Richard Herd
Forum Replies Created
-
[james sandifer] “100% consistent”
It can’t be done. Trying to explain this to a marketing department is useless. I’ve tried many many times. They just can’t seem to grasp it. And there is no good solve. The whole issue arises from additive color v. reflective/masked color models. In Illustrator you can Recolor Artwork > Edit > Display Segmented Color Wheel.
That will give you the range of PMS that RGB can make (or more likely “get close to making”). After Effects cannot recognize CMYK, so it’s best to always always always make a duplicate of the CMYK artwork and call it [file name] RGB.ai. In that duplicate, Recolor Artwork (as above), and then change to RGB. It’s also, I think, wise to make your own color swatches in the .ai document in their own layer, so when you bring that to AE, you have them sitting there waiting for you.
A couple other workflow ideas: rather than import the .ai file and convert to shape layer. Make a solid in ae, copy the ai vector and paste into the ae solid. That requires a bit of planning on your part.
-
Be sure to use the SSD as your Media Cache Database. Preferences > Media > Media Cache Database > Browse…
Your render times will be noticeably faster. If your spinning disks are 7200rpm, you won’t bottleneck those.
-
How many frames are in a second?
-
It’s literally impossible to play at 23.976 file in 29.97 timebase. That’s like saying “if I go 50mph, I am really going 55mph.” It’s impossible because of what things mean and how they are. In some contexts, we manipulate different timebases and playback to create overcrank or undercrank. Mostly, though, the computer assumes editors want persistence of vision, so it will render the clip to whatever the sequence settings are. There are literally more frames of picture at 30fps than there are at 24fps. So the computer has to play that bit faster. When footage is interpreted, then the audio associated with that particular clip will also be interpreted. It will playback either faster or slower. It’s probably unwise to deliver a wedding video of the bride and groom sounding like chipmunks.
Premiere helps us out. In the sequence area, just below the timecode marks, you will see a red, yellow, or green bar.
Red means “this footage needs to be rendered.”
Yellow means “this footage does not need to be rendered but it is processor intensive.”
Green means “this footage is rendered and export will be quick.”We always want green render bars. If they are not green, then render the sequence, and you will create NEW MEDIA of the correct timecode.
Since you will be making new media anyway (the render files of the time code), it is BEST PRACTICE to transcode first.
If you have any sync sound (like exchanging vows or first dance, etc) then use the A camera footage as your sequence. That means, all the footage that is not in that exact format needs to be transcoded to that exact format of the A camera..
This is a relatively painless process, but it can be shocking because your amount of media is about to double (but again, it will double anyway because you will make renders of it all anyway, and if you do any color correction you will make renders of the renders and so on).
TRANSCODE FIRST USING ADOBE MEDIA ENCODER. If you’ve already started cutting. Stop. Transcode.
Here’s how:
1. Memorize, write down, or screen cap the target codec from Premiere’s Project Pane.
2. Open Adobe Media Encoder. In the bottom right is a window called WATCH FOLDERs. If you can’t see it, then turn it on in the window menu.
3. Hit the + button and navigate to a place on your media drive (probably in the same folder as your project)
4. The drop down arrows allow you to chose the EXACT FORMAT from step 1.
5. From the finder/explorer, drag 1 and only 1 (for now) of your original media into the newly created watch folder. (Open Adobe Media Encoder, if it isn’t already.
6. In Premiere Pro, use the Media Browser to import the newly transcoded files into Premiere and compare their settings. If you did everything correctly, then the settings should all match. Right click a piece of the targeted media and choose “New Sequence from Clip.” It should be either green or yellow. We can live with yellow. Then add a file from the newly transcoded footage into the sequence. It should also be yellow. If it is red, something is mismatched, and you’ll need to transcode again. When you have the settings correct, drag all that footage to the watch folder.
7. Grab some lunch while the footage transcodes.
8. It all lands in the output folder, so now Media Browser that folder into Premiere.Believe it or not this is an overview. Haha! So if you need more detail, just ask, and I will post pictures.
-
[Daniel Stone] ” ask a camera operator to also take stills”
What kind of stills? If it’s “behind the scenes” type of stuff that’s gong on facebook or whatever, just snapping them, then that’s different entirely from a photographer’s deliverable of Image Frames on the Mamiya Leaf.
-
DOH!
Groovy controller.
-
[Tero Ahlfors] “AE is kludgy and slow when it comes to grading”
I second this!
Export the final video out of premiere, and import it into AE. Use Adjustment layers timed exactly to the clips you want to effect.
But really it’s worth learning speed grade. It’s very easy and extremely powerful.
-
That’s cool too, but that’s UI on an iPad. I would like something industrial design-y. Camera guys get all their nifty form factors. Where’s mine?
-
I’m glad it wasn’t boring. And one more bit: if anyone ever wants to fool an AI listening device you can always tap the table when you speak, because the computer cannot know what sound is human speech and what is Walla-Walla. Humans can do that easy, like in crowded rooms. We can simply aim our consciousness at another human and understand what he or she is saying.
In reality, what we find is that human learning is so quick and seamless that we can adapt our speech to a computer’s input requirements much easier, like navigating a call center. It can be a funny thing to imagine a room full of people talking to their iPads. 🙂