What on earth bit me?????

[Greg Miller]

Mr. Andrada,

I assume that you tried playback with the card in more than just one device. If you're really lucky, some buffer somewhere might be at a marginal state, and the audio might play OK on some particular card reader.

Mr. Crowley,

Since you pointed out that it's [almost] exactly ten blocks, I've abandoned my idea that it was some sort of RF hit that made the Sound Devices hardware misbehave. It's almost impossible that the "hit" would have been the exact duration -- down to a microsecond -- to cause that. Too improbable to be a coincidence.

It's more likely that some of the bits in those ten blocks were permanently stuck in the "1" state, rather than saving the data that should have been written to them. Apparently some of the least significant bits were saved successfully, so we can see some of the waveform, but not the "big picture."

Re-formatting the card might help, but that would still be a big gamble IMHO. (Of course a brand new card might have a few bad areas, too.)

The nice thing about acetate tape was that you could always put it back together if it broke, with no discernible data loss!

[Jim Andrada]

Hi Greg

I didn't play anything from the card - just copied the files to the hard drive. And it was a new card I had never used before. I sort of treat Flash like tape - I use it once and archive it.

In the course of around a two hour concert there were a total of five hits. One was early in the Tchaikovsky Piano Concerto #1. The pianist came from Japan just for this concert so it was something of a big deal.


I'll see if the folks at Sound Devices have any ideas - I sent them a problem report last night and they're pretty good about quick response.

[Greg Miller]

Hi Jim,

What I was trying to ask: did you try reading the card in a few different card readers (when you copied the data to your HD)?

It occurs to me that if the card is bad in such a way that the buffers are sensitive to voltage or current levels, you might find one reader that will read the data correctly, when another reader will not.

Of course if the actual data on the card is bad, you can't fix that. But if the actual data is good, and the problem is with the process of reading the card, there's a slim chance that you might get lucky with a different card reader.

For that matter, maybe you should try playing the audio directly from the card. If there is a speed problem with the card, it might play OK since the data rate for playback is slower than the data rate when you copy the data directly to your HD.

This is all speculation and I certainly don't guarantee anything, but it surely wouldn't hurt to try.

[Alexey Lukin]

Jim, this is probably caused by the faulty recorder or the memory card. Try reading the same stuff from the card again. The problem is a digital skip in the file. It can be restored like this with a skip correction and some RX: NoiseRestored.wav.

If this file is valuable to you, I can fix a few more of these — just send them in a PM.



(click to see the animated result)

[Greg Miller]

Alexey,

Molodets!!! What program did you use to do that?

[Alexey Lukin]

Just a text editor to delete a byte of binary data. Then RX to fix the clicks at the boundaries.

[Paul R Johnson]

Now that is the power of the internet! Very impressive!

[Greg Miller]

Excellent! I guess I need to get a text editor that can handle large binary files. What do you recommend?

[Alexey Lukin]

I have used EditPlus for this one, but I don't really know if it'll work with really large files. I'm sure there's a lot of freebies to do that.

[Jim Andrada]

Hi Alexey

Thanks very much

Yes, it is quite important. From what I can see there are 3 or 4 in the Tchaikovsky Symphony # 5 and 1 in the Piano Concerto #1. The pianist flew from Japan for this concert. So it was rather a big deal.

I can send you the problem spots from the symphony, but the pianist has the original file for the concerto in Japan with her so I'll have to ask her to send the disk back to me.

Could you explain a little more about how you repaired it? I use RX, but how do you find which byte of data to erase? If I can understand that I can probably fix the problem myself instead of bothering you.

Thanks again

[Alexey Lukin]

I have just deleted a random byte closer to the beginning of the file. This has damaged most of the file, but repaired the previously damaged section. Then I have pasted the repaired section into the original file and RX'ed the transitions. I don't mind doing 3–5 more of these if you don't have luck with other spots.

[Jim Andrada]

Thanks Alexey.

I've tried it and it seems to be working successfully so far. 3 more to go

[Jim Andrada]

All done!

Thanks to everybody who helped out I'll be able to get the CD's out shortly. Video will take a bit longer considering we were running 7 cameras.. The audio files are uploading now and I'll post links when the uploads complete.

[Christian Brown]

Originally Posted by Jim Andrada

So whatever is going wrong must be going wrong internal to the 702's digital stage as the analog signal out seems to be just fine

Hi Jim -- You are right on track here. The distortion you are seeing/hearing is not from the mics or power or anything else you have connected. It is just in the recorder itself, confirmed by the clean signal that ran to the Atmos and three cameras. If you record often enough enough to digital storage, especially the same storage, you will eventually encounter corrupted sectors. This can happen on SD cards, CF cards, SSD, HDDs, etc. Pretty much anything.

To help avoid this: reformat the card and HD before *every* concert. And if it happens twice, change to a new HD or card. I have had Transcend SD cards that I had to retire after several years because of little blips.

Good luck!