Joseph Owens
16
COW Coins
16
DaVinci Coins
Forum Replies Created
-
[Russell Anderson] “Are there any ways to increase transcoding speeds on an iMac?”
More efficient software, and higher speed data throughput between media storage devices.
I find that Adobe Media Encoder can transcode a frame rate conversion, for example, in almost “real time”, that is – a 45-minute 23.98 episode takes just under an hour to export a properly interlaced 3:2 cadenced broadcast master. Apple Compressor will consume about 14-15 hours to accomplish the same thing. In the former case, a Solid State e-SATA or Thunderbolt Drive would be an asset, but in the latter case would be a waste of capital because chiseling the digits on a stone tablet would be almost adequate.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
There might be two approaches that would be helpful, depending on where your comfort level is with managing the ‘split’ process.
Perhaps the simplest to grasp would be to clone your clip onto a higher layer and use the alpha channel to mask the grading process that you would apply to the lower plate. The foreground animation would be essentially “re-keyed” back into the master clip.
The second approach might be to introduce a parallel node tree to split out the background image from the animation — then your grades would be applied into the parallel process and the “separated” animation re-supered in the summing node.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
Still having some difficulty visualizing your image structure… Are you superimposing the animation with a separate track/layer enabled with an alpha /mask, or do you have a single image layer with the animation already composited into the image?
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
If what you are trying to describe is…
You have a timeline that you would like to apply a power grade to except for a few clips — the exception being your animated opening that you would prefer not to affect?
One possibility would be to group the entire timeline except for those instances and then apply a post-correction. That would be valid for only the clips in the group, and would not be applied to any of the clips outside the group.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Leonard Levy] “Are there any settings like “RT” in the old FCP 7 that facilitate real time playback in Resolve.”
No. WYSIWYG for playback speed is determined by your system build.
There is a cache hierarchy for complex corrections or add-on FX filters, but you need to play the clip through once for it to build. Search for “smart” cache.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Kris Kohuth] “I’m trying to find an effective workflow for using ProRes under Windows.”
Rewriting or rendering ProRes under Windows is not currently supported. It is somewhat unlikely to be enabled under Version 12, but I haven’t seen it yet.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
The alpha channel is not part of the normal Y’CbCr color channels so what you will need to do is split out the alpha so that it is accessible to Resolve for correction processing. Probably means gong back to the generating application and exporting a travelling matte movie that you would attach to the foreground plate. Not the workflow you would have preferred, most likely.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Shane Ross] “In short…export>Generate NEW AAF. That solved it.”
You will get a new timeline, true… but its not really a solution for seamless round tripping, because your new AAF is basically a virgin project and will not contain any of the unsupported Media Composer functions that would ordinarily be carried back and recreated in the original MC project sequence.
Better than nothing, though.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Lili Chin] “How do I read the scope and tell if I am going too far and inserting noise . Ie. how can I look at a scope (waveform/ vetor/ parade/ histogram) and read if there is noise in the image. The only way I know is if I drop the saturation to 0 and look at the vector to see if there are any dancing pixels,”
First of all there is a big difference between “noise” and “grain”… and both will be present in your image.
If you are cranking the image up by more than a stop or so, you will be compromising the picture integrity no matter what you do. Scopes (waveform and vectorscope) don’t really have a specific “noise” or “grain” measuring function — what you need to use is your judgement interpreting how “granular” the traces become. Modern software enhancement in the digital domain is very clean. You are unlikely to be “introducing” noise — what is probably the case is that your grade is amplifying a very large grain signature that is inherent in the original image. Super8 is an extremely tiny image to be blown up to anything more than SD and it would take a very capable scan to get it into HD or beyond. I expect its also reversal, so — dense to start with.
Its always worth a try, to contact the transfer facility and just hang it again to make sure everything that was in the emulsion made it out of the machine.
Otherwise, you will be into Noise Reduction.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
The traditional approach, but this has only been in practice since the mid-80s is to grade the picture lock — if you have issues with media-handling, there is a proxy workflow with link-back to source media. But if you can edit with source, then export an AAF, it minimizes some overhead-loading with file size. Even if you intend to use every frame, there is merit to grading a picture lock in a contextual story-flow, rather than simply grading everything “best light” and hoping that it will cut against each other.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.