Jump to content

Jay Rose

  • Content count

  • Joined

  • Last visited

About Jay Rose

  • Rank
    Hero Member

Profile Information

  • Location
    Boston US
  • Interested in Sound for Picture
  • About
    Sound designer and industry author. Member CAS and AES. Humor, articles, and studio info at www.dplay.com.
  1. Re-edit of film years later using 5.1 stems

    Your plan makes sense, except I wouldn't totally throw out LFE on the stereo mix... there might be something important there. And there might be some level trimming necessary if the scene-to-scene flow has changed. The unanswered question, of course, is will the new edit disrupt any of those stems? I don't know what cutting they did, but music frequently gets mangled both inside and across scenes, and bgs can have abrupt shifts if there are edits within a scene. That might take some bandaids -- or at least, offsetting the edits -- before you mix.
  2. The "Best client/producer/agency quotes" thread

    Or brag that you know a pix editor with a new 'visual noise reducer' plug-in, and they should be using your friend instead of the editor they've booked.
  3. The "Best client/producer/agency quotes" thread

    VO for large financial firm, making their first foray into the Salt Lake City media market. Ad manager, straight arrow Boston financial type: "That's good, but can you make it sound more Mormon?"
  4. Content Jobs

    Forgive an old postie who stopped doing location mixing a long time ago: When did the term become 'content job'? A web search on that exact phrase suggests it refers to people who manage, edit, grab user-generated stuff from the web... but not a kind of shoot. (Without 'exact phrase' clicked, it gives me a bunch of job listings for things like web content creator.) Or is this a special kind of job that's supposed to give you contentment for little money, because it's "new media"...
  5. The "Best client/producer/agency quotes" thread

    "Can you make her sound like I love her?" see article...
  6. Sound sync advice using film 0.01%

    I won't address why you're shooting micro-budget on 16mm. Surely stock / lab / fxr costs are going to be significant. But that's your choice. (The last true low-budget indie I worked on that used film was in 2002... it was a very short piece, and the experienced videographer specifically wanted to shoot 35mm for the film experience.) So let's address the pull-up/ pull-down: This is necessary when 24 fps film gets transferred to video, which runs just a tiny bit slower than 30 fps. (The 24<>30 conversion introduces its own strangeness to motion, but at least that's integral so audio sample rates don't change. It's the 30 <> 29.97 fps that means you have to slow down the sound.) 1. It's not absolutely necessary that you compensate. If you're doing short takes, you can sync each one. The drift at the end of 30 seconds will barely be noticeable. If you then have cuts so individual camera angles are even shorter, you can nudge to compensate. I've worked for major post houses where, in the era when spots were shot S16 and finished on NTSC, they didn't even bother with pull-down. 2. Using FCP can be an issue. Years ago, FCP had a known, uh, feature where it would apply .01% changes whether you wanted it to or not, depending on the drop-frame settings. (Drop frame rates have nothing to do with pulldown. The frames are the same length. It's how those equal-length frames are numbered that makes the difference.) Many productions were brought up short -- with complaints of 'your mix is out of sync' -- until we got a handle on this. I don't know if it still exists in the version you're using. I also don't know if other NLEs have had this problem. FWIW, just about every professional DAW and most audio editors have a function to compensate for pull-up or -down. 3. If you do need to compensate, you don't have to do it during production. You can do it in post, before editing. 4. If you're going fully old-school, shooting film and then editing 16mm workprint against fullcoat, it's a don't-care. So long as the production track playback speed is the same as what you recorded when you transfer to fullcoat, you'll be fine.
  7. Neural Networks for Audio: how they work

    Mike, Do you mean it sorts character A's voice from character B's? Or that it sorts dialog from other noises like footsteps and bird calls (which are usually immune to conventional noise reduction)? Both are theoretically possible with NN, but I haven't heard of anyone doing the former. It would take a lot of training. The latter is commercially available in a few products now.
  8. iPhone 7 mic

    I picked up a Focusrite iTrack ($130) so I could use good mics with my phone and iPad. Works brilliantly... just be aware it has to be running before you launch the audio app.
  9. Neural Networks for Audio: how they work

    NewEndian, thanks for the link. That's incredible stuff. Off the top of my head, I suspect iZotope and Audionamix didn't use GAN because 1) it's bleeding edge and these products have been in the works for a year, 2) the infrastructure for commercial development -- like easily purchased AWS training -- isn't there yet (I'm sure it'll be available soon), 3) the challenges of time-variant audio are so different from the xy arrays of image processing, and 4) the immediate market for image manipulation is so much bigger than that for audio manipulation. Visual bias strikes again!
  10. 96khz to 48khz converter

    Sound Grinder is a fast and powerful batch conversion utility.
  11. Audionamix's TRAX Pro SP and the Dialog Isolate module in iZotope RX6 are kind of amazing: they use Neural Networks to clean up production tracks in ways we've never been able to before, and they can even give you a stem with the stuff they took away from dialog (like a clean bg track, or just the environmental music). Far better than any of the multiband expansion noise reducers or other algorithmic techniques we've been using for a couple of decades. They can also seriously screw up your track. Just like any other processing. Both manufacturers graciously gave me large chunks of Skype time with their senior developers, so I could write an article about the techniques for CAS Quarterly. The article appears online today, and will be mailed soon. We've also posted a web page where I've put downloadable samples of actual processing of production elements. (If you do the web page, please download the AIFFs. The posted mp3 files are just for orientation, and distort some of what the processors are doing.) Fellow CAS member Stephen Fitzmaurice added a sidebar with initial impressions of the Audionamix in his mix theater. Detailed reviews will be coming in a future issue. Article is in the Quarterly at cinemaudiosociety.org, or as a downloadable pdf at jayrose.com. This stuff has been blowing my mind. Please comment. (On the technique, not on my mind; that's a lost cause.)
  12. Another make-an-actor-say-anthing app

    Add me to that list. I first proposed it about a dozen years ago, in a DV Magazine column. Cleaning up bad production dialog is one thing. But these apps are (or soon will be) capable of making a convincing recording of anybody saying anything you want to type as an input. All you need is enough samples of their voice to use as training material. And as I reported about a month ago, a different app can take video of someone, and make an absolutely convincing new lipsync video of them saying a new audio input. Demos already online, and of course that's also still in Beta. "Oh brave new world, that has such [non existent] people in it!" -- Shakespeare
  13. Another make-an-actor-say-anthing app

    Bab414, Other artificial speech apps -- some I've posted here -- have controls for prosody and inflection. [I'm assuming you weren't being sarcastic in your post, but actually looking for information...] What hasn't been done ... or rather, what hasn't been published yet ... is training a neural net to generate those wrinkles automatically. It still has to be done by a human operator. But as soon as someone comes up with a training library that's properly keyed for these elements, it'll happen. The mass market is there, for digital assistants with an edge. And building the library won't be hard, as soon as someone develops a consumer app that provides a benefit for users tagging the subtext.
  14. Lyrebird.ai is a Canadian company doing AI creation of new speech from sample recordings. They have an online demo, where you record a minimum of one minute (they guide you through sentences, so the samples have a key), it runs the samples through a neural net, and then it'll create your voice saying anything you type. Pros: almost real-time, with a web interface. Cons: still somewhat artificial sounding, but a lot better than previous while-you-wait examples. ... and this is just a first-gen beta demo, with a really small training set and no ability to tweak the results. The company isn't posting anything about their technique, so I'm just guessing (from the operation and from the principals' bios) that it's NN. They seem to be interested in selling their sdk to other developers, rather than offering a service to filmmakers... but that's also just a guess.
  15. Reverb in exteriors

    It easily could have been. They were so far away, you couldn't see their mouths. Chances are likely they were recorded wild (I won't guess whether it was was during production or post, but post have certainly been cheaper). The verb was definitely added in post. I'd prefer to believe the director requested its wetness and wasn't happy until it sounded that way, because the rerecording mixer probably knew better.