DAWとかVSTとか

Slate Digitalから、長々としたメールが届きました。約束の日を過ぎても、VCCとFG-Xのアップデート(AAX対応版?)がまだリリースできていないことに対するメールです。その半分以上(というかほとんど)は、なぜ現状に陥ったか、説明と言うか言い訳がつづられています。抜粋すると、

When the framework was complete, we had to decide where to start. So let’s just get this out of the way and say it was an obvious choice with VBC and VTM. VBC was written INTO the new framework. Converting it to AAX would be the easiest and more straight forward. Than came the VTM. This was in our old framework and had an insanely complex algorithm, but several of our very talented team members were able to get the port done in the Winter. And now we are left with the VCC and FG-X. Oh boy. Here it goes.

(前略) フレームワークが完成したとき、どこから手をつけるか決めなくてはならなかった。明らかにVBCとVTMが最初の選択肢であった。VBCは、新しいフレームワークの上で書かれた。AAXにコンバートするのは簡単で、単純な仕事だった。次はVTMだった。古いフレームワークで書かれていて、とてつもなく複雑なアルゴリズムであったが、チームの才能あふれるメンバー達が冬のうちに移植を終わらせることができた。そして、VCCとFG-Xが残された。

I think it’s fair to say that the algorithms, meaning the audio processes, of the VCC and FG-X are some of the best out there for what they aim to do. Well, at least I hope you think that since you bought at least one of them. These were the very first two plugins that Fabrice and I worked on and we’re proud of them. But as great as these algorithms sound, the way the code was written back then had a bit to be desired. Ok, they had a lot to be desired. Alright fine, they’re a mess.

VCCとFG-Xのアルゴリズムは、つまり音声信号処理は、その目的とする領分において最高であるといって良いと、自分は思っている。少なくとも、あなたはそう思っていると望んでいる(なぜなら少なくともVCCかFG-Xのどちらかをあなたは買ってくれたのだから)。この二つのプラグインは、Fabriceと自分が初めて一緒に作ったもので、我々はとても誇りに思っている。しかし、アルゴリズムがどんなにすばらしく鳴ったとしても、中身のコードの書かれ方には少々問題があった。OK、少々ではなく、大きな問題があった。いや認めよう、コードはぐちゃぐちゃだった。

There was no way we could use the same code for these plugins in our new framework, because we owe it to you, the customer, to improve them. With a cleaner code, they would be bug free, and use dramatically less CPU. Furthermore, there are aspects of our new framework that the VCC and FG-X code is not 1:1 compatible with so it would have taken just as long to convert the old code into the new framework. So the choice was obvious. We started last year to rewrite FG-X and VCC with fresh code, new features, and new algorithms. But again, going back to my first points, the team was overloaded. There were members bouncing back and forth to all the projects… one week it was VMR, then VTM AAX update, then VBC bug fixing.. it was insane. And to top it all off, we parted ways with two of our team members in the Spring, making a bad situation even worse.

これらのプラグインに対して、同じコードを再利用するという選択肢はあり得なかった。我々は、顧客のため、製品を改良する義務があるからだ。もっときれいなコードなら、バグフリーで、劇的に少ないCPU負荷になるはずだ。さらに、VCCとFG-Xのコードと新しいフレームワークは1:1に対応しないため、古いコードを移植するのは(書き直しと)同じぐらいの作業量になると思われた。我々は、昨年からFG-XとVCCを新規のコード、新しい機能、新しいアルゴリズムで書き直し始めた。しかし、最初のポイントに戻るが、チームは過負荷状態だった。メンバーは、さまざまなプロジェクトの間でいったりきたりしなくてはならず(後略)

———————–

というわけで、FG-Xのコードはぐちゃぐちゃであった、と御大自らがお認めになられました。しかも、一から書き直した方が早いぐらいのどうしようもなさ、という。DAWホストを巻き込んでメモリを破壊する明白なバグを直せなくて、サポートは開き直って仕様だと言い張っていたからなー。さもありなむ。

メールの続きには、どうやってこの難しい局面に対処するかと書いていますが、省略。

何というか、IT業界に身をおくものとして、割とよく聞く類の話ではありますが、ここまでぶっちゃけるのも珍しい。
だんまりになった末トンズラされるよりはよほどましですが、道のりは厳しそうです。FG-XよりはよほどクリーンなはずのSD4のリリース時も相当トラブルが後を引いたしなぁ。ウザいくらいの熱意はありそうだけれど技術力に不安を覚えるのは自分だけでしょうか。

いずれにせよ、今後のSlate Digitalの動向からは目が離せないようです。(ときれいにまとめてみるw)

——————-

メールの全文はこちらになります。

Dear FG-X and VCC Slate Customer,

This is one of the hardest emails I’ve ever had to write. The reason for this, is because I really love what I do. But one of the main reasons I love what I do is because I get to make products that make people happy. But clearly, there are a lot of you out there who are quite unhappy, due to the missed deadlines of our VCC and FG-X updates.

First, and this goes without saying, I feel terrible about this. I never intended for this to happen, and I read all the negative comments, and it breaks my heart. Because I failed you guys.

This company means the world to me, and I have always strived to make mindful decisions that would ensure that it remains healthy. But unfortunately, despite the fact that I am often seen around Hollywood Blvd wearing a cape, I have no super powers and am just a mere mortal who makes mortal mistakes. And clearly, I’ve made mistakes this past year. Let me tell you about them.

As many of you know, we work a long time on our products. Two years ago when AAX was announced, we were just beginning the development of some pretty intense new products, which are all as of this email, unreleased. Two of them you know about, the Virtual Mix Rack, and the Virtual Microphone System. Both of these were new and exciting products that required a lot of effort from the entire team, and so we began to start the R&D. And then, we get hit with AAX.

We were not prepared, nor did we have a big enough team to simultaneously develop big new products while updating old products. My first mistake was realizing that fact too late. What should have happened is a pause in all production to hire more developers to help with our new 64bit AAX framework, pause all new development, and then proceed with full force on all of our legacy updates. Instead, I became fearful. This fear was derived from the fact that we are one of the newer games in town as compared to our competition, and I felt like new product development was crucial enough that we had to keep some part of the team on it while another part worked on the framework and updates. This compromised the productivity of both teams since neither side had enough manpower.

When the framework was complete, we had to decide where to start. So let’s just get this out of the way and say it was an obvious choice with VBC and VTM. VBC was written INTO the new framework. Converting it to AAX would be the easiest and more straight forward. Than came the VTM. This was in our old framework and had an insanely complex algorithm, but several of our very talented team members were able to get the port done in the Winter. And now we are left with the VCC and FG-X. Oh boy. Here it goes.

I think it’s fair to say that the algorithms, meaning the audio processes, of the VCC and FG-X are some of the best out there for what they aim to do. Well, at least I hope you think that since you bought at least one of them. These were the very first two plugins that Fabrice and I worked on and we’re proud of them. But as great as these algorithms sound, the way the code was written back then had a bit to be desired. Ok, they had a lot to be desired. Alright fine, they’re a mess.

There was no way we could use the same code for these plugins in our new framework, because we owe it to you, the customer, to improve them. With a cleaner code, they would be bug free, and use dramatically less CPU. Furthermore, there are aspects of our new framework that the VCC and FG-X code is not 1:1 compatible with so it would have taken just as long to convert the old code into the new framework. So the choice was obvious. We started last year to rewrite FG-X and VCC with fresh code, new features, and new algorithms. But again, going back to my first points, the team was overloaded. There were members bouncing back and forth to all the projects… one week it was VMR, then VTM AAX update, then VBC bug fixing.. it was insane. And to top it all off, we parted ways with two of our team members in the Spring, making a bad situation even worse.

I do NOT expect sympathy from you. I do NOT consider any of this an excuse. I will stand here in front of you and say we messed up. But rather than drown in our sorrows, I’m going to tell you what we are doing to make things better.

First and foremost, with the exception of the few members finalizing VMR, the entire rest of the team is focused on FG-X and VCC. We will do NO OTHER DEVELOPMENT ON ANY PRODUCT until these two AAX updates are done. Second, we are hiring hiring hiring. We are going to put ourself in a position to be able to execute faster, more efficiently, and we’ll be able to meet deadlines and schedules.

So some of you are saying “Enough of the drama, when the hell are we getting our plugins?”. And just in case you didn’t hate me enough already… I can’t give a date YET. Because unlike last time, I will not give you a date until I’m ready to put my life on the line because I’m so sure we’ll make it. But what I am going to do is make you a promise: You will have an inside look at the productivity of these two projects with weekly emails of information. And I also promise that as soon as I have the confidence to give you a release date that I know is a reality, I will state it in one of these weekly updates.

Last, I’ll say that we are working on a possible temporary solution that could get you guys up and running in Pro Tools 11 while we finalize VCC 2.0 and FG-X 2.0. And to end on a positive note, these 2.0 updates will be amazing. Please keep that in mind. With the VCC, it’s goiong to be simpler, have new options, and include a new console emulation that I predict will be your new favorite and most used console emulation yet. The FG-X will have a new metering algorithm but that will likely get overshadowed by the new FG-LEVEL algo. This new algo has been over three years in the making and utilizes a brand new loudness process that is so ridiculously transparent that you’ll think it’s defying the laws of physics.

Ok, so that’s all I have for you today, but you’ll be hearing from me again soon. I know this is not the email you really want, but I hope you’d agree it’s better than silence. You can email me at slate@stevenslate.com.  It might take me a bit to get back to you because it’s so busy here, but I promise I will. Have a great weekend,

Steven Slate

広告

コメントを残す

以下に詳細を記入するか、アイコンをクリックしてログインしてください。

WordPress.com ロゴ

WordPress.com アカウントを使ってコメントしています。 ログアウト / 変更 )

Twitter 画像

Twitter アカウントを使ってコメントしています。 ログアウト / 変更 )

Facebook の写真

Facebook アカウントを使ってコメントしています。 ログアウト / 変更 )

Google+ フォト

Google+ アカウントを使ってコメントしています。 ログアウト / 変更 )

%s と連携中

%d人のブロガーが「いいね」をつけました。