Qualitatitve differences between 9800 & 9900

[BobDavid]

Aside from the addition of orange and green inks, and the flexibility of switching between MK and PK, how does the output of the 9900 compare to the 9800?  What are the real world differences in terms of output? Is there a noticeable difference in gamut, tonality, shading, etc?  I'd love to hear from people that are familiar with both generations of Epson printers. I prefer not to hear from vendors, but from actual users that depend on these machines for income.

[Wayne Fox]

Aside from the addition of orange and green inks, and the flexibility of switching between MK and PK, how does the output of the 9900 compare to the 9800?  What are the real world differences in terms of output? Is there a noticeable difference in gamut, tonality, shading, etc?  I'd love to hear from people that are familiar with both generations of Epson printers. I prefer not to hear from vendors, but from actual users that depend on these machines for income.

Difference between those two printers is visually apparent most of the time as long as you haven't limited your gamut by your workflow's working space.  In some colors and tones the difference is apparent in all of the areas you mentioned, especially if you have very good files. The 9800 is now 2 generations behind the 9900 in head technology, inkset, and screening technology.  Side by side comparisons of most images will reveal improvements.

The difference between the 9880 and 9900 is not as apparent, but the 9900 quality is still visually better some of the time when comparing these two printers.

[MHMG]

The 9800 is now 2 generations behind the 9900 in head technology, inkset, and screening technology.  Side by side comparisons of most images will reveal improvements.

Well, Epson staff (or a trusted company on Epson's behalf) prepared an obviously subject-chosen colorful image of an anorexic looking caucasian model in vivid floral dress for the Photo Plus East show last year using K3, K3VM, and HDR inksets in identical size. The prints were displayed side-by-side and well-lighted in order to demonstrate the differences in the three inksets and their respective printer/screening patterns. IMHO, each print had certain aspects to it that were in a very subtle way better than the other two prints. In other words, trying to precisely control a single colorful image (e.g., the model's dress had lots of vivid magenta and green floral color patterns in it) plus skin tone reproduction from highlight through shadows is no small feat. Each printer/driver/profile combination seemed to accel in certain image areas. That said, when members of my AaI&A digital print research program send me samples of my 12 hues_for_sRGB colorspace image target for light fade testing, it is pretty easy to see that the 7900/9900 printers are edging out the other Epson printers in subtle yet demonstrable ways. Whether once can coax this subtle additional color gamut and dot screening precision into one's real prints is another matter. The bottom line is that in day-to-day printing a 7800, 7880 or 7900 can all do outstanding work in the hands of an accomplished printmaker, but over the long haul an accomplished printmaker is probably going to prefer the 7900/HDR system.

Cheers,
Mark
http:www.aardenburg-imaging.

[Schewe]

Well, Epson staff (or a trusted company on Epson's behalf) prepared an obviously subject-chosen colorful image of an anorexic looking caucasian model in vivid floral dress for the Photo Plus East show last year using K3, K3VM, and HDR inksets in identical size.

Uh, that would be a shot made by Douglas Dubler (see: http://douglasdubler3.com) so I don't think it would be fair to call him "Epson staff" ya know? The fact that Epson chose that image may be telling...but Douglas is a pretty good shooter and print on his own.

I prefer not to hear from vendors, but from actual users that depend on these machines for income.

What does that mean, "Vendors"? You think a bunch of people are laying in wait to pounce on you and sell you a line of goods?

You don't want to hear from the likes of Michael Reichmann or maybe myself? I had a 9800, then a 7880 and now a 7900 but apparently since I have a "relationship" with Epson, you don't want to hear from me? Am I a "vendor"? Hum, should I be insulted?

Oookay...good luck finding out bud.

[MHMG]

Uh, that would be a shot made by Douglas Dubler (see: http://douglasdubler3.com) so I don't think it would be fair to call him "Epson staff" ya know? The fact that Epson chose that image may be telling...but Douglas is a pretty good shooter and print on his own.




What does that mean, "Vendors"? You think a bunch of people are laying in wait to pounce on you and sell you a line of goods?

You don't want to hear from the likes of Michael Reichmann or maybe myself? I had a 9800, then a 7880 and now a 7900 but apparently since I have a "relationship" with Epson, you don't want to hear from me? Am I a "vendor"? Hum, should I be insulted?

Oookay...good luck finding out bud.

With all due respect Shewe, you are being just a little too sensitive here. I meant no disrespect. I simply meant it was a deliberately hand picked image to show off the best quality of the respective systems. I would expect nothing less from any good marketing effort.  Nothing more, nothing less.

Kind regards,

Mark
http://www.aardenburg-imaging.com

[Schewe]

With all due respect Shewe, you are being just a little too sensitive here.

The name is Schewe...

And if you didn't find out who did the shot, it would be presumptuous to presume that "Epson staff (or a trusted company on Epson's behalf)" did anything other than select the image. Pretty sure Douglas even did the prints (he's kinda anal about that stuff). Also pretty sure the shot was not done on behalf of Epson but a result of a job, although I would need to check with Douglas about that–shall I?

[MHMG]

The name is Schewe...

And if you didn't find out who did the shot, it would be presumptuous to presume that "Epson staff (or a trusted company on Epson's behalf)" did anything other than select the image. Pretty sure Douglas even did the prints (he's kinda anal about that stuff). Also pretty sure the shot was not done on behalf of Epson but a result of a job, although I would need to check with Douglas about that–shall I?

My sincere apologies for misspelling your name. Forum participation tends to bring out my typing and spelling inadequacies.  

The three image were prepared on behalf of Epson. Epson staff, Epson management, some Epson employee who knows the photographer, etc were involved in this selection. Let's not get lost in the details. The three prints, side-by-side were meant to show image quality differences between the three systems. A very tough assignment for the Epson marketing people in charge with procuring these prints for the trade show (the prints didn't show up by accident). On the one hand Epson couldn't produce images that made the older systems look bad. On the other hand, Epson marketing folks were trying to promote the virtues of the new technology. Geez... let's leave it at that!!! I applaud Epson's efforts, but what I saw on display (and yes, it's just one person's opinion whose been involved in image quality assessments for over three decades) is that the prints were very close in overall image quality indeed.

My apologies if you are offended. I meant no such offense.

Kind regards,
Mark
http://www.aardenburg-imaging.com

[Schewe]

...Epson marketing folks were trying to promote the virtues of the new technology. Geez... let's leave it at that!!!

That's fine...usually I get to see the stuff in the Epson booth at PPE but last year I didn't go due to book deadlines, so I only know about the HDR comparison second hand (I've seen the shot printed on a 7900 but not on the other ink sets). But I just wanted to get across the fact that Epson works with top photographers and doesn't try to go out and produce their own images–they don't have the talent for that, they stick with making printers...

As far as the HDR inks sept and the x900 class printers, obviously the gamut is larger and the printers can hit colors due to the 10 colors that earlier Epson printers can't. Whether that matters to an individual's own images is tough to say. But the other thing many people overlook is that unless you have Pro Photo RGB images, the odds are you won't see much difference in the HDR inks because, well, Adobe RGB and sRGB are simply too small. The HDR ink set goes well beyond Adobe RGB in some portions of the color gamut.

Also add the fact that there are 360 nozzles/inch and you have not only a substantially faster printer but one that can produce detail that even the x880 series can't match. Of course, you need to have the real resolution to be able to take advantage of that.

But it's really not too clear exactly what the OP wants to hear...and whether Mike nor myself would be considered "vendors", since he states he doesn't want to hear from them.

[BernardLanguillier]

As far as the HDR inks sept and the x900 class printers, obviously the gamut is larger and the printers can hit colors due to the 10 colors that earlier Epson printers can't. Whether that matters to an individual's own images is tough to say. But the other thing many people overlook is that unless you have Pro Photo RGB images, the odds are you won't see much difference in the HDR inks because, well, Adobe RGB and sRGB are simply too small. The HDR ink set goes well beyond Adobe RGB in some portions of the color gamut.

Just printed on my new 9900 some Angkor moss images on Photorag Baryta (using the Hanhemhule supplied profiles), and was very impressed by the intensity of the greens, they have the near fluorescent quality that I recall the actual scene had. Very impressive.

I didn't do any direct comparisons with older Epsons since my 4000 is already gone (and was using an old inkset anyway), but I don't think that it could get much better than this. I might try to build my own profiles to see if I can gain a bit more, but I don't anticipate a major gain, it seems that the supplied profiles are very good.

Cheers,
Bernard

[BobDavid]

Uh, that would be a shot made by Douglas Dubler (see: http://douglasdubler3.com) so I don't think it would be fair to call him "Epson staff" ya know? The fact that Epson chose that image may be telling...but Douglas is a pretty good shooter and print on his own.




What does that mean, "Vendors"? You think a bunch of people are laying in wait to pounce on you and sell you a line of goods?

You don't want to hear from the likes of Michael Reichmann or maybe myself? I had a 9800, then a 7880 and now a 7900 but apparently since I have a "relationship" with Epson, you don't want to hear from me? Am I a "vendor"? Hum, should I be insulted?

Oookay...good luck finding out bud.

Four different vendors contacted me after I wrote a post about my Epson's getting fried. Two of them used pressure tactics. I suggest you look up the word "vendor" in the dictionary. I don't think Michael earns his living selling Epson printers and P1 cameras. He's a guy that runs a blog and conducts workshops. I'd classify him as a photography pundit, not a VAR.  

I don't think you should be insulted.  You just need to be more careful with your thought process.

[Gemmtech]

Uh, that would be a shot made by Douglas Dubler (see: http://douglasdubler3.com) so I don't think it would be fair to call him "Epson staff" ya know? The fact that Epson chose that image may be telling...but Douglas is a pretty good shooter and print on his own.




What does that mean, "Vendors"? You think a bunch of people are laying in wait to pounce on you and sell you a line of goods?

You don't want to hear from the likes of Michael Reichmann or maybe myself? I had a 9800, then a 7880 and now a 7900 but apparently since I have a "relationship" with Epson, you don't want to hear from me? Am I a "vendor"? Hum, should I be insulted?

Oookay...good luck finding out bud.

Jeff, Please don't misinterpret this post, I truly appreciate your level of expertize.  you need to relax, you are going to get arrested by a cardiac     Vendor means what it means (http://dictionary.reference.com/browse/vendor) and yes there are vendors that hang out here (Neither you or MR are vendors) trying to sell some goods (I don't have a problem with it) You have a lot to offer people who want to learn the art of photography/printing and it's nice that you do so here, I thoroughly enjoyed your FCTP video with MR.  Sometimes when we become very proficient and well known in our vocations we can become pompous and arrogant.  Lighten up, I'm being very impartial here and I believe you misinterpreted what the OP wrote.  My father was a mathematical genius and engineer; one of his greatest traits was his ability to teach, he could teach the most complicated math to anybody and never have an attitude.  Selflessly share your gifts with others.

[jukeboxprint]

Hi Bob, we have a large number of the 9900's and personally think they are the best machines on the market!!!  Compared to the 9800 we noticed the 9900 significantly faster,  HUGE improvement over color.. less gradation and images come out rich and vibrant (big difference compared to the 9800 - especially green and orange colours).  Best of all replacing rolls takes no time compared to the 9800. No more breaking your back trying to replace a 44" roll.

One thing also, we had 2 - 9900's running for a SOLID 1.5 months (night and day) and did not encounter any problems mechanically or loss of image quality. Registration was always bang on from start to finish.  This is something the 9800 would of never accomplished!  If  your looking for speed, reliability, exact color re-production and bang for the buck this is truly the machine for you!


TJ
Juke Box

[Wayne Fox]

Well, Epson staff (or a trusted company on Epson's behalf) prepared an obviously subject-chosen colorful image of an anorexic looking caucasian model in vivid floral dress for the Photo Plus East show last year using K3, K3VM, and HDR inksets in identical size. The prints were displayed side-by-side and well-lighted in order to demonstrate the differences in the three inksets and their respective printer/screening patterns. IMHO, each print had certain aspects to it that were in a very subtle way better than the other two prints. In other words, trying to precisely control a single colorful image (e.g., the model's dress had lots of vivid magenta and green floral color patterns in it) plus skin tone reproduction from highlight through shadows is no small feat. Each printer/driver/profile combination seemed to accel in certain image areas. That said, when members of my AaI&A digital print research program send me samples of my 12 hues_for_sRGB colorspace image target for light fade testing, it is pretty easy to see that the 7900/9900 printers are edging out the other Epson printers in subtle yet demonstrable ways. Whether once can coax this subtle additional color gamut and dot screening precision into one's real prints is another matter. The bottom line is that in day-to-day printing a 7800, 7880 or 7900 can all do outstanding work in the hands of an accomplished printmaker, but over the long haul an accomplished printmaker is probably going to prefer the 7900/HDR system.

Cheers,
Mark
http:www.aardenburg-imaging.

Hard to draw any definitive conclusions from 1 image.  I'm sure it was a great image, but I find it hard to believe the 9800 had any areas that were superior even with a subjective analysis.  The OP (or anyone else for that matter) wouldn't be well served to make a decision based on looking at this single image.

As Jeff mentioned, this goes beyond gamut and color, it also has to do with detail and resolution.  I believe Michael mentioned in his review that this printer really does justify using the 2880dpi option on very good files, and I would concur.  I just reprinted a 7ft panorama made from 6 p45+files stitched together and was amazed at the detail visible in the print. The improved screening technology cannot be overlooked.  My 3800 produces better output than my 9800 did despite the same inkset.  The 11880 produces better output than my 9880 did despite the same inkset.  The improvement in head technology allowing for more accurate dot placement allows much improved screening technology which can be seen with some images and good files.

My observation is from printing several dozen of my landscape images, some of which are old enough files they have been printed on all 3 generations of printers, as well as ipf6100 and z3100 printers.  The 7900 is richer and sharper .... I am constantly amazed at the look of these files as they come out of the printer.  Granted I don't have old copies to compare side by side, but I have a pretty good eye and the quality improvement is definitely there. I have compared a few with 11880 files and even there I can see some differences.  Added to all of the other features of the machine, such as the fantastic roll feed system and the substantial speed improvement,  this is the finest inkjet printer I have ever used.

I'll agree the 9800 can do outstanding work, and I also agree that an accomplished printmaker will get more out if it than others.  But as you mentioned the same printmaker will get even more out of a 9900, as long as there isn't an artificial limit in the file.   So my answer to the original posters question, based on my experience printing my own landscape and portrait images is that the 9900 is visually superior in nearly every case to the 9800 output as long as the files are being printed from Lightroom or are converted from RAW using ProPhotoRGB.  Gamut limited files (aRGB and sRGB) will probably show very little difference.

[digitaldog]

So my answer to the original posters question, based on my experience printing my own landscape and portrait images is that the 9900 is visually superior in nearly every case to the 9800 output as long as the files are being printed from Lightroom or are converted from RAW using ProPhotoRGB.  Gamut limited files (aRGB and sRGB) will probably show very little difference.

Exactly! You got to feed em good, wide gamut data, then you see a difference (which I think Douglas image did illustrate on the models dress).

[MHMG]

Exactly! You got to feed em good, wide gamut data, then you see a difference (which I think Douglas image did illustrate on the models dress).

Yes, the colorful dress (comprised of vivid greens, reds, and magenta floral patterns) did showed the most differences, more going from the K3 to K3VM than from K3VM to HDR, but nonetheless, as you have said the colorful part of the image revealed the transition that Epson was hoping people would see. I concur. However, that image also had large expanses of caucasian skin tone colors and low chroma near neutrals (i.e, the white dress with subtle shadows throughout) ranging from high key highlights to deep shadows in skintones and other image elements. It was here in these tonal transitions where hue and chroma WAS NOT pushing the color gamut limits of any of these printers that I found variations in the three renditions with some areas being slightly more pleasingly rendered than others. Each print had some preferred tonal rendering of various areas in the image, IMHO. This is no doubt a consequence of sending presumably the same digital data through three different ICC profiles. Not an easy task to get three different printers with three different ink sets to match output precisely even in areas that would not be challenging the color gamut of any of these systems.

I also concede that it was just one image, but it was an image constructed with enough skin tone and low chroma color areas also present in the scene to fairly represent a broad number of other photographic images as well. My point was not that the three printers aren't different in technical output quality which can be measured and to varying degrees documented, only that the differences are subtle enough that I wouldn't be compelled to trade in a 9800 or 9880 based on image quality alone. Other factors like ease of use, output speed, improved reliability maybe, but not simply for bragging rights on image quality.

[BobDavid]

Hi Bob, we have a large number of the 9900's and personally think they are the best machines on the market!!!  Compared to the 9800 we noticed the 9900 significantly faster,  HUGE improvement over color.. less gradation and images come out rich and vibrant (big difference compared to the 9800 - especially green and orange colours).  Best of all replacing rolls takes no time compared to the 9800. No more breaking your back trying to replace a 44" roll.

One thing also, we had 2 - 9900's running for a SOLID 1.5 months (night and day) and did not encounter any problems mechanically or loss of image quality. Registration was always bang on from start to finish.  This is something the 9800 would of never accomplished!  If  your looking for speed, reliability, exact color re-production and bang for the buck this is truly the machine for you!


TJ
Juke Box

Thanks for your feedback. I appreciate it. ... Impressive looking setup too.

[Ray]

It was here in these tonal transitions where hue and chroma WAS NOT pushing the color gamut limits of any of these printers that I found variations in the three renditions with some areas being slightly more pleasingly rendered than others. Each print had some preferred tonal rendering of various areas in the image, IMHO. This is no doubt a consequence of sending presumably the same digital data through three different ICC profiles. Not an easy task to get three different printers with three different ink sets to match output precisely even in areas that would not be challenging the color gamut of any of these systems.

This is a good point. When comparing a specific quality of equipment performance one needs to eliminate other variables as much as possible which may influence the result. Presumably one needs to use some sort of profile to make a print. We all know that profiles can vary in their success, depending on the equipment and expertise used to make the profile. If the profiles are generic, there may also be some slight variation amongst printers of the same model.

How does one eliminate the variation of the quality of profiles used when comparing the output of different models of printers?

[digitaldog]

How does one eliminate the variation of the quality of profiles used when comparing the output of different models of printers?

If you have the software (something like ColorThink), you can plot the deltaE to the profile which is a start. You basically plot the deltaE of the predicted values against the measured value.

I highly suspect that all the profiles for the Epson prints discussed where carefully generated. But going back to the above, it is interesting to see this deltaE and, when you send patches of various color spaces, its interesting to see the results of this deltaE report of which the original color space can play a role.

[Ray]

If you have the software (something like ColorThink), you can plot the deltaE to the profile which is a start. You basically plot the deltaE of the predicted values against the measured value.

I highly suspect that all the profiles for the Epson prints discussed where carefully generated. But going back to the above, it is interesting to see this deltaE and, when you send patches of various color spaces, its interesting to see the results of this deltaE report of which the original color space can play a role.

Okay! But supposing you are using profiles made by others? For example, I am still using the 'freely available' Bill Atkinson profiles with my Epson 7600. They are certainly much better than the Epson profiles that shipped with the printer. That's why Bill went to the trouble of improving upon them.

But, supposing I were to upgrade to an Epson 7900. I would no doubt notice an improvement in gamut and accuracy of tonality and color rendition. The improvement might well be obvious, using ProPhoto color space, which I always do.

I wonder if such obvious improvement might be far more subtle if I had taken the trouble and expense to get the best current profile available for my specific model of the 7600 before making comparisons.

I should add that I have never made a printer profile in my life, although I do possess the means to make one since I own a ColorMunki. I'm just too busy to see if I can improve upon the old Bill Atkinson profiles. I'm grateful to Bill for doing the work and making his profiles freely available. He's saved me a lot of stuffing around.

[MHMG]

If you have the software (something like ColorThink), you can plot the deltaE to the profile which is a start. You basically plot the deltaE of the predicted values against the measured value.

Except that delta E doesn't measure direction of visual change. So, for instance, take two neighboring colors that should have a tonal gradient in the image of 2 delta L in absolute L* difference between the two colors. Now apply a one delta E error to just one of the colors. If this one delta E error happens fully as a result of a shift in lightness, we could now have the two side-by-side tones become 3 delta L apart, or they could become 1 delta L apart. The delta E analysis dutifully reports in both cases that one color shifted one delta E. Yet, this evaluation tells us nothing about the direction of shift between the two side-by-side colors. In other words, applying a conventional Delta E analysis says the perceived 1 delta E visual error, when comparing the "before" and "after" change in that individual color, is the same no matter what direction that color shifted with respect to its near neighbor color. However, in one case, a 50% increase in visual contrast has now occurred between the two neighboring tones, in the other case a 50% reduction in the visual contrast has occurred between the two neighboring tones.

The conventional application of Delta E to images is indeed "a start" but it doesn't account for near neighbor relationships. The visual impact on the image can be dramatically different with the same delta E outcome. That's why I also use the I* metric which tracks changes in both image lightness and image contrast.

Sorry Andrew, I couldn't resist  

cheers,
Mark
http://www.aardenburg-imaging.com