I recently flew about 1,200 miles to shoot a high profile series of interviews for the digital promotion of a feature film. Used a couple of 5D MKIIIs, Tascam DR40 for audio. Begged studio for a sound mixer but they cheaped out and said no. Shot interviews with several subjects, checked playback, was monitoring with headphones as I shot. All of the audio for the interviews turned out great, except for the one main subject. His files are No 15 below. I am completely baffled how the recorder could have started a header for the files, but then just "not saved" the files? They show a different icon then the rest of the intact files that show the standard Apple iTunes icon. What makes it weirder is that all of the interviews before and after this subject recorded great, sound great. I have been shooting with the DR40 and sometimes the Zoom H4N for a couple of years, I have never seen this happen.

http://i1078.photobucket.com/albums/w489/capybara84/MissingAudio_zpse9a16af5.jpg

High quality Sandisk SDHC media was formatted in the recorder. The executive producer is understandably not happy about it, it makes me look bad and all we have now is the camera audio from the 5D MKIII, which is low quality to begin with. The camera was about six to seven feet from subject, room was fairly quiet with the exception of some HVAC that couldn't be heard on the Tascam mics (Countryman B6 lav and a an Audio Technica AT875R) but the camera mic, of course, picked it up nicely. The noise can be removed, but the reverb and room characteristics can't. Studio will either have to pony up bucks for me to go back or live with the crappy camera audio. Have any of you ever seen this happen where the recorder writes a header, captures the audio (I was monitoring with headphones) and I pushed record, adjusted levels a bit as we shot and hit stop, no different than any of the other clips. I am baffled, it makes me want to throw the DR40 into the trash, how can I trust it again for other shoots? Do you think it could have been a bad sector on the SD card or the recorder or?

I am going to take this up with Tascam. The damage is done. I know, cheap recprder, you get what you pay for, buy a Sound Devices, always use two recorders, etc. but any ideas on how or why this could have happened? It strikes me as incredibly unlikely since all of the files on the card are intact , recorded and played perfectly yet one set of files just "didn't record" yet left perfect header information? Murphys Law that the only file it did this to was the MAIN SUBJECT! I told the producers, "Get me a Sound Mixer for this shoot" and they wanted to but the studio was too cheap and now we are paying the price. Although to be fair, this could have happened with a sound mixer as well.

Advice or comments?

I had a situation like this with a DR-100 (1st gen) that didn't save a clip when the battery died. That day, I had recorded over 70 short interviews (basically, saying one's name on camera). I missed a couple of clips due to user error.

In my case, the faults were all explainable. And fortunately, it was a casual piece. I think I cut one (very introverted, uncomfortable) person, and included the other two with less-good audio. The hiss/echo lasted for less than a couple seconds each, so it was gone as soon as it was noticed. It was kind of a cute result, with cuts from face to face with the spoken name. I put it over a lively beat, and cut to the rhythm.

Anyway, bummer about losing a longer interview. Hopefully, with noise reduction and equalization, you can rescue the piece.

Thanks Jon. I told the producers that IMHO, the camera audio is usable, it needs some noise reduction, EQ and compression but problem is, cutting directly to and from the other interviews, the difference in sound quality is going to be jarring, all of them close-miked except this one with the 5D MKIIIs crappy audio quality. If they can isolate this person and nort cross cut him with the other interviews, might get away with it. It does not sound good but it is all they have unless the studio hears it and decides to fly me back out to reshoot. We'll see, will keep you all posted. Hit up Tascam Service, we will see what they say. It is so frustrating when I do my job correctly, am fastidious about double miking and making sure that the recorder is recording and it still fails.

Sorry for your trouble Dan.
Puzzling for sure.
Were you running on battery or USB/external power? A sudden power loss can result in a file not being written, either by battery removal or external power loss. Dying batteries will cause the recorder shut down but not before saving it, up to the battery warning display point anyway.
Maybe the header was not written, but that usually results in no file name and display either.
Have you checked the card with file recovery software? Look for files with the same date/time, and would likely have a cryptic name.
Another recovery attempt: when I run into file issues, I change the extension to <.raw> and open in Sound Forge Pro. This usually works with a header problem.

I once lost a zoom h4 continuous recording of a full hour when I accidentally bumped the recorder (which was mounted on a mini-tripod) and it fell over, as the tripod was not even 10cm high the recorder didn't receive a hard hit when it fell but I did cause a recorded file with no bytes, the file was just empty and left me with nothing, nada as I had no other source of recording, I also monitored the sound during the record.

For very important audio recordings I always have a backup system now with 2 different recorders close to the source, it's not always fun to learn the hard way.

I had a similar thing happen on an H4n, but it was because the battery died about 30 minutes into the recording. We ended up having to go back and reshoot the interview, as the camera audio just couldn't be made close enough to the other interviews to cut successfully.

Granted, the whole audio path in a DSLR is rather less than mediocre. But AT LEAST feed the proper lav/shotgun audio INTO the DSLR and never use the 87-cent electret capsule in the camera body. That would have at least given you a running start at salvaging good-enough audio out of the DSLR audio track.

I would be surprised if you even get a response from Tascam. As a big corporation, they have undoubtedly insulated themselves very well from the slightest hint of liability.

As far as power, I was changing out the batteries pretty regularly, I am confident they had at least 60-70% power, if not more.

Recorder was not jarred or dropped during recording.

I am beginning to think it was simply a card error that corrupted the file. We still have this card, asked producers to consider having a data recovery place look at it. It is possible the data is still there but just not reading.

I guess I wasn't being redundant enough, I always bring a spare camera and spare recorder but I just have those in case my main units go down. That did happen with my first DR40 about a year ago, but I had my spare H4N with me, which saved the day. I think I am going to switch cards and begin recording interviews with two recorders. Losing audio like this is a nightmare.

Thanks for all of the ideas and words of wisdom.

Why don't you give this a shot. it helped me recently when I lost the entire card. Luckily it recovered file.

Recuva - Download (http://www.piriform.com/recuva/download)

"I recently flew about 1,200 miles to shoot a high profile series of interviews for the digital promotion of a feature film. Used a couple of 5D MKIIIs, Tascam DR40 for audio. Begged studio for a sound mixer but they cheaped out and said no."

I find your first sentence the most interesting and I hope you remind the production that they risked their whole shoot for the sake of a soundie.

Using low cost semi pro kit on faraway shoots is never a good idea and it highlights the failings of the DSLR shooting workflow.

Did you re-format the memory cards in the tascam before each use as just deleting clips can cause directory problems on any system?

I think it's interesting that the previous pair of files, #14 in the sequence, have the same time stamp (4:05 PM) as the bad files. This makes me question the button-push sequence. Did you really start recording #15 that soon after you finished recording #14? (I notice that all the other files are several minutes apart... the 14-15 pair is an anomaly.) How does this compare with the time stamp on the camera files?

Is it possible that the recorder had not finished the process of finalizing and saving #14, when you pushed the button to start recording #15, and so the recorder wasn't able to start recording #15 properly?

You might want to examine the file properties individually (i.e. right-click in Windows) so you can also see the seconds digits of the time stamps. Perhaps that would give you an additional clue. You might also find that there is a file, just a few Bytes long... small enough that it shows up as Zero KiloBytes in the directory.

This certainly highlights the difference between spare ("I had a spare recorder in my bag") and redundant ("I had a second recorder simultaneously making a redundant backup recording").

I hope you (or someone) can find and recover the data. Good luck with that!

The first time I used my original Zoom H4 with a new, larger, faster SD card I ran into something similar. Fortunately the Zoom was only the backup as a second recorder. The boom mic audio was going first into a Mix-Pre, which split it to the camera and the Zoom, so nothing important ended up being lost.

Hopefully either your camera audio will suffice with careful sweetening, or the files (if anything is really there) can be recovered.

Is it possible that the recorder had not finished the process of finalizing and saving #14, when you pushed the button to start recording #15, and so the recorder wasn't able to start recording #15 properly?
> Interesting observation Greg.. and entirely possible.
---------------------
I used Recuva as well to salvage files from a clients H4n, in which the batteries went dead whist recording.

I used Recuva as well to salvage files from a clients H4n, in which the batteries went dead whist recording

That is hugely valuable to know, thanks. I had no idea that file recovery software could be used where the H4n had run out of battery power before completing a file. That had always made me nervous about using it. It must indeed be worth a shot by the OP.

Pete

It sometimes requires a 'deep scan'. The files will have cryptic names as well, so it takes some detective work. If the user re-starts recording after the power loss w/o changing the card, the data 'can' be overwritten.. i.e. not recoverable.

Addendum:
In any case the files will likely not have header info either, so as I stated previously, enter the <.raw> extension and attempt opening in Sound Forge Pro or another audio app that opens RAW files. (many do not)

I've used Recuva, too, although sometimes other recovery programs have found things that Recuva had missed.

My fear about this particular situation is that, if the file system did not set aside space for the "missing" files, and the OP then went on to record more files in sequence, the recorder might have over-written any space which had been used to write the "missing" files.

No point speculating, though, the key is to scan the card ASAP with Recuva and possibly other software, before anything else is written.

I hope we'll hear some news from the OP before too long...

Tried Recuva as well as two other programs. No luck. We ended up using the camera audio, enhancing and noise reducing it. It still sounds lousy but usable.

As has been said before feeding the camera sound from the good mics for an interview is a good idea even if the quality is not as good as the external recorder. Trying to avoid single points of failure is part of the business.