Nevin Styre
Forum Replies Created
-
I’ve heard CS6 can support multiple(nvidia) GPUs on some level in after effects for raytracing, but I don’t know if it applies to GTX cards or just quadro cards, and if so I don’t know if SLI is necessary or not. If you already have 1 580 stick with it for now though, ATI really isn’t an option in the CS6 world. If you just need to have more monitors you could always get a cheaper GPU like a gtx 550 or lower to add more displays, real CUDA work will still be done on the single 580, just make sure it’s another gtx card and the single driver should handle both with no conflicts.
Good choice not going with the 680/690, since they aren’t currently supported in AE CS6 for raytracing.
-
When you import your AI file select Composition but not the retain layer size option.
-
Adjustment layers are so powerful in AE I’m glad they brought them to premiere. Really though doing photoshop work I wonder why they don’t make the photoshop adjustment layers more like AE, right now you can do individual image control/colour correction adjustment layers but it would be great if they gave us blank adjustment layers we could add stackable filters to like in AE. That way I could add an adjustment layer then throw a darkening curve and a blur filter, then mask out the center to make a custom vignette… like I do in AE today.
-
With windows installed on the mac pro a newer CS6 supported GPU should work on the machine(in windows), though I wouldn’t spring for a powerhouse card as the PCI-E 1.1 interface will limit it compared to using it on a more modern motherboard that supports pci-e 2.0
-
While it likely has an anti aliasing filter like many/most DSLRs, the filter is designed for edges/detail on the full 18 megapixel resolution of stills on the camera. It isn’t designed for the sub-2 megapixel resolution of HD video. This is why you are supposed to take this after-market video anti aliasing filter off when taking stills, as it would likely make your 18mp stills pretty blurry as it is designed to “soften” the edges on a 1080p image.
-
It is moire that you are experiencing. Basically it is a limitation of DSLR’s because when they record video they are skipping lines on the sensor to record your 1080p image rather than taking the whole 18mp image and scaling it down(which would require way too much processing in the camera to do at 24-30fps). Fine details on things like shutters, bricks, power lines, and even some clothing will exhibit this rainbow pattern that you see because of the limitation
Bad news, it’s hard to fix the issue in post. Good news, there is a company making a product for exactly this limitation, it’s an optical anti-aliasing filter that you put behind your lens & in front of the sensor (you would want to take it off when shooting photos).
https://www.mosaicengineering.com/products/vaf/7d.html
-
Nothing wrong with it, but most of the time my source footage in any given timeline isn’t usually the same format(or even sometimes framerate-ugh), so I will select a format for my timeline that is higher quality than my source footage, like AVC-I, so preview files I render when editing are high quality & re rendering at the end is not necessary(for me). If I have a longer form project where most all of my source footage is all one of the supported formats then yeah I’m sure it will help, but in my mish-mash format AE roundtripping heavy workflow it’s not going to be the biggest deal to me.
-
Considering you can use AVC intra 100 as a timeline option, the extra generation doesn’t really mean a whole lot in over 99% of the work I do(non-broadcast). Going from DSLR h264->avc-intra->h264 or DVCPro->Avc-intra->H264 is fine with me for almost everything I do. Especially considering on average 20% or less of the public actually use HD stream options when viewing online video. If a project negates I need highest quality export I save that process it until approval & my final render, all my proofs I will do with previews. But for most things that will be watched online then forgotten after a week or 2(ie event promos), shaving 90%+ off the time exporting is worth it for my bottom line & my clients’.
-
This isn’t new or isn’t 100% new, in CS5.5 as long as I checked “use preview files” in the settings dialog when exporting my timeline, the sections already rendered with a green bar on the timeline were used for creating the exported file rather than re-rendering(sometimes the difference between a 20 minute export and a 1 minute export). That’s basically the same thing final cut is doing with rendered sections of timelines.
-
Yeah I was hoping that OpenCL would be just as good as CUDA as I was eventually get a new macbook pro to go alongside my windows tower & replace my aging 2006 MBP, but apparently it’s worse than using just the CPU with just “software” mercury playback(which I wouldn’t want to be stuck with after dropping $2.5k+ on a new laptop).
Considering the price of a decent MBP, HP workstation laptops aren’t priced a whole lot differently, spec wise you probably get more per dollar considering quadro GPU vs ati midrange gaming GPU, looks-wise and formfactor… well yeah they aren’t macbooks.