A finished reel can look correct and still contain fewer labels than the counter shows. Clear-on-clear constructions are particularly unforgiving because the sensor has to distinguish a transparent label from a transparent liner at production speed. The counter display is only useful after the sensor, web position, counting direction and waste-correction method have been proved with the actual job. Treat that proof as part of setup, not as an assumption built into the machine.
Define the quantity that the customer is buying
Write the finished quantity as usable labels per roll. State whether the leader, trailer, setup labels, marked rejects, splices and any labels removed during inspection are excluded. If a job has more than one lane, decide whether each lane must carry the same quantity or whether rolls may be completed independently. A total web count can hide a shortage in one finished lane.
Keep length and label count as separate values. Web length is useful for material control, but it does not prove the number of usable labels when pitch changes, gaps vary or waste has been removed. For booklet, coupon or irregular labels, identify the feature that represents one finished product and where the counter is expected to detect it.
Give the operator a tolerance and a hold rule. If the order calls for an exact count, a mismatch must be reconciled before packing. If an agreed overrun is permitted, record it on the roll label rather than leaving the packing team to estimate quantity from reel diameter. The job ticket should say which number goes on the carton.
Match the sensor to the label and liner together
Put the actual construction in the setup trial: facestock, adhesive, liner, varnish, print, metallised area and label shape. Sensor selection depends on the contrast or material difference between the label and its carrier. Technical guidance for label sensors makes the same point: both the label and carrier matter, and transparent labels on transparent liner may need a sensing method designed for that combination.
Do not assume that one clear job proves every clear job. A thin clear film on a glassine liner behaves differently from a clear film on a clear PET liner. Printed areas, foil, heavy varnish, two-layer booklets and uneven gaps can change the signal seen by the sensor. Keep retained samples of the approved construction with the recipe and identify their material revision.
PRATI's published machine configurator describes a multitask counter for labels or length and a counting photocell intended for opaque and clear labels. That is a configuration statement, not a guarantee for every construction. Confirm the sensor supplied on the exact machine and prove the real substrate during a factory or site acceptance test.
Teach the sensor on a stable part of the web
Position the sensing point where the web is flat, laterally stable and properly supported. Keep it away from a fluttering span, an edge that moves with guiding corrections or a printed feature that changes from label to label. If the sensor can move across the web, mark the approved position for each lane or job so the next setup does not begin from memory.
Teach the sensor using the actual label and gap according to the installed sensor procedure. Run several labels slowly and watch the switching indication. Then test the least favourable area, such as a small gap, clear-on-clear section or printed block near the sensing line. A clean signal at threading speed may become intermittent after the web accelerates.
Record sensitivity, recipe, sensor position and any threshold or teach status available from the machine. Clean the sensor faces and inspect the slot or web path before adjusting sensitivity. Adhesive dust, paper dust and coating deposits can reduce the signal margin, and increasing sensitivity can temporarily hide a cleaning or alignment problem.
Test start, stop and reverse movement
An inspection rewinder does more than run forward at steady speed. The operator may slow down, stop at a defect, pull material by hand or reverse the web for correction and re-inspection. Establish what the installed counter does in each condition. It may add, subtract, pause or create a second count event depending on the control logic and sensor arrangement.
Make a controlled test strip with a known physical quantity. Run it forward, stop twice, reverse past several gaps and return to the original position. Compare the displayed count with the starting quantity and the web position. Repeat the test at the lowest and highest planned production speeds, including the automatic slow-down and stop used to finish a roll.
If the counter does not compensate for every reverse action, define the operator correction. The correction must be visible in the roll record and tied to the reason for the reverse movement. Do not ask operators to remember an unwritten offset at the end of a long job. Save the count before and after each defect correction where traceability is important.
Reconcile removed labels before the next roll starts
When a defect is cut out, record how many label repeats left the accepted web. Include any labels lost in the splice and any extra material used to restart. On multi-lane work, one damaged lane may lose a different quantity from the others. Split or rebuild the rolls according to the order rather than accepting one master counter value for every lane.
Resetting the counter at a new core is necessary but not sufficient. Check that the roll-break position, label pitch and programmed slow-down point produce the intended usable quantity after the leader and trailer are added. A late stop may create an overrun; an early stop followed by manual pulling can create an unrecorded difference.
Keep a simple equation on the production record: incoming usable quantity minus setup waste, rejected labels and splice loss equals accepted output plus retained samples. Investigate any balance that does not close. The purpose is not paperwork for its own sake. It prevents a packing-line shortage from being discovered after the roll has left the converter.
Release the count with a physical check
For the first roll of a clear-label job, verify the counter against a physically known sample or another approved counting method. Check the first roll again after it has been removed from the shaft and labelled. For critical work, repeat the comparison at planned intervals and after any sensor cleaning, recipe change, material splice or construction change.
The finished-roll record should identify the job and revision, material construction, parent roll, lane, sensor or recipe, target quantity, displayed count, manual correction, rejected quantity, splice count, finished quantity and operator. Hold the roll when the counter signal was unstable, the balance does not reconcile or the material differs from the approved setup sample.
Royal Coat represents PRATI inspection, slitting, rewinding and finishing technology across India, Bangladesh, Sri Lanka, Nigeria and Kenya. For a clear-label counting discussion, send the Royal Coat team the label and liner construction, web width, lanes, label size and gap, speed, finished quantity, permitted splices, existing sensor details, reverse-inspection method and samples of the missed count.
Questions from production and purchasing teams
What to ask about clear label counting and rewinding
Why can a label counter miss clear labels?
The signal difference between a transparent label and its liner may be too small or unstable for the selected sensor, position or teach setting, especially at production speed.
Can the same counter recipe be used for every transparent label?
No. Facestock, liner, adhesive, print, varnish, metallised areas, thickness and gap can change the sensing condition. Prove each construction or controlled material family.
Should the counter use labels or metres?
Use label count for the customer's usable quantity and retain length as a separate material-control value. Do not treat one as proof of the other.
What should happen to the count when the rewinder runs backwards?
Confirm the behaviour of the installed counter by test. It may subtract, pause or need an operator correction, and that rule should be written into the job procedure.
How should the first finished roll be approved?
Compare the displayed count with a physically known quantity or another approved method, reconcile waste and splices, and save the result with the roll identity.