Printprism: I have bought it to test its capabilities and quality of the rgb profiles

[ohtoulouz]

Yes, I know. But space on my desk is limited ! By the way, I bought PrintPrism and made profiles with the XP-970 and the result is excellent. Specifications of XP-970 et ET-8500 scanners are exactly the same on Epson website: I am pretty sure they use the same hardware.

 

[pharmacist]

OK good to know. So you use the ET-8550 scan engine of the PrintPrism to generate the profile. Maybe Mike Chaney would be pleased that the XP-970 scanner has the same results because it uses the same scanning engine.

 

[pharmacist]

That is strange: I tried to use my Epson WF-7840 scanner if it would give a good profile (officially not supported) and to my astonishment the profiles are almost the same:

I tried the WF-7840 printer as scanner and using the ET-8550 scanning engine in PrintPrism:

Beneath softproofing with a PrintPrism generated profile: left with my Epson WF-7840 (not supported) and right with the supported ET-8550....They look similar to me.....

Now the 3D plot on iccview.de:

test.icc is the profile made with the officially not-supported WF-7840 scanner:

This is very surprising to me....Could it be the WF-7840 is using the same scanning hardware als the officially supported ET-8550 ???

 

[ohtoulouz]

Yes, exactly. There is no scanner restriction. I choose Printprism ET-8500/8550 option and when the Scan2 window open, I choose XP-970 scanner. Problem is I cannot compare my results with those of an approved scanner. I just can tell that the results with XP-970 are good. I do not know if they are as good as possible unless I buy a Canon Lide 300 or an Epson V39II. So maybe it is too early to tell Mike Chaney to put the XP-970 on his list of approved scanners.

 

[ohtoulouz]

I am sure that there is not that much different scanner hardwares around ! Business is business. If one hardware fits all, it is all benfit for the company !

 

[ohtoulouz]

However, looking at the spec Of the WF-7840 scanner on the Epson website, it not execaly the same specs: (2400dpi max instead of 4800). But it does not mean that the hardware is different.

 

[W. Fisher]

I own a copy of SilverFast Studio AI 9 that I use with an Epson Perfection 600 scanner. It uses a it.8 Version 2 target for a reference prior to scanning for a printer profile. Not a cheap software with the it.8 target, maybe $400.

I also bought a Canon LiDE 300 scanner and Print Prism given the hype.

Honestly, the Print Prism produced a better test image than the SIlverFast Studio AI did, and was significantly cheaper too. The SIlverFast needed an edited profile as the ring-around color test patches were off a lot. The edited version was no better. A third edit version to correct it was no better either than the first one out of Print Prism.

Print Prism was very fast, less paper and ink waste for the test, and easy to use.

Later, I just made a print profile from a 3rd party dye ink (BCH) loaded into one Epson 3880, and the normal pigment ink in the other 3880 made using Print Prism. The dye ink was larger color gamut in the wireframe than the pigment ink, which I sort of expected as shown below.

 

[ConradH]

I've been fooling with Print Prism for a couple weeks now. Also used to use Profile Prism until it became difficult to get decent IT-8 targets with good data files. IMO, IT-8 isn't good enough anyway. My scanner is LiDE 400, not 300, but it works perfectly. My normal target size is a half sheet, 5.5x 8.5, but you get better results with a full sheet. Very few variable or out-of-range patches. For some reason it's critical to get the target dead-nuts square in the scanner. That's the key to getting just a couple clipped or out of range patches. The latest version of the software automates just about everything, so it's near impossible to make a mistake. Turns out you can still access the scanner settings if you hold Ctrl while clicking on the scan button (per Mike, I haven't tried it).

Visually my results are excellent, with both the Eval print everybody uses, and my own stuff. Some subtle color shifts I've been fighting are gone now. Mike says not to obsess over soft proof gamut warnings and profile volume, but the data geek in me still wonders. If I use ProfileMaker/ProfileEditor 5.0 (gamut view) to compare the 2D and 3D gamut displays using my Prism profiles, my i-One spectro profiles, and Red River factory profiles, the Prism profiles are definitely quite a bit smaller, especially at lower L* values. I need to do more tests, but so far I can't see anything in the prints that suggests any shortcomings in the Prism profiles.

This is certainly fool-proof, faster and way more paper and ink efficient than messing with the spectro.

 

[mavtop]

someone has tried with the latest 1.4.0 version ?

it seems it works better with 5 passes (averaging all the passes) and 60 seconds delay between each pass (to get the scanner lamp less hot), it will need about 5 minutes to scan and 2-3 minutes to generate the profile.

if someone has time and want to give a try to see if the resulting profile is better than the previous

 

[ConradH]

I compared and found only the slightest difference at low values of L*, probably because the noise is slightly better with the 5 passes and cooling. I don't think anybody could ever see it and Mike said he was just after the last couple percent improvement that could be had.