Joseph Owens
16
COW Coins
16
DaVinci Coins
Forum Replies Created
-
Most of these items are ones that I have real concerns with as well — unfortunately many imbedded filters are not re-transmitted either accurately or at all as part of the AAF round trip. If there are issues with aspect ratio or re-sizing, sometimes the PTRZ translations are not exact, either — typically, though these should flow-through the round trip without being affected — the caveat being that for example, a previously non-square clip will, of course behave differently if it has been transcoded through the re-render process. The whole round trip strategy works better if there isn’t a lot of re-scaling going on because that math gets weird, and of course there are some processes that are just opaque between manufacturers.
Generally, though the direct import of a roundtrip AAF with single-directory exposure of the graded clips should be less problematic as a re-conform than trying to do absolutely everything from scratch with an attempted re-link. If there are really that many imbedded filters… to a certain extent… how would you know that the correction that was being applied was either valid or contributing to improving the final product? Granted there is not the transparency between NLEs with the ability to copy and paste attributes like that found in FCP7 — at the very least, its not as though you’ve been forced into a position of returning a baked complete timeline. At least, not yet.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Dave Williams] “Does anyone know if there is a way to point to a folder with sub folders holding individual clips from Resolve and have them relink to the show”
Maybe I am missing something here, but — what are they doing?
Did the exported AAF from Resolve not work?
Wa the exported graded media not in a single MXF directory and numbered according to its position on the timeline?
There is a ball being dropped somewhere.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Andy Winter] “this might sound strange, but i would love to have a HUE vs COLOR TEMPERATURE curve!
HUE vs HUE is good in certain situations, but it is limited to just change the hue angle. if i could pick a certain
color-range and make it warmer or colder, a lot of skin-tweaks or look-changes would go a lot easier!
i am sure this would be a great creative new colorist-tool.
-a color temperature slider.
i know i am a colourist and should be able to set this with my balls (no pun intended) correctly (and yes i can!) but
this would not only be a faster and cleaner way, but would also empower me to communicate with my client or the dop
in an easy way we both understand. (eg. can you make the picture 200 kelvins colder?)”It is true that a “colour temperature” control is featured in a number of applications that people use for color correction. There is an implementation of it in the RAW adjustment settings section for the appropriate
formats where metadata is available. A heated debate broke out about a control of this type a couple of months ago on the BMD forum.The gist of that went along the lines of… Resolve attempts to be an exhaustively correct instrument that works within the actual color science, rather than provide a bank of Fisher-Price knobs that, while providing the illusion of making the correction that a user is telling their clients they are accomplishing is merely synthesizing a warm-cool balance that has nothing whatsoever in common with a CIE-Planckian locus transformation. This is fine for wedding videos — but is a pretend transformation that is outside of the color science that other frameworks depend on, such as the ACES workspace, or a LUTted predictive display for example if you wish to work in a P3 colorspace using a 709 display. All of which is a bit more involved mathematically — and basically impossible to implement with media that has no innate base-point color temperature framework. In order to make a Planckian transform, you have to know where the media started — what the reference Kelvin point was when the original image was created. This is part of the issue with (no) respect to that bloody dress. What you are really asking for, as I responded early on in the BMD debate, is a mired-shift control.
Maybe my industry experience is a little skimpy, but within the more recent twenty years working exclusively as a colorist, the only “other” language I’ve ever heard a DP use to warm/cool a shot is to ask for a “couple of points” one way or the other. Without a specific stock/colorimetry densitometric relationship (and that would be relative to LAB aim points for a particular stock) the only practical way to respond is by feel, and the target is acquired when the client says “woof.” Two hundred K colder might mean +3 or 4 points of blue… or minus that many red, depending on where you are relative to 25-25-25… still a stab in the dark.
The show build of Resolve 12 featured a number of refinements that look a little bit like some of the things you would like to see, including redefining how some of the nodes are represented — the recombining ones, for example, and you will be able to lasso a complex node tree and essentially “nest” that correction into a single node… stuff like that.
I do agree that (while I am not happy to be essentially re-editing an erstwhile finished project) constantly having to re-position PTRZ up and down a timeline and not really being able to have the exact match immediately available or figuring out which pan/tilt tool is appropriate — between the Edit and Color pages — is irritating.
I think help is on the way though.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Terence Morris] “do I then need to manually convert them to an editable format like Pro Res before adding them to the Media Pool? Or is this something that Resolve can do on the hop, like FCPX?”
It becomes a matter of processing power and convenience. If Resolve recognizes your media, you are good to go, although there may be some caveats with respect to how much energy the application will have to expend on building the intra frames if it is a Long-GOP format. Occasionally there have been issues with secondary processes involving tracking and variable speed frame remapping.
ProRes is optimized as an intra (all frames are real) format, so the bitstream doesn’t have to be digested multiple times by the machine before you get a solid playback, so it doesn’t have to bother re-mapping the b/p fields so that all other optical processes are much easier and faster at the CPU level.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
Unless Apple changes its mind on whether or not you need or want more than 2 GPUs, and certainly not if they are nVidia, then I don’t ever expect any change on that matter.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Enrique Londaits] “without actually recompressing the footage.”
So this is where I am suddenly not secure in my understanding of what you are trying to accomplish.
The way Resolve works is that you create (render) new, color-corrected media, which is a kind of recompression. The color corrections do not travel with the source media as a filter, effect or metadata. If you make editorial changes in the timeline — depending on the source edit, whether it originated in FCP or PP or MC, you do have options regarding export of the sequence as a modified version of the original XML, AAF or as a completely new composition — which is mostly the case in trying to round-trip an AAF. Some editorial changes will nullify the ability to use the original with revisions — it will corrupt on export.
Where that happens it is my experience that exporting a new XML from Resolve, re-opening that XML in a new project and then exporting a new AAF will often result in a good timeline.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Sascha Haber] “I tend to save out an EDL and apply that to a source file, i never use Clip to mediapool as this kills the option to fade from one grading to another.
“That would seem to be the wiser methodology — the clip approach even makes it impossible to slip an event if the marker is wrong by a frame.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
Here’s another one: DIY with things you have around the house —
https://www.faculty.idc.ac.il/smozes/8mm.htmljPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Michael Gissing] “Marquess of Queensbury “
Handbags at 3 paces, I should think.
Used to be the convention in F1, anyway.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
built-in graphics? strongly doubt it would get past the splash page.
jPo“I always pass on free advice — its never of any use to me” Oscar Wilde.