Can AI Really Fix Bad Podcast Audio? Adobe's Answer

Todd and Rob in the afternoon.

Hey, afternoon delight with Todd and Rob.

Oh yeah, oh yeah. And here we are with the new media show with Todd in the Philippines, and we're on a new streaming setup this time using OBS. So I'm looking at the interwebs and everything looks Good, Rob Greenlee. Welcome to the show. Yeah, so I'm looking at the stream and I'm not hearing— I'm not getting your guys's audio. So let me turn on another audio source. I don't know what happened here. Let me go and unhide Oh, lovely. There's always something that, that happens here, and I don't know why. So let me do this. Unhide all. All right, now let's try. All right, go ahead and talk, Rob.

Yeah, well, Mark, I know that the Adobe Podcast platform has been around for, what, a couple of years now, right? Why don't you tell us a little bit about kind of how, how the whole thing kind of came about at Adobe and what your connection was at the very beginning of this and, and kind of then we can dive into the details.

Yeah, absolutely. So I joined Adobe in 2018 when a company I had founded called SaySpring was acquired by Adobe. And SaySpring was a voice design and prototyping platform. So if you're building stuff for voice assistants, which a lot of people were at the time, you needed new design tooling in order to do that. There was a lot of synergies with Adobe. So we had the opportunity to join. I've used Adobe and Macromedia tools my entire career. So I was absolutely thrilled to join the company. And one of the goals of that acquisition was to come up with a broader voice strategy for Adobe. And so kind of the way our team took that is we stepped back and looked at, in a world where there was so much smart audio hardware, right, where everybody was wearing AirPods now, already podcast consumption was, you know, sort of growing rapidly. We looked at sort of what does it take to create good voice content, right? Good spoken audio content. And as we saw in the beginning of this, it's hard. It's hard to get everything set up correctly, especially if you're new to it, right? There's a lot of challenges with it. At the same time, we had an opportunity to meet a lot of the amazing researchers that we have in Adobe Research and see what amazing audio technology they had been incubating and working on themselves. So kind of I guess it was around 2020, early 2020, we pitched to Scott Belsky, who was Chief Product Officer at the time, let our team kind of roll out and focus on what is an AI-first audio strategy that we could put together that was, you know, based around spoken audio and sort of the mission that we always say is helping people share something as fast as it is to create it., right? So just say something and, and sort of cut down on the editing process, cut down on a lot of the work that's involved. And so we also, because we had been acquired, we were still a small team of engineers, designers, and a product person. So Adobe and Scott sort of helped us set up as an independent team and basically get started and build what has become Adobe Podcast. It's funny, at first, Podcast wasn't thinking about remote recording. And then the first week that we all worked on it together was March 2020, which was the beginning of the shutdown. And all of a sudden, remote recording became a big requirement. So we sort of built out this platform, started rolling out pieces of it. We were in a private beta, and to get some more people into the private beta, we decided, hey, let's take two of the core AI services that are part of the platform, which is Mic Check, which is sort of an online assistant that helps you set up your mic correctly and make sure the reverb in the room is good for recording. And then Enhanced Speech, which makes spoken audio sound like it was recorded in a studio. And we created these sort of single-serve pages. And just the idea was, oh, people are going to get a little excited by this, join the beta, kind of see everything that we're building. Enhanced Speech just took off. I've been building stuff for 25 years. I've never seen something just virally grow as fast as Enhanced Speech did. It took over TikTok, it took over YouTube really quickly while we were a team of 5 people. Quickly had 1 million users using it. That just started the journey, then got us to get some more people to join the team to build it out and start what has become the Adobe Podcast journey. But yeah, it was sort of humble beginnings of incubating something small that just took off because of sort of the genius of Adobe researchers coming up with Enhanced Speech.

That's kind of a— that's amazing transition from one thing to a product now that podcasters are using to record their episodes and do everything that Adobe Podcast or Adobe Podcast Premium because it's more of a package.

Yeah, it was always intended to be an end-to-end platform and product that helped anyone, you know, sort of record, edit, share spoken audio content, podcasts. And then enhanced speech just sort of became our wedge to help everybody know what we were up to. And so, you know, enhanced speech is still its own tool, is now embedded in Adobe Podcast Studio, which is the end-to-end experience. And so, yeah, so now, you know, we have people who do group recording there, do sort of the end-to-end text-based editing. You know, the promise of podcasts, the platform, is that you don't need to be an audio expert, that it's easy and accessible to use. You know, we don't even have sort of a multitrack editor. It's all text-based editing. So just way more lightweight than sort of, you know, some of the other workflows that we see people using out there. So it's also been a fun experience to just learn more about how people are creating these things, finding ways to kind of bring that functionality to people who aren't creating these things yet and are relatively new to it, but also the unique opportunity to do it in a place like Adobe where you know, we have Audition, we have Premiere.

Yeah.

So we, we don't have to go sort of pro and high end.

Yeah.

So for the, you know, life of podcast, we'll be able to sort of focus on a certain sort of simplicity that, you know, so the product lifestyle always ends up making you, uh, drive your product to be a little bit more complex cuz that's how you go after bigger customers and, you know, listen to pro users. And so, I think the benefit of sort of being at a place like Adobe is you have a bunch of that already. So we can just remain sort of focused on simplicity in, in again, a way like in my career, uh, I haven't had the opportunity to do. So it's a, it's a fun challenge of not only what do we build to bring to our users, but also like what don't we have to build? Like what can we, you know, drive to a different place in Adobe when that's what you're looking for and that's what you need.

You know, one thing, yeah, it's just a, Oh, go ahead.

You know, you know, I have been an Adobe Audition user for— I mean, I mean, like, probably since I started podcasting in 2004. So whatever version it was then, I, I can't remember using anything else. I may have used a free product for a few episodes, but I was like, I need it. And I'll be honest with you, I don't use 95% of the things in Audition that is available. You know, I use a few of the tools, so for me it's like, I, I, it's completely overkill. Um, it's an overkill product for me. Um, and then, you know, when I see Adobe Podcast, I, I kind of giggle a little bit because it really gives me everything that, um, all I use in Adobe Audition. So, you know, so you're cutting You know, the leveling that's available, you know, just a few things. And I'm thinking to myself, I'm, you know, I'm paying for an Adobe, of course I use Premiere too, so I need Premiere, you know, so I'm like, wow, maybe, maybe I could lower my bill a little bit. But for a new podcaster, this is one of the reasons why, you know, Blueberry did it, the, you know, the relationship and, and opened up this ability for Blueberry podcasters to get a 60-day free trial is we wanted podcasters to understand that they don't need a high-end product like Audition or Hindenburg or some of these other products are out there to get started. You get started, get your feet wet, and then maybe you graduate into Audition. And I think too, a lot of times, uh, new content creators are very, very intimidated by the editing process, and maybe they don't have a budget for an editor. So the text-based editing to me was like, it's no more complicated than really, um, you know, editing a doc file. You know, you just, you kind of scrub what you want, you delete it. It's transcripts. Yeah, it's easy.

And it really Editing a transcript is what you're doing.

Yeah, it's very, very, very easy to utilize. And, um, you know, I'm hoping from this, you know, what you guys are doing, obviously you're going to be improving your product over time, but I, I love the simplicity at this point. And I'm, I guess, you know, from my standpoint is, you know, the— you guys are obviously using some some AI, I'm assuming, in the tool, or am I wrong?

Yeah, no, there's several different pieces of AI that sort of power things, right? I mean, obviously, like, the transcription that also powers it is AI, but then we also do some smart things just under the hood, right, that aren't necessarily AI, but to just help make edits sound more natural. And actually, when you cut out a word, actually finding the right place to sort of make it as seamless as it can be when you use text-based editor. You know, again, enhanced speech is a core part of the recording and editing experience. It was funny, we did a— we looked into how many people are exporting something out of Adobe Podcast without turning on enhanced speech, and it's like 3%.

Wow.

Like, everybody who uses that tool uses enhanced speech to make it sound great. And so, you know, because we were like, oh, this is an opportunity to share this feature users might not know about. And it's not on by default. But that's, you know, I think that the ability to just sound professional, right, and have your content sound professional grade, I think is— it's a really hard problem to solve, both, technically and from a user experience perspective. But again, like that's the thing that our team just kind of is obsessed with and focused on is how do we continue to just keep that simple and make it so anyone can use it. If you go to Twitter and look at, or a social media platform and look at Adobe Podcast, you're gonna see lots of teachers posting stuff of kids, K to 12, who are using it, who are creating podcasts for the first time. So it sort of has that level of ease, but can create sort of professional-grade stuff. So yeah, it's been a super fun product to work on.

One thing that I was also, we talked to a number of folks and I think the thing that really set me apart and why we wanted to, you know, to align was that you guys are, you guys are a monster. You're a big company, and this is what you— this is what Adobe does. You know, this is the lifeblood of the company, you know, audio, video, uh, graphics, all the things, you know, all the tools that everyone's familiar with. So creative tools, yeah, of all the creative tools and the creative suites So for me, it just made sense because I'm thinking, you know, you've had this powerhouse name that is— they already know Core Audio. This is, you know, I don't know how many total years Adobe has done audio, but it's a process audio, you know, going all the way back to Audition and all the other video editing tools, but I just knew that over time that you're, you're probably your internal roadmap just says the product is just going to get better. So for me, the, the understanding is, is I can, I can be guaranteed this product's gonna be around tomorrow. It's going to continue to, um, improve over time. And obviously there's some integration now with some of the other products. So I'm pretty, I'm pretty jazzed with the product overall. I don't know, Rob, you want to jump in here? I'm hogging up the mic.

Yeah, I think it's really, really interesting for a company like Adobe to be moving into this direction of a, of like a single-use kind of tool, right? I mean, it's an entry-level kind of use case, but I On the other hand, it may be a way for you guys to develop AI capability that is appealing, right, in a single use case that then can be attractive, that maybe will help a consumer transition over to using like a full product. Or is the company, do you think, Marc, looking at this from, you know, these type of products like you've created within the Adobe Podcast area that that, that is kind of more, more of a glimpse of the future, right? As you think about AI and AI agents and things like that, because I can see where those 3 tools that you've built in Adobe Podcast so far are very much kind of aligned with each other from a functional standpoint. So, so I'm just kind of curious on your thoughts on that as you look to, to the future of AI agents and how that plays a role in what you're doing here.

Yeah, absolutely. So I, I think there's sort of two pieces to it, right? So you were first sort of talking about, you know, as, as people are getting used to new tooling, uh, and are, are new to different workflows. I mean, you see the same thing in other parts of the Adobe business too, right? So we have Adobe Express, which is our graphics and video platform that sits alongside our pro-level tools. Uh, sort of the same idea, either You, you don't need the powerhouse of Photoshop to make certain graphics, and Express is exactly what you need. You're new to graphic design overall, and you know, uh, that's the right tool for you. You're in EDU, that's the right tool for you. And so I think in order to, you know, our, our, our vision of creativity for all, I think that part of that is making easier-to-use tools for those kinds of users, right? And so I think podcast is that for audio. But I think you're seeing that across the company, right? I also think that you're seeing sort of across the digital landscape just the power of design and design becoming a more and more important part of all kinds of work and content creation, right, of marketing, of sort of what we have traditionally called knowledge work. I think design is a more and more important part of all of that. And so again, that sort of brings a whole different type of user to it. When it comes to AI, you know, I think there's— I mean, one of the benefits of, of AI, at least the way we've approached it in, um, podcast, is we very intentionally sort of looked at what is sort of the end-to-end workflow of creating audio content. What are all the things that either take time or take a high level of skill in order to complete? And then how can we use AI just to start chipping away at all those workflows, right? So, or just build it in a way that AI is not as important but still chips away at that part, right? So I would say the first is mic check, making sure you're starting with a good recording, super important part of it. When you record in Podcast, we're also recording locally on your machine and then sort of feeding that recording up into the system. So the audio that you have isn't the conversation the two people had with each other in real time, right? It's the local recording, so it's much higher quality. Most of our users have no idea that that happened, but it just ends with really good sounding audio, right? Then you have AI to transcribe it and give you a transcript to edit with. You have AI and enhanced speech to be able to make it sound good, right? And all the next things that are going to come from Adobe Podcast are, you know, how do we cut down on the editing process? How do we do, you know, automated filler word removal? How do we help you find the best clips to share on social, right? Like throw captions on those things, like all of the things that you then need to share that work with other people, right? I think AI has a place to fit in all of those things. And one of the things that I find sort of exciting and rewarding about, about what the way we focused on it is none of this is sort of like replacing creativity, right? It's like accelerating creativity. You know, I think that, you know, when we look at some of the other things that are out there, people keep asking me like, our podcast's going away, like, can't we just do like text-to-speech? I was just, I mean, like, can't you just have a robot write a book? Can't you just have a robot create an image? Like, I think that there will always, you know, there'll be a subset of work that I think, you know, AI might fill in. Uh, but I don't think creativity could be replaced, human creativity. And so I think when you can use AI to accelerate that process, I think that becomes super exciting, right? And it gets more people to be able to share more stories with more people out there in more ways, which, you know, to me is kind of the whole point.

On the enhancement of the audio process, obviously you all have perfected on the software side. I mean, there is a— you go into effects in Audition and you can really manipulate the audio in a big, big way. So on the enhancement of the audio, did you use the knowledge of, okay, we need to know to set this, this, and this? You had specific— I guess a better question is, was there a set of logarithms that was gained from knowledge on the core software product, or was this a brand new effort on the enhancement of the audio? Or did you know that 'Hey, we need to touch this piece, this, this, and this.' And we kind of know the algorithms already, that's generally good, or— and so I guess maybe the question is, is you had to sample the audio first to know what to change. That had to be relatively complex, I would think. Am I wrong on that, or what was that process?

So Yeah, so it is a completely different approach to enhancing audio. And again, this was, you know, developed within Adobe Research. We have some brilliant PhDs there. And it was funny when I first heard about enhanced speech. I don't know if we ever shared this, but the research had called it Awesome-izer. And so the Awesome-izer, when I I first saw it and I was enthralled by it. I talked to research and there was at that time no plan to put it in a product. Oh. And one of the reasons for that was when people were talking to customers, no customer was asking for it. Like, no customer was asking for just an AI magical switch that just makes it sound better. But it's also because customers didn't necessarily know that you could just do that and like it would work well. Uh, and so, you know, we, we brought it to market because of that. But the way it actually works— and I am not a PhD, so I'll give you my, my dumb guy version of this. So what it actually does is it was because, you know, the one of the benefits of Adobe is, uh, Adobe Research was able to get a big proprietary set of, uh, training data that was all people in studios speaking. So to basically train a service to say, this is what it sounds like, these are the audio characteristics of spoken audio when they're recorded in a professional studio. And then what happens is when somebody uploads audio, it looks at that audio and it just says, if that audio was recorded in a professional studio, What do I think its audio characteristics would look like? And then it just generates a new version of it, essentially. And so it doesn't generate voices. And it's kind of a black box. So we have a slider that kind of lets you mix it back with the original and stuff. But that's kind of all it's doing because you get basically one result out of it. And it's all based on audio. So it's language agnostic. It doesn't know what anyone's saying in it. It doesn't know what language is in it. And sometimes you'll get weird artifacts in it. Like, some of them might sound like they have a lisp, but it's actually because it's like cutting off a little bit of the audio because there was a loud reverb and it doesn't know what to do in that section. So it's kind of guessing. And so that's what it does. It basically extracts out the spoken audio from the background and says, this is what I think it would look like, sound like if it had been recorded in a studio. So wildly different approach to audio and very much based on like image generation. Like it's sort of looking at like the spectral representation of that audio and then extracting it out.

You should call it magic mastering or something like that instead of—

I know.

Well, the audio people will tell you it's not really mastering. It doesn't really master, right? But yeah, I mean, it is wild. And so we actually— I think we started chatting around November. That's when we launched V2. And so what V2 does now is also it separates out the background and then takes just the isolated speech, then enhances that, then gives you two sliders back that says, let me slide the speech between being isolated versus enhanced. And how much of the background do I want in? So now there's another AI service called SoundLift that's embedded in it that does the source separation and actually pulls the voices out above the background, but then lets you mix the background back in.

See, Rob, this is why there's other companies out there, but they don't have PhDs working in a research lab coming up with this kind of stuff.

Well, I wanted to ask you too. I mean, that's very much tangential to this is, are we getting to a point with this technology of noise reduction and background sound removal that the whole need that creators need to be worried about of having a really high quality microphone are slowly starting to disappear? So like Todd and I have these very expensive microphones. Are we quickly coming to a time when the quality of the microphone is somewhat irrelevant to getting studio quality audio?

Yeah, so it's a good question. I, I don't think quite irrelevant because I think the better audio you start with, okay, the better something like enhanced quality and quality, right?

So kind of, right?

Yeah. So it's, you know, it's, it's not an accident that we also launched a Mic Check service at the same time to like Hey, start with the best possible recording you can have. However, kind of one of the things that we always describe is Enhanced Speech is excellent at intentionally recorded audio. So if you're like recording outside in the world and there's a lot of stuff going on in the environment, that can be harder to work with. And sort of any AI, you may have some artifacts, some effect to it. But if you can like, sit in a room, even with like a MacBook laptop mic, and you can like look at it the right way and then speak to it specifically, enhanced speech and AI services in general will do like a great job with that. So I think we're getting to a point where, not that the microphone is irrelevant, but you don't need it, right? And it becomes much more accessible. So if you're you know, a teenager in Brazil using an Android device. If you take that Android device and you like put it on your desk and it's as quiet as you can make the room, even though it might be reverby because you have tile floors and whatnot, but you intentionally record it that way, it will come out great. It will come out professional quality.

Yeah, or even like an iPhone kind of cable microphone that kind of thing, it usually does a pretty good job. Right. And, and so I just wonder if these really, really high-end microphones are really going to be unnecessary.

I think for the average creator, we're already seeing that. Yeah.

It's like, you know, these $50 mics will suffice if you're using a sound optimization type of a tool like, like what you guys are offering. They may not have to invest $600 into a Shure SM7B.

So this is what I'm saying. So absolutely on that, that part, you know, I think that again, as long as you have a mic, right, a mic that is like set up and you're specifically recording into, uh, I also think, you know, so we haven't released it in any way, but on fast enough hardware, enhanced speech works in real time. And so I think it's only a matter of time before not only is this just like a post-production process, but just audio always sounds good, right? Every phone call eventually will just sound like it's coming from a professional studio, right? So I think just the world of— you give it 10 years, I think the world of bad audio will kind of go away.

Well, I think one example is like a RODECaster. Mixer has built-in, you know, noise reduction and all these kind of filters, dynamic range enhancement and all sorts of things that you can turn off and on in the software. Do you see at some point a tool like yours being built into hardware too?

So that's my dream.

I mean, it's, you know, not really Adobe business to be built into the microphone, right?

Yeah. I mean, I mean, we also, you know, on the Firefly side, Adobe Firefly, which is a host to AI, services we have. We've just also launched the ability to translate content, you know, in the speaker's original voice. You know, I think that it's inevitable that we will have a TV that will play any movie ever made in a voice that mimics, you know, the talent on screen that makes it sound good if it's noisy. You know, I think like so much of this right now is at the point of creation and post-production, I think it's not too far until it's all in the world of consumption and listening. And then you can just do it to anything ever. It doesn't matter what it was. The creator didn't need to do it.

The listener can. It's funny because at home, same microphone, I have an $800 preamp that it's attached to. It goes through an old traditional rack system with an EQ. A compeller, a compressor. It has, um, I have an Apex Big Bottom, and you know, I've got that thing dialed into exactly where I want it. You know, I've got sound panels, and you know, so the sound is good. But here I am, uh, 8,000 miles away from that studio, and I'm in a RODECaster right now, and guess what? It's got the Big Bottom built in from Apex into the actual, um, software panel that this thing offers. So I would imagine if I went and did an A/B test on how I sound right now raw compared to what I sound like raw at home, there's probably not too much of a delta. There'll be nuances of it. At the same time, if I was to switch the microphone out to, you know, an Audio-Technica ATR2100, there would be a different nuance in the tone. But again, I think in the end, with what you guys are doing here, the— it shouldn't matter. Shouldn't matter as long— again, Christian, if you have a Blue Yeti and you're talking into the top of it, well, then we can't fix that because, you know, they're not talking into the diaphragm. They're, you know, they haven't— they're using it incorrectly. So, um, even here, Rob and I talking a little bit off axis, the mic will support it. But if, you know, if I was directly into it all the time, it would be better. So There are certain things you can't fix. Background noise, those types of things that happen are still going to show up. I remember in Honolulu, in my office studio, I didn't— I had wall units for air conditioning, and there was no way I could run that while recording. So I opened the windows to get air, and of course, if someone started a, you know, a blower across the street, a leaf blower, then it completely jacked my recording. But Still, in the end, the product came out really nice. So I think that we're living in an age right now where there's a lot of forgiveness on the input. Of course, if it's bad, it's bad. We've all heard, you know, unrecoverable audio before. You know, part of our challenge of doing live events was it took us a number of attempts to finally get to the point where we could get the, you know, the roar of a background in a trade show down to the point where we actually could, you know, there was just a muffle of that sound. It's really hard. And people, I think today more than ever, and you tie video component into this, if the audio is good, sounds good, they will stick with you much longer than they will if the audio is bad. 'Cause people would be very annoyed by it. You know, I think about the early days when we were encoding in 16K, you know, that audio was pretty tinny. And, but again, that's what we had to do in those early days. Yeah.

Yeah. I think, I mean, I think bad audio kills content, right? I mean, I think especially, you know, any of us who were listening to podcasts in 2020 when everyone needed to start recording remotely and were using, Teams and Zoom and weren't really ready for it. You listen to some of those podcasts like they're unlistenable. You can't listen to them, right? Like podcasts I remember religiously listening to and then hearing sort of that level of quality. It just, it kills the experience. On the flip side, I also think as creators, there's a tendency to get obsessed with quality and to think, you know, that better is just, you know, better is better. And it's, you know, everything needs to be in 4K., you know, oh, I now I need to, you know, my phone can record 8K and so everything needs to be in 8K. Yeah. And, you know, people are listening to content on AirPods in the subway and, you know, kind of putting it on while they're driving. So I think that, you know, I think there's sort of a, and not formally defined, but I think there's sort of like a base level of quality that if you are below it, your content's dead, that you need to be above. But then I think sort of at the top there's a little bit diminishing returns, you know. So I think that, you know, the— we, we always call it the knobs and sliders. Like, can we just give people a toggle that you can just turn it on and I promise you it's going to sound good enough? Yeah, like 98% of people will like it. I think there's also, you know, it's funny, even as I, I work with the audio research team, I mean that team, those ears are all so finely tuned. You know, they can listen to something and point out all the problems with it. And I'm just like, no one is ever going to hear that in a million years. Like, no one is ever going to notice that, right? It's the same thing with— I think a lot of people, you know, test like, like whether or not they can create a voice model of their own voice and then they hear it and like, that doesn't sound like me. Well, it's like, it sounds like you to other people. It just doesn't sound like you to you, you know? So I think that also understanding like who the audience audiences and kind of what you're planning to do with this content is a super important part of like creating good stuff and not just getting obsessed with like better and better.

There's a good friend of mine, I haven't heard from him in a number of years, but his name is Paul Figgiani and he was involved in some of the early days working with some pioneers in the podcasting space coming up with audio processing. Audio engineer. Yeah, he's an audio engineer and Paul was a fan of the show and he was like constantly And he finally heard one day that I was using Audition. And like 3 days later, he emailed me like 3 files, says, you need to put these in, you know, this filter, this filter, this filter, you need to use this in post-processing. And to be honest with you, I could barely tell the difference. But Paul had an ear that was just— and some people have it, it's just like wine. Wine to me is wine. You know, I don't taste the difference. I might know the difference between a Merlot or, you know, our cab, but that's good wine and bad wine, diminishing returns up at the top. Right. So, and I think other people, you know, and there's also, I've been in events where, you know, you go in and there's a, you know, you have the hardcore audiophile that is, you know, they got the $5,000 amplifier and they're wearing, you know, the $2,000 headphones and they're immersed. But, and I'm thinking, yeah, but like I said, most of us are just listening to music on AirPods. So, um, and again, obviously the quality on them have improved over the years, and but it's, it is, uh, I think there is a final lower quality, right? Yeah, for sure. And do you really need to have that deep unspoken word?

Yeah, and just, and knowing sort of also what you're trying to accomplish and what story you're trying to tell, right? Like you brought up sort of the roar of a trade show.

Yeah.

You know, our conference floor. When we first sort of had released enhanced speech, we had a bunch of example videos that were recorded at Adobe MAX, which is our big conference. And so it's on, you know, the studio floor.

Yeah.

And you toggle it on and it sounds super clean, like it was recorded in a studio. Wow. But when you look at it, you're like, it's like uncanny valley. You're like, well, you need the background noise. I see that person is not in a studio.

Right. Right.

So you kind of like need some of that background noise, which is what led to us kind of adding that stuff. So, you know, all, all again, sort of like understanding what it is you're trying to, to con— to, to get across to the listener is an important part of the process.

But at the same time, I've been in some events where I've been holding a microphone. And a person that kept moving off axis, and they're driving you crazy, and you have a huge number of people. So sometimes you just need to get the stuff isolated so you can hear it. And obviously Rob and I are in controlled environments, and most podcasters are as well. But then again, there's podcasters that are doing, you know, podcasts in bars, and, you know, so you want a little bit of that, like you said, that background ambience You know, they're not drinking beer in this sterile environment or doing shots or whatever they may be doing. So, um, yep. Yeah, so there's— it's nice that you've added that. Um, Rob, go ahead.

Yeah, I was just going to kind of expand on this topic too because I've done some videos, um, that I, I've run through the audio optimization processes for the videos and the videos were like filmed either in my car while I'm driving or something like that. And it's maybe it plays into exactly like what you guys were saying is that sometimes the expectation of the audio quality doesn't match with what the actual audio quality is. And that's what it gets into this conversation around. Well, if I'm doing a recording in my car while I'm driving and people are seeing a video, but they can hear me talking at the same time. And the audio sounds like I'm speaking into an SM7B, that there's kind of like a, maybe a disconnect there from the standpoint of, well, his audio shouldn't sound that good.

Being there.

Yeah, absolutely. Right. So I've had that experience too. And it's great. You know, I can listen to my video and I can say, well, that audio sounds terrific, but it does seem a little odd to have that quality of audio in the car while you're driving. So anyway, and I've worked with these tools for a while now down this path of the record anywhere and you just run it through this algorithm and it cleans it up and makes it sound perfect. And which is great for a lot of situations that you can get into. But I also know that your platform on your premium plan is now working with video too. So you're doing that same thing where you're processing a video, but you're fixing the audio in the video and making it better.

Is that right? Yeah, absolutely. So it's funny, you know, in Especially at Adobe too, right? People tend to think audio and video, but yeah, audio is a very important part of video, right?

Right.

And so, you know, when we first launched Enhance Speech, people were yelling at us like, this needs to support video. I use it for a lot of video. And big spoiler, we're going to be bringing video to our recording. We're going to be bringing video to our editing because it's just what we hear from people. That's kind of what they need.

That's a hot thing now, right?

Yeah, it's funny too. It's— I— so we're like, we're researching it a lot around like recording too. And kind of what we're learning is, so, you know, we do like the local recording of audio right now when you record an Adobe Podcast. To do that for video is, you know, a big technical challenge. It's solvable, but it's a big investment. But so we've been analyzing, well, could we just do cloud recording of video but the local audio, like, are people as particular about the video quality as they are about the audio quality? And as we talk to users, like, it turns out that for stuff like what we're doing, right, like talking head videos, podcasts, the audio is actually way more important than the video, that people are pretty forgiving of, you know, a couple artifacts on a talking head video or two people having a conversation. If the audio sounds good. And so it kind of becomes this flip of when it is audio-first content or videos, the audio needs to be better than the video because you can't do it the other way. So it's, yeah, just interesting to kind of learn what people actually enjoy consuming and listening to.

So I was using a variety of tools to do this show from here. When I'm home, I have a very high-end system. I have a TriCaster. It's completely overkill, but I get incredibly good video captures and the audio chains really, really good as well. But when I came to this type of situation, which is more like most content creators are doing now that are on the podcasting side and having the video portion, my challenge really was I was getting one or the other. I was getting either great video or I was getting great audio, and I wasn't really getting both with some of the tools that are out there. And it was really driving me insane. So this is one of the reasons why I've changed up what we're doing when I'm in this location and how I'm recording. Yeah, I'm back. I'm recording the, you know, just so everybody knows where I'm using Zoom, but that's not where the master recording is happening for the video. The master recording's happening in OBS now, but I am doing a backup recording in Zoom in case something goes sideways, which is, you know, backup. But I would not use the Zoom video that is— because it can get weird, it can get pixelated if someone's connection's a little bit better, and it could happen here too. But I think your idea of if you can have it so that you record both the video and audio locally and upload even if the person has to wait until a little bit of time after the live event, that to me is the best solution for getting both a good quality audio recording and video. Because if I say something and the connection goes a little bit sideways, the computer will catch it. The live will suck, but the master will still be good because we still have the master recording and same thing on each of your screens. Me, I think that is the way to, to, to really get that. Well, now sure, you could, you can build a cool interface similar to Riverside or StreamYard, but again, I tried both of those tools and they have their own limitations. StreamYard's cloud recording sucks, it's horrible, where Riverside has great video recording, but the studio tools weren't that good. So I think if you can get a combination of both, you'll knock it out of the ballpark. You'll compete heavily in that regard.

Yeah. So we haven't made decisions on exactly how we'll roll it out, but I will say that in the kinds of— so it's hard to fill gaps in audio when recording. It's not that hard to record gaps in video. That's true. If you have Jaggies, if you have a stutter, that's where I think there's a role for AI to be able to like fill those gaps and make it sound good, right? So that won't work on a motion picture you're editing in Premiere maybe, right? But it definitely works for sort of the kinds of stuff that people would be creating and working on in Adobe Podcasts, right? So again, I think that, you know, we kind of always joke about like, what's enhanced video going to look like? To sit alongside enhanced speech? What is enhanced music going to sound like? So I think there's a lot of opportunity to kind of solve some of those problems because, yeah, local recording might be great. But also, you know, if you're that Brazilian teenager on a low-power Android device, that can't hold the whole thing, right? It can't— the memory on that can't hold that local recording. So there's a big opportunity for, you know, with the distribution and the reach that Adobe has to make something that is like super accessible to a whole bunch of storytellers that, you know, have a lot of trouble creating those kinds of pieces of content now. It's just like— I think it's all a balance and we'll figure it out.

It's just like right now I've got a light adjusted wrong, so I'm getting a little refraction in my glass and it's driving me absolutely crazy. I can't fix that. It's there. It's going to live. I can fix it in— I can fix it after I get done and adjust the light. I can't— well, you said maybe you can fix it, but right now, you know, it's doing live. There's no way to fix it.

Yeah, I mean, live is definitely trickier. I will also say I had a call earlier today just as the sun was going down, and I noticed the reflection of the screen in my glasses. So 20 minutes before I joined this call, I put my contacts in so that I wouldn't have this when I talk to you both.

Yeah.

So there's still things you're gonna need to do to make the content better.

Yeah. I would love to be able to actually do, do my show entirely with a lavalier, but the problem with the lavalier, it is that it just doesn't have the same kind of quality. So I've got a little lavalier and a lot of people now are just doing videos now holding their little tiny microphone in front of their mouths now more and more. That's the, been the hot thing to do. But that's kind of what I'm talking about as far as being able to do this with just a small mic that you maybe put on your lapel or something like that. It could be wireless. It could be anything like that. And we could actually just get rid of these things, right? And take them out of the picture. And I think it would improve the quality of the video too. So So we don't have that big hardware right in our faces. So I was just trying to show that.

It's, yeah, absolutely. It's funny. As a product creator, it's really easy to fall into the trap of like having a narrative in your head about like who the users are and who the people are using your tool and kind of what they have at their disposal. Right. And when you do it right and you talk to a lot of people, you learn a bunch of things you would've never realized or thought. I was talking to a correspondent who works with BBC and they were telling me about, you know, when they have correspondents all over the world, the different regions that they're in bring different audio challenges, right? So you're in the Middle East, there's a lot of marble there, a lot of reverb. In the interior spaces where you tend to record. Right. You know, Brazil, Latin America, similar. There's a lot of, you know, tile everywhere because of the climate. In India, a lot of spaces are very small. And so again, a lot of reverb. I talked to somebody who made education content. They had a creator in Africa and they would record outside at night because that's when it was the quietest. But there were crickets in the background, right? So like the crickets were always in all the content. It's always something. And so you just, you know, you start to realize that, you know, hey, even if you can get the fancy mic, a lot of, you know, even if you're a professional, even if you're, you know, a BBC journalist, you're gonna be in places where you're still gonna have challenges, right? And so again, I think that's to me the perfect place that like AI can help create you know, or help people create content that would be hard to create otherwise. We also always hear the word save. It is so common that somebody goes to record something with somebody and the lavalier isn't turned on or wired correctly, and the only audio they have is the mic audio from the camera.

Yeah, we've all done that.

And so that's what they, you know, that's what they're stuck with. But yeah, you can enhance it and you can still use it. We've had— I won't name them, but there are several big-name podcasters who have been saved by enhanced speech plenty of times because of hardware setups and glitches that happen throughout the show.

Yeah, it's going to happen. It's funny because you guys may saw me moving around here. I figured out where the glare was coming from. It was coming from the notes that Rob Greenlee sent over. That white screen was enough to cause the glare. Wasn't the light at all. You know, it's an interesting time now with so many people trying so many different things, you know, when it comes to audio. And you look at TikTok and some of those, believe it or not, some of the audio on TikTok is pretty bad. Because, you know, they just— oh yeah, this is where they're, you know, they're trying to record into a ring light and, you know, maybe they have a mic, maybe they don't. Um, but I think there's lots of opportunities out there and, um, I, I'm kind of excited. So on Adobe Podcast Premium, it ties in Adobe Express. How are you seeing that utilization? Are the people using the Adobe Podcast product and then they're going straight into Express, or what, what does the usage look like between those two, those two products?

Yeah, so far what we see is, is sort of maybe adjacent workflows or like people supporting kind of one way or another. So a hobbyist who uses Adobe Express to make videos and graphics to support their Etsy business decides to record voiceovers and podcasts or make a podcast about their hobby. We also see sort of the other way around, right? So people who record something and then you need cover art. So cover art in Adobe Express is a great place to do that. You know, there's a captions tool within Adobe Express. And so, you know, I think over time you'll see us do a better job of sort of integrating workflows across the two tools. But one thing that we've noticed sort of more broadly just in the creative ecosystem is, you know, you sort of mentioned TikTok, like there's a new generation of creator who is used to not having end-to-end workflows, who is used to using 3 or 4 tools to do each one of those tasks, right? You know, so they record somewhere, they edit somewhere, they use this tool to like create all the clips that they put on social, they use this tool to add the captions to them. And so, you know, I think that our first point is to help expose the people who are using podcasts to the things that Express can do for them, even if, you know, I think same thing. Like, I think we can get obsessed at Adobe with like, it needs to be that you can like easily just transfer the project over here and keep working on it. But I think the starting point is also just the education of just like making sure that, you know, people know everything that they have available to them. You know, for— I'm a little biased, but you know, for $9.99 a month to have all of podcasts and all of Adobe Express available to you.

Yes. It's amazing. It is amazing. And if you do an A/B comparison for people that are listening or watching this, and I think it's important to watch budget for, for creators, because what we see is usually the hosting part of the bill is the least expensive part of their entire budget. They're spending X number of dollars for an AI tool. They're spending X number of dollars for an editor. They're spending X number of dollars for maybe a headline or a Canva or whatever it may be, and the next thing you know, they've got $100 wrapped up in additional charges per month. Whereas that doesn't have to be the case. And I'm pretty— I, I, I, I raised 5 kids, so, I mean, I was living in Hawaii at the time, so I was super budget, budget conscious. And because we had to make the dollars go a long ways. So in order to reduce spend and have a product that is just as good, to me is important. So that's why, you know, I applaud you guys on that $9.99 price point. I think that's where people should really— maybe their eyeballs open up here a little bit, um, in, in thinking about this. If you're starting to be, you know, especially with economy the way it is right now, be budget conscious. And I'll be honest with you, I'm a— I can draw stick figures, you know. I could barely find my way around, you know, when it came to editing stuff in Photoshop or whatever years ago. It's just like I would screw things up more than I would fix things. So for me, a tool that is simple, templated, and gives me what I need maybe isn't a masterpiece if I was an expert in all those other tools. But, you know, you can get pretty, pretty doggone good and something that looks really super professional, um, at $9.99 a month. I think that's the— for me, that's the thing that makes me say wow, you know.

Yeah, I mean, I think that's the, you know, the product challenge for us too. You know, you mentioned earlier that, you know, you use 10-15% of any given tool. And I think with Pro Tools, that's, you know, definitely the case. But I think that's the case with most software. Just everyone kind of uses a different 10 or 15%. Right. So, right. So our game is sort of, can we, across all the things you need to do to create this stuff?

Yeah.

Can we have a good opinion on like what we believe the most important 10 or 15% of functionality across all those kinds of tools are? And then give just that to people, right? At like an accessible price point to help them create whatever it is they need to create. Right? So it's, it's, you know, we all spend a ton of time, you know, everyone on our team has, has a pretty big background in just tooling. And so we all are very opinionated on kind of what, you know, how do we deliver on that promise? And again, knowing that we'll always be able to maintain a certain level of simplicity. Because we have the pro-grade tools, you know, uh, a lot of, a lot of our tools are fighter jets, right? It's like getting in the cockpit of a fighter jet when you open Photoshop. Yeah. It is an amazing piece of technology, right? It has created, you know, countless, you know, um, important works that are, you know, part of our culture. Uh, but there's a learning curve. And I'm just— so how do we— and I'm just putting that right, 10 or 15%.

And I'm just putting text on an AI-generated image, you know? So, you know, I'm using 0.001% of Photoshop's ability where it can be done in Express. And, you know, so—

Yeah, absolutely. Yeah, so it's a fun challenge as being somebody who, you know, builds products.

Yeah. So this topic is really interesting because I was watching a panel of AI kind of experts and where they were seeing the tooling that would be driven by AI as you look to the future. And this one person I was hearing talk up on stage on a panel was from Hollywood, right? Was from the movie industry. And so what he was doing was trying to create AI tools that matched with the expectations of the studios and how they like to produce movies, right? So, you know, there's separate steps that go into every movie production, right? And if you apply AI tools to do some part of each step of this, that will fit in better with the human workflow process. So are we doing this really because we're trying to fit in with the human workflow process, but eventually we will come around to maybe not having so many steps and trying to connect those steps together. Is there any thoughts on that?

Yeah, it's funny. Um, we have this debate all the time on our team too, because I am very aware as somebody who's in their mid-40s and has been using creative tools forever, uh, that I just have a certain way of working, right?

Yeah.

I almost never drag files. I'm a file picker guy. I like the— we've had big debates where my, you know, our experienced designers, like, I drag and drop everything onto the canvas. I was like, I don't think I ever do that. I always like go to file, click import, right?

I'm a picker too.

Apple Pie. Yeah, totally. I think, right? So I, I think that there's, there's parts where, uh, I don't, I don't wanna be too much the generational guy, but I think like there are just when people are new to a workflow, there's an opportunity to like kind of rethink what it is, right? And so I think when you see, you know, Hollywood studios and the way they create stuff, and then opportunities for AI to accelerate that workflow, that's obviously a huge opportunity. But then you see, you know, movies with a $6 million budget clean the award shows that was edited in Premiere, right? You know, last year, the editing award went to, you know, another film that was also edited in Premiere. Those creators are going to have like completely different workflows and different ways of thinking about this. And they're now going to, you know, sort of rise up in the industry and they'll all be, you know, having huge productions at some point in their career. And so I think like it will also change, right? And so I also think there's an opportunity for sort of these like to— it's an overused term, but like AI first, Right? Of like, there's just gonna be opportunities to go record a bunch of stuff out there and just make it sound all as good as it can sound and then start there. And you don't have to worry about making sure that you're lugging around all this audio equipment. I think we're gonna have the same thing for video, right? I think like there's gonna be ways to kind of iterate much faster on like what shots look like that, that an individual does in a tool like Premiere, right? Versus sort of, uh, more of a Hollywood studio kind of workflow. So I think you're going to have both, right? And I think— I also think one, one of the things that I'm super excited about is because of the nature of AI, it gets trained on what has happened in the past, right? And yeah, I think what's going to happen is certain AI outputs are going to then just seem like they're AI. So even if they're not, right? Like, I think that, you know, when you look at something like NotebookLM, right? And like, it's like, it's very sort of, you know, NPR style conversational podcasts of two people in a studio. Yeah. Even if you now have two people in a studio, I think people are going to start to think some of that sounds like NotebookLM and be like, oh, this is AI. So I think people are going to start creating like different approaches to stuff in different formats. Like, to me, almost like the, the notebook element of a— right, it's a picture saying it's breaking the mold, right? I think that, you know, AI-generated images that are sort of like fantastical and have a lot of like light blues and light pinks in them, and that everybody kind of looks like AI now.

Yeah.

And so I think you're gonna have, eh, that sounds like AI, that kind of looks like AI video. And so I think you're actually gonna have a whole bunch of people doing new stuff, you know, this whole like— I think AI avatars and the social media videos, all that tooling is going to make all of that stuff look like it's AI-generated, which is going to make people create something completely different.

Right.

And so I think it is going to like— I think people are going to just continue to break the mold. And so I think social media content will look different. Podcasts will sound different. Movies will start to look different because everything is, you know, a whole bunch of things are going to start to feel like AI and people are going to want to break that mold. You know, I think so. I think that's kind of what I think is super exciting about, about where we're headed with this.

If we take—

Do you see Adobe playing a role in that?

Yeah, I think, you know, I think Adobe can play a bunch of roles. You know, I think Adobe can definitely play the role of there's a Hollywood end-to-end workflow and we can help make it more efficient in a bunch of places. Uh, I think we're gonna have a, a bunch of tools that let anybody create really professional looking content. And then I think, you know, back to kind of where we started, we're going to have the tools that people are going to help break the mold with. One of my favorite things is, I guess Illustrator now is, I don't know, 34 years old, 35 years old. When Illustrator first came out, its tagline was, we don't even know everything it can do for you. And like, that to me is just like the guiding light of a creative tool of just like, I don't even know what people are going to do with this stuff. When we launched Enhanced Speech, people started taking it and like adding adding the artifacts to ambient music and like EDM because like they actually like the roboticness of like the, the service getting something wrong. And so they like mixed it into a whole bunch of tracks. So there's a whole bunch of weird stuff like people will create and break the mold. And so I think Adobe can sort of play a role in all of it, which is, you know, again, not to, not to drink the Kool-Aid too much, but it's a super fun business to be a part of.

But if we bring it back down to basics, I'll be be kind of frank, especially for new creators, this is the people that I'm worried about the most, is they get really overwhelmed. We're talking a lot of real cool stuff here, but I think what I've seen, and at least my mindset is when it comes to Adobe Podcast and Adobe Premiere, um, and excuse me, Adobe Express, is the ability for someone that again is getting started, they're new, They're gonna get in, they're gonna record their podcast, they have a simple editing tool, they got this great way to get enhanced audio out, and it's a process that doesn't take a lot of time. A lot of, a lot of content creators today get so far in the weeds. Oh, I edit for 4 hours on my show. If I had to edit for 4 hours, I would have quit 19 years ago. I just couldn't have done it. So we want people to focus on the content Allow the tools to, you know, get them to the point where they've got a great sounding podcast without having to spend 5 hours editing something and they get wrapped around the axle.

You let it—

let's get— let them get 5 years into the content or 3 years into the content before they get completely wrapped around the axle and doing 3 hours of editing or have someone to hire that does that editing. I think there's got to be a point in the beginning, and I think the thing I always tell content creators, just hit record. Let's get it recorded. Do minor editing, let's get it out the door. The product's going to change. Your first step, 10th episode's not going to sound like your first, and so on and so forth. And I think this was my major appeal with Adobe Podcast, was that it wasn't, to me, wasn't overwhelming. I took a new gal that we had working for us, say, hey, uh, do a demo video real quick. Just go in, figure out How to use Adobe Podcast. And like, she's like, well, that's easy. And I think that's the key. It's real easy and you'll get product.

Yeah, I think it's, I think getting, you know, it's, it's like how do you write a great novel, right? You just like write the first page and get started. So I think that, you know, we need to make it as easy as possible for people to create stuff. Uh, Again, I also think that the, the tools and simplicity, um, sometimes changes the, the form, right? Like, I think now when you watch a lot of the YouTube videos, like, you'll see a lot of jump cuts because a lot of people do text-based editing, right? And, and it's just become part of the format, and I think everyone's okay with it, right? Um, and so I think that, yeah, you kind of get started and you start in, and, you know, even back to some of the, the AI stuff, like, I think the new creative journey is usually about mimicry. It's usually about sort of wanting to create something that seems like the thing that you enjoy. Yeah. Lin-Manuel Miranda tells a great story of like the first time he ever went to go see a live musical was Rent. And he was like, and I immediately came back and wrote a whole bunch of things that sounded exactly like Rent. All right. And then you sort of find your own voice. And so I think that easily accessible tools and AI help people start the creative journey. And usually that's going to be mimicry of some sense, right? Of like seeing the kinds of videos they see on YouTube, the kinds of podcasts that they listen to and get started there. And then, and then hopefully instead of 3 or 4 years down the road where they get obsessed with the, the, the craft of editing, you know, maybe they get obsessed with like the craft of creation. And because the editing is so much easier, they can just like push themselves and try different things, right? And once you can create content much faster, you can just try a lot of different stuff. And so I think that that's again, sort of what is, I think, exciting about this whole AI revolution, if we want to call it that, is I think it's going to make it so much easier for anyone to create stuff. And then I think it's going to make it so much easier to like break the mold and bring out new stuff no one has seen or thought about before.

You talked a little bit about the Ellen Notes. I think that's what it was, right? The Notebook LM? Yeah, yeah. And, you know, I could see right now you can hear that audio and just know immediately where it came from. But what I would foresee happening with that product is they may dumb it down a little bit to make it make people make the voices make a mistake. And you'd be like, okay, the AI would not normally make a mistake. Was that an AI or was it not?

So I Oh, it's just a setting.

Yeah, you have to do—

yeah, more human-like errors.

Yeah, have a sneeze in there or a cough or something like that, or too many ahs and ums, you know, have that, uh, have that typical thing that we have with— but again, at what point, just like you guys and some of the tools that you've created, is, you know, I, I record my voice for 3 hours and— or whatever it requires to train a prompt. And the next thing I know, I'm speaking French on some—

Yeah, I don't know. I'm pretty optimistic about this kind of stuff. Like, I think when you create this kind of AI content, I think you should not try and like mimic that they're humans, right? I think that falls apart fast, right? Again, like, even now when you listen to it and it's, you know, kind of been in the public consciousness for, what, 6 months? It has a sound, right? And so I think that nobody wants to listen to AI that they think are humans. But, and, and I think some of the conversational artifacts that they put in there with the ums and the ahs, I actually think that goes away because what I want out of a tool like that is like, I want to dump like all my educational content into it. And then I want to like listen to that when I walk to class, right? Or like when I'm driving to like be able to consume information in a more efficient way. Way or a more enjoyable way. So I think that that's where that stuff goes, right? And then I, and then I think that again, the human beings are going to lean into creating stuff that feels very different and authentic. And I think that we may be in this middle ground where people are trying to get away with AI avatars that people think are humans and stuff. But I think like, as we all kind of get exposed to it, I actually don't think that's what people want. I think it's, it's, it's attractive because of like, oh, content production costs cut down so much and, you know, but I think that stuff gets relegated to online learning and, and, you know, the, the, the training materials that we have to watch all the time working in a big company. Uh, and then I think on the sort of social media side, there's a place for it, but I think it, it changes. I, and I think like, again, 5, 10 years from now, I don't think mimicking, you know, AI mimicking humans is kind of where sort of entertainment plays itself out.

I definitely hope it gets separated out and marked. That's something that's important to me is to have it marked so that, you know, now I can tell if I watch a YouTube video and it's, you know, it's that voice you hear every time, I immediately back out. I think we're all going to be seeking original voices. I think it's going to become more important as time goes on. So podcasters, YouTubers, Anyone that's creating original content, I think people are going to seek that out because like you said, I agree with you too. The AI stuff should be relegated to, you know, the 1-hour training content you have to listen to every quarter or whatever it is from, from your company.

So yeah, who knows?

Maybe. Yes.

But I, but I think if, if I had the option to, you know, watch shows from another country in another language that I couldn't watch otherwise, or like, I love watching cricket highlights that aren't in English and now I can listen. So I think that, and like, I'll know it's an AI voice. I'm totally fine with that.

I'm good with that too.

But it now makes it more accessible. I think, again, like, I think this like makes the world sort of more fluid and accessible and kind of open stuff up to other people in a way that, you know, wouldn't happen without AI versus, you know, people trying to use AI to just kind of, get one over on you and pretend it's a human being.

There was a time when I was watching some foreign films, Italian specifically, kind of these dark dramas that came out of Italy for a while. And of course it was all subtitled. I would have loved to have been able to hear that in English.

So I think, yeah, there's a place. There's some stuff that will get a budget that they can localize across different languages., there's a lot of stuff that doesn't. And so if you can use AI to build audiences and, you know, again, and I think, and going the other way too, you know, again, I think we always, in a lot of these conversations, we think about like English-centric and then being able to like put it in other languages. So, uh, people in other countries can listen to our stuff, but the other way around, right? Like, again, there's tons of creators in, in Brazil and India and, you know, Mexico, Turkey. Who could, you know, which are by the way, like the 4 countries that are the biggest for Adobe Podcast after the US, um, that can create a bunch of stuff that is then in English. And then not only can English audiences appreciate it, but then you can also run English ads against it so those creators can, you know, make better ad rates, like a whole bunch of kind of stuff that just like changes, I think, what the sort of online economy around content creation and viewing looks like.

And there's a series of Portuguese podcasts that are— they get massive numbers, a million listeners per episode, designed for factory workers. And I've always wanted to listen to one of those shows and to understand what are they saying to keep these people listening 3 hours to a 3-hour program every day. You can hear the excitement in the voice, but I don't speak Portuguese. So I just, I kind of have to to wonder, you know, is it half comedy? Is it, you know, and again, so I think there'll be the ability then to learn from other content creators and learn new ideas from different countries too.

Yeah, yeah, absolutely right. And then, and then every country has their version of the podcast for factory workers that is based on sort of, you know, what I mean, we see with television all the time, right? There's, you know, America's Got Talent is in every country in every language. And so I think you're going to see that down the stack at the podcast level, you know, as, as kind of this stuff becomes more prevalent.

Rob, you got any more questions for Mark before we let Mark go?

Well, yeah, I just, I was just thinking about something too, something that I've seen happen a lot with video on, on YouTube lately is these compilation videos, right, where people are pulling real-world video from libraries and then tying them together with a story of sorts, right? That changes and transitions and things like that. And so I think those videos I've seen that actually are real human characters in there, but maybe an AI voice or a compilation that is generated by AI of human voices that are actually talking about a specific topic and how AI can kind of like mix and match and create something new out of something that's preexisting or that already has been produced by a big creator or something like that is what I'm seeing. Like I've got up on my screen right now a video that's doing the exact same thing as that. It's got 100,000 views or that kind of thing. So I think that there is kind of AI-generated content that will drive audience. It's just, you know, what's the topic, what's the relevance of the topic, and how trusted is the source for that, and how good is it produced is going to be the other element here. And is some of those video and some of that audio AI generated or, or not. So I think that there, there may be a hybrid of what we're talking about here, which is the combination of AI plus human creators kind of collaborating together to create content.

So it's gonna be interesting. Yeah, I think, you know, yeah, I think the AI workflows we was gonna talk about with this human in the loop, right? Where there's like a human who's part of it. I think again is, yeah, super necessary and I think these kinds of tools like are going to open up new kinds of content that wouldn't have been made otherwise. I do also think that it's going to also lead to a lot of slop, right? There's going to be just a lot of—

that's the new term now, slop, right?

Yeah, it is. I remember, so my daughter is now 11, and I remember when she was, I don't know, 3, we made the mistake of putting on YouTube. It was the first time she ever watched it, and it was some like animated little, you know, nursery rhyme video. And then like 2 videos in, there was just like a creepy version of that kind of thing with like an obvious like synthetic voice singing a song and just obvious like bad 3D graphics. And I was just like, oof. And so that, that was the beginning and the end of her YouTube watching career. So there's also going to be a lot of that, right? Because for a lot of the, you know, possibility of making one of those videos that gets 100,000 views, a lot of that's going to come from making like a million of them, right? And so there's just going to be a lot of stuff out there that is also completely unwatchable.

I call it the new forum spam, you know, that's what it really is. It's, it's, it's video spam.

Video companies are going to be using this, this technology to produce promo videos for all sorts of different things, I'm sure.

Yeah. Yeah.

And this is also, I don't know, I guess this is also just the nature of media and the nature of creation, right? I mean, this is, you know, there's always gatekeepers who want the studios who were in charge, owned all that actors and actresses, right? And kind of then TV, then cable, like it just, it always goes down.

Right.

And so I think that you're, we're definitely living in a world of like abundance. When it comes to content. And I think, you know, a mix of the platforms and audiences and people kind of just continue to sift through it. So there's always a learning curve. There's going to be a lot of slop out there for a while. But I don't know, I think we're, I think we're going to get better new original sort of stories and things that we've never seen before sort of come out the other side of this.

Well, it's definitely going to be an exciting time, uh, for sure. And, you know, any projections then on Adobe Podcasts and Adobe Premiere— I mean, Adobe Express, excuse me. Again, I keep— some reason I keep going back to Premiere. Uh, what it's looking like for the next year? Yeah, yeah.

I would say that the two big focuses we have, uh, is, yeah, I think video is going to become a bigger and bigger part of the workflows and the platform itself. And then back to sort of the way I had described our AI vision of just like, let's look at all these things that take a lot of time, either setting up or in the editing process. And how can we use AI to solve that for people and make it much easier? So, you know, people have that across a whole bunch of different things when it comes to creating clips, enhancing video, right? Making it look good, you know, solving dropouts. So I think that, you know, we'll continue to collaborate with Adobe Research pretty tightly and making sure that, you know, we can just help people create great stuff without needing to be a pro or have tons of time or big budgets to get that stuff done.

Awesome. Rob, you have anything else for—

Yeah, I was just going to ask, let's Let's talk about kind of how a person would sign up and what they would get as part of, you know, it looks like that there's two plans. There's a free plan and then there's a premium plan. You know, if you could talk a little bit about the free plan and what people get with that and how that works.

Yeah, I mean, it's, it's a pretty affordable and open platform in general, so it's a freemium product, right? So there is a free version that can be free forever that you can use. And then there's a premium version if you upgrade, and even that has a 30-day trial on it. So the free version, you can record unlimited record, but exporting up to 30 minutes, up to 2 projects a day. On enhanced speech, you can enhance up to an hour's worth of stuff every day for free. If you, you know, record in studio, you can enhance it there. If you have something you've already recorded, a piece of audio, you can upload it. And then premium is really about sort of almost like how much time do you have, you know, like there's a lot of stuff you can do in the free version if you have a lot of time on your hands. But if you, you know, need to work faster, get stuff out quicker, That's where in premium we have additional sliders. We have video support so you can upload a video directly, not just the audio like you can in free. Unlimited recording, unlimited exports. We launched something called Audiogram. So if you have a piece of audio, you can share that on social. It creates a video with the animated transcript. So, you know, and then of course the $9.99 plan also comes with Adobe Express. And Adobe Express Premium has, you know, hundreds of thousands of templates for logos, for graphics, a video editor, motion graphics. I mean, pretty much. And you can build an end-to-end marketing department with Podcasts and Express for what you get for, for $9.99. So it's a pretty, pretty solid deal that I feel pretty good about.

So is the service Based on a browser usage model, or is there an app associated with it?

Yeah, so it's actually all browser-based. Yeah, so it's all based on the web. So, you know, Express as well. So, you know, Express has a mobile app you can download, but the rest of the Express experience is on the web. So yeah, there's no software to download. And so it's really easy to use. The whole recording and editing experience for podcasts lives on the desktop browser. But you can use enhanced speech to take any video off your phone, make it sound really good if you have premium. So yeah, it's a mix of capabilities that, again, sort of, I think because of the way we launched this, everything that we've added has been based on listening to users and what they need and learning how they were using it. And then how we've decided to charge for it was you know, you get an hour free of Enhanced Speech every day, you know, and you can come back every day and get another hour. So, you know, 30 hours a month for free, that's a pretty good starting point. It is. But people who are, you know, again, sort of, I think, working intentionally and have things they need to get done, I think Premium is kind of the place for them.

And for $9.99 a month, come on. It's the best deal. It's a fantastic deal. Yeah, yeah, that's can't be it. Yeah. Mark, I want to thank you for coming on. Is there, is there, I don't want you to necessarily give your email, but is there a good way to reach the Adobe Podcast team, or is anything, or just contact?

Yeah, so you can, you know, at, yeah, podcast.adobe.com is where the, um, app is located. Uh, there's a contact section there, but anyone can always email us at teampodcast adobe.com, and we read and look at every email that we ever get, so feel free to reach out.

Hey, thank you for being with us late tonight. We definitely appreciate it, and, uh, much success. And we'll, we'll check back in a year and see where the product is and, uh, look forward to the additional enhancements.

Yeah.

All right.

Yeah, I appreciate the conversation, guys. Hey, thank you so much. Thank you. So Rob, uh, any specific news that, uh, we have seen that we need to cover before we bust out of here today?

Um, I don't know. I don't think so. I mean, I think there's only like 5 minutes left, so I don't think we have enough to go too deep on anything.

All right, um, I guess we can just call it a a day here. Yeah, I guess what we'll do is, um, yeah, so everyone, thanks for, for tuning in today. And, uh, here I got this on the screen. I think I do. Let me move this over here. Uh, getting used to this, uh, as—

so yeah, I'm fumbling, sorry. Okay. The only thing that's been bubbling is this whole discussion about Spotify and video.

Yeah, we'll spend some discussion on that next week for sure.

Keeps spinning up there and there's a lot of— this is a pretty important topic for the podcast landscape.

Yeah. And if you don't know what we're referring to, there's an article on PodNews that links to a LinkedIn article that has gotten a lot of attention today.

Yeah. Yeah, and it's all about the issues that we've been talking about in this show for, I don't know, probably a good close to a year now, probably about how Spotify is allowing people to upload video to replace their audio.

And that's—

And some people losing money.

Yeah, so yeah, which, you know, and I've had some conversations with people off the record that don't want to go on the record that are, you know, heavily concerned about this particular issue.

So, but anyway, and I saw that— yeah, and I saw the Evergreen Podcast Network launched an experience on Roku on the video side.

Oh, really? Some people are back to building Roku apps.

That's funny. Well, I know you spent a lot of years doing that too in the early years.

Yeah, we were testing. So we were app 12, 13, and 14 on the Roku. So third-party app. Um, yeah, it's, that's funny how we're coming full circle again.

Here we are. The whole video thing is really becoming something, and I don't think we've seen the end of this either. I think it's going to continue to spin.

Yeah. All right, everybody, I'm Todd at blueberry.com. I'm @geeknews on X. I'm at geeknews at geeknews.chat.

Uh, Rob, I'm on X, um, @RobGreenlee, on YouTube @RobGreenlee, um, on Facebook @RobGreenlee. And so I'm easy to find on LinkedIn as well. And if you want to send me an email, you're certainly welcome to send it to robgreenlee@gmail.com, and that'll get to me.

So I don't have our AI disclosure, but just, uh, raw here, any AI tool can use and index our our content here and use for your training models, we authorize you to do that. So awesome. I think pretty good run on this first try here and going to make a few tweaks next week, but otherwise I think it went smooth. Yeah, I think it went smooth. We'll see how the recording turned out. Little bump at the beginning with audio. I need to check that out, but I fixed it really quickly. So I've got backup upon backup of recording today, so we got a good capture one way or the other. So, but anyway, thanks for being with me, Rob, and we'll see you a week from today. Okay, sounds great.