Issue 002 → Section 05

Embroidery and colour.

One machine has three top speeds depending on the frame, and thread can cap all of them. Colour tolerances come in formulas rarely named. Nearly all figures are sellers' own.

ZSK’s new RACER II 6S is listed at 1,000 rpm on a border frame, 1,400 on a tubular frame and 1,000 on a cap. That is one machine with three top speeds, and the sheet does not say which one a quote should use. Then the thread decides. Coats says trilobal polyester can run above 800 stitches per minute and that rayon usually cannot exceed 600. A speed on a catalogue page and the speed on the floor are separate numbers.

Embroidery: what the spec sheet says

The multi-head field sits in a narrow band. Barudan’s Japanese lineup page gives its multi-head series 9 to 15 needles, 10 to 28 heads and a maximum of 1,000 to 1,200 rpm. Happy’s HCR3 lists 1,000 to 1,200 rpm on tubular frames and 850 to 1,100 on caps. Tajima’s TMEZ-KC multi-head is rated at 1,100 rpm. ZSK’s 1,400 rpm on tubular, from its January 2026 news issue, is the highest rpm figure among them, and it is a tubular-frame number. Melco’s SUMMIT is quoted in a different unit, a 16-needle machine at up to 1,500 stitches per minute. Its page calls it “one of the fastest” commercial machines available, then adds that embroidery performance is more than stitch speed. Verdict 68 confirms the 1,500 and the wording, and says what it is: Melco’s own sentence, not an independent ranking.

Sources: ZSK news issue 01/2026 · Barudan (Japan) · Happy HCR3 · Tajima TMEZ · Melco SUMMIT · Coats · Verdict 68

Tension is where the wording drifts most. Tajima says its i-TM system supplies the right amount of upper thread “in just 0.05 seconds” and that its presser-foot system reacts to fabric thickness in the same 0.05 seconds per stitch. The TMEZ-KC page claims an estimated 30% cut in processing time compared with conventional machines, with no comparison machine named, and says the time for trial sewing and adjustment “can be reduced”. Hirsch Solutions, a distributor, lists “eliminate test runs” among i-TM’s efficiency gains in its March 2025 release. Verdict 67 rates “eliminates test runs” partly true: the stronger words are the distributor’s, and the maker’s are weaker.

Thread cost has one useful number, from Coats: about 6.0 metres of needle thread and 2.3 metres of bobbin thread per 1,000 stitches. Coats calls it an average that stitch length, tension and fabric thickness will move. It is still the figure to put in a quote, rather than a vendor’s speed.

Sources: Tajima i-TM · Tajima TMEZ · Hirsch Solutions release (printing.org) · Coats · Verdict 67

Superlatives from the maker

Three vendor sentences in this issue’s checks have the same shape. Melco: one of the fastest. Coloreel: its thread-colouring unit “can print over 12 million colors”, roughly 80% of the Pantone palette, in the company’s FAQ. Xaar: its TF Technology “is unique to Xaar and is the best ink recirculation technology available”, with other makers’ methods unable to match its reliability and print quality. Each was quoted accurately. Verdict 66 confirms the Coloreel figure as the maker’s statement. Its reseller ZSK adds that certain tones fall outside the CMYK spectrum Coloreel uses, which fits an 80% figure rather than contradicting it. For Xaar, verdict 58 rates the claim partly true. Ricoh’s printhead page lists several ink-recirculating models, so Xaar is not alone, and neither page measures reliability or print quality against the other.

Sources: Coloreel FAQ · ZSK on Coloreel · Xaar TF Technology · Ricoh industrial inkjet · Verdict 66 · Verdict 58

Colour: a number without its formula

A Delta E 2000 of 1 is often given as the point below which two colours look the same. A Techkon page, published through Datacolor, says “generally” 1 is the threshold. A 2024 study in the Journal of Clinical and Experimental Dentistry measured it directly, with 20 dentists and 20 patients judging ceramic shade tabs under D65 light, and put the 50% perceptibility threshold at Delta E 2000 of 2.29 for dentists and 2.27 for patients. That is a dental study and not a print one, but it is a measurement, and it is over twice the rule of thumb. Verdict 69 rates the claim partly true: defensible as a rule of thumb, not as a measured threshold.

The instruments do not help a shop compare numbers. X-Rite’s eXact 2 lists four selectable Delta E formulas (1976, CMC, 2000 and 1994). The inter-instrument agreement it states, 0.25 on average and 0.45 at most, is given in delta E ab, the 1976 formula. Datacolor’s Spectro 700 gives agreement of 0.15 average and 0.30 maximum under a “CIELAB” heading and does not name a formula. G7 press control, in a data sheet hosted by IDEAlliance, reports weighted delta-L and delta-Ch for gray balance, with the formulae in IDEAlliance’s TR015: a different measurement from any of those four. G7 Master certification carries an $800 application fee per facility ($550 to renew) plus per-device fees, is valid for one year, and lists commercial, packaging, digital production and wide-format printers among its typical facilities. Garment decoration is not on that list. That is an absence from a list, not an exclusion.

Fabric adds its own error. Konica Minolta says a glossy object dyed a navy blue falls in a darker, more saturated region of colour space than a rough-surfaced object dyed the same blue, so for instruments of that geometry, surface texture moves the reading. Barbieri sells a Textile Edition of its spectrophotometer whose camera-based chart recognition is built to find charts distorted by textile stretch. The viewing standard for prints and proofs, ISO 3664:2009, sets 2,000 lux plus or minus 500, in the words of a light-booth maker that publishes it.

The platforms do not offer a figure to hold them to. Printful says it cannot guarantee 100% colour accuracy even with a calibrated monitor. Gelato says a 100% match is not always feasible. Neither gives a Delta E or any other tolerance.

Sources: Techkon (Datacolor) · J Clin Exp Dent 2024 · X-Rite eXact 2 · Datacolor Spectro 700 · IDEAlliance G7 data sheet · G7 Master · Konica Minolta · Barbieri Textile Edition · VeriVide on ISO 3664 · Printful colour disclaimer · Gelato support · Verdict 69

Three labels that say more than the page

AATCC’s test method TM117 is titled “Colorfastness to Heat: Dry (Excluding Pressing)”, and its scope covers textiles of all kinds and forms. A test-equipment vendor’s listing calls it a colour-fastness test for sublimation; AATCC’s page does not use the word. Verdict 64 rates the claim partly true. Little CMS’s new calc ICC app, reported on 7 September, is a colour conversion tool that imports existing profiles for conversion, soft proofing and gamut checks. It does not build or calibrate profiles, and verdict 92 rates the profiling-and-calibration description false. Which profile each platform wants embedded in an uploaded file is a separate question, answered per process in the sublimation guide and the wall decor guide. Enfocus’s Griffin page says its true-shape nesting can “reduce waste by up to 50%”. The claim as it circulated said material costs. Verdict 88 rates it partly true: waste and cost are different quantities, and the page gives no percentage for cost.

Sources: AATCC TM117 · Test-equipment vendor listing · Digital Printer (calc ICC) · Enfocus Griffin · Verdict 64 · Verdict 92 · Verdict 88

What this section cannot do

No figure above comes from a test by someone with nothing to sell, apart from the dental study. The embroidery numbers are makers’ and resellers’ pages, mostly undated, and the colour numbers are instrument makers’ specification sheets. The useful output is the comparison itself: which pages say the same thing in different units, and which headlines say more than the page under them.

✓ Confirmed “Melco SUMMIT runs at up to 1,500 stitches per minute, “one of the fastest stitching commercial embroidery machines available””

◐ Partly true “Tajima’s i-TM automated tensioning system eliminates test runs on the embroidery machine”

✓ Confirmed “Coloreel’s thread-coloring unit can produce over 12 million colors, roughly 80% of the Pantone palette”

◐ Partly true “Xaar’s TF Technology is unique to Xaar and is the best ink recirculation technology available; other manufacturers’ recirculation cannot deliver the same performance.”

◐ Partly true “A Delta E 2000 value of 1 is the threshold below which a colour difference is imperceptible.”

◐ Partly true “AATCC 117 is a standardized test method used to evaluate colour fastness to dry heat or sublimation in synthetic textiles.”

✗ False “Little CMS unveils mobile ICC profiling app for digital color calibration workflows”

◐ Partly true “Enfocus says Griffin’s true-shape nesting and gang-run printing cut material costs by up to 50%.”