Comparison of Epson 7900 versus HP Z3200

[neil snape]

From my (very) limited printing experience view I would agree, even with profiles you still need to learn how a paper will look when printed. So limiting your choice of paper is the first step, I don't see the need to spend a fortune on print profile calibration machinery when you only need a small number of papers and profiles. Find a paper you like and stick with it, unless something super wonderful comes along that really improves your printed images, then get a profile made and learn that one.

Kevin.

There are not many users that will use the profiling part of the printers for a lot of papers. For the amount of profiles needed often not more than 5-10 will be all that will ever be needed. With an i1 and the sliding ruler it's not so hard to do, and for these few profiles will avoid having something like a full IsIs etc.
Where the onboard spectro becomes interesting is for controlling the calibrations and verification of the before and after states of the printer.

Epson are known to be more stable over thermal heads, but you'd be surprised to see variations in relatively extreme conditions.
Calibrating the printer per paper before jobs is a great thing. From what we read you can do so with the 7900 as you can with the HP Z. After the calibration is done, the profile will be able to repro any job with that much less deviation.
On the outside, reading of control strips is a big plus, exactly why Epson describe the 79+9900 series as a prepress printer.
Of course this doesn't have to be a prepress only thing, photographers and FA printers are someday going to want to validate the printing state at the job time. Digigraphie with Epson in Europe is a very strong argument for an archival print, yet print verification could be a strong addition to Digigraphie.

[Scott Martin]

Is it doable for the user to design the right CMYKxxxxx profiles with the suitable profile creators ?

RIPs allow for CMYK profiling of 6-12 ink printers. The final 6-12 ink separation is done after the CMYK conversion (in some cases on the printer's processor(fast!), in other cases in the RIP(slow)). This separates the very different roles of color space conversions and complicated multi-ink separation parameters. It's easier for the end user and the results are fantastic. No need to make multicolor profiles.

External spectrometers and no support for the integrated spectrometer ?

They all support external spectros. GMG brings a new level of hands-off workflow automation though it's support of the on-board Z spectro. The response to that has been fantastic. HP has done well and Epson is scrambling to get into that game ASAP. I hope we'll see more products taking advantage of on-printer spectros. With prices coming down the way they are, we might one day see spectros in every large format printer. I personally like the consistency of being able to calibrate and profile all the printers in a large shop with a single software and hardware solution.

[KevinA]

There are not many users that will use the profiling part of the printers for a lot of papers. For the amount of profiles needed often not more than 5-10 will be all that will ever be needed. With an i1 and the sliding ruler it's not so hard to do, and for these few profiles will avoid having something like a full IsIs etc.
Where the onboard spectro becomes interesting is for controlling the calibrations and verification of the before and after states of the printer.

Epson are known to be more stable over thermal heads, but you'd be surprised to see variations in relatively extreme conditions.
Calibrating the printer per paper before jobs is a great thing. From what we read you can do so with the 7900 as you can with the HP Z. After the calibration is done, the profile will be able to repro any job with that much less deviation.
On the outside, reading of control strips is a big plus, exactly why Epson describe the 79+9900 series as a prepress printer.
Of course this doesn't have to be a prepress only thing, photographers and FA printers are someday going to want to validate the printing state at the job time. Digigraphie with Epson in Europe is a very strong argument for an archival print, yet print verification could be a strong addition to Digigraphie.

So is this variation a paper batch thing like we would of had with colour paper in the wet days or a machine drifting off spec more like a chemical thing. I don't for one minute count myself as anything but a novice dabbler at fineart printing but if I do a reprint from something that was done several months ago it still looks the same to me. I can see it's need in the proofing environment where there is much more everyday variety. Most of these machines sold will be run by one guy making a few prints a day that learns how the softproof relates to the paper print and any slight variation over time will be compensated for without realising it.
I can see for large commercial batch runs ongoing checks would be a bonus, but for most here would be overkill.
For me to get prints of a standard I would like I don't see the need for constant paper profiling with the consistency of materials companies like Epson make, in fact that is one of the big improvements over wet printing for me.
If there are big variations in before and after printing having a Spectro will not solve the problem only prove it and if you need to measure it to see it, does it matter to the average photographer/Fineart printer? For me it would be just a load more information to worry about.

Cheers,

Kevin.

[neil snape]

So is this variation a paper batch thing like we would of had with colour paper in the wet days or a machine drifting off spec more like a chemical thing. I don't for one minute count myself as anything but a novice dabbler at fineart printing but if I do a reprint from something that was done several months ago it still looks the same to me. I can see it's need in the proofing environment where there is much more everyday variety. Most of these machines sold will be run by one guy making a few prints a day that learns how the softproof relates to the paper print and any slight variation over time will be compensated for without realising it.
I can see for large commercial batch runs ongoing checks would be a bonus, but for most here would be overkill.
For me to get prints of a standard I would like I don't see the need for constant paper profiling with the consistency of materials companies like Epson make, in fact that is one of the big improvements over wet printing for me.
If there are big variations in before and after printing having a Spectro will not solve the problem only prove it and if you need to measure it to see it, does it matter to the average photographer/Fineart printer? For me it would be just a load more information to worry about.

Cheers,

Kevin.

Variations in inking, inks, paper, humidity, conditions of the print heads, all are part of the parcel.
The idea of a spectro onboard is for those who want to control and verify. There is no obligation, and it's a good thing that Epson do not do so.
I'm not sure if it is overkill or not. I do know that if you have to reprint limited edition prints then it makes the difference between what is a better guarantee of that edition in the eyes of all concerned.
For one offs , pleasing colour has always been fairly easy to achieve. Accurate and repeatable colour though needs extra steps. Nothing says it has to be an onboard spectro BTW, as this has been done off line for a long time too. The onboard part takes out uncertainty by it's repeatable consistency in measure.
If the variations of the prints are large enough for you to see , they are indeed larger than what you would want on sellable prints.
Not all prints need this kind of precision, nor do users all need this kind of tolerance in  printing, but those looking at the 7900 +9900 might well be the few that do.

[KevinA]

Variations in inking, inks, paper, humidity, conditions of the print heads, all are part of the parcel.
The idea of a spectro onboard is for those who want to control and verify. There is no obligation, and it's a good thing that Epson do not do so.
I'm not sure if it is overkill or not. I do know that if you have to reprint limited edition prints then it makes the difference between what is a better guarantee of that edition in the eyes of all concerned.
For one offs , pleasing colour has always been fairly easy to achieve. Accurate and repeatable colour though needs extra steps. Nothing says it has to be an onboard spectro BTW, as this has been done off line for a long time too. The onboard part takes out uncertainty by it's repeatable consistency in measure.
If the variations of the prints are large enough for you to see , they are indeed larger than what you would want on sellable prints.
Not all prints need this kind of precision, nor do users all need this kind of tolerance in  printing, but those looking at the 7900 +9900 might well be the few that do.

Neil,
I can see why you would need it. I also think a lot will think they need it and buy it when they don't need or know exactly what they are trying to achieve with it. Your needs are not far removed from what Epson is saying it's aimed at.

Kevin.

[neil snape]

Neil,
I can see why you would need it. I also think a lot will think they need it and buy it when they don't need or know exactly what they are trying to achieve with it. Your needs are not far removed from what Epson is saying it's aimed at.

Kevin.

Actually I can see your point. True, the vast majority of users would never be able to distinguish a difference between runs, and for good reason: the inks are stable and the paper mills between batch lots are very close.

I think I get caught up in this trip all too often... LOL I better get back to printing on my less than ideal A3+ printer for my portfolio pix that are supposed to get me work.

[Mussi_Spectraflow]

Finally got the Epson SpectroProofer to work last night. Very interesting unit. Don't know if this has been answered but it look as if the included utilities do not allow for profile creation. Yes this seems totally counter intuitive but this was backed up by the documentation. Of course this could change. What it does do out of the box is perform linearizations. The process of linearizing a media for all of the resolutions takes about 45minutes. I was surprised by the speed of the spectro, simply measuring strips it is blazingly fast, however to advance to the next line of a patch takes several seconds meaning that in the end it's not a lot faster than the Z series spectro. The chart size however is drastically smaller, somewhere around the size of an i1 chart. By default the printer will also print measure and then save a CGATS measurement file which can then be dropped into your profiling package to produce a profile. Also I thought that the bronzing on the 7900 was not much different than the 7880. Granted this is based on making prints on the premium luster 260 paper, but I was a bit let down. Not that it's bad in any respect, I just didn't notice any improvement. I'll go back and review this, and likely this could improve with new firmware, but that's my observation at this point.

[alan a]

Finally got the Epson SpectroProofer to work last night. Very interesting unit. Don't know if this has been answered but it look as if the included utilities do not allow for profile creation. Yes this seems totally counter intuitive but this was backed up by the documentation. Of course this could change. What it does do out of the box is perform linearizations. The process of linearizing a media for all of the resolutions takes about 45minutes. I was surprised by the speed of the spectro, simply measuring strips it is blazingly fast, however to advance to the next line of a patch takes several seconds meaning that in the end it's not a lot faster than the Z series spectro. The chart size however is drastically smaller, somewhere around the size of an i1 chart. By default the printer will also print measure and then save a CGATS measurement file which can then be dropped into your profiling package to produce a profile. Also I thought that the bronzing on the 7900 was not much different than the 7880. Granted this is based on making prints on the premium luster 260 paper, but I was a bit let down. Not that it's bad in any respect, I just didn't notice any improvement. I'll go back and review this, and likely this could improve with new firmware, but that's my observation at this point.

I hadn't been able to print my images on the same or similar paper at Photokina but can say that what I saw confirms gloss differential is controlled by HP to the point of a non issue for GD, but don't forget the increase in cost and printing time, and fragility of the surface. Epson has very GD by nature, but is not completely eliminated, nor is it for any pigment printer I have seen to date.

It is extremely variable though on different surfaces. For example, at the Innova stand they had all three brands of printers. Kind of surprising to see , they had many great images printed on bad combinations of media and image style. On certain media the Epson VM inkset and the 7900 had substantial GD on glossy Innova (can't remember the name), as did the Canon too. HP with GE there was little or rather none in the highlights . . .One of the reasons I felt when the Z 3100 was released the GE was not marketed as an important element as it should have been. It is a good solution, one that works. IF not for GE, the inkset they have is not up to par with the others as far as gloss diff goes..

When I began this thread, I acknowledged in an exchange with Schewe that I was leaning towards buying the Epson 7900.  I'm not so sure based on these reports.  From my perspective, as only one consumer, these reports represent two strikes against the Epson 7900 and in favor of the HP Z3200/3100 series with Gloss Enhancer (GE) and a built-in spectro that can perform paper profiling.

We now have two reports, from two experienced reviewers, who have used both printers.  

Strike 1 --  Both report that Epson still suffers from gloss differential (GD).  Julian reports that it has not improved from the 7880 to the 7900, and that "I was a bit let down."  (The last Epson I used was the 4000, and it is my frame of reference.  I don't know how GD compares between the 4000 and the 7880.  Julian, is GD on the 7880 and the 4000 about the same?)  Neil reports "substantial GD" with the 7900 in the case of one paper type and with HP "there was little or rather none in the highlights" with the same paper type.

From my own experience with the Z3100, the paper type used by Julian, Epson Luster, is not one of the worst for gloss differential by any means.  Luster would not be the paper choice to test GD. Others are even better examples of gross GD without the HP GE and almost no GD when HP GE is used.  So I assume that the 7900 looks even worse with other papers than what Julian is reporting.  If I was to buy the 7900 I'd be kicking myself everytime I looked at a print on photo paper with GD, remembering how well controlled GD is with the HP GE.

Strike 2 --  Julian addressed the issue of the Epson spectro with the included utilities.  So the jury is still out as to whether it could make profiles with Profilemaker 5 or some other software.  But note that the MSRP on that is $2,500 on the Xrite web site, and B&H lists it as a discontinued item in any event.  

To again acknowledge one point, the spectro is not the most important issue for a printer.  A bad printer with a spectro is still a bad printer.  Other issues, such as GD, GE, and overall print quality, are more important.  I also recognize that opinions vary on the value of an on-board spectro for profiles based on the above posts, but I strongly suspect that some of those who have posted have never actually used such an HP Z series printer and haven't learned what a wonderful convenience it represents. Some who have used the Z series, such as Neil, point out that you can do profiling with various hand-held and manual packages.  But why bother when automation is available and included with the HP Z series?  

Having said the above, Epson shot themselves in the foot on this one.  As I previously argued, for Epson to produce and market a spectro, but not include profiling software to drive it, is an illogical marketing decision by Epson.  Epson is handing that sales pitch and feature to HP and the Z series.  Common sense says that Epson is better off to be able to compete with HP with regards to that feature, than not at all.  If the 7900 and Z3200 are about equal as printers, consumers might opt for the one with the very nice additional feature -- a spectro with software to profile papers.  The Z3100 provided tangible proof that it is an important consideration for many consumers, all other things being equal.  

I also wondered, in an earlier posting, if the Epson engineers might be shortsighted enough and bad enough at marketing to build an expensive spectro but not include the necessary software to drive it and fully utilize its potential.  We have our answer.

Different people have different needs and will buy different printers as a result.  For me, as one consumer, this represents two big strikes against the Epson 7900.

[JimGoshorn]

By all accounts, this printer was not originally designed for the photography/fine art market but for the press market which works through RIPs. Since the printer is going to be supported by the major RIPs, it is there where you are likely to find the profiling feature if it is going to be found anywhere. Epson's literature says "When driven by the latest software RIP front-ends, the Epson SpectroProofer can automate virtually any color management process in your workflow".

Not defending the decision given the HP Z printers have a spectro, but given Epson's main target audience for this printer, it's not too surprising.

Jim

[Ernst Dinkla]

By all accounts, this printer was not originally designed for the photography/fine art market but for the press market which works through RIPs. Since the printer is going to be supported by the major RIPs, it is there where you are likely to find the profiling feature if it is going to be found anywhere. Epson's literature says "When driven by the latest software RIP front-ends, the Epson SpectroProofer can automate virtually any color management process in your workflow".

Not defending the decision given the HP Z printers have a spectro, but given Epson's main target audience for this printer, it's not too surprising.

Jim

Jim,

It could be more complex though. And "automate virtually any color management process in your workflow" is as vague as it is broad in its definition.

Before you actually can create profiles with an x900 there may be more steps needed: Epson x900 + optional spectrometer + optional RIP + optional profile creation software.

In some cases the last two come in one package or the profile creation is optional for the RIP solution. Some RIPS do not have an extension for profile creation. There's another (unlikely) possibility where the profile creation software can approach the printer independently, sending it targets and let the spectrometer read them. But I doubt you could get RGB-device like profiles that way, usable for the normal driver like the basic HP solution does. If it creates CMYK(XXX) style profiles you need a RIP anyway to use them on that printer.

I think it is wise to start the quest for an Epson integrated package at the other side of the chain, let the RIP or Profiler manufacturer demonstrate that the profiler and RIP work on the Epson x900 and ask the price of the total there. That's not unusual in the pre-press market.


Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

[Mussi_Spectraflow]

Honestly, and I'll argue against myself here, I wouldn't make any decisions yet. Remember that the prints were working with have not even shipped yet. The Z3100 in it's early state was almost unusable on matte paper. I expect the Epson and the HP to be further improved by the time they are shipping in quantity. Let me also add that the bronzing on the 7900 is not bad by any means, it's still really limited to the highlights and there is just a hint of gloss differential in the blackest black.... Okay I made some new prints this morning using a different profile and I'll add that there is some improvement over the same print on the 7880, I notice it in the green going to blue highlights. After staring at prints for the last 10 minutes I'll say that as usual the problem still mostly affects the magenta regions of a print. So depending on how picky you are, you may not have any issue with it. I'm sorry I cant be any clearer than that. I can also confirm that you can easily make a profile with the cgats file that the 7900 spits out. Interesting to note that this option is now included in the Z3200 color center as well(measuring and saving a cgats file).
On last thing. While this printer may not have been designed for the photo fine art market, it's going to be a very capable high volume photo printer. In this case it's likely that a RIP will be used or that some form of print automation will be employed. In this respect the spectroproofer should be well suited. It's a very interesting product, looks well built, includes a print dryer, records the ambient temperature, although would have liked it to be a bit faster. It's also possible that Epson may include some sort of profiling option as well. There are also free profile creation utilities (Argyle) that you could use to generate a profile with. That said this is a vastly different approach to color management than HP has taken. I actually have a lot more respect for the color center and it's ability to manage, create and maintain paper presets.

[Roscolo]

In my prints, the z3100 eliminated gloss differential. YMMV

The spectro on the z is not just handy, its value is not just in its accuracy, but its ease of use and convenience. It doesn't require one to have any special knowledge or skill, and it doesn't take any of my time. With the z, you literally just click and the z3100 does its thing and prints the target, scans the target, creates and installs the profile. While it's doing this, I can be doing other work, then come back 20 minutes later and your paper is profiled and your prints are a perfect match, provided you have a well calibrated monitor.

[Ernst Dinkla]

Honestly, and I'll argue against myself here, I wouldn't make any decisions yet. Remember that the prints were working with have not even shipped yet. The Z3100 in it's early state was almost unusable on matte paper. I expect the Epson and the HP to be further improved by the time they are shipping in quantity. Let me also add that the bronzing on the 7900 is not bad by any means, it's still really limited to the highlights and there is just a hint of gloss differential in the blackest black.... Okay I made some new prints this morning using a different profile and I'll add that there is some improvement over the same print on the 7880, I notice it in the green going to blue highlights. After staring at prints for the last 10 minutes I'll say that as usual the problem still mostly affects the magenta regions of a print. So depending on how picky you are, you may not have any issue with it. I'm sorry I cant be any clearer than that. I can also confirm that you can easily make a profile with the cgats file that the 7900 spits out. Interesting to note that this option is now included in the Z3200 color center as well(measuring and saving a cgats file).
On last thing. While this printer may not have been designed for the photo fine art market, it's going to be a very capable high volume photo printer. In this case it's likely that a RIP will be used or that some form of print automation will be employed. In this respect the spectroproofer should be well suited. It's a very interesting product, looks well built, includes a print dryer, records the ambient temperature, although would have liked it to be a bit faster. It's also possible that Epson may include some sort of profiling option as well. There are also free profile creation utilities (Argyle) that you could use to generate a profile with. That said this is a vastly different approach to color management than HP has taken. I actually have a lot more respect for the color center and it's ability to manage, create and maintain paper presets.

Alright, so with a profile creation program at hand one can make profiles already without the need of an integrated profile creation program ?  Based on the targets available in the printer/driver or feeding the printer any target (any cgats file) from the external program ?

There was a volunteer who liked to make cooperation between the Z3100 and ArgyllCMS possible. I had some doubts whether that would work with the interface the Z3100 uses but with the cgats output of the Z3200 it should be quite easy. Is it already in color center and not in an optional extension of it ? Upgrading to the Z3200 becomes more interesting every day.

Strange that both companies more or less cover up the cgats feature in their specs of the machines.


Ernst Dinkla

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/

[stevenf]

I am not sure if clogging is a factor any more with the epson printers.

I had an epson 7800 and had no end of problems with clogging. I now have had the z3100 for a year and have never had an issue with clogging. The only issue is the printer wakes my wife up at night when it cycles. My office is above our master bedroom.

I would be interested to know if the clogging is still an issue with the new epson 9880/9900 series of printers. I tend to go away for a month at a time for photo shoots so the printer doesn't run on a daily basis.

I print for other photographers (they bring their own paper as they have differing paper needs compared to mine)- I  have to say the profiling ability of the z printers is a huge benefit.

Steven

www.friedmanphoto.com

[alan a]

The spectro on the z is not just handy, its value is not just in its accuracy, but its ease of use and convenience. It doesn't require one to have any special knowledge or skill, and it doesn't take any of my time. With the z, you literally just click and the z3100 does its thing and prints the target, scans the target, creates and installs the profile. While it's doing this, I can be doing other work, then come back 20 minutes later and your paper is profiled and your prints are a perfect match, provided you have a well calibrated monitor.

An excellent explanation of the practical advantage of the spectro on the Z3100.  

Some of the above postings have argued that the 7900 is only intended for prepress shops and similar commercial operations.  That had better be the case, because if the printers are demonstrated side-by-side to photographers, the advantages of the spectro on the Z series will be immediately obvious, and just as Roscolo described it.

Epson can certainly release software in the future that would fully utilize the capabilities of their expensive spectro, and neutralize the marketing advantage now held by HP with regards to having a spectro that can profile papers very conveniently as described by Roscolo.

And then we have the issue of GD using a paper that is vulnerable to it, with a photo file that really makes it obvious -- like gray mountains or rocks against white snow, and a comparison of both printers in that regard.

If the 7900 is more vulnerable to GD than the HP Z series with GE, there probably isn't anything that Epson can do about it at this point, since Epson chose not to utilize an ink like the GE used by HP.

[tony wyeth]

I read this on X-rites site.

With the Epson SpectroProofer, the following color control workflows can be achieved automatically and very precisely: color calibration, color certification, daily color checks, profile creation, spot color matching, color shift tracking and target measurement. Furthermore, a color certification label, like the Fogra wedge, can be printed to support the ISO certification process. The SpectroProofer allows initiation of a printer color calibration test either automatically, at a predefined interval, or at the request of a service technician.

A compact, carriage-mounted in-line spectrophotometer, with completely self-contained optics and electronics, X-Rite’s ILS platform enables seamless non-contact measurements for fast moving media and provides direct feedback for color calibration/management and control of Epson Stylus Pro color printers. It quickly sends the calculated color data to a host RIP such as the Epson SpectroProofer software, via a direct USB interface link, providing on-the-fly calibration and closed loop, accurate color management of the print process and patented remote control capabilities.

The newly launched Epson 24” Stylus Pro 7900 and 44” Stylus Pro 9900 set new standards for the professional large format printer market. Improved ink formulation, a new printer mechanism and a raft of new technologies make these two new additions to the Epson Stylus Pro range ideal for production and contract proofing, as well as for the highest quality fine-art and photographic applications. They will also enable printers to extend their business into the demanding packaging proofing and flexographic markets.

It seems to answer some of our questions.
Thank you all for your posts, please keep them comeing.

[alan a]

I read this on X-rites site. . .It seems to answer some of our questions.

Before you actually can create profiles with an x900 there may be more steps needed: Epson x900 + optional spectrometer + optional RIP + optional profile creation software.

In some cases the last two come in one package or the profile creation is optional for the RIP solution. Some RIPS do not have an extension for profile creation. There's another (unlikely) possibility where the profile creation software can approach the printer independently, sending it targets and let the spectrometer read them. But I doubt you could get RGB-device like profiles that way, usable for the normal driver like the basic HP solution does. If it creates CMYK(XXX) style profiles you need a RIP anyway to use them on that printer.

I think it is wise to start the quest for an Epson integrated package at the other side of the chain, let the RIP or Profiler manufacturer demonstrate that the profiler and RIP work on the Epson x900 and ask the price of the total there. That's not unusual in the pre-press market.

Unfortunately, the X-Rite PR piece does not answer our questions.  We've already read similar vague summaries, none of which specifically explain which of these functions are performed by the software included with the spectro versus other RIPS.  Nor do any of these summaries explain specifically which RIPs could do this and whether they require additional profile creation software to accomplish this.  The combination of all of that could add thousands of dollars to the price.  Ernst is correct -- we need the RIP or profile manufacturer to demonstrate an integrated package that works with the Epson spectro and then, and only then, will we know how it works, whether it works, and what it costs.    

According to Julian the included Epson utilities do not accomplish some of these functions, especially paper profiling.  The X-Rite description appears to be a great piece of company propoganda to applaud their role in creating this device.  But what X-Rite declines to disclose is their likely role in creating the utilities that drive the device, and that those utilities failed to provide profiling of papers.

Remember that the Epson spectro already adds $1,500 to the price of the printer.  I can envision some photographers buying the Epson solution for an additional $800 as compared with the cost of the HP Z series that includes a spectro in the price of the printer.  (Not sure what the Z3200 with APS will cost, but the MSRP on the HP web site says $4,700.  If the 7900 and spectro is $5,500, then the price difference is $800.)

But that assumes that the included Epson software would perform paper profiling.  It does not, and while X-Rite is busy patting themselves on the back, they didn't design the utilities to perform that function according to the report from Julian.  So as Ernst said, we need to know what the actual added cost would be with additional software.  Such a software solution is probably very expensive.

P.S. --  Before the usual responses come in that we don't really need spectros; that hand-held laborious manual solutions work as well; or that the PQ of a printer is more important -- see the above postings from Roscolo and me, as well as others, that address those points.  No need to repeat all of that here.  I will again repeat that the printing capability of a printer is more important than a spectro.  A bad printer with a spectro is still a bad printer.  The point is that an automated spectro for profiling on a high quality printer is a wonderful feature, all other things being equal, and all the more so if the price is $800 less.

[alan a]

My original posting included a long list of issues that must be considered when looking at the HP Z series versus the Epson 7900, so let me focus attention on three others that have not been commented on.

Mechanical reliability; Software and driver reliability; and Customer Service and tech support.

That was one of the reasons that I was leaning to Epson, and honestly disclosed that in an earlier response to Schewe.  It is impossible to quantify customer service and tech support.  What we need is the equivalent of a consumer report survey for printers, with thousands of responses, so we get an accurate representation of consumer experience.

Lacking that, each of us needs to make his or her own gut determination.  That requires setting aside anecdotal and isolated reports from a few people, and looking at the big picture.  Probably only people with problems post in this forum, and people who have no problems do not.  Even if that is true, this forum, in only the last several months, let alone the last two years, has had more reported problems with HP than Epson.  

Those problems with HP, and there are many more such reports as compared with Epson, are related to mechanical issues, software and driver issues, and customer service and tech support.  I based that just on a quick scan of the topics that are posted.

If Epson included a spectro and utilities for paper profiling, and even if Epson was a little worse on GD on some papers, I would lean to Epson even if costs $800 more.  Because I am guessing that the 7900 might prove to be superior in other areas as compared with the Z3200, and my own gut impression is that Epson is clearly superior on mechanical reliability, software and driver reliability, and customer service and tech support.

Unfortunately, Epson dropped the ball on the spectro; can't compete with HP in that regard; and might be significantly worse on GD.  

(On GD I'll take a DVD with a photo file to one of the Epson Print Academies and see if they will make a print on Epson paper that I'll bring with me, so I can make my own determination.  As noted in a previous posting, they won't do that from a USB memory stick, presumably due to viruses, so I don't know if they will from a DVD.  But I will nonetheless try, as my local area will never have a 7900 set up to demo.)

[neil snape]

Again I think the questions will be answered about the spectro , which applications will drive it, and eventual options on using it.
The biggest part of the first development was by GMB, and I can imagine there are a lot of patents out on the process. Since a lot of this is till in the same hands , I'm quite sure there will have to be some concessions before Epson can integrate in ways that HP have locked up.
It took GMG quite some time to tune their RIP with the HP, but when done I'm told it has worked very well since. The same will be true with Epson, it will take time and some rounds of updates. Even the APS package was released a little prematurely.
Image quality was and is still something of it's only flavour on each brand. Where one can time pondering which can do one task well, the decision on which printer does well at most things takes less time.
Tech support is not something I know much about as I didn't or haven't had any problem that needed any attention or onsite repair (on any shipping model that is). I read reports here though which go to both extremes.
Build quality is important though, and attention to details. So comparing price alone doesn't cover build quality. I didn't find the HP to have a bad overall build quality, just the opposite, things like the stand were way ahead of others. Yet certain details , little things like a switch or wiring, guides, catch tray all take away from being the best they could be.
Epson have shown that they are very serious about photography, photographers, and FA printers. It's not about how much money you throw at names, but the care going out to users, their choice of who represents them, and the connection between them all. Funny though , that the less than perfect history is quickly forgotten as soon as the problems are solved. Not true for tech support though as the contact you have is a determining point that will be remembered long after the fact.

In the end to hold some prints in your hands will tell you what is right for you. But only since through this forum there is a lot of shared knowledge.

[MHMG]

Just returned from Photo Plus East in NY today. Not to muddy the waters any more, but I will share a few thoughts. First, I applaud Hp for often enlisting its engineering staff from Spain to help out on the trade show floor. It's great to talk to highly knowledgeable people about products they have worked on and are justifiably proud of. I happened to remark to one such Hp engineer that I was a potential customer but the on-board spectro actually seemed redundant because I already own a Spectroscan with Profilemaker 5 software. My remark prompted a very interesting and refreshingly honest reply from this Hp engineer. He said essentially that the spectro was necessary to achieve the printer's closed-loop calibration consistency, but my standalone version of Profilemaker 5 would actually build an even higher quality profile than those produced by the printer software, even the APS option. He then advised "you can import your own custom reference target data to print, use the on-board spectro to generate your measurements, then export the measured color patch data as a text file to build an even better profile with your external application". That advice confirms others' comments about the Z3200 spectro's data exporting feature, and presents a justification why one might want to do just that. Kind of reminds me of Bill Atkinson's "bouquet of profiles" made with different profiling apps.

I had the following exchange with an Epson rep about the HDR ink in the 7900:  I said "I assume the orange ink in the HDR set is being blended and swapped for yellow and magenta inks to make skin tone colors". He replied, "Absolutely!  The orange ink improves the skin tone reproduction and our beta testers have told us they see noticeably better skin tones".  I then asked" How does the new orange ink affect the light fastness of the skin tones". He replied - "It's been tested by....(an independent lab)". I said "but this lab's published testing methodology doesn't include any colors other than two levels of cyan, magenta, yellow, and gray. Do you happen to know if the testing was modified to handle the new HDR ink set?"  He then answered "Well, the orange ink is not the limiting factor". Our discussion had quickly reached what I know to be the current limits of light fade testing in the industry.  The legacy testing procedures of the traditional color print era which are still in use today for inkjet prints are not well equipped to rate the fading response of new multi-colorant systems.  I am looking forward to getting some samples into my own tests which can differentiate skin tone performance, but the printer's limited availability probably mean it's going to be a while before I can get some test targets printed.

Epson also had a very nice set of three prints mounted side-by-side that were made from the same digital image file on a 7800, 7880, and 7900. The image was one of those anorexic looking models dressed in a bright "floral pattern" dress and vivid lime green hat to illustrate the respective color gamut differences in the K3, K3 vM, and HDR Ultrachrome ink sets. The dress pattern had a dizzying array of magenta, red-magenta, violet-magenta, and lime green color regions. The dress pattern was obviously chosen to accentuate and reveal the differences in color gamuts of the three ink sets. My take was this: Yes, you could indeed see subtle increased color vividness in the magentas, and greens in the expected order from K3 to K3vM to HDR.  But the difference was extremely subtle even for this image purposefully chosen to illustrate the differences. These differences would not be considered significant by the vast majority of viewers or even noted in anything other than a direct side-by-side comparisonof prints. Moreover, there were other more significant lightness and contrast differences in the prints which were probably attributable to the three translations of the file rendered by three different profiles. Thus, there were various local regions in all three prints that I favored slightly. Likewise, for skin tone reproduction.  Kind of like making three "identical" custom prints in the old darkroom days, and preferring one slightly better than the other two because the hand dodging and burning came out ever so slightly differently!  I think the printmaker(s) did a good job with a very challenging assignment, ie., to take a single digital file and reproduce it exactingly on the same paper type using three different printers in order to illustrate very subtle differences indeed between the three printers and their respective ink sets.

With respect to GD, bronzing, and Gloss enhancers, by examining a variety of prints not under glass, I think that neither the Z3200, nor the Canon, nor the Epson offerings have totally eliminated the need to choose substrates wisely, and in some cases resort to post coatings if residual GD and bronzing issues are still of concern. In theory, a gloss enhancer should eliminate GD and bronzing for all glossy/semigloss papers, but this pigment-free clear ink can sink right into a number of third party papers, in some cases making image surface appearance worse rather than better.  Hp to its credit has recognized this tendency of the clear ink to sink right into some image coatings rather than stay on top, and therefore more user control over GE inking levels has been added to the Z3200. The HP engineer I mentioned above conceded that despite the added control over GE, some third party papers are still not always responding well to the GE.  Again, I found his candor very refreshing.
 
My sense from the 7900 samples I saw, is if you already own a 7800 or 7880, IQ improvements from the HDR ink technology on the 7900 are subtle and not a compelling reason to upgrade, but increased speed, likely improvements in clog-free reliability with Epson's teflon nozzle coating technology, plus easy PK/MK switching are all a nice step forward. I don't pay much attention to what the official customer demographic is for the Epson 7900. Photographers and fine art printers are going to embrace it for sure. Schewe, your 7900 print with the image of clasped hands superimposed with the world map that was on display at the Epson booth was truly captivating, and it also left no doubt about the technical precision of the 7900.

Lastly, I was amazed to see learn how deeply various dealers at the show were willing to discount the price of a Canon ipf X100 series printer.  This printer line has the requisite internal calibration/linearization technology to handle printer/environment/media variability, but not the full spectro/profiling capability. However, at it's current street price point, one aught to take a good hard look at the Canon printer line before settling on a Z3200 or Epson 7900. You can buy a lot of custom made profiles for the price differential!