Subscribe to the Adobe Audition Podcast
Never miss another episode of the podcast for Adobe Audition users by Adobe Audition users!
: Hey I'm Mike Russell from Music Radio Creative. And welcome to the Adobe Audition podcast honoring 25 years of Adobe Audition in this series of twenty five episodes. I will interview power users of this awesome audio editing software. We'll reminisce back to the Cool Edit Pro through to the introduction of multi-track editing and bring you right up to date with Adobe Audition CC and features like the Essential Sound panel. The show is brought to you by the Awesome Audio Gear Giveaway. If you'd like the chance to win my perfect audio creator studio setup. Head over to MRC dot fm slash win and enter now. Will be many prize draws every month with a final gig giveaway taking place at the end of September. That's MRC dot FM slash w i n for a chance to win. Good luck. Is my guest on this show is David Johnston. He received his Bachelor of Science Degree in Computer Science from the University of Washington in 1992. Right now he's a principal software design engineer with the audio and acoustics research group in Microsoft Research Labs joining in 2011 birds. The reason you may know David Johnston is in the early 90s or even I would say perhaps and we'll find out from David the late 80s. David started to create coolheaded that was a. Originally it was a stereo audio editor for Windows. I even believe it might have had a DOS version as well. And in 1995 CINT Trillium came about. David cofounded centrally software developed the multitrack studio. Also Ed. I remember him multi-track editing came. It was just amazing. And that of course was cool as a pro of course initially 1.0 all the way up to version 2.0 one and then he finally sold the company to Adobe in 2003 and continued to work on more became Adobe Audition. Right the way through until 2010. Now I'll delve into this a little bit during the show with David but currently understand David is working on things that include spatial audio for hollow lens and windows are really exciting stuff exciting guest Davis. I'm so excited to have you home.
: All right. Thanks for having me.
: It's a real honor and a pleasure. So let's start with a trip down memory lane. Let's go back to the late 80s the early 90s and what you were doing then Dave is cool. Edit not even cool let it grow. How did that first come to be. What was the idea. How did that spark initiate.
: Yeah well actually back probably back in the 80s I was layaways. I was using a program on their Macintosh called sounded and that was so much fun being able to like record your voice and change it and add effects and everything but I had a PC at home and now was unable to do all that. So I had to write the stuff myself to do the same thing and then yeah. And so I had done small little programs like that and then around 1991 and that summer my windows wasn't even really much of Windows Yellowstone version to point one which was really hard to use. So I do use make my own user interface and windows and dialogues and everything. And then being able to record sound and then play back and again back bank pieces are really small I had about maybe one and a half seconds of memory in ram that you killed that audio. So I had to have everything running off of a hard disk. So I had to be very efficient at reading an Audi off the desk and playing it and you know and throwing away and getting more out and so on. Yeah. Anyway so it was quite a struggle back then. But I did get something running in Windows three point one and then the first release of shareware was in January and basically in 1992. So hence that 25 years
: Like you mentioned that it was playing initially with the software all Mac and thinking hey I'd like something like this on Windows. So going back to those days you mentioned Windows three point one was it was it three point one or was it Windows 98 where there was a piece of software called sound recorder. So it was cool at it before or after the native kind of sound recorder that was there.
: Yeah. And so you this is about five years before that and there was a I think there was just something to record where your voice and playback but it didn't have any effects or do anything. So it wasn't so much how it was even in there. Again this was a 16 bit version of windows. Well before the first Windows 95 and everything. But it was really cool that it actually worked. And so actually that's why I called it I'm cool at it because it was like sound at it but cooler.
: So that's how the name came about. That's an awesome story. So yeah cool days. And I understand initially it's not a very basic single track editor and then it progressed to the multitrack editor is that is that right. Maybe you can take us through the story of how that happened.
: Yeah that's right yeah. It was just to track Ed.. Because I was really in stereo audio and stereo Ofex and I was playing around was like this brainwave synchronize or effect I think Jason was talking about earlier in early episode there and just like I'm basically just an audio playground just experimental right the Coden and you know it sounds and then if something worked out well I would posted online write up to post a DOE bulletin board because this was before we had the Web. And then after them when Amway came out it was a lot easier because I could just have a hosted up on a BYB site and people could download it and actually that's when sales started really taking off.
: So I definitely want to get into a little bit more behind the success and how cool it is and then cool it took off but I just heard you mention that the brainwave synchronize I wish you write in an earlier episode. Episode 2 of this podcast I was talking to Jason about to tell us the story behind that and did it actually work. Did it bring you into a state of flow or anything like that.
: If nothing else there was a great placebo effect. It may have actually worked. I've had people who say they have tried it and hook themselves up to an EEG and measured it and showed that between mamm that's somewhere between Faida Wavves and alpha waves like 5 Hertz and 10 hertz it was actually working. They said
: Yeah someone needs to bring that back in one form of that. That was an interesting experimental feature. So let's talk about the the secret behind cool and cool let it pros success. Now certainly from my professional career before I now myself I'm a podcast. But I did work in radio for a couple of decades and it seemed to be the standard audio editor in every radio station I ever worked at. So how did you get so prolific with that software. How did that happen.
: Know I was about just as amazed as you are. I guess we started getting these registrations from you know people from radio stations even people here. Now I'm in the Pacific Northwest and in Seattle and even got our registration from was it was Norm Gregory I remember being a Cummo radio or something like what KOMO Radio uses this and you know about ten years later we did a tour of Como studios and they were still using it. So that was pretty interesting. It just turned out to be. I tried to write the programs are easy for me to use and you know if something kind of got in the way of the flow then I would try to fix that and make it easier. And it must have worked for more people than just myself.
: Definitely. So yeah. What started as a project I guess a passion project for yourself David turned into something
: Quite amazing.
: Been working out of my kitchen actually. Yeah it started out with us working at a day job at Microsoft. But in the evenings and weekends I spend my whole time using the computer in our kitchen trying to write the code and then also try to handle all the people sending and shareware checks and that took a lot of time to which is why I had to start a company and had a partner and everything so able to do things a lot quicker with their help.
: That's my kind of project. So radio obviously a big part of your user base but who else in the early days what kind of other people or other organizations were using Coolheaded. Did you understand from the early days
: Yeah well we're pretty excited that we are that the BBC are using it and that's also radio. Some interesting things we found where researchers people who are doing like birdsongs or will calls or dolphin whistles are using the special features of audition in order to actually see what the sound of the animal is and then they can you know identify animals by the sounds they are making and things like that and I thought that was pretty interesting.
: Thats really cool. Yeah using this so much that it might you say you can see in the Sprague's spectral view. I mean now this is fantastic for noise reduction of the terrible powers all schoolings and you can kind of paint overheads and take it out without damaging too much of the surrounding sound. So
: Yeah yeah that was my idea I thought that be kind of cool and make it kind of like photoshop. You know where you can just see if you can see the sound. Why can't you just modify it or delete it or do whatever you want with it.
: I love that and yeah it saves so many recordings just being able to light you say paint over it like Photoshop style. So lets talk about okay so you mentioned about starting off from your kitchen when he was still working at Microsoft and then eventually I think it was around 1995 that you cofounded Sintra. It was 2003. The company was sold to Adobe. So in 1995 when that came about and it became a little bit more than just a kitchen project. You had developers working for you. Eventually I believe you you were managing a team of five developers in by 2003 so it grew and grew and grew. And you must have come across a lot of different challenges as you were developing the software no doubt you were getting feedback from users saying it would be good if we could have this or if something like this could exist. So what were some of the real challenges that you came across during those interleaved days.
: Now starting the company was a lot of fun it actually made it less challenging when actually I had a partner to do a lot of the work and help out so I could focus fully on development and that's why we were able to do they call it a pro. And so yeah one challenge was just working on that multitrack trying to get it to work on those slower machines of the day. And you know say you want to back then if you had say 10 tracks that's about all you could do. Live if you're going to be doing mixing but we want to do a lot more. So we came up with a system that could do 60 or 100 tracks if you needed to do a lot by doing a lot of pre mixing of stuff in the background by the time you actually want to play the audio. Everything was already mixed in and if you want to like move one track and time your cut and mix that from the whole multitracks so you can move it to mix it in live in real time and anyway a lot of these tricks were put in there so that it could be efficient. And then when that actually worked it was a pretty awesome to hear. Just to be able to build a really large inho sessions like you'd actually use if you had your own band and
: So cool let it go let it go. I seem to remember it for just the fantastic voice changing capabilities and the fact you could do all these pitches and stretches and add flanger modulation chorus all that great stuff. Was there any particular effect that you worked on that was your real passion something that you were really passionate about what was your favorite effect beyond just the multi-track editing capabilities. Anything that particularly interested you.
: Yeah I really enjoyed just experimenting with all those different effects and so you can find a fact which is in there for a number of years allowed you take a photograph and then convert it to actually turn it into a picture and spectral view. So whatever the picture showed you actually see it if you saw it in spectral view and just coming up with the math for that and then making it sound good and so on was really fun. Other effects like the AMA called The Fall River. So incepted the standard reverb you could control how the each frequency decays over time you could have a more more Bhoomi or more high and you know echoey and stuff like that. Just yeah all those effects in general were just a blast to work on them. And then just and also being able to publish it and put it out there within a few months after writing it so people could try it out I can get feedback right away and they liked it I can work more on that effect and make it better. So someone.
: The photo you can go to a picture inside the spectral view of cool let it cool it it probe any interest in that. So how did a photo sound.
: I could probably don't know if you can hear this so that's a picture of my dog.
: Wow. Can we can we do that today.
: So if you have a look at inspect or below you'll see a dogood
: Wow. Is it is it still possible to do something like that. I would like to do something like that
: Well I know Ed.. Three still have that. And actually I do use three quite often just because there is a lot of extra features in there which haven't quite made it. No not just the brainwave synchronized but many other you know there's a spectral pan view and I'm not sure that if this spectral save feature or suspected Ramin's and there are so magicians still has it that you can get your hands on it. It's fun to play with
: Definitely. That's that's him. You can you can create pictures in the spectral frequencies. That's just absolutely blowing my mind.
: You can even exploit it edit and Photoshop and then bring it back in to see what it sounds like.
: Photoshop editing was so so cool and I seem to remember audition three. That seemed to have something wasn't audition three or was earlier when you had the musical notes and you could put musical notes
: Oh yeah.
: On and you could kind of pitch shifting to a tune right.
: Yeah in fact I think that might not have already made it all the way to audition 3 but that was and you know all the cool call it a pro and so on. And that was yeah that was on though very early things I think I put that in around 94 or so and that was kind of based off again backed with sound at it that if you can take a sound and then assign it to you keys and you can hit the keys on the keyboard and play like a piano or whatever your sound was like a dog barking Jingle Bells or something. And so I thought well take this a step further since I have the pit shifter and stretch functions take whatever sound you want and then just pitch shift and stretch it and so on to match the score that you put on there. And it actually worked. It was kind of you know the user interface was wow looks like 1993 and it didn't really prove much sense then but it did sound good.
: I remember those little I think they were blue musical notes.
: So you
: Were maybe. Yeah you are. You are heading towards a kind of auto tune really I don't remember when auto tune initially came around. Maybe that was in the late 90s or something like that possibly so you kind of almost heading in that direction.
: Yeah. We were doing some auto tune like stuff some face vocoder stuff and so on to about the same time share guitar song with auto tune vocals and everyone's like auto tune.
: But yeah we were we were able to do stuff like that and I think there is a similar future arc. You've been in it was our recent audition three I remember
: Wow really really cool. Yeah you absolutely right. The late 90s Cher believe I think Eiffel 65 blue dabit D. That was another one. It was definitely a big trend. So. Wow really really interesting so I want to dig into more of your knowledge obviously of the cool days but also look ahead to the future with you David as well. Obviously now you're working on cutting edge audio technologies at Microsoft including projects with the halo lens which sounds amazing it's just let's talk about that for a moment and look into Holland then so I think I understand what Hollonds is but I'm sure you'll explain it better. Is it a mixed reality kind of visor
: Exactly. I mixed reality so you can see through it so you can see the real world at the same time. You can put computer images in front of your eyes are in place in the real world. So actually you know like holograms if you put them on you'll see a glowing object somewhere in your environment. And the stuff that I was working on with my other researchers was to get the sound to come from that exact location. So if you had a bird or something somewhere and it was chirping you could walk around it and this will be coming directly from that exact spot. So you know they needed something that would have spatial audio you need if you can see the hologram easy with a visor. But if there is an object behind you or on the side you can't see it but you need you have your attention drawn to it. The only way to do that is with sound and that has to sound like it's really coming from that location so you can turn your head to view it. So we were able to achieve that
: How would someone actually create sounds in certain areas around the head is the fall. I don't know. Game developers will people developing for this software for Hollo lens do they have an editor where they can place objects
: Yes. It depends on the kit you're using like there's unity plug ins for example so you can build your app in unity and you can also assign sounds to your objects in unity in 3D space and the sound will just come from there automatically
: That's brilliant. Other other possibilities I mean could you could recreate a whole musical composition in this kind of sound.
: Well you could I mean I've experimented with taking some say 20 tracks and then just placing the objects in 3D space just to hear how it sounds. You know as if you are like in the middle of the band playing and you know and just setting it up pretty quickly. It does have some really nice. It sounds pretty cool. You know cause you can like walk around and you know sit next to the bass player or stand next to the vocalist and it sounds pretty good. So there's a lot of possibility there. And again folks out there could get the whole Enzio devolvement get or the windows mixed reality. You can start building these things start placing audition 3d space. Put music out there.
: That's great. So some of your current work. What about something you'd like to work on in the future. Maybe a really cool project that you really want to get involved in. Coming up in your future what would you look towards.
: Now is more work going on with machine learning and neural networks and so on. So it seems to be still plenty of applications where that could be quite useful for audio. One thing's you know we've been experimenting with just really moving certain background noises and you know if you move one sound but not another or just clean up speech better you can use a neuron to train. I know the background sounds or to recognize sounds in the background. One project we worked with worked for someone last summer on her that was they were using machine learning just to identify sounds of the environment for people who are deaf. So if a fire alarm or fire truck is making a sound or you know there is a dog barking outside or someone knocking on the door their phone will alert them to all the sounds that are going on in their environment. All of them with machine learning. So it just seems to be a lot of possibilities going that direction. So the latest thing I also heard he made it to Google was creating basically music from scratch you give it a lot of examples of music and Maniel just started generating this music of that style or something like that just out of its own like a deep dream. But for music and some of those examples are sounds quite promising. But they take a whole lot of compute power
: Bad but really really interesting projects for the future. Like you mentioned machine learning a big part of the future. So do you see artificial intelligence machine learning becoming more and more involved in the audio creation of the future and perhaps maybe not replacing producers because I think that worries that but maybe taking away the boring job like the like you say like cleaning up dialogue. Maybe. I mean the biggest thing I hear every single day is how do I make my voice sound good. What what do I apply. How do I compress. So it sounds good. When I compress it it brings the breaths up too much. Do you think artificial intelligence will be able to help us with things like that and maybe Mull back to the interview in a moment but if you want to in my perfect audio create SSL head over to a dot fm slash when.
: Yeah I think you'd be able to just work with it hand in hand. Probably won't replace anybody on us it's a really mundane kind of work but you can work with the person using it. So for example let's say you have a radio announcer and you really like the way they sound while you should be able to then analyze the way that sounds. And then with artificial intelligence. Build a model of that and then maybe apply it to your own voice or any other any other sound to make it sound like the thing you wanted. So anyway there are some possibilities there even if you have like a really cheap microphone you could synthesize all the other harmonics and everything you need to make it sound like you have a really good microphone. Perhaps someone has to work on that.
: A fifty dollar microphone suddenly sounds like a Norman TLM I'm 103 and you're like oh wow that's interesting. OK we're getting into interesting territory here. I assume you have seen the adobe Voko project. I'd love to know where you can record. I think it's 20 minutes worth of speech of anyone so somebody could record a speech of a celebrity speaking and then this is just for anyone listening who's not aware of the projects and then apparently you can then make them voice say anything just by typing text into a kind of word processor. First of all what are your thoughts when you see that going. Is it good. Is it dangerous. I'd love to know what you think.
: Yeah I actually heard about that last year myself and listen to some of the examples. And that's you know definitely machine learning where you're just learning all the characteristics of one person's voice. You know if you know that person says in a this way and so on maps it all out and then the text to speech. It sounds like the person's voice but you know currently you can a person can still tell it's computer speaking in person. You know there's still work to be done there. It could get kind of a kind of scary overall they might need to put some watermarking in there if it gets too perfect so that and software that does this you can control under frequency analysis or something like that you know that it was actually synthesizing not the real person. When you have to think about how to do all that.
: Find some way because you have photoshop definitely changed the landscape Photoshop with Babel to photoshop an image. So yeah the same thing coming to Orio is as mindblowing
: It's basically exactly the same thing. Now you have to deal that all over again with audio
: So you look at something like the Alexa or google home and the Siri even they have voices and I I assume from what I understand I may be wrong. A voice artist initially records the kind of voice profile footprint and then the assistance voice is based off a real human voice is that what you understand you know much about that particular.
: Yeah that's pretty much what I understand in fact I hear that if you want to. There are ways you can just go in and do that with your own voice going for a few hours of recording and you can have it you know scenery be you I guess but I'm not sure of the process. I'm sure some ticklers got up but it would be really handy if there's a person who is said to have a disability where they lost their voice and if they have enough recordings of their voice they can actually synthesize or actual voice back. And so I do see some pretty cool uses for this technology as well
: This is great. I'm loving this. I'm actually making notes as I speak. Give your take and your knowledge your broad knowledge on audio is just amazing. So this is certainly going to be a replay episode for those listening to the show so okay let's go in a different direction now. Let's go back to your childhood David and look at your. Maybe one of your first memories around audio that made you smile and kind of got you started on this journey obviously then you got to the Mac and the editing software and thought I would like software that came along for Windows but way before that. Well it was really kind of made the fire inside you light up when you heard audio what was that for you.
: I think I was 7 or 8 my grandmother got me when I was cassette recorders like a RadioShack cassette recorder. And it was I started like pretending I was being an announcer and you know I'd get a book and start reading the book. And so that's a fun tape you go back and listen to when you're. But then I got to find out that if I put a little piece of foil over the erase turn up turned out to be the erase head on there. I couldn't start doubling my voice and tripling it I give out like 10 of me talking at once. That's kind of fun. And then I open up the back of the recorder and I found out if I put my thumb on the on the belt I could slow it down and record it that way then I play back it comes out really fast. I can't control you know the pitch for watching and recording that way. So I even soldered a little post in there. Slow run and a slower speed do some recordings and then undo it and play it back and anyway me and my friends will just have a blast. You know just pretending to do commercials and things like that and then play it back all really high pitched him.
: And again I even found that tape to send it to my friends and they had a fun time listening to it again. So it just started without being able to manipulate audio as a lot of fun. And in high school I was the kid of race where he was in the corner playing on the computer when no one else knew what a computer was. And I'd also build synthesizers and things from scratch. So it turned out you know I built a nice sequencer which I used to play like Pac Man music you can program in due to Type noises and then play back that a guitar amplifier there so we go to like a basketball game in high school I plug this into the guitar amp and I get the Waka Waka Waka brewery. All those noises and so on and and that was a lot of fun until you know that the power went out on my thing and had to reprogram within like one minute or so to get the sound effects in again. But it got me on the band bus and everything so I got to go on these long trips and so on from
: That's great. So you were doing cool effects before cool existed.
: Not just in the hardware alone you know
: That's amazing. So interesting features do you see today existing in software like Adobe Audition and other software that you never could have imagined back in the late 80s and 90s what really blows your mind today in audio technology.
: Man. Most just about everything I does today would have taken hours to do just a few seconds worth of audio back then. So it's all so amazing. I look at it sometimes and you know I just miss being able to work live in the spectrum of you know that would just be unheard of. You know in the and say like in 94 95 and 96 no on that work kind of stuff is. Yeah wasn't even invented back then. So just the amazing things you can get with that. I'm just looking at a copy of Just Like audition 3 and CC in front of me here. Yeah the restoration tools that we have now are just way ahead of what used to be there. And actually at work at Microsoft I do a lot of work with 64 channel audio so just like one recording 64 channels to get the full sound from all directions and just being able to manipulate that playback in real time and apply and HRT to it so a person listening to it sounds like the right the place where the recording was made and they can't tell the difference as between if they're are actually there live or they're listening to the recording. Things like that and just couldn't have even imagined. I did have a small HRT half module that I don't think we ever shipped but I had it running to experiment with. And you know I can kind of get sounds to sound like they're coming from your left or above your head and things like that. But it wasn't wasn't quite perfect you know. And 20 years later doing the work for hollow ends it's the exact same thing it's the age art which allows you to externalize the Aveo outside your head. And but now it actually works. And all the science behind it you know makes it work perfectly. So I would never imagine that having that before.
: So some some fantastic features that exist today. Do you have any features or any audio technologies that you'd like to see in the future that maybe don't exist that you dream about.
: There's the maybe just applying one music style to another recording. So let's say I have a sample of a bunch of symphony music now using that for the training. Now I just want to go whistle a song and then come back out turn it back into a symphony based on that whistle. Something like that. So it's like the with images you can do neural style transfer. So this would be like an audio style transfer be pretty cool and that requires a lot of work to do that
: Definitely. My goodness this is a really inspiring conversation. So OK so as we get towards the end of the podcast I'm going to ask you for a little bit of advice to people who are working or aspiring to work in the industry. But just before I get to that I'd like to know what's the next great project that you're working on right now so you are heavily involved in Hawler lens. Anything else kind of on the horizon or that you're working on right now.
: You know it's it's essentially that kind of work you know. Spaceflight is a big part of it. Like I mentioned the 64 channel recording to capture all Southfields and then playing that back and just other ways just more work and getting things is more realistic you know where you are so hollow lands let's say you're using a hollow ends in a small room. You know you'd want the objects to sound like in a small room or says whether you're outdoors you want the object to sound like you're in a larger space or just things like that trying to figure out how to get all the math to work behind that. And there's a lot more to be done. No not everything has been invented just yet. When it comes to audio.
: Indeed Absolutely. And on an early episode I was speaking to Maxim Jayco who's very into virtual reality mixed reality and he mentioned the term I'd never heard of before called real reality show. You've heard of the term and so you will be the audio guy when we eventually get into our real reality. You will be the guy writing the code for that right.
: I hope I have some hand in it that sounds like fun.
: Definitely. Awesome. So exciting. All right so a lot of people listening to the show right now. Let's talk to that one person right now who is aspiring to be an audio created to get into the audio industry in some form or another with your experience over the years. Creating cool at it starting up your own company in 1995 eventually selling that company to Adobe in 2003. Working with the audition team until 2010 and of course working I believe also Google Chrome OS and now your current work at Microsoft for Hollo lens and windows. So take all of that knowledge and try and condensed it down into a single piece of advice if you can and give it to the listener right now who's aspiring to be in this industry what would you say to them.
: Well basically look at what is you're doing and when you're doing your work what is it your enjoy doing and see if that you can apply that to we work on. For instance when I was doing cool at it it was all just done on my spare time because I actually had an actual daytime job and at Microsoft. But it was just so exciting. I just kept at it and if you really have a passion for what you're working on you know nothing's going to stop you you're going to be able to go out further than you expect. And I think it also it also shows when other people see the work you do they can see the amount of passion and effort you put into it. And in the end even if you don't get rewards rewarded right away just keep at it. You know if you're doing a podcast and no one's listening right away it just just keep doing it. If you're having fun and they'll come
: Really really solid advice I like that. David thank you so much for joining me and for anybody who's interested now to find you online. What is the best place for them to go.
: You can just email me at my zen dog software dot com. So just do David at Sundog software dot com. So that's a side company I did to sell some of the other programs we had it's in Trillium after we sell to Adobe.
: Well David I really appreciate your time today. Thank you so much for joining me on the
: All right.
: Sure. Nevada has a lot of fun. Can't wait to listen to this episode.
: That concludes this episode would you like an extra chance to win the awesome audio gear giveaway. Subscribe and review this podcast. Then melody tells the podcast at MRC Dotts FM for an extra entry into the awesome Olio giveaway. Good luck.