We examine some Epson 2100 print samples

10th May - 2002
By Ian Burley

We analyse colour and mono print samples from Epson
INTRODUCTION and colour print analysis

As there has been so much excitement generated by the recent launch of Epson’s new Stylus 2100 Photo printer, but a long-ish wait expected for review units and units to buy, we thought we’d use some of the latest print samples from Epson to glimpse at the potential of this interesting pigment ink printer.

Below are some magnified close-ups, in colour, from the new 2100 and also the dye-based 950 A4 model also announced by Epson, plus Canon’s S900 dye-based contender.

It must be stressed that we aren’t able to compare like for like as we haven’t been able to produce our own test prints on our own 2100, but it’s an interesting enough exercise nonetheless.

Has Epson fixed the problems previously endured with pigment ink?

Normal viewing

Before we look at the magnified views, it has to be said that viewed in the hand the colour sample print on semigloss paper looked very good. Colours are vibrant, there is little to reveal that it’s a digital print. My only word of caution is that it’s a sample print and no-doubt the image was selected meticulously to show off the best of the printer.

All scans were made using a HP 7450 2400x2400ppi flat bed scanner.

Directly below is a thumb nail of our test print model’s right eye close up:

 


And below is the Epson 2100 6x8 inch sample thumbnail, printed on Epson Premium Semigloss at 2880x1440dpi plus its selected close up portion:

 

The 2400ppi scan of the Epson 950 print shows just how small the 950’s 2 picolitre dots are. The half-toning makes dot-spotting pretty difficult. Horizontal micro-banding at this intense magnification is characteristic of the 950.

We were originally highly impressed by Canon’s ability to hide its dots, but the S900 evidence above simply reinforces the remarkable feat of the Epson 950. While the Epson lays down horizontal micro-bands, the Canon S900’s half-toning clearly makes vertical columns. These are much more discernible than with the Epson 950 and visible, if you squint, to the naked eye, especially where there is solid dark colour.

And here (above) is the 2400ppi selection from the 2100 print - remember, this is a pigment ink print. The dot arrangement is very diffused. It’s not quite as fine or smooth as the 950, but the dots are better hidden than with the Canon S900. It’s difficult to tell how much detail is retained without being able to compare like with like, but the road sign makes an excellent impression of retained detail.

At 600ppi, we get a more general view of the fine detail. The lack of dot visibility remains a significant feature in the Epson 950 print above. Unlike the Epson 895 (see further below), there is no clumping of the dots. Detail preservation is very good. The horizontal structure of the half-toning is just visible here.

Here (above), the S900 scores well in preserving detail, but the half-toning is more evident thans the S900. It’s a good result, but the Epson is better.

At this magnification (600ppi) there is just a touch of evidence of some microbanding with the 2100 - a bit like the 950 above. The grain structure isn’t quite as refined as the 950. It’s more comparable with the S900 above.

Just to make it complete: below are also soe 600ppi samples from Epson’s 895 and a Hewlett-Packard Photoret III based printer:

The Epson 895 (above) does a good job of hiding its bigger (4pl as opposed to 2pl with the 950) dots, but the price paid is some obvious clumping. The result is also slightly softer.

Here is a scan from a Hewlett-Packard psc 950 all-in-one scanner and printer. It shares the same Photoret III engine as other HP photo ink-jet printers (our tests with a Photosmart P1218 printer produced virtually identical results). Here you can see the dots clearly, the detail is blurred and the result is too red.

 

[Back to top of the page]

 

| Feedback | Main news page | Discussion forum | Newsletter | Home |
 

 
 

DPN News service

www.dp-now.com