Topic: Edirol R-09 bad file

[hobbes4444]

So, in doing some research here, it looks like I might be able to save/recover the file I recorded last night.  I have a 1.2 Gb file according to windows (i transferred the file to my PC HD), but it is not recognized as a .wav file.  The Edirol shows this file as "Improper song" but thankfully it shows that the 4Gb card has 675mb remaining (there are 3 files in total on the card, each at 1+ Gb.).  so i know i have a file, just likely one with a corrupt header.  So is the best way to try to recover this file to get audiohack?  The only other program I have is audacity.  Could i import the file with that?  I'm very unsavvy when it comes to using command line prompts, etc., so any handholding to work this through would be appreciated.  And if anyone is interested, the corrupt file is Suzanne Vega's late set at the RegattaBar last night.  Happy to share it with folks who can help before I hopefully get it on dime (DPA4061>R-09).  IT should have come out better than the first set since the bar was mostly empty for the late show, and she was definitely looser. . .

Thanks as always,
Dave

[udovdh]

Rename to .raw and try to import the file. Was asked before in the past few days and worked fine as a solution.

[hobbes4444]

Rename to .raw and try to import the file. Was asked before in the past few days and worked fine as a solution.

I thought that solution was in relation to Soundforge. . .  though after a little more research it seems, as you suggest, that any program that can accept .raw files could work with this out.  will try this solution with audacity.

Thanks. . .
Dave   

[hobbes4444]

Rename to .raw and try to import the file. Was asked before in the past few days and worked fine as a solution.

importing raw file didn't work. 

tried audiohack.  not sure what exactly happened.   i have a file that is 44 bytes, and i'm not sure what to do now.  ..

[Brian Skalinder]

If the file has valid audio data, but no or a hosed header, CD-Wave should open it.  IMO, best to try on the original file, not the one that's been processed by audiohack.  (Dunno if audiohack creates a copy on which to work, or if it works on the original file.)

[live2496]

Rename to .raw and try to import the file. Was asked before in the past few days and worked fine as a solution.

importing raw file didn't work. 

tried audiohack.  not sure what exactly happened.   i have a file that is 44 bytes, and i'm not sure what to do now.  ..

Audiohack creates two files. Do both have only 44 bytes? If so, then Audiohack could not determine the start of the data chunk.
To fix this manually, you would have to edit the header and then run it through Audiohack. This solution is only viable if you are comfortable with editing using a hex editor. It is not that difficult, really.

Or you could import the audio as raw into an audio editor.

Gordon

[hobbes4444]

Rename to .raw and try to import the file. Was asked before in the past few days and worked fine as a solution.

importing raw file didn't work. 

tried audiohack.  not sure what exactly happened.   i have a file that is 44 bytes, and i'm not sure what to do now.  ..

Audiohack creates two files. Do both have only 44 bytes? If so, then Audiohack could not determine the start of the data chunk.
To fix this manually, you would have to edit the header and then run it through Audiohack. This solution is only viable if you are comfortable with editing using a hex editor. It is not that difficult, really.

Or you could import the audio as raw into an audio editor.

Gordon

yes, both files are 44 bytes, and the cmd line indicated input file has no data chunk.  so i'm trying to create a PCM file with audiohack.  same result.

audacity has the capability to import raw files, but i couldn't get it done properly.  the original file is 24/44.1 on the R-09.  there seem to be a lot of options in importing raw and i'm not sure which to use.  tried 24 bit signed PCM and little endian and that yielded noise.

no clue what i'd be doing re hex editor. . .

i think i have cdwave, i'll try that next.

thanks so far, for the help . ..

[hobbes4444]

cd wave said cannot open RIFF chuck, file is not a valid wave file.  so does that mean i have a f'd header that will have to be manually edited?  i think a friend of mine has cool edit, so maybe i can get the file to him to try.  and i can still try chkdsk i guess, tho i tried that on the file that i copied to the HD, and nothing appeared wrong.

[Teen Age Riot]

FWIW, I had the exact same problem and I haven't been able to solve it. So far.
http://taperssection.com/index.php?topic=68518.msg944528#msg944528

[udovdh]

audacity has the capability to import raw files, but i couldn't get it done properly.  the original file is 24/44.1 on the R-09.  there seem to be a lot of options in importing raw and i'm not sure which to use.  tried 24 bit signed PCM and little endian and that yielded noise.

If you have a hosed WAV header the 44 bytes are not dividable by 6 (2 times 3 bytes per sample) causing noise.
So you need to add 4 bytes to the header or lose 2 bytes.
Or copy over a 44-byte wav header using a hex-editor. You could then even fill in the length etc for the data chunk.

no clue what i'd be doing re hex editor. . .

Get one.
Have a look at the start of a valid WAV file. Have some info on the WAV header at hand. E.g. http://www.sonicspot.com/guide/wavefiles.html
Now open the problem file. See if it has traces of a WAV header. Fill in the blanks or missing info.

[live2496]

cd wave said cannot open RIFF chuck, file is not a valid wave file.  so does that mean i have a f'd header that will have to be manually edited?  i think a friend of mine has cool edit, so maybe i can get the file to him to try.  and i can still try chkdsk i guess, tho i tried that on the file that i copied to the HD, and nothing appeared wrong.

I will continue on the RIFF header edit scenario.  What you have to do is make the first 44 bytes of this file match that of a good file recorded using the same settings.

Download Ultra-Edit from www.ultraedit.com .  Open up your file with it and it should switch to the hex view because it will detect it as being a binary file.  Now open a good file. Set the Window so that you can compare the files in the same window. (From the menu this is Window->Tile Horizontal or Window->Tile Vertical )

Make sure that the first 44 bytes of both files match exactly. Ultra-Edit will allow you to change values by positioning the cursor on the bytes you wish to change. Save the RIFF corrected file.

Now redo the audiohack procedure and you should get two files. The first file will have the RIFF header counter and data chunk counter corrected based upon the file size.

If you get a file that can be opened in an audio editor but it contains white noise, then the actual offset to the data is incorrect. With many software pacakges the riff header should take up 44 bytes, but this is not always the case if the fmt chunk is larger or there is a pad chunk inserted into your file by some process.

In that case, you would need to import the file as raw and specify the start of the pcm data. Ultraedit can help you determine the start of the data chunk. I use Samplitude 7 for raw data import because it has a way to specify a starting offset to the samples. With 24-bit data one of six offsets would be correct. I would try separate attempts with 45, 46, 47, 48, 49, or 50.

Audiohack also has an option to strip the RIFF header from the file, but it has to be able to determine where the data chunk starts. For your file this can't be determined. However, I guess I need an option so that the user can specify a certain number of bytes to strip from the beginning of the file. That way it could be subsequently imported to any software package that has the raw import capability.

Gordon

[live2496]

If you have a hosed WAV header the 44 bytes are not dividable by 6 (2 times 3 bytes per sample) causing noise.
So you need to add 4 bytes to the header or lose 2 bytes.
Or copy over a 44-byte wav header using a hex-editor. You could then even fill in the length etc for the data chunk.

Hi Udo,
The data chunk only needs to be a multiple of 6 bytes (for 24-bit stereo).

For many softwares, 44 bytes of RIFF+fmt chunk is normal.

Gordon

[Zaphod]

Well I am having the same issue. Apparently the R-09 has a problem saving more than one 1+ GB file on the same 4GB card. But I don't think its a predictable occurance, or is it?

I recorded one set @ 24/48 and the file is 1.05 GB it saved fine

I recorded a second set again @ 24/48 the file is 1.4 GB and it the R-09 flashes "improper song" when I try to play it. I copied the file to my mac and it won't open in Spark XL or with the Quicktime player.

The questions is, does anyone know of a good free wav header editor for the Mac?

[udovdh]

Well I am having the same issue. Apparently the R-09 has a problem saving more than one 1+ GB file on the same 4GB card. But I don't think its a predictable occurance, or is it?

Perhaps other people exprienced the same?
Maybe some more testing can do the job?
(just set for 24/48 and let it roll...)

[Zaphod]

Well I am having the same issue. Apparently the R-09 has a problem saving more than one 1+ GB file on the same 4GB card. But I don't think its a predictable occurance, or is it?

Perhaps other people exprienced the same?
Maybe some more testing can do the job?
(just set for 24/48 and let it roll...)

I've only had one other file the R-09 couldn't recognize like this other one. It was a 2.8 GB file (3 hour show @ 24/48) it too said "improper song", but when I transferred it to the mac it opened up fine in Spark XL. I think the only reason the R-09 wouldn't recognize it is because it was over 2GB.

So when it said "improper song" I thought I'd just transfer the file and open it up in Spark XL like the last one, but nope it won't open.

So I suspect the problem here is that the R-09 sometimes won't write the .wav file header correctly if there is already one file over 1GB on the card.