First episode post-listen comments:
I'm probably not going to listen to this myself, but I appreciate the work that must have gone into it.
Upvoting to acausally allow something indistinguishable from myself in as-yet-undefined ways to listen to this before reading the text version.
Also, thanks!
Have you ever thought “I’d love to read Harry Potter and the Methods of Rationality, but I just don’t have the spare time. I wish it was available in audio format.”
Yes, that is precisely what I think every time HPMoR pops into my head. Thank you.
Considering the addictive effects the story is known to have a audio might just serve as the first hit, and then make you transfer to the text anyway. Sleepless nights are reported.
Do people really consider podcasts to take less time than conventional reading? For me the opposite is true, though I understand how podcasts/audiobooks can be convenient under certain conditions.
I find most spoken audio very comfortable at double speed. Whether this matches reading speed depends on your reading speed. I also find spoken audio less fatiguing than reading for most fiction and most popular nonfiction. I am less prone to being derailed from the book. This is where I read for a bit and then find something else more compelling, such as a blog or a rationalist forum, or where I pick up the book and then put it back down having resolved to continue tomorrow. Even if reading-time is shorter when I read, calendar time is often very long.
Interesting and informative post, thanks! I'll have to try the double-speed method. I suspect that this would still not match my reading speed (which is very fast indeed), but would still make audiobooks and the like more efficient in the cases where they are convenient.
For those who want a program that does this well, I find that "mplayer" is decent at speeding up the things it plays. '[' and ']' alter the speed, and the intial speed can be set with eg. "-speed 2.0" on the command line, though I find that too fast.
One generally wants to correct for pitch changes with the "scaletempo" audio filter. This can be enabled with "af=scaletempo" in a config file, or "-af scaletempo" on the command line.
(One can also use the "mencoder" version of this to create sped-up files for use on other devices).
Anything I could comprehend at double speed would probably not be worth listening to.
Audio may take longer, but perhaps that enables me to remember it better. Any studies on that?
Anything I could comprehend at double speed would probably not be worth listening to.
A few weeks ago I listened to Dawkins's audio reading of Darwin's Origin of Species at double speed.
The simple fact that many people read at double or more speed compared to speech suggests that doubling the speed need not reduce comprehension at all.
There is an optimum speed for comprehension, and for me it seems to be closer to double speed for something like the Origin of Species. To see that slower is not always better, take it to an extreme to make the effect obvious even in imagination. Suppose the audio is at 1/4 speed. At the end of a long sentence, you may have trouble remembering how it started.
If I fail to understand something to my satisfaction or my mind has wandered, I use the "jump back 30 seconds" button once or a few times. Another useful feature.
If you try double speed, at first it is likely to seem overwhelming. But give it a chance. After I have listened to a book at double speed for a long time I find that standard speed tries my patience.
Do you think you remember more than half as much when listening at double speed?
Do you have any programs that can convert an MP3 to double speed?
I have noticed no difference at all in long term retention at different speeds, provided that I understood what was being said initially. In the short term, I find that double speed typically enhances my ability to follow what is being said. I liberally use "jump back" to re-hear and "pause" to mull over something if I find myself wanting to.
I am sure there are programs for the major platforms (PC, Mac, Linux) that can double speed but I use an iPhone/iPod function which doubles playback speed without changing pitch (so it sounds like a human, not like a chipmunk).
There is one major drawback with audio, and that is that I can't highlight. I rely heavily on highlights and notes to retain the contents. However, lack of highlights obviously applies to all audio, not just double speed.
I think I remember more than as much at double speed, because my mind doesn't wander so much. The downside is that when it does wander, the speaker has more often moved on to another point by the time I come back, so I have to rewind.
Like Constant I have not noticed any retention problems at increased playback speeds.
To increase the speed at which you comprehend I recommend using a program that can increment the speed by 10% at a time that way you can ease your way to 2-4 times speed.
The simple fact that many people read at double or more speed compared to speech suggests that doubling the speed need not reduce comprehension at all.
Not if you don't process the information in the same way.
Humans have faster tactile reflexes than visual; you can respond more quickly to a movement that you can feel than to one that you can only see. A person might be able to perform chi sao very quickly, but not be able to respond to attacks of the same speed when they're not in contact with the assailant.
Similarly, I'm pretty sure that I can process written words at a speed at which spoken words would be pure gibberish.
Not if you don't process the information in the same way.
I said suggests, not proves. The point of the word "suggests" is precisely to acknowledge compactly that one can easily raise objections such as the one you raised.
Similarly, I'm pretty sure that I can process written words at a speed at which spoken words would be pure gibberish.
You seem to be referring to speeds at which the phonemes become indistinguishable. But that doesn't happen at double speed, which is what I was talking about. Nor is that the pitfall we were talking about, because the pitfall was that certain content rich texts might be too hard to comprehend at double speed. Rich content isn't a fact about the phonemes.
Tangent: Listening to audiobooks is better practice in identifying some kinds of bad writing.
I often find places in audiobooks that I just can't listen to - I have to replay them 3 or 4 times before I am able to pay attention, because there's one particular sentence that is so boring or devoid of information that my mind always wanders off to something else. I never notice this while reading.
Would it be possible to put this somewhere where iTunes is not required? I would rather not install it.
Yup, this page: http://hpmor.libsyn.com/ It updates every time I publish a new episode and you can download the mp3 directly from it.
Great, thanks! Here is the direct download + rss feed page (referred to as podcast's website in the above link).
I've run into a problem. How do I give you more karma for producing an excellent episode #2?
:)
HA! I had just sent you a message via your blog's contact page and decided to check back here, as it was related, even though I rarely come back to posts more than a week old. Imagine my surprise to find you'd already tried to contact me! For an alternate karma-equivalent idea, please see your site's mail. :)
Thank you! These will be great. I much prefer books to podcasts, but I can't read books while I'm walking to uni :(
Have you ever thought “I’d love to read Harry Potter and the Methods of Rationality, but I just don’t have the spare time. I wish it was available in audio format.” Fret no more! I present to you the HPMoR Podcast! First chapter out now, and another one added every Wednesday.
http://itunes.apple.com/us/podcast/harry-potter-methods-rationality/id431784580