Descript Q&A: AI Editing, Multicam, Layout Packs (Full Transcript)

[00:00:00] Speaker 1: you You the Hello, everyone. Welcome. It felt like the longest countdown in the world. We might have to shrink that down a little bit more. 30 seconds feels right. 30 seconds feels right if we have the sort of pre-roll welcome. But welcome, everyone, to the Descript Q&A. I'm Trevor. I'm a member of the support team. I'm Sal, and also a member of the support team. And we also have in the chat Alex, a member of the support team, who's going to be helping out with some links and answering some questions in the chat. I think we can get the ball rolling here. Let's get started. This is the Descript Q&A. We are here every Tuesday at 11 a.m. with members of the support team. Usually it's us two and Alex in the chat, but we sometimes have some others that join us. This is for anybody and everybody who is a Descript user, whether you're just getting started on your Descript journey, you are a seasoned user and you're just getting back in the swing of things, or if you've never heard of Descript and stumbled upon this by accident, this is for you. And the goals are to answer your questions live. So if you see that YouTube chat function where we have a few messages already in there, please send your message over there and then we will be able to highlight them like this and answer them live. We'll share a screen, sorry, in a few moments and go over some Descript tips and tricks as well as answering and demoing some of those questions. Some other resources that we have are our Help Center, which has walkthroughs, tutorials, troubleshooting guides, a bunch of feature overviews. Whatever question you have, whatever feature you're looking to get more information on, there is at least one page on there. If you're watching this on YouTube, which I think is the only way to watch this, we have other video tutorials and reviews and other past webinars there as well. This webinar, as well as every other webinar, is available on YouTube to re-watch as many times as you want in our live tab. We also have a growing community of Descript users on Discord, Reddit, and we have an official Facebook group now. Please join and submit your feedback. Say hi to everyone. We got a lot of great users there that are happy to collaborate and give tips and tricks. And a lot of us from our, a lot of us from the team also hang out there. And we also have some other live events. As I've mentioned, this is here every Tuesday at 11 a.m. We have one on Wednesday at 1 p.m. currently in our Discord, but we might be swapping over to our YouTube channel as well. And then we have lots of official sort of longer form webinars for a variety of topics. And we also collaborate with some other teams as well. If you have a specific question that we can't quite get to, you can always reach out directly. If you're in the app, just press that question mark at the top right and press contact support to get in touch directly. And with that, we are taking questions. Please, again, use that chat box there. There are no wrong questions. We will take a beginner question, an advanced question, pretty much anything. And we have, we've taken a variety of things. Drop them in there. We'll give folks a few moments to ask a question there, but if not, we can just jump right into a project and start demoing some stuff.

[00:05:51] Speaker 2: Yeah, I'd be happy to get my screen shared here. Huzzah. All right. So, this is the home tab of your Drive view. This is usually where you start when you sign in. And if you're looking for projects you've recently opened, we have this projects tab here. But basically, we have them split up into two sections. We have a private workspace and a drive workspace where all of your projects will live. I'm going to open up this one that we were playing with last week. It's a fun little podcast that we had about what if the ocean were purple. And I believe Trevor just popped it into the random topic

[00:06:30] Speaker 1: generator. We got some good stuff out of it. It was like, I Googled random topic generator, pressed the first result, and then kept generating one until there was something we could talk about. It didn't take many times. And this was one of the first ones. We got 12 minutes out of it. So, I think it was a success. About as much as anyone could talk about a purple ocean.

[00:06:52] Speaker 2: Yes, exactly. So, this is a rooms recording that we recorded in Descript's collaborative room recorder. So, each of us have a separate feed for our recording, which is really neat. Gives us a lot of flexibility. But I like to just start by looking at the script and see what can be cleaned up as far as our recording goes. And if you haven't messed around with the script editor yet, it's pretty straightforward. If there's anything you want to remove or edit out, it's just like highlighting in a text document and deleting. So, if I want to get rid of this section right here, we don't have an official name. So, let's just get rid of that. I'm going to highlight it and delete. Have for this podcast. Simple as that. So, that's a really fun way to edit without having to scrub through your video. You can basically just read through the script and remove stuff you don't like. And with that comes some really handy tools because we are script based. We have some tools here that can help us edit our video without having to lift a finger. Well, maybe a finger or two. So, to start, we're going to go to the AI tools panel here. And we have a variety of features that you can use. Under sound good, we have some options here. We have edit for clarity, studio sound, remove filler words, remove retakes, shorten word gaps, and add chapters. So, these are pretty straightforward as far as naming goes. And I like to start with edit for clarity. It sort of combines a number of features into one. What it does-

[00:08:24] Speaker 1: Before we jump into that, it looks like we do have a question.

[00:08:27] Speaker 2: Awesome.

[00:08:27] Speaker 1: From Love Rains Society. If we want to take this one and then we can probably add this to our workflow here. Hi, is there a way to use a particular part of a GIF or video that you can forward to that part and then add to the scene instead of just accepting what they give you? Gotcha. So, it sounds like instead of just adding from the beginning and playing all the way through, you only want to get part of that and then use that part. So, you only want the second half or something. There are a few ways to do this. We can demo all of them. We should have time

[00:09:01] Speaker 2: if you want to find something. Let me just add a scene marker here to the beginning. That way, it's not going to be too crazy. But let's find a short video here, some B-roll that I could add into this first scene that I created. Let's do something pretty long. So, this 20-second clip, I'm going to add to my first scene. And we'll let that import for just a moment. So, there's no way to sort of predetermine what section you're going to import right away. But we do have some ways to quickly edit this down. So, I have this video into my scene. If I go to properties, I can change the start point of this video. So, instead of it starting right at zero seconds, I can choose a start point from here. If I shorten this a little bit, I could even trim it, I believe. Okay. So, maybe this is just a start point. So, if you have a longer...

[00:10:00] Speaker 1: You have to trim it first before either the scene or the clip before the start at will really

[00:10:08] Speaker 2: take effect. Okay. Awesome. So, let's open up our timeline. I'm going to expand it a little bit. And now we have this full clip in view. By trimming it, we're just going to grab the end of that clip and then drag it in. So, now I only need eight seconds of it. So, now I have more flexibility as far as the start point here. And I wanted to start halfway through this clip.

[00:10:33] Speaker 1: All right. For this podcast, but we have three members of the support team.

[00:10:39] Speaker 2: Another way that you can choose the start point of this clip, if you're happy with the placement and the length of it, you can switch to our slip tool. You could also press Y on your keyboard. Basically, this is going to just change the start point as well. It's a little bit hard to tell because this is just a tree, but you'll be able to adjust that as well. I do want to call out,

[00:11:00] Speaker 1: there is a way from your project files to add from a specific point, but you can't do it directly from the media library. You have to first add it to your project. So, if we go to that stock media and select this one, if you move on this, this is sort of the preview. There's an end point here. And if you press the arrow next to add new, add as new layer. Next. So, that'll add as the whole layer. But if you go back to that preview and then the down arrow, sort of the dropdown next to Underlord, you can add from the end point.

[00:11:39] Speaker 2: Oh, wow. See, I didn't even know this.

[00:11:43] Speaker 1: Yeah. It's not used very often because often when you're adding from the stock media, you just add it to your project immediately. And typically, you'll be using scenes. Scenes are the best way to control your visuals. You typically don't want to have too many layers that only last part of your scene. It just, the way Descript works, using scenes is really going to save you time and also make it so much more seamless between these. But the more you learn.

[00:12:11] Speaker 2: That is great. Yeah, the more you learn. That's really cool. So, for example, if I want to shorten this scene a lot and you drag in this media. Let me experiment with this fun little thing I learned. It'll only span. Oop, wrong scene. But it should only span the length of the scene if I was under the correct one. We were so excited to try the end points that we didn't choose the right scene.

[00:12:38] Speaker 1: I know. I got excited. But yeah, the end point, this is definitely something that is in a more traditional editor. People use the end points and out points. And we have that. But the thing that separates Descript is using this scene boundary so that you can really just contain everything and then control from there. Like we are showing the timeline right now, but you don't even need to see that. You can just collapse the timeline for your entire editing experience.

[00:13:07] Speaker 2: Yeah. Let's get rid of that. I'm going to add a GIF right to this section here. And this scene, and it only spanned the scene here. So GIF is a little bit different since it's a looping element.

[00:13:23] Speaker 3: Immediately call the police. That'd be my first instinct. No, thanks. No, thank you.

[00:13:30] Speaker 1: Actually kind of fits. Purple Ocean? No, thanks. I don't know. I don't know what's

[00:13:35] Speaker 2: going on there, but I want no part of it. Very fun. So yeah, for GIFs especially, I think using scene boundaries is the most accurate way to contain how long you want it to be. Especially because they're like three seconds most of the time.

[00:13:52] Speaker 1: Yeah. We do have a follow up from them. It's sort of unrelated, but we'll tackle this one next because it's the next in line. If there is a way to upload a video and not transcribe it, then if you want certain parts transcribed to just do that part once that part is shown. It's a good question. Great question. You want to tackle this one?

[00:14:14] Speaker 2: Sure. With transcription, I think previously we had a system where transcription time really mattered. Now that we've sort of moved on to this media minute system in Descript, so anything you upload to your project files, as far as video, audio, images go, and if you record into Descript, those will count against your media minutes. So transcription time isn't really counted anymore. So even with the prior system, there was no way to transcribe just a portion of it. So unfortunately, to answer your question, there's no way to select a portion of your video to transcribe. But with that said, you can transcribe all you want and it won't count towards another pool of transcription minutes. It'll just be media minutes. So yeah, and with that, it'll be easier to edit since you'll have a full view of your video's transcription or your audio's transcription in context. But yeah, unfortunately, there's no way to transcribe just a portion of it.

[00:15:17] Speaker 1: I guess the best way to do that is if you don't, you can remove it from the script. So any video file that's on the script, that's going to be transcribed. So for a sequence, anything in a sequence is going to have to be in the script because that's just how you want to edit sequences. But let's say that camera, we were using some of those stock media. Let's say that they did have spoken word audio. If you add them to the layer, they won't be transcribed as just an additional layer. So if you want to just do parts of it that are transcribed, you can have the video clip partially on a layer for part of it for the first minute. And then the second minute, you can have it on the script track. And if you just want it to be shown for part of it, as in the final project, the final export, if you're using captions, the way that captions work is you add a caption layer separately. And I mean, any spoken word, if you use something like YouTube or other video hosting providers, they just do auto captions anyway. And sometimes they're just auto-generated. So our caption layers are the ones that actually will take from the script and then automate captions and animate captions from there. So you could just not have a caption layer for a certain part. Like right now, we don't have any caption layers. If we added a caption layer, it'll go throughout the entire project. But you can choose this scene or all scenes. And I hope this is answering if there's anything more specific about this. If there's a specific use case, I'd love to hear a little bit more about it because we could definitely cover that as well.

[00:17:03] Speaker 2: Yeah, that's a great question. And to sort of visualize what a caption layer does. So I've added a caption layer just to the scene. And since I do have a transcript or something transcribed on the script track, it's just going to display what's on the script here. So welcome to we don't

[00:17:18] Speaker 1: have an official name for this podcast. If I want to just trim this and maybe

[00:17:24] Speaker 2: caption out the welcome part, I can just really fine tune the caption layer to really.

[00:17:30] Speaker 1: Yeah, there we go. All right. Welcome to we don't have an official. And then the captions stop. So

[00:17:36] Speaker 2: the captions pretty much take what's on your script and displays them. So you could cut them up. You could remove them in certain sections or just have it play through the entire video if you'd like. So yeah, that's a great question. Yeah. No, thanks for your questions so far. Again, if you have a

[00:17:56] Speaker 1: question, there are no wrong questions. We'll be happy to answer anything. Just send it in the chat. But you were talking about edit for clarity.

[00:18:08] Speaker 2: I was. I was.

[00:18:11] Speaker 1: You were. But also we just got a new question.

[00:18:14] Speaker 2: Perfect. I love it.

[00:18:16] Speaker 1: So we will wait on clarifying the edit. Do you recommend when adding AI B-roll to highlight all parts in the whole script to add it, then use AI tool or go one by one, then add my own personal B-roll for storytelling as afterwards?

[00:18:34] Speaker 2: That's a good question. I guess it depends on your workflow.

[00:18:39] Speaker 1: And I guess I would. I would probably go one by one just because if if you are either well, two cases, if you're generating media, you definitely want to go one by one because generating media will use the most amount of credits, I think out of any AI tool right now, especially depending on your model. Some of our models will use a lot more credit. Some of them will use a little bit less. But for that case, you want to go one by one because if it generates a bunch of media and you don't like half of it, that's a lot of credits that you just use to to generate those images or videos. If you're just asking Underlord to add B-roll, even then I would probably just go little by little and see if it can find it. But you can use the stock media to find your own B-roll from there. The cool thing about Underlord using our AI agent Underlord is you can ask Underlord to check in your project files for existing B-roll and stock media and say, add B-roll to my project and it will do the best it can to add B-roll images, videos, other GIFs and things as much as possible that relates to your script. It might not be perfect, so there will be some massaging to do, but it can do that. It can get you a really, really great first draft. But it will use AI credits, so that's sort of the

[00:20:09] Speaker 2: two-sided coin there. That is, yeah, that is probably what I would say, too. I could use Underlord to just have it generate all the media in my entire script, but like Trevor said, it can either be amazing or some of it you might hate. And that's, let's say, 20 images that I created that you might not like. So going to AI tools, AI tools, and then heading to generate media will give you the most control over what you're creating. And you can really fine-tune the model, the parameters of your video or image. So instead of spitting out four variations, I could keep it to one to save on credits. I could choose my AI model and the style here. If I'm going for studio lighting, for example, and then changing it to my image generation model. And then I can just prompt it individually. So you could start with Underlord and be like, hey, pull in some relevant stock media, which doesn't, which that edit in itself doesn't cost AI credits, but the Underlord working to edit does. So stock media, it could just pull it in and then you can just replace what you don't like. That could be a way to sort of kickstart your edit and then replace anything with individual generations, which is going to give you the most control.

[00:21:38] Speaker 1: Yeah. Tons of ways to do that. Awesome. Thank you for your questions so far. All right. Do we want to go over what edit for clarity does for the third time? Don't feel bad about stopping what we're doing. We've done this a million times. Yeah. Please interrupt us. We're just demoing some of the things we like.

[00:21:59] Speaker 2: Yeah. We honestly love all the questions, so please keep them coming. Edit for clarity. Yes. So this is a great tool for just cleaning up your script overall, and you can control how intense this feature is. Some might start with low, moderate. It's usually a good in-between of heavy-handed edit and keeping it natural. So I'll just keep it a moderate and submit. Basically, it's just going to go through my entire script and get rid of any junk that it deems unnecessary from my project. So already we have 90 edits here that are suggested. And that looks good to me. A lot of it is just filler words. I know personally, I say uh in Omelot, so it's getting rid of that, any repeated words, and any tangents. So I feel comfortable with this. So I'm just going to click done right here at the bottom, and you can see the edits are going to reflect on my script. If any of these you decide that I want to keep later on, you can just hover over that edit and just click this restore button. I think that's a lie. With that strikethrough, it's just going to skip over it, which is nice to see. I think that's a lie. So yeah, let's keep that out. But really quick way to edit down your video, already went from 12 minutes and 20 seconds to about 10 minutes with all of those edits. And I think we were looking at some other features here under sound good, and studio sound is one that I like to use a lot, which is really helpful if you have a less than ideal recording setup. In our case, we just used our laptop microphones. It looks like I do have a mic here. I think you had the best

[00:23:44] Speaker 1: microphone, and I used, me and Gabe used our laptop microphones. Mine was not plugged in. Just to demo what studios, oh yours was not. Okay, we all used macbook microphones just to show, but watch this, it's going to sound like we used our microphones. Exactly. So you heard a little snippet of what

[00:24:04] Speaker 2: we sound like. Rainbow, I guess, indigo, or violet. Yeah, it's a mix of. It's fine, you know, there's a little bit of a background hum, our computer's whirring away in the background. So let's just slap on studio sound here. And it should apply it to all of the layers if I do it under AI

[00:24:22] Speaker 1: tools. So let's hear it now. Well, then rainbow, I guess, indigo, or violet. Yeah, it's a mix of

[00:24:28] Speaker 3: colors. So it's a mix of colors. Yeah, anything that's like a mix of colors has a good chance.

[00:24:33] Speaker 2: So already, sounds like, sounds like, what's that one guy, Howard Stern. Howard Stern. You know, so yeah, it just brings out the vocals, makes it a little bit more present, and gets rid of any background noise to clean up your recording. Really, really cool feature that I like to use. And yeah, that's as easy as that. If you want to adjust the parameters of that, I'm going to click on one of our layers here, and go to properties. And you can adjust the intensity of it after the fact. So right now, it's at 100%. Definitely a lot. It's a lot. And this is a sequence I keep forgetting, where did my computer freeze? Looks like it might have frozen a little bit.

[00:25:23] Speaker 1: Do you need to unshare and share? Nope, we're good. There we go. Nothing happened. Nothing happened. We're all good. Part of the fun of working with an editor and also sharing your

[00:25:36] Speaker 2: screen and also streaming. Streaming, just my internet company is just screaming at me right now. So 80% is usually a sweet spot, and in most cases, keeps things pretty natural, and also enhancing your recording. I think I might just refresh just for safety. All right, trucking along. How'd you just refresh there? I just did it via my browser. So I'm on the web app right now on Chrome. So if you click on the refresh button on your browser, you can do that. If you're on the desktop app, I believe it's Command-Shift-R to just refresh. Aren't you worried about losing your save? Good question. So everything is consistently synced to the cloud when you're editing in Descript. And you could verify that by looking at this little synced icon here. So if there's a checkmark there, any edit that you do is saved to the cloud. Now, what is the cloud? The cloud. We don't have to talk about the cloud. The cloud is just, it's a feeling. It's where it rains.

[00:26:50] Speaker 1: Exactly. It's a state of mind. That's what the cloud is. It's a state of mind when your head is in the cloud. So where are your project edits? Google. I just say Google all the time. Google. Awesome. So yeah, we have Studio Sound. We tried out Edit for Clarity. What else can we try out here to basically edit this project without having to do too much manual finessing here, which we obviously can do. But I notice right now it looks like just us three talking in the same thing. But on my end, because we have three people, we have those black bars on either end of me. Is there any way to make this look nicer?

[00:27:32] Speaker 2: Let's make it look nicer. And let me make my screen share look nicer too. There we go. Yeah. So there's a couple of ways we can do this. We explored scenes just a minute ago. So I could break my project into scenes by adding a scene boundary, which all that is, is just clicking forward slash on your keyboard. So I'm adding scene boundaries here. And let's say for this section, I want it to be just Gabe. I'm going to click on Gabe and I'm going to fill the scene editor. So when it gets to the scene, it's just going to be Gabe.

[00:28:02] Speaker 3: Would immediately call the police. That'd be my first instinct. I also was thinking, you know, the Romans used to talk about like the wine red sea, because I guess they didn't have a name for the color blue, but.

[00:28:16] Speaker 2: And then at the end of that scene boundary, it goes back to the three of us. So I could even switch it out to be Trevor when this new scene starts and get rid of myself and Gabe. So I could do this. I could do this manually and it could take hours, but we have a feature called automatic multi-cam and this works. If you have a sequence with multi-tracks, which we do in this case, since we recorded in rooms and basically what it's going to do is going to, it's going to take your speaker labels here. So whenever the speaker label changes, it's going to go to their camera and to make it the, the view for that layout or that scene. So let's just go ahead and click on that under AI tools and the style we have automatic or show only active speaker. Automatic is great. It kind of does a little bit of, you know, focusing the speaker at the, at the current moment, or it could do a split view or it could do reaction shots. So it's sort of a great all arounder, but if you wanted to just keep it to show the active speaker, you could do that as well. But I'll keep it on automatic and by default, it's going to have a layout pack selected, which is going to be the multi-cam full screen. So let's go ahead and just

[00:29:28] Speaker 1: click submit. We'll let that load up. It takes a while. While this loads, we did get a new question here. When adding or changing a sentence in the script, then using AI, an AI voice that you recorded with the script, the script provided, and it doesn't sound like you, how do you fix sound or tonality? So this is a great question. It sounds like you're referring to using regenerate. We have a couple of ways to generate AI text to speech. You can just generate text to speech from the ether using magic of software to just type in. And if you go to write mode, if you see at the top there, you'll see a button that says write with a little feather pen, featherhead. Featherhead?

[00:30:19] Speaker 2: Featherhead.

[00:30:19] Speaker 1: In Quill.

[00:30:20] Speaker 2: Yes.

[00:30:21] Speaker 1: Yes. That is so you can write, yeah, Quill, you can write your own text to speech. Whatever you write will then generate to text to speech when you have an assigned speaker. The other way is if you highlight something, let's see, you know where it says it kind of feels like a video game. Let's instead highlight, it kind of feels like, or you can highlight all of it and, and we could just change this to it feels like. And if we press regenerate, we can choose it feels like, looks like for some reason regenerate video is blocked on ours, but, oh, okay. We don't have a voice for me. Okay. Let's pretend we did have a voice for me and let's pretend we did have a voice for me and we used regenerate. You can do this for audio or video now with video regenerate. If you don't like how it sounds, you can try to regenerate it again to see if the new results will, you'd like, you can add punctuation to change the tonality and change some of the timing of it. Or if you really don't like that, you can use a different model. We have actually a couple of different models to choose from for AI speakers, or you can try to recreate your AI speaker altogether. So all of those are different sort of ways to do this. I like to say that you should record your AI speaker in the best environment possible while creating an AI speaker is very, very easy. Having a good recording environment is not. So if you recorded it in a very loud room or in a room with a lot of echo, it might have some artifacts during the training process, and it'll try to remove those. And in doing so, it might not sound exactly like you. And also it's better to do this for shorter phrases rather than longer phrases, because you start to notice that it's not a person talking when you are doing an entire paragraph or

[00:32:19] Speaker 2: story. Yes, I agree with all that. And to sort of add to that point in regards to recording your AI speaker, I like to have multiple versions of my voice, and you can set up as many as you want. One thing that I've seen happen to me, at least if I did a recording, like for this example, we have this room's recording, and I just use my laptop microphone, and my AI speaker was created using this $400 microphone. It's going to have some tonal differences. So you could have, you know, an AI speaker using the same setup as the recording just to match it even better. And just having different versions of inflections and sort of, you know, attitude as far as the recording goes, you could have like an excited one, you could have a sad one. So yeah, just experimenting with different AI speakers that you make can help sort of match it as well. So another tip that you can help that to blend easier.

[00:33:26] Speaker 1: A great question. Yeah, great. Great question. Great call out on some of these. If you were to create a speaker, how would you do so? Yeah, let's do that for me. Let me, let's try to regenerate. Actually create a speaker, but we can show the steps to do it.

[00:33:43] Speaker 2: I'd be all over that, honestly. There we go. There's me. So I'm going to click on my speaker. And if you click on the three dots right next to your name, you can enable speech generation for the speaker label. So once you do that, you're going to have this pop up asking you to read this consent statement. And you would choose your microphone from here. And basically just click record and read, and it'll train your voice and create a version of it for using text-to-speech or regenerate. So it's pretty straightforward.

[00:34:19] Speaker 3: Oh, sorry. Were you saying something? No, I was going to say that choose a file. Is that for

[00:34:24] Speaker 1: if we have someone that's not here? Like Gabe, if we wanted to use Gabe, would we be able to have Gabe record this statement and then send it to us and we can choose that file and create it?

[00:34:36] Speaker 2: Yeah, that's a good call out. If you don't have someone with you physically and you want to create an AI speaker for them, they can certainly just read this and record it. As long as it's an MP3 or WAV file, you just upload that file and you can create an AI speaker for them. So that's really helpful for collaborating as well. That's awesome. So for Gabe, yeah, very, very useful.

[00:35:04] Speaker 1: Awesome. All right. So we have the automatic multicam. Do we go over everything with it, you think? Or is there more to kind of go through this? It looks like it loaded and it created a bunch of scenes. Anywhere that there's that thumbnail is the scene boundary where it created

[00:35:19] Speaker 2: one. Yeah, it looks like there's a bunch here. Let's see how many we have. We have now over 100 or just 100 scenes, which is amazing. I didn't have to do a thing, which is really cool. Let's add some captions. I think captions can help keep people engaged for these kind of things. So if you just go to the caption tab to the right here, you can start with one of our caption templates here. The one word is always a favorite of mine, so I'll just use that for now.

[00:35:56] Speaker 1: There we go. Three members of the support team where we have used a... Let's go. I mean,

[00:36:06] Speaker 3: it sounds like a mind virus to me. So once you have a caption layer added, you can customize

[00:36:15] Speaker 2: this even further. So you don't have to stick with what we have given you here. If you go to properties, you can adjust the elements of this caption layer. So I could change the highlight color to be red, the actual text color to be yellow, which is going to hurt my eyes. But we also have the settings icon here where you could further customize your caption to fit your needs, which is really, really interesting. That is yellow. There we go. Yeah, so you can really get into the look of your caption layer. As for anything, if you click on a layer on Descript and you're under properties, you can affect some parameters that are associated with that layer. So I have Gabe's layer selected here. I can do some color adjustments for Gabe. This is a slightly different color. I'm going to make it a little bit more I can do some color adjustments for Gabe. This is a slightly newer feature, which I absolutely love.

[00:37:13] Speaker 1: Yeah, having just these little presets for color adjustments is really nice. And you can adjust them even further, too, by pressing that toggle. And then you can change the saturation, the temperature, the tint, the white balance, just some other things. Well, that was certainly something wacky there. It's a 2002 edit. Exactly.

[00:37:38] Speaker 2: And if I wanted to do this for all scenes that Gabe is in, before I make these changes, I'm just going to click on all scenes and then I can go ahead and do that. Bam.

[00:37:52] Speaker 1: So now this is a really...

[00:37:54] Speaker 2: So you'll see Gabe is flagging away.

[00:37:57] Speaker 1: Gabe is artsy here, this sort of...

[00:38:01] Speaker 2: Somber, somber Gabe. This is somber Gabe. Yeah, I'm going to rename.

[00:38:09] Speaker 1: Do we want to try and add a little intro and outro layout?

[00:38:15] Speaker 2: Let's do it, yeah. I'm going to go to the start of my project here and insert a blank scene. So at the start of my project, we have an option to add a new scene at the very beginning. So I'm going to go ahead and just do that. And by doing so, this inserts a five second gap clip and nothing else. So this gives me a canvas to work with as far as having an intro. So just to visualize this, I have this gap clip and I could extend this or make it shorter. Let's say we want a 10 second introduction. I'm going to go ahead and start with that. Bam. So this is where you would insert your intro recording or maybe if you have a music file or an image or some sort of graphic, this is where you would build that. I'm going to go ahead and start with some music here. Let's go to media and flip over to the audio tab here. And I'm tired of Elevate Your Love, so let's find something different.

[00:39:28] Speaker 1: Yeah, we do this. We do this all the time, that one. How about Higher to the Sky? Higher to the Sky sounds promising. What's great about the stock media is that you can play them, you can try as many as you want. These don't actually use any of your media minutes, so you can add as much as you want to it.

[00:39:54] Speaker 2: Yeah, you can just preview these before even inserting them into your script. I'm so confident that I'm going to love Higher to the Sky. So you can see that it ended up inserting the whole song, which is expected. So I'm just going to trim this by clicking S on my keyboard. Then deleting the rest of that. And like we explored earlier, maybe I wanted to start in the middle of the song, so we can go to properties and change the start point to something else. Ooh, I like that. That's very nice. I like that a lot too. Cool. So that's sounding great. Let's add a little graphic. Let's go to my media. Oh, stock media. I think this is fitting. We're going to go with the clouds here. The clouds. I think this is the video too.

[00:41:00] Speaker 3: So we'll have to add some text, some fancy text. Stunning. Fancy text.

[00:41:07] Speaker 1: Just so you know the name of, like look at the name of this file. When you're searching your stock media, you can search for these buzzwords like stunning, amazing, grand, atmospheric. You can search for these more adjectives and find something. You don't have to search for something tangible. Yeah. Yeah. Ethereal.

[00:41:30] Speaker 2: Ethereal. I wonder what you'd find if you typed in ethereal.

[00:41:34] Speaker 1: I want that. If you typed in ethereal. I want like an ethereal forest. It kind of nails it. All of these are ethereal. That's ethereal.

[00:41:46] Speaker 2: They're pretty ethereal. Yeah. I like that. Winner. We're committing. Today is a day of commitments. I'm committing with natural park. Stunning. Okay. So that looks cool. Let's play it back a little bit. See what we have. You know what? I kind of wanted to zoom in gradually for this video. That's nice. Let's click on this layer in the scene editor and do animation. Let's do a little, let's do a little zoom in pan action. I wanted to start at the beginning and then just kind of end up in the middle.

[00:42:33] Speaker 1: Whoa. We are being taken on a ride here.

[00:42:40] Speaker 2: This is a journey. Support Q and A journey. Yeah, that's really, it's really fun to just mess with the layers on the scene editor.

[00:42:50] Speaker 1: Now when it zoom ins, do you want to add some text for the name of the podcast?

[00:42:55] Speaker 2: Let's do it. I'm going to go to elements and then let's add a title element. Let's have it come in like a couple of seconds in. There we go.

[00:43:13] Speaker 1: Purple Ocean. Purple Ocean. The Purple Ocean Podcast. Okay. I love it.

[00:43:24] Speaker 2: Let's animate this show.

[00:43:25] Speaker 1: As we show clouds.

[00:43:28] Speaker 2: Let's do a little reveal. Let's animate by word. Let's do character. Let's get funky with it. Let's do character.

[00:43:36] Speaker 1: There's all these different types of animations here. The easy.

[00:43:41] Speaker 2: Oh, cool.

[00:43:44] Speaker 1: Let's play it back. So now we have an intro to a podcast. We're going to be a top five podcast about purple oceans. Number one. Number one on the charts. Now we make tons of episodes specifically about purple oceans, but it doesn't always happen to me. How would we save this if we wanted to use it on other projects?

[00:44:14] Speaker 2: Good question. Let's commit to this. This is a scene by itself that we made this intro in. So we have all these layers contained into the scene. Now I can save the scene as a layout, which I can reuse and reapply to future projects as much as I'd like. So if I right click on the scene thumbnail here, you could also do it on the script. I can save the scene as a new layout. So once you do that, you'll have this little pop up to the right here and you can name your layout. So I'll call this my intro layout with clouds, just so I know. Intro layout with clouds.

[00:44:53] Speaker 3: Intro layout with clouds. It can't go wrong.

[00:44:59] Speaker 2: Okay, great. I did it. So I cleaned out our layout pack. So I'm going to create a new layout pack. Oh, nice. This is my first one and I'll just call it my layouts just to keep things simple. And I can continue adding new layouts that I create or save into this layout pack called my layouts. So let's create.

[00:45:20] Speaker 1: What is the difference between a layout pack and a layout?

[00:45:24] Speaker 2: Layout pack is a collection of layouts. You can have your intro, outro. You can have your certain caption styles you want saved in there. So it's basically just a collection of layouts.

[00:45:39] Speaker 1: So I shouldn't have a different layout pack for each individual layout. You'd say I should have a layout pack with a bunch of layouts for a specific project. Yep. Okay, cool. Good to know. That'll help a bunch.

[00:45:56] Speaker 2: If you're working with a lot of different kinds of projects or like, you know, different clients, for example, you could have different layout packs for that. So great question. Okay, so that's saved. You can sort of see now when I click on the scene, it's going to show you the name of the layout that's applied. And now if I just delete this entirely. So let's pretend we're in a brand new project and I add a new scene at the beginning. I can just apply my intro layout. And everything that we created is going to pop up into that scene. Purple Ocean Podcast trademark. Awesome.

[00:46:40] Speaker 1: Awesome stuff so far. Feel free to ask any more questions. We still have a little bit of time, but we can always close things up here a little early because we'll be back next week. We're here every week at the same time. But don't be shy to ask any questions. I know we're always available on support if you want to ask us questions, but we're here because this is a great way to get your questions answered live and also see a visual of what it is. And also just having a way to communicate with us directly and see our faces and see a project and see us working in Descript.

[00:47:19] Speaker 2: Yeah, it's a ton of fun for us. So please continue to drop them. We're happy to answer. I'm going to make an outro while we're just hanging out here. Sounds good.

[00:47:32] Speaker 1: You're very welcome. Thank you for joining. Love rains society. I almost said love ruins society. I'm like, that's not it. The opposite of what it does. Opposite of that. But thanks for joining. Thank you for your questions. Feel free to join us next week or in any other future week.

[00:47:53] Speaker 2: Absolutely. So same thing. Yep. Easy peasy. We're just going to go to the end of our project and add a new scene at the end. And you can see that there's a blank scene there now. So I can definitely build one. But you know what? I'm just going to apply the same intro layout that we had and just edit the text.

[00:48:14] Speaker 1: Thanks for tuning in.

[00:48:17] Speaker 2: And then I can even save this as my outro layout. So I could reuse that. All right, kids. Beautiful. I think I'm ready to send this to Spotify. What do you think? I think it's ready. It's ready.

[00:48:39] Speaker 1: So how do we do that?

[00:48:41] Speaker 2: Let's up top. We have this export button. And we'll have a couple of options once you click that. Since this is a video project, I'm going to keep it under video. But you could export just the audio if you want. If you work at other editors, like we have a couple of supported options here like Pro Tools, Audition, Final Cut, Premiere, Resolve. This can export a timeline XML file, which you could import into a different editor and continue working. There are some limitations with that. So I would look at our article if you have specific needs. You can export the transcript or a subtitle file. I'm going to keep it back on video. And destination is where you choose what kind of export it's going to be as far as the video goes. It's going to default to a web link, which is recommended for the most part. It's going to create a private URL that you can watch and share with your friends. And you can also download the video from there. Or you can switch to local export, which will save it directly to your computer, which gives you a bit more control as far as the settings go. Resolution, quality, some audio settings here. And then we have a few integrations with other platforms like YouTube, Google Drive, Headliner, HubSpot, Podbean, and more. So a ton of options for you. But for the most part, local export will be your bet or web link. And then there you are. It's going to save your computer and you can share it with the world. Out it goes. Almost 1% at a time. Yeah, let's give my poor computer a break and cancel that.

[00:50:31] Speaker 1: Any other questions we have? We don't have to sit here for that. But any other questions while we're here? We're happy to stick around for a few more minutes while we wait for any questions to come in. But that's sort of our demo here. I feel like we did a good job showing just some general things to get you started, especially if you're working on a multi-track recording. There's a lot of other types of projects you can use when working in Descript, whether it's just a single-person recording, whether you don't even have a person. We have something called avatars to use. If you don't want to show your face, you can hide it under an animated avatar. Or you could just use entirely B-roll or screen recordings and other things.

[00:51:16] Speaker 2: Yeah, lots of ways to start. So we can explore that in different sessions for sure. Yeah.

[00:51:24] Speaker 1: Well, I think that might wrap it up for this week. Feel free to ask any questions and we can try and get to them in the chat. Oh, we do have one more question. We can definitely take this. Let's see. I'm having some audio quality issues. Do you have any advice to reduce the demand on the CPU while recording? That's a good question. Definitely a little bit more technical. And this may be a good situation to reach out to our team directly. Because if you reach out to support, we can actually take a look at some of the logs of your computer information. And also, if you're experiencing any error messages, we can actually see what's happening. We have some internal systems that we can see. But if this is just the demand on the CPU while recording, the first thing is make sure you are using a device that meets our system requirements. We do have a Help Center article that has a list of requirements. And if you're not meeting that, that's definitely an issue. Alex just sent along an article about recording performance issues, low frame rate or sync problems that can happen, and some handy dandy sort of recommendations if you're experiencing that and troubleshooting steps. As far as the quality issue, this might be something we might have to look at the project to see what the audio quality is. If it is a sync issue, that could be one solution. If it is some audio artifacts, that could be something to do with the microphone and some compatibility issues. We do have a list of unsupported devices in software that run that can also cause some issues and also a higher demand on the CPU than required. Anything I mentioned or anything I didn't mention that is worth mentioning on this kind of case, Sal?

[00:53:19] Speaker 2: That covers most of it. As sort of a general rule of thumb, I think closing out any unused tabs can help speed things up in any programs that are running in the background, like Zoom, Teams, or anything that's using a lot of CPU. Just making sure everything's closed to give your computer the best chance to handle that recording and processing it to the best quality.

[00:53:41] Speaker 1: I do know that as far as recording, because the recordings are really high quality, if you are recording in 4K and your computer might not meet the demands of recording in 4K, either on the screen, your camera, etc. If you're recording video game gameplay or something, it can definitely have unneeded demand. I've tried that just to see what we could use, basically how much we could demand from Descript and the computer. And there's some mixed results there. So that's another thing to check is just what are you recording? What is the settings of your recording? And if you lower the quality, does it help or change anything? Good tips. All right. I think that'll wrap up this session. We will be back here next week, Tuesday, 11am, back on YouTube and ready to answer your questions and talk about Descript.

[00:54:44] Speaker 2: Yeah, we look forward to it. It's always a fun time. And in the meantime, reach out to support if you run into anything else. And have a great rest of the Tuesday. The best Tuesday.

[00:54:56] Speaker 1: Bye. Bye. Bye.