Each work is a pair. The left panel is a photograph as it was made. The right panel is the same photograph carrying one intervention derived from its own measurements. Side by side, the difference often lives at the threshold of vision, and it is stated exactly, down to fractions of a single tonal step. The numbers are the work, and anyone may measure them again. You are offered almost nothing to see, and everything to check. · · · Not one pixel has been altered. The frame has moved thirteen pixels so its center falls between the two flowers, each now the other's antipode. The midpoint of the two daffodil heads sat 13.30 pixels from the center of the frame, 1.50% of its width. Had the two flowers fallen independently and uniformly in a square this size, the chance of their midpoint landing that close to the center is about 1 in 362. The world very nearly arranged them. The measurement is stable. Sweeping the detection threshold across its whole reasonable range moves the midpoint by 0.3 pixels. So the frame moved instead of the flowers. The largest square on the original pixel grid whose center lands on that midpoint is 859 × 859, taken at (27, 11), keeping 94.0% of the area at a new offset of 0.44 pixels. Because the crop is integer-aligned, nothing was resampled, and the maximum difference between the altered file and the corresponding region of the original is 0.0 levels. The two panels therefore show different extents of the same pavement, and that difference is plainly visible at the edges. It is meant to be. What is attested is stricter and stranger, that on every inch the two panels share, the difference is exactly zero. Every other work in this series changed values. This one changes none. Every pixel it keeps is exactly the corresponding value in the source file. The only thing that moved is where it stops. Presentation. The two panels are different sizes, 2658 and 2577 pixels at print scale, and must remain so. Same magnification at different extent is the entire work. They hang aligned on the flowers' midpoint, so the same square inch of pavement sits at the same height in both, and the apertures visibly fail to line up by the amount of the original error. The mount is RGB (63, 60, 69), the joint's own color, sampled from the dark seam running down both panels. The picture is two slabs with a dark line between them, and the presentation is two panels with more joint around them. There is no rule. The work is about an edge, and drawing a second one around it would be the one unforgivable move. To check. On everything the two panels share, the difference is exactly 0. Download this work's original file, 09_antipodal.png, not a screenshot or a marketplace preview, attach it to an AI that can run code, and paste this: Run this exact line on the attached file, changing only the file path if needed. Show me its exact output, and say whether that output is exactly: difference 0 python3 -c "import numpy, PIL.Image as I; c=numpy.asarray(I.open('09_antipodal.png').convert('RGB')).astype(int); a=c[800:3458,800:3458][::3,::3]; b=c[833:3410,3858:6435][::3,::3]; print('difference', abs(b-a[11:870,27:886]).max())"