can't you just insert your own custom tags into the wav file?
or use some of the existing chunks?
http://www.borg.com/~jglatt/tech/wave.htm
nctsoft looks very cool, but I'm not willing to pay $300 for that one function. However, I'd probably pay $25 for code or a control that can extract X minutes from a WAV file.
BTW, the plugin DLL system would be for how to detect the song/artist from whatever software is playing (e.g. XM, XM Online, iTunes, etc.)
---Mike
can't you just insert your own custom tags into the wav file?
or use some of the existing chunks?
http://www.borg.com/~jglatt/tech/wave.htm
It does this already. It logs all the songs, and then I have a player that plays backs the WAV while displaying the title and artist (I am adding a big display with big buttons.) The only thing (so far) that I can't do is click on a song and have it save out that portion of the WAV as its own file.Well if your app knows a new song is comming, then have it log each song into a database with the location of the song in the grand scheme of things.
This is probably nicer than a separate log file. I didn't know there were chunks of data one could embed until I visited the nctaudio page.can't you just insert your own custom tags into the wav file?
or use some of the existing chunks?
Note this is mostly for my own use, although I will be releasing it as freeware in case anyone else finds it useful. So, it won't be anything fancy, and I don't want to invest too much time into it.
---Mike
Seems like you're making this harder than it really needs to be by trying to stick to the way you're currently doing things. Wouldn't it be easier to break the files up while recording, then use a database to link them and play them in the order they were recorded. This would let you replay the whole night in the order it was recorded if you wished, then if you chose to save a segment of the night, you then just move copy that segment from the nights location to whatever location you choose to save it in. This will help with system resources as well, because you can mantain a small database and work with small audio files much easier than you can work with one large audio file and still have to have the database to maintain positions.
2006 Chevy Colorado: VIA M10000 EDEN, 1 GB RAM, 80 GB 2.5" Seagate HDD, USB Slim Slot DVD/RW, Holux GPS, MobileVU 10.4" LCD (touch not working yet), VOOMPC Case (blue), 70W DC-DC supply.
The problem with saving while recording is that I cannot be accurate when one song ends and another begins. However, if I could be recording while saving (double-buffered, in effect), then that would work. I was not able to get that work, but perhaps I shoud try again. It certainly would be easier.
---Mike
Can't you trigger the save when the display data i.e. song name changes? That should be fairly accurate. I'm not experienced with XM at all, but I know Sirius changes the song an instant before it starts playing. You could then offset the saving appropriately (record for x seconds after the change or trim x seconds before teh change). I think you said that you are already logging the time the song changes accurate to within a couple of seconds. I would have to guess that you become less accurate as the record length increases? If you're off by .1 seconds on each song, you would rapidly be off time. This would too would be solved by breaking the files up.
2006 Chevy Colorado: VIA M10000 EDEN, 1 GB RAM, 80 GB 2.5" Seagate HDD, USB Slim Slot DVD/RW, Holux GPS, MobileVU 10.4" LCD (touch not working yet), VOOMPC Case (blue), 70W DC-DC supply.
With XM, the name changes +/- a couple seconds. I'm not sure what causes it to vary. Also, they usually fade one song in while the other is playing.
I did a little experimentation, and found that two instances of Cool Edit (for one) can be recording at the same time, so maybe I can too. Then I can tack on 5 seconds to beginning and end of each song. (I like only one in 20 songs on most channels, so I don't need it totally automated.)
---Mike
Bookmarks