Note: Landa recently announced a new version of their press, the S11P. The most significant difference is speed. Where the presses I have tested print 6,250 two-sided impressions per hour, the new Landa press can print at twice that speed. This puts the S11P squarely in competition with most B1-size offset presses (28 x 40 inch). Other benefits accrue as a result.
This is the second in a series of posts about the Landa S10P press and its capabilities. To read the first post, please click here.
In the previous post I explained how I was attempting to profile the color capabilities of a Landa S10P press, using its seven-color gamut on an excellent coated paper (same paper used when profiling for GRACoL in the U.S.).
I had the printer, Brodnax 21C, in Dallas, Texas, print two different press sheets on their machine. One is populated by RGB source files, the other primarily with CMYK images.
Today’s discussion is about profiling the Landa S10P, and specifically about making an ICC profile for that press printing from both an RGB test target set, and one in CMYK, both generated by X-Rite’s i1 Profiler software. The RGB target is a “standard” 2033-patch set with the colors designated as RGB values. I have run this target many times, on many devices, all ink-jet printing machines that expect RGB as source files. The CMYK target is an ECI target.
While teaching in Germany in 2021-22, I used a Landa press to make a similar test sheet, then I visited the labs at FOGRA, Germany’s printing industry research organization, to measure the results and make the profile. I declared it a “failure” because the profile was almost identical to sRGB, which is the second-smallest color gamut in the industry. I was convinced by that experience that it was my error, and that I had not prepared something correctly, I had not used the correct target, or had made some other error.
Despite my profiling “failure” in Munich in 2021, the rest of my test sheet revealed several important things, one of which was that the color space we had chosen to use – ProPhoto RGB – was not acceptable, because the Landa machine, and its Fiery front-end, did not manage the color in that space correctly. It made our photos look muddy, We switched, mid-project, to Adobe RGB, which is a less voluminous color space, and which has many fewer potential colors inside its envelope. That one discovery saved our project, preventing a disastrous printing job.
This year’s test was similar. I put a standard i1 Profiler target set of 2,033 patches on the RGB sheet, and sent the file to Brodnax. The visual impact of that target set is lovely. It looks great. The CMYK test target set is similarly nice, though it doesn’t look as beautiful.
I also had a benchmark for this year’s test sheet – a reproduction of that same 2,033-patch target set printed on another ink-jet machine with more than four colors of ink. This machine is the Epson P9000, which prints with CMYK, Orange and Green (it also has “light magenta,” light cyan and two “light blacks”). The profile that my students made from that target set is glorious. It shows what an expanded-gamut printer can do.
When in Germany, I had nothing against which to compare the performance of the Landa S10P press. This time it would be different. Both are ink-jet machines, both use water-based inks, and both are printing on a premium paper (the Epson paper is glossier, but still a comparable medium). If anything, the absence of a blue ink on the Epson should give the Landa an advantage.
My testing process is common: run the test sheet, then read the sheet with a spectrophotometer (an X-Rite Isis 2 XL), then use i1 Profiler to calculate and create the profile, using ICC v4 standards for the calculations.
I run my analyses in ColorThink Pro 4 from Chromix. That software is the latest version for color profile analysis and editing. It is capable of making visual comparisons in CIELab color, CIEYxy, LUV, and two other less common color spaces. I used both Yxy and Lab for these analyses. The data are the same.
On Landa’s web site they show a CIELab graph of the color of their press, claiming 86 percent coverage of the Pantone library. I have no reason to doubt their claim, and was hoping that my profiling would prove their claim to be correct. (Note: in October, 2024 I profiled a BenQ RGB display that claims 98 percent coverage of Adobe RGB. I was skeptical, but found that their display is slightly better than they claim!)
The graph used on the Landa site is shown in CIELab color (but, oddly, flipped horizontally). I have flipped their chart here so that the colors in my analyses are represented in approximately the same coordinates, and along the axes commonly used in color analysis (red at the top with 0° hue angle, green 120° counter-clockwise, and blue 240° counter-clockwise from red. This also aligns with the CIELab charts in ColorThink Pro, in i1 Profiler, Adobe Photoshop, and others).
I made a profile of the seven-color output on the Landa press, taken from the RGB test patch set of 2,033 color patches. Then I made a profile of the CMYK target (printed with all seven colors) using the ECI 2002 CMYK random patch set. In both cases, I measured the output with an X-Rite Isis2 XL spectrophotometer.
Once the profiles were made, I opened each for analysis in ColorThink Pro 4, the most recent analysis tool from Chromix. Using the CIELab color space for these tests, I compared the RGB results to those I got from an earlier test of our Epson P9000 ink-jet printer on Epson glossy photo paper. This paper is glossier than the C2S SAPPI paper used to print my test sheets on the Landa press.
The gamut of colors of the Landa S10P seems to be limited to, or very closely aligned with the sRGB color space, which is disappointing, because sRGB is the second-smallest common RGB color space in the industry. This is exactly what I experienced at FOGRA in Munich in 2021; then I assumed that I was at-fault, that I had chosen the wrong RGB target patch set, or that I had handled the target file incorrectly, converting it accidentally to sRGB (default on most installations of Adobe Photoshop). It turns out that the same target patch set printed on the Epson P9000 generates a dramatically larger color gamut. This proves that it’s the management of the color on the Landa prepress system (a Fiery) and not my handling of the file, that caused this constraint.
The CMYK color patch set, the ECI 2002, produced a much different result: it aligns almost exactly with the GRACoL color gamut, that used for measuring and certifying the output of offset presses in the United States. I will test this against one of the FOGRA 37 and 39 CMYK color gamuts in another blog post..
And here is the mystery of this experience, so far:
With such a potentially huge color gamut (as promised by Landa), what is preventing the S10P from printing colors outside the gamut of sRGB on the RGB front, and outside of GRACoL on the CMYK front?
As a closing thought, I made a chart of exclusive colors on both sheets – remember these are printed on the same machine, using the same seven colors of ink, at the same time on the same paper – there are patches of color that are exclusive to each color space, but ironically inside the opposite color space. These are colors that can be printed with CMYK, and colors that can be printed by RGB (converted to CMYK-Red-Green-Blue in the Fiery) that cannot be printed by CMYK colors (and also processed by the Fiery to the gamut of colors on the Landa). This is illogical, since they obviously can be printed by the machine on that substrate at the same time.
Why would a printing press have colors that it can print not print?
I’ll leave that for us to ponder until the next post.
