LBA vs PBA Explained: Why SMR Drives Are Treated as Physical Drives When Formatted or Deleted
Summary: An SMR drive can wipe deleted data within minutes. Its overlapping tracks can only be rewritten as whole zones, and to avoid slow rewrites later, the firmware quietly clears freed zones in the background. That leaves a very short window for recovery. If you notice data loss, switch the drive off straight away and contact a specialist recovery lab.
Delete a file from a traditional hard drive and, most of the time, nothing dramatic happens. The operating system simply forgets where the file was, and the data stays on the disk until something else is written over it. That is why recovery tools usually work.
SMR Responds Differently
An SMR drive responds to a delete command differently. Within minutes, its firmware can start erasing the affected zones, which shortens the recovery window considerably. Recovery is still possible with methods designed for SMR, though it works quite differently from standard drive recovery.
In this guide, we look at how logical and physical block addresses work on each drive type, and why SMR drives react as they do when files are deleted or a drive is formatted.
How Does a Hard Drive Translate Logical Addresses to Physical Locations?
When your computer asks a drive for a file, it never says where on the platter to look. It asks for a numbered block, and the drive works out the rest. This table is one reason recovery works on some drives and fails on others, which is why the LBA vs PBA relationship is worth understanding before we look at SMR.
Logical and Physical Block Addressing
Logical Block Addressing (LBA) is the face the drive shows to the computer. To the computer, the drive looks like one long numbered list of blocks, starting at LBA 0 and running to the full capacity. Physical Block Addressing (PBA) is the reality behind that list, the exact spot on the platters where the data is stored, identified by head number, track position and sector offset.
Linking the two is a firmware module called the translator. Give it any LBA from the computer, and it finds the matching PBA. Almost everything in hard drive data recovery depends on how well that link survives.
How Does the LBA-to-PBA Map Work on CMR Drives?
On Conventional Magnetic Recording (CMR) drives, tracks run side by side with small gaps between them. The gaps make it less likely that the write head will accidentally touch a neighbouring track. They also let the drive overwrite a single sector in place, without relocating a large area every time one block changes.
A Map That Rarely Changes
The result is a map that rarely changes. On a healthy CMR drive, an LBA keeps pointing to the same physical sector unless the drive finds a fault. Then the firmware steps in through defect management. The primary defect list records sectors that were faulty from the factory, and the grown defect list records those that fail later in service. When a sector becomes unreliable, its LBA is redirected to a spare.
Apart from these cases, logical and physical addresses stay in step. That steadiness is why experts can find deleted data on a CMR drive by scanning the logical address space. Why does that matter? Because SMR drives take that steadiness away.
How Do SMR Drives Change the Standard LBA-to-PBA Relationship?
Shingled Magnetic Recording (SMR) increases drive capacity by placing tracks closer together. Each new track overlaps part of the previous one, much like roof shingles.
Track 1 ==========
Track 2 ==========
Track 3 ==========
This works because of a physical difference between the two heads. The write head must be wider, as it needs a larger contact surface to produce a magnetic field strong enough to flip the grains on the platter. The read head only detects existing patterns and can be much narrower. It can therefore read the thin strip of a track that remains exposed after the next track overlaps it. The overlap enables higher capacity, but it also changes the SMR drive’s physical block address landscape in ways that affect every recovery attempt.
SMR Drives Organise Overlapping Tracks into Zones
Overlapping tracks create a write dependency. An SMR drive cannot safely update one sector in isolation, because rewriting a sector in the middle of a shingled group may overwrite data in the neighbouring track.
To control this, the drive divides its recording surface into bands, or zones. Each zone contains a run of overlapping tracks, with guard regions between zones to stop a write in one from affecting another. Data within a zone must be written sequentially from one end to the other.
If one sector in the middle of a zone needs updating, the drive performs a Read-Modify-Write (RMW) cycle. The firmware may need to do the following.
The Read-Modify-Write Cycle
- Read the valid contents of the zone.
- Modify the requested sectors in memory or temporary storage.
- Rewrite the affected sequence of tracks.
- Update its mapping information.
This zone-level constraint is why the physical address of any piece of data is far less stable on an SMR drive than on a CMR drive.
Why Does the Physical Location of Data Keep Shifting on SMR Drives?
A full Read-Modify-Write (RMW) cycle for every random write would slow an SMR drive badly. Device-managed drives avoid this with a persistent media cache, a region of the platter (typically 20 GB to 30 GB) formatted as CMR, where incoming random writes land first at full speed.
During idle periods, background processes called destaging or compaction move the cached data into the final shingled bands. A secondary translator records whether each LBA is in the cache or in a shingled zone, along with its exact physical coordinates.
As a result, the physical location of any LBA keeps shifting during routine housekeeping. If power is lost mid-move, the translator can be corrupted. The drive then cannot find data that is still on the platters, and standard LBA vs PBA data recovery methods that read through the logical interface return nothing usable.
CMR and SMR at a Glance
The table below compares how the two drive types record, update, and erase data, and what each means for recovery.
| Feature | CMR Drive | SMR Drive |
|---|---|---|
| Track Layout | Separate tracks with gaps | Overlapping tracks in zones |
| Update Method | Overwrite a single sector | Read-Modify-Write across a zone |
| LBA-to-PBA Mapping | Stable | Changes continuously |
| Effect of Delete or Format | Metadata change only | Firmware may erase zones physically |
| Typical Recovery Window | Until new data overwrites the sectors | Minutes, until firmware housekeeping completes |
Why Does Deleting or Formatting an SMR Drive Trigger Physical-Level Erasure?
The difference follows directly from the architecture described above. On a CMR drive, deletion is a purely logical operation. On an SMR drive, it starts a physical process that can destroy the data within minutes.
What a Delete or Format Does on a CMR Drive
- A file deletion or quick format changes only the file system metadata, such as the Master File Table (MFT) or the file catalogue.
- The operating system marks the sectors as available for reuse, but sends the drive controller no instruction to alter the magnetic bits.
- As the LBA-to-PBA mapping is static, the controller has no reason to act on the freed space until the operating system issues new write commands.
- The physical block addresses keep the original magnetic patterns. Experts can scan those LBA ranges, locate file signatures and reconstruct deleted content for as long as the sectors stay untouched.
What a Delete or Format Does on an SMR Drive
On an SMR drive, the same command sets off a chain of firmware actions that begin erasing data at the physical level almost immediately.
- When the operating system deletes files or formats a partition, it sends a TRIM command (on SATA interfaces) or an UNMAP command (on SCSI and UASP interfaces) to the drive controller.
- The secondary translator drops the mapping entries for those LBAs from its tables.
- From then on, the firmware intercepts any attempt to read those addresses and returns zeroes, even if the magnetic data still exists at the original physical locations.
- Within seconds or minutes, the drive’s autonomous garbage collection begins reclaiming the orphaned zones.
- The firmware moves any remaining valid sectors to new zones, then physically overwrites the old bands to prepare them for future sequential writes.
Why Does SMR Firmware Prioritise Physical Erasure After Deletion?
The aggressive response is a direct result of the RMW penalty. If the drive kept stale data in shingled zones, every later write to those zones would force a full Read-Modify-Write cycle. By clearing zones that TRIM has marked as empty, the firmware removes that overhead and can write to them sequentially at full speed.
What Decides the Recovery Window
For recovery, this creates a critical difference from CMR drives. The recovery window on an SMR drive is not decided by whether new user data happens to overwrite the old sectors. It depends on how quickly the drive’s own firmware completes its housekeeping. Once the translator drops the LBA mappings and garbage collection rewrites the zones, LBA vs PBA data recovery through the standard host interface becomes impossible, regardless of the technique used.
What Makes Data Recovery from SMR Drives Difficult?
SMR drives bring obstacles that do not exist on CMR hardware.
Zero-Fill Problem Caused by Translator Corruption
The most common failure is corruption of the secondary translator, typically after a power loss during background destaging. The drive still reports its full capacity to the operating system but returns zeroes for every sector read. The magnetic data remains on the platters, but the firmware has lost the index that connects LBA ranges to their physical addresses. Standard recovery software cannot get past this zero-filled layer, as it works entirely through the host LBA interface.
Shorter Recovery Window from Background Firmware Activity
An SMR drive is never truly passive. The moment it receives power, its firmware begins destaging data from the media cache and running garbage collection across orphaned zones. If the translator is already corrupt, or the heads are degrading, these processes can write corrupted mappings into the shingled regions or overwrite valid data with stale metadata. Experts must therefore apply a firmware-level hardware write lock before the drive completes its power-on initialisation, freezing the platter state and preventing further damage.
Risk of Rapid Mechanical Degradation
The RMW cycle magnifies the effect of any mechanical fault.
- On a CMR drive, a failing head affects only the sectors it cannot read.
- On an SMR drive, one unreadable sector forces the firmware to read, modify, and rewrite an entire shingled band of 15 MB to 40 MB.
- A marginal head is then under sustained thermal and mechanical stress across a far larger range of data.
- This can trigger a cascade failure, in which the head degrades further or damages adjacent healthy tracks.
- A single bad sector can grow into a platter-wide failure within hours of the drive being powered on without intervention.
Wrapping Up: Talk to Stellar About SMR Drive Recovery
At Stellar Data Recovery, we recover data from SMR drives in our ISO 9001- and ISO 27001-certified labs, using hardware- and firmware-level tools that bypass the standard host interface entirely.
How Stellar Approaches SMR Recovery
- Our Stellar data recovery experts reconstruct corrupted SMR translators by reading raw magnetic data directly from the platters and rebuilding the LBA-to-PBA mapping manually where the firmware no longer can. This restores access to every recoverable physical block address.
- Every SMR drive that reaches our facilities is write-locked at the firmware level straight away, before background processes can alter the platter state.
- Our in-house R&D team at Stellar keeps support updated for current SMR firmware revisions across all major manufacturers, so our procedures reflect the translator architecture of each drive model.
If you are unsure whether your data can be saved, please call Stellar Data Recovery on 1800 102 3232 for a free SMR drive recovery consultation. The sooner you call, the better the chance of recovery.
Frequently Asked Questions
1. Can Windows Tell Me Whether My Hard Drive Uses SMR or CMR?
Not normally. Windows does not identify a drive’s recording technology. Look up the exact drive model in the manufacturer’s specifications or a trusted compatibility database. Knowing this distinction is also useful when assessing LBA vs PBA data recovery, as CMR and SMR drives handle physical data mapping differently.
2. Does a Full Format Affect SMR Recovery Differently from a Quick Format?
Yes. A quick format mainly recreates file system structures, although later firmware activity may still reclaim deleted zones. A full format also writes across the storage space, which makes recovery far less likely on both SMR and CMR drives.
3. Does Powering Off an SMR Drive Immediately After Accidental Deletion Improve Recovery Chances?
Yes. It is the single most effective action. Powering off stops the garbage collection and destaging that would otherwise begin overwriting the orphaned shingled zones.
4. Does Cloning or Imaging an SMR Drive Before Recovery Preserve Data the Host Interface Cannot See?
No. A standard clone copies only what the firmware serves at the LBA level. If the translator has dropped mappings or returns zeroes for deleted zones, the clone captures those zeroes and not the underlying magnetic data. Recovery from an SMR image requires specialised hardware that reads below the translator and captures raw data from the shingled bands.
5. Can an SMR Drive's Translator Be Rebuilt Without Specialised Hardware?
No. The secondary translator is proprietary firmware stored in a reserved service area on the platters. Consumer-grade software cannot rebuild it.