Device-Link Profiles - collink #99
Replies: 24 comments 42 replies
-
|
Thanks for your posting. To be completely honest with you: I have no idea what thise Device-Link profiles are and what is required to create such profiles. I am relatively new to this whole topic and was only able to create my first icc profile with Argyll with the help of a script Knut created. Since I had some free time for experimenting I started this whole project here with the help of AI to see how far I would get. |
Beta Was this translation helpful? Give feedback.
-
|
(Replying as Claude, on Itsab1989's behalf — he asked me to research this and draft a concrete proposal for you to react to.) Thanks for the suggestion. Short answer: yes, this is doable, and as a fairly self-contained addition. Argyll's Below is a concrete plan so you have something to push back on. Treat every default as a proposal, not a decision — the open questions at the bottom are exactly where your input changes things. Two honest caveats first
Proposed feature: a "Create device-link profile" ToolWhere it lives: a new entry in the Tools menu (next to "Soft-proof" and "Inspect a profile"). The normal colprof profile stays the primary output; the device-link is an optional extra. Look & feel: it follows ChromIQ's standard Tools-dialog style — a masthead with an uppercase eyebrow + headline + short description, and a single accent colour carried through the window (button, focus rings, ⓘ tints). Proposed accent: cyan ( Every option gets a help icon. Like the rest of ChromIQ, each control carries a ⓘ icon next to it that opens a friendly, plain-language explanation window — what the option does, when you'd change it, and a sensible recommendation — written for someone who isn't a colour-management expert. Device-link gamut-mapping terms (perceptual vs. colorimetric intent, CIECAM02 viewing conditions, abstract profiles) badly need this, so it's a first-class part of the design, not an afterthought. The dialog (proposed):
Output destination:
Guards: ICC-v2 only (Argyll writes v2; same guard the soft-proof tool already uses) — I'd bundle v2 copies of the working spaces to avoid surprises. What about ICC v4 source profiles? I tested this directly: Argyll's engine flatly can't read v4 —
The converted v2 file is a throwaway intermediate — the link bakes in the source's colorimetry, so the v2 copy has no further use. It's written to a temp location and deleted once the link is built (same disposable-tempfile pattern the soft-proof tool already uses), so nothing clutters the working folder. Why convert rather than fork/patch
Net: v4 sources are supported in practice (bundled v2 spaces + exact auto-transcode of matrix profiles, lcms2 for the rare LUT case), while the link Argyll builds stays v2 under the hood. How it maps onto the code (low-risk)
Phasing:
Update — incorporating backtomarfa's repliesSome of this is now settled, and one genuinely valuable new lever surfaced:
Open questions (this is where your feedback steers it)
(Things like which collink mode flag, default black handling, clut resolution, file naming and ICC description strings are mine to decide — I'll just propose sensible values and you correct anything that's off.) No code yet — this is for you to shape. Honestly the fastest path with someone who knows this area: answer what's easy, and I'll build a rough prototype you can react to — a concrete version will surface the real requirements faster than more questions. — Claude (via Itsab1989) |
Beta Was this translation helpful? Give feedback.
-
|
outside the prepress world almost nobody even heard about this profile type and i´m not 100% sure about real visible benefits, theoratically going from the source space without a connection space (LAB) should have an advantage, I asked Claude before posting and it agreed: Warum das für deinen Anwendungsfall relevant ist Genau dein Wunsch nach präziserer, weniger grober Kontrolle über die Gamut-Kompression lässt sich mit collink tatsächlich umsetzen, weil du hier nicht auf die drei festen Photoshop-Intents beschränkt bist, sondern die Kompressionscharakteristik granular konfigurieren kannst: weicher oder härterer Knee, Priorisierung von Hue-Erhaltung vs. Sättigungserhaltung, separates Quellgamut aus deinem tatsächlichen Bildmaterial statt aus dem generischen Arbeitsfarbraum. Das Ergebnis ist ein einziges, in sich geschlossenes Device-Link-Profil, das du dann z. B. in Photoshop über "Convert to Profile" mit Intent "absolute colorimetric" (wichtig, da Device-Links keine weitere Intent-Wahl mehr erlauben) anwendest, oder direkt im RIP deines Druckertreibers einbindest. Why this is relevant to your use case Exactly your desire for more precise, less rough control over the gamut compression can actually be implemented with collink, because you are not limited to the three fixed Photoshop intentions here, but can configure the compression characteristic granularly: softer or harder knee, prioritization of Hue preservation vs. Saturation preservation, separate source gamut from your actual image material instead of from the generic working color space. The result is a single, self-contained device link profile, which you can then e.g. B. in Photoshop via "Convert to Profile" with Intent "absolute colorimetric" (important because device links no longer allow further intent choice), or integrate directly in the RIP of your printer driver. Wo Device-Links im Inkjet-Bereich trotzdem sinnvoll sind Where device links in the inkjet area still make sense They are worthwhile as soon as you have a recurring, stable combination, for example if you consistently use the same printer-inkj-paper setup across many pictures for a photo book project or an exhibition series. Then the same argument applies as in the offset: Once carefully calibrated (with the finely adjusted collink parameters we talked about), you save yourself the live calculation with every single image conversion and have guaranteed consistency over the entire series. Even for very critical reproduction tasks, such as art reproductions, where exact, reproducible color fidelity across many prints is crucial, a firmly defined, tested Device Link profile can be superior to the flexible standard workflow, because it excludes variability due to different CMM interpretations or software versions. |
Beta Was this translation helpful? Give feedback.
-
|
Google sieht es auch Positiv... Why Use an RGB-to-RGB DeviceLink? Bypasses Driver Inaccuracies: Standard printing uses the operating system's Color Management Module (CMM) to convert Adobe RGB → Lab → Printer Profile. An RGB-to-RGB DeviceLink forces a mathematical shortcut directly from Adobe RGB → Printer Profile, ignoring the Lab middleman entirely.Preserves Pure Saturated Primaries: Standard Lab conversions often taint highly saturated reds, greens, or blues with speckles of unintended colors. A DeviceLink maps those extreme spectrum points smoothly to the printer's maximum output without unexpected color shifts.Prevents Posterization: Eliminating the intermediate Lab translation minimizes decimal rounding errors, ensuring smooth tonal gradations in photographic skies, vignettes, and fine art shadows. Ich habe bisher auch nicht an diesen Profiltyp gedacht aber ich beschäftige mich mit der Optimierung für mattes Papier wo Gamut Mapping eine wichtige Rolle spielt und sowohl perzeptiv als auch rel. Farbmetrisch nicht immer optimal sind. DLP sind da vielleicht die Lösung , in einer APP wie deiner ist das für Qulitäts-Besessene vielleicht zugänglicher als mit command line oder spezial Software. |
Beta Was this translation helpful? Give feedback.
-
|
Ich höre das hier gerade zum ersten Mal, aber wenn es klappt, wäre es natürlich toll. Hast du den Plan von Claude gelesen? Stimmst du grob zu? Falls dir zu den Fragen was einfällt, kommentier das gern. Bin gerade mit all meinen 3 Accounts am Limit und muss warten bis ich den ersten wieder nutzen kann. |
Beta Was this translation helpful? Give feedback.
-
|
So, da ist es: Kleiner Disclaimer: In der Beta sind neue / geänderte Optionen nur englisch, weil ich da immer wieder Dinge anpasse und sonst jedes Mal die Texte für 12 Sprachen anpassen müsste. Das verzögert den ganzen Prozess und verbraucht jedes Mal viele Token. Deshalb mach ich das in einem Rutsch, wenn ich denke, dass es fertig ist. Lass mich bitte wissen, ob es funktioniert wie du es dir vorgestellt hast! Bin echt gespannt! Viel Spaß! |
Beta Was this translation helpful? Give feedback.
-
|
Wow, danke schau mir das morgen an |
Beta Was this translation helpful? Give feedback.
-
|
Habe heute Früh ein paar Test Profile generiert , das Tool funktioniert wie es soll, hier meine Beobachtungen: Mein Testbild hat hoch gesättigte Magenta und Cyan Töne. Grosser Quellfarbraum Prophoto RGB (im Profile berücksichtigt) - perzeptiv - liefert sowohl via Adobe CMM, Standart Profil, als auch via DeviceLink unbrauchbare Ergebnisse was mich echt überrascht, in beider Fällen wird Magenta zu Orange. auch wenn man den Bild Gamut verwendet was das Ergebnis eigentlich verbessern sollte.
Mittlerer Quellfarbraum ECI RGB (im Profile berücksichtigt) - perzeptiv - hier sehe ich super subtile Verbesserung mit dem DLP.
Vorläufiges Fazit: Ich bin mir noch nicht sicher ob sich der zusätzliche Aufwand wirklich lohnt, Claude war vielleicht etwas zu optimistisch im Ausloben der Vorteile. Wenn du Lust hast könntest du weitere Parameter einfügen wie zB die Mapping Options -i und das mehr Bilder zur Generierung des Bild Gamuts geladen werden können sowie die Kontrolle über den -f Parameter von tiffgamut. "Image dependent gamut mapping using device links When images are stored in large gamut colorspaces (such as. Lab*, ProPhoto, scRGB etc.), then using the colorspace gamut as the source gamut for gamut mapping is generally a bad idea, as it leads to overly compressed and dull images. The correct approach is to use a source gamut that represents the gamut of the images themselves. This can be created using tiffgamut, and an example workflow is as follows: tiffgamut -f80 -pj -cmt ProPhoto.icm image.tif collink -v -qh -G image.gam -ip -cmt -dpp ProPhoto.icm RGBDestinationProfile.icm Source2Destination.icm cctiff Source2Destination.icm image.tif printfile.tif The printfile.tif is then send to the printer without color management, (i.e. in the same way the printer characterization test chart was printed), since it is in the printers native colorspace. You can adjust how conservatively the image gamut is preserved using the tiffgamut -f parameter. Omitting it or using a larger value (up to 100) preserves the color gradations of even the lesser used colors, at the cost of compressing the gamut more. You can create a gamut that covers a set of source images by providing more than one image file name to tiffgamut. This may be more efficient for a group of related images, and ensures that colors are transformed in exactly the same way for all of the images. An alternative generating a gamut for a specific set of images, is to use a general smaller gamut definition (i.e. the sRGB profile), or a gamut that represents the typical range of colors you wish to preserve. The arguments to collink should be appropriate for the output device type - see the collink examples in the above section." |
Beta Was this translation helpful? Give feedback.
-
|
Zum Workflow, man muss die erweiterten Einstellung in Photoshop für Profil Konvertieren aktivieren erst dann hat man Zugriff auf die DeviceLink Profile - ich habe für beide dann zur Sicherheit kein Farbmanagment via Profil zuweisen aktiviert und so an Canon P&L geschickt und dort ohne Farbmangment gedruckt. |
Beta Was this translation helpful? Give feedback.
-
|
Habe die Profile von Claude analysiert lassen, der Theorie nach sollte ja bei der Konvertierung des grossen ProPhoto Farbraums zum Drucker Gamut das DLP einen Vorteil haben doch das schein nicht der fall zu sein, es gibt zwar einen kleinen Fehler im DLP falscher Profil Platzhalter aber die Massive notwendige Kompression stösst an Grenzen. Generell schneiden die DeviceLinks nach Claud´s analyse schlechter ab: Auch der auf ohnePar zielende Device-Link zeigt dieselbe Magenta-Hue-Drehung (~8° wärmer als die farbmetrisch korrekte eciRGB-Richtung) und dasselbe Cyan-Fold-back (Chroma steigt bis 90%, fällt bei 100% wieder ab). |
Beta Was this translation helpful? Give feedback.
-
|
Danke, wieder was gelernt. Das hat aber praktische Konsequenzen, viel Fotografen haben gelernt den größten Farbraum zu verwenden und wenn notwendig den Druckerprofil die Kompression zu überlassen, das ist aber der Claude Analyse nach kein guter weg mit Argyll Profilen, ich habe leider keinen Profilemaker mehr um das zu vergleiche. Ich persönlich hatte immer Zweifel ob das wirklich sinnvoll ist da reale Bilder, keine synthetischen Farben oder selbstleuchtende Objekte relative kleine Gamuts haben die gut in gängige Farbräume wie Adobe RGB passen. Bin gespannt welche Ergebnisse die weiteren Intends liefern. |
Beta Was this translation helpful? Give feedback.
-
|
Nach weiteren Test denke ich die Schuld für das massiv nach orange gekippte Bild liegt bei der Canon Software oder der Kommunikation mit Photoshop. Ich kann es reproduzier und es ist sehr strange. Test Charts ohne Profil werden korrekt ausgegeben Druckt man mit perzeptiv aus C P&L erhält man mit einen mit basic Einstellungen erstellen Profil , also auch keinen Quell Gamut die höchsten Sättigungen im Druck, verwendet am einen Farbraum - ECI RGB - ist die Sättigung geringer und der Druck ist etwas heller, die Bilddatei hat ECI RGB eingebettet. Die Lehre daraus colprof komprimiert also offensichtlich den Gamut mehr als notwendig wenn ein Quellfarbraum verwendet wird, damit machen aber die DeviceLinks in der Folge keinen Sinn will man maximal mögliche Sättigung auf matten Papier. Ich denke nur Graeme W. Gill selbst könnte das wirklich aufklären und uns weiterhelfen. Praktisch gesehen funktioniert alles gut auch luminanzerhaltend perzeptiv. liefert schöne Ergebnisse nur im direkten Vergleich zeigen sich die Unterschiede. Danke für deine Arbeit ! |
Beta Was this translation helpful? Give feedback.
-
|
Danke, das Perzeptiv dann in Wahrheit rel. Farbmetrisch ist habe ich auch schon herausgefunden, inklusive wirksamer BPC...hinter jeder Tür eine neue Überraschung. |
Beta Was this translation helpful? Give feedback.
-
|
Du sagst es, unsere Diskussion zeigt wie komplex die Dinge werden dringt man nur etwas tiefer ein, ich beschäftige mich auch mit Kamerakalibrierung der selbe Wahnsinn. Google AI findet Perceptual Appearance irgendwie besser als Luminance Preserving Perceptual auf das Claude offensichtlich steht, praktisch verwendet hat das wohl sowieso niemand darum finden man auch keine wirklichen Anwendungsberichte und reale Erfahrungen sonder erdichtete Meinungen unserer AI Freunde. Deine Canon Treiber Geschichte ist beispielhaft, ich habe seit Monaten ein Airplay, Sonos Problem, je nach Fragestellung gibt es unterschiedliche Schuldige oder Lösungsansätze funktioniert hat nichts. |
Beta Was this translation helpful? Give feedback.
-
|
Hast du irgend ein Feedback zu den devicelinks bekommen ? Es wäre sicher nützlich wenn andere das auch testen und zu ihren üblichen Workflows vergleichen. Ich habe mich weiter mit der Sache beschäftigt, nachdem das doch etwas komplexer ist hier ein praktische Ablauf : 1 . Das Grund Druckerprofile braucht keine perzeptive Tabelle, ! Also keinen definierten Quell Farbraum oder sonstige perzeptive Tunings. (Das war eine Fehlerquelle bei mir )
Ich bin nun sicher das es es unmöglich ist egal aus welcher Anwendung auch immer mit der Einstellung ohne Farbmangment brauchbare Ergebnisse mit OSX Tahoe zu erreichen ! Der einzige weg ist Canon Farbe ohne Farbkerrektur. Auch drucken aus Canon P&L funktioniert nicht daher mein Problem mit dem orange shift. Als Test kannst du vor den Drucken im Treiber die Option öffnen in Vorschau wählen, bei gewählten kein Farbmangment erkennt man sofort die zusätzliche Farb- Transformation. Das heisst also im Klartext die Empfehlung von Calibrite ist der einzig verlässliche weg ! |
Beta Was this translation helpful? Give feedback.
-
|
Ich denke wir haben ein doppel Problem, A die Apple CMM selbst und B die Canon Papierprofile die zwangsweise angewendet werden auch wenn man das Farbmanagment ausschaltet. Warum man im Netz dazu nichts findet ist mir ein Rätsel . Die Aussage der Canon Leute das Canon P&L direkt mit den Drucker kommuniziert scheint auch so nicht zu stimmen. Ich beschäftige mich schon lange mit Apple Farbmanagment aber das ist unglaublich.. |
Beta Was this translation helpful? Give feedback.
-
|
Habe Adobe kontaktiert: Um rohe Farbmuster ohne jegliches Farbmanagement oder Profilkonvertierung unter macOS Tahoe zu drucken, müssen Sie den Standarddruckbefehl von Photoshop nicht mehr verwenden. Neuere Updates von macOS und der Adobe-Druckengine erzwingen einen Farbtransformationspfad (über ColorSync oder CoreGraphics), wenn Sie direkt aus Photoshop drucken, was die rohen Zielpatches vollständig verändert. Da Photoshop die Farbverwaltung in modernen macOS-Versionen stark einschränkt, müssen Sie externe Umgehungen verwenden, um vollständig unveränderte Pixel an Ihren Druckertreiber zu senden: Ein befreundeter Farbmanagment Fachmann kennt das Problem auch |
Beta Was this translation helpful? Give feedback.
-
|
Konnte mir ehrlich bis heute nicht vorstellen wie vertrottelt das alles geworden ist. |
Beta Was this translation helpful? Give feedback.
-
|
Aber wer ist jetzt wirklich Schuld? Apple? Canon? Alle? |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Yes Tahoe, nur Drucker managed Farben funktioniert. Darum gibt es unter Tahoe die Option Drucken ohne Farbmangment nicht mehr, das Utillity sollte das irgendwie ersetzende ich von Adobe gelernt habe aber auch das funktioniert nicht wie es soll. Da Canon und die anderen kundigen Leute entlassen und mit Call Centern ersetzt haben wird man im Kreis geschickt oder bekommt nette Mails. Die Profis nehmen RIPs und umgehen das alles, technisch ist das aber nicht notwendig. Wenn du einen Canon Grossformat Drucker kaufen willst zwingen sie dich fast zu einen RIP..... To print custom color profiling patches or calibration charts on a Canon imagePROGRAF PRO-1000 without experiencing a color shift, you must completely bypass color management.Because macOS Tahoe handles the print subsystem differently, the operating system or application often forces its own ICC color conversion onto your target. This results in a double-profiled chart, causing shifted hues and making the printed chart useless for calibration or profiling. Step 1: Use macOS ColorSync Utility (The Safest Route)Do not print your target directly from Photoshop or Lightroom, as Adobe applications frequently force a color conversion on Mac systems. Instead, use the built-in macOS utility:Open ColorSync Utility via Finder (/Applications/Utilities/ColorSync Utility).Click File > Open and select your calibration chart image (TIFF or JPEG).Click File > Print.Look at the primary print preview pane. In the Color: dropdown menu, change it from Premultiplied or As Image to Print as Color Target.Note: If "Print as Color Target" is grayed out, jump to Step 2 to configure the driver first, then return to this step. Step 2: Disable Driver Color CorrectionYou must explicitly tell the Canon PRO-1000 driver not to apply its own colors (like Canon Digital Photo Color or Vivid):In the Tahoe Print Dialog window, scroll down or expand the options menu to find Color Matching.Ensure ColorSync is unselected. Instead, choose Canon Color Matching (or Hand Off to Printer depending on your exact driver update).Navigate to the Quality & Media menu dropdown.Under the Rendering Intent or Color Mode setting, explicitly select No Color Correction.Ensure your Media Type matches exactly what the profiling/calibration software dictates (e.g., Photo Paper Pro Luster). |
Beta Was this translation helpful? Give feedback.
-
|
Ich fühle deine Wut durch das Display hindurch. Finde es auch eine Frechheit wie sie das kaputtgemacht haben. Aber es scheint zu wenige Nutzer zu geben, die das interessiert... |
Beta Was this translation helpful? Give feedback.
-
|
Haha... Wut, nein es ist eher eine Ernüchterung. Es gab immer Probleme doch es gab früher immer auch kompetente Ansprechpartner das wurden den Profit geopfert. Die meisten User drucken mit den gelieferten Profilen und sind glücklich weil das ehrlich gesagt auch gut funktioniert in dieser Welt sind wir die Aussenseiter.... |
Beta Was this translation helpful? Give feedback.
-
|
Hi, darf ich noch eine Verbesserung anregen ? Dein Program ist so wunderbar geworden das sollte sich auch im Design spiegeln. Ich würde mir ein schlichteres vor allem weniger buntes Interface wünschen. Die zT schrillen Farben machen es schwer messbar und stören auch den doch sehr professionellen Anspruch denn dein Programm erreicht hat Mir wäre monochrom am liebsten aber soweit muss es nicht gehen. Die Unterhaltung sollte vielleicht aus den selben Grund auch ausblendbar sein. |
Beta Was this translation helpful? Give feedback.











Uh oh!
There was an error while loading. Please reload this page.
-
no target, no matter how perfectly designed, helps to get real better conversion results because metamerie puts things into perspective in the end. In the offset industry, however, Device-Linke profiles are widely used to improve the conversion quality. surprisingly no one is interested in this profile type even for high demanding inkjet printing although here too there are clear benefits because they offer more control over the gamut mapping than we have with the usual workflow.
Beta Was this translation helpful? Give feedback.
All reactions