Another Ink Replacement Failure

It's been a maddening week here. I'm about to leave for a short camping trip, immediately followed by a trip east to visit my parents. As often happens, client work has come in steadily, all of which required quick turn-around while I should have been preparing for these trips. Clients don't care about my trips, commitments I've made to other clients, or my desire to relax for a day or two. Fair enough; I do my best to accommodate them.

The 7900 had been idle for a couple of weeks, and has been busy since. Every day's start-up began with a nozzle check that showed the need for cleanings. Every day. As my printer currently has six inks at 5% or lower capacity, cleanings invariably require swapping one or more of those inks for new cartridges. When a clean nozzle check print finally emerges, I replace those fuller cartridges with the originals, most of which are at 1%. This is standard procedure for the 7900, but is tedious and takes a lot of time. What should be a quick printing job can often require an hour or more. Such is life with a 7900.

OK, to the real issue for this posting: In my last post, waaaay back in April, I wrote about the printer stopping in mid-print because an ink cartridge had reached the empty state. The printer allows one to replace that cartridge with a new one, after which the print continues. This is widely reported to work just fine, and in fact has done in my own experience. Then I had the issue reported in the previous posting.

For my most recent job I'd been hired to reproduce several very large oil paintings. I photographed these, color-corrected the resulting files, and started making prints. The job required a number of prints, which I spread out over three days. On the final day I had only one print to make, this one fairly large. The start-up nozzle check showed LLK missing completely. A standard cleaning of the Y/LLK pair resolved this, so I loaded the roll paper and printed.

After printing several inches of the image the printer stopped, reporting an empty VLM cartridge must be replaced. I did that, and the machine continued printing. With horrible results. I canceled the print job. Below is a small section from the "master" file:

Here is a photo of the failed print. I made no effort to color-correct the jpeg for this posting, but you can clearly see the band of magenta in the lower half of the crop:

I removed the roll paper and ran a nozzle check as usual on a sheet of bond paper. This showed the LC channel missing completely. This should have nothing to do with the VLM cartridge swap, and it would seem the LC channel was fine before the machine required that VLM replacement.

Frustrated and more than a little angry, I quit for the day; it was getting late, and I really wanted to wrap up the job, but I knew better than to keep going. Next morning I ran a nozzle check and found no change, with VLM still absent. A powerful cleaning of the LC/VLM pair cleared most of the VLM nozzles. I set up and ran the print, which looked good.

Yet another new experience with the 7900, yet another way to produce scrap and waste time. Some days this machine is not good for my blood pressure. I've now been bitten twice by this "bug." I think the lesson here is that I can't trust the printer to continue properly when an ink cartridge must be replaced mid-print. Instead, I'll cancel the job, scrap the print, replace the ink, run a nozzle check, do whatever cleanings seem necessary, and then, finally, start the print job anew. Yes, this sucks, and shouldn't be necessary, but at least with my machine, clearly is.

  --Jay

LK Mid-print Ink Replacement Fail

I'm in one of those periods in which several of my inks are showing 1% remaining. I have inventory for all, and I've occasionally needed those cartridges when cleanings are required. On those days the cartridge shuffle is the norm, a situation in which the printer won't complete the cleaning cycle until the low inks are replaced by fuller cartridges. When I recently did a swap from MK to PK, after which the machine cleans all channels (as far as I can tell), I simply replaced them all, let the cleaning finish, and then put the one-percenters back in place. This is standard operating procedure.

A week ago I got a job to print ten 8x12 inch and one 16x24 inch black and white photos for a local portrait photographer. She's putting together a display for marketing purposes; she chose a range of subjects, all photographed in the same outdoor location, the doorway of an old barn on her farm.

I made a couple of prints a day, on a 24-inch roll of Epson Luster. The machine started each day without issue, a perfect nozzle check, and then I made the prints. As I've said here many times, the 7900 can be a joy to use when things work well, and the week's work only proves the point; using the machine frequently and regularly keeps everyone happy.

I finished the job by making the large print. I started with a nozzle check, as usual; it was fine, so I fed the roll paper, and started the print job. Several inches into the print the 7900 stopped, displaying a message that LK must be replaced. It's been well publicized that these machines can stop mid-job, ink can be replaced, and then the printer will continue with no visible evidence an ink change had been done. I know this to be true from my own experience. But I'd never had to do it in the middle of a black and white print.

When the message appeared, I removed the empty LK, replacing it with one that had been used for several cleanings. It had 92% of its ink remaining. Swapping the ink took no more than 90 seconds. I closed the ink cartridge door, the system pressurized, and printing resumed. Cool.

The print finished and was cut from the roll. I placed the print on my work table and immediately saw a dark band in the black background. Below is a photo of a part of the affected area. You can't miss the flaw through the horizontal center of this snippet. This is, of course, exactly where printing stopped so the LK could be replaced.

I let the print dry 24 hours, but found no improvement. Interestingly, the dark band is obvious under florescent lighting, but almost impossible to find in today's cloud-softened daylight. In any case, the print is scrap. I made another a few minutes ago. A normal print, no messages from the 7900, no need to swap inks.

Although my sample size is only one, I'll take this as a lesson that when printing black and white, perhaps especially those with large, very black areas, it is NOT cool when printing stops for ink cartridge replacement. I suspect the same would be true for any color image that has large, solidly-colored areas. Should this happen again I'll cancel the job so as to not waste any more paper and ink than necessary, replace the empty cartridge, and start the job over.

  --Jay

Epson Hot Press Natural

I've mentioned here that I rarely print on mat (or "matte", if you insist) papers. It's not that I don't like mat papers; it's more that I really love some of the baryta and other soft-gloss papers and prefer the higher D-max typically found with gloss papers and MK ink. I like fairly contrasty images, and have been very happy with defunct Ilford's GGFS and now with Canson's Infinity Baryta Photographique. And a few others of a similar ilk.

Most of my printing on mat paper has been for clients who have that preference. I recently wrapped up a printing job on mat paper, which required the swap from PK to MK. It had been a little over a week since I'd last used the 7900. I did the ink swap, and the machine then auto-cleaned as usual when I attempted to print the nozzle check. After that cleaning the nozzle check was fine, as were the client's prints.

A few days later I printed my B&W photo on Epson's Hot Press Natural (HPN). This is a warm-toned (no OBA), very stiff paper. It comes off a 24-inch roll with a very strong curl. It's like dealing with a roll of aluminum flashing, so stiff it's almost dangerous, and a challenge to flatten.

I made the photo (above), of a abandoned and crumbling outhouse in the Ninepipe National Wildlife Refuge here in northwest Montana, on a beautiful, calm March morning, just as the sun appeared over the Mission Mountain range east of the NWR. I love these old derelicts, with their deep wood grain and patina, cracked and missing boards, exposed nails and other hardware. The strong side-lighting helps bring out that detail.

The print is lovely. Deep, rich blacks, excellent detail, and the smooth, buttery look of HPN. Sharpening these kinds of images can be a challenge, but is worth the effort to get perfect.

A crop at 100% ("actual pixels" in Photoshop-speak) showing some of the detail in the old shack.

I used Epson's "canned" profile for HPN. While I think the print is excellent, it's not a great match to the monitor (NEC's MultiSync PA272W, calibrated with NEC's SpectraView II software and the NEC-branded colorimeter). The print is warmer than the soft-proofed version on the monitor. My color printing work has been a spot-on match to the monitor, so I'm assuming the canned profile isn't Epson's best work. I've posted a query to friend Dean, who also sometimes uses HPN with his 7900, and who also has a calibrated NEC monitor. I'll post his comments when I receive them.

  --Jay

New Format for Ink Expiration Dates?

Today I received two new ink cartridges, Y and LLC. The first thing I do when new ink arrives is check the expiration date. Long ago I got some new ink with an expiration date barely a week away; given that one sometimes needs a new ink cartridge only for cleanings, after which the older ink (below 5% of its original capacity) can be returned to the printer and used for many more prints, I often need to buy ink that will sit in inventory for weeks, even months.

The ink expiration date ("Best before") is marked on the outer carton, and also on the cartridge itself. It's always been straight forward, obvious, and easy to understand. Until today. The inks I got today have two dates on the carton. The bottom date is from last year, the top date two years later. What's one to make of that? How can both dates be "Best before"? There's no new marking on the carton, nor is there any internal paper with an explanation.

Older ink with a single date at top, newer ink with two dates at bottom.

I assume the older date is a manufacturing date, and the second, later date is the expiration date. Perhaps this should be obvious, but the older date is additional information that's not really of any use. It's an unexplained change I wish Epson had made clear. I don't like guessing Epson's intentions. I haven't had time to look around on Epson's support site, but I'd be shocked to find there any mention of this change.

The ink carts themselves are marked with only one date, as always. This is the later date that appears on the box. No confusion there.

  --Jay

A Delayed Firmware Update

As mentioned in my previous posting, my 7900 had been idle for some time before printing the most recent job. During much of that time I'd been busy with other things; making prints, or doing anything else photographic hadn't been on my to-do list. I therefore hadn't looked lately to see if Epson had posted any updates to firmware or software for the 7900.

A couple of days ago I did that, and was surprised to find, on the U.S. Epson site, a firmware update dated 29 October 2014. As usual, Epson provided no clues as to what changed between the newer firmware version HN110EA, and the previous release, HN111E8. I've no idea how I missed the posting of this. I am certain I've checked the update page since last October. In any case, I downloaded the newer firmware.

I also downloaded an update of the Remote Panel 2 ("RP2") software for Macintosh OS X (I'm still using version 10.9.5 "Mavericks" of OS X, having found no compelling reason to update to the latest).

RP2 has always been twitchy for me. It seems to have trouble locating my printer. When RP2 eventually succeeds after several attempts, it seems to work fine. I'd hoped the newer RP2 (version 3.01) would improve, but alas, it hasn't. My printer is networked. There are reports that RP2 is more reliable with directly-connected (USB) printers. I've never had mine directly connected, so I can't report on that. Also, for me, RP2 never finds newer firmware; the search always fails. To work around that I download the firmware and then mount the image. I point RP2 to that, which works fine, and then proceed with the update, which also has always worked as expected.

I launched RP2. On the third try it located my printer and displayed its status. I selected the option to update firmware, navigated to the downloaded update file, and waited. RP2's status doesn't show much, only that the update is in process. The printer's LCD is a bit more informative, eventually showing a progress bar of asterisks. When this completed the printer restarted. All appeared to happen normally, and a check of the firmware version on the LCD indicated the new version.

However, when I returned to the computer, RP2 had raised a dialog box indicating the update had failed. No explanation, no numbers, only "Update failed." I dismissed that. RP2's status window showed a small blinking icon of the printer overlaid with an X.

Since the printer rebooted, and shows the proper firmware version installed, and since I can navigate through the menus and all appears normal, I suspect the firmware installation went fine and all's well. I've not made a print since the update, but I have a small job to do in the next few days. If that proceeds as expected and the print looks good, I'll forget about this. If not, I'll report back here.

  --Jay

Printing Hassles, Nothing Unusual

Winter in northwestern Montana: often a time to get outside to play in the snow. It can also be a time to start (and perhaps finish!) projects that are set aside in better weather. I live in a town with an economy supported mostly by tourism. It is located between major ski resorts; we see little traffic from either. The tourist season here is short, essentially from July 4th through Labor Day (in the U.S., that's the first Monday in September). During those two months there's traffic, the sidewalks and shops are crowded with people from places near and far, and every weekend is chock full of events drawing large crowds. This is a time when art galleries do very well, the outdoor art shows occupy local parks and town squares, and artists are busy making prints, framing their work, hanging shows, and attending opening receptions.

But winter! Winter is different. A long season of short days. Cold and snowy (although less so every year, it seems). I've spent much of this winter installing some new tech here, solving the mysteries of my new QNAP NAS (network attached storage), and working to process some of the raw files made last year. I've done very little printing. I last fired up the 7900 on 19 December. Until a couple of days ago when I got a job to make 20 prints for a local artist, a painter who has had several of her original works scanned, and for whom I've made prints before. It seems she's planning ahead for this summer.

After being idle for 30 days, the printer's nozzle check showed about one quarter of the LK channel missing, and a few tiny voids in LC. I ran a standard cleaning of the Y/LK pair. I'm not sure why I did this, as the "standard" cleaning almost never clears the problem. That didn't improve this time. After the cleaning LK was almost entirely missing, and LC was entirely  missing. Before continuing I replaced the VLM cartridge, with an ink level of only 1%, with a new cartridge. I then did a standard cleaning of the LC/VLM pair, after which LC had a few scattered voids and the bottom two lines missing. As expected, LK was unchanged. Next I did a powerful cleaning of Y/LK. This fixed the LK channel and left all others unchanged. Finally, I ran another standard cleaning of LC/VLM, which cleared the LC clogs. I then reinstalled the 1% VLM cartridge.

After more than half an hour of this nonsense I was able to print the job. LK is now at 1%, Y is at 4%. I don't have either of those colors on hand, so it's time to order ink.

These kinds of problems seem to be less severe when I print more often, but even when I print every three or four days these stubborn "clogs" (or "ink delivery problems", which aren't necessarily clogged nozzles) can happen. As I've said so often in these postings, that's life with a 7900.

  --Jay

A Bit of a Scare

Anyone who's used this breed of printer from Epson knows the drill: there's a tiny fear every time you run a nozzle check print. Could this be the one? Could this be the start of a head clog, or ink delivery problem of other cause, from which the machine never recovers?

This is scary because we've all read accounts of exactly that happening. One color channel prints an incomplete test pattern, it gets worse with time, and no cleanings, whether standard or "powerful", resolve the problem. The "clog" is permanent. As with hard-drive failures, it's not a matter of if it will happen, but when. These machines become our personal sword of Damocles. Some people call for service, while some attempt the repairs themselves. These are complex machines; a bad nozzle check can have many causes, sometimes multiple causes. There's never a cheap fix. Some people simply scrap their machine or sell it for whatever value its parts might have.

Early this month I got a job to print 17 small prints and two larger ones. The small prints were of a black-and-white photo and would be made on letter-size sheets of Epson luster paper. The larger prints would be on Epson luster (24-inch roll).

The 7900 had been idle for 28 days. As usual I started with a nozzle check. I found scattered tiny areas of the VM and PK channels missing, so I ran a standard cleaning of the C/VM pair and the PK/LK pair. Another nozzle check showed the black channel at 100%, but VM was now missing fully half of the check pattern, much worse than before the cleaning. Next I ran a "powerful" cleaning of the C/VM pair. Another nozzle check showed a half-dozen tiny bits of the pattern still missing.

To meet the client's requirements I needed to print right away, so I did. While it's common for even tiny flaws in ink delivery to be obvious in prints, in this case the prints looked good. I suspect the VM channel cleared as I was printing the black-and-white image, but I didn't run a nozzle check after the job to confirm.

Nine days later I set up to print another job. The nozzle check showed most of the VM channel missing. Because I rarely get a good result from standard cleanings, I started with a powerful cleaning of the C/VM pair. This left a tiny missing spot in the middle of the VM channel. Once again, I ran the print job, finding no problems with the output.

Since that job I've run three more, on the 13th, 15th, and 19th of December. The nozzle checks have all been perfect, including the VM channel.

Looking back through the nozzle check prints I'd run in the past several months I see the VM channel has been troublesome far more often than any other. It's a concern, something to watch. But to put this into context, most of my nozzle check prints show no problems with any channel. "Clogs" happen, but they aren't typical. Perhaps they only happen when I'm in a hurry.

  --Jay