KIWA & VoiceQ: NexGen Storytelling & Localization feat. Jordan Koziol Repia

Tucker: Hello my name is Tucker Johnson and I am your host today as we experience Nimsy live where we talk about the latest and greatest in translation localization internationalization culturalization and all that fun stuff global companies need to delight their international customers On this program we invite guests who like to have fun and have some value to add for our audience of globalization professionals I'm always eager to provide a platform to those with a good story or a good data set So let us know if there are any topics you'd like covered or guests we should reach out to for future episodes If you haven't already done so make sure that you are following and or subscribed to Nimdziy Insights We are coming to you live today on YouTube Facebook X and probably some other and LinkedIn of course where a lot of folks watch us A different time zone today I normally don't do these in the afternoon I'm here in Seattle so it's afternoon for me But we are talking to someone that will get into it very soon in a different time zone rather than the boring old US Europe time zone So um without further ado I'd say let's just get into it here We're talking to someone from Kiwa Digital Kiwa is a true pioneer in digital storytelling media localization and the preservation of indigenous languages Whether it's through interactive learning tools multilingual media or cutting edge voice technology their mission is clear To break barriers connect cultures and amplify voices that have too often gone unheard And at the heart of this work is today's guest Jordan Kosio Repia senior project manager at Kiwa. Jordan's journey is a remarkable one rooted in peaceful towns of Woria New Haka Forgive me I'm trying my best with the pronunciations here Jordan of Mayor and New Haka And shaped by a deep passion for media storytelling and Mayori culture Over nearly 15 years at Kiwa Jordan has gone from audio engineer to product leader driving global dubbing innovation with voice Q the gold standard in localization software used by Netflix Disney Amazon and more. His work blends audio engineering project management and cultural insights into powerful solutions that shape how stories are told around the world Jordan welcome to the show. how are you very good Thank you Thank you Like I said this is a different time zone for me I'm not used to to doing it in the afternoon So and you're over in New Zealand correct yeah down here in Tomak Makoto in Oakland Very cool Very cool Well welcome Welcome We'll see if we can get some people in um in this time zone joining us today If not of course people are watching watching the recording So Jordan I tried my best to write an intro here uh about what we want to talk about What I want to talk about today is of course that I want to talk about voice Q I want to talk about multimedia localization and workflows because I'm a process guy Um I'm an operations guy I want to hear all about those workflows and all of that stuff but I also really want to hear about the what you're doing to um help lift up voices from different cultures underserved languages and all of that stuff So just kind of give you a sneak peek on the questions I'm going to be throwing your way today But first why don't you tell us a little bit about Kiwa tell us a little bit about Voice Q and your role. Jordan: name is Jordan Kosipio and I'm a senior product manager at Kiwa Digital So for the past 15 years I've had the privilege of being part of Kio Digital here in Alteor New Zealand um that's been from its inception as a used to be a video production media house and then around the time of the iPad when that was first introduced around 2009 uh the entire company pivoted to educational content it actually started with children's books So we're actually adapting one of the first companies I think in the world to adapt children's story books into digital storytelling with animation and full live voicing and multilingual integration so that was one of the first key points of that from there it evolved into what you're seeing here on screen which is um a whole bunch of CQ is what we call it It's cultural intelligence applications So here in New Zealand we have a very rich culture with Tadel Maldi but we also support the Pacifica community you know the Pacific Islands that' be Fiji Samour Tonga etc So here in New Zealand we have a lot of those different languages So just we support those in businesses So allow them to basically translate their business culture into an application to essentially onboard but also teach them the culture that they're coming into So we have we have a lot of different apps in that space Uh everything from corporate you know we do have consumer apps as well So a little more fun stuff But yeah if have you have a look at our deal digital app store you'll notice there's a huge range of everything for Android and iOS But myself I've been part of what's called VoiceQ so I've been helming that since about 2016 and we have a really strong relationship with a lot of our clients especially in the dubbing space you know Netflix Disney other SBO providers but we also support smaller indie and indigenous groups as well to help revitalize language That's a key point of difference on that So most of discussion today I'm really happy to talk about is around voiceQ like you said you're into systems so those dubbing workflows how AI is coming in to change all that and how we're adapting with that but also how our company's always been at the forefront of technology but also very much keeping with what we call human in the loop So yes AI is coming in Yes it's taking over dubbing but we're not all about hey let's replace everyone with robots Let's keep this human interaction Let's keep that human storytelling But hey why don't we use these tools to basically take all the medial jobs and medial tasks that are traditionally done in dubbing and you know bring those down to a very low cost so that way the creatives actually have more space to work in so that's really what we're doing there cover cultural view So yeah it's quite a range of things we do. Tucker: it's it's interesting to me because I was checking out what you guys do and all that stuff and researching it and it's interesting to me the um you do language services and you have technology solutions right and you're embracing AI and you're embracing the human in the loop and the human perspective of it and lastly it's you have these scalable big how do I put this like you're you're working for Netflix right you're working for Disney Um and I'm not breaking any NDAs there It's on your website and yet you're also working with smaller local regional equally important sure but these languages of lesser diffusion I guess is the the term that people use underserved languages underserved cultures right and to me it seems like chemo it's it's largely I mean there's a lot of like contradictions in there Not contradictions but things that you don't see together every day right. I've talked to a lot of people that have worked with Netflix You know a lot of companies that have worked with Netflix and they're these big companies They don't do a lot of work with underserved communities right. Conversely I've talked to a lot of companies that work with underserved communities That's what they specialize in They don't do a lot of work with Netflix right and and you're kind of in the middle of the ven diagram with both And I think that's kind of a unique position for you to be in Jordan: It really is And I think it's what's like you said it's intrigued a lot of people about Kiwa and where it sits in that landscape because it is quite interesting you know you see obviously we travel up to the United States quite a bit you know we do all the shows and we go visit you all these different offices and things like that and I I understand where you're coming from with the the corporate side and then it's quite funny we come in it's like hey we built this well it's like hey we built this this tool that you know helps bring your costs down So obviously the financial guys are happy but it's also about how can we serve the creatives And I don't think I don't think the voice cue application will be where it is today without us actually like hopping on calls and different time zones to actually listen to the creatives and see exactly how they want to use the tool So by listening to the creators we've built a tool for directors and adapters and audio engineers specifically So we listen to all of those workflows We're like "Hey what features work what don't?" And that helps with that And then in terms of the management side you know the finance guys are really happy because obviously we're bringing your bottom line down but the creativity goes up right so we've had a lot of huge success with a lot of obviously Netflix television dubbing which you know just before 2020 all the way through Um but you know like like you said in terms of indigenous communities you know we we helped the Cherokee Nation basically you know they wanted a way to dub I think it was Amazon's Rings of Power into um the Cherokee Nation language So we're like look we'll help with that We helped um I think the Wolf Willow Studios in in Dakota to translate and dub the Avengers into the Lakota language and you know we've helped other communities as well like like you said it's um those are the ones we're actually quite passionate about there's a really good story and like even here in New Zealand we do the titty or moldi versions of large Disney productions and doesn't seem like much but to see those communities interact with that content in their own language it means quite a bit to them and it's very impactful Um as opposed to just you know like a a Korean drama into English but even that's quite impactful as well because again that's reaching a whole new audience So yeah we're really happy to be part of that But like I said we're really about helping those creatives to deliver good dubbing Yeah Every We don't want you to deliver the old kung fu film style dubbing It's all about how can we get that perfect lip sync That's what they're all about They're all striving for that perfect lip sync but they still want voice actors in the studio to be able to you know do their job So what voice Q really is at the heart of it is we really call it karaoke for dubbing But behind that like you said there's a workflow you know everything from translation adaption QC management to recording and mixing all in the background of that karaoke um screen for the actor Yeah Tucker: And I you know you talk about uh localizing the like Disney movies for example And to me that's something that's super important because like so many in this day and age a lot of us have just become used to watch consuming content with subtitles Like if I were to watch a Korean drama I'd probably just watch it with subtitles because I don't want to watch the dubbing Um but those Disney movies and they they're for kids right kids don't read so good yet right so it's super important I think for especially for the younger generation that they have access to to that dubbing in in whatever language it might be And you're doing it for these local languages Um I I do want to get into Yeah I want to I want to take a look at the tool I want to take a look at voice cuz I want to see um hear a little bit more about this with the voice uh with the lip syncing and the dubbing and all that stuff But before before we get into that I just want to talk a little bit about this technology versus culturalization right because I think there's this perception out there that well the machines can't help us The machines aren't going to replicate culture The machines aren't going to be able to help us with culture Yeah maybe we can machine translation some stuff but we're always going to need humans And and you agree with that You said it We're keeping that human in the loop But I want to ask you about how is technology not just changing the way that we translate not just changing the way that we dub and do lip syncing and all of that technical stuff but how is technology actually helping us culturalize content Jordan: it's really about those I guess the cultural entities actually taking part in the technology So I think I think the unique part of New Zealand is the multi businesses here are very integrated into technologies So we have very great you know telco partners here you know we roll out internet to like remote you know places and and like literally in the middle of the forest there'll be mai with a group of people and you know in New Zealand we have this unique opportunity to roll out internet there I know over in the United States they're doing the same and it's about how do we educate these communities to you know I guess feel empowered to use technology to push their language So you know what we are actually in that space where large language models are being created for indigenous languages Um I think the biggest discussion is about who has ownership over that right um definitely the people that create it have ownership over it So one of Tucker: there's been some controversy particularly out of New Zealand if I remember correctly um reading some case studies about this about there being a little bit of um a kurfuffle let's say about a techn I think it was Microsoft coming in and creating a a data set using a local language correct me if I'm wrong or if I'm barking up completely the wrong tree but the idea is that indigenous people should have ownership of that language and the second a big company like Microsoft for example I'm just using them as an example Or Google starts using that language data to create engines large language models whatever it is This kind of gets into the philosophical debate Are they stealing the language from the local people Jordan: no as long as they're involved in it I think that it's fine I mean we've we have a good relationship with Microsoft and but a lot of our technology is built on Amazon's AWS and they actually worked with a whole lot of different um you know multi- influencers and people in the top in terms of um language to actually write what's called a multi data lens uh which is all about data sovereignty and ownership rights for you know digital content in terms of that indigenous space and I know in Australia and other places that they're definitely writing those briefs as well Uh I do know the one you're talking about Um it's all about the usage So obviously you're going to have times where it's misused Here in New Zealand we have good cases where like you know alcohol brands or other brands would just use a moldy word because it sounds cool and not actually understand the meaning of it Um I'm sure in different countries that happens all the time as well as you know someone just thinks "Oh this word's cool but it could mean something terrible and they just don't understand it." but yeah in terms of ownership of the language yeah it is as long as the people are part of it I think that's you know the indigenous groups need to be a part of that conversation then it's fine But again I think that's where the data sovereignty comes in It's like okay so whose language are you using so here in New Zealand you know we have all these different tribes and it's similar to you know other places around the world So their dialect and their way of saying a word may be different you know up the road as opposed to where I'm sitting right now so that's when you get into that large language model training It's okay so what tribe is training the large language model so when I you know I can go into Google and do Google translate on hello you know similar to the speech I said at the start I can type into Google that it may give me a completely different translation than what Microsoft has because obviously they they're talking to someone different so then that brings up that question of who who's in the right and at the end of the day it's like well no one's no one's in the wrong but no one else is in the right either because there you need to respect these different levels of language as well Tucker: Well let's go over to I want to take a look at the tool here so voiceQ um sponsor or funded by created by um supported by Kiwa um but its own standalone tool here used by a bunch of different people out there I'm going to bring up the website here on the Nimdzi 100 which just recently came out This my own shameless plug for myself there Um what is it what does it do just give us a quick intro here And I I saw also maybe it makes sense I saw that you had a an intro video Maybe we can just play that No I can't find it of course though So oh here we go Let's check this out Yeah sure You might not be able to hear it but they can hear it. Voice Q 8.0 Yeah that's that's a language release Mac OS application design is a complete pre-recording and recording tool I'm trying to narrate here It supports over 900 languages and you can import over 15 script file types Oh you're going too fast for me All right I didn't realize this was not um voiced over which come on with a tool like Voice Q your your promotional material should be voiced over Come on there's a missed opportunity there Well why don't you just tell us about it Jordan: all right No so um yeah it's it's been a tool that's been around since the early 2003 Um so well before my time Um but we decided to resurrect it in 2016 after you know I think one of the senior managers were like "Hey what's this voice Q tool that people are still using?" So yeah people were still using the tool like after all those years Um so we reached out to a whole lot of you know a bunch of the studios are still using it They're like "Yeah we use Voice Q on the daily to do dubbing." They were like "Okay this is interesting to see." So we had a few conversations there and then we released Voice Q 2.0 back in 2016 and we went up to NAB in Las Vegas It was one of the first times we went there and introduced voiceQ writer which was our translation tool So prior to that there had always been a high quality dubbing tool used in studios right So we needed a break into the translation adaption market So we we brought in a lowcost version of the application that integrates straight into the studio one So that way an adapter could download this and create a whole script and basically send it to a studio and an actor can just come walk right in and just record And in terms of recording we integrated very specifically with ProTools So about let's just say 90% of our studios all use ProTools every day So we integrated very heavily with Avid Satellite and also with um traditionally with MIDI So we're locked in really tight with that So we have a really tight sync between both So if I press play in ProTools press play in voice Q everything's locked in and synced and then the the way it works is that the video output comes from voice Q There's a karaoke overlay that basically goes across the screen that the adapter actually needs to sort out themselves So they do the assignment and the alignment for the translated script that comes in and then they do what's called a trans adaption where you basically translate But let's say I'm speaking to you directly on a screen My lips are moving like this Basically what their job is is to adapt it for the local territory to make sure that the lip sync matches what's being set so that's really where that grew from Um and then the next part of the workflow obviously is the recording So recording the actor and then mixing that down and then on top of that we have uh what's called our voice Q cloud system So we've got the Mac OS app which is our flagship application but also the cloud system communicates all that together which you're seeing here on screen in the back there that gives your project management people like an oversight and overview of like where things are at So you know one of the biggest challenges is like hey man we've got 20 episodes that Netflix want recorded by next week help so there is to figure out you know how do we cast everybody how do we manage all the lines how do we book it you know how is it priced so everything fairly priced and set up um so we help with all of that so you know it was really awesome to create what's called the voice cloud grouping so what it is I can take a series of um television scripts So let's just say 12 episodes of a TV show I know hey Tucker you got to come in You've got um let me just press a button You've got you know 600 lines you need to record And all of them have a varying degree of duration but we could do a duration check So it's like okay you got 600 lines and that duration totals about 30 minutes So you be coming in to record 30 minutes of dialogue We've got that you know and then you as an actor have that power to be like "Okay cool I'm going to charge you x amount depending on where you are." So it empowered the actors It empowered the studios as well because now they're not just shooting in the dark Because prior to that you got paper scripts or you're just using Excel sheets You kind of got to you bring an actor in you know your one hour session ends up being three hours So now you're paying them two hours extra Well. Tucker: I remember at working on these you know back when I used to have a real job and I didn't just talk into a microphone like I remember working on projects where we would pay for the voice actor to come in and we'd also pay for what we called like a voice coach to come in and that person was there to kind of help them with the timing help them with the pronunciation There's nothing worse than having a studio session and then a week later getting feedback from the client that you mispronounced something or um something like that and you got to do it all over again right um and it sounds to me like technologies kind of alleviated that need for having a voice coach necessarily in the studio or is this is this all happening like in studios or are people using home studios to record this because I know that was a big thing that a trend that's happened in the last 5 10 years is we're starting to see more and more voice actors use home studios Jordan: Yeah So remote dubbing is 100% possible voice Q Um we promoted it pretty heavily around the the 2020 time right um so through 2020 to about 2021 you know we promoted remote recording using our systems You could do that on cloud or on native and then send that through to studios to do mixdown So because a lot of our clientele are big studios we really get it to say "Hey look you just do a recording and then push a button and send it to the studio and they'll mix it they'll process it they'll chop it they'll do the fades." So we took a lot of that off of the actors based on feedback and it worked out really well because obviously you know they're still producing content over those two three years Remote recording is still a big part of some um actors businesses I actually think it's quite cool how you could be anywhere in the world Um someone sends you content you record it and then send it back. One of the cool things with Voice Q cloud um which we've recently done is you know we added DRM for those people to make sure that you know content is safe So yeah I can share with you a project like a full live project uh for a limited time that has all the protections on it So you know you can't screen record and etc And you can do your recording send it back to me and then I'll just pull the project off you straight away So So it's about how do we do Tucker: Sorry for for the listeners out there DRM digital rights management, Why is that important? why is that why is that something that um people are worried about? Jordan: they're worried about it because of mainly because of leaks obviously you know you have content that's you know do on cinema or do on streaming Tucker: the studios that's worried about it Jordan: Yeah So it mainly comes from the top So top ends make sure they protect their content to make sure that it you know it's streamed correctly um and safely So we we alleviate a lot of that stress by you know letting them use their own S3 buckets you know in terms of AWS for S3 that's like their own little hard drive they can use in the cloud uh where they have full control over the content That's really helped a lot of the big studios and smaller studios to you know use cloud and remote recording Um because I'm a real advocate for remote work in general I think it's I think it's great I think it's just really helped out in you know it allows people in smaller communities even to you know become part of a larger group Tucker: right you don't have to live in the big city anymore.Yeah You can have your home studio and you know raise your family in your hometown And there there's a lot of there's a lot of advantages to every a lot of stuff being remote these days I would say. Jordan:_Exactly 100% Tucker: Now we've been talking a lot about voice actors like and I don't know if if you've got into the whole synthetic voice thing yet but um I want to ask your opinion about synthetic voices. I'll give you my opinion they're quite good right they're they're they're getting quite good Now I'm not saying they're like good enough for the latest episode of the Mandalorian right but for general purpose things the synthetic voices have really come a long ways So I kind of want to get your two cents on where do you how do you view synthetic voice and where do you think it's going where do you think we'll be 5 10 years from now with synthetic voices will it ever get to that point where Disney can dub their movies or streaming content with synthetic voices instead of voice actors um yeah There's my my open-ended question Jordan: Yeah I mean who's saying they're not already we just don't know it's that's really where technology is at nowadays no I mean Kiwa's stance on AI is that you know we want to use it as a tool to empower our users That that's the real key stance So we do have um synthetic voices through Excelate and veritone inside of our system Um basically you can tap into those voice data banks and create um content through that Um Tucker: like have you developed those synthetic voices or are you tapping into like Amazon Poly or using somebody else Jordan: yes we can tap into different things that we want Um so we're not personally not creating our own voices things like that Okay Um because obviously there's a lot of great third party services that already provide that and they've you know invested very heavily inside of that That's not something that we really want to focus on in our business We're really Tucker: it's a big barrier to entry It's a lot of investment into that Jordan: Yeah Yeah And there's a lot of obviously there's been a lot of um legal um stress around it as well as like you know what's right what's wrong things Tucker: like that That's kind of why I'm asking you because I I wanted to make sure if you were doing it or not because now my follow-up question now now that I know you're not actually building those yourself is is there a lot of concern that you hear from the voice actors for example about their voices being used to train engines because that's something that I've been hearing a lot like from voice actors is that well the more that I record then these big companies are just going to use my voice to train an engine and that's eventually going to replace me what have you been hearing around that Jordan: i mean I know a lot of the strike was about that and then I think it was more about you know these actors going to their agents and you know reconfiguring their contracts to make sure that that's all covered Um because it was a bit you know there's a bit of a gray area prior to that strike Um I think they got quite a you know a good turnout from that Um basically you know a lot of these large companies have said "No we're not going to replace humans in terms of dubbing uh for content which is great." Um there is a lot of stress around it I know a lot of actors are worried Um but again it's I think it's about who you work for and what what contracts you sign because at the end of the day you know you have a right over your own voice and we stand by that Um Tucker: but there was a time when that statement wasn't there was a time when that statement wasn't taken for granted though right when you say you have a right over your own voice like yeah now we do It took some strikes and took some negotiation as a society to figure that out And you know it's still not 100% clear I think to everybody out there but I'm glad that that's the way the pendulum has swung and hopefully it's going to stay on that side where you do have a a right to your own voice and your own likeness Jordan: Oh 100% Uh and and you know like I said there are a lot of cool thirdparty technologies out there that yeah they allow you to clone your voice pretty quick and you know you can make content and all that but yeah always read the fine print That's what I always say when you use any of this technology If it says that it's going to store your voice maybe don't use that system Um we actually partnered with Veritone and with Excelate because they have really good rights around voice ownership So you can create your own custom models You own that custom model and you know it's not shared with anybody else So I could record my own voice in the veritone I have my premium voice and I can use that for my content but no one else can use it unless I decide to license it out to somebody So that's why we chose um those companies. In terms of its use case in terms of voice Q there's actually two use cases we use at the moment for synthetic voice One is for audio description Um it's really helped in the audio description market because obviously audio description is just you know you just saying hey look guide jumps on screen right you know bell bell toll thing like that There generic stuff So replacing voice actors on that it's not a big deal And you know there's not a lot of emotion inside of that either So those flat voices work perfect for that. same goes for um you know replacing voices on you know real generic content So training manuals things like that That's all all been done inside of voice Q They can use that Um but when it comes to dubbing we're actually using the flat voice as a way to track lip sync Yep So let's just say I've got a Korean language script that's been sent to me and I want to translate it to English and adapt it So the adapter can go through change it translate it They can use machine translation or can use you know manual translation Completely up to them Once they translated it they can actually generate a voice for it and then use that flat voice to basically create that lip sync um that I was telling you across the band and then when an actor comes in they can use that as a follow guide to record Yeah Tucker: So it's kind of like a rough scratch track to so that you're getting timing right and everything That that makes a lot of sense because why not it's you're not paying someone to do it right and that to me is what I see with um generative AI whether we're talking about synthetic voices whether we're talking about generated content is that there a lot of the use cases aren't necessarily things that are replacing work done by humans previously It's stuff that just didn't make sense to pay a human to do before right but now it's like yeah why don't we just use this that makes a lot of sense to me And I I I did want to drill down into this about the because I don't come from a multimedia background Yeah sure I've worked on some multimedia projects before right but I it's kind of like its own unique space over there especially when you're talking about multimedia like entertainment multimedia Um and so it's kind of a blind spot for me and I've can't recall that I've ever actually worked on a project that requires lip syncing before so I want to ask you a little bit about that but first of all for the audience and for me too um can you kind of walk us through like the different you know what's what's the difference between because there's all these terms that I think lay people we use interchangeably whether it's you know the difference between subtitles and closed captions for example the difference between what is dubbing versus voice over versus all of these different things Can you kind of give us a brief intro course to voice over dubbing 101 here Jordan: yes So voice over is more traditionally narration So narrating tracks I mean you could say in terms of you know narrating over a podcast or something like that that'd be voice over or voice over on a video. What dubbing is is it's about getting that original content and using that original mouth movement um like I've discussed before and making sure that the translated language let's just say we're going from English to Spanish you want to have the Spanish words still lip sync up to how I'm speaking the words now Um so that's the huge difference with between dubbing and voice over So voiceover is very loose in terms of its sync as opposed to dubbing which has to adhere to a whole lot of rules So I think you talked about earlier about having a voice coach in the studio We actually promote having a director in the studio So the director is a voice cue master who knows the system and can adjust the sync and edit the script on the fly So a lot of studios in LA actually do this They actually have a director whose specific job is to guide an actor to say the right word And if the word doesn't fit the mouth movement they actually adapt it and switch the word out for something else It's basically like going through a thesaurus finding a word that fits that that actually works with like a MB or P or how you know how I'd say like a pop or something like that That's the director's job to sync that up for the actor So that even the great thing about that is with voice Q because it's all live editing the script once you're done you can export that out and that's the as as recorded subtitles that are you know put onto the the the platform So you actually have a dubsync script as opposed to just a subtitle one So again that goes against what you're talking about with subtitles and closed captions is with voice Q you can create a what's called an asrecorded script which is completely different to a subtitle So a subtitle you know that would traditionally be a complete you know transliteration of what's said in the original language So if you're watching something like let's say anime or things like that that's a straight transliteration but a dub potentially would just it would be I guess unique to let's just say it's done for the US right so it have unique wording that in terminology that's specific to the US as opposed to if it's dubbed in Australia they'll have completely different terminology for you know slang and things like that so that's kind of where the dubbing comes in It's like I said before it's trans adapted for the territory that it's being localized for Huh Tucker: I think you just answered a question that I've had for years which is why when I watch Netflix and I'm watching a Spanish show I can turn on the subtitles and I can also turn on the English dubbing and the subtitles don't match the dubbing And it's because there's actually two different scripts One's adapted for the dubbing to make it more to fit more the the dubbing and one's just a straight translation of the actual script Jordan: Yep So most content nowadays will have a as recorded script So that matches the the English dub and then there's the original translation which is more closer to what the original content was about. So the original content subtitle is usually done by actually from a different a different studio or a different business will actually do the subtitle as opposed to the dub. Tucker: Interesting Very interesting And and yes so this is super interesting for me to understand Um like I said because it's largely been a blind spot for me and how they actually get the lip syncing to work They're actually changing the script so that it more closely fits the um the the lip movements of the actors And yeah I was always curious like do they actually somehow like change the lips is that even a thing like using like CGI or anything or especially like with animated shows or with video games or something is that even a thing that happens where they'll actually like go and edit the lips Jordan: yeah So um especially in video games you know video games are a bit unique we we've got companies that do um cutscene recordings So a lot of them are just straight dubs for those cutscene recordings because a lot of them are baked into the system Um maybe they're not in engine films But a lot of the in-g engine stuff that's recorded using voice cue or you know just in general um the great thing about a lot of the game engines nowadays is you know you can just put the waveform in and it just the the mouth will match that perfectly but you know if you got static CGI trailers let's just say like a trailer for a new game that's coming out most of the time that's not going to be you know takes hours to render one scene They're not exactly going to render it for each individual you know translation they do So that's where the dub would still come into play on that So you'd still try to match the lip and adapt it. In terms of general just general content yeah it's um it's it's an I know there's a few AI third parties out there that are looking into it there's some really cool stuff where basically I can speak and then it you know replaces the mouth It's getting better every day obviously as the technology evolves that may be a thing I've seen some early development on that Um I think it was like lip dub AI was was one Um and you know we've been in discussions with a whole lot of these different players but yeah that's it's not something that's completely mainstream yet but like you said in 5 10 years it could be the norm Tucker: It's moving so fast Everything's moving so fast right and that Well Jordan: and it's generational as well Like I think because the kids nowadays are growing up on absorbing so much content it's almost like look I'm just going to talk about my age now Our generation's more about the quality and can actually see that difference because we know quality from you know the films and television that we've all watched But the children nowadays they absorb so much content at such a fast pace that it's almost like they don't care It's like so I believe the audience in like 10 15 years is going to be completely different to the audience now And you're probably going to be able to get away with a lot of this AI integration because you know because content is being consumed so quickly it's almost like they don't you know they don't really mind it Yeah Because they grow up on it right Tucker: this is what I say darn near every episode that I do here And I feel sorry for people that watch like the followers that listen to all of these podcasts cuz I just repeat myself over and over again But like when it comes to the technology like when it I usually say this within the context of machine translation but we can also use it within the context of synthetic voice for example or text to speech or whatever you want right is like two things are happening One technology is getting better The quality is getting better right but two which is what you just touched on is that societal expectations are coming down because we're more desensitized to it And you're absolutely right like my generation growing up we grew up with professional everything and we can really tell if something is a synthetic voice right but I watch my kids on their iPad you know watching the YouTubetubes and stuff and you know and the synthetic voices synthetic voice on social media is just used so much to the point that I don't think my kids or your kids are they're not going to notice it when they're our age it's just going to be completely different And secondly they're native to artificial intelligence They're going to be just in the way that I always grew up with the internet and the computer It was a slow internet and a slow computer but I always had the internet and a computer right my parents didn't have that so I was native to that They call me for tech support when their computer is not working right and it's like what am I going to be calling my kids for tech support for in 20 years right and it scares me It scares me because like oh my gosh I'm already losing touch Um you know next this weekend I'm going to go down and teach a course I I I do guest lecturing and workshops and stuff at universities And I look at these university students and I'm just like you guys are in a completely different age right now Like when I was in college I remember I'd go to the library to check out a book to read the book and take notes in a paper while sitting in the library And nowadays these kids have generative AI And it's like I don't know who to feel sorry for the kids or the professors right because how do you grade papers when generative AI is a thing and people are just using that to write their their essays and they should be using it to write their essays right i don't think forbidding people forbid like I I have my kids using chat GPT supervised mind you but I'm been encouraging my kids to use chat GPT because why wouldn't I that's the technology of the future and I think forbidding it is not setting them up for success. So anyways tying it back in to to our industry like you you mentioned you revived voice Q when it was you created voice Q 2.0 I think we're at what 8.0 now Jordan: Yeah For voice All right So we've pushing pretty hard I think we've done about thousand new features since we did that iteration So we just really been pushing hard Tucker: Well congratulations I mean I think your guys's reputation um speaks for itself with all the companies that you're working for Um it's largely an industry industry standard it sounds like I think a great kind of to start wrapping this up is what's in 9.0 What's on the road map like what are what are the things that you've identified for the future that this is where this industry is going and therefore this is where voice is going Jordan: Yeah So we have a unique take on AI and that that's really what we're pushing with the next version of voice Q Um so one is how do we close the gap on transcription so if I've got a I've got no script and I've got media you know transcription is so good nowadays Oh yeah Um you can just drop content in and just push a button it transcribes it We're doing that on device now So for even higher security you can just put it on an Apple device push a button it transcribes the whole film And then from there that's where you can create your translation script So again we're looking to save you know 80 to 90% of time just setting up the project Uh which is good the next part of that is what we call um we're using vision AI to create video analysis So that's going through and finding where a character appears on screen where they leave screen Uh it's doing scene analysis finding where a scene starts you know camera shot changes things like that That's all being automated now And the next stage on that is um lip detection So actually seeing you know how how an actor's lip is portrayed on screen and then creating a marker on our timeline to show that So when an actor goes in they know okay so I need to have my mouth a little bit like this or like this or maybe it's a closed mouth um in order to portray that better So it's again that's trying to help the actors portray better as opposed to you know using synthetic voice And then the final part of that is um you being able to export it quickly So with our cloud system that we built out that's all part of the ecosystem So I'm able to you know synchronize a project from here in New Zealand with someone in LA someone in Miami all at one time So that's really what we're looking at is is full globalization on a project So I'm able to send a project to you in real time You edit it I see what you've changed and I can record that straight in the studio without any downtime And like the goal of that is you know you can have directors working remotely sync into the studio actors there as well We can all just communicate on video chat and record in different spots but having that content secure is you know one of the biggest concerns that a lot of the partners have So we've been really focused on increasing that security So getting those security ticks from everybody Tucker: Well very cool I mean it's like I said things technology is advancing so quickly it's hard to keep up It. It's really hard to keep up right and in the meantime you have like those boring old requirements like the security requirements which are just becoming more more complex or more nuanced I should say And some companies are kind of fighting just because they're a little bit less they're a little bit more risk adverse I'd say but generally the conversations that I'm having with um clients enterprises is whereas a year ago I'd talk to people on the client side and they'd say things like well we're not using chat GPT because our CEO is forbid any use of generative AI from anybody We're not allowed as block from our systems Right i'm not hearing that as much anymore I'm hearing more about like how you know how are we going to be leveraging this technology and mandate from the top like I don't care how but you better be using generative AI for these things So times are a changing um before we Jordan: I mean as long as you use it well Tucker I think you know it's it's fine like yeah use it correctly don't just use it because you've been mandated to use it That's really our stuff Tucker: That's the conversation I'm having nowadays right where whereas a year ago it was um like I said people were saying "Oh we're not allowed to use it." And the conversation I would be having with people is like "No no no You should be encouraged like don't change your whole process." But now is the time to be testing prototyping playing around with it figuring it out Right so I was encouraging people to use it. Now the conversation that I'm having more often is "okay you want to use AI Why right" and I'm not saying you shouldn't use AI but start with the why And don't use it Don't use generative AI just to use generative AI There's got to be start with the use case and then figure out how AI can help you rather than saying I'm going to use AI and then trying to mash a system together based upon Jordan: Exactly This is that's us to use it as a tool to enhance your own work That's it Tucker: Exactly So and I I use it every day you know I use it I said it before I'll say it again I use it to write the scripts Um your introduction you know everything you know Kiwa Digital It's like "Hey AI here's a LinkedIn page Please write me an introduction for Jordan." And it writes me like nine paragraphs And I'm just like "Yeah okay We'll just take the first one for Jordan Hope he's not too offended." But before before we get any further I just got to say I've done a really bad job of getting to chat today But Gio I see you over there in the chat As usually if you don't know Giovana Giovana is great Um Gio I need to just start inviting you to co-host these things If you ever see something on the schedule for an MC live Gio that you want to co-host just reach out to me We'll put you we'll give you a block here on screen But yeah it's good to see you in the chat Gio Sorry I didn't do a great job of responding in real time. Jordan um I think we're we're getting to the end here Um tell us where can we find you um people that want to who should reach out to you people that want to know more about Kiwa people that want to know more about Voice Q where can they find you how can they reach you give us all of that good information here Yeah Jordan: So so we do a lot of bespoke um custom uh development So yeah if you want to get in touch with us go definitely go through u.com there's a really good portal on there for um contacting the team for that Um but if you're specifically looking for a dubbing or enterprise solution please feel free to contact you know contact me directly if you want to or go through voiceq.com There's you know we do full like 10day trials on voice Q so you can easily just load up and go and yeah you know it's not like a trial where you just go in and press a button and you're left to yourself You can actually talk to us We like talking to new clients about what their use case is where their pain points are and then figuring out solutions for that Um so we're really open to doing all that So yeah if if you're in that in the localization industry and you are interested in remote recording or just you just want to enhance your dubbing um workflows please feel free to reach out Tucker: Very nice Thank you And I just brought it up on screen here but if you're watching listening to this I encourage you to go follow Jordan on LinkedIn so you can be up to date with all of his stuff go that's going on Kiwadigital.com is the address for kiwa and voiceq.com if you want to know more about the tool itself right there And with that I will take us out Stick around Jordan and we'll chat right afterwards But in the meantime ladies gentlemen chat We are out of time for today Thank you so much for joining me on this episode of Nimdzi Live.