To OtsLabs friends: the features that radio guys expect

[milky]

I use the Aussie sounding "Nicole" on my home entertainment system, simply to tell me the name of a track if I'm outside and someone asks, "Who is this?" Whe is the best I've heard, but the announcements are still bland and often the inflections are wrong. A presenter needs to be spontaneous and, even when voice-tracking is used, sound enthusiastic. I couldn't imagine Nicole exhorting the multitudes to flock down to the car wash/fundraiser.

[JAylmer]

I use "Nicole" mostly also and agree with your comments, languages are complex.
Lane calls it a performance, and yes "Nicole" can't do that.  Would need an inflection parameter for each sentence, but it still would not be 100% and it doesn't answer mpanda'a real time considerations.  And Jeffs comments "You can still tell its not real", that's the yardstick, so they still have a way to go

I would be interested at a technical level to hear how you did that with your home system.  Is that using OTS, Winamp etc?

I use a mix of myself and Nicole, which helps to overcome the shortcomings of each other.

[milky]

I would be interested at a technical level to hear how you did that with your home system.  Is that using OTS, Winamp etc?

I use Ots exclusively. It works this way. 1) S&L logging is active, so each track that plays creates an entry in the daily log folder.
2) I have a "folder watcher" program which launches a VB script as soon as it "sees" the new entry and isolates the track and artist from the log entry. This happens each time a new track is launched.
3) I use an X10 wireless remote and AutoHotKey (the "L" version) script to send Play, Stop, Pause, Forward, Previous and "Announce" commands.
4) The Announce option pauses the music, launches Nicole who reads the Artist/Track line created in step 2, and then the music continues.

[JAylmer]

Excellent!

[Audioot]

Hm, that (http://www.ivona.com) won't work with foreign languages at all. Especially since songs are mostly English in titles (and bands) and those will then be attempted in the local language .. .
For the Dutch : try this :" Zometeen: de pretenders met I go to sleep."  That doesn't sound right.. does it?

[Dj Buik]

Hm, that (http://www.ivona.com) won't work with foreign languages at all. Especially since songs are mostly English in titles (and bands) and those will then be attempted in the local language .. .
For the Dutch : try this :" Zometeen: de pretenders met I go to sleep."  That doesn't sound right.. does it?

What you are trying to achieve is mixing two languages in the same sentence. That wil not work, because the Ivona program doesn't know what is Dutch or what is English.
The Ivona application does not have a dictionary, it only pronounces words in the selected language, due to certain program rules.

But everybody in Holland understand the following english sentence: 

Coming up: The Pretenders with I go to sleep

[Lane]

Milky, have you considered having the music keep playing but duck down in volume during the track announcement? Perhaps your preference is not miss any notes!

[milky]

Yes, Lane. That would be my preference (similar to radio), but the only option I could easily program was to pause, announce, and unpause.

[Darryl]

Having a marker (similar to the ACI mix points) for the point where it is safe to start talking at the end of a track along with the vocal start point (i.e. where to stop talking on the way in) plus the associated start/end/length of the talking bit using similar logic to the IntelliFade (R) mixing in OtsAV and it would sound like a real jock live on air talking from the end of one track into the start of the next with the same relatively small effort we currently have setting up our libraries for ABM.  It seems like the next logical step for Ots AV automated mixing to me.

[Lane]

I could live with just the time to post marker (forget what ots is calling it) that has an ACI chunk already. Talking can start at the mix out point and up to to the post.  Third deck needed for this I figure.

[mpanda]

... Third deck needed for this I figure.

The crucial function would be 'overlapping'.
Only by overlapping outro and intro you can run voice tracking.