Most people look at the blue or yellow health box and stop. Technicians scroll down to four specific rows and read one column — the raw values. Here is what 05, C5, C6 and C7 actually mean, in plain words.
Open CrystalDiskInfo and you get a table of twenty or more attributes with confusing columns — Current, Worst, Threshold, Raw Values. Almost all of it is background noise. Four attributes tell you whether the disk surface and its connection are healthy, and one column — the raw value — is the only number you need to read. This page teaches that reading, the technician way.
👆 Example readings. On a healthy drive the first three raw values are 0. C7 is the odd one out — it points at the cable and connection, not the disk surface.
CrystalDiskInfo shows each attribute twice: a normalised score (Current / Worst / Threshold) and a Raw Value. The normalised scores are the maker's own grading and they move slowly and politely. The raw value is the actual count the drive firmware recorded — sectors damaged, errors seen, events logged.
One setting matters here: under Function → Advanced Feature → Raw Values, choose 10 [DEC] so raw values show as normal decimal numbers instead of hexadecimal. After that, the rule for the big three is brutally simple: raw 0 is healthy, raw anything-else needs attention.
When the drive finds a sector it can no longer read or write reliably, it retires that sector and quietly substitutes one of its spare sectors. That substitution is called a reallocation, and 05 counts how many have happened.
Think of it as the drive's own confession: this many spots on my surface have already died. A small number that appeared years ago and never grows can be lived with. A number that climbs between checks means the surface is actively failing — drives rarely recover from that, because the same wear that killed the first sectors is spreading. Growth matters more than the total.
A pending sector is one the drive has failed to read and has put on a watch list. It has not been replaced yet — the drive will decide when something next tries to write to it. Until then, any file sitting on that sector is effectively unreadable.
C5 is the most urgent of the four. It means damage exists right now and data is at risk right now, not historically. If C5 is non-zero, back up anything important on that drive before running long scans — a full surface scan forces the drive to work through exactly these weak spots.
C6 counts sectors that could not be recovered even after the drive used all its error-correction tricks. In plain words: data written in those places came back wrong or not at all.
Any non-zero C6 means real files have already been damaged — this is the count that turns a slow PC complaint into a data-loss conversation. A drive with C6 above zero should not be trusted with the only copy of anything, whatever the big health box says. C5 and C6 often appear together, and together they are the classic CrystalDiskInfo Caution pattern.
C7 is different from the other three: it counts transfer errors between the drive and the computer — data that arrived corrupted over the cable. It says nothing about the disk surface.
A high C7 with clean 05/C5/C6 usually means a loose, bent or dying SATA cable, a poor connection, or a port problem — not a dying drive. The fix order: power down, reseat or replace the SATA data cable, then watch whether the C7 count keeps rising. The old count never resets — it is a lifetime total — so judge it by whether it grows after the cable change, not by the number itself. Laptop owners can mostly ignore C7; there is no cable to change.
Video me ek asal kharab disk ke SMART values live padh kar dikhaye gaye hain — kaunsa number dekh kar technician turant faisla leta hai.
▶️ Watch Full VideoThere is no single magic number, because the pattern matters more than the total. A count that has sat unchanged for a year is low-risk. A count that grows between two checks a week apart is a failing surface, even if the total is still small. Growth is the danger signal.
Set CrystalDiskInfo to decimal (Function, Advanced Feature, Raw Values, 10 [DEC]). In hexadecimal mode a raw value like 000000000010 means 16 events, which misleads many people into thinking the count is enormous or tiny.
No honest tool does this, and it would prove nothing if it did. The counts live in the drive's firmware as a permanent record. Software claiming to clear SMART or restore Good health is editing the report card, not the disk.