Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: Geoff Wittig on December 31, 2010, 08:03:54 pm
-
Nice review; all the stuff a potential owner would want to know before pulling the trigger.
I had to chuckle at his photograph of Letchworth State Park; I live about 20 miles away from the park, and have photographed from that exact location many times. I am very familiar with that particular tree in the left foreground! Because of its location right at the edge of a striking vista, you almost have to include it as a framing element.
-
...Because of its location right at the edge of a striking vista, you almost have to include it as a framing element.
Unless you use it as a rappel anchor!
;-P
I also enjoyed the review; thanks Mark. I never knew there was a way to make the bottom margin smaller than 0.56 in an Epson. Does that work with a 7900? I need to check around in the driver...
Dave
-
Excellent review, Mark.
Eric
-
Thanks for the feedback guys and all the best for 2011.
Dave, I think it should work for the 7900 also - it depends on what paper options they have included in the driver. If they do not have an equal border option for the paper sise you use, the alternative is to create a custom paper preset which does have all four borders equal, ad load it into your printing preset. That should work.
-
Mark,
Pretty good review...the one thing you don't mention (as far as I can see) is the real importance of running the head alignment. The Auto alignment does a better job of trying to do a manual alignment. I also note that you stated the manual said to print out with High Speed Off? You sure? It's been my experience that if you have properly aligned the heads, the print quality with High Speed on or off is the same. Obviously, if you have High Speed on, the printer is 2x the speed since it prints in both directions. To be accurate, this is my experience when working with a 79/9900 printer and not the 4900 which I don't have yet. So, this is something you should test on your 4900.
The other thing I would mention is that the head, with 360 nozzles/inch can handle high resolution image detail very, very well. With my P65+ back images, I routinely feed images through Lightroom at high PPI numbers up to 720PPI. Some images with high frequency texture may benefit from actually upsampling in Lightroom to 720PPI and then output sharpening (a huge benefit if printing from Lightroom as you don't need to spawn off different sized image iterations)...this is another thing to test for yourself.
-
Mark,
Pretty good review...the one thing you don't mention (as far as I can see) is the real importance of running the head alignment. The Auto alignment does a better job of trying to do a manual alignment. I also note that you stated the manual said to print out with High Speed Off? You sure? It's been my experience that if you have properly aligned the heads, the print quality with High Speed on or off is the same. Obviously, if you have High Speed on, the printer is 2x the speed since it prints in both directions. To be accurate, this is my experience when working with a 79/9900 printer and not the 4900 which I don't have yet. So, this is something you should test on your 4900.
The other thing I would mention is that the head, with 360 nozzles/inch can handle high resolution image detail very, very well. With my P65+ back images, I routinely feed images through Lightroom at high PPI numbers up to 720PPI. Some images with high frequency texture may benefit from actually upsampling in Lightroom to 720PPI and then output sharpening (a huge benefit if printing from Lightroom as you don't need to spawn off different sized image iterations)...this is another thing to test for yourself.
Hi Jeff, thanks and all the best for 2011. I did not run the auto-alignment, because according to page 125 this is recommended upon seeing banding or graininess in the prints. There is no evidence of any of that. On High Speed Off - yes - I'm sure - at least according to the manual and based on discussion with Epson. Haven't really felt the need to experiment with it, but out of curiosity I shall print the same image both ways one one of these days. I suspect the visual difference to be de minimus. And yes, it handles high frequency detail out of LR or PS very well indeed.
-
I did not run the auto-alignment, because according to page 125 this is recommended upon seeing banding or graininess in the prints. There is no evidence of any of that.
Doesn't matter what the manual says...what I know from experience is that a newly installed (or moved) printer SHOULD have the heads aligned...you seriously DO want to do this regardless of what the manual states (cause the manual is wrong if that's what it says). Do it before you do any serious printing...and I'm not kidding. And once done you really only need to redo it if you notice any micro-banding or once in a while (say 2-3x a year at most).
Actually, I'll have to give the product manager some shyte for not stating in the manual that at least an initial auto head alignment required...
-
Hmmm, there's certainly no harm in doing the alignement, Jeff, I wouldn't say it's a mandatory thing to do - plenty of printers (and from my direct experience last year with basically every model, including several 4900s) gain no benefit.
That said, the 4900's head alignment process is faster than previous models, so I can't really argue against doing it :-)
As for high-speed (bi-di) versus not high-speed (uni-di) I agree 100% with Jeff. Except at draft resolutions, you just won't see the difference and the new LUT, the TFP head and other hardware and firmware improvements really make that comment in the User Guide redundant.
-
Doesn't matter what the manual says...what I know from experience is that a newly installed (or moved) printer SHOULD have the heads aligned...you seriously DO want to do this regardless of what the manual states (cause the manual is wrong if that's what it says). Do it before you do any serious printing...and I'm not kidding. And once done you really only need to redo it if you notice any micro-banding or once in a while (say 2-3x a year at most).
Actually, I'll have to give the product manager some shyte for not stating in the manual that at least an initial auto head alignment required...
Jeff, How "required" it is I guess "depends". So I'll do two things: (a) raise the alignment question with the same product manager - I'm sure we're talking to the same guy, and (b) I'll test the high speed "on" option to see for myself whether their advice is purely academic or makes a practical difference. My previous experience with head alignments on the 3800 (because I have done it on that printer when I had an issue which I thought it may address) has shown zero difference before and after. (The issue I had turned out to be unrelated and got resolved otherwise.) I can imagine that if the printers got seriously banged in transit, which one wouldn't know, a head alignment is good insurance, so I'm not saying you aren't making sense here - only that unless one sees evidence of the need for it, I wonder how important. I have a penchant for leaving well enough alone, and the print quality, detail rendition etc I got out of the box is so good I didn't see the need for it. Anyhow, I hear you and TBD.
-
I can imagine that if the printers got seriously banged in transit, which one wouldn't know, a head alignment is good insurance, so I'm not saying you aren't making sense here - only that unless one sees evidence of the need for it, I wonder how important.
I view doing an auto head alignment as a "best practice" for Epson printers simply because unless you actually DO a head alignment you don't know for a fact they are aligned–even if you don't see visible mis-alignment. Since it takes little time and paper I guess I wonder why you are reluctant to do it and know for sure your heads are aligned?
Look, I have a lot of experience dealing with a lot of Epson printers over the years-some that have been crated up and shipped from location to location (some just sitting in my studio) and I would never print any serious work until I did a head alignment (easy and fast with auto more time consuming and twiddly doing it manually). Often the auto head alignment results in no alignment needed but you don't know that for a fact till after you run it...and an accurate head alignment IS required for running the printer bi-directionally and getting the same quality as uni-directional. Which I think, since it doubles print speed, is useful.
-
...................
Since it takes little time and paper I guess I wonder why you are reluctant to do it and know for sure your heads are aligned?
..................... Often the auto head alignment results in no alignment needed but you don't know that for a fact till after you run it...and an accurate head alignment IS required for running the printer bi-directionally and getting the same quality as uni-directional. Which I think, since it doubles print speed, is useful.
I'm not at all reluctant to do what I think needs to be done, but based on the results I've seen so far, I haven't been the least bit motivated to do it; but since I haven't printed high speed either, the need has been even less apparent. Anyhow, my mind is open on both issues and I'll report back anon.
-
Hi Jeff,
Not to highjack this thread, but can you comment on whether it makes any sense to do the head alignment using one's "standard" media? My 7900 manual says to do it using either the supplied Doubleweight Matte or Premium Luster. But I'm wondering if there might be any logic in doing the alignment with the actual paper that I use most of the time (which for me is either Ilford Gold Fibre Silk or Canson Infinity Baryta Photographique). Does the media play any signifcant factor in the alignment procedure?
Thanks,
Dave
-
Does the media play any signifcant factor in the alignment procedure?
No...
-
I think "Mostly, no". I wouldn't attempt to do it on canvas and I wouldn't attempt to do it on something that was very close to the working limits of the printer (very thick, for example) or even if it required a platten gap adjustment. You want to do all printer adjustments with the printer in "default" setup.
-
Good point...yeah I agree that you shouldn't use a media at the "edge" of the thickness range. Any standard Epson media should be fine like Luster, Matte, etc...
-
Anyhow, my mind is open on both issues and I'll report back anon.
Good...do the Auto head alignment and know for sure...
-
Mark,
Nice review.
Some questions:
Images shown next to one another, one with the wrong profile, the other with the correct one, may not tell correctly what the gamut difference in practice is between a CcMmYKkk and CcMmYOGKkk printer. But those images and the article's gamut shape illustrations confirm my conclusion that the OG inks do a lot for photography as well while there have been messages that the x900 models are mainly there for proofing tasks. What do you think with both printer versions available to you?
There was one reported observation in this forum that the 4900 showed more color inconstancy with changing light ("metamerism") compared to a 7890 printer output. Julian Mussi will probably check that too but maybe you can give an answer already? Given a RIT research report I came across last week it becomes even more unlikely. An Epson supported research to find algorithms for N-color ink mixing (LUTs) that will reduce color inconstancy. One would expect that Epson used that for the entire x900 range and the GS6000.
http://www.art-si.org/PDFs/Printing/CIC03_Chen.pdf
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +190 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
Mark,
I wish to know if you have the model# or manufacture's type of table this is that you show the printer resting on in your review.
(http://www.luminous-landscape.com/articleImages/Epson4900/Fig_3_SP4900-1096.jpg)
Thank you.
-
Mark,
I wish to know if you have the model# or manufacture's type of table this is that you show the printer resting on in your review.
Thank you.
IKEA. "Galant".
-
Mark,
Nice review.
Some questions:
Images shown next to one another, one with the wrong profile, the other with the correct one, may not tell correctly what the gamut difference in practice is between a CcMmYKkk and CcMmYOGKkk printer. But those images and the article's gamut shape illustrations confirm my conclusion that the OG inks do a lot for photography as well while there have been messages that the x900 models are mainly there for proofing tasks. What do you think with both printer versions available to you?
There was one reported observation in this forum that the 4900 showed more color inconstancy with changing light ("metamerism") compared to a 7890 printer output. Julian Mussi will probably check that too but maybe you can give an answer already? Given a RIT research report I came across last week it becomes even more unlikely. An Epson supported research to find algorithms for N-color ink mixing (LUTs) that will reduce color inconstancy. One would expect that Epson used that for the entire x900 range and the GS6000.
http://www.art-si.org/PDFs/Printing/CIC03_Chen.pdf
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +190 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
Hello Ernst. Thanks - glad you enjoyed the review.
I don't do proofing. I use these printers for making final photographs; many professional photographers and service bureaus who print for fine art/commercial photographers use the x900 series printers for final output on inkjet fine art papers. They are meant for this purpose as well as for proofing.
I really haven't been concerned about colour inconstancy since the Epson 4800. It was definitely a big deal with the 2000P - you may recall a short ten years ago or so it was the first in the Epson line of desktop pigment printers promising archival properties, but at the cost of a narrower gamut and very noticeable colour inconstancy. This was dramatically improved but not eliminated with the 4000, and by the 4800 I stopped paying attention to it as an issue. I have no basis to compare with 7890 since I don't own one and never seen its output, but I would expect the behaviour of all the x900s to be very similar. Maybe I'm not the best judge of this, because mentally I expect to see some difference in a print when I am looking at it under warmer versus cooler illumination - I know this is a different matter, but it affects overall perception to the extent, especially with these newer printers, that you really need to look very "analytically" beneath this overall effect to see whether you can detect colour inconstancy. When it gets that good, I think the issue becomes a bit theoretical. I did some interesting B&W stuff last night, so as we speak I looked at the image under real daylight, then took it to an incandescent bulb and for the life of me, I couldn't detect any corrupting of the grayscale, whereas in the days of the 2000P I could clearly see a change from grayish to greenish. No more.
-
There was one reported observation in this forum that the 4900 showed more color inconstancy with changing light ("metamerism") compared to a 7890 printer output.
You mean metameric failure don't ya? :~)
I tend to discount that one observation as an outlier...I've printed with every Epson printer since the 10000 and can tell you that with each round of head and ink improvements, metameric failure is continually reduced. I don't see any metameric failure with my 9900 and I have several viewing booths to use for comparison. I would also note that gloss differential also seems better with the 9900 at least on the papers I print on; EFP, Luster, Semi. GD isn't an issue with fine art paper.
I would also say that Epson has tended to understate the usefulness of O & G for photos...originally designed for proofing to hit spot colors, Epson was actually worried that the added O & G would make it more difficult to profile (they had the same worry back when they released the R800 with red and blue inks). But I don't think the 9900 is any more difficult to produce profiles for than any other printer. The new dithering produced in association with Munsell and the internal separations from input to Cc, Mm, Y, K1 K2 K3, O, G seem very reliable for both photo and proofing as long as you print at 1440 or above (there is an issue when printing out at 720 but I won't get into that here).
-
I saw no issues creating a custom profile for the 4900/Ilford Gold Fibre Silk combination.
-
I have completed further consultation with Epson about the two matters Jeff raised: High speed on versus off, and head alignment.
On the speed issue I am advised: "There is no perceivable visual quality difference between high speed on and off for continuous tone color photography at normal viewing distances. If you are going to get into a dot placement "pixel peeping" accuracy discussion for those looking with a loop then there can be a visible difference .......... where there may be some overshoot/spray of the dot on the return pass. .......if your system is incapable of keeping the printer at full speed (it pauses in between passes) then the recommendation is to shift to uni-directional.......We also use Uni-d on media types that have higher ink loads and longer dry down periods. Again this is mostly proofing applications with Clear and Metallic films. ...............High Speed On and Off is a tool to be used as needed depending on the circumstances and or personal preference."
On the alignment issue: "If you are going to do your own profiling then it is best to start from a known reference and run a head alignment."
Thanks for raising these matters Jeff. I think this is now sufficiently clear on both counts and your advice is consistent.
-
You mean metameric failure don't ya? :~)
No. For that you need two patches made with different colorants and they shouldn't match under a light source. I didn't mean a metameric match either as that is about two patches with different colorants that make a match under a light source. I did mean color inconstancy under changing lighting and added ("metamerism") as that is the term normally but incorrectly used in forums like this one. I do so because sometimes there are readers that want to correct the incorrect use of ("metamerism") but I add ("metamerism") to get the message across to people that are less strict on terms but still know what it more or less means in practice. But your response means I should take yet another route to get there :-)
Yes, that single observation isn't supported by other observations so far. The fact that the prints were created on an Epson booth with Epson people at hand and the writer very sure about the conditions the prints were made under and his decission to go for a 7890 instead of a 4900 for that reason gave it some impact. Still curious what actually happened.
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/
-
I have just finished reading this extensive review. I have a 3800 but need to move to a roll capability printer so I am considering the 4900. Mark's excellent review has given me a lot to think about and has answered a lot of questions. Great review .. thanks a lot
-
No. For that you need two patches made with different colorants and they shouldn't match under a light source. I didn't mean a metameric match either as that is about two patches with different colorants that make a match under a light source. I did mean color inconstancy under changing lighting and added ("metamerism") as that is the term normally but incorrectly used in forums like this one. I do so because sometimes there are readers that want to correct the incorrect use of ("metamerism") but I add ("metamerism") to get the message across to people that are less strict on terms but still know what it more or less means in practice. But your response means I should take yet another route to get there :-)
Yes, that single observation isn't supported by other observations so far. The fact that the prints were created on an Epson booth with Epson people at hand and the writer very sure about the conditions the prints were made under and his decission to go for a 7890 instead of a 4900 for that reason gave it some impact. Still curious what actually happened.
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/
From all I've read on this distinction we are indeed discussing "color inconstancy". That was the issue going back to the 2000P and that is what one examines for ever since: the same colour changing appearance under different illuminants. I'd also be curious to know how that apparent outlier result occurred. Any way to find out?
-
From all I've read on this distinction we are indeed discussing "color inconstancy". That was the issue going back to the 2000P and that is what one examines for ever since: the same colour changing appearance under different illuminants. I'd also be curious to know how that apparent outlier result occurred. Any way to find out?
The original thread was wiped by accident but I tried to pin down in that thread possible faults in the printmaking process but none could be found. Same paper was used etc.
Some miracles shouldn't be explained.
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +190 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
I have just finished reading this extensive review. I have a 3800 but need to move to a roll capability printer so I am considering the 4900. Mark's excellent review has given me a lot to think about and has answered a lot of questions. Great review .. thanks a lot
You are welcome - glad you found it useful.
-
I'll be back in the office next week and can print on both a 4900 and a 7890 and a 9900. It'll take me a couple of days to catch up to have time.
Does anyone have a suggested image for testing this issue? I'd be happy to mail out the results to someone like Ernst for analysis.
-
Phil, this should be very interesting if you can manage to do it. I think any image which has a good representative tonal scale when converted to B&W in as neutral a manner possible would probably be ideal. To normalize the test conditions the profiles for all the printers/papers (I would suggest one matte and one gloss) should be done with the same spectro and software, and of course you would be using the same papers in all printers, and in each case the Epson driver with as similar settings as possible between the printers. With as much as possible normalized, only the printer vintage and inkset should be the determinative factors of any differing outcomes.
-
Gloss and matte is a good idea and I'll do that. I intend to use Epson media with Epson canned profiles and appropriate driver settings, using a standard colour managed workflow. I think that should minimise errors and variations.
Perhaps I should do both colour and B&W both with and without ABW. 3 printers, 2 media, three outputs = 18 prints.
I'll get back to this middle of next week, once I'm caught up with work, which will no doubt be overflowing upon my return!
-
A gloss paper with a wide gamut and an Atkinson evaluation image (including the B&W greyscale range) printed in color mode should be a worst case scenario. Epson paper and profiles. If possible a paper with little FBA to exclude that effect on color constancy and a paper with more FBA. 6 prints only needed. If that total doesn't show problems then there are no problems. Don't look for an absolute match under 5000-5500 K as there will be no match (at least not between the 7890 and x900 prints), it is about unexpected shifts per print when lighting is changed. You will notice the differences if there are differences. I observed it between the Z3100 prints made with the HP driver (good) and made with the Wasatch SoftRip, same inkset, same paper, different media presets and profiles.
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +190 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
Mark,
can you tell me about the back of the 4900? I have a 4880 and I print on canvas. With my 4880 I can put through the back of the printer a 18.25" piece of canvas and print a 16x20 with about a 1" bleed of canvas. (Of course I have to take off the roller attachment to do this) A 1" bleed of canvas is sufficient for me to stretch the canvas on a stretcher bar. This is great when I only need to print one 16x20. I only print 16x20's on my 9900 when I need pairs. My local Epson dealer doesn't have a display model. I would opt for one of these if I knew I could do the same with the 4900.
Thanks,
Gar
Pittsburgh, PA
-
Thank you for this excellent review. It is a great service to 3800 owners like myself who have been considering the move. Personally, your review has saved me a lot of money...well, not really. It has just redirected it to a new lens instead of a new printer.
Thanks,
Carl
-
Mark,
can you tell me about the back of the 4900? I have a 4880 and I print on canvas. With my 4880 I can put through the back of the printer a 18.25" piece of canvas and print a 16x20 with about a 1" bleed of canvas. (Of course I have to take off the roller attachment to do this) A 1" bleed of canvas is sufficient for me to stretch the canvas on a stretcher bar. This is great when I only need to print one 16x20. I only print 16x20's on my 9900 when I need pairs. My local Epson dealer doesn't have a display model. I would opt for one of these if I knew I could do the same with the 4900.
Thanks,
Gar
Pittsburgh, PA
With the Epson 4900 you do not remove the roll-holder. For thick media you use the front manual feeder, making sure there is some room behind the printer for the paper to adjust to the print position. The print will come out from the front resting on the tray. All very convenient.
-
A gloss paper with a wide gamut and an Atkinson evaluation image (including the B&W greyscale range) printed in color mode should be a worst case scenario. Epson paper and profiles. If possible a paper with little FBA to exclude that effect on color constancy and a paper with more FBA. 6 prints only needed. If that total doesn't show problems then there are no problems. Don't look for an absolute match under 5000-5500 K as there will be no match (at least not between the 7890 and x900 prints), it is about unexpected shifts per print when lighting is changed. You will notice the differences if there are differences. I observed it between the Z3100 prints made with the HP driver (good) and made with the Wasatch SoftRip, same inkset, same paper, different media presets and profiles.
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +190 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
Yes, I agree - the Atkinson printer test page would be an excellent test object for this exercise. It has everything one needs and then some. The one option I would perhaps recommend however would be to also select one matte paper, to test for whether the paper surface makes a difference to colour constancy appearance under different illuminants. Ernst, if you know for sure from your previous work that it doesn't, then this suggestion could be skipped, saving time and paper.
-
Thanks Ernst and Mark for your continued suggestions for this.
-
Mark, thank you for your very thorough review. Sounds like an excellent printer - I've been waiting quite awhile for this, patiently nursing my ancient SP4000 along...
Sorry to come out of left field here - don't want to hijack the thread, but had to ask:
You mentioned that you've successfully used the old Pulse ColorElite system with OS X 10.6.4. This was among the many things I lost when I upgraded my old Dual G5 to a 2009 quad-core 'Nehalem' machine. Running 10.6.5, the install disk won't run.
How did you get ColorElite installed on your system? Is your Mac hardware comparable to mine (I'm wondering if it's a h/w issue instead of Snow Leopard)?
Thanks again, & best New Year wishes.
John Lund
JWL Images
-
Mark, thank you for your very thorough review. Sounds like an excellent printer - I've been waiting quite awhile for this, patiently nursing my ancient SP4000 along...
Sorry to come out of left field here - don't want to hijack the thread, but had to ask:
You mentioned that you've successfully used the old Pulse ColorElite system with OS X 10.6.4. This was among the many things I lost when I upgraded my old Dual G5 to a 2009 quad-core 'Nehalem' machine. Running 10.6.5, the install disk won't run.
How did you get ColorElite installed on your system? Is your Mac hardware comparable to mine (I'm wondering if it's a h/w issue instead of Snow Leopard)?
Thanks again, & best New Year wishes.
John Lund
JWL Images
XRite doesn't support this and they were very surprised when I told them I got it to work. Seems as if it is hit and miss. I am still on OSX 10.6.4. - haven't installed the update to 10.6.5. I'm using a Mac Pro with a pair of 6-core Intel Xeon processors with hyperthreading, 2.66GHz. I inserted the program disc into the tray and just installed it in the usual way one installs programs on a Mac. No special manoeuvres, and so far so good.
-
XRite doesn't support this and they were very surprised when I told them I got it to work...
-- Hmmm, well, who knows about some of these things. Thanks for the response, tho.
Maybe I'll try a network install from the Dual G5 machine.
Regards,
John
-
Hi Mark or anyone else willing to help,
I enjoyed reading your review of the new 4900. I am an amatuer and new to this forum. I have had an epson 4800 since 2005 and have recently moved to the new 4900.
I would like to give the Illford Gold Fiber Silk paper you like so much a try with my new 4900. I used Epson's Premium Luster almost exclusively in the past with the 4800.
Ilford has not yet made IGFS ICC profiles available for the 4900. I do not have the capability or no how needed to create my own profile, as you have done. Can I use Ilford's Gold Fiber Silk ICC profiles for the 7900 and 9900 printers in the 4900? If not, they claim Epson's own built in Epson glossy paper ICC profiles may work. Have any of you found that they work?
-
Please send me a private message through the messaging facility of this Forum giving me your own direct email address and I shall help you with this.
-
Right, I am now getting close to being up to date enough to run these tests. I'm going to use Epson Luster and Epson Archival Matte across a 7890, a 9900 and a 4900. Hope to have it completed Monday and will then report back.
-
Right, I am now getting close to being up to date enough to run these tests. I'm going to use Epson Luster and Epson Archival Matte across a 7890, a 9900 and a 4900. Hope to have it completed Monday and will then report back.
I am curious to your findings, Phil.
I was the one reporting here my observations after having printed some of my images at an Epson event on both a 4900 and the 7890. Prints done on the same paper with the very same settings in the driver. What we saw with prints coming from the 4900 was color inconsistency under changing light conditions (I then called that metarism, I now know that that is not correct) . This effect was quite noticeable. That surprised me a lot and something I had not expected at all.
The prints from the 7890 also looked more photographic, something hard to put in words. In the thread that was mistakenly deleted by the OP, I suggested that anybody interested in these printers have a look for themselves by printing a couple of their images to find out whether they got the same result as we had.
So I am looking forward to your results !
Might be that this exemplar of the 4900 was out of spec.
Cheers,
Remko
-
Well, I've completed my tests. 4900, 7890 and 9900. I'm not seeing any issue or any significant differences between the machines. There are some differences between the 10 and 8 colour machines, but you have to look for it and really it's helped by having them side by side.
Mark - if you want to PM me an address, I'd be happy to send a set of samples through as I think it would be useful to have you review and comment - you likely will have a more discerning eye than me and on top of your review of the 4900 it may be useful.
-
Well, I've completed my tests. 4900, 7890 and 9900. I'm not seeing any issue or any significant differences between the machines. There are some differences between the 10 and 8 colour machines, but you have to look for it and really it's helped by having them side by side.
Mark - if you want to PM me an address, I'd be happy to send a set of samples through as I think it would be useful to have you review and comment - you likely will have a more discerning eye than me and on top of your review of the 4900 it may be useful.
Hi Phil, that's really good of you to offer and thank you very much. I'll be in touch. I expect to see what you saw, nonetheless it is often good for at least two pairs of eyes to look at stuff like this, especially when the differences don't hit you in the face.
-
Fantastic review, Mark. Thanks for putting in the time and detail on that. Just seeing the size alone of this beast gives me some pause, given my small office. 100+ lbs nothing to sneeze at either. OTOH, the improved color reproduction and other improvements over prior models are tempting so I'll have to mull this investment over a bit longer, but as usual LL reviews trump all others. Great job.
-
Thanks Dan, glad you found it useful. And yes, the beast kind of crowds my office too!
-
Well, I've completed my tests. 4900, 7890 and 9900. I'm not seeing any issue or any significant differences between the machines. There are some differences between the 10 and 8 colour machines, but you have to look for it and really it's helped by having them side by side.
That is very interesting, Phil, especially since you ran your test on the same paper as I printed my photos on: Epson's Premium Luster.
How did you find the consistency of the Luster 260 icc profile among the 4900 and the 7890 ? We found the print coming from the 4900 to be a bit too redish and having more contrast. The latter gave the impression of a bit sharper print, but on close inspection that was due to the difference in contrast.
Cheers,
Remko
-
Hi Remko,
I found them to be very similar and I didn't notice a particular cast from one to the other. There are differences, particularly when compared side by side or under a loupe looking at a solid colour.
Susnset orange is, not surprisingly, improved with the 10 colour ink set, but you do need to be looking closely. There's a similar advantage when looking at greens. Again, not surprising.
Interestingly, on my prints here I see slightly more magenta in the skin tones from the 7890 than the 4900. Neither are bad and the difference is very small. I've sent a set of prints to Mark for him to review and comment, so it will be interesting to see what he sees, too.
B&W (desaturated RGB) I'm seeing slightly darker and more contrast on the 4900 in the aforemention skin tones.
The match from the 9900 to the 4900 is extremely close (as you'd expect).
I'm keen to hear Mark's report once he receives them and has some spare time.
-
Hi Remko,
I found them to be very similar and I didn't notice a particular cast from one to the other. There are differences, particularly when compared side by side or under a loupe looking at a solid colour.
Susnset orange is, not surprisingly, improved with the 10 colour ink set, but you do need to be looking closely. There's a similar advantage when looking at greens. Again, not surprising.
Interestingly, on my prints here I see slightly more magenta in the skin tones from the 7890 than the 4900. Neither are bad and the difference is very small. I've sent a set of prints to Mark for him to review and comment, so it will be interesting to see what he sees, too.
B&W (desaturated RGB) I'm seeing slightly darker and more contrast on the 4900 in the aforemention skin tones.
The match from the 9900 to the 4900 is extremely close (as you'd expect).
I'm keen to hear Mark's report once he receives them and has some spare time.
Looking forward to seeing them Phil. They've got a good distance to travel, so it will likely be a while.
-
Courier picked them up yesterday, so with any luck you'll have them early next week.
-
Hi Remko,
I found them to be very similar and I didn't notice a particular cast from one to the other. There are differences, particularly when compared side by side or under a loupe looking at a solid colour.
Susnset orange is, not surprisingly, improved with the 10 colour ink set, but you do need to be looking closely. There's a similar advantage when looking at greens. Again, not surprising.
Interestingly, on my prints here I see slightly more magenta in the skin tones from the 7890 than the 4900. Neither are bad and the difference is very small. I've sent a set of prints to Mark for him to review and comment, so it will be interesting to see what he sees, too.
B&W (desaturated RGB) I'm seeing slightly darker and more contrast on the 4900 in the aforemention skin tones.
The match from the 9900 to the 4900 is extremely close (as you'd expect).
I'm keen to hear Mark's report once he receives them and has some spare time.
Hi Phil,
Thank you for the test and the feedback, I appreciate it :)
I now do think that the 4900 on that event was a bit out of spec, as we found that the greens on a forest landscape print were better (more differentiated) on the 7890 print - which was a surprise of course. That is when comparing them side by side.
The skin tones on the 7890 print looked fine, if anything than maybe slighty a bit too yellow on very close inspection. Go figure.
On the 4900 the skin tones were definitely a bit too redish.
cheers,
Remko
-
One may wonder though what caused the different ink lay down of the first Imageprint 8 4900 driver. Did they simply copy a 7900 configuration that didn't behave the same on the 4900 or was it something else? Colorbyte's comment was "minor differences between the 4900 and 7900/9900 printers which causes slight overinking in the dark areas." Epson will enhance image quality and other aspects if they get a chance and there has been quite some time since the 7900/9900 were introduced.
That said >>>"metamerism"<<< seems to be at the same low level compared to the 9900, something we did expect. Thank you Phil.
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +230 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
Thank you for the long review, Mark. I feel much better about being "stuck" with my 3800 for the foreseeable future.... :)
-
Sorry to say it, but I tried a 4900 and the pizza wheel marks are unacceptable (to me) on Museo Silver Rag and H Photo Rag Baryta. Scratches are clearly discernible when viewed with strong light at an angle to the paper. The pizza wheels come down six inches from the end of the paper (to eject it) and create columns of marks. The only printers that don't have this problem are the 7xxx and 9xxx Epsons (they have other sheet feeder problems though) and the 4xxx series if you retract the pizza wheel plate manually. Canons are unacceptable also.
It appears that the 4880 may be the best printer out there for sheet feeding these thicker glossy papers (if you retract the pizza wheels manually).
The 4900 would make a great matte printer though and perhaps it would work fine for thinner weight glossy papers.
-
Different papers will see different issues. 260gsm isn't light, but it's not super heavy - the Epson Luster doesn't suffer from pizza-marks.
What issues are you seeing with 7900/9900 sheet feeding (other than it being only 1 sheet at a time)?
Also, how are you feeding the 4900? Rear or front?
-
The 4900 ejects the paper using the pizza wheels if the sheets are loaded from the top or from the cassette. 7xxx and 9xxx printers have a difficult time (often misfeeding) when trying to load the above mentioned papers if they are in sheet form. Other thinner papers should be fine.
-
The 4900 doesn't drive the paper with the pizza wheels. I was trying to obtain information from you to see if the issue was more or less prevelant based on the paper feed method with those papers.
As for the 7900/9900 - I have to say, I love the sheet feed compared to previous models - I certainly don't see how you can have paper feed issues during load. Can you detail the problem?
-
"Catchall" - I haven't seen this problem withthe top feed, the roll feed, or the cassette feed, using papers with caliper up to 315 micron. I have not tried much heavier papers. For those the manual specifies to use the front feed. Have you tried adjusting the platen gap and the paper thickness settings in the Epson driver? For thicker papers you may need a paper setting of 4 and a "wide platen gap, or depending on thickness and weight of paper a bit more of each.
-
I have a 3880 and have not noticed any pizza wheel marks on either Museo Silver Rag or Ilford Gold Fibre Silk. As Mark noted, it's important to get the paper settings right and a paper setting of 4 and wide platen gap seem to work well for these two papers.
-
The pizza wheels only come down at the end of printing a sheet but they engage the surface of the print about 6 inches from the end of the paper. On my paper of choice, H Photo Rag Baryta, they are clearly visible in the printed area (with bright light at a severe angle), thus marring the print. They seem to be there to eject the paper. There are two separate sequences of these pizza or star wheels. The first seems to dig into the print to move it into place for the second set of pizza wheels that move the paper along until its completely ejected from the machine. This is different from the 4xxx series printers where there is only one set of pizza wheels (but two rows of them). You can see this happening as you look into the printer’s paper path from the front. After it ejects the paper, it then retracts the plate back out of the way so that the pizza wheels don't affect the entire sheet. This happens on every sheet.
As for the H Photo Rag Baryta sheet handling problems on the larger series printers, my only experience is in working with a local professional lab. They've had problems with thicker papers feeding using the sheet feeder. The platen was set at the "wider" setting. Some sheets would load fine; others would not feed at all unless pushed a bit from the top. Neither of us could come up with a procedure or setting that solved this problem other than using the push method. There may also be variances in the manufacturing of each machine, particularly if others aren't experiencing this problem with H Photo Rag Baryta. Thinner papers work fine.
-
Hi Mark, Thank you for your very interesting and helpful review of the 4900.
This may be a characteristic only of the soft surface of certain thick glossy papers (in my case H Photo Rag Baryta). These marks are clearly there when looking at the print using strong light at a severe angle with the platen gap at "wide" and the thickness adjusted for the wider paper. They are only there for the last 6 inches of the paper and are more visible if the print has broad areas of black. The Epson dealer and I tried thickness settings of 3, 4, 5 and 6. Still no luck.
-
On my 3800, marks are there on both H. Photo Rag Baryta and Silver Rag. I confirmed this with prints made on other 3800s owned by friends. There are no marks on matte papers or certain other thinner glossy papers. The platen gap is set to wide, paper thickness is tried at 3,4,5 and 6. Also experimented with longer drying times between passes of the printer head (that's excruciating). Still no luck. A bright, angled light source is needed in order to see these. It's particularly noticeable if the print has large areas of black in them.
-
Mark, Was hoping to use the cassette as I can with my 4800. Does the front loading using the pizza wheels?
-
I've seen the pizza wheel marks on Silver Rag going through my 4880. It doesn't bother me though, you need extreme lighting conditions to see them, but they are there. I get printing problems with a wider platen gap though. At first I thought they were paper faults, but after going back to a normal platen gap, the tiny white marks (where no ink had been laid) that occasionally appeared on a wide platen setting, - disappeared. Still using thickness 4.
-
Interesting! On dark photos with lots of black it's more of a problem. I assume you can manually retract the 4880 pizza wheel plate (by pushing the plate in and up) just like you can on the 4800 and that the plate stays up for the entire print job (unlike the 4900 which retracts and engages for each print). Can you confirm that? Thanks!
-
I assume you can manually retract the 4880 pizza wheel plate (by pushing the plate in and up) just like you can on the 4800 and that the plate stays up for the entire print job (unlike the 4900 which retracts and engages for each print). Can you confirm that? Thanks!
I'm not sure if you can manually retract it, at least I haven't tried. I'll be printing in the next day or so, and I'll have a look.
-
(with bright light at a severe angle)
Honestly, how often are your prints viewed in such conditions?
I'd recommend that you try Mark's suggestion of the front feed (because it seems, and perhaps I'm misreading this, that you've only tried rear feed?).
With regard to 7900 and 9900, it sounds like the correct sheet feed method isn't being used. Press the paper change button and wait for the printer to be ready (ensure you've also changed the paper mode to sheet), then feed the paper down until it stops. Then press the down arrow once and the paper will load. The only thing I've ever had a slight issue with was actual board - much heavier and thicker than the paper you're using.
-
I realize that this is usually not a problem for most people, particularly when prints are hung on a wall, framed behind glass. But these are portfolio prints where the purchaser may handhold the print under different lighting conditions. Thus having a machine that scratches prints is not acceptable to me. But that's my personal preference.
Re: loading technique for 7900 and 9900, that's exactly how we loaded them. Again, this is not a problem for most papers.
-
Mark, Was hoping to use the cassette as I can with my 4800. Does the front loading using the pizza wheels?
Front load is the only one I hadn't tested, so I can't advise how it works. I suggest you send an email to Epson tech support with that quetion. They are very responsive, from my experience. I would NOT use the front feed anyhow unless your paper is of the thicknesses for which they recommend that feed only in the manual. I would NOT use the cassette for anything heavier than Epson Premium Luster, which is not all that thick. For the remainder I recommend the top feed, paying attention to the paper thickness and platen gap settings.
-
Hi Mark,
What is the maximum width of canvas or paper that the top of the 4900 cant take? On my 4880 I can put a 18.25 width of canvas through the back of the printer. You mentioned that you can't take the roller assembly off the 4900. Loading canvas from the top, can the 4900 take a width wider than 17"? If you have time could you measure the maximum width that the printer can take.
Thanks,
Gar
-
Hi Mark,
What is the maximum width of canvas or paper that the top of the 4900 cant take? On my 4880 I can put a 18.25 width of canvas through the back of the printer. You mentioned that you can't take the roller assembly off the 4900. Loading canvas from the top, can the 4900 take a width wider than 17"? If you have time could you measure the maximum width that the printer can take.
Thanks,
Gar
Gar, I'm not in a position to do this until the coming weekend, but in principle as you know it is specified for 17 inch width, so anything more you try to stick in there would be "at your own risk". Best thing to do may be to send an email to Epson tech support and ask them about it. They are usually prompt and helpful.
-
What is the maximum width of canvas or paper that the top of the 4900 cant take? On my 4880 I can put a 18.25 width of canvas through the back of the printer. You mentioned that you can't take the roller assembly off the 4900. Loading canvas from the top, can the 4900 take a width wider than 17"? If you have time could you measure the maximum width that the printer can take.
I just pulled through am 18.25" wide piece of paper (I cut it - I don't have media that wide normally) and it JUST fits. The area of concern is actually the exit path - the paper was just touching either side as it came out, so that might be a concern. I couldn't guarantee it would work, but it would be close. The roll holder itself won't take media that wide - it's limited at about 17.5" or just under.
Best bet would be to find a demo unit somewhere and test it.
As Mark says, definitely "at your risk" and may have warranty concerns.