Gulf Coast Fungi

Updated 2026-10-27 · ToupLite, OMAX A3550U3 camera (U3CMOS05100K), 2560 × 1922 px

Calibration in use

Method

Pixel measurement of the stage micrometer (0.01 mm = 10 µm divisions). Line centres are found in horizontal bands; a linear fit gives pixels per division; px/µm = pixels per division / 10.

Log (newest first)

• 2026-10-27done WRJ_0218 write-up, record file and iNat note updated with the cystidia and stem-base results (microscopy addendum); spore CSV citation now points to the 2026-10-26 archive; independent review of the note run and its corrections applied. The two WRJ_0218 open items are closed.
• 2026-10-27done color_correct.py fixed after the independent check: saves at maximum JPEG quality (100, no colour subsampling); --background now refuses a field with no clear neutral background (selected pixels below 150 or a gain outside 0.80-1.25; tested on a brown field: refused); outputs always .jpg. WRJ_0218 copies regenerated (earlier quality-95 copies moved to _superseded_q95). 218_40X_base_teased refused. Note: redo.jpg has no burned-in scale bar; background mode also removes a stained mountant's tint, so stain reactions are judged on originals.
• 2026-10-27done Colour correction after capture (Will chose option 2). Script tools/microscope_calibration/color_correct.py: --background uses each image's clear background (brightest 5%, unclipped pixels) to make it neutral grey; --reference uses a blank-field frame. Originals untouched; corrected copies only. A 40X blank reference (white_reference_40X_2026-10-27.jpg, R/G 0.87) over-corrected the 100X images (their backgrounds are R/G 0.89-0.93), so the cast differs by objective or brightness; those outputs were moved to _superseded_40X_reference_on_100X. Re-run with --background on the 11 WRJ_0218 100X images from 2026-10-27: backgrounds now neutral within about 1%. 218_40X_base_teased.jpg not corrected (62.8% of pixels already clipped). Rule for references: one blank frame per objective, same light and settings. The ToupLite save landed in WRJ_0082's folder by mistake; moved to Microscope_calibration/white_reference/.
• 2026-10-27done White balance, second try (Will had reset to Defaults: 6503/1000, Exposure Target 120). Blank field is bright but not clipped; red is about 10% below green and blue (R/G 0.90). One-push White Balance changed the numbers to 7887/1195 but the picture measured the same (R/G 0.895). ToupLite's histogram shows the same channel split (one channel around 100-140, the other two around 150-175 of 255). So brightness is not the cause, and ToupLite's white balance is not correcting it. Options given to Will: check for a coloured filter in the light path; or correct colour after capture from a blank-field frame per session (candidate for the Microscope project's automation).
• 2026-10-27done ToupLite white balance, blank slide at 40X. The one-push 'White Balance' button gave a cyan cast every time (Will once at 4X, Claude three times at different brightness; it set 7474/1183, 6865/1322, 7893/1242). The manual sliders do work: Temperature higher = warmer (4966 turned the field blue), Tint lower = pinker (451 and 624 magenta). Closest neutral measured: Temperature about 5944-6449 with Tint 836, red still about 6% low. Some slider clicks from Claude changed the number but not the picture, so final tuning is left to Will by eye. Exposure Target lowered from 120 to 49 (Auto Exposure on) so the white field no longer clips; Defaults restores 120. Illumination is uneven (darker lower left). Settings left: Temperature 6449, Tint 836, Exposure Target 49.
• 2026-10-27note Correction from Will: no stain was used on 218_40X_base_teased.jpg. The dried stem base turns any mounting liquid dark brown where concentrated, lightening to reddish and then yellow as it dilutes; that pigment, plus the clipped capture, is why the frame is yellow. Recorded as an observation of this collection. The write-up quotes S52 for P. rubricitrinus: 'deep red pigment present at base of stipital context'; whether it leaches is not stated in the quoted text, so not used as evidence. Next: set ToupLite white balance with Will.
• 2026-10-27done 218_40X_base_teased.jpg: the dried stem base is hard (Will: 'consistency of pine bark') and does not tease. Image inconclusive: colour clipped (red channel saturated in most of the frame, blue near zero), so it reads solid yellow; faint wavy interwoven threads, none measurable; a few thin-walled round outlines about 11-18 µm, two dark bodies about 16 and 23 µm, and small dark spore-sized bodies; none identifiable. Scale bar matches 40X. Teasing the base is not needed for the sclerotia question (sclerotia are visible lumps on the outside). Independent check run; corrections applied.
• 2026-10-27done Stem-base photos 218_tomlov_stem and _2 (dried, no scale): no orange sclerotia visible. Surface maroon with ochre-yellow patches showing through on the lower stem (fits 'red punctae over a yellow ground colour' and the fresh photo PA120028). Pale buff cylinder in stem.jpg: same colour and texture as the grass blade beside it, most likely a rolled grass leaf. Whitish strands in stem_2: flat, glossy, fraying ribbons, most likely plant fibre. Two small yellow-tan granular patches in stem_2 worth a look under the scope. Meaning: P. sclerotiorum is 'sometimes with orange sclerotia', so none seen is neutral. Independent check run; corrections applied.
• 2026-10-27done Redo images (redo, redo_2 to _4; no scale bar). Result recorded: cystidia NOT OBSERVED in any WRJ_0218 mount (KOH 3%, KOH 10%, Congo red; 8 images 2026-10-26, 7 + 4 on 2026-10-27). Will (at the scope): every sampling method gives the same result and there is no transparency issue. The one in-focus hymenial cell (redo; same as red_glycerin_4) is club-shaped, widening upward from about 7-8 to 12+ µm, about 30 µm long, with two spores at its tip: basidium-like, not the tapering cheilocystidium shape. Limit: no intact pore edge is shown in the photos, so the write-up's Q3 test ('No cystidia in a good pore-mouth mount') is not met by photographs alone; the result is not used to decide the species. Independent check run; corrections applied.
• 2026-10-27done Reviewed 7 new 100X images (KOH 10%; 2 'blue', 5 'red_glycerin'). Scale bars correct (239 px). No cystidium confirmed, and no hymenium (row of basidia) is visible in any frame, so these frames do not test cystidia presence. Candidates to reshoot: (1) red_glycerin bottom edge, a shape projecting into clear mountant tapering from about 11 to under 4 µm, out of focus, tip off the frame; (2) red_glycerin_4, a free septate hypha with a terminal cell about 29 µm long, neck about 7 µm, two spores lying over its upper part: basidium, cystidium or hyphal end cannot be told. Blue pair: no blue in the tissue. Independent check run; its corrections applied.
• 2026-10-27done Reviewed the two Tomlov photos of the WRJ_0218 pore-surface slice (lit from above, dry): a slab of tube mouths about one to two pore widths thick, the kind of slice used for the earlier slides. Advice given (sources: QMS stain procedures, BMS reagents, boletales.com, ZEISS Köhler page): soften the dry block in Clémençon's reagent (80 IMS : 20 0.88 ammonia : 1 glycerol) and cut much thinner; make a squash of torn tube walls as a second mount; stain with Congo red and draw KOH or ammonia under the coverslip until the background is clear; phloxine (1.5 g/100 ml water) for cell contents if obtained; close the condenser iris for contrast; avoid lactic acid for bolete spores. Independent check run before the reply; its corrections were applied.
• 2026-10-27done In-use table cleaned up at Will's request: exact ToupLite px/m values (40X 9,250,142.86, was shown rounded), a px per 10 µm column (computed from px/µm) on this page and the public-site builder, short notes. No calibration value changed.
• 2026-10-26done MICROSCOPE_REFERENCE_2026-10-26.md reduced to equipment facts and a pointer here, so the values exist in one place.
• 2026-10-26decision Before a reply reports microscope work to Will, a second instance that has not seen the work checks it against the files, and the reply says what the check found.
• 2026-10-26done Second-instance check of the day's work (Will's instruction, 20:18). Findings applied: the entered 10X value is a centre-of-field value (full-field 2.2966, 0.16% lower; right third 0.7% lower); the two 10X slides agree within 0.02% by the same method, so the earlier '0.2% lower' was a line-selection artefact; 40X and 100X values depend on method (±0.3%) and colour channel (±1-2%), so their stated precision was too high; two 40X bands on the second copy were defective. No stored value changes.
• 2026-10-26done The old Downloads copy of MICROSCOPE_REFERENCE_2026-10-26.md (wrong 10X value) is gone: Will deleted it about 19:17. Removed from this list.
• 2026-10-26done Reviewed the existing pore-mouth mounts (KOH 3%, no stain: 10X_01, 40X_01 to 03) and the 100X Congo/glycerin images (1 to 4). No cystidia identified in any of the eight; spores and debris dominate. Correction: the 100X set is KOH with Congo Red and glycerin, so not all WRJ_0218 micro images are unstained KOH.
• 2026-10-26done Note for future measurements: in the WRJ_0218 40X image the fine lines sit in the lower frame (rows about 1040-1900). The upper rows show only the long 50 µm lines, so measure the lower rows.
• 2026-10-26done 4X confirmed on two slides: slide 1 (same file in both folders) 0.90976 px/µm, new slide 2 in the WRJ_0218 folder 0.90982 px/µm. Stored 0.91001 is 0.02-0.03% high. No change.
• 2026-10-26decision Record copy for 40X: the WRJ_0218 file. Its band-to-band spread is 3x smaller (0.006 vs 0.021) and its residuals are about 1 px, not 2 px. Will delegated this choice. The other copy stays as the earlier source.
• 2026-10-26done 40X second copy (Microscope_calibration) re-run with 10 bands: 9.21788 px/µm, replacing the 4-band 9.2197. Within 0.35% of stored.
• 2026-10-26done 40X re-measured with 10 bands on the WRJ_0218 copy: 9.22709 px/µm (spread 0.006). 0.25% below stored 9.25014. No change.
• 2026-10-26decision Standing rule: all microscope work is recorded on this Microscope page and kept current here. No separate index.
• 2026-10-26done calibration_log.csv repaired: the 10X status field had an unquoted comma. Every row now has 11 columns.
• 2026-10-26done Archived as not certain: the superseded reference file, placeholder and automated CSVs, unverified August AmScope/ToupLite copies, generic ImageJ steps. Originals in Will_original are untouched.
• 2026-10-26done 100X oil micrometer measured 23.97913 px/µm (8 bands, blurry lines). Stored 23.90025 kept.
• 2026-10-26done 40X slide measured 9.2197 px/µm (4 bands, looser fit). Stored 9.25014 kept for now.
• 2026-10-26done Second 10X slide measured 2.29565 px/µm (10 bands), 0.2% below the entered value. Entered value kept.
• 2026-10-26done 10X corrected: the stored 11.49164 px/µm was 5x too large. Measured 2.30033 px/µm from the old 10X slide; Will entered 2,300,326 px/m in ToupLite.

All calibration measurements (14)

Text and data on this site: CC0 (no rights reserved), the same as Will's iNaturalist observations. · Built 2026-11-01 18:25.