TUAW Interview: Andrew Welch on WireTap Studio

After our interview about iToner last week, Ambrosia Software president Andrew Welch sat down with us again to talk about his company's newest product, WireTap Studio, which was released yesterday. WireTap Studio will supercede WireTap Pro allowing users not only to record any audio source on their Macs, but also to edit that recording in a lossless manner. As you can probably tell from the interview, Andrew is clearly excited about the product and expects it to make a big splash in the market.

What kind of user did you have in mind for WTS?

It's actually interesting that you lead in with that question; because it's the same question I asked myself when we started designing WireTap Studio.

What we did with WireTap Studio was take a step back, and think of what people actually do with audio. What we found is that there are essentially two kinds of people when it comes to audio: consumers and producers. Consumers listen to audio (music, podcasts, whatever). Producers make it. What we found in interviewing people was that the tools they were using were either far too simplistic, such as recording podcasts directly to MP3 format using QuickTime Pro, or they were using very very high end programs, but only using a very limited subset of the functionality that those programs offered.

So the answer to your question is that we went for the sweet spot right in the middle. We went for an easy to use product that does 90% of what people do with audio, and does it very, very well.

So what kind of particular challenges did you run into while developing WTS?

I just had a conversation about this last night with a developer friend of mine. We concluded that very few pieces of software are actually designed. Most software is just written. The distinction is important - when you design something, you look at the big picture, from a user-centric point of view. You try to build solutions that work well enough that the tool disappears, and the user can focus on the work they are trying to accomplish.

So the most difficult challenges we faced were really challenging ourselves. Challenging ourselves to rethink the way an audio recording/editing program should work. We spent an incredible amount of time thinking about how the whole process should work.

Another difficulty was figuring out what to cut OUT of the program. It's pretty easy to write a program - and I'm sure you've seen them - that have options/preferences for everything under the sun. To me, this is a cop-out. What it means is that the developer couldn't figure out how to design the software, so they punted and left the decision up to the user. We tried very hard to make our software lean and mean so it does what people really want, without a lot of confusing junk thrown in.

So the most ground-breaking thing about WTS seems to be its lossless nature. This means that every recording it makes includes a high-quality master behind the scenes, right? What is the quality of the master recording? How much storage space is it going to require?

I think the most ground-breaking thing about WireTap Studio is the way the whole program works together. I can't tell you how many people I've shown WireTap Studio to, and had them say to me "Ah! Of COURSE! That's the way it should be done? Why hasn't anyone made something like this before?"

Obviously this is a bit of an over-dramatization, this is no cure for cancer, but I think it's really important to look at the program as a whole to see why it's innovative. For instance, why in every audio program under the sun do I have to choose what compressor, bit rate, khz rate, etc. my audio will be before I even record it? This would be akin to wanting to create an image for a web page in Photoshop, and Photoshop doesn't let you even paint the first brush stroke until you tell it what file format and compression level you want for the image. It makes no sense.

So the approach we took is the same approach that recording studios use. You record a master version of the audio, and from that master version can make as many compressed/edited copies as you like. But we did it in a transparent way, so the user has no idea of the magic that's going on behind the scenes.

Here's a perfect example of why this is important. Let's say you record a piece of audio for your podcast, and you decide that the quality is too low. You have to options: you either throw out the recording and do it over again, or you "ship it anyway" and delivery crappy sounding audio. With WireTap Studio, it's no problem. You tell us what format you want to record in, and we'll indeed record in that format... but you can change your mind after the fact to upgrade the quality of the audio recording, or even do something like change a mono recording into a stereo recording, while actually gaining quality.

The decision to use the concept of a lossless master recording also allows us to do ground-breaking things like our LivePreview function [which is patent pending]. This allows you to listen to your audio, and on the fly switch from one compression level to another, or even to different file formats, and instantly hear the difference in the sound quality and size of the recording. It floors people when they see it, because it's something they've longed for, but never knew it!

But if every recording I make is in such high quality that's inevitably going eat up lots hard drive space, etc. That's why we use compression codecs in the first place.

We don't record everything lossless; it's actually compressed but in a lossless manner, so we aren't talking about raw audio files here (which can indeed be huge). But I think the disk space is largely a non-issue. Disk space is incredibly cheap, and is growing cheaper and larger every day. What we don't have more of is time. I can tell you for certain that if WireTap Studio saves your hour long recording from going into the trash even once, you'll find the way our lossless master recording works very, very well worth it.

However, if you really don't want the feature on, you can turn it off - we let you record directly to any compressed file format as well. In fact during a recording, we do a dual encode. We record into the file format you say you want, and we also record into the lossless master recording.

We do it this way so that you can instantly do something with your recording after you're done, without waiting. It's ready to go be uploaded as a podcast or what have you. If you make changes to the file by cropping it, changing the format, etc., we transparently update the compressed version in the background so it's always ready to go and you don't have to wait for it to encode.

So if I want the absolute highest quality output I can get from WTS (for instance to import into a high-end editor) what's that going to be?

That depends on your input source! What WireTap Studio does is it queries the device you're recording from, and it figures out what the highest quality that device or application can offer in terms of audio. And that's what it records at. We don't waste space by over-recording... if a device only supports 16 bit audio, that's all we'll ever record from it, for instance.

So are all the edits non-destructive? How does this work? How are these edits stored (some kind of metadata, etc.)?

A good way to think of it is that the master recording is your pristine original source; any edits that you make to the master recording can be thought of as the "recipe" that creates the final compressed audio file. That recipe can include normalization, any number of audio unit effects, cropping the audio, silencing portions of the audio, deleting portions of the audio, fades, etc. Lossless original + edits = final compressed version - and the great thing is not just that you can do this, it's the way the editor works.

The editor is so natural it's a joy to use. Want to crop some audio? Just drag to crop it! Decide you cropped too much? Just drag it back. It's editing without fear, because there's nothing you can to do damage your original pristine master recording.

Do you have any comment on WTS with regards to Rogue Amoeba's competing products, Audio Hijack Pro and Fission?

I think Fission is a useful niche product; if you have already compressed audio that you need to edit, then it (and MP3 Trimmer) are useful tools. And I think it will always be a useful tool. However I think we took a look at things from a more interesting perspective, and have ended up with a much more compelling solution for capturing, storing, editing, and distributing audio.

I am so pleased with how WireTap Studio turned out - even the caveat that it's a 1.0 product - that I would love for people to download it and play with it side by side with any competing products. I'm really confident people are going to love WireTap Studio. We've been showing the product to people for some time, and most recently at the Podcast and New Media Expo - and this was the kind of reaction people had to it. They were smitten.

What is the relationship of WTS to WireTap Pro? Does it simply replace it or will the two products co-exist? Will users have an upgrade path?

WireTap Studio replaces WireTap Pro completely. We're offering a discounted upgrade path to existing WireTap Pro customers. WireTap Pro cost $19... WireTap Studio retails for $69, or a difference of $50. But we're offering an upgrade to our existing registered users for just $30.

And I think it's going to be a no-brainer update for most people. We've made a product that's better in every way, but orders of magnitude, than WireTap Pro. Honestly, there's no comparison at all. The combination of LivePreview, the lossless recording technology, a fantastic built-in editor... it's a home run.

Anything else you'd like to add about WTS?

I like to think of WireTap Studio as the audio version if iPhoto. It makes recording and editing audio so easy that it will make it accessible to people who have previously been scared off of trying to do anything with audio. Our guiding principle in the design was to make it easy to use, but that actually caused us to end up with a very powerful program that uses new paradigms for editing audio. And new paradigms were needed. How many times can people redo "SoundEdit 16" over again?

Most of all, I'd say that people should see it in action to "get it" - have a look at some of our tutorial videos that we have made available here. Watch a few. You'll get it.

Ambrosia seems to do a lot of remarkably different things: games, Snapz, Dragster, WireTap, etc. In fact, it's hard to think of another Mac dev with a similar range of products. Is there some kind of organizing principle to the software you offer?

Dragster was actually never even meant to be a product. It's a funny story, so I'll relate it really quickly for you. We wanted an export framework for use with WireTap Studio (and other upcoming and unannounced products), so that the user could easily upload their audio clips to a remote server, to their iDisk, via email, etc, etc. Well, I decided that the only way we'd really get an export frame tested in a robust manner was to build a product around it. So Dragster was created as my little pet product for uploading files, and we ended up with a very robust exporting framework for use in WireTap Studio and our other products.

As for an overall focus, we work on software that we find interested. We are finding audio/visual media really interesting right now, and we do games because it's fun to make games now and again. It's a horrible thing as a business plan, but it does wonders for the morale of people working here that we try to work on products we really have faith in and enjoy working on.

Okay I have to ask. I still fondly remember spending much of middle-school computer lab time playing Maelstrom; it seems like the iconic Mac game of that era. What happened to it? Will it ever come back?

Maelstrom is close to my heart as well. I wrote it over the summer on break from college (I was a photojournalism major at RIT). I wrote it on my Mac ][si for two very good reasons:

1) Because I wanted to hone my skill in writing assembler (Maelstrom was almost entirely written in assembly code) 2) Because someone dared state that high speed full color animation couldn't be done on a Mac at the time. That pissed me off, and I was determined to prove 'em wrong.

My friends and I had a blast making the game, with many friends of mine contributing to the sound effects, and telling me what they thought was lame and what was fun, etc. I have very fond memories of that program, and the people who helped me make it. We had some fantastic artists in Mark and Ian working on it too.

As for what happened to it, that's a good story too. We had a fellow approach us years ago asking if he could use Maelstrom for his PHD thesis. He was writing something called Simple Direct Media Layer, to allow for easily writing cross-platform games. I took some sick perverted pleasure in saying "yes" to this gentleman, named Sam Lantinga, because I thought it would be amusing to see him try to grok the 68K assembler code. So he ended up writing SDL using Maelstrom as the basis for this doctoral thesis, SDL has gone on to be wildly popular for writing cross platform games (or often even single platform games), and Maelstrom is now Open Source and has been for some time.

Oh and by the way, he is now a lead developer for World of Warcraft, so I think things worked out well all the way around!

Thanks to Andrew and best of luck to Ambrosia on the release of WireTap Studio.