Jump to content
tomsalyer

SD 633 WAV File Shorter Than Expected ... Recovery?

Recommended Posts

5 minutes ago, Constantin said:


That's not a hard drive comment. Since you are simply inferring what I "meant" from what I wrote, you might've noticed I used the plural "media". So yes, one medium was a hard drive, but the other was another cf card.
Besides, this whole hard drive issue is really beside the point and you still didn't explain what you actually meant by your original comment...

Constantin,  All the 788t's that I have seen only had a hard drive and a compact flash. That's where I made the assumption that when you said your CF card failed, but your other media didn't I just assumed it was a hard drive you were referring to. I didn't realize you had a special 788t that had more than one compact flash card hole in it.

Sincerely, Martin 

Share this post


Link to post
Share on other sites
Constantin,  All the 788t's that I have seen only had a hard drive and a compact flash. That's where I made the assumption that when you said your CF card failed, but your other media didn't I just assumed it was a hard drive you were referring to. I didn't realize you had a special 788t that had more than one compact flash card hole in it.
Sincerely, Martin 

I don't have a special one. I have one with the regular Firewire port, to which I have connected a firewire cf card reader. It's pretty simple really.
What's increasingly difficult, and possibly futile, is to get an answer out of you what you meant by your comment:

"Hello, The failure is BEFORE the CF cards, and in his case infinitely before the hard drive"

I'd be interested in hearing how you got to this conclusion, and in general would like to steer the conversation away from hard drive amd back to CF card failures and other possible causes for the OP's problem

Share this post


Link to post
Share on other sites
Just now, Constantin said:


I don't have a special one. I have one with the regular Firewire port, to which I have connected a firewire cf card reader. It's pretty simple really.
What's increasingly difficult, and possibly futile, is to get an answer out of you what you meant by your comment:

"Hello, The failure is BEFORE the CF cards, and in his case infinitely before the hard drive"

I'd be interested in hearing how you got to this conclusion, and in general would like to steer the conversation away from hard drive amd back to CF card failures and other possible causes for the OP's problem

Constantin, The comment you referenced above it's not a conclusion it's a fact. The sun came up yesterday. That's also not a conclusion, that's a fact. So somehow you think that because you connected something that's NOT part of a Sound Devices unit, that's somehow relevant. It's extremely irrelevant. I don't know if you can plug anything like that into a 633, and even if you could, how is that relevant to what happened INSIDE his 633? You are not referencing what happened INSIDE your 788t. Yes, if he had hooked up 5 other recorders to the outputs of his 633 he would have a backup recording. But that's still not relevant to what happened INSIDE his 633. Why would you even mention that a recording outside of your 788 is somehow relevant. If a frog had wings.

Sincerely, Martin 

Share this post


Link to post
Share on other sites
1 hour ago, MartinTheMixer said:

Constantin, The comment you referenced above it's not a conclusion it's a fact. The sun came up yesterday. That's also not a conclusion, that's a fact. So somehow you think that because you connected something that's NOT part of a Sound Devices unit, that's somehow relevant. It's extremely irrelevant. I don't know if you can plug anything like that into a 633, and even if you could, how is that relevant to what happened INSIDE his 633? You are not referencing what happened INSIDE your 788t. Yes, if he had hooked up 5 other recorders to the outputs of his 633 he would have a backup recording. But that's still not relevant to what happened INSIDE his 633. Why would you even mention that a recording outside of your 788 is somehow relevant. If a frog had wings.

Sincerely, Martin 

Martin, first of all: Relax. We're all colleagues here, ok? Secondly, personally I can do without your analogies. There are no children here, everyone knows what a fact is.

And what you stated is not a fact. It is what you believe to be one, but no-one can know what the reason (or where, if you prefer) for the problem is. Without official confirmation, everything else is just guessing. You cannot know if the CF card is the reason for the malfunction or not. As I wrote twice now, it is possible (and I have heard of such occurences) that a corrupt CF card could cause problems to the operating system, which seemingly have nothing to do with the CF card itself.

The CF card reader connected to my 788T is not exactly outside the recorder. You know, it doesn't matter where the drive is located physically, the data still originates from within the recorder. Once the recorder has processed the digital signal it needs to send it somehwhere for storage. In the case of the 633 it can do that on two recording media, a CF and an SD card. The recorder can either send an independent mirror to both cards or it could send the file to one card first and then on to the other. Maybe there are other options, too, but in all cases the data to be stored will come from a single data stream from the recorder. So it doesn't really matter if the recorder sends the data to an internal hard drive, to an external DVD recorder, or to a card reader built into the device, it all comes from the same data stream. If this stream has a problem, unnoticed or not, the problem will most likely be the same on all drives. Why there is a problem in the stream to begin with is another matter, but as I mentioned before, it could be because of an upset OS, caused by a faulty card.

Now, I mentioned my 788, because here the data stream appears to be of the first scenario I mentioned (a mirror, i.e. an identical file will be sent independently to all connected drives). So inspite of a failure of one recording card, the machine just went on recording on the other drives. So Sound Devices know how to do this, and it would be strange if they didn't implement the same strategy on the 633. That's basically all I was trying to say. Didn't think I would have to use so many words for it.

Share this post


Link to post
Share on other sites
2 minutes ago, Constantin said:

Martin, first of all: Relax. We're all colleagues here, ok? Secondly, personally I can do without your analogies. There are no children here, everyone knows what a fact is.

And what you stated is not a fact. It is what you believe to be one, but no-one can know what the reason (or where, if you prefer) for the problem is. Without official confirmation, everything else is just guessing. You cannot know if the CF card is the reason for the malfunction or not. As I wrote twice now, it is possible (and I have heard of such occurences) that a corrupt CF card could cause problems to the operating system, which seemingly have nothing to do with the CF card itself.

The CF card reader connected to my 788T is not exactly outside the recorder. You know, it doesn't matter where the drive is located physically, the data still originates from within the recorder. Once the recorder has processed the digital signal it needs to send it somehwhere for storage. In the case of the 633 it can do that on two recording media, a CF and an SD card. The recorder can either send an independent mirror to both cards or it could send the file to one card first and then on to the other. Maybe there are other options, too, but in all cases the data to be stored will come from a single data stream from the recorder. So it doesn't really matter if the recorder sends the data to an internal hard drive, to an external DVD recorder, or to a card reader built into the device, it all comes from the same data stream. If this stream has a problem, unnoticed or not, the problem will most likely be the same on all drives. Why there is a problem in the stream to begin with is another matter, but as I mentioned before, it could be because of an upset OS, caused by a faulty card.

Now, I mentioned my 788, because here the data stream appears to be of the first scenario I mentioned (a mirror, i.e. an identical file will be sent independently to all connected drives). So inspite of a failure of one recording card, the machine just went on recording on the other drives. So Sound Devices know how to do this, and it would be strange if they didn't implement the same strategy on the 633. That's basically all I was trying to say. Didn't think I would have to use so many words for it.

Constantin, The fact referenced was that a hard drive in his case is infinitely before his hard drive, because there's not one. That is a fact not an opinion. That is an engineering term. infinitely meaning never, ever and forever. If you look on a camera lens you will see a funky little figure that looks kind of like an 8. That's the infinity figure. If you want to focus on Mars or Way Beyond Mars that's the setting that you would put the camera lens on. So yes it is a fact that the hard drive in his case, it's located in Infinity meaning never. I don't know how else to explain this. 

Sincerely, Martin 

Share this post


Link to post
Share on other sites
Constantin, The fact referenced was that a hard drive in his case is infinitely before his hard drive, because there's not one. That is a fact not an opinion. That is an engineering term. infinitely meaning never, ever and forever. If you look on a camera lens you will see a funky little figure that looks kind of like an 8. That's the infinity figure. If you want to focus on Mars or Way Beyond Mars that's the setting that you would put the camera lens on. So yes it is a fact that the hard drive in his case, it's located in Infinity meaning never. I don't know how else to explain this. 
Sincerely, Martin 

Martin! That's what you were talking about??? Is there or isn't there a hard drive? Look, I try not to swear in public, but for f**ks sake we could have cleared that up a long time ago: Indeed there is no hard drive on the 633. But that is not relevant. You could read the rest of my previous post and offer an opinion on that, (warning: analogy coming up) but you are like a political pundit: you state your opinion and if someone offers a different opinion you sidetrack them on a benign issue of no particular pertinence to the discussion (I've noticed that in other threads where you contributed, too).
But we are not having a political argument here, we are trying to have a discussion

Share this post


Link to post
Share on other sites
Freeheel   

The important consideration here is the question of whether a failure on one media can damage the other one.

On my SD 744t, if the hard drive was full, the recording would fail on BOTH the CF card and the hard drive.

Which I considered a terrible issue and was one of the reasons I jumped to get a 633 when they first came out.

I have always backed up both of my 633 media files (SD and CF) when handing in files to production in case of media errors.  If a media error in one causes an issue in the other than the "redundant system" is not so redundant...

I'm phoning Sound Devices right now...

Cheers,

Brent Calkin

Share this post


Link to post
Share on other sites
Freeheel   

So after talking to SD tech support, a media failure on one card of a 633 should not affect the other one.  You can test this by pulling a card out of its slot while recording. (they suggest that you reformat the card after trying this)

Any failure that would create a problem on both cards would have to originate in the data buffer or further up the recording chain inside the 633.

"Infinity" and "winged frogs" notwithstanding.

Cheers,

Brent Calkin

 

Share this post


Link to post
Share on other sites

Way out of my depth here, but here I go anyway. The 633 lets you configure the CF or SD cards to record at a different format simultaneously ie: SD could be MP3, CF WAV. Does this not imply that the 2 media have 2 different "pipelines" feeding them?

Share this post


Link to post
Share on other sites

My experiences in both 7xx and 6xx machines lead me to believe in "all of the above".   It is possible on both series to have one media go bad and the other (or the 2 other in the case of 7xx) be fine, and it is also possible to have an issue cause a problem in both media.  Most of the (very occasional) issues I've had were the former type.  There were a few "bad day at the office" types, but maybe only 2 in many many years of use (like from the first availability of both 7xx and 6xx series).  There was a mention of having a full 744 HD cause an issue with the CF card recording--ok, we all know that recording to near-full media is a terrible idea so that can be discounted as human error.  For the OP: the "Occam's Razor" reason for the issue is unfortunately that you stopped recording w/o realizing it: my experience points to most issues being caused by human error.  No shame--it happens to everyone sooner or later.   If your discussions with SD indicate a card issue, then besides pitching BOTH cards in question, I'd consider sending the machine in to SD for a look-see.  I don't like mysteries, and after one 664 mystery an intermittent issue was found with one of the card-slots that was then fixed.

Share this post


Link to post
Share on other sites
53 minutes ago, chrisnewton said:

Way out of my depth here, but here I go anyway. The 633 lets you configure the CF or SD cards to record at a different format simultaneously ie: SD could be MP3, CF WAV. Does this not imply that the 2 media have 2 different "pipelines" feeding them?

yes, for me that implies the same thing.

the other thing is that personally, I trust that a company who's whole existence is based on reliable, high quality recording spends some time on thinking how to make recordings as redundant as possible and they probably come up with a better solution then I could. 

And if in doubt, it's best to clarify with the engineers (as Brent did) rather then wildly speculate. Tempted to try the pulling the card next time I have my recorder fired up ; )

 

Share this post


Link to post
Share on other sites

Yep, stuff happens as has been discussed here in depth. I've had my 744t fail completely. It happened once, lost everything on the drive. SD people were perplexed. I've also rented a 788t from one of the usual suspects which again failed completely very recently. Mid take, live show, not good. I had to pull the battery let it sit for a sec and restart. After explaining what happened they could hardly believe me and they told me that they were going to send in the 788t for a factory checkup. Called a week later to inquire about the factory findings and it turns out they never sent it in because they couldn't replicate the problem I had. Luckily, I had the 744t on backup and I was able to use something from it to cover the scene. I guess I come back to my opening line: stuff happens! It sucks when you don't catch it in the moment to request a re-take. I never liked how easy it is to accidentally hit the 'stop' or 'rec' button on the SD stuff.

That being said, when I upgraded my recorder I did not go the SD route...or a Zaxcom, though I love the Zax wireless.

Share this post


Link to post
Share on other sites
6 hours ago, Freeheel said:

So after talking to SD tech support, a media failure on one card of a 633 should not affect the other one.  You can test this by pulling a card out of its slot while recording. (they suggest that you reformat the card after trying this)

Any failure that would create a problem on both cards would have to originate in the data buffer or further up the recording chain inside the 633.

"Infinity" and "winged frogs" notwithstanding.

Cheers,

Brent Calkin

 

Brent, That is in complete contrast to what they told me. To use their words I was told that "one card tells the other one to go bad". I completely  that this position is nonsense. So apparently, we now have a clear understanding,with what sound devices is now telling vs. what they told me then, that the second card is not a backup. That conclusion must be obvious since both cards failed and the chances of having both cards fail at the exact same frame number, certainly must be in the tens of millions to one, if not more than that.

Sincerely, Martin 

21 hours ago, Mobilemike said:

I'd be curious to hear what someone from Sound Devices says about what MartintheMixer posted. After all it's all speculation until we hear from them. 

Mike, You're saying it's speculation, on the part of who?

Sincerely, Martin 

7 hours ago, chrisnewton said:

Way out of my depth here, but here I go anyway. The 633 lets you configure the CF or SD cards to record at a different format simultaneously ie: SD could be MP3, CF WAV. Does this not imply that the 2 media have 2 different "pipelines" feeding them?

Chris, I like that theory, but the problem this leaves you with, this still provides no explanation of how both cards failed without the one pipe theory. Since we know this was a total failure to record, almost certainly, wherever that "record" starts happening, prior to sending a particular file type, that is what failed here. Assuming what one poster has stated here is correct, then that would mean that Sound Devices has made two, contrary statements concerning this problem.

Sincerely, Martin 

Share this post


Link to post
Share on other sites
8 hours ago, cbsixty said:

I read the whole topic now and must say: I didn't really understand this infinity-thing. Could you explain it again?

Sure, You've seen those signs on cheesy restaurants that say "Free beer tomorrow"?  

Same thing, that free beer is located at infinity, your never gonna get it. 

Share this post


Link to post
Share on other sites
1 hour ago, MartinTheMixer said:

Brent, That is in complete contrast to what they told me. To use their words I was told that "one card tells the other one to go bad".

That is what you think they told you. After this thread, I tend to believe they told you something completely different and you just inferred something else from that. 

Share this post


Link to post
Share on other sites
17 minutes ago, Constantin said:

That is what you think they told you. After this thread, I tend to believe they told you something completely different and you just inferred something else from that. 

Yes, I think they told me that, because they told me that, and in response to them telling me that is when I made the comment that so if you had 50 card slots you would just greatly increase your chances of having no audio files, since the second card is not a backup, and since one card can tell the other card to "go bad". And the answer was yes. And I guess Brent just inferred what they told him also. If you look back at Brent's comment, you will notice that he stated that Sound Devices stated that "it should not". If that is the actual wording from Sound Devices that's not a very definitive statement. Of course it also "should not" fail to record, but it did. I would be willing to bet if you asked Sound Devices should the 633 fail to record? The answer would be no "it shouldn't." But it did. 

Share this post


Link to post
Share on other sites

Just off phone with Sound Devices tech support again, confirming information mentioned earlier on this thread. The CF and SD cards are not back up to each other, but they both write the same data at the same time. On his machine he hit record, then pulled the SD card, received grey boxed "SD warning I/O error", machine kept recording onto the CF. Full file good on CF, nothing written to the SD. So if my Transcend (no longer approved) SD card had failed, then I should have the full recording on the CF. I did not. Both CF and SD had same six minutes of 18 minute recording. So he can only speculate 1) the buffer was overloaded, but then I should of received a "overflow I/O warning", or 2) a hardware issue above the buffer. He's never seen this before. So no satisfactory answer. Normally I would suspect my own operator error, but in this particular case I recall looking down at the screen, seeing the machine in record and the time rolling, then reporting the 18 minutes to the producer. Thanks for everyone's help.

Share this post


Link to post
Share on other sites
Freeheel   

Hey Tom, sorry to hear that that the files were not recoverable and you have no satisfactory answer to the problem.  This kind of thing is why I still generally refuse to send a scratch track to camera, and always send a high quality 2 track mix.  

Out of curiosity, were you using a higher sample rate than 48 Khz?

 

And Ivanovich, I mentioned in my call to SD tech support that I consider a menu optional "2 handed" Record Stop function to be the most important requested feature on their list.  Anyone who agrees with me please contact SD and request this for the next software update.  It's not so bad when bend over to tie my shoes (with my bag on) and realize that I knocked it into record... but not so cool when the same thing happens when I'm already in record and the machine buttons off.  It should not be so easy to pull a run' gun machine out of record- and it's really easy to do if skiing or running while wearing a 633.

Cheers,

Brent Calkin

Share this post


Link to post
Share on other sites
1 hour ago, tomsalyer said:

Just off phone with Sound Devices tech support again, confirming information mentioned earlier on this thread. The CF and SD cards are not back up to each other, but they both write the same data at the same time. On his machine he hit record, then pulled the SD card, received grey boxed "SD warning I/O error", machine kept recording onto the CF. Full file good on CF, nothing written to the SD. So if my Transcend (no longer approved) SD card had failed, then I should have the full recording on the CF. I did not. Both CF and SD had same six minutes of 18 minute recording. So he can only speculate 1) the buffer was overloaded, but then I should of received a "overflow I/O warning", or 2) a hardware issue above the buffer. He's never seen this before. So no satisfactory answer. Normally I would suspect my own operator error, but in this particular case I recall looking down at the screen, seeing the machine in record and the time rolling, then reporting the 18 minutes to the producer. Thanks for everyone's help.

Ok, there we go. 2nd card is not a backup. So we can forget about the one card telling the other card to go bad. 

Share this post


Link to post
Share on other sites

"Not a back up to each other" is an incomplete and misleading statement, right?  Because of course they are a back up to each other--that's one reason why there are 2 cards.  We have to interpret what the this phrase means in practice: thus: we know you can pull a card during recording (on 6xx) and card 2 continues just fine.  We know a card can go bad in one of a number of ways, thus not properly recording, and the other card will continue to record normally.  What is being said is that if there is a problem upstream of the cards, as in the recorder firmware or the physical circuits routing signal to the card slots, then having two cards will not help you.  This kind of issue can happen (rarely) in non-SD recorders as well, computer-based systems etc etc (was a bugaboo of live-show laptop recording in FW400, for instance).  If you have approved cards in good shape and make sure you don't get twisty-fancy with your menus while you are rolling, you will be fine except for that one day when the sunspot-activity or etc gets you.  If one can't accept this small but real risk inherent in file-based recording, then one should consider the risks (and disasters) we had with DAT, DVD-RAM and 1/4" CTTC recording in the field.  The current gear is way better.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×