[00:48.050 --> 00:50.890]  Welcome again to the final afternoon of HOPE.
[00:52.110 --> 00:57.210]  For attendees that are here, keep in mind at 6 o'clock, we do have our closing ceremonies.
[00:57.690 --> 01:01.290]  Overflow is going to be in room 206 and not the little theater.
[01:01.690 --> 01:04.710]  Room 206 will be where we have overflow.
[01:04.710 --> 01:09.990]  So please come and join us, whether you're in the main room or in the overflow room.
[01:11.750 --> 01:13.570]  And now on to our next talk.
[01:14.770 --> 01:19.730]  Whether we're talking about commercial products like... sorry, allergies.
[01:20.690 --> 01:34.570]  Non-profit projects like LibriVox or a community project or projects for the community, such as talking books, such as Reading for the Blind, audiobook production is nowhere near as straightforward as it might seem.
[01:35.070 --> 01:41.710]  To tell us more about what happens behind the scenes and how it gets done, our good friend Neo.
[01:44.070 --> 01:44.570]  Thank you.
[01:44.750 --> 01:45.330]  Thank you all.
[01:46.430 --> 01:54.530]  And just before we begin, there was a talk, I think it was here yesterday, one of the rooms where we were talking about the history of Hotel Pennsylvania.
[01:54.530 --> 01:56.870]  We omitted something rather important.
[01:57.010 --> 02:01.650]  We forgot to thank the staff of Hotel Pennsylvania for hosting us and putting up with us for all those years.
[02:01.990 --> 02:04.010]  So, corrections out of the gate.
[02:04.910 --> 02:05.570]  All right.
[02:05.690 --> 02:08.210]  So, thank you for showing up.
[02:08.510 --> 02:11.850]  What we're talking about is writing for the ear.
[02:11.850 --> 02:16.490]  Now, writing typically means ink, pen, typewriter, print, paper.
[02:16.630 --> 02:17.410]  It's a visual medium.
[02:17.950 --> 02:23.810]  When we're writing for the ear, we are transmuting print into audio.
[02:25.990 --> 02:28.310]  So, the scope of the talk covers a lot of things.
[02:28.450 --> 02:33.810]  You'll hear phrases like talking book, audiobook, radio dramatization, lectures.
[02:33.810 --> 02:40.290]  I want to address them because we sometimes use the word talking book to cover a lot of things or audiobook to cover a lot of things.
[02:41.190 --> 02:44.830]  So, I'm going to start with a couple of definitions about what is in scope of this document.
[02:45.290 --> 02:48.770]  First and foremost, what's within scope is the talking book.
[02:49.310 --> 02:59.070]  The point of the talking book, very simply, is to transmute the words from the visual medium into audio such that the printed copy is not necessary.
[03:00.170 --> 03:05.870]  Now, that's an important point because not only is a book its content, but it's also navigable.
[03:06.170 --> 03:08.950]  So, some of you will have older dictionaries at home.
[03:09.150 --> 03:17.510]  And you'll remember those little finger creases that you can put on the side so you can rapidly get to letter X or Q or H or O or P or E.
[03:18.410 --> 03:20.750]  That has to be translated into audio as well.
[03:21.970 --> 03:24.010]  A talking book will focus on that.
[03:24.010 --> 03:29.150]  And I happen to have a record here for an example of Jorge Luis Borges as a talking book.
[03:30.490 --> 03:38.590]  An audiobook, on the other hand, makes the written word available in addition to or instead of print.
[03:39.630 --> 03:43.550]  So, Douglas Adams read his own work and it's not a bad read.
[03:43.570 --> 03:47.810]  It's not as fun as the radio play, but he read his own work and that is an audiobook.
[03:49.730 --> 03:56.650]  On the border of what's on the scope of this talk is going to be recitations and recitals.
[03:57.330 --> 04:05.090]  So, Borges reading his own work, reading his own poetry, or George Carlin's famous modern man.
[04:05.310 --> 04:08.750]  I'm a modern man, digital, smoke-free, ready for the millennium.
[04:08.890 --> 04:10.570]  That's more of a recital of poetry.
[04:11.050 --> 04:14.690]  But they do fall under the same heading and there is some overlap between.
[04:15.390 --> 04:20.130]  What's beyond the scope are dramatizations, making words come to life.
[04:20.270 --> 04:24.870]  And there you have the BBC radio series for the Hitchhiker's Guide, which everybody is familiar with, I'm sure.
[04:25.790 --> 04:26.270]  Lectures.
[04:27.110 --> 04:30.090]  Borges, again, I harp on Borges for a couple of different reasons.
[04:30.510 --> 04:33.210]  But he gave some lectures called This Craft of Verse.
[04:33.350 --> 04:34.770]  Basically, it's the same thing that's happening here.
[04:35.090 --> 04:41.950]  Speaking about metaphor and lecture and talking about the uncouthness of certain phrases in translation.
[04:43.190 --> 04:46.990]  And finally, with a shameless plug to Rob and Gila, who are not with us.
[04:47.090 --> 04:50.790]  And they are hope institutions.
[04:52.170 --> 04:52.610]  Podcasts.
[04:52.770 --> 04:54.970]  Because podcasts, you don't start with the written word.
[04:55.410 --> 04:57.510]  Podcasts just are the audio.
[04:58.850 --> 05:02.530]  So, with that in mind, and while I bonk a second laptop out of position.
[05:03.210 --> 05:07.850]  We're going to move on to the next part, which is pre-production.
[05:08.150 --> 05:14.930]  So, in any audio, video, image, what have you, you have different phases.
[05:15.210 --> 05:17.430]  You have pre-production, production, and post-production.
[05:18.090 --> 05:23.470]  In the pre-production phase, you have to make certain decisions like, is this book worth reading?
[05:24.050 --> 05:28.490]  And I'm going to just start with the hot five that you should never read.
[05:28.610 --> 05:29.670]  First off, Bibles.
[05:29.810 --> 05:30.770]  We are sick and tired of Bibles.
[05:31.490 --> 05:32.410]  Bibles are everywhere.
[05:32.650 --> 05:34.030]  Everybody's got a free copy of the Bible.
[05:34.190 --> 05:35.150]  You can get it for five cents.
[05:35.350 --> 05:36.090]  You can get it for free.
[05:36.250 --> 05:37.250]  LibriVox has a copy.
[05:37.870 --> 05:41.030]  Whatever your denomination is, I assure you, even the Koran has been read.
[05:41.230 --> 05:42.430]  The Tanakh has been read.
[05:42.530 --> 05:43.230]  Everything has been read.
[05:44.910 --> 05:46.330]  Another one, children's books.
[05:47.130 --> 05:48.730]  Children's books are very demanding.
[05:49.590 --> 05:51.570]  To read to a child is like teaching.
[05:51.570 --> 05:54.050]  In order to teach, you have to do a lot of prep work.
[05:54.350 --> 05:58.890]  And you have to enunciate from time to time and know when to emphasize.
[05:59.790 --> 06:01.030]  Leave it to the professionals.
[06:01.150 --> 06:02.190]  You're just going to drive yourself crazy.
[06:03.130 --> 06:04.630]  What's the demand for the book?
[06:04.910 --> 06:06.370]  Does the book already exist?
[06:06.570 --> 06:07.750]  Who's your target audience?
[06:08.050 --> 06:10.710]  Do you have the time to complete these books?
[06:11.950 --> 06:14.270]  Do you have the talent, the funding?
[06:14.790 --> 06:16.430]  Do you have the copyright clearances?
[06:17.290 --> 06:22.470]  And when it comes to fanfic, theoretically you have your own clearances, but trust me, guys, listen to me.
[06:23.830 --> 06:28.230]  Nobody, nobody wants your copy of Fifty Shades of Great Beard with the NFT Tattoo.
[06:28.450 --> 06:29.170]  Don't read it!
[06:31.690 --> 06:33.090]  And finally, hot books.
[06:33.190 --> 06:39.130]  And this is where tech comes into a bit of a problem because hot books means, these are books that are going to be updated regularly.
[06:39.130 --> 06:45.630]  If I write a book about cybersecurity, if I write, by the time it hits the press, it's out of date.
[06:46.330 --> 06:50.750]  By the time you get the audio book recording of it, it's not only out of date.
[06:51.030 --> 06:52.210]  Academia is ahead of you.
[06:53.030 --> 06:55.470]  So you might want to avoid doing those things.
[06:56.030 --> 07:01.470]  And consider books that might be worth reading, like magazines, periodicals, serialized works.
[07:01.470 --> 07:09.810]  The reason for that is your time commitment, your energy commitment to complete these products is a big commitment.
[07:09.950 --> 07:11.590]  So you have to really think about it.
[07:13.650 --> 07:18.530]  From that pre-production position, we now move on to how to read and record.
[07:18.530 --> 07:24.050]  So let's say that you've been through your work, you've decided that the book that you have in your hand is worth the recording.
[07:24.710 --> 07:26.870]  Well, first off, you have to read clearly.
[07:27.370 --> 07:30.230]  You can hear a little bit of dryness in my mouth.
[07:30.450 --> 07:31.630]  Just a slight amount.
[07:32.490 --> 07:38.570]  It has to be clear, consistent, coherent, and the word really you're looking for is authoritative.
[07:38.790 --> 07:40.450]  You are a representation of the book.
[07:41.170 --> 07:45.790]  Going back to our example of the Hitchhiker's Guide, Simon Jones is the voice of the book.
[07:46.530 --> 07:49.770]  When you hear Simon Jones's voice, you hear the book.
[07:49.870 --> 07:52.870]  And it's presented as, I know what I'm speaking about.
[07:53.450 --> 07:57.490]  You are speaking on behalf of the author who may or may not be reading it themselves.
[07:59.690 --> 08:02.150]  The next up is, we have two pieces of guidance.
[08:02.390 --> 08:13.990]  Now, I work in accordance with the guidelines of the National Library Service of the Library of Congress because I produce talking books for the blind, the visually impaired, the reading disabled, and people who physically cannot hold a book in their hands.
[08:15.050 --> 08:23.990]  So, the guideline that I've been following, way up here, is Billy West's The Art and Science of Audiobook Production.
[08:24.290 --> 08:26.070]  It's a great crash course.
[08:26.070 --> 08:27.770]  It gives you some background information.
[08:28.950 --> 08:30.810]  And it's a very high standard.
[08:31.190 --> 08:33.390]  And it's intentionally to be high standards.
[08:34.410 --> 08:36.930]  The other one with LibriVox, they have their own little wiki.
[08:37.030 --> 08:38.690]  They have their own policy pages and guidelines.
[08:38.690 --> 08:40.930]  But you will find the Storyteller's Recording Guide.
[08:41.370 --> 08:42.550]  So, refer to them.
[08:43.290 --> 08:49.670]  In the case of Art and Science of Book Production, NLS reorganized their website recently.
[08:49.850 --> 08:54.930]  But fortunately, somebody decided to archive it on Internet Archive.
[08:55.070 --> 08:56.850]  So, it is there if you want to read it.
[08:58.710 --> 08:59.890]  So, those are your guidelines.
[09:00.110 --> 09:02.010]  But we have to conform to rules.
[09:02.530 --> 09:07.930]  So, LibriVox has its policies regarding recording, text, music and sound effects.
[09:08.070 --> 09:08.930]  I'll spare you a click.
[09:09.110 --> 09:09.490]  No!
[09:11.010 --> 09:11.490]  Languages.
[09:11.850 --> 09:13.750]  They have their guidelines, their rules.
[09:14.310 --> 09:16.170]  So, does the National Library Service.
[09:16.170 --> 09:19.570]  And those specs that you see up there are what I deal with every day.
[09:19.890 --> 09:30.710]  Ranging from narration, talking book mastering, construction of the books, construction of magazines, and preparation of audio files from pre-recorded commercial works.
[09:31.310 --> 09:37.050]  And even in the commercial world, Amazon and Findaway Voices will have their audio requirements.
[09:37.570 --> 09:41.130]  Before you begin recording, learn these requirements.
[09:41.150 --> 09:42.710]  Those are your goal posts.
[09:42.950 --> 09:44.730]  It's not just about the reading.
[09:46.290 --> 09:48.410]  And you have to learn from others.
[09:48.650 --> 09:49.890]  And I'm just throwing it up there.
[09:50.510 --> 09:51.270]  Klingon Tich!
[09:51.770 --> 09:52.930]  Shuvuik Demek Pa!
[09:53.910 --> 09:57.410]  Which I can't pronounce properly, so I should not read that book.
[09:57.810 --> 10:00.570]  Instead, we'll leave it to Michael Dorn and Roxanne Dawson.
[10:00.670 --> 10:03.990]  Because they are friggin' awesome and Klingon is a very interesting language.
[10:07.660 --> 10:13.460]  So, once you're familiar with all these rules, we start talking about the hardware you're going to use.
[10:14.060 --> 10:16.920]  So, I've broken this down into four sections.
[10:17.780 --> 10:21.640]  First and foremost, don't use all-in-one field production units.
[10:22.120 --> 10:23.240]  Just don't.
[10:23.960 --> 10:26.040]  And I'll cover a couple of reasons why not.
[10:26.260 --> 10:30.140]  But the basic one is, there are many programs where you are recording from home.
[10:31.580 --> 10:34.800]  Even in this room, there is a little background sound.
[10:36.000 --> 10:38.180]  I can hear the projector up ahead.
[10:38.520 --> 10:40.820]  I can hear some fans and ventilation.
[10:41.200 --> 10:43.200]  The room has an echo to it.
[10:44.200 --> 10:49.000]  If you record at home, and this is pretty well soundproofed and reinforced by the speakers on the side.
[10:49.780 --> 10:52.440]  In your own home, where do you have such silence?
[10:54.880 --> 10:58.060]  And also, most people when they record, they don't have the microphone facing them.
[10:58.140 --> 11:04.340]  They'll have a little object on the table pointed towards their mouth while they're reading a book off at an angle.
[11:05.720 --> 11:07.020]  Now, I'm off at an angle.
[11:07.560 --> 11:08.480]  I'm reading the book.
[11:08.620 --> 11:10.380]  I'm trying to focus on the print in front of me.
[11:10.600 --> 11:13.740]  You'll notice that nobody can hear me because I'm not on the microphone.
[11:13.740 --> 11:16.560]  I have to turn in order to face the microphone.
[11:16.700 --> 11:17.480]  My levels go down.
[11:17.720 --> 11:23.100]  And if you're recording alone, and you don't have the audio feedback, and you don't have a director, it becomes very difficult.
[11:23.720 --> 11:26.660]  The other issue is most people put those devices on a table.
[11:27.180 --> 11:29.600]  A table has echo to it.
[11:29.900 --> 11:37.420]  It will reinforce, create reverberation, and you end up creating a very decent reading, but a very bad recording.
[11:40.140 --> 11:42.240]  So that's one of the reasons why I don't recommend it.
[11:42.240 --> 11:45.540]  The other is, a lot of the units don't have the specifications you need to know.
[11:46.060 --> 11:50.240]  You need to know, is this microphone omni-directional, picking up everything around it?
[11:50.560 --> 11:53.400]  Is it hypercardioid, facing one direction?
[11:54.060 --> 11:57.520]  You need to know what noise itself it brings into the chain.
[11:58.740 --> 12:02.320]  So all of these details tend to be missing from the recorded home units.
[12:03.880 --> 12:07.120]  And the most important part is, how do you know if it's powered on?
[12:07.540 --> 12:09.600]  How do you know if you've run out of storage space?
[12:09.600 --> 12:12.080]  How do you know if it's working at all?
[12:13.160 --> 12:16.760]  You may record three hours and not have anything to show for it.
[12:18.520 --> 12:20.420]  So the easier parts are storage.
[12:20.580 --> 12:21.740]  Now, storage is very simple.
[12:21.880 --> 12:25.180]  Anything capable of sustaining one megabyte per second.
[12:25.260 --> 12:26.360]  Actually, it's 1.1.
[12:26.360 --> 12:33.820]  In those first specs, I'm taking DJ standards, which is a sample rate of 192 kHz, 24 bits stereo.
[12:34.400 --> 12:37.800]  You will mostly be recording at CD quality or DAT quality.
[12:38.100 --> 12:41.680]  The requirements, as you can see, are greatly reduced.
[12:42.780 --> 12:45.340]  But also, you'll probably be recording in mono.
[12:46.160 --> 12:49.360]  With microphones, we have a prime example of a microphone right here.
[12:49.520 --> 12:52.680]  It's the generic Shure SM58.
[12:53.460 --> 12:54.420]  It's a classic.
[12:54.600 --> 12:56.060]  It picks up in one direction.
[12:56.800 --> 12:57.300]  Dynamic.
[12:57.620 --> 13:00.440]  At the place where I work, we use Shure SM7Bs.
[13:01.300 --> 13:03.120]  They're unidirectional dynamic microphones.
[13:03.240 --> 13:07.700]  Very popular with streamers, YouTube channels, Instagram.
[13:08.440 --> 13:09.540]  You'll see it everywhere.
[13:10.240 --> 13:12.860]  We used to use the Electro-Voice RE27ND.
[13:13.180 --> 13:15.020]  It was a cardioid dynamic microphone.
[13:15.140 --> 13:16.100]  It had neodymium drivers.
[13:16.560 --> 13:20.580]  And it was very close to what we would call a condenser mic in terms of its pickup.
[13:20.760 --> 13:22.720]  It did not forgive anything.
[13:23.320 --> 13:25.680]  If you had lunch an hour ago, we knew.
[13:26.360 --> 13:28.100]  If you farted, we knew.
[13:28.640 --> 13:35.160]  If the elevator behind the studio room where we were recording was going up or down, we knew.
[13:36.000 --> 13:39.000]  Your cell phone is basically a condenser microphone.
[13:39.460 --> 13:40.600]  It is unforgiving.
[13:40.780 --> 13:41.120]  It's cheap.
[13:41.260 --> 13:41.880]  It's small.
[13:42.040 --> 13:42.900]  It picks up everything.
[13:43.060 --> 13:48.160]  And if you want an example of it, try to make a phone call to somebody who you really need to speak to on a busy avenue.
[13:48.320 --> 13:51.460]  And try to hear them over the background noise.
[13:52.340 --> 13:55.080]  Most of the all-in-one units are condenser microphones.
[13:55.320 --> 13:57.120]  So splurge the extra if you can.
[13:57.960 --> 13:58.400]  Borrow.
[13:58.700 --> 14:01.360]  I'm not going to say steal, but beg if you have to.
[14:01.680 --> 14:03.480]  And get full equipment.
[14:03.480 --> 14:05.980]  The last item is not a microphone.
[14:06.640 --> 14:09.740]  It's actually a Focusrite 2i4 second generation.
[14:09.980 --> 14:14.780]  The reason we have is because these microphones have a connector called an XLR cable connector.
[14:15.300 --> 14:19.040]  Basically, signal, negative signal, ground, and shielding.
[14:19.160 --> 14:20.560]  And that's the connector right here.
[14:20.940 --> 14:28.280]  We use these Focusrite devices to convert from the analog connector of the XLR into a USB device.
[14:30.000 --> 14:31.500]  It just digitizes for you.
[14:31.740 --> 14:32.460]  Nice and easy.
[14:33.840 --> 14:34.560]  Finally, headphones.
[14:34.780 --> 14:36.640]  Because you need to be able to hear yourself.
[14:36.840 --> 14:44.780]  If you're working in a team, maybe you happen to have access to a studio, you need to be able to hear your director, giving you instructions, and you also need to hear yourself.
[14:46.200 --> 14:49.680]  So, typically, in our studios, we'll use an MDR 7506.
[14:49.840 --> 14:50.480]  They're very comfortable.
[14:50.780 --> 14:52.160]  They are soft padded.
[14:52.300 --> 14:53.640]  They're not open to the world.
[14:53.960 --> 14:55.520]  It's not like having an open speaker.
[14:56.620 --> 15:03.160]  The cable is corded, and it allows you some freedom of movement, so you can reposition as you're reading.
[15:04.960 --> 15:06.580]  They're not expensive.
[15:07.400 --> 15:08.560]  I'll be honest about that.
[15:08.720 --> 15:10.840]  They might be a little more...
[15:11.640 --> 15:13.620]  Let's say a project would start at about $250.
[15:14.860 --> 15:20.560]  So, of all that equipment, you can pretty much do a lot of it getting away on the cheap and easy the first couple of times.
[15:21.360 --> 15:30.160]  But do not, and I really have to emphasize this, do not use open headphones, anything that causes feedback.
[15:30.980 --> 15:34.920]  So, for example, I happen to like my cheap speakers, $29.
[15:36.260 --> 15:38.360]  And if they break, who cares, because they're only $29.
[15:38.880 --> 15:41.660]  But they have enough frequency response that you can hear your own voice.
[15:41.880 --> 15:45.620]  You don't need studio engineering speakers that go from 20,000 to 20,000 hertz.
[15:46.120 --> 15:49.880]  You need a narrow range, roughly between 100 and 5 kilohertz.
[15:51.460 --> 15:54.100]  Don't use in-ear earbuds.
[15:54.980 --> 15:56.340]  That's going to leak all the audio.
[15:56.460 --> 15:57.560]  It's going to come back to your microphone.
[15:57.640 --> 15:58.200]  You get a feedback.
[15:58.980 --> 16:01.080]  The final one, and I know a lot of people like headsets.
[16:01.160 --> 16:03.320]  I happen to like headsets, but don't use a Bluetooth headset.
[16:04.120 --> 16:04.840]  Please don't.
[16:05.660 --> 16:07.380]  Bluetooth is notorious for its latency.
[16:07.580 --> 16:09.080]  I know it's getting better and better all the time.
[16:09.260 --> 16:24.980]  But by the time you start a sentence, have it transmitted over Bluetooth, arrive at the computer, commit it to disk, have the computer play it back to you, go over Bluetooth and come back to your ear, you have already lost anywhere from 20 milliseconds to 100 milliseconds very easily.
[16:25.260 --> 16:26.620]  It will distract you.
[16:26.740 --> 16:27.900]  It will slow you down.
[16:28.040 --> 16:31.140]  And you'll just be frustrated to the point where you can't continue.
[16:33.360 --> 16:34.880]  Software, very simple.
[16:35.200 --> 16:37.960]  And most of you probably have two out of these four.
[16:38.340 --> 16:41.020]  The first one is the APH Studio Recorder.
[16:41.020 --> 16:45.400]  It's popular with our studios because it's very easy for volunteers to use.
[16:45.600 --> 16:49.060]  It's sort of the dumbest recorder you can possibly have, and it works.
[16:50.160 --> 16:53.980]  For post-production, as well as recording, we can use the Hindenburg Audiobook Creator.
[16:53.980 --> 16:58.040]  We use a specific version made for the National Library Service and affiliates.
[16:58.440 --> 17:03.340]  That's what allows us to take the recordings and turn them into full-fledged talking books.
[17:03.340 --> 17:11.300]  It's digital talking books ready to be serviced and given to patrons of the Library of Congress who are blind, visually impaired, and reading disabled.
[17:12.180 --> 17:12.860]  Audacity.
[17:13.180 --> 17:14.180]  Ah, yes.
[17:15.080 --> 17:19.780]  The audio editing software made by programmers for programmers.
[17:20.600 --> 17:22.340]  But there ain't nothing else.
[17:22.340 --> 17:26.380]  I mean, realistically speaking, you could use SOX.
[17:26.500 --> 17:31.760]  I mean, you could just say, poll, dev, input, mic, zero, and commit to disk.
[17:32.140 --> 17:34.640]  And just do live recordings and edit it afterwards.
[17:34.900 --> 17:36.880]  But Audacity really is the only game in town.
[17:37.120 --> 17:41.280]  The other audio software that exists on Linux is mostly driven for music.
[17:41.580 --> 17:44.800]  And you're not necessarily making music.
[17:44.800 --> 17:48.840]  You are making a recording of audio, of voice.
[17:49.320 --> 17:50.740]  It's a very different beast.
[17:52.380 --> 17:53.940]  So we've talked about hardware.
[17:54.300 --> 17:55.260]  We've talked about software.
[17:55.440 --> 17:56.500]  Time to talk about underwear.
[17:56.680 --> 17:57.380]  Your talent.
[17:58.420 --> 18:01.600]  Now your talent breaks down into basically four people.
[18:01.880 --> 18:04.660]  A narrator, director, and an engineer.
[18:04.900 --> 18:08.080]  So the narrator translates the written word.
[18:08.140 --> 18:10.600]  And I'm going to read off of this guy because my other laptop is shot.
[18:11.200 --> 18:18.780]  The narrator translates the written word to the spoken word in a way that is as consistent as possible with the intent of the author.
[18:18.980 --> 18:20.260]  Thank you, Bill West, for writing that.
[18:20.320 --> 18:21.020]  You saved me a lot.
[18:21.660 --> 18:24.800]  So the job of the narrator is to be the author.
[18:25.080 --> 18:26.220]  Be authoritative.
[18:26.860 --> 18:35.500]  The engineer controls all technical operations during a recording session, allowing the narrator to devote full attention and effort to the act of narration.
[18:36.440 --> 18:38.020]  So, checking your levels.
[18:38.140 --> 18:40.480]  Making sure that you're not over-plussing the mic.
[18:41.420 --> 18:44.020]  That you are, in fact, talking to the microphone.
[18:45.620 --> 18:46.500]  That's his job.
[18:46.600 --> 18:47.340]  Or her job.
[18:47.460 --> 18:48.000]  Or their job.
[18:48.200 --> 18:49.460]  Making sure that the computer is running.
[18:50.120 --> 18:51.820]  Making sure that you have storage space.
[18:52.240 --> 18:52.960]  That's their job.
[18:53.360 --> 18:55.140]  The director has a harder job.
[18:55.360 --> 18:56.660]  The director is sort of like a driver.
[18:56.820 --> 18:58.540]  You know, when you drive on a road, you have to look at the road.
[18:58.580 --> 18:59.340]  You have to look at the traffic.
[18:59.520 --> 19:01.600]  You have to remember what the lights mean.
[19:01.800 --> 19:04.080]  What the rules of the road are when you try to make a left turn.
[19:04.240 --> 19:05.300]  Make sure you're not speeding.
[19:05.540 --> 19:06.840]  The director has multiple roles.
[19:06.840 --> 19:12.340]  And basically, the director is responsible for quality assurance of the content and its continuity.
[19:12.960 --> 19:16.540]  They are the first listener of your audiobook.
[19:17.580 --> 19:21.900]  Now, I'm saying listener for a reason because there's also a first reader of the book.
[19:22.080 --> 19:23.540]  These books are read.
[19:24.020 --> 19:33.300]  So the director is concerned with tonality, inflections, pronunciation, enunciation, energy levels, whether you are reading correctly.
[19:34.380 --> 19:37.160]  They are also looking for digestion noises.
[19:37.500 --> 19:41.080]  Or if you pick your nose and it happens to show up on the microphone, they are going to stop you.
[19:41.080 --> 19:43.760]  If you sound tired, if you sound dry.
[19:44.820 --> 19:47.400]  And they also basically keep the recording record.
[19:48.020 --> 19:50.000]  That is to say, metadata about the recordings.
[19:53.300 --> 19:54.640]  Finally, there's quality control.
[19:54.780 --> 19:59.580]  But before I continue to quality control, because it's another stage, don't try to do all three of these things at once.
[20:00.180 --> 20:03.540]  Because you can't be, and I'm going to just try this right now, and you're seeing it right now.
[20:03.680 --> 20:04.540]  I'm trying to read.
[20:05.340 --> 20:06.480]  Stay on the microphone.
[20:06.940 --> 20:07.600]  Check my levels.
[20:08.020 --> 20:08.680]  Check the clock.
[20:09.020 --> 20:09.820]  Make sure that I'm okay.
[20:10.080 --> 20:11.420]  I've lost track of where I was.
[20:11.560 --> 20:12.120]  Where was I?
[20:13.740 --> 20:17.640]  You end up doing a lot of frustrating work trying to just stay on track.
[20:19.100 --> 20:26.220]  And quality control checks for errors and irregularities in the recording and notifies the talent of what needs fixing.
[20:26.220 --> 20:29.920]  So, quality control should be an unbiased ear.
[20:30.240 --> 20:31.700]  It should not be you.
[20:31.860 --> 20:33.740]  You should not check your own work.
[20:34.100 --> 20:35.740]  Have someone else do it for you.
[20:36.220 --> 20:50.600]  If you are working remotely, you might consider doing a Zoom session, or a Skype session, or a team session, or just have someone on the phone listening in with their own copy of the book that you're trying to read, and follow along.
[20:51.860 --> 20:53.260]  It does not take much.
[20:57.430 --> 20:59.910]  Next up, we have the production environment.
[21:00.170 --> 21:03.030]  So, hardware, software, underwear, and just plain wear.
[21:05.350 --> 21:08.170]  Noises come in different forms, and I've covered a lot of them.
[21:08.430 --> 21:12.250]  You know, fans, if you're recording in your kitchen, it's a refrigerator typically.
[21:13.090 --> 21:15.110]  If you're by a busy street, you have traffic.
[21:15.310 --> 21:24.950]  If you're out, you know, near JFK or LaGuardia or wherever, you have air traffic, lawnmowers, and your neighbors, because your neighbors are never cooperating with you ever.
[21:25.510 --> 21:25.570]  Right?
[21:26.970 --> 21:30.910]  But soundproofing is a way of abating extraneous noises.
[21:31.150 --> 21:33.590]  Now, there are two kinds of noises.
[21:33.810 --> 21:35.610]  There's the kind that the structural noise.
[21:35.850 --> 21:39.950]  So if, you know, the ground shakes, you can't do much about it, the building carries all that sound.
[21:40.470 --> 21:43.170]  If the air has noise to it, you can abate it.
[21:43.170 --> 21:50.430]  And those of you familiar with Radio Statler from here at HOPE will remember that we had polystyrene walls with some graffiti on it.
[21:50.510 --> 21:52.270]  But we're focusing on the polystyrene.
[21:52.870 --> 21:53.790]  It acts as a buffer.
[21:54.110 --> 21:58.830]  So you can have basically Styrofoam walls, a small enclosed studio.
[21:59.910 --> 22:02.170]  The other possibility is use drapes.
[22:02.490 --> 22:03.370]  Use curtaining.
[22:03.750 --> 22:06.190]  Use a heavy blanket.
[22:07.410 --> 22:08.130]  But don't!
[22:08.650 --> 22:09.010]  Don't!
[22:09.350 --> 22:09.710]  Don't!
[22:09.870 --> 22:10.810]  Read in the bathroom!
[22:12.870 --> 22:13.830]  Why do we say that?
[22:13.930 --> 22:14.990]  Okay, so a lot of people...
[22:14.990 --> 22:16.550]  Who thinks they sing great in the bathroom?
[22:18.050 --> 22:19.570]  Okay, I've got a lot of honest people here.
[22:21.170 --> 22:24.050]  So, you think that the bathroom is a great place to read.
[22:24.430 --> 22:28.990]  Because you're enclosed, you're comfortable, you have that lovely tonality to it.
[22:29.050 --> 22:30.530]  Well, again, that's the room noise.
[22:30.530 --> 22:33.350]  That's the reverberation and the echo of the tiles.
[22:34.850 --> 22:35.570]  Don't do it.
[22:35.990 --> 22:40.330]  Because you sound good to yourself, but a microphone does not forgive.
[22:40.850 --> 22:53.110]  The same way that the human being basically discards most of its information and focuses in on those narrow things that it really has to pay attention to and remember, your ears do the same thing, your eyes do the same thing.
[22:53.270 --> 22:56.790]  Most of the information that you pick up, you just immediately discard.
[22:56.790 --> 23:00.030]  The brain is a lossy medium.
[23:01.570 --> 23:02.670]  Microphones don't forgive.
[23:03.270 --> 23:04.950]  So, don't do it in your bathroom.
[23:05.570 --> 23:06.830]  Also, somebody else might need the bathroom.
[23:06.950 --> 23:07.330]  You never know.
[23:08.950 --> 23:15.950]  So, once you've gone from the production phase, we now go into post-production, where you have editing and then you have distribution.
[23:17.070 --> 23:22.290]  With editing, basically, you are dealing with a 16-bit wave, maybe a FLAC.
[23:22.810 --> 23:27.030]  You're dealing with CD-quality or DAT-quality audio.
[23:27.730 --> 23:31.150]  And you're going to perfect it, and you're going to keep it under 2 gigs.
[23:33.090 --> 23:37.030]  The reason for that is, one, if your recording is longer than 3 hours, you're doing something wrong.
[23:37.990 --> 23:42.810]  And two, you should always expect legacy hardware in use.
[23:42.810 --> 23:46.430]  In fact, I make a mention of Neo's mantra later on.
[23:47.130 --> 23:49.050]  But expect embedded devices.
[23:49.850 --> 23:54.710]  Expect a 32 plus or minus 1 bit limit.
[23:56.550 --> 24:03.330]  Your workflow for this is going to basically be, make sure that you have all the components of the audio book, before you sit down to edit.
[24:04.070 --> 24:06.150]  Label all the logical sections.
[24:06.470 --> 24:09.830]  Look at the table of contents.
[24:09.830 --> 24:13.970]  I mean, it's stupid, but you open up a book, you see a table of contents, typically.
[24:14.470 --> 24:18.350]  In pieces of fiction, it'll be, you just jump right into the first chapter.
[24:18.770 --> 24:19.890]  Know what the hell you're reading.
[24:20.430 --> 24:21.850]  Find out how many chapters you have.
[24:21.970 --> 24:23.710]  How many parts you have to it.
[24:23.750 --> 24:27.890]  I mean, we got one book where the first chapter was numbered 1.
[24:28.210 --> 24:30.070]  The last chapter was number 36.
[24:30.490 --> 24:34.570]  We got all the audio down to the engineer's booth where we were editing it.
[24:34.750 --> 24:37.770]  And we discovered that chapters 24 through 28 were missing.
[24:38.370 --> 24:41.770]  If you just check the first and the end, you think everything's great.
[24:43.190 --> 24:44.950]  So make sure you have all your materials.
[24:45.950 --> 24:48.770]  Make sure you have preliminary header and concluding materials.
[24:48.870 --> 24:49.690]  Now, this is your metadata.
[24:50.350 --> 24:54.730]  This is the part where you would say, name of the book by author.
[24:55.330 --> 24:57.270]  If you have a serial number, you throw it in there.
[24:57.410 --> 25:03.310]  And you say, narrated by, directed by, engineered by, recorded on such and such a date.
[25:03.310 --> 25:05.110]  For the benefit of.
[25:05.310 --> 25:06.970]  Under the auspices of.
[25:09.870 --> 25:14.810]  That keeps a record of not only who recorded it and who's responsible for it, but who took part.
[25:15.190 --> 25:18.890]  And there are some projects where you can contribute for money.
[25:19.310 --> 25:22.010]  And, you know, make a decent living.
[25:22.270 --> 25:23.890]  Not a great one, but a decent one.
[25:24.070 --> 25:25.970]  Nothing to laugh at.
[25:27.490 --> 25:31.550]  But, ultimately, you are finessing the audio quality last.
[25:32.830 --> 25:38.570]  For the last 75 years, the focus of talking books has been to convey the information.
[25:39.290 --> 25:40.850]  Not make it entertaining.
[25:41.270 --> 25:44.110]  Not to make it an audio experience.
[25:45.470 --> 25:48.430]  All you want to know is, how the hell do you spell a word?
[25:48.730 --> 25:51.250]  You have to be able to navigate through a talking book to get to it.
[25:51.690 --> 25:54.450]  Or, I just want to know what the epic of Gilgamesh is.
[25:54.550 --> 25:55.130]  I don't need an...
[25:57.870 --> 25:59.070]  And Gilgamesh did...
[25:59.070 --> 26:00.210]  Who cares about that?
[26:00.290 --> 26:00.990]  Just read the story.
[26:02.330 --> 26:05.270]  So, we get to the point first, and then we finesse the audio.
[26:05.970 --> 26:12.070]  And one of the ways you can do that, I'll mention later on, in the do's and don'ts is, Riding the Silence.
[26:14.650 --> 26:15.010]  Foreshadowing.
[26:16.470 --> 26:19.290]  Sometimes, you're going to have to transmit files from here to there.
[26:19.890 --> 26:22.850]  Now, admittedly, ideally, you would be all working in a studio.
[26:23.730 --> 26:25.070]  All your equipment would be present.
[26:25.190 --> 26:26.090]  All your people would be present.
[26:26.250 --> 26:28.690]  But sometimes, you have to work across the globe.
[26:29.370 --> 26:31.190]  So, how do you transmit big waves?
[26:31.350 --> 26:33.050]  Well, one thing is, you can knock it down to FLAC.
[26:33.830 --> 26:36.030]  FLAC is recommended under these circumstances.
[26:36.990 --> 26:39.810]  Another thing you can do is, you can TAR-XZ your waves.
[26:39.810 --> 26:43.090]  In both cases, you save about 25% of the file space.
[26:43.230 --> 26:46.850]  And the sooner you can transfer files, the closer it is to being done.
[26:48.650 --> 26:54.910]  For the love of God, TAR-Ball, your entire Audacity project.
[26:56.950 --> 27:03.070]  We received a very difficult book, a very lengthy book, to edit.
[27:05.030 --> 27:13.370]  And it consisted of about 65 kilobytes of the edit definition list from Audacity, because they sent the AUP, not the audio that came with it.
[27:15.750 --> 27:26.910]  If you're working in Audacity, and I know version 3 has a, you know, save compressed format, and single file format, expect other people to still be using 2.4 for various reasons.
[27:27.830 --> 27:30.610]  TAR-Ball your entire folder before you transmit it.
[27:30.870 --> 27:31.550]  Don't just send.
[27:31.670 --> 27:35.190]  Open file X at this time, and include file Y.
[27:35.890 --> 27:39.810]  And do not, under any circumstance, send MP3 to editors.
[27:40.150 --> 27:41.630]  And the reason for this is very simple.
[27:41.630 --> 27:45.390]  Once you lose quality, you cannot get it back.
[27:46.730 --> 27:49.270]  There's a saying, we'll fix it in post, we'll fix it in post.
[27:49.430 --> 27:50.970]  No, post can't fix things.
[27:51.810 --> 27:56.430]  If there's an omission, a gap, or a loss of quality, you can't fix it.
[27:56.510 --> 28:01.750]  You cannot recover it, because it was never made, or because it was outright destroyed.
[28:03.410 --> 28:08.330]  MP3 is very popular for a distribution format, but it is no good for editors.
[28:08.330 --> 28:15.470]  You're working on finessed, pristine, master versions of your project.
[28:16.270 --> 28:23.970]  And finally, use 8.3 filenames, because like I said, expect embedded devices from the year dot to still be in use.
[28:24.090 --> 28:27.110]  I mean, you can use an Atari ST to still cut audio.
[28:27.610 --> 28:29.830]  So, in comes Neo's mantra.
[28:29.830 --> 28:45.010]  Never underestimate the ubiquity of, and reliance upon, irreplaceable, and unupgradeable, well-past, end-of-life equipment, fulfilling mission-critical production roles.
[28:47.490 --> 28:51.330]  Expect old shit to still be used.
[28:52.670 --> 28:55.050]  And some of that, believe it or not, is DOS-based.
[28:58.380 --> 29:04.780]  Once you have finessed the audio, you can account for all of the recordings, you have all the metadata, we are ready for post-production distribution.
[29:04.900 --> 29:13.880]  This is the part where you take your master reels, and you give it to the public, or your patrons, or your club, or whatever.
[29:15.120 --> 29:19.120]  National Library Service uses 3GP AMR wideband.
[29:19.400 --> 29:21.980]  That's their codec of choice for their talking book medium.
[29:22.820 --> 29:27.740]  LibriVox will send out MP3s as a final product, not as an editing product.
[29:28.100 --> 29:32.800]  At 64 kilobits, if you go to Internet Archive, sometimes you can get it for 128 kilobit.
[29:33.540 --> 29:37.340]  Commercial proprietary formats go from AAC to M4B, etc.
[29:37.540 --> 29:39.960]  But basically, it's DRM'd, encoded audio.
[29:40.120 --> 29:42.200]  So, it may or may not apply to you.
[29:43.620 --> 29:45.880]  But this is the far speaker.
[29:46.780 --> 29:48.540]  That is to say, the last...
[29:48.540 --> 29:50.360]  We're familiar with the notion of the last mile.
[29:50.580 --> 29:54.600]  The last mile of telecommunications, the bit between the home and the trunk line.
[29:55.160 --> 29:58.040]  Well, here is your last mile.
[29:58.860 --> 30:03.600]  You can get away with a lot of stuff simply knowing what speakers you're going to be playing against.
[30:04.580 --> 30:06.440]  This will give you 20,000 to 20,000 hertz.
[30:07.580 --> 30:10.520]  These speakers here might cover a little bit more of the range.
[30:10.520 --> 30:13.120]  Some people might be listening with in-ear buds.
[30:13.460 --> 30:16.660]  You can knock down the codec based on what your finished product is.
[30:16.860 --> 30:20.980]  But you can expect at least one person to have a good audio quality system going on.
[30:21.580 --> 30:26.960]  So, always provide a higher quality than you expect the device to be.
[30:30.180 --> 30:35.280]  Always produce a higher quality end product than the device you expect to be playing it on.
[30:38.460 --> 30:40.320]  So, now we get to talk about the workflow.
[30:40.500 --> 30:42.140]  We've covered pre-production, post-production.
[30:42.480 --> 30:43.580]  Was the book worth it?
[30:43.580 --> 30:44.740]  How are we going to distribute it?
[30:44.860 --> 30:46.080]  How are we going to make our master reels?
[30:47.100 --> 30:50.240]  The first part of any project is to commit to the project.
[30:51.540 --> 30:54.760]  If you don't commit to the project, there's no work to be done.
[30:55.940 --> 30:59.200]  You have to assemble your talent, your time, and your money.
[30:59.820 --> 31:00.660]  Yes, money.
[31:00.980 --> 31:04.920]  Because even if you're doing this on a volunteer basis, time is money.
[31:05.320 --> 31:06.440]  It comes at a cost.
[31:06.440 --> 31:08.320]  Maybe you can't watch the ball game.
[31:08.660 --> 31:13.100]  Or maybe you have to give up your Wednesday afternoons to listen to somebody else's recording.
[31:14.940 --> 31:17.780]  And there should be a little bit of compensation for all this work.
[31:18.120 --> 31:20.860]  At the very least, something more than just plain gratitude.
[31:22.460 --> 31:34.100]  But as you're dividing up the book into logical chapters or sections to be recorded, and let the table of contents guide you, divide up the work evenly and take responsibility for each part.
[31:34.100 --> 31:36.200]  Especially if you're working in a group setting.
[31:37.540 --> 31:50.900]  A production team which consists of a narrator, a monitor, a reviewer, typically requires five and a half hours of staff time to produce one hour of error-free product.
[31:51.560 --> 31:54.940]  This comes from the Bill West document I mentioned earlier.
[31:54.940 --> 32:06.160]  And a typical two-track original master, with just under three hours of listening time, requires not less than sixteen and a half hours of production time.
[32:07.100 --> 32:12.460]  So when you budget your time, you're being told to expect five and a half to six and a half to one.
[32:12.660 --> 32:14.300]  I would raise it to ten to one.
[32:15.060 --> 32:16.760]  Especially if you haven't done this before.
[32:19.580 --> 32:26.760]  Now, you can record a section on its own, send it off to your quality control person, and have them report back as soon as you finish recording.
[32:27.600 --> 32:36.480]  It does speed up the process, and it's easier for the person who's doing the narration, if they're told, while they have the content fresh in their mind, what they have to go back and fix.
[32:37.540 --> 32:43.100]  In a lot of places, you will record, then review the whole book.
[32:44.080 --> 32:45.780]  That can take weeks.
[32:47.200 --> 32:49.320]  Do it while the iron is hot.
[32:50.280 --> 33:02.660]  So I recommend, actually, either through GitHub, or some other file sharing system, or at least some annotating system, convey what errors you find as soon as you find them, as soon as you possibly can.
[33:02.660 --> 33:07.060]  Because the next recording session, someone might come in and say, oh, shoot, I have to fix five things.
[33:07.140 --> 33:09.480]  Well, let's take care of that before I get to, you know, the rest of the work.
[33:17.140 --> 33:24.760]  Once you've taken care of QC, and there are no fixes left, you might want to run through another pair of ears just to listen to it.
[33:25.140 --> 33:29.900]  See if you can hear when different recording sessions took place.
[33:31.040 --> 33:35.640]  At the beginning of a talk, or at the beginning of a lecture, or at the beginning of a recording, you come in with a whole bunch of energy.
[33:35.840 --> 33:36.680]  You're ready to read.
[33:36.680 --> 33:37.660]  You are ready to go.
[33:38.120 --> 33:43.640]  And then after about an hour, you start to talk a little slower and lose energy.
[33:44.560 --> 33:50.780]  So if you make a flub, excuse me, error, while you are talking slowly, that's what it's going to sound like.
[33:50.820 --> 33:52.160]  It's going to sound disjointed.
[33:53.240 --> 33:56.040]  When you make a mistake, you don't want to call attention to the mistake you made.
[33:56.160 --> 34:04.340]  You know, if you say, for example, in the beginning, dog made heaven and earth, and you go back and you have to fix it and you say, in the beginning, God made the heaven and earth.
[34:05.360 --> 34:09.200]  And your listener is going to say, why is he emphasizing that word?
[34:09.390 --> 34:10.760]  He must have screwed up.
[34:11.930 --> 34:14.280]  So you want to be aware of your continuity also.
[34:19.000 --> 34:26.720]  In studio recordings, and that's sort of the on-site system that we have, the narrator and the director work as a team.
[34:26.980 --> 34:27.820]  They really do.
[34:28.080 --> 34:33.660]  The engineer and the director are immediately able to provide quality assurance.
[34:34.540 --> 34:36.740]  Now, quality assurance is the prevention of problems.
[34:36.920 --> 34:39.660]  Quality control is detecting problems.
[34:40.740 --> 34:46.220]  When an engineer and director work together, they're able to prevent problems from recurring.
[34:46.460 --> 34:47.580]  They address it on the spot.
[34:48.480 --> 34:50.440]  This means you don't have to go back and re-record.
[34:51.840 --> 34:55.120]  The director and quality control cooperate with each other.
[34:56.320 --> 35:04.300]  Quality control will say, this has to be re-read, or I heard a sneeze here, or the guy lip slacks in the middle of, you know, page five.
[35:04.580 --> 35:08.000]  And you can deal with it as soon as the narrator comes back.
[35:08.760 --> 35:11.200]  And the errors are caught in a timely manner.
[35:11.300 --> 35:12.640]  You don't have to wait weeks for it.
[35:13.440 --> 35:24.760]  Generally speaking, about 80% of the books to 90% of the books that we record in studio as a group, on-premises, continue on to post-production.
[35:26.220 --> 35:27.820]  It's a very high success rate.
[35:29.340 --> 35:32.320]  Road sessions, on the other hand, are a little different.
[35:32.960 --> 35:35.980]  About 80% of the books fail to make it out of production.
[35:36.260 --> 35:37.660]  And here are the reasons why.
[35:38.200 --> 35:42.200]  If you're doing this solo, there's no director to sustain quality assurance.
[35:42.200 --> 35:45.760]  You have to rely on your own bias to decide if you've done the right job.
[35:46.300 --> 35:48.920]  There's no engineer to maintain technological assets.
[35:49.140 --> 35:53.000]  In other words, you really can't keep an eye on whether the computer is in good shape or not.
[35:54.280 --> 35:56.860]  There's no quality control because the deadline is more important.
[35:58.040 --> 36:03.680]  Sometimes you're on a deadline, and you're on a schedule, and you have to make the schedule, and you have to sacrifice, and that sacrifice happens to be the quality control.
[36:03.680 --> 36:06.400]  So occasionally you hear a book with a burp in it.
[36:06.940 --> 36:07.740]  It happens.
[36:08.800 --> 36:09.720]  Lack of communication.
[36:09.740 --> 36:12.960]  If you've got people who are supposed to be taking their part, they don't respond back.
[36:13.240 --> 36:14.860]  We suddenly don't trust these people.
[36:15.440 --> 36:16.360]  Life happens.
[36:16.820 --> 36:17.260]  You know?
[36:17.980 --> 36:19.120]  People lose interest.
[36:20.080 --> 36:22.460]  And if they lose interest, they drop out of the project.
[36:23.200 --> 36:34.540]  So to quote someone who I remember way back in, you know, back when the year started with 1980, don't allow the task ahead of you to be greater than the enthusiasm behind you.
[36:35.540 --> 36:42.640]  Like, I don't want to scare anyone from this project, but if you're going to commit to this, commit to it enthusiastically knowing what you are dealing with.
[36:43.660 --> 36:46.000]  Recording at home has a very high failure rate.
[36:46.760 --> 36:52.300]  Some of the books that I've heard from LibriVox would never have passed quality control for an NLS standing.
[36:52.560 --> 37:00.020]  In fact, some of the books from Audible that I've heard, even though I'm not technically a client, they would not pass for the highest standards.
[37:00.480 --> 37:03.920]  For example, you need 48 decibels of signal-to-noise ratio minimum.
[37:05.100 --> 37:05.580]  Minimum.
[37:06.680 --> 37:12.200]  If I were to take an audio recording sample right now of this room, you would see a small flat line.
[37:12.560 --> 37:14.780]  You would think that the air is clear.
[37:15.120 --> 37:20.340]  But when you actually measure the number of decibels, no, this sounds a bit more like 42 dB.
[37:21.760 --> 37:22.660]  Sorry about that.
[37:22.660 --> 37:24.160]  I don't know why I'm going ding.
[37:28.480 --> 37:32.590]  But participants will drop out of the project, you know, if they don't have a success rate.
[37:33.070 --> 37:34.300]  People will be turned down.
[37:34.690 --> 37:36.590]  And also, you need qualitative voices.
[37:37.800 --> 37:50.320]  Now, there is one caveat to all of this, which is that books narrated from home, using home studios, tend to pass the rudimentary quality control requirements, but tend to require greater post-production.
[37:50.730 --> 37:54.960]  That is to say, if you have a home studio, odds are you know what you're doing to begin with.
[37:55.980 --> 37:57.320]  So, go on.
[37:58.090 --> 38:04.440]  But because there's no one there with an unbiased ear, you may have to do more fixes later on.
[38:08.720 --> 38:11.060]  Therefore, I don't want to say don't do it, but do it.
[38:11.320 --> 38:13.900]  If you have talent and you want to share it, do so.
[38:15.940 --> 38:24.280]  But I can't tell you how many times I've heard this in the middle of a book from LibriBox, where I can tell that the narrator has a home studio.
[38:25.580 --> 38:27.100]  It'll just piss off your engineers.
[38:27.460 --> 38:33.980]  So, post-production takes a lot longer while you try to remove those defects, rather than focusing on the content.
[38:36.280 --> 38:42.060]  And for those who are recording from home, by the way, your success rate is pretty much 60%.
[38:42.060 --> 38:43.440]  I mean, it's not bad.
[38:43.540 --> 38:45.180]  It just dies in post-production.
[38:46.740 --> 38:52.140]  So, the final slide is the do's and don'ts.
[38:52.200 --> 38:53.600]  And there are quite a few here.
[38:54.500 --> 39:00.740]  So, I want to start off with what a workflow would sound like if you were doing this at home for your first time.
[39:00.740 --> 39:02.180]  Maybe you're working in a team.
[39:02.420 --> 39:05.400]  Let's just name a book that you happen to think of off the top of your head.
[39:06.300 --> 39:06.460]  Anyone?
[39:08.880 --> 39:09.240]  Okay.
[39:09.520 --> 39:10.040]  No participation.
[39:10.220 --> 39:10.380]  Fine.
[39:12.180 --> 39:14.420]  Whatever book you have, you're going to sit down.
[39:14.480 --> 39:15.020]  You're going to record.
[39:15.380 --> 39:16.740]  You've got your microphone in front of you.
[39:16.780 --> 39:20.060]  You want to make sure that the microphone is exactly where your voice is.
[39:20.140 --> 39:24.500]  You'll notice that if I talk down here with the microphone above me, you're not hearing anything.
[39:24.920 --> 39:27.680]  If I talk up here, you're hearing my voice very well.
[39:27.860 --> 39:28.040]  Why?
[39:28.140 --> 39:30.100]  Because I happen to know my speech is here.
[39:30.100 --> 39:31.760]  It's concentrated in front of my teeth.
[39:32.040 --> 39:33.300]  A little bit below the jaw.
[39:33.760 --> 39:35.060]  Not off to the side.
[39:35.540 --> 39:36.780]  Not off at an angle.
[39:37.520 --> 39:39.340]  So, I will know where my axis is.
[39:39.620 --> 39:46.020]  If I want to read this screen, for example, I would position the microphone where my voice is going to go.
[39:46.780 --> 39:49.980]  Where I can see the monitor and read.
[39:50.760 --> 39:51.980]  Not off to the side.
[39:53.700 --> 39:55.900]  And project my voice downwards.
[39:56.700 --> 39:59.600]  A lot of streamers, they work like this.
[39:59.800 --> 40:00.040]  Yeah.
[40:00.740 --> 40:02.240]  Why do they work like this?
[40:02.420 --> 40:03.240]  Feel very attacked.
[40:03.380 --> 40:04.540]  You are attacked.
[40:04.880 --> 40:05.520]  And you know why.
[40:05.980 --> 40:11.760]  Because you are using a cardioid microphone where you expect people to talk upwards.
[40:11.760 --> 40:13.660]  The focus is not on reading the screen.
[40:13.920 --> 40:15.920]  The focus is on looking at others.
[40:17.960 --> 40:19.960]  It's at looking at the camera, at the zoom.
[40:20.360 --> 40:22.360]  And who cares if there is a microphone in your way.
[40:22.680 --> 40:23.680]  You are looking at a monitor.
[40:23.840 --> 40:24.580]  You are showing off.
[40:24.720 --> 40:24.780]  Right?
[40:25.240 --> 40:26.100]  Damn, I look good.
[40:28.260 --> 40:30.760]  Realistically, you want the microphone beneath you.
[40:30.760 --> 40:34.460]  You are not going to be screaming and orating into the microphone.
[40:35.000 --> 40:36.620]  You are just going to be reading off a screen.
[40:37.180 --> 40:38.040]  Do's and don'ts.
[40:38.640 --> 40:40.080]  This section is focused on narration.
[40:40.980 --> 40:46.440]  Tips for directors, engineers, post-production, markup, and collaboration will be found at...
[40:47.340 --> 40:48.320]  And you just read.
[40:49.520 --> 40:59.720]  When I do my sections at home or if I am on the road or if I am giving new talent advice, I usually say start your recording thusly.
[41:00.260 --> 41:01.280]  Name the book.
[41:02.020 --> 41:03.120]  Who is the author?
[41:03.580 --> 41:04.600]  Who is reading?
[41:04.780 --> 41:05.320]  Who is directing?
[41:07.780 --> 41:12.700]  Put in the day and time that you began your recording.
[41:13.240 --> 41:14.320]  Very important.
[41:14.600 --> 41:16.380]  Because it will help you keep a log.
[41:17.020 --> 41:19.620]  It will tell you how much time and effort you have put into work.
[41:20.980 --> 41:23.760]  And then shut the hell up for ten seconds.
[41:32.920 --> 41:34.120]  Why would you do that?
[41:34.360 --> 41:38.700]  The answer is you want the microphone to actually pick up all the background noise.
[41:39.560 --> 41:46.400]  As an engineer, we have worked with a lot of books that have recorded back in the 1980s, 1970s, 1990s.
[41:46.620 --> 41:48.060]  All sorts of background noises.
[41:48.660 --> 41:57.540]  If you have a good sample of the noise floor and the room tone where you are recording, Audacity has this nice little feature called noise reduction.
[41:57.900 --> 41:58.920]  We love it.
[41:59.400 --> 42:04.320]  But in order to work properly, you have to give it a sample of noise.
[42:04.840 --> 42:07.780]  Don't make me extract it from snippets of audio.
[42:08.440 --> 42:10.320]  Give me the sample at the beginning.
[42:11.720 --> 42:17.340]  So, you do your intro, you walk away from the mic for a few seconds, and then you begin recording.
[42:18.200 --> 42:21.640]  So that's a top notch pet peeve, as you can tell.
[42:22.580 --> 42:26.860]  Another one, ride the pause.
[42:27.440 --> 42:32.940]  If you have a book and you're going to have paragraphs and sentences, read the frickin' punctuation.
[42:33.100 --> 42:34.300]  I mean, you're not Christopher Walken.
[42:34.700 --> 42:41.020]  Even Christopher Walken would respect the notion of a sentence is a logical unit of information.
[42:42.380 --> 42:48.580]  If you have a list, A, B, C, and D, ride the commas.
[42:49.160 --> 42:50.280]  Why would you do that?
[42:50.400 --> 42:55.200]  Well, for one thing, it makes it a lot easier to pick up in case you make a mistake.
[42:56.280 --> 43:00.040]  If I say, you know, lettuce, tomato, onions, underwear.
[43:01.100 --> 43:05.620]  I can easily cut after tomatoes and replace underwear with cheese.
[43:07.220 --> 43:09.220]  If you have lettuce, tomato, onions, and underwear.
[43:09.720 --> 43:09.900]  Cheese.
[43:10.820 --> 43:15.320]  I don't have enough gap as an editor to separate the cheese from the underwear.
[43:16.840 --> 43:17.680]  Ride the pause.
[43:17.680 --> 43:19.320]  You can always tighten up a pause.
[43:19.600 --> 43:24.340]  You can't stretch the two apart without sounding processed.
[43:25.460 --> 43:28.100]  Speaking of sounding processed, compression.
[43:28.580 --> 43:29.840]  A lot of people use compression.
[43:29.840 --> 43:34.420]  They try to rise the levels, make it sound a more full sound of audio.
[43:34.880 --> 43:36.400]  Guys, this is not music.
[43:36.660 --> 43:40.860]  If you adjust the human voice, it's going to be audible.
[43:41.000 --> 43:41.620]  It's going to be noticeable.
[43:41.800 --> 43:45.100]  It's going to become the uncanny valley of your eardrums.
[43:46.200 --> 43:49.040]  So, it'll sound a little more like auto-tune.
[43:49.420 --> 43:50.220]  And auto-tune...
[43:50.220 --> 43:50.540]  I know it's...
[43:50.540 --> 43:51.440]  I know it's 2022.
[43:51.440 --> 43:54.040]  I know we have really good deep fakes with audio.
[43:54.400 --> 43:55.420]  But if...
[43:55.420 --> 43:55.960]  No.
[43:56.100 --> 43:57.100]  It's not going to sound right.
[43:57.220 --> 43:58.460]  It's going to sound absolutely wrong.
[43:58.980 --> 44:01.440]  It actually is horrific.
[44:01.960 --> 44:03.520]  So, do not.
[44:03.700 --> 44:05.400]  Do not use compression.
[44:05.660 --> 44:10.120]  Try to prevent needing compression in the first place by getting your levels high.
[44:10.120 --> 44:16.860]  You want to keep it between negative 6 to, let's say, negative 12 decibels.
[44:17.140 --> 44:19.700]  It's kind of hard to watch because you need a view meter.
[44:19.920 --> 44:21.740]  And you're trying to focus on reading.
[44:23.080 --> 44:24.140]  Speaking of reading.
[44:25.740 --> 44:26.380]  Please.
[44:27.500 --> 44:28.140]  Please.
[44:29.160 --> 44:32.640]  Read the damn book before you record.
[44:33.640 --> 44:34.280]  Please.
[44:34.800 --> 44:36.200]  I beg you.
[44:37.100 --> 44:41.240]  Because if you don't read the book ahead of time, and you're reading cold, you...
[44:41.240 --> 44:41.940]  It will show.
[44:42.100 --> 44:42.620]  It will sound.
[44:42.740 --> 44:44.060]  It will be a mechanical reading.
[44:45.140 --> 44:45.860]  Be familiar.
[44:46.160 --> 44:46.760]  Not just familiar.
[44:46.900 --> 44:48.720]  Be authoritative in what you read.
[44:49.540 --> 44:57.620]  People are relying on you, and you, and you, and you, and you, and you, and you, and you, to present the book like a finished product.
[44:57.620 --> 44:59.600]  Not like, I'm just learning this for the first time.
[44:59.600 --> 45:00.340]  Excuse me.
[45:01.620 --> 45:03.020]  Just read the book.
[45:03.740 --> 45:06.060]  And then read it to yourself aloud.
[45:07.500 --> 45:09.680]  You need to know what words you're going to trip over.
[45:09.920 --> 45:11.880]  What pronunciation guides you have.
[45:12.720 --> 45:13.820]  If you don't know...
[45:13.820 --> 45:19.920]  If you are so naive that you think it's written, it's pronounced knave, don't read the book.
[45:21.120 --> 45:24.620]  Because if you pronounce naive as knave all throughout, guess what?
[45:24.620 --> 45:27.040]  There is no such thing as find and replace in audio.
[45:28.740 --> 45:33.660]  And not only that, sometimes you're going to say, naive, naive, naive.
[45:34.040 --> 45:37.560]  And you're going to replace it with naive, naive.
[45:39.520 --> 45:42.700]  Gee, Jim was sounding very naive all throughout.
[45:44.560 --> 45:46.100]  Don't make your edits noticeable.
[45:47.140 --> 45:53.340]  If you screw up a sentence and you have to go back to re-record it, don't just record the word you screwed up or the sentence.
[45:53.340 --> 45:54.860]  Record a sentence before.
[45:55.260 --> 45:57.180]  Record a sentence after.
[45:58.000 --> 46:00.380]  Try to make it dovetail with the correction.
[46:00.920 --> 46:02.780]  This is Audio Engineering 101.
[46:03.200 --> 46:04.640]  And it's very easy to do.
[46:04.780 --> 46:12.000]  And if you ride the pauses between sentences, you don't need to find where those words are.
[46:12.860 --> 46:15.580]  So, for example, if I have a piece of audio...
[46:15.580 --> 46:21.700]  Unfortunately, I do not have an example on me, but if you follow up on GitHub later on, probably next week, I'll give you a prime example.
[46:22.480 --> 46:35.340]  Where the audio goes on for 24 hours, because some books go on for 24 or even 36 hours straight, I have to, as an engineer, without a copy of my book, find where the corrections have to be made.
[46:36.400 --> 46:41.880]  If you give a five-second pause between chapters, you have just made my life so much easier.
[46:42.260 --> 46:43.960]  Because all I have to do is look for the big gaps.
[46:44.800 --> 46:46.800]  One, two, three, four, five...
[46:46.800 --> 46:47.960]  That's got to be chapter five.
[46:48.120 --> 46:48.880]  Chapter five.
[46:49.280 --> 46:52.080]  Now I can narrow down my search for what needs to be corrected.
[46:53.320 --> 46:56.860]  When you go from paragraph to paragraph, take a breath.
[46:57.640 --> 47:00.860]  You don't have to read a sentence in a single lungful.
[47:02.620 --> 47:04.180]  You have the ability...
[47:04.180 --> 47:05.280]  I mean, we're not doing this live.
[47:05.540 --> 47:12.480]  What we are doing instead is committing to tape, or in the case of the 21st century, committing to computers.
[47:13.580 --> 47:16.000]  You have so much storage capacity.
[47:17.440 --> 47:18.740]  Use it to your benefit.
[47:19.360 --> 47:20.420]  Take it slow.
[47:22.040 --> 47:25.440]  That will speed up the recording, the post-production, the corrections.
[47:25.980 --> 47:27.380]  Your levels will be good.
[47:28.760 --> 47:30.660]  Ultimately, you will be published.
[47:31.800 --> 47:34.420]  But you have to keep your standards high at all times.
[47:34.700 --> 47:41.800]  You have to make the investment in the equipment, the time, the effort, the ability to take criticism.
[47:42.540 --> 47:47.300]  Like if somebody is very nasal, you got to go back and say, you have to rerecord this chapter.
[47:47.500 --> 47:49.700]  You had a booger up your nose the whole time.
[47:50.560 --> 47:53.020]  And take that criticism realistically.
[47:54.480 --> 47:57.240]  There are some people who will read sentences mechanically.
[47:58.360 --> 48:01.380]  They have a distinct pattern to the way that they speak.
[48:01.880 --> 48:04.420]  You are hearing an example of it right now.
[48:05.060 --> 48:06.720]  This is a major turn-off.
[48:08.040 --> 48:14.480]  You need to sound as normal, conversational, as you could possibly be.
[48:15.120 --> 48:17.160]  There are a few times when you have to exaggerate.
[48:17.240 --> 48:20.120]  I mean, if you see an exclamation point, you have to exclaim!
[48:21.020 --> 48:24.400]  If there's a question mark, can you ask the question?
[48:26.080 --> 48:29.400]  And especially, finally, because this is one...
[48:29.400 --> 48:30.120]  This is a big problem.
[48:31.240 --> 48:33.940]  There are a lot of books that are read mechanically by computers.
[48:34.200 --> 48:36.000]  Now, like I said, this is 2022.
[48:36.540 --> 48:37.540]  Deep fakes exist.
[48:37.980 --> 48:45.640]  There are great ways of futzing with audio to make things sound like what someone said, even though they never said it, because it's a computer-generated thing, but for the love of...
[48:45.640 --> 48:46.160]  No.
[48:47.160 --> 48:47.980]  If I give...
[48:47.980 --> 48:49.300]  Who here watches PBS?
[48:49.520 --> 48:51.020]  Are you familiar with the phrase, viewers like you?
[48:52.180 --> 48:53.020]  Okay, okay.
[48:53.280 --> 48:55.020]  A cold reading of viewers like you.
[48:56.180 --> 48:57.020]  Where's the emphasis?
[48:57.020 --> 49:00.860]  If I say, viewers like you...
[49:00.860 --> 49:01.500]  Great!
[49:02.940 --> 49:05.280]  If viewers like you...
[49:05.280 --> 49:05.840]  Great!
[49:06.680 --> 49:08.020]  Viewers like you...
[49:09.040 --> 49:12.740]  These are three different readings of the exact same phrase.
[49:13.020 --> 49:16.000]  If I say, I love you...
[49:16.000 --> 49:17.340]  I love you.
[49:17.740 --> 49:18.880]  I love you.
[49:19.880 --> 49:20.060]  I love you.
[49:20.180 --> 49:22.860]  Three different readings of the exact same thing.
[49:23.140 --> 49:26.100]  Computers are shit at figuring out what the emphasis should be.
[49:26.700 --> 49:28.080]  But for humans, it's normal.
[49:28.200 --> 49:29.040]  Especially in English.
[49:29.760 --> 49:34.800]  We emphasize the word that needs to be dressed, to be brought to our attention.
[49:36.380 --> 49:36.940]  Emphasis.
[49:37.380 --> 49:39.180]  Something that's not really good for computers.
[49:39.520 --> 49:46.740]  So, please, if I ever hear that you ran any piece of text through text-to-speech, especially that TikTok voice.
[49:46.980 --> 49:47.960]  I'm gonna kill you.
[49:48.560 --> 49:49.480]  I'm gonna kill you.
[49:49.540 --> 49:50.580]  I'm gonna beat you over the head.
[49:50.760 --> 49:51.980]  I'm gonna take away your microphone.
[49:52.300 --> 49:54.680]  But most of all, we're not there yet.
[49:55.280 --> 49:56.780]  AI is not that good yet.
[49:56.980 --> 49:58.740]  And it's gonna take more years for that.
[49:59.280 --> 50:05.380]  So, for the human nuance, and also for the familiarity, we still need human voices.
[50:06.180 --> 50:09.900]  And I'd like you to join in this quest to become the next set of voices.
[50:10.560 --> 50:11.920]  You know what you gotta do.
[50:11.960 --> 50:13.480]  You know what you shouldn't do.
[50:14.100 --> 50:15.480]  Thank you for being here.
[50:15.740 --> 50:20.780]  And I hope you enjoy trying to become the next set of voices reading to someone.
[50:21.080 --> 50:23.300]  Because it really does change lives.
[50:23.860 --> 50:27.240]  You are being a gap filler.
[50:27.540 --> 50:30.400]  We have a great degree of illiteracy in this world still.
[50:30.640 --> 50:33.520]  I mean, it is the 21st century, but illiteracy is an issue.
[50:34.420 --> 50:40.920]  When you become a book for someone, suddenly it doesn't matter if they're illiterate.
[50:41.580 --> 50:43.380]  They're reading by their ears.
[50:43.880 --> 50:46.380]  And you will be writing for the ear.
[50:46.940 --> 50:47.260]  Thank you.
[50:53.800 --> 50:56.240]  And I see we've made it exactly on time.
[50:57.040 --> 50:57.520]  So...
[50:57.520 --> 51:00.340]  If you have any questions, come over here to this microphone.
[51:03.930 --> 51:04.410]  Wow.
[51:04.410 --> 51:06.210]  Did you render everyone's speeches again?
[51:06.650 --> 51:07.610]  I think I did.
[51:07.810 --> 51:08.410]  I mean...
[51:13.640 --> 51:14.960]  I don't think I have that much cord.
[51:15.540 --> 51:17.560]  Well, there's not a lot of questions, but let's find out.
[51:25.890 --> 51:27.510]  Now that we've offered the opportunity.
[51:27.510 --> 51:28.310]  Right.
[51:30.350 --> 51:31.610]  I am an audible...
[51:34.290 --> 51:35.270]  What do you want to call me?
[51:35.490 --> 51:35.930]  Subscriber.
[51:36.070 --> 51:36.910]  I've got hundreds.
[51:37.930 --> 51:44.790]  One of the things that I've noticed that I used to just pull off books, but then I found read by author.
[51:45.790 --> 51:47.970]  And oh my God, does that make a difference.
[51:48.250 --> 51:51.170]  I happen to prefer works that are read by the author.
[51:51.890 --> 51:55.910]  This is why I make a great enjoyment, actually, of Borges.
[51:56.310 --> 52:04.130]  While he was in the United States for a fairly brief period, he did give lectures with the Smithsonian and the Library of Congress.
[52:04.130 --> 52:05.330]  He read some of his own works.
[52:05.430 --> 52:10.970]  A lot of his recitals I've committed to heart better through the ear than through reading by eye.
[52:11.310 --> 52:19.830]  One of the things that I liked was I noticed that the author would emphasize a phrase or a situation different than a cold reading.
[52:20.330 --> 52:23.430]  Not only a cold reading, but even in translation.
[52:23.650 --> 52:30.490]  For example, the short story, Borges and I, which is sort of an autobiographical joke or jest.
[52:31.110 --> 52:33.550]  The English translation is a fairly good read.
[52:33.710 --> 52:35.630]  I believe it's available from Simon & Schuster.
[52:36.130 --> 52:40.970]  But then when you hear it in the original Spanish, or actually I think he wrote it in English originally.
[52:40.970 --> 52:41.570]  I'm not sure.
[52:42.530 --> 52:44.150]  But this is the problem with Borges.
[52:44.250 --> 52:46.690]  He spoke too many languages and just wrote prolifically.
[52:46.690 --> 52:52.290]  But in his own reading, he emphasized, and I'm going to try to get this right.
[52:52.290 --> 52:54.330]  I'll have an example up on the GitHub pages.
[52:55.970 --> 53:00.790]  He ends, No se cual de los dos escribe esta página.
[53:02.310 --> 53:08.350]  So, I do not know which one of the two, of the two Borgeses, writes this page.
[53:09.050 --> 53:10.450]  Sort of a matter of fact.
[53:11.310 --> 53:17.910]  The translation that came from Simon & Schuster was, I'm not sure which one of the two reads this page.
[53:18.950 --> 53:20.470]  Those are very different readings.
[53:21.630 --> 53:24.490]  You're not there to inflect for the author.
[53:24.750 --> 53:28.030]  You're there to convey what the author has already said.
[53:29.170 --> 53:32.810]  The other issue is, there's a lot of music thrown in for, I don't know why.
[53:32.810 --> 53:34.430]  I think it's just for decor.
[53:40.720 --> 53:41.300]  Challenges?
[53:41.520 --> 53:41.860]  Questions?
[53:42.200 --> 53:42.540]  Comments?
[53:42.700 --> 53:43.140]  Cvetches?
[53:44.000 --> 53:44.580]  Coffee?
[53:45.620 --> 53:51.700]  You were talking about including metadata at the start of your recordings, before that ten second gap.
[53:52.980 --> 53:58.500]  One thing I wanted to add to that, because this is something that I experienced in the last few months.
[53:59.480 --> 54:05.940]  As you do more of this, you're going to struggle to like the sound of the recordings that you make.
[54:06.340 --> 54:15.540]  And eventually, if you own a lot of equipment and experiment with different microphones, you might accidentally create one that you really like the sound of.
[54:16.660 --> 54:24.820]  And there's nothing worse than going back over your work, and discovering that, and realizing that you have no information about how you did that.
[54:24.820 --> 54:38.800]  So, when you are including that metadata at the start of your recordings, you might also want to include things like where you positioned the microphone, which microphone you were using, what your general analog equipment settings were.
[54:39.040 --> 54:41.700]  Because that would have saved me a lot of time, personally.
[54:42.380 --> 54:42.920]  Absolutely.
[54:42.920 --> 54:48.020]  When you start a project, keep your hardware and your software and your positioning consistent.
[54:48.600 --> 54:50.320]  This is environmental factors, obviously.
[54:51.600 --> 54:57.020]  The other thing that is worth mentioning is, when you finish a recording, don't just, you know, hit cut and you're done.
[54:58.820 --> 55:04.300]  Take five seconds, and just say, I'm finishing at such and such a time, on such and such a page.
[55:04.620 --> 55:08.020]  It makes it easier to start the next section as you go along.
[55:09.180 --> 55:12.760]  Also, one thing that I think I forgot to mention, but record natural.
[55:13.520 --> 55:16.600]  Don't use EQ as you're recording.
[55:16.600 --> 55:20.880]  That you can do a little more finessed in post-production than you can live.
[55:21.100 --> 55:23.520]  And you can keep it consistent all throughout your work.
[55:23.700 --> 55:28.860]  So anybody, I don't know if anybody listens to the Brian Lehrer show, because on WMAC he's got that nice deep voice.
[55:29.020 --> 55:32.040]  But then you hear him on television, he's got this little titty voice up here.
[55:32.480 --> 55:35.220]  That's because the EQ is being ridden.
[55:36.000 --> 55:36.920]  Record flat.
[55:37.240 --> 55:38.040]  Record natural.
[55:39.200 --> 55:46.230]  You know, I noticed that you recommended FLAC as the file sharing kind of audio format.
[55:47.120 --> 55:48.310]  Why not WAV?
[55:49.080 --> 55:52.180]  WAV is a very large file format.
[55:52.740 --> 55:56.780]  So, I mean, just to start with, one CD is 650 megs.
[55:56.920 --> 55:58.190]  That should get you about...
[55:58.740 --> 55:59.560]  What was it?
[56:00.340 --> 56:01.860]  I think it was about...
[56:01.860 --> 56:02.400]  Sorry.
[56:02.820 --> 56:04.480]  Let me start from just another direction.
[56:05.340 --> 56:10.610]  Let's talk about the two gig file limit, because some DOS programs are still in use that do audio adjustments.
[56:11.440 --> 56:13.240]  That should get you about six hours of audio.
[56:14.230 --> 56:20.640]  That six hour of audio at two gigs would fall down to about a gig and a quarter as a FLAC.
[56:20.860 --> 56:22.960]  And also, FLAC is lossless.
[56:23.280 --> 56:25.260]  You're not losing quality along the way.
[56:26.240 --> 56:30.980]  So, you're shaving off about 750 megs of transmission on a two gig file.
[56:30.980 --> 56:35.700]  And also, the other thing is FLAC is now pretty much well supported by everybody.
[56:36.260 --> 56:38.080]  OGG is not as well supported.
[56:38.660 --> 56:41.500]  So, FLAC can be recommended at that point.
[56:42.880 --> 56:49.720]  But, basically, I tested against Tarek zipping WAV files and FLAC and they're about the same.
[56:56.590 --> 56:57.650]  Now I'm in trouble.
[57:05.720 --> 57:22.080]  So, I think you kind of very briefly touched on this, but my background is not really an accessibility thing, unfortunately, to some degree, is that, like, as someone who was a big Star Wars fan for a while with a lot of their book series starting with the original audio books,
[57:22.080 --> 57:37.600]  because it's based on a movie series, they often, instead of just relaying or just excitingly read the information, they would do, as you say, make it an audio experience where they would add in the movies, reuse the movie soundtracks, put in sound effects.
[57:37.740 --> 57:39.480]  That's no longer an audio book at that point.
[57:39.680 --> 57:50.660]  That used to be the thing that was kind of unique to Star Wars and some sci-fi, but now there's just pure audio projects where basically they're bringing back radio serials instead of trying to, like, put out information.
[57:50.660 --> 57:54.140]  I just wanted to know if you wanted to elaborate, like, that sort of proliferation.
[57:54.400 --> 57:59.780]  Do you see that as, like, muddying the waters of the type of work you're trying to do, or is it something where it's become now more accessible?
[57:59.780 --> 58:01.080]  I'll put in another parallel.
[58:01.300 --> 58:06.640]  It's like saying the difference between Watchmen the movie and Watchmen the graphic novel.
[58:08.040 --> 58:12.640]  Whenever you transmute one medium into another, there's a certain adaptation that has to take place.
[58:13.920 --> 58:19.620]  What the Star Wars series is doing is being entertainment, or a continuation of the entertainment as if it were a radio play.
[58:19.620 --> 58:26.320]  If you were reading a book, then it is superfluous and you're distracting from the message.
[58:27.300 --> 58:34.120]  There is one last thing I am going to add to this because I don't think I covered this quite in my slide deck, but I want to talk about foreign language.
[58:34.660 --> 58:35.260]  Excuse me.
[58:35.720 --> 58:36.920]  Foreign language.
[58:37.300 --> 58:38.760]  Because this is the United States.
[58:38.900 --> 58:40.520]  Half of all homes speak another language.
[58:41.900 --> 58:46.260]  There is a big difference between a native speaker and a native reader.
[58:47.960 --> 58:52.460]  You can be very... you can be proficient in a language and still flub it.
[58:52.540 --> 58:55.280]  And if you don't believe me, how many of you were here for Hurricane Sandy?
[58:56.200 --> 59:00.880]  And how many people remember Michael Bloomberg's god-awful, atrocious Spanish?
[59:03.000 --> 59:05.960]  That's the difference... that's sort of the difference I'm going at.
[59:06.180 --> 59:08.860]  You can be very proficient, but sound horrific.
[59:09.400 --> 59:11.700]  If you have an accent, it may become a problem.
[59:12.220 --> 59:18.700]  And then finally, because I see the time is going, native speakers don't necessarily be...
[59:19.460 --> 59:22.360]  they don't necessarily become native readers.
[59:22.860 --> 59:24.180]  You need fluency.
[59:24.460 --> 59:29.960]  So if you're not fluent in a language and can't read on a fluent level, let someone else do it.
[59:30.180 --> 59:34.060]  You can help on the editing, QC, just don't narrate yourself.
[59:35.780 --> 59:37.640]  All right, I think we are out of time at this point.
[59:37.760 --> 59:39.640]  We are out of time, so thank you very much.
[59:39.640 --> 59:42.580]  I know Cory Doctorow is still probably spieling as we speak.
[59:42.940 --> 59:45.640]  And for you to come here, I really appreciate it.
[59:45.740 --> 59:46.360]  So thank you all.
[59:52.980 --> 59:56.080]  We'll be just a few minutes to set up for this next time.