Joseph Owens
16
COW Coins
16
DaVinci Coins
Forum Replies Created
-
The other issue is that the 4,1 may not support a newer nVidia card at all; its a BIOS problem, and you will have to try to run a newer OS, Mountain Lion was the first to start supporting non-EFI nVidia cards and then the support is ambiguous as far as number of cards supported and so on. In any event you will have joined the rest of us fighting the constant battle with Apple over Web Drivers and version compatibility. Its a nightmare.
jPo, CSI
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
A technique that many compositors use (mostly to make “clean plates”) is to re-mix a negative reference image with the original — when the register is perfect, you will only see a blank grey frame. Anything out of kilter will show up as a “difference” edge… jimmy the PTRZ with key frames until all the edges go away and the bump will be gone.
jPo, CSI
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
Seems like a GPU failure, or it is a CPU decoding deficiency. The iMac is not particularly well suited to above-HD resolutions, and you have loaded it down with a fairly heavy media extraction task. While most grade software has evolved from .cin and .dpx sequential image formats, the original hardware platforms were Linux/Unix servers. Commodity-level PCs seem to deal more readily with .mov formats, with the best compromise centering around intra- flow codecs like ProRes and DnX — and this seems to be a defining difference between producing motion media as an animation/photographic methodology vs. a cinematographic one.
jPo, CSI
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Ben Dittrich] ” between two cuts, the Viewer of the GUI Monitor has a short freeze on every edit”
Do you have any cache-ing or proxy options enabled?
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Michael McCune] “In practice this is usually not an major issue.”
It would be a major issue, though. Banding, level discontinuity, is normally associated with lower density chroma-sampling and is very common with 8-bit codecs and monitoring. It is not inherent in Resolve process computations which are carried out in 32-bit float.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Paul Campbell] “Is transcoding to ProRes going to make my product look better?”
Not really. Unless you do some chroma-sampling tricks and macro-block smoothing. The whole notion behind recoding H.264 to a codec like ProRes is to make the CPU usage more efficient. H.264 is a Long-GOP format, which means that the system needs to make whole “intra” frames on-the-fly out of the i-b-p sequence. Intermediate density codecs like ProRes and DnX are natively “intra” which means every frame is a whole frame and not a bi-directional or predictive “difference” entity. The faster you skip around in a Resolve timeline, the harder the CPU has to work in order to keep the video smooth.
Eventually you start getting gaps and skips, black holes and decoding errors when the processing overloads the system’s ability to keep up the bitrate and – above all – do the translation into integer-float so that the application can do its work.
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Matthew Kingdon] “So this is where it gets really weird, the first 4 clips on my sequence are all from the same original long file, the first 3 have connected, but the 4th hasn’t. If I right click and click on “Find in Media Pool” it doesn’t find it, whereas if I click on the clip before and do the same, it shows it in the media pool. They’re both the same damn original clip!! “
Starting to have the feel of a bad reference to a spanned series of source clips. Your reference to an original “long” file makes me suspect this. Sony A7 can and does span media over a series of otherwise succeeding camera clips. I just hit this myself conforming a feature length documentary where the cameras were allowed to run for extended periods of time. In a number of instances the whole trim duration fell inside a specific source file, but if it ever crossed the hand-over, the clip went off line or incomplete. This might suggest why the overall “take” is described as a Compound Clip — because it is composed of several files meta-patched together as a spanned series. If my theory holds any water….
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
What is your media type that the reference clip is exported in?
jPo
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
Well there you go and for the price of a cup of coffee. World Vision must be feeling the heat of competition.
What I did come away with is maybe a better appreciation of what a loaded term “properly exposed” is these days.
It does seem to illustrate the dangers of depending on a single reverse-transform LUT to restore the viewfinder values over a range of exposures and lighting conditions. Which is something we colorists seem to be battling every time we roll into a new scene. Without the EI index as metadata, the LUT gets applied to the wrong exposure range — and I think anyone who has had to deal with a baked-in ProRes or DnX S-Log source clip has experienced this and realized that the offset has to be applied before linearizing or expanding the compressed dynamic with a reverse LUT. My first guess would be that if you have ever had a scene blow out the top of the scope, that’s why.
Two things… the article does not go into how this is affected in ACES.
And can’t resist the notion that whoever that guy is that wrote his connoisseurship article on “Why Marvel movies look like mud (or something)” could absorb a few lessons. The main benefit, of course, in writing an article of his type is that it definitively illustrates the perils of a little bit of knowledge. And BTW, did it break LIftGammaGain…? Its been down for days.
jPo, CSI
“I always pass on free advice — its never of any use to me” Oscar Wilde.
-
[Tim Wilde] “(Rec709 assumed)
“Seems like a Photoshop kind of question.
https://en.wikipedia.org/wiki/Middle_gray
Its not discussed much in cinematography. Its a legacy thing and with display-referenced grading, the emphasis has not been on hitting absolute RGB values.
In short, the 8-bit RGB values are roughly 128*3 in Gamma 2.2 and about 119*3 for gamma 2.4 and if you want that in 10-bit, the math is fairly straightforward.X-Rite also publishes the target values represented on their reference charts.
https://xritephoto.com/ph_product_overview.aspx?ID=824&Action=Support&SupportID=5159
jPo, CSI
“I always pass on free advice — its never of any use to me” Oscar Wilde.