I’m having a problem with multiple export and I dont know if its something I’m doing (probable) or something this progamme cant do. I’m installing voice onto my R/C transmitter so it speak to me instead on just beeps and cheeps. Using ver 2.0.1 and Vista, I convert text to speech and save as an mp3 file then open it in Audacity, normalise, equalise and adjust gain. Then using the sound finder locate the individual block of text/words. This works OK with a bit of tweaking to remove some unwanted bits so I end up exporting just what I want which is lots of files.
The problem is exporting these files as 16Khz 8bit wav, I can do that and they will all play in Audacity and win media player but when I install them onto my SD card and put them into the Tx I only get a couple of clicks. I have managed to export these files by one by one using the export selection function and manually re-naming the files but this is a boring and time consuming job. I have been told by the chap who developed the audio mod for the TX that multiple export is adding some data into the file which is preventing it from playing in the Tx.
(these are the replies from designer (copied from er9x forum)
I wonder if it is the export multiple causing the problem. I just looked at your files and the header is different from the files I created when I did an export. Your header is longer and includes a LIST specifier. Try loading one of you files in, and exporting it by itself (other uncompressed files, WAV (Microsoft) unsigned 8-bit).
OK, I just tried that on one of your files. When you export it, as well as asking for the filename, it then pops up a dialog box asking for information. Your files include a track number at this point. I deleted the track number, so the dialog box had no information in it, then clicked OK. The header of the resulting file was the same size and content as my files.)
So after all this waffle, my question is can Audacity do multiple export without this extra data, and if so how?
Important: The RIFF specification (where RIFF WAVE defines the WAV file format) explicitely allows unknown and even illegal data chunks in a WAV file and explicitely requires the playback software to ignore and skip all unknown data. This means that if the playback software can’t play a WAV file only because it contains a LIST chunk (a legal WAV chunk, containing the data from the Audacity metadata window) then this is definitely a bug in the playback software, that can’t play WAV files correctly.
I have tried it myself and there seems to be no other way than manually delete all data from the Audacity metadata window to export WAV files without a LIST chunk. Unsetting “Edit > Preferences > Show metadate editor” does not stop Audacity to export WAV files containing LIST chunks. This means that if you have an Audacity track split into into 100 parts via 100 labels, then exporting the track via “File > Export Multiple > Split files based on labels”, you have to manually delete 100 times the data from the metadata window.
Maybe somebody else knows an easier way, but I have found no better.
Please note that this is a bug in the playback software, not a bug in Audacity.
edgar
P.S.: The original discussion can be found in the 9x forum under 9xforums.com
RFT = Radio / Television Broadcast Technician (German abbreviation, that’s why the letters don’t match), I’m an audio / video hardware repair something.
I’m not an Rx specialist, but I’ve already read MikeB’s VoiceHowTo.pdf that tells me that the software that plays the audio data is contained in this MP3 SD Card Sound Module:
Is this really true or is this just an obsolete predecessor version? An MP3-Module can’t play WAV files. I would need more information from anywhere what WAV files can be played and what not or at least a description of the playback hard- or software.
The particular problem here is that WAV is a “container” format, what means that WAV only describes HOW the data is stored, but not WHAT data is stored. People often think that a WAV file must contain audio data, but this is not true. According to the RIFF WAVE specification, a WAV file can contain virtually any arbitrary data, from audio, video, text, up to really everything.
Wave File Chunks - many commonly used non-audio data chunks contained in WAV files
Important: It’s task of the WAV audio software to detect, ignore and skip data chunks it cannot handle.
WAV audio software is required to detect, ignore, and skip all kind of data in a WAV file it cannot handle during playback or processing. This is an essential part of all WAV audio software, explicitely required by the RIFF specification.
WAV audio software that is incapable to detect, ignore, and skip non-audio data during playback or processing is no WAV audio software according to the RIFF specification.
With WAV audio software that cannot detect, ignore, and skip non-audio data you will run into endless trouble with ALL audio software, not only with Audacity.
With a small SD card player there is only a very small amount of memory available, so everything looks as if the WAV audio software has no routines to identify and skip non-audio data implemented.
All you need to do is to write a little command-line loop (depending on the operating system) to process all exported files with SoX, and if you’re lucky, the “copy-xxx.wav” versions can be used for the SD card player.
Thanks, Edgar. That looks more like a SoX bug than a feature?
While you were brewing up that solution I was brewing up another solution, using a patch that a developer gave me a while ago (see below).
I am a bit confused by that. Deleting the Track Title so that its “value” field is empty should indeed mean that no tag is written (not just an empty tag). But Audacity does not support writing Track Number in WAV files (see Missing features - Audacity Support ), so there is no need to delete Track Number from Metadata Editor .
In the case where you had to delete multiple tags, you could press “Clear” in Metadata Editor (but you still have to do it for each window that appears).
We can’t allow hiding Metadata Editor to prevent tag export, because entering the metadata common to all tracks before export, then unchecking “Show Metadata Editor” is the only way you can Export Multiple with common metadata but without entering that metadata for each and every song. This problem is under discussion with the developers, so we shall see.
@billl516 (at your own risk) you can try gaclrecords.org.uk . What you want to do is extract “audacity_no_metadata.exe” from the zip, drop it into the folder you launch Audacity from, quit Audacity, then launch “audacity_no_metadata.exe” instead of “audacity.exe”.
When you Export Multiple, choose “Do not write metadata” in “Prompt for metadata” - there will be no Metadata Editor windows and no metadata will be exported.
You can also choose “Before first file only”, and click OK on one Metadata Editor window, which will be empty. This will also export no metadata, but is of course not what the developer was intending, which was to pass the Track Title and Track Number through to the first window, thence to the files.
Gale thanks for that link. I tried it and it seems to work, I just did a quick split in Audacity and and multiple export with no metadata and saved as *.wav. I had to re-number the files so the Tx played the correct file and I got the welcome message at power on and minute message after a minute had elapsed. So I cant see any reason why it should not have solved the problem. Its going to save a whole lot of typing and clicking.