Ad
Ad
Ad
Pages: « 1 [2]   Bottom of Page
Print
Author Topic: CS5 Plugin for Win X64 Platforms Released for iPF 6300/8300  (Read 6886 times)
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1312


WWW
« Reply #20 on: November 16, 2010, 12:27:32 PM »
ReplyReply

I'm not sure how you can say that.

1) Well, it's just an opinion and was presented as such.

2) I've made Perceptual vs RelCol prints and have traveled around with them in my print samples box showing them to clients and photographers at workshops, lectures, etc. I've repeated the test on their printers with their images again and again. If the perceptual rendering is made a certain way, 100% of the people I've polled (over 500) agree that it's preferable 100% of the time.

3) I should also say that, IMO the perfect perceptual rendering isn't found when the default perceptual rendering intent settings are used. I'm not saying what I use but I do tweak the settings for a sweet spot that I've spent a lot of time discovering. The fact that the defaults are disappointing is also a failure on the manufacture's part considering the potential that's there.

4) It bears repeating that different manufacture's get to design their own perceptual rendering - it's not defined by the ICC as the other intents are. So RelCol from GMB, XRIte and basICColor profiles will perform about the same but Perceptual can be shockingly different. Some of them perform Perceptual rendering much better than others, and some are quite disappointing. I'd like to humbly suggest that those that don't like perceptual rendering, or don't think it's for all fine art printing purposes simply haven't seen the best perceptual rendering that's possible.

FWIW, I find I often prefer the RelCol intent.
[respectfully] Then you probably haven't found the optimal perceptual rendering settings. Once you do I doubt you'd ever like RelCol for fine art purposes. If there's something that you're liking about RelCol that would be fun to discuss and I think Perceptual has an answer for it.

I've talked with the engineers involved with making MP's perceptual rendering intent and I think it's one of the crown jewels that XRite obtained with the Monaco purchase. One of many that they are finally putting into good use...
Logged

digitaldog
Sr. Member
****
Offline Offline

Posts: 9311



WWW
« Reply #21 on: November 16, 2010, 02:19:57 PM »
ReplyReply

[respectfully] Then you probably haven't found the optimal perceptual rendering settings. Once you do I doubt you'd ever like RelCol for fine art purposes. If there's something that you're liking about RelCol that would be fun to discuss and I think Perceptual has an answer for it.

I've talked with the engineers involved with making MP's perceptual rendering intent and I think it's one of the crown jewels that XRite obtained with the Monaco purchase. One of many that they are finally putting into good use...

I frankly don’t see (and haven’t seen) where the current technology, which looks at single, solid color values without placing them in any context can make a decision like this better than I can by viewing the different intents based on image content. By the time the Perceptual or RelCol intent is upon us, the software is just looking at a big pile of numeric values, Lab values no less. It doesn’t treat a black dog on coal, a white cat on snow or a very colorful scene any differently, it has no idea about how these thousands or millions of solid squares react next to each other as a visual context.

I’ve got access to the same newer tools as you, at least in terms of those crown jewels, nothing has convinced me that I should blindly pick one rendering intent.

Here’s an image, its nothing special but its an image composed of lots of colored pixels and a rendering intent makes them look different. Its my image, so I have a strong feeling about which rendering I prefer:


Here’s what a profile knows about this image and nothing more! Its simply the unique Lab values extracted using ColorThink (I decided it wasn’t necessary to provide all 2667 unique values, hopefully we all get the idea here):

SampleID   SampleName   Lab_L   Lab_a   Lab_b
END_DATA_FORMAT
NUMBER_OF_SETS 2667
BEGIN_DATA
90856   "1"   15.6204044   21.9335938   13.4414062
81515   "1"   16.7064951   24.90625   21.5703125
90211   "1"   16.4537377   26.0820312   25.0351562
82780   "1"   17.578125   0.1757812   1.46875
82162   "1"   19.4301471   25.15625   20.0742188
81839   "1"   18.950674   19.7851562   20.390625
82171   "1"   20.2925858   8.75   8.4453125
80225   "1"   19.3167892   17.2578125   13.59375
89564   "1"   20.6387868   7.9335938   17.484375
81838   "1"   22.9856005   17.1679688   19.1171875
82172   "1"   24.364277   6.8320312   8.1171875
83104   "1"   24.0211397   0.1328125   1.4140625
81516   "1"   26.5686275   31.3359375   22.1523438
82495   "1"   26.4476103   12.7226562   9.3164062
82779   "1"   27.5827206   0.1523438   1.375
89888   "1"   28.7699142   27.6914062   23.5195312
82818   "1"   30.9957108   16.0585938   10.7304688
80226   "1"   31.2178309   9.2890625   8.90625
89886   "1"   33.2153799   11.9140625   7.390625
82808   "1"   33.6672794   23.3359375   38.3476562
83464   "1"   34.8682598   15.7695312   16.1875
80219   "1"   36.4966299   9.90625   6.4179688
83787   "1"   35.8838848   16.2070312   19.4492188
81197   "1"   36.9148284   12.7421875   12.1015625
81846   "1"   36.3909314   7.7421875   7.4570312
84367   "1"   36.9056373   1.2460938   4.265625
81522   "1"   35.6755515   4.8203125   7.578125
79897   "1"   38.6488971   1.6484375   2.4960938
82781   "1"   37.8691789   0.1132812   1.3085938
79575   "1"   37.5505515   7.3632812   6.6328125
83142   "1"   39.4286152   14.0585938   14.4960938
81198   "1"   38.560049   4.484375   8.625
83141   "1"   38.9583333   15.4765625   12.0351562
82819   "1"   41.0202206   12.4101562   11.796875
83143   "1"   40.401348   11.8710938   7.2148438
80872   "1"   39.5174632   14.3242188   15.8554688
80220   "1"   40.5193015   15.1953125   9.9414062
79577   "1"   40.4764093   9.0078125   7.2382812
83103   "1"   41.5946691   0.1210938   1.2773438
82487   "1"   41.6145833   24.7773438   26.3789062
79900   "1"   42.0036765   16.078125   9.6757812
95346   "1"   43.1724877   19.3242188   25.8203125
80227   "1"   42.901348   7.1484375   6.4453125
82496   "1"   42.5811887   7.4726562   7.1992188
80876   "1"   43.0208333   2.4765625   6.4804688
82494   "1"   43.9154412   14.0507812   9.3320312
81521   "1"   42.8354779   5.0117188   9.9140625
84110   "1"   43.7806373   14.5351562   21.1484375
81844   "1"   42.3284314   6.2226562   11.0429688
82163   "1"   44.4883578   29.5976562   25.1757812
80541   "1"   44.5189951   5.625   2.7109375
82491   "1"   44.263174   8.9101562   13.8867188
95347   "1"   44.2830882   19.0078125   23.2148438
89563   "1"   45.6341912   13.96875   17.1875
82164   "1"   46.4705882   22.25   21.6445312
90239   "1"   44.7518382   2.8671875   6.5976562
80551   "1"   45.713848   6.9257812   6.9140625
63078   "1"   46.0278799   0.0039062   -0.0078125
83105   "1"   45.2665441   0.1171875   1.2578125
84364   "1"   46.596201   0.3515625   5.625
82486   "1"   44.8820466   30.5585938   28.6992188
76324   "1"   46.4828431   0.3320312   -1.1328125
76002   "1"   46.0784314   0.328125   -1.1289062
83137   "1"   45.8287377   10.2460938   13.7695312
82173   "1"   45.5560662   2.859375   6.5703125
84435   "1"   45.1501225   9.1054688   15.140625
85005   "1"   45.1531863   2.8632812   6.5820312
85403   "1"   46.8259804   9.2382812   16.328125
68897   "2"   46.885723   0.3320312   -1.125
81845   "1"   46.6421569   5.6875   8.2226562
79898   "1"   46.8673407   11.4960938   6.9296875
66629   "3"   46.8367034   0.0078125   -0.0078125
85080   "1"   47.432598   15.8125   22.6914062
84359   "1"   47.5199142   4.2695312   8.2656
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
digitaldog
Sr. Member
****
Offline Offline

Posts: 9311



WWW
« Reply #22 on: November 16, 2010, 02:29:11 PM »
ReplyReply

As I understand it this is not about running Photoshop in 64 bit mode with the ipf x300 plugin, it's about running under CS4 and CS5 that are installed on 64 bit versions of Windows.
Mac users aren't missing any kind of critical update here as far as I can tell.

The version I had (4.00) wasn’t 64-bit savvy but yes, it appears that this 5.00 version just released yesterday is. Going to download and give it a try in CS5.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #23 on: November 16, 2010, 02:42:30 PM »
ReplyReply

[respectfully] Then you probably haven't found the optimal perceptual rendering settings. Once you do I doubt you'd ever like RelCol for fine art purposes. If there's something that you're liking about RelCol that would be fun to discuss and I think Perceptual has an answer for it.

Well, I think it depends a great deal on the printer's gamut and the dynamic range of the paper whether or not Perceptual or RelCol will look better for a particular images and yes, of course it depends a lot on who made the profile and what settings were used when making the Perceptual tables...

For my work printed on an Epson 9900 usually with EFP and images in ProPhoto RGB (oh, BTW, the working space of the image plays a huge role on Perceptual vs RelCol as well) I rarely prefer a Perceptual rendering because of the impact in the overall tonal range Perceptual forces on an image...the Perceptual mapping not only maps colors but tones as well. Considering the large gamut of the 9900, I rarely need to worry about the colors that end up out of gamut. I'm usually much more concerned with the tonal relationships in the image.

If you are often printing to low dynamic range watercolor papers whose D-Max is poor, I suppose the tonal adjustments of Perceptual may actually help certain images regardless of the gamut compression.

Personally I think it's better to tell people that the only best rendering intent is the one that makes YOUR particular image look the best. Blindly telling people to use one or the other I think is wrong.
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 9311



WWW
« Reply #24 on: November 16, 2010, 02:45:06 PM »
ReplyReply

The version I had (4.00) wasn’t 64-bit savvy but yes, it appears that this 5.00 version just released yesterday is. Going to download and give it a try in CS5.

Well the good news is, the plug-in does run in CS5 on Mac under 64 bit. Bad news is the same CMM error message posted below. Love to get that fixed somehow.
Other issue is the older ViseX installer (3) that Canon is using crashes upon typing your admin password IF you have a non U.S. System loaded in Language & Text. Its a known bug that was fixed in the ViseX version 4 installer.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1312


WWW
« Reply #25 on: November 16, 2010, 03:45:05 PM »
ReplyReply


Andrew - great image for evaluating highlights!

Personally I think it's better to tell people that the only best rendering intent is the one that makes YOUR particular image look the best. Blindly telling people to use one or the other I think is wrong.

Fair enough! Well said. Results will vary and experimentation is encouraged.
Logged

keith_cooper
Sr. Member
****
Offline Offline

Posts: 470



WWW
« Reply #26 on: November 16, 2010, 05:50:47 PM »
ReplyReply

...Other issue is the older ViseX installer (3) that Canon is using crashes upon typing your admin password IF you have a non U.S. System loaded in Language & Text. Its a known bug that was fixed in the ViseX version 4 installer.
Minor correction: Non 'English' I believe rather than non US?  There is no country defined in the Language section of these particular prefs. Formats shows UK as region on the machine I installed the plugin on yesterday.

As to choices over RI, fully agree that looking at the options and deciding what looks best is usually the best way.

Does anyone know if a 64 bit version of the Adobe CMM is in the works?
Logged

digitaldog
Sr. Member
****
Offline Offline

Posts: 9311



WWW
« Reply #27 on: November 16, 2010, 06:27:52 PM »
ReplyReply

Minor correction: Non 'English' I believe rather than non US? 

In my case, I have custom input source name that replaces the U.S. (flag) banner that will produce the crash. It has to be set to the US icon to avoid the crash.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #28 on: November 16, 2010, 10:08:52 PM »
ReplyReply

Quote
IMO, that represents a failure on the profile manufacturer's behalf. Excellent perceptual rendering will trump RelCol for fine art printing. The profile manufacturer plays a big part here and Monaco Profiler and it's upcoming successor have the best Perceptual rendering. Once you use it you'll never want to go back to RelCol again. Better "edge gamut" saturation without loss of detail, "middle gamut" colors stay the same, better gray balance, better shadow detail, etc.
I would tend to agree with your for matte prints, where the compressed tonal range and limited gamut can be problematic. My experience is that with fiber-gloss papers, these problems just don't exist for many images.

I always try both options when soft-proofing. But what I find is that for fiber-gloss prints a lot of times the perceptual intent boosts the shadows too much. Maybe for folks who would otherwise be complaining about prints being too dark this would be a good thing, but I have things calibrated so that I don't want or need an extra boost in the shadows.

I know you're a big fan of MonacoProfiler (and reading between the lines, the new i1Profiler). Maybe their perceptual intent is better, and I'll be able to try it for myself once the new software is available. But the truth is rel-col often gives me a very good screen-to-print match for fiber-gloss prints, whereas from what I've seen the perceptual intent will be more likely to look different in some way. I guess it's partly a matter of philosophy; I don't want the profile to do any "enhancements", I just want it to give me a good match between screen and print.

Logged

deanwork
Sr. Member
****
Offline Offline

Posts: 749


« Reply #29 on: November 17, 2010, 01:12:30 AM »
ReplyReply

Keith,

What in world is going with these measurements? I"m totally confused, and must be reading something wrong or there must be some typo.

You are listing a dmax on the Innova Ultra Smooth gloss at 1.61?? If that is accurate it makes no sense at all. Even on a Photorag type matte media my Z3100 is giving me 1.8+ on a daily basis, and even my Cone K7 inks that do not have great dmax by any stretch are giving my 1.67 on MATTE media. My Z on this Innova gloss and semi gloss paper and the rc photo medis is in the realm of 2.4 or so and Epson is up there about 2+ also with Pk. Something is not right. I must be reading this table wrong? 1.62 is not Black it is gray and on the Innova F gloss it is flat gray, no where close to black.

Before I buy an 8300 please tell me it ain't so.

john
Logged
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #30 on: November 17, 2010, 01:24:55 AM »
ReplyReply

Dean, I think those measurements of Keith's were for a matte paper.

That said, if DMax on matte papers is a big concern for you, sticking with HP may be your best bet as they currently seem to be champs in this area. I've never gotten 1.8 from a matte paper with my Canons; maybe it's possible with the right paper and media settings, but I haven't found typical results for matte papers to be around 1.65 +/- .05.

Glossy/semi-gloss papers that use the photo-black ink are another story though, most of them will be in the 2.4-2.65 range with an x300.

To be honest, the DMax performance on matte papers has not been an issue for me. For images where a DMax of 1.7 isn't good enough, 1.8 still isn't going to be good enough, so I'm just going to use a fiber-gloss paper anyway.
« Last Edit: November 17, 2010, 01:33:28 AM by JeffKohn » Logged

JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #31 on: November 17, 2010, 01:26:45 AM »
ReplyReply

Quote
Does anyone know if a 64 bit version of the Adobe CMM is in the works?
I sure hope so. But I looked at AdobeForums and this very question was asked 6 months ago and went unanswered, which I don't take as a very good sign.
Logged

keith_cooper
Sr. Member
****
Offline Offline

Posts: 470



WWW
« Reply #32 on: November 17, 2010, 03:16:20 AM »
ReplyReply

Worry not :-)

The measurements were for Innova IFA 11 - Smooth cotton natural white. I use the IFA designation since I've numerous boxes of sheets and rolls of Innova paper and the names are often quite similar and may be different internationally.
Logged

Scott Martin
Sr. Member
****
Offline Offline

Posts: 1312


WWW
« Reply #33 on: November 17, 2010, 08:35:27 AM »
ReplyReply

I just want it to give me a good match between screen and print.

And that's what you'll get with really exceptional perceptual rendering. Better match than with RelCol! Lots of people haven't seen his yet because they're messing around with products with crappy perceptual rendering, IMO. Anyway, something to look forward to... enough for now.
Logged

deanwork
Sr. Member
****
Offline Offline

Posts: 749


« Reply #34 on: November 17, 2010, 09:34:26 AM »
ReplyReply

Keith,

I did read that wrong I though you were referring to the Innova Ultrasmooth gloss. And from my experience Innova has slightly less dmax than Hahnemuhle of a similar surface but they are very close.

You know, working with different inks all the time now I've kind of changed my thoughts about all this in the last few years. I used to accept the 1.64-1.66 damx of the original Ultrachrome  and the same for Piezography as just fine on matt media. But you know after a few years on the Z, I have altered my thinking. This is especially significant to my own work that includes large black backgrounds on about half of what I do now. The old inks just look gray to me side by side. One exception to the non-hp black that is really good is the Ebony carbon black sold by MIS. It will get very close to the HP black, in the realm of 1.77 or so on Photorag. It's fade tests are excellent on the Aardenburg site so that is what I'm putting in my K7s this month as a boost.

Its funny because I used to be the one arguing on these forums for not being concerned about dmax on matte media, that it was way over rated, and many jobs it clearly doesn't matter, but when you have two big prints with black backgrounds sitting side by side and one is just killing the other, you kind of have to recalculate your needs. At least I do. And for what its worth my matte media prints have a density that looks darker and much richer than any fiber gloss. It is an illusion of course, it is the lack of reflections make it look darker.

Having said that the Canon is excellent on the fiber gloss media and apparently beats the Z series  and everything else too on gloss differential issues, which is ironic considering it has no separate gloss enhancer channel like the hp that works so well on rc media. ( I still end up spraying my Z fiber gloss prints with uv coat) So, I"ll still go with the IPF8300 for the next one most likely and renew the warranty on my HP once more because I'm addicted to that look now on the Canson matte media for bw.

john
Logged
Pages: « 1 [2]   Top of Page
Print
Jump to:  

Ad
Ad
Ad