Nokia E51 MP4 video format problem

[Peter]

I have done some short videos.
A very short one
http://www.zen74158.zen.co.uk/temp-files-ph/hsi-ok.mp4
plays back OK with VLC media player, but a slightly longer one
http://www.zen74158.zen.co.uk/temp-files-ph/hsi-fault.mp4
fails, due to some error.
Can anybody illuminate this issue?
Very many thanks.
The 2nd video is actually quite important and I would very much like
to retrieve it...

G-Spot says no playback size, video or audio codec info can be found in
the faulty file. I'd guess at corrupted or missing metadata. Has the fle
been edited, or did you notice any problem when saving it?
A Hex Editor shows the same info in the start blocks of the files, but
totally different info at the end, which I take to mean the faulty file
has been corrupted while saving at some point. In particular, the file
creation date and recommended playback size are missing from the
metadata, so it may just be truncated, as these are the last pieces of
data to be written. Someone with better Hex editing skills than me *may*
be able to resurrect it, but it won't be easy.
From reading about similar problems on various video forums, I'd not be
too optimistic, unfortunately.

It must be corrupted because the phone won't play it either... but
quite possibly just at the end. I wish someone could help me recover
it.

The file was saved to a 4GB micro-SD card which had plenty of room on
it so the corruption is a mystery. It was not edited.

[The dog from that film you saw]

Post by Peter
It must be corrupted because the phone won't play it either... but
quite possibly just at the end. I wish someone could help me recover
it.
The file was saved to a 4GB micro-SD card which had plenty of room on
it so the corruption is a mystery. It was not edited.

tried loading it into virtualdub and then re-processing it as a new video
file? ( full processing mode ).

[Frank]

On Sun, 16 May 2010 16:20:02 +0100, in 'rec.video.production',
in article <Re: Nokia E51 MP4 video format problem>,

Post by Peter
It must be corrupted because the phone won't play it either... but
quite possibly just at the end. I wish someone could help me recover
it.
The file was saved to a 4GB micro-SD card which had plenty of room on
it so the corruption is a mystery. It was not edited.

Peter, I looked at your files the other day when you first posted
about this problem, but I didn't reply at the time because I was busy
and didn't have any good news to report to you anyway.

Based upon my analysis, it appears that the 'moov' atom is corrupt. In
fact, at least one of the programs that I used with your bad file
reported that there was no 'moov' atom in the file. A quick look at
the file with a hex editor program revealed that there is a 'moov'
atom present, however.

Given that the writing of the 'moov' atom is one of the very last
things that happens at file creation time, I don't think that it's too
likely that you'll be able to recover any useful data from the file,
at least not by using general purpose MPEG-4 video file manipulation
tools.

If you can locate a program that's been specifically written to
recover MPEG-4 video files from your particular make and model of
phone, then you just might have a chance at successful recovery.
Otherwise, I think that you'll just have to call it a loss.

The one final point that I can make is this: most programs/devices
that create MPEG-4 video files place the 'moov' atom at the end of the
file rather than up at the beginning of the file.

Why is this a consideration, you may ask? It's because many people who
create MPEG-4 video files wish to pseudo-stream them (with progressive
download effect) from an ordinary Web server (usually using a
Flash-based player that runs in the visitor's browser) and it just so
happens that the file will not progressively download if the 'moov'
atom is located at the tail of the file. It must be located at the
head of the file in order for the progressive download effect to take
place.

Upon examining your two files, I found that in both cases the 'moov'
atom was located at the tail of the file and not at the head of the
file and would therefore have to be relocated to the head of the file
if you wanted to serve them off of an ordinary Web server with
progressive download effect. There are freeware programs out there for
doing this and since they don't re-encode the audio/video data within
the file, there's no quality loss.

[Peter]

Frank (and others),

Many thanks for your detailed reply.

What I find puzzling is that I have ~ 4MB of mpeg video data and it
cannot be recovered.

I did some googling on "moov atom" and there is a lot of stuff out
there on it, but beyond my capability to follow.

I can re-create the video but it will cost me about 100 quid. It shows
an obscure aircraft avionics failure mode which would have been
valuable to document.

Post by Frank
On Sun, 16 May 2010 16:20:02 +0100, in 'rec.video.production',
in article <Re: Nokia E51 MP4 video format problem>,

Post by Peter
It must be corrupted because the phone won't play it either... but
quite possibly just at the end. I wish someone could help me recover
it.
The file was saved to a 4GB micro-SD card which had plenty of room on
it so the corruption is a mystery. It was not edited.

Peter, I looked at your files the other day when you first posted
about this problem, but I didn't reply at the time because I was busy
and didn't have any good news to report to you anyway.
Based upon my analysis, it appears that the 'moov' atom is corrupt. In
fact, at least one of the programs that I used with your bad file
reported that there was no 'moov' atom in the file. A quick look at
the file with a hex editor program revealed that there is a 'moov'
atom present, however.
Given that the writing of the 'moov' atom is one of the very last
things that happens at file creation time, I don't think that it's too
likely that you'll be able to recover any useful data from the file,
at least not by using general purpose MPEG-4 video file manipulation
tools.
If you can locate a program that's been specifically written to
recover MPEG-4 video files from your particular make and model of
phone, then you just might have a chance at successful recovery.
Otherwise, I think that you'll just have to call it a loss.
The one final point that I can make is this: most programs/devices
that create MPEG-4 video files place the 'moov' atom at the end of the
file rather than up at the beginning of the file.
Why is this a consideration, you may ask? It's because many people who
create MPEG-4 video files wish to pseudo-stream them (with progressive
download effect) from an ordinary Web server (usually using a
Flash-based player that runs in the visitor's browser) and it just so
happens that the file will not progressively download if the 'moov'
atom is located at the tail of the file. It must be located at the
head of the file in order for the progressive download effect to take
place.
Upon examining your two files, I found that in both cases the 'moov'
atom was located at the tail of the file and not at the head of the
file and would therefore have to be relocated to the head of the file
if you wanted to serve them off of an ordinary Web server with
progressive download effect. There are freeware programs out there for
doing this and since they don't re-encode the audio/video data within
the file, there's no quality loss.

[John Williamson]

Post by Peter
It must be corrupted because the phone won't play it either... but
quite possibly just at the end. I wish someone could help me recover
it.
The file was saved to a 4GB micro-SD card which had plenty of room on
it so the corruption is a mystery. It was not edited.

I've had random corruption for unknown reasons on all sorts of media
over the years. Sometimes all you can do is shrug your shoulders and
move on.

In this case, if, as you've said, it's only going to cost a hundred
bucks or so to re-shoot, I'd recommend that, as if you pay someone,
it'll probably cost you more even if they fail, and depending how you
value your time, it'll still be quicker and cheaper to re-shoot than
fiddle round yourself. The exception, as always, is if you're using it
as a learning experience so it will be easier next time.

Sorry I can't be more hopeful.

In the meantime, I've tried all the programs I know,and come up with
precisely zero results. The first sections of the two files are
identical, but as others have said, the video data seems to be
completely mangled.