Sir Maxwell
Forum Enthusiast
I've seen Phils (?) - is that it?
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Where are the R1800 reviews?
- Thread starter Sir Maxwell
- Start date
Walter #37034
Forum Enthusiast
Look for R800 reviews. The R1800 is just an A3+ version of the R800, that why we won't test it again. Many others may feel the same way.
--
Johan
Sir Maxwell
Forum Enthusiast
The photo-i review i saw already - I think I was thinking that was Phil's - maybe it is - dont know.
As far as the R800 - it has reviewed out pretty good, right?
general concensus.
As far as the R800 - it has reviewed out pretty good, right?
general concensus.
Russell Williams
Active member
Two differences: paper handling, and color gamut. Epson is claiming a larger gamut on the R1800 because of different RGB -> 7 color conversion in the driver. I would be interested in a comparison to try to verify this.
Larger than normal CMYK or CcMmyK printers, yes. But not larger than the R800, because the R800 uses the same inks.
--
Johan
Micah Scheu
Well-known member
Vincent Oliver at Photo-i touched base about the wider color gamut a little. Here is an Epson article that goes into more detail about the software engine of the R1800 producing a wider gamut than the R800.
http://www.epson.co.jp/e/newsroom/tech_news/tnl0411single.pdf
Micah
colourgeek
Veteran Member
Actually the driver was clipping colors, an ongoing trick to cover. The inks ,yes are the same , but the software will allow better conversions according to Seiko. A lot of people found it a benefit to use a third party driver or rip such as ImagePrint to get more out of the printer than the Epson driver would allow. I do hope they achieve this with the R1800.
--
Neil Snape photographer Paris http://www.neilsnape.com
The way I understand it, this only applies if you were using Epson's color conversion engine (PhotoEnhance) with the R800. The color gamut of the R1800 is not really different (how could it be with the same inks?), it's just that PhotoEnhance for the R800 only utilized 79% of the gamut. The new PhotoEnhance uses 98%.
However, most professionals do not use Epson PhotoEnhance anyway, they use Photoshop's color management. From this article I see no reason why using Photoshop's color management would also give a different result.
--
Johan
Russell Williams
Active member
You're misunderstanding the article. I say this based both on my reading of the article as one of the implementers of Photoshop's printing code, and my conversations with my fellow Photoshop team member who's talked to the Epson engineers about this technology.
The PhotoEnhance improvements and the color conversion engine improvements are separate and independent things. I don't use PhotoEnhance either, but if you use Epson's driver, you use the color conversion engine -- even if you use Photoshop color management.
Whatever color management you do in Photoshop -- even if you use a custom profile created for your specific printer and paper -- in the end, Photoshop sends RGB data to the Epson driver. The driver has to turn those RGB values into locations and sizes for droplets of seven different inks. That's a difficult problem, and one that's the subject of much research. Here are some considerations in addition to maximizing total gamut size:
- For many of the possible RGB input values, there will be many different combinations of those 7 inks that will produce the same perceived color.
- Some ink combinations may be equivalent only under certain light -- different ink combinations may be equivalent in some light but produce different levels of metamerism
- You don't want sudden changes in ink density for small changes in inputs -- you don't want to switch suddenly from producing a pure red with C + Y to producing it entirely with R. (For one thing this would make metamerism far worse -- something that was a gradient under one light would have a sharp dividing line in another light).
- The RGB input has 16 million possible values. You want the printed output from those inputs to be smooth -- you don't want a huge jump in the color produced when you go from input of 250, 0, 0 to 251, 0, 0.
Thus it is completely credible that they could have gone from
Since I haven't talked to the Epson engineers myself, it is also possible that they've hooked things up internally in a different way so that you only get the the gamut expansion if you use their "Adobe RGB" printing mode -- where you send the driver Adobe RGB data and it does the conversion to the appropriate paper profile, rather than doing the Adobe RGB (or sRGB) to paper profile conversion in Photoshop.
What this change means in real-world images is still unknown. This is why I'd still like to see comparison tests with the R800.
You may be right in what you are saying, but in that case the article is not correct or at least not complete. The article does not talk about the process of converting RGB data to printer dots, it only talks about converting one color space (AdobeRGB) into another one. Let me quote a few lines:
"But with the introduction of our new color conversion engine—software that translates or converts one color space into another—we have come close to overcoming these differences,” says Nakajima."
"To print Adobe RGB data correctly, however, a user needs to understand how to use ICC (International Color Consortium) Profiles, notes Nakajima. These are data files that describe how particular devices (e.g., monitors, digital cameras, scanners, printers) reproduce color and also help translate color data created from one device into another device’s color space. “Choosing which Profiles are best for the devices involved, as well as dealing with a series of associated processes, is a complicated matter, to the point that this has become a barrier for many users,” says Nakajima. “So we have provided an easy way to print out Adobe RGB data that entirely bypasses the need to deal with ICC Profiles.”
That's why I concluded that this article only suggests that the new technology may be easier for non-'color management savvy' users, but does not necessarily also improve the results for those who are using icc-profiles already. But that may indeed be an incorrect conclusion.
--
Johan
"But with the introduction of our new color conversion engine—software that translates or converts one color space into another—we have come close to overcoming these differences,” says Nakajima."
"To print Adobe RGB data correctly, however, a user needs to understand how to use ICC (International Color Consortium) Profiles, notes Nakajima. These are data files that describe how particular devices (e.g., monitors, digital cameras, scanners, printers) reproduce color and also help translate color data created from one device into another device’s color space. “Choosing which Profiles are best for the devices involved, as well as dealing with a series of associated processes, is a complicated matter, to the point that this has become a barrier for many users,” says Nakajima. “So we have provided an easy way to print out Adobe RGB data that entirely bypasses the need to deal with ICC Profiles.”
That's why I concluded that this article only suggests that the new technology may be easier for non-'color management savvy' users, but does not necessarily also improve the results for those who are using icc-profiles already. But that may indeed be an incorrect conclusion.
--
Johan
Basil Meddings
New member
Dear Mr. Williams,
I have just read your post as I am in the middle of revising some notes for a "Digital Photography" course that I give to some of my fellow Seniors. May I have your permission to include, with proper acknowledgement, that section of your post dealing with color rendition in inkjet printers ?
Yours truly,
Basil meddings
Whatever color management you do in Photoshop -- even if you use a custom profile created for your specific printer and paper -- in the end, Photoshop sends RGB data to the Epson driver. The driver has to turn those RGB values into locations and sizes for droplets of seven different inks. That's a difficult problem, and one that's the subject of much research. Here are some considerations in addition to maximizing total gamut size:
--
BasilM
I have just read your post as I am in the middle of revising some notes for a "Digital Photography" course that I give to some of my fellow Seniors. May I have your permission to include, with proper acknowledgement, that section of your post dealing with color rendition in inkjet printers ?
Yours truly,
Basil meddings
Whatever color management you do in Photoshop -- even if you use a custom profile created for your specific printer and paper -- in the end, Photoshop sends RGB data to the Epson driver. The driver has to turn those RGB values into locations and sizes for droplets of seven different inks. That's a difficult problem, and one that's the subject of much research. Here are some considerations in addition to maximizing total gamut size:
- For many of the possible RGB input values, there will be many different combinations of those 7 inks that will produce the same perceived color.
- Some ink combinations may be equivalent only under certain light -- different ink combinations may be equivalent in some light but produce different levels of metamerism
- You don't want sudden changes in ink density for small changes in inputs -- you don't want to switch suddenly from producing a pure red with C + Y to producing it entirely with R. (For one thing this would make metamerism far worse -- something that was a gradient under one light would have a sharp dividing line in another light).
- The RGB input has 16 million possible values. You want the printed output from those inputs to be smooth -- you don't want a huge jump in the color produced when you go from input of 250, 0, 0 to 251, 0, 0.
--
BasilM
Rodger124714
Member
We have just received 200 R1800's and I will post a report on Tuesday (4/5/05) in the picturelineNews.
Rodger at pictureline.com
--
[email protected]
Rodger at pictureline.com
--
[email protected]
Russell Williams
Active member
Yes, you can use my writing, but note that while I know what the conceptual steps are for printing within the print driver, I do not know how they've actually implemented things. For instance, conceptually there are always two color space transforms involved: RGB document space-> RGB printer profile-> 7 color printer space, with the first two color spaces represented by ICC profiles. The first transform can be done by either Photoshop or the driver; the second transform is always done by the driver.
For instance, in the past, some versions of Epson's drivers have performed the first transform (depending on selected Photoshop and driver options) without any actual ICC printer profile appearing in any of the profile files that shipped with the driver.
The R1800 driver has an "Adobe RGB" mode that assumes its input is in that color space. I don't know if they actually do two steps in this case or if they use a separate, one step Adobe RGB-> 7 ink path that alone contains the "new color conversion engine".
For instance, in the past, some versions of Epson's drivers have performed the first transform (depending on selected Photoshop and driver options) without any actual ICC printer profile appearing in any of the profile files that shipped with the driver.
The R1800 driver has an "Adobe RGB" mode that assumes its input is in that color space. I don't know if they actually do two steps in this case or if they use a separate, one step Adobe RGB-> 7 ink path that alone contains the "new color conversion engine".
colourgeek
Veteran Member
Same here, I can't say for sure as I don't work with Epson. Yet you are probably right in theat they are just adding tags to the privte sauce in the profiles to link the media profile to the color space (assumed in this case as ithe intention is the user need not think what conversion should be used.
Before they had this with Epson Print or similar name for a linked color space profile that was linked to the media profile.
Not a bad thing to simplify the workflow. HP have done the same with the driver for the 8750, probably implemented for all Photosmart printers soon after. It's just a work flow simplification in the case of HP. Epson and HP did have some clipping of color space data like it or not when running through any other than no color adjustment. Problem with Epson was no color adjustment then had too much inking to create a good profile, hence the grid points in the shadows had to bend around such small differences that many problems occured result of the non linear inking with NCA.
I can't say for sure but I do think the new motor is using a better assumption of the color space and using a performance boosted processor to correlate the source space data to the maximum gamut.
Sooner or later there will be more and more machines that will use V4.1 ICC spec workflow . At that point intelligent mapping and remapping will provide better subjective color. Perhaps Epson are using some of this technology that was discussed at ICC meeting in Vancouver 2 years ago?
--
Neil Snape photographer Paris http://www.neilsnape.com
Basil Meddings
New member
Thank you very much indeed.
Don't worry about any qualifications !
My problem is to get Seniors, with an average age of about 75, to think enough about the printing process to realise that matching any one of 16 million colours by mixing only 4, 5, 6, 7 or even 8 inks is not going to be simple. Your words will make them realise that very quickly.
Regards,
Basil
--
BasilM
Don't worry about any qualifications !
My problem is to get Seniors, with an average age of about 75, to think enough about the printing process to realise that matching any one of 16 million colours by mixing only 4, 5, 6, 7 or even 8 inks is not going to be simple. Your words will make them realise that very quickly.
Regards,
Basil
--
BasilM
Keyboard shortcuts
- f
- Forum