diff --git a/docs/EXTERNAL_TEST_DATA.md b/docs/EXTERNAL_TEST_DATA.md index 98bd50b27..ca0a0308c 100644 --- a/docs/EXTERNAL_TEST_DATA.md +++ b/docs/EXTERNAL_TEST_DATA.md @@ -23,8 +23,8 @@ the table below; the repositories themselves are cited in reprocessing results; no quantity measured by Jungfraujoch appears on this page. - **Detector is read out of the image files themselves** - the NXmx `/entry/instrument/detector/description`, the miniCBF `# Detector:` header, the marCCD - instrument header or the SMV key block - because the detector named in a PDB entry is often - only approximate. Where the two differ, the difference is listed below the table. + instrument header or the SMV key block - which is authoritative where the PDB entry names a + different detector. - Anything that could not be established from one of those sources is left blank. ## Datasets @@ -306,11 +306,7 @@ The authors of the second letter also published their own reciprocal-space recon 6RLR data as a separate Zenodo record, [10.5281/zenodo.6961763](https://doi.org/10.5281/zenodo.6961763). -## Detector: image file vs PDB entry - -For 163 of the 164 PDB-coded rows both the image file and the PDB entry name a detector. (For -8XTG neither can be compared - the header reads `PILATUS XXX, S/N XX-XXX`.) The table above uses -the file value in every case, because the entry's label is often approximate. +## Detector column for marCCD and SMV files marCCD and SMV files name the detector differently - or not at all. A marCCD file names no model: its instrument header states the image dimensions and the pixel size, from which the plate size @@ -321,50 +317,6 @@ Detector column carries what the file itself establishes: the plate size (`marCC plate`), the comment's name where one is present (`Rayonix MX-300`), or the serial (`SMV, S/N 930`). -**Eighteen of the 163 genuinely conflict** - the two sources name detectors that cannot both be -right: - -| PDB | PDB entry says | Image file says | Conflict | -|---|---|---|---| -| 5MLN | MARMOSAIC 225 mm CCD | PILATUS3 2M, S/N 24-0118, ESRF ID23 | model / size | -| 6FWC | DECTRIS EIGER X 4M | PILATUS 2MF, S/N 24-0109-F | model / size | -| 6JGJ | DECTRIS PILATUS3 6M | PILATUS3 300K, S/N 3-0226 | model / size | -| 6P8P | DECTRIS PILATUS3 S 6M | PILATUS 6M-F, S/N 60-0112-F | generation | -| 7ATG | DECTRIS PILATUS3 S 6M | PILATUS 6M-F, S/N 60-0117-F | generation | -| 7N2S | DECTRIS EIGER X 16M | PILATUS 6M, S/N 60-0101 | model / size | -| 8R5R | DECTRIS PILATUS 6M | Dectris EIGER2 CdTe 16M | model / size | -| 8RUD | DECTRIS PILATUS 6M | Dectris Eiger 16M, E-32-0107 | model / size | -| 9H0Q | DECTRIS EIGER X 16M | Dectris EIGER1 Si 9M, E-18-0102 | model / size | -| 9HNC | DECTRIS EIGER X 16M | PILATUS 6M-F, S/N 60-0117-F | model / size | -| 9KHR | MAR CCD 165 mm | marCCD, S/N 35, 3072 x 3072 pixels of 73.242 um | plate size | -| 9O0H | DECTRIS EIGER X 16M | Dectris EIGER2 Si 16M, S/N D021324 | generation | -| 9Q41 | DECTRIS PILATUS 6M | Dectris EIGER2 Si 16M | model / size | -| 9SL0 | DECTRIS PILATUS4 X 4M | Dectris EIGER2 Si 9M | model / size | -| 9UPT | RAYONIX MX300-HS | SMV, S/N 930, 3072 x 3072 pixels of 102.588 um | plate size | -| 9VX7 | DECTRIS EIGER X 9M | PILATUS3 6M, S/N 60-0133 | model / size | -| 9Z44 | DECTRIS EIGER X 9M | Dectris EIGER2 Si 9M, S/N E-18-0131 | generation | -| 9Z72 | DECTRIS EIGER X 16M | Dectris EIGER2 Si 16M, S/N E-32-0127 | generation | - -For 9SL0 the file is decisive and the entry is wrong: 3108 x 3262 pixels of 75 um on 450 um -silicon, written by EIGER2 firmware `release-2022.1.2`, is an EIGER2 9M and not a PILATUS4 4M. - -For 9KHR and 9UPT the file names no model, so the comparison is on geometry, and it is decisive -both times: 3072 x 3072 pixels of 73.242 um is a 225 mm plate, not the entry's 165 mm one, and -3072 x 3072 pixels of 102.588 um is a 315 mm plate, which no 300 mm detector has. The 6FWC -frames were written by the same PILATUS 2M-F, S/N 24-0109-F, that wrote the 5NW5 and 6TOC frames -at SLS X06DA, although the entry deposits an EIGER 4M at ESRF MASSIF-3. - -The other 145 agree with their entry, up to how much each side states, which is not a conflict. -An NXmx `description` often gives the model and size but no generation (`Dectris Eiger 16M`) -where the entry names one (`DECTRIS EIGER X 16M`); a miniCBF header sometimes names a generation -(`PILATUS3 6M`) that the entry leaves off (`DECTRIS PILATUS 6M`), as for 6YQF, 7PH1, 7QIS, 7YZX, -8XTE and 9YZK. Other rows differ as `DECTRIS EIGER X 9M` against the file's `Dectris EIGER1 Si -9M`, `MARMOSAIC 300 mm CCD` against a comment reading `Rayonix MX-300` (the same detector under -its later brand), or by a serial number or an `-F` suffix the entry leaves off. The ADSC entries -(`ADSC QUANTUM 315`, `ADSC QUANTUM 315r`) have SMV files of 3072 x 3072 pixels of 0.1026 mm -(0.102592 mm for 6NEN), a 315 mm detector; the marCCD files are 225 mm plates where the entry -names a 225 mm detector and 300 mm plates where it names a 300 mm one. - ## Deposited models and structure factors 164 of the 171 datasets have a released PDB entry, and RCSB