Karl Chelette
Forum Replies Created
-
Rod, sorry, been busy working on taxes. Get me the sd card image and image of the test recording and I’ll see what I can do. The first eight minutes aren’t necessarily gone, but might be. You can email me at karlc37@dslextreme.com
-
Are you able to load the file as raw in audacity and hear bits of the music? Besides the disk image will need a sample recording in the same format. Same number of channels, bit rate, sample rate, etc and it would be helpful if you would write zeros to the SD card first, so I don’t have to search so hard. I would like you to record on the first channel, silence, the second channel, loud noise. Repeat for the remaining channels. Need 1-2 minutes only.
Karl
-
Adam, ok, dr-40. What I would do to find the data block size is make a test recording with channel 1-pink or white noise, channel 2- silence. Load into audacity as raw, find the section where the silence begins, expand out the timeline and trim off the excess noise leaving very little before the silence starts. Then stretch out the timeline to the very max and you can find the last byte of noise before the silence byte, a diagonal line coming down to the center line. Trim off the noise. Do the same thing at the end of the silence. Save as wav file, load into hxd, trim off the 44 byte header and do a byte count. It should line up symmetrically, the first byte starting at byte 0 and the last one ending at byte F. Use “goto” to change the hex number to decimal for chunk size. Load the lost file section in audacity and hopefully you’ll be able to figure out the start of data byte, which is where the recorder started recording at, trim off the excess. Save as wav file, load into hxd, save the 44 byte header into a separate file or write it down, delete the header, split file with hxd into chunks of the right size, combine every other chunk and you’ll end up with 2 mono files. Copy the wave header you saved earlier and paste onto the beginning of each file. Done. Headache yet?
-
Hi Adam, yes, I can do this for you. The first time I spent more than 40 hours to figure it out and this is after tech support at Zoom said it’s not possible! The hardest part is to find where the data starts because there is no header written. Any headers you find are old file headers. Is this an H4n? If it is, I’ll need a disk image of your test recording. If it’s a R16/R24 won’t need it. And of course a disk image of the sd card. Are you able to load it into Audacity as raw and hear the interleaved recording?
-
Hi Adam, yes, I can do this for you. The first time I spent more than 40 hours to figure it out and this is after tech support at Zoom said it’s not possible! The hardest part is to find where the data starts because there is no header written. Any headers you find are old file headers. Is this an H4n? If it is, I’ll need a disk image of your test recording. If it’s a R16/R24 won’t need it. And of course a disk image of the sd card. Are you able to load it into Audacity as raw and hear the interleaved recording?
-
Hi Ben, Sure, I can help. I would say recovery is almost certainly possible unless you recorded something else on the sd card. You will need to get me a disk image of the sd card. My dropbox is only about 5 gigs so anything bigger you’ll have to mail to me, unless you have a bigger dropbox or cloud storage. Get it to me and I’ll start working on it right away. When I’m done I’ll be able to upload them to my dropbox for you. How long are the tracks you are trying to recover?
-
Hi Jackson, what I would do is make an image of the sd card, then import all of it into Audacity as raw, experiment with the start offset, 0-4 until you get something that plays. Then you could search through it to find out if you recorded over it or not. You could then maybe use Audacity to trim off the beginning and resave it. That way when you go looking for the start of data in the hex editor you won’t have to sift through so much data. In order to recover, you have to find the exact start off data and know the data chunk size, so you can manually un-interleave it. In my case there was a whole lot of 0’s before the data started. Let me know what you find. I’m just now starting to work on a set of lost files from a Sony DR680 and I don’t think it is going to be as easy as the Zoom.
Karl -
Chris, I don’t think it’s possible to automate this process because the start and end of the data chunk must be determined and this info is lost when the unit loses power. On any given chip there might be half a dozen data chunks floating around from long ago deleted recordings. If you can make a disk image of your sd card and make a disk image of a sample recording in the same format, since I don’t know the data chunk size of the H4n, put them on a usb drive and send it to me, I’ll recover it for you, put it on your usb drive and mail it back to you and call it a merry christmas to you! P.O. Box 2387 Nevada City Ca, 95959
-
Hi all, I’ve just cracked the 4 and 8 track corrupt audio file problem, which is a 3 second echo of each track after importing raw into Audacity. This happened from sound guy turning powerstrip off with no batteries in unit. Here are the steps I used: First I made image file of the sd card with the lost files, then I took an old sd card and wrote all zeros to it, formatted it. Then made a short test recording with the same format, saved it and turned it off properly.
Then I made an image file of the test recording and opened it up in HXD hex editor. I found out that with my zoom r24 recorder, when recording, it records 393216 bytes of each channel sequentially and when done writes the wav header with data location, recording format etc. When turned off improperly the wav header is not written and the data that is there is in alternating blocks of each channel.
Using HXD hex editor I found the start of the data and saved this as a new file, which I then used the split option in HXD and split the file into chunks of 393216 bytes, named 0001 and up. Then I opened the folder with the file chunks into it and squised the screen down till I could view 2 rows of files, about 4,000 of them. Then I selected the odd row on the left and cut and pasted it into a new folder. Then using HXD, I concatenated (joined) the odd file numbers together, then did the same for the even ones. This left me with in my case, the data for both mono channels. Then I imported the test recording file in HXD and pasted the recovered data into the proper space, which was at 60000 hex location. The next thing I had to alter was the last 4 bytes before the data which tell the file length data. Repeated the same for the other channel then voila, both channels recovered, no echo, synced properly, 3 hours worth back!!!!!