Some notes on different rendering intents based on experiments with an Epson 2100, custom profiles and driving from Picture Window Pro - using the LCMS colour engine.
Jo-1 recently generated some profiles for me using the latest release of PFP. For the first time the greys are pretty neutral, previous releases have given terrrible greys - so bad that the profiles were unusable. I need to give the prints time to dry properly, but compared to the standard Epson profile which gives very magenta greys on my machine, it's light years ahead. Profiles were not tweaked or edited in any way.
Paper was Canson Arches Canvas. Matte Black ink used. Paper setting was Archival Matte. Test chart was in Adobe RGB colour space.
I used N Koren's step chart - which the PFP people say isn't suitable for assessing profiles as the gradations show imperfetions that wouldn't be noticeable in a photo print
The Epson profile (SP2200 Enhanced Matte 1440MK) was printed using PWP's maintain full gamut, which is equivalent to PS' Perceptual intent. This gave solid blacks, but (as expected) a magenta caste to the grey scale. Before anyone shouts, it does exactly the same on my printer using the correct epson paper. Colours were OK, but a little muted.
The PFP profile was used with 3 intents for comparison, Maintain full gamut, Maintain Identical colours with BPC (=PS' Rel colorimetric + BPC) and PFP's recommendation - Preserve saturation (=PS' Saturation).
Maintain full gamut gave the best print overall, with neutral greys and blacks that matched the Epson ones, while still maintaing all steps in the step chart. On the fine light grey scale, Grey was distinguishable from V=253/254 - pretty good, but no better than epson on the wrong paper. Colours showed a better gradation and yellows, greens and reds were noticeably better than the epson ones, a tendancy I've noted before with PFP profiles. What was less pleasing were some strange patterns in the blue/cyan areas, at saturations from about .7 upwards, moving slightly into the greens. However they were muted with this intent. I need to print a strong sky to see if it's an issue in real life - I suspect it may be. Another efefct of these profiles is that the primary colours and their complementary colours all showed a narrow, clearly defined band of that colour - compared to the epson profiles AND some custom profiles for the printer.
Preserve Identical colours + BPC (Rel colorimetric + BPC) produced the worst print. Although it sorted out the funnies int he Blue/cyan area, it raised the blacks to such an extent that the darkest grey corresponded to a density of 1.4 when compared to the previous print. As a result there was less shadow detail and a very flat image. Grey's remained neutral, but colours were flat and lifeless.
Preserve Saturation (Saturation) Emphasised the funnies in the high saturation areas, but kept Good blacks and grey scale. Similarly the colours were good - except that it strongly emphasised the effects of the funnies in the cyan/blue areas of high saturation.
I also took a good look at the older prints I've retained. The PFP profiles aren't as good s the custom profiles I had made for the machine (on high end equipment) for HahnemÃ¼hle photo rag, but are considerably better than the stock epson ones - with reservations in some areas. I'm guessing, but will experiment later, that the reason for the narrow strong colour areas is to increase apparant contrast, so giving a punchier print with PFP profiles.
Experiment with profiles, rendering intents and test charts.
Need to watch for funnies in prints with big smooth blue/cyan tones, even in the green cyan areas. Some issues in prints may be down to profiles/rendering intents.
PFP still lags behind high end profiling equipment, but may do what you need, especially on older photo printers which seem to vary more than newer models - or where paper manufacturer profiles aren't available - or are poor.
Consider custom profiles on high end kit for critical work.
What works for me may not work for you.
Thanks Jo-1, now to look at the other profile (tomorrow).