📱

Get Our Mobile App

Take your business learning on the go!

Download on the App StoreGet it on Google Play

09/18 - HeyGen Product Updates: Avatar IV Digital Twin & video editing with Nano Banana

HeyGen44:02

Transcription

Good morning, good afternoon, good evening. We are live. I'm just going to check that we're all set. It looks like we are.

Well, it has been an exciting week at Hen. We've had uh triple announcements this week. uh rebrand, we had video agents beta launch um and we announced uh an exciting acquisition uh joining the team from Alyssa. So you can check out all those announcements um on LinkedIn blog posts in the community.

Um, and today we're going to dive into product updates with Adam and April. Um, and then if you check out our other events on September 30th, um, our CEO is going to have, um, a live Q&A session about all of the announcements from this week. Um, diving into kind of the future of Hey Genen, um, and where we're going next. So, that's on September 30th. Uh, you can find that on our events page. We'd love to see you there, too.

And today, we're going to dive into, yeah, exciting product updates. what's been happening on the product team, what you can go into Hey Junen today and start testing out. So without further ado, I'll hand it over to Adam and April to take us through those updates. Thank you both for joining us today.

>> Thanks Audrey. Um, yeah, super excited to tell you about some awesome product updates. Uh, we first have you know this big new video agent launch which uh April will uh go through and then also have a few other just exciting updates around nano banana for look editing a major upgrade to our photo avatar unlimited mode and a revamped avatar navigation. Um, so we'll hand it over to April to go over some video agent.

>> Awesome. Um, hey everyone I'm April and I've been working with the team to build agent. It's a major initiative we've been working on and it brings to life an all new uh kind of a gentic video creation workflow and it'll enable you to bring videos to life with a single prompt. Um, so for those of you interested I I wanted to give a quick um kind of walkthrough of what you can expect uh that's new in the platform. Um, I guess high level there are entry points for agent now all over hey genen. You'll be able to find it through the home banner navigation bar and through your projects. And then once you're in agent, you'll be able to create your first video with a single prompt. And then once you have your video with agent, you'll be able to edit that video in studio. So if you go to the next slide.

Yep. So you can see here some of the entry points. This is um dark mode, which actually I guess I'm giving you guys a sneak peek. I don't know if it's actually live yet. Um, but here are some of the entry points. Um, we have the hero image on the home screen. We selected these images essentially to showcase all of Hen's um various avatars. So you'll have a wide range of avatars to star in your video when you create create um your first agent video. And on the left navigation, you'll see something called agent in beta. And uh, you can also tap on the clock icon to the right of the video agent tab in the left navigation to be able to see all your past s uh sessions once you start uh start creating your first creations. Um, we can go to the next slide.

Once you tap on the video agent page, you'll be able to see a prompt box. So if you guys are familiar with like some of the other AI native tools, probably things like chat GBT, you'll be very familiar with this idea. you can just drop your video idea and then agent will create your first video for you. Um, below the prompt box you'll see real videos um as samples from our community. Once you click the video you'll be able to see also the chat um prompts that created those videos. I know in our early kind of interviews a lot of users were asking what kind of prompts perform best. So you'll be able to take a look at that um here in the gallery.

And then if you go to the next slide here, um, once you click generate, you'll have like a short interaction with our agent chat and then you'll get your first video. Um, once the video has been drafted, you can make uh edits. So there's essentially two ways you can make edits. The first one is through the chat interface on the left side here. So you can just type what you want to kind of update. And then on the right side here where you see script, avatar, voice, music, that's what we call the light editor. And in both of these places, you'll be able to make kind of global simple changes to your video, to your script, your avatar voice, and easily regenerate the video that you want. Um, one thing that we really focused on actually is to make sure that if you did want to do um have like fine grained controllability over the video, let's say you want to fix the pronunciation on your script or you want to do a scene by scene uh avatar swap, you can actually take this video to studio. Um, if you go to the bottom right there, you'll see edit in studio. And if you go to the next slide, if you click that, you'll be able to make a lot of the more kind of specific um changes. Like for example, if you wanted to change the avatar in scene 2 and scene 4, you can do that here. You can do um change the pronunciation, which we get a lot of requests for in the script panel on the left here. And then finally, generate. Um, once you do generate the video, if you're all using the Hunen platform, it'll live in the projects page just like all of your other videos. Um, and you'll be able to access your sessions from the projects page as well as all of your videos. Um, so that's it. That's the core experience. We're super excited for everyone to try it out and um, yeah, excited for all of your feedback. Thanks.

>> Amazing. We'll have time for uh, questions at the end. Um, yeah, if anyone has more questions for April on how to use it, what to use it for. Um, but yeah, obviously very exciting um for everyone at Hen as this gets us, you know, a huge step closer towards making video creation accessible to all. Um, because as I'm sure all you guys know, uh creating a video is more than just filming the raw footage. All the polish uh ends up taking a lot of the time. And the whole point of Hey Jen is really save time um creating videos. And so with this, we're, you know, really excited to hopefully save you a lot of time and still help you, you know, create polished, professional, ready to go videos that help you, uh, build your presence.

Awesome. Uh, so we have a few other cool updates um, to go over. One is we are rolling out Nano Banana to support uh, editing your looks. Uh, so on your avatar tab, you'll now see a nano banana option in our new prompt bar. And you can use this to precisely edit uh any of your photo looks if you want to, for example, change the background. Uh, here, you know, I do an example where I'm changing the hair color. Uh, you can add products. Um, what this is really good for is if you want to make specific changes to an existing shot while still maintaining, you know, the character and everything else in the shot that you don't ask for uh to be changed. Um, another really big benefit of Nano Banana, as um I'm sure many of you guys are already familiar with, is it can support multiple image references. So, if I want to add a product to the shot and I also want to dictate the outfit and I want to add my logo, uh, Nano Banana supports adding multiple image references and that'll then combine with your avatar look and create a new shot that, um, you know, incorporates those image references. And it works, you know, via natural language. So you can really just tell it, you know, in your own words what you want it to do and it'll make that edit change for you. Kind of understanding um, you know, what you're referencing between. So for example, if I upload this bottle and I'm like make my avatar hold the bottle um, it knows that I'm talking about this bottle because you know, it sees one of my images as a bottle.

Um, so now we actually have two ways to generate uh looks on on hey genen. One is nano banana and the other is our traditional flux lura away. Uh, so we'll provide a couple uh recommendations to help understand when choose uh which model. So flux is best when you want to generate entirely new scenes. Uh, so if you want to generate something where none of your existing photos match that scene at all. So, you know, if I'm uploading like photos of me like this, but I want to create uh a new look of me as a surfer, like lying down on a surfboard, where the pose changes, the outfit changes, the whole background changes, uh, Flux is really good for those because it'll generate cohesive scenes where everything, you know, is there to match that scene. Uh, so if I, you know, yeah, ask to be, you know, make me a surfer, it'll change my outfit, my background, um, you know, all the elements in the scene to really match that. Whereas Nano Banana is great for those precise edits. So if I'm if I already have a picture of me on the beach, but I want to add a surfboard, then Nano Banana would be great for like add a surfboard, you know, make him hold a surfboard or, you know, change it from day to night or change the background from, you know, a beach to a lake. Uh, so if you want to work off an existing photo, Nano Banana, you know, tends to work best. Whereas if you want to, you know, create a scene from scratch, particularly, you know, if you want it really stylized, uh, flex works great. Um, and one way to think about it is is in terms of pose changes, like if the scene is going to require a completely different pose change, uh, then often flux will work better because that requires nano banana to, you know, essentially generate that image from scratch rather than use any sort of like inpaint thing. Um, so yeah, if there's going to be massive pose changes, worth trying Flux. Um, if not, worth uh trying Nano Banana first. But again, like with all things generative AI, nothing's fully deterministic. Um, so would encourage you to experiment with both.

Um, in terms of some of the details of how they work, Flux trains on all of your photos that you provide. Generally, it's the more photos, the better. Uh, 20 plus for best results. So um, you know, when creating your avatar you can upload you know multiple photos for the same identity and ideally for flux you upload photos in different poses different angles really so it can understand you know how you look like across all your different kind of angles emotions shapes um and this gives it the flexibility to then, you know, generate you in entirely new scenes. Whereas Nano Banana, it requires a base look or a base photo to edit off of. Um, so it'll take one of your photos and then it'll edit off that photo. Um, in terms of speed, uh nanobin is faster. It's about like 50% faster per image generation. And uh, flux also requires a two-minute training. Uh, so that only needs to happen once. Once you train it once, then it's 30 seconds going forward. But if you're looking to move fast, create with lots of different characters, Nano Banana is typically the quicker way to iterate.

Um, yeah, so that is Nano Banana and now have another really exciting announcement uh with a major upgrade coming to our photo avatar unlimited mode. Uh, so this one I think the the easiest way to explain it is is to show it. Um, and just for context by photo avatar unlimited mode we mean uh when you create with a photo avatar on hen by default you'll now be using our avatar 4 and you get 10 minutes of avatar 4 a month and once you run out of those 10 minutes you can switch to our unlimited mode which is you know our technology that we offered before we launched Avatar 4 and you can use that unlimited uh so you still get unlimited video creation plus you get the 10 minutes of Avatar 4 every month. Um, and what this changes is it's a massive upgrade to that unlimited mode. Um, so I'll show you the two videos. So this before video is uh what it looks like today if you upload a photo on hey gen and then you you know type in a script and generate uh in the you know avatar unlimited mode. Um, and actually I don't know if you guys can hear. Maybe I will. I'm going to try reshing. All right. So, this is the before.

>> Welcome to our world of innovation and excellence. Discover how we can help you achieve your goals with our trusted solutions. Join us today and experience the difference. Your success starts here.

>> Welcome. Uh, so as you can see just the head moved everything else um, you know, the background the tree the waves >> to our world of innovation and excellence. Discover how we can help you achieve your goals without >> and so with our upgraded photo avatar unlimited mode. Uh, this is >> welcome to our world of innovation and excellence. Discover how we can help you achieve your goals with our trusted solutions. Join us today and experience the difference. Your success starts here. Welcome to

So, as you can see there, you have full body, full background movement. The tree, the tree is swaying. Um, you know, the hands are moving. Um, you know, even the necklace moves as his body moves. Uh, so it really has that that full body movement and realism that you've, you know, come to enjoy with Avatar 4 that you get with our video avatar. Um, and so now officially all all avatar types on Hen have full body movement, background movement. This is rolling out um, starting to roll out this week. So some of you guys might might see it. Um, we're running it, you know, as as experiment. Um, but by next week, everyone should should be getting this. And what's awesome about this too is there's no additional work required on uh your end to use it. Uh, all you do is still just upload a photo. You can type in your script and generate. And the difference is instead of getting you know this where only the face moves, now the whole body will move. Um, one call out is uh you don't need to add motion first. Um, so today you do have the option of after you upload the photo to add motion as a kind of uh preset up step before you go to video creation. With this uh that's not required. You can upload the photo immediately just generate a video and that motion will come with it. Um, and under the hood how this is working is we essentially generate that motion once using avatar 4 and then we treat that as the training footage for your uh for your video avatar. And we've actually seen um, you know, a lot of or some of our users actually do this themselves and that was some of the inspiration for this project. We saw users you know creating Avatar 4 videos and then re-uploading that, you know, to create digital twins from those videos because, you know, they want that full body background um motion. So, I think it's it's another one of those cases where we're always so inspired by, you know, uh seeing what our users create and it's very much this two-way street of um, you know, shipping something then seeing how people use it and uh, you know, shipping new features based on um, you know, coming up with these new ideas that that users really came up with.

Um, yeah. So, then finally, um, not quite as exciting as the last two, but we also have a revamped avatar navigation, uh, with the goal of really just helping you do the day-to-day things a lot faster. Uh, we know your time's very valuable. You know, in a lot of cases, time literally is money, you know, when you're billing by the hour. And so, we don't want you wasting time, you know, doing unnecessary clicks or waiting on load screens. Uh, so with this new navigation, we've really centralized everything into this one avatar page. you can do all the key actions uh without ever you know navigating out of here. So you have this single entry point right here for creating new avatars and new looks. Doesn't matter whether you're creating you know a new identity or you're adding a look to an existing identity, you can do it all right here. Uh, we also just have this centralized page for viewing and managing all your avatars and your avatar looks. Uh, so rather than having to like click into each avatar to view that avatar's looks, uh, you can view them all on this one page and then you can tap on any of these circles to filter for just that avatar's looks. Um, so just a lot less clicking, a lot less load time uh to be able to quickly navigate to find the avatar you're looking for. We're also working on avatar looks search um so can look forward to that um coming soon as well. And then finally, we have this unifi prompt bar where you can both generate entirely new looks with flux and um make precise edits to look with nano banana. So yeah, uh looking forward to your feedback um both on this are upgraded in limited mode and uh nano banana with all these features. We're you know we put them out quickly because you know we we learn faster by seeing how people use it. Um, so we you know plan to iterate quickly to improve but really very much would love uh to have you guys you know test it and let us know how can we improve it.

>> Awesome. So yeah with that said uh I guess we can open up to Q&A.

>> Awesome. Yeah. Thank you Adam. I yeah unlimited photo avatar mode very exciting. Um, I put a link in the chat um for video agent videos when you start trying out unlimited mode. If you have feedback, if you have videos you're making as examples, we'd love to see them. You can always post them in the forum um and share them with us there. We've got a community videos thread. We've got a product feedback area. Uh, we'd love to hear what you guys think and what you're creating. Um, if you make a really cool video agent video and you got a great prompt, share it with the community so we can all learn together from that. Um, would love to see what you're all you're all making. Yeah, we can jump into some Q&A. Um, I've saw some folks putting questions in the Q&A tab. Um, if you put them in the chat, if you can help us by copying them over um, posting your question in the Q&A tab, that'll make it the easiest for us to answer some of those. Um, I think we can start uh with a couple Nano Banana questions, Adam. Um, does Nano Banana work only with video or audio too? Can it make the voice deeper?

>> Uh, today Nano Banana works only with photos. Um, so you can use it to edit any uh photo look. You can also use it to edit one of your, you know, video looks that you upload though it will edit it as a photo. Um, but again with, you know, our improved with Avatar 4 plus the improved unlimited mode, even if you're working off a photo, you do get uh full body background motion. Um, in terms of audio editing, um, you know, we have different options for how to edit your voice. Um, we definitely recommend check out our resources, but if you click on a voice, like from our voice tab, brings up this voice detail where you can try out different voice engines. We offer um, 11 labs. I'd recommend trying the new 11laps V3 engine uh which we recently offered which is more expressive. It's also controllable with audio tags. Um, so one I would try different engines and the other is uh you can also update the settings um like on 11 Labs there's stability style. Typically every voice needs to um need just a little fine tuning to like make those settings sound right. Each voice those settings you know should look a little different. But once you find the settings that work well for your voice, uh typically you shouldn't have to change it again. So would would recommend trying to tweak those a bit, see if that helps, and then once you u you know get something you like, you should be good going forward.

>> Awesome. There's another question around nano banana, I think around like um how much maybe the limits of how much you can use that. Um, someone saying you know could they replace their nano banana subscription with using this within hen um today?

>> Uh, good question. So right now the limits here work the same way as the current uh look generations with flux. Uh, so you get 100 a month on the paid plan included and then um we also offer that generative add-on pack and it's one credit per uh image.

>> Cool. Awesome. April, I see a question around agent for you similarly. Um, is there a limit on how much we can use agent? If so, what happens if we go over the limit? Um, and and how that all works?

>> Yeah, that's a great question. So, for free users, you'll be able to create three videos using kind of our best models. So, this is what we call quality mode. Um, and then after that, you will be able to choose between two modes. versus quality mode and unlimited mode. Um, on the quality mode, you'll be able to uh use it up to I have to get back with the exact numbers because we just finished our payw wall, but um, it's basically there's like a time limit and a credit limit on that quality mode. And then um, once you use all of your credits or time on that quality mode, you can top up and add more credits to continue on quality mode or you'll always have access to what we call unlimited mode. And in that mode, you'll be able to create videos with your stock with our like kind of stock avatars, your stock media. And um, yeah, so that's that's kind of like the high level. I mean, I can follow up with a one pager on the pricing and kind of plans here.

>> Awesome. Sounds great. Yeah, I know it's very new, so uh just just getting started with with the launch. Um, I think I know we get a lot of questions around access and limits and I think it's just good to remind everyone that hey Genen is across all of our pricing and and subscriptions really committed to making it so you can test out all of the features even on a free plan and then there's they really enable you to continue creating. Um, but our fastest and newest models uh do have like limits on them. Um, but really it's intended so there's always a way that you can create and keep creating um with Hey Gen. And as you can see, you know, feedback is influencing Adam is working on updated uh unlimited mode for photo avatars. Um, so when you if you hit your avatar 4 limits, um, you can buy more credits or you can keep creating unlimited. So I just want to remind everyone you can still keep creating. We want to make sure you can keep creating. Uh, you might just might need to switch around what mode you're using. Good questions on those. I know a lot of people have those. Let's see. Um, this is a question around the unlimited photo avatars. Adam, um, do the motions in unlimited in the new unlimited photo avatars um mode. Are they in a loop? So will they loop in the background for those?

>> Welcome to our world of innovation and excellence. They they will um not the same way the current motion avatars do where those are based off like a five or up to 10 second loop. Uh, these these will be closer to how uh your like instant avatar or essentially your video avatars in unlimited mode uh work. So, uh, essentially if the scene and and this does vary, um, you know, depending on a few different factors, but rule of thumb is if the scene is uh like greater than 30 seconds, you might start to see looping. Um, but less than 30 seconds, you typically shouldn't see looping. One other just pro tip and and we are going to be rapidly iterating and improving on this but if you create with a photo and you don't love the motion uh because that motion is set on that photo uh if you create again with that same f photo avatar you'll get the similar motion style. So if you don't love it uh you can try recreating with a different photo or just re-uploading that photo and then creating again. And we're going to be doing um improvements over the coming weeks here. But yeah, just uh pro tip there is um if you do want to mix up the motion for the same photo, try uploading it again.

>> Cool. Cool.

>> Awesome. Thanks, Adam. Um, this is a question around agent. I see. Um, Craig says, "I uploaded a story that involved a main character. However, that character was completely different in each scene. Is it possible to maintain character consistency across scenes?"

>> Yes. Uh, we should be doing that now. I wonder if you could send me that session maybe. Um, I I'll throw my email here in chat. Um, I could take a look. If we're if it's using one of our avatars, we should be using the same avatar across every scene. Um, but I'll dig in post call to to see what's going on there.

>> Awesome. All right. Okay, we got a lot of questions. I'm just looking through uh someone asked about video agent. Um, are the prompts and results generated? Do they remain in the public gallery? So good question around with the the community gallery and agent are people if they create something uh does it go is it there or is it private for their in their account?

>> Um, great question. So if you go to the next slide after this, Adam, you'll see up top on the left, you'll see like a public button. So uh the sessions, the chat sessions are public by default, but if you click that, you'll be able to set it to private. So we'll never use any of your private sessions in our gallery and um um without essentially your permission there.

>> Awesome. Cool. Um, another question around agent around um reusable templates uh in agent. Are you able to provide a reusable template for this new video agent?

>> Yes. So, that's actually on the road map. We're working on enabling it. Um, don't have a timeline yet, but we know that a lot of our customers want like simple, easy ways to kind of swap things out. So, we're we're we're working on that.

>> Awesome. Okay, let's see. Someone asked, "Wolf asked, can we buy additional Avatar 4 minutes?" Um, yes, you can. You can purchase generative credit packs that will give you more Avatar 4 minutes. Adam, maybe you could show us um on uh Jeff asked, "How can we choose which mode it's using so we don't burn through our minutes in testing?"

>> Yeah, absolutely. I can share my screen. Um, and when you say mode, you mean Avatar 4 verse unlimited?

>> I think so. Yeah.

>> Yeah. Um, so what you can do is when you go to create a video. So if I choose this guy, uh, the mode appears here. Um, so if you toggle it to unlimited now on unlimited mode, um, and I can see I have like five minutes left of avatar 4. And then also if I like write a script, hi, this is a test script. And then I submit here um, and it's on if it's on avatar 4 mode, it will show uh, this. So this will use two seconds of avatar 4 time, tell you how much you have remaining. Um, and then we have our quality mode. So um, because this is on I've set it to our quality mode which costs the 2x credits. It'll also um mention that here of just the you're using quality mode which has the 2x. We are we are um experimenting with placement. So if you see this actually appear over here. Don't be too surprised. Um, but yeah working on making sure it's both clear to everyone and yet not um yeah I guess disruptive to the experience. And uh, yeah, then again, and if you do run out of time here, it will uh especially give you an option to switch to unlimited to continue creating the video. Uh, so you should always know uh whether you'll have enough credits to complete that video and if needed, you can switch to unlimited um which you know has unlimited credits. So you can create as many videos of any length in that mode.

>> Awesome. I see another question uh here around digital twins and maybe some tips, Adam, that you could share from Damian. Um, asks, "What specific steps would you advise to create the most realistic digital twin avatar that has the most realistic natural body movements and facial expressions with all of these new updates in mind?"

>> Yeah. Um, so a lot of the instructions I would say still hold true. Um, you know, our video avatar instructions. We have a lot of different materials here of like what is the best way to record that video avatar. Um, I I would read through those. A lot of it is just getting the right setup. Um, you know, good environment, good lighting, ensure there's good contrast between you and the background. Um, talking just naturally. I would say one one difference between the uh Avatar, you know, basically our old unlimited mode and now our new Avatar 4 digital twin is digital twin is able to adapt the movements to the script. And so what this means for the recording footage is it can support much more expressive uh gestures um and movements and it still won't look unnatural. So, one tip that we would often give when you were filming your video avatar, you know, prior to Avatar 4 digital twin was like very subtle hand motions. Um, so like generally like kind of keep them in this anchored position and do, you know, subtle hand motions like this that could kind of make sense no matter what you're saying. Um, with Avatar 4 digital twin, it's able to adapt motions, you know, based on the script. And so one change of advice is just be expressive and natural. Um, you know, this type of you know expressive hand motions that you know really emphasize what I'm saying and you know bring it life. Uh, that works great with Avatar 4 digital twin whereas kind of this more muted um and kind of back to base position style works better for our previous model and our unlimited model. Uh, so that's one way to think about it. If you do plan to use both models, I you know might even try recording both types of footage. Um, and the other thing just to call out is we are you know rapidly iterating on our upar for digital twin model. Um, and so we expect those these instructions to evolve over time. Um, potentially in our next one we might actually ask that you you uh give us different emotions that really help cover the full spectrum. Um, but for now, for this model, it's still generally better to just have one consistent style. Um, so, you know, the whole thing can be like friendly, natural, or it can be, you know, serious if you want like very different emotional styles. We'd recommend just creating those as separate looks right now.

>> Cool. Awesome. And, uh, okay, I see a couple questions here. I think back to the video agent side. Um, first question, April, is can, this is from Regina. Can the agent videos be longer than 30 seconds?

>> Uh, yes, they can. So, if you go to the prompt box, Adam, I don't know if we have a screen. Oh, I have it in the close mode, but yes, you can have it uh longer than 30 seconds. I think the longest option we have right now is 5 minutes. Um, on the free plan, there is a duration limit. I think off the top of my head it's probably one minute. Um, but if you are on the paid plan, you can choose from any of these options here. Oh, it's three minutes. Yeah.

>> Awesome. And then Ross asks, "Can I provide my own script?" And Hen generates a video that incorporates both an avatar plus images representing the content.

>> Uh, sorry, could you repeat that question?

>> Yeah, Russ is asking. I think this is the answer Ross I think is is yes this is what video agent um will do for you. Uh, if you provide if Ross provides a script uh will hun generate a video that incorporates both an avatar and images representing the content.

>> Yeah. Yeah, that that would work really well. And you can um you can just drop that script and uh you can even put like a description of the avatar you want and then it'll it'll create that first draft for you.

>> Awesome. Okay, we have a question from Wasib. What about multiple people in scenes? Can we upload multiple characters and make them come in the same scene?

>> Um, unfortunately not. Although we do get a lot of requests for this. So, I will talk to um the team to figure out if we can uh put this on the road map.

>> Awesome. Yeah, I know we get that a lot and I I think I saw another question around video podcast. Um, you can you use the beta video podcast feature today. Um, I know some people are using it that way, but we do get a lot of requests around this. So, I think um can't say anything about the road map, but I know I know we do hear a lot that uh interest from people to have multiple avatars in the same scene. Um, Camille, I see a question around what's the best audio engine for people with accents. Um, I believe, and if either of you know better than me, I I believe fish often works um better for some accents. Um, I know especially Australian accents perform the best with the fish voice engine. Um, but there are different voice engines you can switch uh to try out specifically with different accents. So, I think it depends on what dialect and what accent um you're using. You might want to test them. But definitely try the fish voice engine for accents.

>> Yeah, I would um yeah, for accent, I would definitely try fish as as Audrey mentioned, we see uh particularly for yeah, for the Australian um Indian-American accent. Um, so if you English with an Indian accent, uh, fish tends to work great. Uh, even like an American southern accent, we've seen people really like it for that. Um, would definitely recommend trying out the different engines. Uh, we have 11 laps V2, uh, B3. Um, we have Starfish and Fish. Um, so can try out those, see what works better. Um, you can also try adjusting the similarity setting on 11 Labs V2. uh that you know or the the stability setting um, you know, that can tend to make the voice sound closer to you or you know more kind of stylistic. Um, so I would try all of that. And then finally if um we also just have an accent selector. Uh, so if the concern is like you really just want this accent to be accurate um, you can choose the accent selector in in Hey Jen. Um, and this might come at the expense of some similarity to your voice texture, but it can guarantee that the accent is like perfectly accurate. Um, so yeah, accent should be yeah available here.

>> Awesome.

>> And the voice engines you can by default we selected on auto. Um, but here you can try out the different engines. Um, 11 Labs V3 is is the new one that 11 Labs has launched. Um, and this one this one has uh uh tags that it supports. So you can do like um so if you wanted to like laugh when saying something um, you can do tags like that.

>> Yeah. Awesome. Thanks for showing that, Adam. Yeah, we had uh 11 Labs team uh do a a live session with us last week. So, you can find the recording if you're interested in seeing more about how that V3 works and some tips. And um I know we had questions last week around um what which tags are available. And 11 Labs, if you look up, they have on their website a list of all of the different uh prompts you can use in V3 um that you can use in Hen. Awesome. Okay, back to this is a question around video agent again. Um, how do we use our personal avatars in agent mode?

>> Um, you can use your avatar ID. I I I know this feature is in in beta, but one way you can do it is to put like use avatar ID and then put your ID in the prompt >> or in the light editor, which let's see if I can if you could go to the light editor view, Adam. Um, it's the next slide after this. Yeah. So on the right side once you click that avatar tab on the light editor on the right side you should be able to see a um kind of column called my avatars and your avatars will show up there.

>> Awesome. Let's see. I think we've gotten through a lot of the questions. Just looking through uh people are asking about recordings uh from today's session. This is recorded. You can always find all of our live session recordings on YouTube on the Hijen YouTube page under live or back in the community community.hen.com under events. We upload the recordings later the day of the session. So you can always find them on the events page for past events. [Music] Let's see. [Music]

Okay, so another question from Ross and April around video agent. Will it pull in um is it possible will video agent also include like stock videos or um as well as stock images that go with the script?

>> Yeah. So it does include our like stock media library. Um, so on the free mode, sorry, on for your first three videos, it'll just choose from generative or stock. And then once you switch to paid, if you stay on quality, you'll have access to both stock and generative. If you switch to unlimited mode, you'll always have access to the stock media. So I guess in every scenario, you'll always have the stock stock media.

>> Awesome. Ken, I see your question around um in video agent you added jewelry photos um but the photos are coming out too big um in the video. Is there a certain size you need? April, is there anything around like yeah adding media um and video agent that works best?

>> Um, I don't know about the specs. So, I'd have to look at the session. Um, yeah, you feel free to send me that session as well and I'll I'll take a look.

>> Awesome. Cool, Ken. Yeah, if you share that with us. Uh, and yeah, this is the beta public beta launch. So, this kind of feedback is great. Um, if you share your video ID with us, um, and you're Yeah, we can we can take a look and it's always very helpful to hear hear this feedback. [Music] You can always connect with our team in the forum. If you leave a post there, we look at it um every day. We can try to get you an answer uh soon. Of course, our support team is the quickest. If you ever need technical support, please do reach out to support.com or use the chat in hijen um for technical support. But in the forum, uh, you can send us feedback, share your video examples. We love to see that there.

Awesome. Well, I think we've gone through a ton of questions. We probably can't get to every single one today. So, please do share share your comments with us afterwards. Um, you can reach out to me uh in the forum, too. You can find me and send me a message uh if you didn't see April's email and and we talked about following up. Um, thank you Adam and April for joining today. Uh, I know there's I mean it's the speed of page end. There's always exciting things coming out so quickly and I know it makes um such a big difference to hear the tips and the latest things to try out from you as they're being released. Um, if anyone joined today and is new to Hey Genen, we really dove into the updates. You can find recordings of walkthroughs. You can head over to hey genen academy uh and find intros um make it really easy for you to get started um and also always just recommend go in and get started creating and and see from there that's the best way to really learn but you can find hen academy tons of resources in the community um and you can find this recording later today so thank you everyone hope you have a great rest of your Okay.