Hadrian, (and copied to the list in case someone else can still help) I've written up detailed instructions for the process that I use for this. (Attached). Have a look, and see what you think might be missing and/or poorly explained.
The process is fairly simple, and should take roughly 2-5 minutes per room/day, and then an additional 2-5 minutes per session. So, for a 7 session day, it might take as much as a half-hour, but I think it's likely to be a lot faster once you get the hang of the process. --Rich On 05/05/2016 11:27 AM, Hadrian Zbarcea wrote: > Hi Rich, > > I will be in Vancouver starting the morning of the 11th. I can help and > I think I can provide (at least temporary) storage for the data and some > processing time. I am estimating some 75M/hr for audio and somewhere > under 1G/hr for video. I may bring a 4k camera, not sure how much that > uses per hr. > > I would like to use this opportunity as an experiment to see if the > process could be crowd sourced and automated for future events. If we > could find some 15 mins to chat to educate me on what worked in the past > it'd be appreciated. > > Cheers, > Hadrian > > > On 05/05/2016 08:29 AM, Rich Bowen wrote: >> As some of you know, we do not have a sponsor to handle audio/video >> recording at ApacheCon this year. Thus, we'll be doing it ourselves - >> just audio, not video. >> >> So, I need volunteers for the following things: >> >> If you are willing to video presentations and put them on YouTube, >> please step up and do this. Keynotes, in particular, would be >> appreciated. >> >> More importantly, I'll be dealing with 8 tracks x 9 hours x 5 days of >> audio that will need to be edited. This isn't complicated, it just >> requires: >> >> * Computer with Audacity ( http://www.audacityteam.org/ ) installed, and >> your own headphones >> * Lots of patience >> >> The process is: >> >> * Dump huge files off of SD cards at the end of each day >> * Split large recording into sessions >> * Strip off leading, trailing noise >> * Encode to mp3 >> * Upload to feathercast.apache.org >> * Write brief boilerplate blog post (Name, Session title, Abstract) >> >> If you are willing to do this for even an hour, please let me know. I >> will be doing this each evening, for the duration of the event, and >> probably quite a bit during the days. If you just drop by the hackathon >> area, and help me out for even a brief time, I'll be very grateful. >> >> Thanks. >> >> --Rich >> >> (Looking towards next event, if you or your company want to sponsor >> audio/video for the upcoming ApacheCon EU, your company will get a lot >> of thanks and retweets out of this.) >> -- Rich Bowen - rbo...@rcbowen.com - @rbowen http://apachecon.com/ - @apachecon
COPY THIS FILE TO YOUR DESKTOP FOR REFERENCE 0) Prerequisites Audacity: http://www.audacityteam.org/ (If you use a different audio editor, you're on your own for step 2, but please follow file naming coventions anyways.) Laptop: I'm assuming that you're bringing your own computer. It doesn't matter what OS. Everything here should work on any OS. 1) SAVING FILES: Rich has a 1TB USB drive, henceforth referred to as PASSPORT, since that's the drive label. Plug in the PASSPORT drive to one USB port, and the SONY recorder to another. Switch on the SONY recorder. In a moment, both drives should mount. Files are typically in /VOICE/FOLDER01/ Please each audio file in the directory corresponding with the day it was recorded. Name it in such a way that it will be obvious which track/room it was. Follow the pattern of those that went before you. Once the file has been copied, and you've verified that the two files are the same size, DELETE the file from the recorder, unmount the recorder, and hand it back to Rich or Nick. 2) PROCESSING FILES: Copy the file to your local machine. On the PASSPORT drive, move the file into the WORKING directory so that nobody else tries to work on it. Open Audacity Open the file you're going to work on. Identify the breaks. This may be done visually, if the room was quiet between sessions, but usually you'll have to listen to these points to identify the break. Set a BOOKMARK in the middle of each break. This is done by placing the cursor where you want it, and pressing CTRL-B. (NOTE: This is the Fedora default. If it's different on Windows/Mac, please let me know so that I can update this doc!) Export these tracks with File -> Export Multiple (shift-ctrl-L). In the dialog that comes up, make sure that you have selected: A sensible target directory Format: MP3 Split files based on: Labels Check "Include audio before first label" Name files: Check "Using label/track name" This will result in a bunch of dialogs on which you will just press OK on each, and the result will be N files, one per track, in the target directory. Each track will have the session number appended to the filename, which should make the next step easier. 3) PROCESSING TRACKS Open each track in Audacity. Find the beginning of the talk. Select and delete everything before this. If the room is particularly noisy, select the first 5 seconds of audio before the speaker starts and do Effect -> Fade In to make the start less abrupt. Find the end of the talk. This usually is applause, but there may be questions afterwards that we'll want to capture. (Although these can sometimes be inaudible.) As with the start, delete everything past the end, and fade out (Effect -> Fade Out) the last 3-5 seconds of applause. Export the file, (File -> Export audio). Name the file: day_##_speaker.mp3 For example, the third session on Monday, with a speaker named Shaposhnik should be named monday_03_shaposhnik.mp3 Consult the conference schedule on the website to determine the session number and the speaker name. When determining session numbers, just go by the track number that is appended to the filename you started with. In the ID3 tags dialog, if you want to, Artist Name is the speaker of the session, Album title is "ApacheCon NA 2016" and year is "2016". Back on the PASSPORT drive, place these finished files into the DONE directory under the appropriate day. MOVE the file you started with out of the WORKING directory into the DONE directory. PLEASE ENSURE THAT YOU DO NOT OVERWRITE ANY EXISTING FILES. In the case of naming conflicts, just make the difference obvious, or ask someone.
signature.asc
Description: OpenPGP digital signature
