Hard Disk File System investigation for the Magnavox 537, 535, 533, 515, 513, 2160A, 2160, 2080 &...

I hooked my drive outa my 3575 in here a while back and the OS saw nothing but I didn't try every possibility that might have worked.
I didn't want to blow up the drive or contents till I saw if my upgrade worked, which it still does.
IF somebody does fully figure it out it would be great for folks with multiple drives hooked up in external bays and things, provided they just wanted to do a direct edit and burn easily on their PC rather then in the recorder.
I have some pretty good and easy editing programs that make the job of dumping commercials and choosing the video format a breeze provided you want to pay for the program, and I'm sure there are free versions that work well too.

This question was asked the day the first DVD/HDD recorder shipped 8 years ago, and the answer today is the same as it was then: forget it. You cannot really perform any useful manipulation of video content of a DVR HDD once you remove it from the recorder and put it in a PC. You might as well take the HDD from a 1991 Macintosh SE and plug it into a Dell Windows XP box: same result. The HDD requires the proprietary recorder OS to access the material in any normal fashion.

Recent experiments by several members here with Stellar Phoenix recovery/hex software has resulted in salvage options for some recorder HDDs, i.e. they have failed in the recorder but a PC running the software can extract the file fragments. This usually entails seriously tedious and time consuming reassembly of the video fragments, nothing like normal video handling on a computer. And as you've hinted, the other option is to "clone" an existing recorder HDD, but this results in losing any additional capacity of the larger newer HDD.

In older, more expensive models from Pioneer, Panasonic and Toshiba one at least had the option of formatting DVD or RAM backups in a way that permitted high speed lossless return of the backed up HDD files to any replacement HDD. But I don't think the Magnavox design provides this option- perhaps wajo could chime in here with a confirmation? Maybe the Magnavox can do this with DVD+RW backups? If not, then there is really no way to upgrade its HDD capacity without sacrificing any recordings on the original HDD (or backing them up as normal DVD-Rs). On the plus side, the Magnavox is the easiest recorder in the world to upgrade with another HDD.

I bold-ed one of your remarks above for clarity. This issue of only being able to do a real-time re-encoded copy from DVD back to HDD is probably the root of advanced Magnavox users' frustration. With older machines of other brands, the high-speed lossless copy is both faster than the Magnavox and works in either direction, as long as the backup DVD is formatted as a backup DVD (not a video DVD-R) or you used DVD-RAM discs. For example, my 2005 Pioneer 531 began failing this past winter. This model (and other Pios, Panasonics, etc) can make a high speed lossless straight digital copy of 2 hours of SP material from HDD to backup-format DVD in about 12 minutes, so it took me about 4 hours/18 discs to back up an 80GB HDD. After repairing the unit, it took four hours to restore those files back to the HDD, for a total of 8 hours approx.

While dull work, 8 hours is not really that bad since I might need the backed-up files again later, and I was able to restore 38 hours of "live", editable, lossless material back to the HDD without enduring real-time and video degradation. Had I upgraded to a much larger HDD capacity, this restoration method would have preserved the larger useful capacity while retaining the original file quality (and defragmenting the drive in the process). While not ideal, its at least a reasonable alternative that lets you achieve the goal. Unfortunately new-generation designs like the Mags appear not to have bi-directional high-speed lossless dubbing, it only works in the HDD to DVD direction. This drawback must be weighed against other advantages, however: the Magnavox is half the price Pioneer and Panasonic recorders sold for, and it includes an excellent ATSC tuner not available in any other DVD/HDD model.

The Magnavox HDD replacement interface is not fussy or overcomplicated like some of the Panasonics and Toshibas (which were picky about specific HDDs and locked at their original capacity limit no matter how big the replacement), and the Magnavox uses a built-in service mode to replace its drives (Pioneers require expensive service remotes and service discs to accept new HDDs). If you drop a new 500GB HDD into your 160GB Magnavox, it recognizes the entire 500GB, no problem. The newest Magnavox MDR513 ships with a 320GB drive, which obviates the need to ever upgrade it really. The dream of shoehorning an enormous terabyte HDD into these DVD/HDD recorders loses its appeal when you realize how slow and cumbersome their navigation is: 320GB is the practical limit and I find 500 to already be way too much. You also hit a technical limit of 999 max titles on the HDD with all PVRs, so at a certain point higher HDD capacity becomes moot.

There is no mouse or command line in the recorder OS: navigating many hundreds of titles with the remote arrow keys is hopelessly inefficient even with the sleek scrolling Pioneer interface, the far clunkier Magnavox navigator would have me punching walls at anything beyond 160GB. Be careful what you wish for. (That said, I am in frequent contact with an AVS member who hot swaps terabyte drives in and out of his Pioneers, filling them to full capacity with aviation coverage. Its doable for such very long form material you expect to only view from the HDDs.)



There's no lack of interest, there is HUGE interest, its just that sales of DVD/HDD recorders have dropped off sharply and most of the old hands here long ago gave up the fantasy of easily-tampered-with HDDs. Hollywood stepped in early on, pressuring hardware mfrs to make the HDD raw data as inaccessible as possible outside the recorder chassis. They all colluded, resulting in an assortment of proprietary Linux-based file systems and OSes. No computer can host one of these HDDs and make any sense of them other than data salvage projects: certain software can reveal the individual files, but they cannot be directly manipulated as video. You have to manually catalog all the various pieces, stitch them together (blindly) in the correct playback order, then convert them into a standard video file format usable by and visible on computers (AVI, DiVX, whatever). The reward for all this effort is just not worth the time and trouble, unless of course you have priceless material on a damaged HDD that you must salvage at any cost.


See above: there is little to no interest in harvesting these raw files if they cannot be viewed or worked with as direct video files. Kinda defeats the point.


If you're referring to making a bit-by-bit clone of the original HDD to a larger replacement HDD, while retaining the additional capacity of the replacement drive, this has been attempted without success many times before. I mean, really, REALLY tried: hundreds of times by hundreds of European Sony RDR-HX recorder owners alone. Throw in another few thousand Panasonic, Pioneer, Toshiba and JVC owners and believe me: if there was a way to pull this off it would have been found by now. It can't be done: if you clone an existing drive, you clone the capacity limit right along with the videos. These mfrs are not as dumb as you'd think: they put more effort into enforcing capacity limits and frustrating HDD file access than they do any other aspect of the recorder. Very few recorder models will address a larger drive capacity than they shipped with: put in even a completely blank terabyte drive and the typical Sony, Panasonic, or Toshiba will format it as 80 or 160GB. Most Pioneers will recognize larger capacity up to a terabyte, if you use the stupid service remote/service disk tools. The Magnavox apparently has no capacity detection limit (as long as the new HDD is empty). Replacing the HDD in an older JVC is so absurdly difficult most owners give up and dump them in a closet (not even JVC remembers how to install them).

A few years back, there were two or three one-off, very chintzy low-end DVD/HDD models (RCA, Radio Shack, Polaroid) that supposedly did use near-normal HDD file formats. But these were rare exceptions to the rule, and such crummy-quality recorders that most users tired of them quickly or didn't bother. For all intents and purposes, DVD/HDD recorders are closed systems. If you want flexible, easy to handle video files and spontaneous expansion options, you're stuck with PC-based DVR solutions.

Yeah, hex editors seem to be the only things capable of seeing or extracting any useful material from these recorder HDDs. This was reported once or twice a few years ago by extremely obsessive owners who needed to salvage their damaged HDDs and were willing to pull their hair out in the process (as I recall it took them six months to extract and assemble four hours worth of video). Eventually, one particular hex editor became the "go-to" application for those who want to play with recorder HDDs: the Stellar Phoenix software I mentioned in an earlier post. This app seems to detect more file fragment variations on more different recorder structures. It used to cost something like $300 and was only available for Linux, but recently dropped to $59 and is now available for Linux, Mac, and Windows. I haven't used Stellar Phoenix myself, I can't be bothered, but there are many posts here regarding it (particularly if you search the "Panasonic HDD File Recovery- It can Be Done!" type of threads). AFAIK its still more of a "recovery" process than an everyday use thing, and doesn't help break the capacity limit code on cloned HDDs or allow larger HDDs to be formatted to full capacity in the recorders with capacity lockouts (because their motherboards control capacity recognition: nothing you can do to the HDD boot blocks or directory structure will get around that hard-wired restriction).

Peter,

I'm glad you're investigating this. This project is just too time consuming for me to be able to continue after I've looked at the disk and tried various programs to read it.

I believe the used/free sectors allocation table are stored somewhere (most likely at the beginning) of the disk. And all the sector addresses were calculated at the time we initialize the new hard disk with SKIP 079 command. People have reported that they can use the disk with a different Mag 2160 with all titles intact. And this enable the HDD farming like the setup that auskck has.

If the disk size was just one constant stored in that allocation table then it's not so difficult. But if the whole table was calculated up front using this size, then we'll have to modify each entry to ensure they all correct. And all this is just to be able to clone the disk to a larger size disk. The audio/video files structure could be on another table that stores the sector addresses. So we'll have to figure that out if we want to extract the titles.

I'd suggest you look for the name of one of recorded titles. See if you can find it. If you can't, then the names were possibly encrypted so that people can't find them easily.

Peter, while you wait for the other unit, could you post some of the smaller files you've extracted, like the VBI....DAT and the .IFO files? Some of us may want to take a gander as well.

Could one of the unknown sections be designated for the 6 hour auto-record buffer?

From: VizmayaTech VBI Data Sicing

What is VBI?

VBI stands for Vertical Blanking Interval. In analog video display, such as a CRT, a video frame is shown by drawing horizontal lines from the top to the bottom of the screen. Once a frame is completely drawn, the beam has to travel back to the top left, to start drawing the next frame. Since there is nothing to show while the beam retraces, this interval is blanked and hence called the vertical blanking interval.

VBI Data

During vertical blanking time, no video is shown. So the display is not consuming anything from the incoming video signal. This interval effectively leaves open a chunk of bandwidth that can be used for data communication. To keep things simple, the horizontal lines (referred to as just lines', henceforth) are continued to be transmitted during the vertical blanking interval, but the lines do not contain video. Each VBI line, or group of lines may contain data, encoded as per a variety of standards. The most common used VBI service is the closed caption service, which is used to show text subtitles.

To ensure that VBI data goes wherever the video goes, VBI data is actually encoded as video. This allows VBI data to be transmitted transparently across any carrier media (Over The Air (OTA), cable, satellite, fibre optic cable etc) that is designed to transmit the video signal.

Raw and Sliced data

The hardware that receives VBI data may just sample the VBI lines to collect the data bits, with no knowledge of the type of data it contains. This is called raw data. The raw data may then be processed by software to interpret the contents.

The other possibility is to have the VBI hardware interpret the VBI data, based on the type or standard used on the line(s). The output in this case is extracted information such as closed caption characters, or data packets. This is called sliced data.

Typical VBI data extraction hardware supports both modes. The 'sliced' mode is much more useful, since a lot of low level functions such as data extraction , error correction are handled by the hardware. Raw data is primarily used to work with non standard VBI data, and for re-encoding unmodified into an output video signal.

A note on Interlaced video

NTSC analog video uses interlaced video. This means that even though the video has 30 pictures (frames) per second, the video is transmitted as 60 fields by splitting each frame into an odd and even field.

How much data can VBI carry?

In NTSC, the VBI consists of 21 lines. Lines 1 to 9 are reserved' for television timing signaling. Lines 10 to 21 are available for data transmission. Line 21 is widely used for closed captioning, hence the available lines are 10 - 20, a total of 11 lines.

Each VBI line can transmit up to 288 bits. At a rate of 60 fields per second this amounts to a raw line data rate of 17, 280 bits per second. Note that actual useful data rate would be much lower, due to packetization and error correction overheads.

If the HDD becomes full with saved recordings, will that interfere with the auto-record buffer time? If it does not interfere I would think there has to be some method to reserve space for the buffer.

From FILExt.com

File Extension IFO

The IFO file type is primarily associated with 'DVD Info File'. An info file associated with a .VRO video file. The .VRO file contains the video stream and the .IFO file contains various scene and timing information. This is different from a .VOB file where that information is included in the file. Several manufacturers use .VRO/.IFO files. If present, the file VIDEO_TS.IFO consists of control and playback information for the entire DVD. VIDEO_TS.IFO is known as VMGI (The Video Manager Information file). This file is required to be present on a DVD-compliant disc.



Maybe this is only used for HDD->DVD dubbing purposes?

I found a free windows utility called "Disk Investigator 1.5" on ZDNet. It looks pretty good. It does Text, Hex, and Decimal. It reads raw disk data by cluster or you can read files directly from directories. You can also recover deleted data if you know how to find it. I haven't messed with it too much yet but I ran your IFO file.

Disk Investigator text output of TI000_0TI000_00.IFO file:

\\..w....................................................... ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ .....................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK.................48 Hours:514 Stargate SG-1..........................................-.......<<br /> ..............................9FK............. ...The Monopolar Expedition:223Bi..................................L......... ..................................................
...The Monopolar Expedition:223Bi..................................L......... ......................................................The Monopolar Expedition:223Bi..................................L......... ......................................................The Monopolar Expedition:223Bi..................................L......... ..................................................
...The Monopolar Expedition:223Bi..................................L......... ......................................................The Monopola
*********** End of Cluster ***********

Edit:
That's just the first of 50 clusters in the file as downloaded on my PC HDD.

The ATSC/QAM tuner in both OTA and cable modes has gone haywire in my first 2160A. I had bought the extended service plan from WalMart and am awaiting the shipping label. On the phone with the service company lifetimeservice.com I was told breaking the warranty seal to remove and clone the HD prior to repair is an acceptable practice. After my new 160G HD arrives I'll clone the Hitachi in it now prior to sending in for service in case whatever it returns is missing my 54 titles, but I went ahead and dd'd the first 8 sectors in case anyone here who is more adventurous than I can make use of it. It shows 23h 13m free @ SP. First title is 1:47:34, 54th is 0:41:28.

h t t p : // fm.no-ip.com/ Tmp/ Dfsee/ magn1head.bin

What do you mean by "haywire" and what troubleshooting steps have your taken... Soft Reste, Hard Reset, etc.?

I'm really surprised to hear that they allow breaking the warranty seal. Very forward thinking! I have one machine under Walmart extended service, and I applaud this wise policy. (unless it's a fluke and everyone else will say it's no longer covered.... )