The continuous gradient band

A grid of 25 patches tells you how your process responds to 25 distinct grey levels. It's precise, it's measurable, it's what the app uses to build your curve. But between two patches there's a 4% jump in intensity — and your eye sees 25 separate cells that it compares one by one, not a transition. If your process has a subtle blocking plateau or a zone that flips abruptly between two patches, you can miss it.

That's exactly what the continuous gradient band present on every Calibration Flow test chart is for: a vertical column to the right of the grid, running from pure white to full black without any visible step. The calibration measurement is still made on the 25 discrete patches; the continuous band, for its part, shows you what no discrete-point measurement can show you — how your process behaves between those points.

#What you see on the paper

When you print the chart, the grid of 25 patches comes out as expected — 25 distinct cells, from paper white to the deepest black your process can reach. Next to it, the gradient column comes out as a transition. What that column reveals depends entirely on your process.

On an uncalibrated cyanotype, you'll typically see three zones: an almost-white part at the top (the chemistry doesn't engage at low exposures), a steep transition in the middle (the chemistry responds abruptly between two thresholds), an almost-black part at the bottom (saturation). These three zones exist in the grid too — patches 0 to 6 white, patches 7 to 14 flipping fast, patches 15 to 24 black — but looking at the grid you see cells; on the gradient, you see the mechanics of the process at a glance.

On a calibrated, clean process, the transition is almost progressive from top to bottom. No jump, no plateau. It's the sign that no tonal zone is blocking abruptly.

On an unstable process (drifting chemistry, poorly sensitized paper), you can see bands or rings in the gradient — zones where density inverts locally, plateaus where the chemistry blocks. These defects exist in the grid too but are hard to isolate there; on the gradient, they jump out.

#Why it's valuable

Visual diagnosis in 30 seconds. Before even importing the chart into the app, you glance at the dry gradient. If the transition is even, you know your process is in a calibratable state and you can scan to measure. If you see a jump or a ring, you know there's an upstream problem — chemistry, paper, exposure — to fix before calibrating. Thirty seconds of observation spares you a failed calibration and a pointless scan.

Detecting Mach bands. On some processes (dense carbon, Aquaprint Sanguine), the tonal transition is so steep at a precise spot that you physically see lighter and darker bands around the transition — that's the "Mach bands" phenomenon. The continuous gradient makes them visible before the final image. You know you'll need to soften the slope of your curve at that spot so as not to find these bands again in your prints.

Verification after adjustment. When you print a second chart after changing something (changing sensitizer, changing paper, tweaking chemistry), comparing the two gradient bands side by side tells you immediately whether your change affected the overall response of the process. Faster than mentally comparing 25 patches two by two.

#When you can ignore the gradient

You're running a quick production calibration. If you're calibrating a process you've mastered and whose behavior you already know, the gradient adds no useful information for this session. You scan, the app measures the 25 patches, you export your curve. The continuous band is there, but you don't look at it.

You're a complete beginner. The gradient is a reading tool that requires knowing what you're looking at. At the very start, you won't be able to tell a normal transition from a pathological one — both can look "smooth" to the untrained eye. Focus on the grid for the first calibrations, then come back to the gradient when you start reading the correction curve effortlessly.

Your printer renders continuous mid-tones poorly. On some basic inkjet transparencies, the gradient band can come out with stray printer-driver banding (visible steps that aren't in the chart but in the rendering). These bands can be mistaken for process defects. In that case, prefer the A4 300 dpi dithering version — see A4 300 dpi test chart with dithering — which sidesteps the problem by converting the grid into a structured bitmap.

#Key points

#The test

Print a normal chart on your usual process. Once the chart is dry, look at the gradient band on the right before even scanning it. Follow the transition from top to bottom with your finger: density should increase with no perceptible jump. If you see a spot where density jumps abruptly between two very close zones, or a spot where density stays flat over a visible length while it was progressing before and after, your process has a non-linearity the calibration will need to correct. Make a mental note of where the problem zone sits on the band — you'll find the same zone in the correction curve the app generates.