Profiling Ilford Gold Fibre Silk w/Spyder3Print

[TimG]

Hi,

Tried twice to profile some Ilford Gold Fibre Silk this afternoon, using my Spyder3Print (part of the Spyder3Studio bundle, not the newer SR version) and the results were nowhere near as good as Ilford's profile.

- In Spyder3Print, I chose the following: 229 Patch on 1 Sheet (8.5x11).  I did not choose the Extended Grays.
- Set the paper type for my Epson Stylus Pro 7800 to Premium Semigloss Photo Paper as per Ilford recommendations.
- Turned Color Management off in the Epson driver.
- Waited 20 minutes after the target finished printing before I measured it.
- Measured the profile under my standard viewing light conditions (same as editing, ambient light well within the acceptable range as reported by Spyder3Elite).

The first measurement resulted in a profile which, when viewed in Photoshop CS4 (Mac), with a custom Soft Proof condition for that paper type, using a known neutral monochrome image was very warm, almost sepia.

The second measurement used the 229 target, plus the 238 Extended Grays target; this resulted in an unacceptable profile with severe posterization (looked like I ran a Cutout filter on it), shadow clipping, and an reddish-brown color cast.

The other papers I have profiled (Inkpress Luster, Moab Colorado Fiber Gloss, Epson Premium Luster) have been fantastic - not a single problem.

Any idea what may be causing this?

[KeithR]

Can't help you with your problem, but have you down loaded the newest drivers for SpyderPrint3? Recently released 4.0.1
You might want to check the DataColor group at Yahoo.

[BobDavid]

You really should let the print dry for a couple of days before measuring it for a profile. No wonder your profile came out poorly. Wait at least 48 hours.

[TimG]

You really should let the print dry for a couple of days before measuring it for a profile. No wonder your profile came out poorly. Wait at least 48 hours.

I didn't make an actual print, I soft proofed the image. The Spyder3Print states you only need to wait 15 minutes, not 48 hours.  

Maybe check your facts before making smart ass remarks?

[MHMG]

I didn't make an actual print, I soft proofed the image. The Spyder3Print states you only need to wait 15 minutes, not 48 hours.  

Maybe check your facts before making smart ass remarks?

I don't think it's a smart ass remark.  48 hours may be a bit conservative, but 24 hours isn't, and 15 minutes is very often too short a time frame after printing for truly accurate color.  Short term drift issues with inkjet prints is a very real phenomenon and for highly accurate profiles one must wait at least 1-3 hours after printing for pigmented inks on most papers, 24-48 hours for dye-based systems (and even then some poorly mordanted dye/paper combinations just keep on drifting and drifting due to humidity induced dye migration!).  However, the color drift I'm talking about is generally constrained to about 2-4 delta E for pigmented systems during that critical first hour of "dry down" but can be as much as 10-12 delta E for dye-based systems.

That said, if you were profiling a pigmented ink on paper print system and only waited for 15 minutes the 2-4 delta E error, usually in high chroma colors, isn't going to completely blow your profile nor would these subsequent short term drift errors on real prints show up in the soft proof of that profile. If you had serious profile issues more than likely you had a measurement error caused by improper placement of the instrument on the paper color patches, lifting the instrument head prematurely during a reading, hooking in the color ramp due to improper ink loads, etc.

[BobDavid]

I didn't make an actual print, I soft proofed the image. The Spyder3Print states you only need to wait 15 minutes, not 48 hours.  

Maybe check your facts before making smart ass remarks?

I earn my living making fine art reproductions for picky customers that demand accurate color in their prints. Do you? In Central Florida, it's a good idea to wait a couple days before reading color patches for profiling; 24 hours is probably fine in a dryer environments. FYI, I use EyeOne Match3 for camera and printer ICC profiles.  

You might consider exercising a bit more etiquette online.

[NikoJorj]

However, the color drift I'm talking about is generally constrained to about 2-4 delta E for pigmented systems during that critical first hour of "dry down" but can be as much as 10-12 delta E for dye-based systems.

Ditto : if I remember well, it's one of epson's biggest advertising arguments that the DeltaE is minimal after a short while, for the proofing market. I couldn't retrieve data for K3 inks but here is some for HiGloss2 (R1900) : (PDF 1.5Mo) less than 2 DeltaE after 15mn... but I suppose it's only valid for the proofing paper (and I didn't verify it myself!).

If you had serious profile issues more than likely you had a measurement error caused by improper placement of the instrument on the paper color patches, lifting the instrument head prematurely during a reading, hooking in the color ramp due to improper ink loads, etc.

Can't agree more. Severe posterization would indicate some discontinuity in the measures taken, I'd say?

[digitaldog]

Ditto : if I remember well, it's one of epson's biggest advertising arguments that the DeltaE is minimal after a short while, for the proofing market.

Correct. Here’s my data on Luster paper, Epson 2880 one hour after printing and 24 hours later. Target is an iStar.

Overall - (323 colors)
--------------------------------------------------
Average dE:   0.10
    Max dE:   0.53
    Min dE:   0.01
 StdDev dE:   0.08

Best 90% - (290 colors)
--------------------------------------------------
Average dE:   0.08
    Max dE:   0.18
    Min dE:   0.01
 StdDev dE:   0.04

Worst 10% - (33 colors)
--------------------------------------------------
Average dE:   0.29
    Max dE:   0.53
    Min dE:   0.19
 StdDev dE:   0.09

--------------------------------------------------

[BobDavid]

So how noticeable, Digitaldog, are these differences between best and worst as typical photographic prints side by side?

[digitaldog]

So how noticeable, Digitaldog, are these differences between best and worst as typical photographic prints side by side?

Not sure I understand the question.

All the values above are non visible differences (anything less than deltaE of 1.).