Data Recovery Case File · Legacy & Obsolete Media · A Proprietary Machine
An Interrupted Save Leaves the Project Index Incomplete, Not the Head on the Surface
His enquiry offers a hypothesis alongside the facts, and the facts are better than the hypothesis. A standalone hard disc recorder whose mains plug was knocked out mid-save, where "I'm guessing the record head must be stuck on the disc." That is not what a power interruption does: heads retract when power is lost, and what was actually left incomplete is the machine's own record of where the project lives.
| Media | Standalone multitrack hard disc recorder fitted with an 80GB legacy interface drive — power removed during a save operation; proprietary volume format |
| Reported situation | Studio hard disc recorder in ordinary use · project save initiated by the owner · mains supply disconnected accidentally during the save · machine shutting down mid-operation · head adhesion suspected by the owner · project content required |
| Fault class | Write interrupted during a structural update on a proprietary volume — audio regions written and index update incomplete; mechanism not implicated by the interruption |
| Equipment used | Head adhesion excluded by inspection before any assumption of mechanical damage · drive removed and read independently of the recorder · imaged write-blocked at the block level via a legacy interface bridge · proprietary volume structures interpreted from the image · audio regions carved by signature where the project index did not reconcile |
The decode: what losing power actually does, and where the project went
Why heads do not stick when power is cut: retraction is designed for exactly this. A drive holds enough residual energy in its spinning platters to park the heads as power fails, using the motor's own momentum to drive the retraction.
Why that mechanism exists: power loss is expected rather than exceptional. Drives are built on the assumption that supply can vanish at any moment, and unpowered parking is the first thing they are designed to do.
What his interruption did instead: stopped a write partway. The recorder was updating its record of the project when the machine stopped, so that update is incomplete.
Why that matters more than it sounds: a save on a recorder of this kind writes structure rather than audio. The audio was already on the disc as it was recorded; the save was writing where each part of it belongs.
Why that is the encouraging half: the material is present. Recorded audio occupies the great majority of the disc and was written long before the save — what is missing is the map, not the territory.
What makes this machine harder than a computer: the volume format is the manufacturer's own. A disc from a standalone recorder is not readable by an ordinary computer, which is why nothing sensible happens when it is connected to one.
Why that is not an obstacle to recovery: the audio itself is stored in recognisable form. Regions can be located by their own structure and extracted as ordinary audio, independently of the machine's project database.
What is lost when the index cannot be rebuilt: the arrangement rather than the sound. Track assignments, positions and names live in the project record, so audio may return as separate takes requiring reassembly.
Why the drive should not go back into the recorder: the machine will attempt to make sense of a half-written project on start-up. Whatever it does to reconcile that is a write to the region being reconstructed.
Why the drive being a standard interface type helps: it is an ordinary drive in an unusual role. It reads on ordinary equipment through a bridge, and nothing about the recorder is required to capture it.
On the bench
Head adhesion was excluded by inspection before any assumption of mechanical damage — drives retracting heads on power loss using residual momentum in the spinning platters, a mechanism designed for exactly that event. The interruption instead halted a structural write, the save recording where project material lies rather than the audio itself, which was written during recording. Proprietary volume formats are not interpretable by ordinary hosts. Audio regions were carved by signature where the index did not reconcile.
The outcome
Head adhesion excluded by inspection, the drive read independently of the recorder, and audio regions carved where the project index did not reconcile. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode: losing power does not leave a head on the surface — drives park themselves using the platters' own momentum. What stopped partway was the machine writing down where your project lives.
A standalone recorder interrupted mid-save
Don't put the drive back in the machine and don't power the recorder again, because it will try to reconcile a half-written project on start-up and that reconciliation writes to the region being reconstructed. Set aside the worry about the head: drives retract on power loss by design, using the momentum of the spinning platters. What was interrupted was a structural write — the save records where your audio lives rather than the audio itself, which went down during recording. The material is very likely all there.
Don't power it again — call Manchester Data Recovery on 0161 871 0788; head adhesion excluded by inspection, drive read independently of the machine, audio carved by signature where the index fails.
Request a quote online →
Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.