This is an older version. The current version is 3.5.0. Argyll CMS 3.5.0 · Released:
Installer files 3.4.0
| Architecture | Type | Scope | File size | Download | SHA256 |
|---|---|---|---|---|---|
| 64-bit | ZIP → Portable | — | 38.0 MB | argyllcms.com | VirusTotal |
| 32-bit | ZIP → Portable | — | 40.6 MB | argyllcms.com | VirusTotal |
Link and SHA256 come from the Windows Package Manager community manifest. Compare the hash of the downloaded file before running it.
Release notes
- Added support for Spyder and SpyderPRO (2024)
- Renamed proposed CIE 2012 2 and 10 degree to standard CIE 2015 2 and 10 degree observers, as per CIE 170-2:2015.
- Added -Y c: option to colprof and invprofcheck to allow setting/overriding alibration curves for setting final ink limits.
- Added -Y c: option to xicclu to allow setting/overriding calibration curves for setting final ink limits, and added -T flag to show ink limits.
- Fixed Spyder X and X2 to not make Device Disconect noise (if such sounds are enabled) on MSWin.
- Fixed bug in colprof -nI failing for CMYK profiles.
- Enhanced strip reader patch recognition to reject strips that don't start and end on the media, as well as adding some speed compensation to the patch recognition for non-zerbra ruler measurements.
- Fixed bug in i1Pro3 strip reading where it would return bad values if the zerbra stripe wasn't used.
- Changed chartread so that it will issue warnings of possible bad row or patch reading even when .ti2 reference measurement is not accurate (i.e. no preconditioning profile used in targen.)
- Increased emphasis of making sure that there is contrast between patches at the end of rows in printtarg patch ordering for strip instruments.
- Added -C parameter to targen to allow overriding any calibration curves found in the ICC profile, used to estimage the total ink limit from that in the ICC profile. Also better enforces final raw computes ink limits.
- Modified ccxxmake so that it checks that it has actually found a white seeming patch to use as the L*a*b* white reference, and the patch to de-weight. If there is no white patch (i.e. just RGB patches) then it will use D65 as the L*a*b* conversion reference. ccxxmake will fail if there are less than 3 patches.
- Fixed ICC profile writing so that it clips rather than failing when writing a ColorantTable PCS value that is out of range. A warning to stderr will be issued.
- Changed ColorMunki driver so that it is more forgiving about unexpected version string lengths.
- Fixed dispcal and dispread so that they won't error out if there is no instrument but the -M parameter is provided.
