that should reach $10,000 or more even if I spent $2,000 – $3,000 on the actual server and $2,000 on drives (which I wouldn’t need that many anyway).”
If this is simply for backups, as you say it is, why would you want 20 SATA disks striped together?
Sean
-
[Ben Holmes] “it doesn’t look like the new machines will be any faster per dollar spent”
I would wait for the barefeats verdict before claiming that. There’s a lot more to these new chips. It’s not just number of cores and ghz. For example they have hyperthreading, so one core acts as two.
Sean
-
[Paul Dickin] “The same 113% Turbo boost to a 2.26GHz CPU would bring it up to nearly 2.6GHz…”
I’m talking about making the Mac think the 2.26ghz is actually a 3.2ghz chip. They’re all the same die and come off the same line. Some are just QC’d to run at the higher rates. So it’s frustrating to see such a large price gap.
Sean
-
Sean Oneil
March 5, 2009 at 11:18 pm in reply to: Largest Video Codec Capture to 3x 1TB internal MacPro StripeThe new tech that makes these drives larger also makes them faster. So I’m willing to bet 3x 2TB drives will be fast enough for Uncompressed HD.
But in a critical environment you don’t want to even cut it close. The drives will slow down as they get fragmented. You’re much better of going with an external SATA enclosure with 6 disks or more if you want to use U/C HD.
If your clientele does not pay well enough for you to afford such a paltry expense, you shouldn’t be giving them U/C HD anyways. Use ProRes as others here just mentioned.
Sean
-
People don’t really “choose a codec” based on artistic decisions. It’s not like choosing different film stock. Artistic decisions are made with camera choice, lenses, lighting etc. Codec choice is based on whatever camera you’re using, budget, and other workflow related decisions. Some codecs are higher quality and some are not. A higher quality codec will provide a cleaner image and be easier to color grade, do chroma keying and things like that. But again it’s not like an artistic decision. It has more to do with the camera and the overall workflow, like choosing to shoot on a RED camera vs. HDV. Each one uses it’s own respective codec, but the codec itself not really the basis for artistic decisions. Nobody’s ever going to say “Hmmm, I could shoot RED but I think I prefer the look of HDV for this shot.” It doesn’t work that way.
There are times when lower quality video can be used for artistic decisions (like “28 Days Later”) but that has more to do with the camera than the particular video compression format used.
Sean
-
Thanks, I’ll try to check out your blog from time to time.
[Eric Hansen] “i installed a Mac Pro version of this server setup and the speed tests from the client were between 80 and 102MB/s over Gigabit”
I haven’t been able to achieve that. Using a Mac Pro, I connected to another Mac Pro and mounted it’s MaxDigital disk. Read speed was around 100, but write speed occasionally dipped to less than 40. Could be a number of factors. Maybe that sustained transfer feature on the CalDigit card helps it out a lot.
[Eric Hansen] “i assume that the tests with this G5 would be similar since all the hardware is the same (well, minus the processor, but these transfers are not taking much processor)”
I think the processor actually has a lot to do with TCP performance.
Sean