>I was rather thinking about normal ICC profiles and a CMM which only >honors the relcol AToB1 tables to obtain the device characteristics and >ignores all other tables. however, it seems that your thoughts go even >further ...
If the CMM has to do all gamut remapping and extrapolation, then there is no need to have some data already extrapolated. Moreover, having the training set would give to the CMM the real measurements, else it would have no clue on what points are belonging to the device and what are added by the profiler which generated the AToB1 table. So, it would be tied to the AToB1 space. Is because that I said that smart CMM would have strong impact on how we understand color management today. No profilers anymore, and a set of measured patches for each device would be enough to accomplish smooth color reproduction. This means a lot of companies selling profilers would suddenly be out of businnes. High end devices could be measured individually, whilst consumer level could have a set of "reference" measurements. No ICC profiles, No working spaces, just a small text file containig measured patches. Amazing. >Please don't misunderstand my. Even though the term "dumb CMM" >might be interpreted negatively, I really DO NOT want to imply, that >this is such a bad evil. As you say (and as I also attempted to say >previously) that's just the way how ICC color management works today. >It has some advantages (simpler CMM (though still complicated enough), >fast calculation of the transform) and it also has some limitations, >which could only be solved with a smart CMM approach. Oh, well, the dumb model is right now the only way, so I think the discussion on what would be the best is a bit futile ;-) I also like the smart CMM model, thought. But there are more things to take in consideration aside technical excellence. Anyway, I believe we are going to see some smart CMM implementations in near future, then we could compare. Marti. ----- Original Message ----- From: "Gerhard Fuernkranz" <[EMAIL PROTECTED]> To: "Marti Maria" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]> Sent: Friday, February 27, 2004 12:50 PM Subject: Re: [Lcms-user] Gamut Warning > But the implications of smart CMM are enormous. In a smart CMM > environment you need not anymore the profile as we know today. Just > the characterization data of both devices. i.e. a sheet containing > measured > patches for the scanner and other sheet with measured patches for the > printer. > No AToB/BToA tables anymore, since the CMM would compute the > transfer directly. Amazing. Marti, I was rather thinking about normal ICC profiles and a CMM which only honors the relcol AToB1 tables to obtain the device characteristics and ignores all other tables. however, it seems that your thoughts go even further ... > >Except for a few extended features (like BPC, proofing, gamut warning), > >LCMS is basically also a "dumb CMM" engine, which just applies the > >tables contained in the profiles. > > Yes.. this is how ICC profiles works today. But please note that is not so > evil at all. Please don't misunderstand my. Even though the term "dumb CMM" might be interpreted negatively, I really DO NOT want to imply, that this is such a bad evil. As you say (and as I also attempted to say previously) that's just the way how ICC color management works today. It has some advantages (simpler CMM (though still complicated enough), fast calculation of the transform) and it also has some limitations, which could only be solved with a smart CMM approach. Regards, Gerhard -- GMX ProMail (250 MB Mailbox, 50 FreeSMS, Virenschutz, 2,99 EUR/Monat...) jetzt 3 Monate GRATIS + 3x DER SPIEGEL +++ http://www.gmx.net/derspiegel +++ ------------------------------------------------------- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click _______________________________________________ Lcms-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/lcms-user ------------------------------------------------------- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click _______________________________________________ Lcms-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/lcms-user
