Hi Greg,

Yes, playing back one of the 1/4 speed tapes back at normal speed should work OK (I'll leave it to you to figure out the amount of pitch correction to be applied). Tone detection is another matter. I've spent about 8 years doing that kind of weird specialized work. It can be a hard nut - if you're trying to make it automated - depending on the quality of the source audio and index tones. On the big automated system I referred to earlier tone detection was done in realtime. This last year I wrote an application called Tone Detective (as a work for hire). That software is designed to work with pretty good quality audio and index tones which stick pretty closely to a published specification. It's being used by one of the libraries as a post-production tool to automatically detect index tones in audio that has already been digitized. Here's link to more information on Tone Detective at the KD web site:

http://www.kafkasdaytime.com/tonedetective/

There's some more or less good news on the katieplayer development front which I will post in the katieblog before long. As you know, katieplayer is self-funded by me and proceeds from my other development work. I believe I now have enough put aside to push through the last development steps to complete katieplayer Cocoa. That is, I have enough to commit the time and pay contributing programmers who have been involved in the project from the beginning. I'm going to host a few "programming retreats" which will put us (few) in a room together for a few days at a time and really bang out the last work. I think that more aggressive approach will bring us to beta testing and release dates much more quickly and probably more inexpensively over the long run. I don't have dates yet but things are looking much better now than they were, say, 2 months ago. Obviously, I'm also comfortable discussing this stuff publicly since I'm posting to the list so please feel free to share this information.

Joe

On Dec 28, 2006, at 10:52 AM, Greg Kearney wrote:

Joe;

Let's say I use one of the NLS players playing back at normal tape speed a NLS book which was recoded at 1/4 speed. Would that work? Also would it be possible to detect the tone indexes on the NLS tapes such that the recording could be broken down into seperate recoding? I'm trying to devise a method of converting NLS books into Daisy recordings.

By the way I have been pushing Daisy playback hard at Apple. I've had a few meeting with the documentation people there about it. Any word on when I can try out your latest work? I'm beta testing 10.5 as well as the new NLS Daisy books now.

Greg
On Dec 28, 2006, at 08:43 , Kafka's Daytime wrote:

Hi Greg,

I built a system for one of the libraries that did/does automated conversion of taped audiobooks to digital. To speed the process significantly, we custom-modified tape machines to play at higher speeds and then pitch-shifted the digitally recorded audio to return to normal speed. I don't recall what the threshold was, but you have to find a happy balance between speed going in (and resulting bandwidth loss) and artifacts resulting from pitch- shifting (i.e. if you go in too fast your audio will get mangled in a too-dramatic pitch shift and bandwidth loss will be too great).

For a home version, you might be able to use one of the Marantz tape players (or Narrator model or similar) to play taped audio in at higher speed - into Audalyze, for instance - and then apply pitch shift to slow audio down.

Ripping at Fast Forward speed will not work as all of the audio's frequency information gets mangled when played at Fast Forward and the Fast Forward speeds will generally be too fast i.e. the factor of pitch shift required and bandwidth lost will be too great.

There is more information in the patent abstract if you care to search USPTO.gov for patent number 6,710,955.

Hope that helps,

Joe

On Dec 28, 2006, at 9:42 AM, Greg Kearney wrote:

While we're on this topic I would like to know if there is a way to rip tapes at fast forward speed and then slow them dow to normal speed in the computer. This would speed up the process of ripping taped books as you would not have to play them into the sound recoding software at the normal speed.

Greg Kearney
On Dec 27, 2006, at 22:19 , louie wrote:

Also you can rip NLS tapes using sound studio.

On Dec 27, 2006, at 9:03 PM, Cheryl Homiak wrote:

Our local libraries here in Madison have a lot more on cassette than on cd as far as books and quite a few of them are abridged, which I hate. but for books which are available this certainly works and in fact, with a line-in setup and audio hijack pro it isn't really difficult to make files of cassettes either.


--
Cheryl

"Where your treasure is, there will your heart be also."














Reply via email to