2012 OLF audio: Difference between revisions

From WPLUG
Jump to navigation Jump to search
(add links to raw files, minor edits)
 
(8 intermediate revisions by 2 users not shown)
Line 1: Line 1:
This page is to coordinate activities for editing the audio files from the 2012 Ohio LinuxFest. It is divided into a section for each SD card, which contains a number of talks. To claim a particular talk or set of talks, download the raw audio file from the linked directory and then change the entry in the "Editor" column from "unclaimed" to your name/e-mail address. If it takes you more than a few days to work on them, e-mail [[User:Vance|Vance]], otherwise he may revert it back to "unclaimed" so someone else can pick it up.
This page is to coordinate activities for editing the audio files from the 2012 Ohio LinuxFest. It is divided into a section for each SD card, which contains a number of talks. To claim a particular talk or set of talks, download the raw audio file from the linked directory and then change the entry in the "Editor" column from "unclaimed" to your name/e-mail address. If it takes you more than a few days to work on them, e-mail [[User:Vance|Vance]], otherwise he may revert it back to "unclaimed" so someone else can pick it up.


Contact [[User:Vance|Vance]] for instructions and credentials for uploading the edited files.
More specific instructions on editing will appear at the end of this page. When you're done, name the resulting WAV file as ##.wav with the track number from the table. Compress the WAV files as FLACs (this saves upload time and storage space) and upload them to INSERT LOCATION.

More [[#Editing_Instructions|specific instructions on editing]] appear at the end of this page. When you're done, you should have a FLAC file named as ##.flac with the track number from the table. Then upload to the location you were given.


== Card 1 ==
== Card 1 ==
[http://pair.ohiolinux.org/rawaudio2012/card1/ Link to raw files]


=== Early Penguin Talks ===
=== Early Penguin Talks ===
Line 100: Line 103:


== Card 2 ==
== Card 2 ==
[http://pair.ohiolinux.org/rawaudio2012/card2/ Link to raw files]


=== Track 2 Talks ===
=== Track 2 Talks ===
Line 150: Line 154:


== Card 5 ==
== Card 5 ==
[http://pair.ohiolinux.org/rawaudio2012/card5/ Link to raw files]


=== Track 4 Talks ===
=== Track 4 Talks ===
Line 199: Line 204:


== Card 6 ==
== Card 6 ==
[http://pair.ohiolinux.org/rawaudio2012/card6/ Link to raw files]


=== Opening Keynote Talk ===
=== Opening Keynote Talk ===
Line 253: Line 259:


== Card 7 ==
== Card 7 ==
[http://pair.ohiolinux.org/rawaudio2012/card7/ Link to raw files]


=== Open Source Solutions Stage Talks ===
=== Open Source Solutions Stage Talks ===
Line 295: Line 302:


== Card 9 ==
== Card 9 ==
[http://pair.ohiolinux.org/rawaudio2012/card9/ Link to raw files]


=== Career Track Talks ===
=== Career Track Talks ===
Line 301: Line 309:
|-
|-
|61
|61
|Career Ownership
|The American Dream: Seven Steps to (Career) Ownership
|Janine Moon
|Janine Moon
|SR000F.flac / SR000R.flac
|SR000F.flac / SR000R.flac
Line 313: Line 321:
|-
|-
|63
|63
|2CheckOut
|Direct Hire Recruitment
|Alan Czako
|SR000F.flac / SR000R.flac
|unclaimed
|-
|64
|The Perspective of Recruiters
|Lou Russo and Rufus Smith
|Lou Russo and Rufus Smith
|SR001F.flac / SR001R.flac
|SR001F.flac / SR001R.flac
|unclaimed
|unclaimed
|-
|-
|64
|65
|LPI: Catching the Wave of Open Source Careers
|LPI: Catching the Wave of Open Source Careers
|Ross Brunson
|Ross Brunson
Line 324: Line 338:
|unclaimed
|unclaimed
|-
|-
|65
|66
|The Intentional Career
|The Intentional Career
|David Crone
|David Crone
Line 330: Line 344:
|unclaimed
|unclaimed
|-
|-
|66
|67
|Manta
|Manta
|Steven ?
|Steven ?
Line 336: Line 350:
|unclaimed
|unclaimed
|-
|-
|67
|68
|Rackspace
|Tricia Medina, Deborah Carter, Christine ?, ? Brown
|SR002F.flac / SR002R.flac
|unclaimed
|-
|69
|You Inc. - Taking Control of Your Career
|You Inc. - Taking Control of Your Career
|Ben Blanquera
|Ben Blanquera
Line 342: Line 362:
|unclaimed
|unclaimed
|}
|}


== Editing Instructions ==

Some general notes:
* Audacity works fine, and is what I'll describe. If you like something else better, use that.
* Take care not to change the sample rate from the original. Some files are 44.1 kHz, some are 48 kHz. Using '''File -> Import''' in Audacity will cause resampling to the project's rate. Use '''File -> Open''' instead to avoid this.
* The final product will be a separate 1-channel (mono) 16-bit PCM WAV file for each talk. Compress this into a FLAC file before uploading to save time and space.

There are two methods that were used for recording the talks. The Early Penguin and Career Track talks were recorded using the [http://www.zoom.co.jp/english/products/h2/ H2's] four built-in microphones. This results in two stereo files with the names SR###F.wav and SR###R.wav. The "F" comes from the front microphones and the "R" from the rear ones. The rest of the talks were taken directly from the PA soundboard via the line-in jack. These have only a single stereo file named STE-###.wav. All of these were compressed into .flac files before uploading.

=== Separating talks ===

Load the STE-###.flac or SR###F.flac file into Audacity with '''File -> Open'''. If applicable, also import the matching SR###R.flac file using '''File -> Import'''. Locate the beginning of the talk in question. There may be some random discussion, so find the place where the speaker "officially" starts their presentation. Highlight and delete everything prior to this point ('''Edit -> Delete'''). If you have more than one track on screen, make sure you hit all of them.

Next you locate the end of the talk. This is after the speaker has finished and any open question-and-answer period. Some recordings may capture post-talk conversations that individuals are having with the speaker. These should '''not''' be included as they are not intended to be public. Highlight everything from this point to the end of all tracks, and then delete.

=== Selecting audio channel ===

From the drop-down track menu (the downward-pointing triangle at the upper left of the track), select '''Split Stereo to Mono''' for each track. Now your job is to choose which of these sounds the best. You can isolate tracks using the '''Mute''' and '''Solo''' buttons on each track.

If you're working on a STE-###.flac file, there should really be no difference. If you find that one has more noise or lower signal, then delete that track by clicking the X in the upper left; otherwise, pick one at random to delete.

For the SR###F.flac / SR###R.flac files, you'll probably find that one of the channels is louder than the others. Pick whichever one seems to have the best signal to noise ratio, then delete the rest using the X in the upper left.

=== Normalizing ===

At this point you'll have just one single mono track left. Make sure you do not have a selection made, or alternatively select the entire track. Now you'll want to change the display to '''Waveform (dB)''' using the drop-down track menu (the triangle in the upper left of the track). You can zoom in by left-clicking on the scale or zoom out by right-clicking.

[[Image:Olfcompressor1.jpg]]

You should see something like the above (minus the red and green lines). The red line represents the noise floor: that is, the amount of sound present when no one is talking. For STE-###.flac files, you can probably just set this to -60 dB and not worry about it, since noise shouldn't be a problem there. For the SR###F/R.flac files, you'll have to listen and pick a reasonable value (take note that this can only be adjusted in 5 dB steps). Here we've chosen -55 dB.

The green line indicates the threshold value. This is the point at which the effect starts to tamp down the louder peaks so that the entire track has a more consistent volume level. Set this around the level of the most quiet parts of the speech. Here we're using -50 dB. Try not to set this too close to the noise floor; a 5 dB difference is probably as close as you want to be.

Now we select '''Effect -> Compressor''' from the menu and get a dialog like the below.

[[Image:Olfcompressor2.jpg]]

As discussed, we have set the threshold at -50 dB and the noise floor at -55 dB. Those two settings are the only ones that need to be changed from one speech to another. Make the ratio 10:1, attack time 0.1 secs, and decay time 1.0 secs. Check the box for '''Make-up gain''', and uncheck the one for '''Compress based on Peaks'''. Then click '''OK''' and wait a few minutes for it to work its magic.

[[Image:Olfcompressor3.jpg]]

When complete, you should get something like the above. The times where the speaker is talking should be no lower than -15 dB, and preferably -12 dB or greater. This example is quite noisy, so you don't see as much differentiation between periods of speech and periods of silence. Take a listen and if you're not happy, hit '''Edit -> Undo''' and try again with different settings.

=== Saving ===

Once you're happy with the way things sound, go to '''File -> Export'''. You can select "WAV signed 16 bit PCM" as the format, or if it's supported choose "FLAC" from within Audacity (make sure under "Options" you choose 16 bit). Name the file with the track number (e.g., 07.wav or 07.flac). WAV files can be compressed after the fact using the command line '''flac --best ##.wav''' which will create a file named ##.flac.

Latest revision as of 04:07, 4 November 2012

This page is to coordinate activities for editing the audio files from the 2012 Ohio LinuxFest. It is divided into a section for each SD card, which contains a number of talks. To claim a particular talk or set of talks, download the raw audio file from the linked directory and then change the entry in the "Editor" column from "unclaimed" to your name/e-mail address. If it takes you more than a few days to work on them, e-mail Vance, otherwise he may revert it back to "unclaimed" so someone else can pick it up.

Contact Vance for instructions and credentials for uploading the edited files.

More specific instructions on editing appear at the end of this page. When you're done, you should have a FLAC file named as ##.flac with the track number from the table. Then upload to the location you were given.

Card 1

Link to raw files

Early Penguin Talks

Track # Title Speaker Raw File(s) Editor
01 Freebase: Big Data and Archiving Kirk Kimmel SR000F.flac / SR000R.flac unclaimed
02 SSH The Easy Way Carl T. Miller SR000F.flac / SR000R.flac unclaimed
03 Physical Computing with Arduino and Linux Ethan Dicks SR001F.flac / SR001R.flac unclaimed
04 Computer Reach, Ghana, and Free Software Beth Lynn Eicher SR002F.flac / SR002R.flac unclaimed
05 Software Patents: What You Can Do Deb Nicholson SR002F.flac / SR002R.flac unclaimed
06 MySQL Replication Keith Larson SR002F.flac / SR002R.flac unclaimed
07 The Perfect Storm Jon 'maddog' Hall SR003F.flac / SR003R.flac unclaimed

Track 1 Talks

Track # Title Speaker Raw File(s) Editor
21 Android Custom ROM Development with Kexec Mike Kasick STE-001.flac unclaimed
22 Embedded Development with Linux John J. McDonough STE-001.flac unclaimed
23 Collabograte: Open Source Collaboration Kartik Subbarao N/A; didn't use microphone
24 The Importance of Free Software and Accessibility Jonathan Nadeau STE-002.flac unclaimed
25 Servers So Easy a Caveman Can Do It Christopher Laco STE-003.flac unclaimed
26 The Road to 31 Flavors Todd Robinson STE-003.flac unclaimed
27 Level Up: Linux and the Original Gameboy Dru Streicher STE-003.flac unclaimed


Card 2

Link to raw files

Track 2 Talks

Track # Title Speaker Raw File(s) Editor
31 Free Software for Audio and Video Production Neil Clopton STE-002.flac unclaimed
32 Automating Video Editing with Open Source Raul Suarez STE-002.flac unclaimed
33 Amahi - Powerful, Simple, Home Server Alan Jachimiak STE-003.flac unclaimed
34 Imaging with FOG Richard Gingerich STE-003.flac unclaimed
35 Customizing FreeNAS 8.3 Using Plugins Jail Dru Lavigne STE-004.flac unclaimed
36 Creating a Self-Defending Network with Open Source Steve McMaster STE-004.flac unclaimed
37 Highly Available Applications Brian Likosar STE-004.flac unclaimed


Card 5

Link to raw files

Track 4 Talks

Track # Title Speaker Raw File(s) Editor
41 We Will Be Legion: Decentralizing the Web Deb Nicholson STE-001.flac unclaimed
42 How to Create Your Own Cloud Joe Brockmeier STE-001.flac unclaimed
43 Aeolus: Deploying Across Clouds the Open Source Way Mo Morsi STE-002.flac unclaimed
44 It Takes an Open Source Village to Build a Cloud Joe Brockmeier STE-002.flac unclaimed

Closing Keynote Talks

Track # Title Speaker Raw File(s) Editor
71 Growing Up with Linux Elizabeth Garbee STE-003.flac unclaimed
72 How to Create Ravenously Passionate Contributors Angela Byron STE-003.flac unclaimed


Card 6

Link to raw files

Opening Keynote Talk

Track # Title Speaker Raw File(s) Editor
11 Stopping SOPA and Its Spawn: Free Expression on the Internet Wendy Seltzer STE-001.flac unclaimed

Track 3 Talks

Track # Title Speaker Raw File(s) Editor
12 Bedrock Linux Daniel Thau STE-001.flac unclaimed
13 Introduction to PC-BSD 9 Kris Moore STE-001.flac unclaimed
14 What Are Open Source Communities Like? Samuel Greenfeld STE-003.flac unclaimed
15 Extraterrestrial Adventures of an Open Source Community Sarah White STE-003.flac unclaimed
16 Yes, You Can Run Your Business on PostgreSQL Josh Williams STE-004.flac unclaimed
17 Fund Raising 101 Cat Allman STE-004.flac unclaimed

Card 7

Link to raw files

Open Source Solutions Stage Talks

Track # Title Speaker Raw File(s) Editor
51 Advancements in Open Source Virtualization with KVM Mike Day STE-001.flac unclaimed
52 Bringing Linux Essentials to the Classroom Medina Dupuis STE-001.flac unclaimed
53 Open Source Initiatives at HP Phil Robb STE-002.flac unclaimed
54 Help a Sysadmin Out! Alex Juarez STE-002.flac unclaimed
55 Cloud Monitoring in 15 Minutes or Less Rob Booth N/A; didn't use microphone
56 Arista Networks Extensible Operating System Darrin Machay N/A; didn't use microphone


Card 9

Link to raw files

Career Track Talks

Track # Title Speaker Raw File(s) Editor
61 The American Dream: Seven Steps to (Career) Ownership Janine Moon SR000F.flac / SR000R.flac unclaimed
62 CoverMyMeds Matt Scantland SR000F.flac / SR000R.flac unclaimed
63 2CheckOut Alan Czako SR000F.flac / SR000R.flac unclaimed
64 The Perspective of Recruiters Lou Russo and Rufus Smith SR001F.flac / SR001R.flac unclaimed
65 LPI: Catching the Wave of Open Source Careers Ross Brunson SR001F.flac / SR001R.flac unclaimed
66 The Intentional Career David Crone SR001F.flac / SR001R.flac unclaimed
67 Manta Steven ? SR002F.flac / SR002R.flac unclaimed
68 Rackspace Tricia Medina, Deborah Carter, Christine ?, ? Brown SR002F.flac / SR002R.flac unclaimed
69 You Inc. - Taking Control of Your Career Ben Blanquera SR002F.flac / SR002R.flac unclaimed


Editing Instructions

Some general notes:

  • Audacity works fine, and is what I'll describe. If you like something else better, use that.
  • Take care not to change the sample rate from the original. Some files are 44.1 kHz, some are 48 kHz. Using File -> Import in Audacity will cause resampling to the project's rate. Use File -> Open instead to avoid this.
  • The final product will be a separate 1-channel (mono) 16-bit PCM WAV file for each talk. Compress this into a FLAC file before uploading to save time and space.

There are two methods that were used for recording the talks. The Early Penguin and Career Track talks were recorded using the H2's four built-in microphones. This results in two stereo files with the names SR###F.wav and SR###R.wav. The "F" comes from the front microphones and the "R" from the rear ones. The rest of the talks were taken directly from the PA soundboard via the line-in jack. These have only a single stereo file named STE-###.wav. All of these were compressed into .flac files before uploading.

Separating talks

Load the STE-###.flac or SR###F.flac file into Audacity with File -> Open. If applicable, also import the matching SR###R.flac file using File -> Import. Locate the beginning of the talk in question. There may be some random discussion, so find the place where the speaker "officially" starts their presentation. Highlight and delete everything prior to this point (Edit -> Delete). If you have more than one track on screen, make sure you hit all of them.

Next you locate the end of the talk. This is after the speaker has finished and any open question-and-answer period. Some recordings may capture post-talk conversations that individuals are having with the speaker. These should not be included as they are not intended to be public. Highlight everything from this point to the end of all tracks, and then delete.

Selecting audio channel

From the drop-down track menu (the downward-pointing triangle at the upper left of the track), select Split Stereo to Mono for each track. Now your job is to choose which of these sounds the best. You can isolate tracks using the Mute and Solo buttons on each track.

If you're working on a STE-###.flac file, there should really be no difference. If you find that one has more noise or lower signal, then delete that track by clicking the X in the upper left; otherwise, pick one at random to delete.

For the SR###F.flac / SR###R.flac files, you'll probably find that one of the channels is louder than the others. Pick whichever one seems to have the best signal to noise ratio, then delete the rest using the X in the upper left.

Normalizing

At this point you'll have just one single mono track left. Make sure you do not have a selection made, or alternatively select the entire track. Now you'll want to change the display to Waveform (dB) using the drop-down track menu (the triangle in the upper left of the track). You can zoom in by left-clicking on the scale or zoom out by right-clicking.

Olfcompressor1.jpg

You should see something like the above (minus the red and green lines). The red line represents the noise floor: that is, the amount of sound present when no one is talking. For STE-###.flac files, you can probably just set this to -60 dB and not worry about it, since noise shouldn't be a problem there. For the SR###F/R.flac files, you'll have to listen and pick a reasonable value (take note that this can only be adjusted in 5 dB steps). Here we've chosen -55 dB.

The green line indicates the threshold value. This is the point at which the effect starts to tamp down the louder peaks so that the entire track has a more consistent volume level. Set this around the level of the most quiet parts of the speech. Here we're using -50 dB. Try not to set this too close to the noise floor; a 5 dB difference is probably as close as you want to be.

Now we select Effect -> Compressor from the menu and get a dialog like the below.

Olfcompressor2.jpg

As discussed, we have set the threshold at -50 dB and the noise floor at -55 dB. Those two settings are the only ones that need to be changed from one speech to another. Make the ratio 10:1, attack time 0.1 secs, and decay time 1.0 secs. Check the box for Make-up gain, and uncheck the one for Compress based on Peaks. Then click OK and wait a few minutes for it to work its magic.

Olfcompressor3.jpg

When complete, you should get something like the above. The times where the speaker is talking should be no lower than -15 dB, and preferably -12 dB or greater. This example is quite noisy, so you don't see as much differentiation between periods of speech and periods of silence. Take a listen and if you're not happy, hit Edit -> Undo and try again with different settings.

Saving

Once you're happy with the way things sound, go to File -> Export. You can select "WAV signed 16 bit PCM" as the format, or if it's supported choose "FLAC" from within Audacity (make sure under "Options" you choose 16 bit). Name the file with the track number (e.g., 07.wav or 07.flac). WAV files can be compressed after the fact using the command line flac --best ##.wav which will create a file named ##.flac.