Auto file-creation while recording from tape

[SumGuy]

Does Audacity (any version) have the ability to save each song on a tape or record to it's own file while recording from a tape or record by detecting the silent gaps between songs?

[steve]

Not while it is recording.
In Audacity 1.3.12 you can automatically detect silences after recording the entire album/tape, then split the recording into individual files based on the detected silences.
http://wiki.audacityteam.org/wiki/Split ... ate_tracks

[kozikowski]

If you have noisy tapes or records, it can kill Silence Finder. The show is never silent -- even in the lead grooves. Don't fall in love with automatically breaking up your whole music collection automatically. That almost never works.

Koz

[steve]

I've not found tape noise to be a problem for detecting "silences", it's just necessary to set the "silence threshold" a bit higher.
Noise between tracks on vinyl can be a problem for silence detection as clicks, pops and crackles can have relatively high amplitude, so it is best to de-crackle and de-pop the the recording before attempting to run silence detection.

[SumGuy]

I've found a few so-called "share-ware" programs that come as free and paid or "full" versions that have auto-gap detection and generate individual mp3 files on the fly during recording. One in particular is "Easy Hi-Q recorder". It's crippled or free version only records for max 3 minutes per song. Here's how it's gap-detection works:

You set a "Start" sound level with a slider control, marked in db. This is the sound level that needs to be present to start a recording. If the level is set too low it might record constantly. If the start level is set too high, the recording might get off to a late start or might not record at all.

You set a Silence level with another slider (also in db) which is used to stop the recording. If this level is set too low, silent gaps might not be detected which would keep the recording going. If this level is set too high, normal softer sound passages might be considered as silent gaps and would trigger the end of recording.

A third slider sets the amount of time (duration) to wait once the Silence level is detected to consider it a real silent gap between songs and stop recording. You can set it from 0.1, (1/10th of a second) to 3 seconds. In other words, the Silence level must last for at least the duration set by this slider to be considered a real gap or break between songs or audio sequences. If the silence level is shorter than the Duration setting, then recording continues.

I understand that Audacity doesn't do this in real time. But in a post-process situation, how does Audacity's gap detection and file-splitting work?

Does it also have the same method or controls to detect silent gaps between audio sequences in an automated way like this? And once detected, are new files created automatically based on the sensing of these silent gaps?

[waxcylinder]

See: http://manual.audacityteam.org/man/Sile ... Parameters

WC

[steve]

There are two similar effects in Audacity 1.3.12. Both are in the "Analyze" menu.

"Silence Finder" detects "silence" of a minimum length set by the user. There is also a "Threshold" setting. Audio below the threshold level is treated as silence.
On running the effect, labels are created before the end of the silence - the time between the label and start of the next track is set by the third slider.

"Sound Finder" is very similar but marks the detected sounds where the level is above the threshold with a "region" label. The settings are similar to Silence Finder and should be fairly self explanatory.

The marked regions may then be exported as individual files using "File menu > Export Multiple" (as described in the link that I posted previously.

Audacity is not trial-ware. It is full featured, free, open source software. Rather than going through all the theoretical detail it would probably be quicker to just download Audacity 1.3.12 and give it a try. http://audacityteam.org/download/

[kozikowski]

What kills these tools is real world recordings. If each "silent' groove has a pop or click at the wrong place, the tools will not see the space as "silence" and go around it. In cassette recordings, a plain vanilla cassette, no CrO2 or Dolby, etc. can have background hiss the same level as some of the quieter performances -- particularly one-of-a-kind home recordings. I bought commercial cassettes and they were uniformly awful.

This brings you to capturing, processing, and filtering the recording carefully -- and with custom settings for each record -- before you try splitting them up into songs. So it can be done, but that's the end of real time splitting and exporting.

The system is also actively biased against desire. Your favorite out-of-print record is the most likely to have such high noise and damage that the automatic processes fail straight away. It could be argued that this is the one record you want to handle by hand anyway, but you may have no choice.

Koz

[SumGuy]

Just for the record, here's a piece of software I found that does exactly what I was looking for:

AudioGrabber:
http://www.audiograbber.org/

You'll have to download a separate mp3 codec, which can be found here:
http://www.heinzle.co.at/lame/ag_mp3_plugin_setup.exe

[steve]

SumGuy, please don't link directly to executable files - anyone clicking on the link will have no idea of what they are downloading or whether it is safe to do so.
Anyone trying out AudioGrabber should be aware that by default it will install a "shopping" toolbar in your browser. The installer gives the option of whether to install the toolbar or not.

Although Audacity does not split the recording while it is recording, I think that splitting in Audacity is more flexible and easier to control. I guess it depends on what you want.