[George Stark Participant Posts: 72]

Filter Failed issue

on: October 25, 2020 at 11:43 pm

OS-10.15.7 Catalina
Mac 5k 27in Late 2015
Epson P900
latest up to date versions of PS and LR

Before the P900 I was using the Epson 3880 with Mojave….

I read Mark’s article and it gave me the shakes…ha.. I’ve experience many strange printing issues over time but I have not experience that one…. and I’m crossing my fingers I don’t hex myself here!!

I was just curious if you’ve found a common thread among those who experience the issue?..

Same model of Mac… Same versions of software… specific model of printers ?

When printing from LR with Mojave and the 3880 …. I found one day after running a Mac system update that suddenly my prints were coming out a big mess… It looked somewhat like it was being double color managed… after many ruined prints I found by accident that for some reason if I checked the 16bit box in the Epson Printer Dialogue and the 16 bit box in Lr that was the cause of the issue….If I did not check off that box in Epson’s dialogue all was well…. From all my research on that issue it appeared as if it was a very small group of folks that had this issue…. it is so hard to pin these issues down if they are not consistent across the board…. can’t help but think other things on our computers … Plug-ins etc… could be the culprits ….

I notice the New Epson Driver for the P900 does not contain the 16bit box to check off any more.

Thanks for the heads up hope you can figure this out…

[Mark D Segal Silver Member Posts: 951]

Re: Filter Failed issue

Reply #1 on: October 25, 2020 at 11:44 pm

Glenn, yes you have a point – it may not be limited to Adobe applications, as your Epson Print Layout experience shows it could affect other applications that also use an ICC-based colour managed workflow needing to interface with MacOS. Epson Print Layout depends on the MacOS printing system and allows an ICC-based colour managed workflow. But interestingly, the LaserSoft Imaging application PrinTao 8 application also accommodates ICC-based colour management and does not exhibit this problem at the same time that Lightroom does. So just when you think the problem is nailed by one unique set of printing conditions, you find that it is not – there are contrary indications. But CUPS within MacOs is common to all scenarios and that is why I believe a fundamental solution to this issue needs to start there.

[Mark D Segal]

Re: Filter Failed issue

Reply #2 on: October 25, 2020 at 11:52 pm

The one common thread that is unambiguous is that it happens only on MacOS, not Windows. It could also be that certain image editing applications that support printing (not printer drivers) providing for ICC-based colour management may be hosting this problem in conjunction with MacOS. So for example Lightroom, Photoshop. But insofar as it appears to happen with certain non-Adobe colour-managed printing applications as well, this may not be as unambiguous a common thread, which doesn’t rule it out as contributory.

And no, I cannot definitively figure this one out. I think deductive logic has taken me about as far as I can go. The more people who can contribute on the problem in this Forum, the more possibilities we may be able to rule in or out by applying logic, but basically, the folks with the engineering degrees in printing systems and related code will need to put their minds to it.

• This reply was modified 3 years, 10 months ago by Mark D Segal.

[Pekka Riikonen Participant Posts: 10]

Re: Filter Failed issue

Reply #3 on: October 26, 2020 at 6:00 am

Mark,

The “filter failed” is a generic CUPS error and can have many many different root causes.  Only way to try to figure out what the reason is in your case is to check your CUPS error log.  So, let’s tart with that:

– Launch Terminal
– Give the command: cat /var/log/cups/error_log

And then copy-paste here what you see…

[Mark D Segal]

Re: Filter Failed issue

Reply #4 on: October 26, 2020 at 9:41 am

Pekka Riikonen wrote:

Mark,

The “filter failed” is a generic CUPS error and can have many many different root causes.  Only way to try to figure out what the reason is in your case is to check your CUPS error log.  So, let’s tart with that:

– Launch Terminal
– Give the command: cat /var/log/cups/error_log

And then copy-paste here what you see…

Thanks Pekka, here we go: not sure this will help because I have been only using the P900 I now have under evaluation and there have been no Filter Fail incidents yet. But it’s good to have this Terminal command string, thanks so much; next time I encounter the error I shall copy the same log and post it. So far from all I have been able to deduce I think you are correct that there is a generic CUPS problem, but I’m wondering whether it is the only problem insofar as some applications allow printing and there is failure with others.

(Edit Note: I deleted the content of the log because it was far too long and it’s purpose has been served.)

 

• This reply was modified 3 years, 10 months ago by Mark D Segal.

[Tommy Weir Participant Posts: 27]

Re: Filter Failed issue

Reply #5 on: October 26, 2020 at 9:59 am

I suspect your fingering of CUPS is correct.  More than once this year I’ve had to reset my printing system and add printers anew.  Like Glenn, I’ve had this error message via Epson Print Layout.   First I’ve heard of PrintTao, will explore.  A comparative review of the various printing applications might be of interest too many.

[Mark D Segal Silver Member Posts: 951]

Re: Filter Failed issue

Reply #6 on: October 26, 2020 at 10:34 am

Tommy Weir wrote:

I suspect your fingering of CUPS is correct.  More than once this year I’ve had to reset my printing system and add printers anew.  Like Glenn, I’ve had this error message via Epson Print Layout.   First I’ve heard of PrintTao, will explore.  A comparative review of the various printing applications might be of interest too many.

Hi Tommy, I did review PrinTao 8 back in the LuLa days. It should still be posted there. It’s a very good application, but with the limitation that it cannot print from raw files in the way Lightroom can.

The problem with reviewing applications is that within each there are a number of printing conditions, so the matrix of factors to compare could be a lot to handle. If you print from a number of applications not using ICC colour management there is no filter failed problem at the very same time the printing system produces this problem when using application-managed ICC-based colour management (except in the case of PrinTao 8 which also uses ICC-based colour management). We also know it is a Mac and Linux only problem, not Windows. Mac and Linux use CUPS. So I think Pekka’s observation that this is fundamentally a CUPS problem is correct, but it could also involve the front-end applications it is working with in a manner that isn’t wholly consistent, so it’s hard to untangle.

[Glenn Kroeger Participant Posts: 15]

Re: Filter Failed issue

Reply #7 on: October 26, 2020 at 10:52 am

It’s hard for me to fathom how CUPS could be impacted by the choice of color management. I would think Lightroom/PS/EPL do the color conversion if requested, so the only thing CUPS needs to know is whether or not to use ColorSync. It’s hard to see how that binary software switch could be so problematic.

Epson really needs to be the driving force behind diagnosing and fixing this issue. They have software engineers, presumably debugging hooks into CUPS, and most importantly, a vested interest in people being able to use printers!

I am sure Mark has better connections in Epson, but my conversations with tech support suggest that they haven’t a clue about how serious this issue is,  or how to begin diagnosing the problem… they just throw suggested fixes against the wall and hope something sticks.

[Pekka Riikonen Participant Posts: 10]

Re: Filter Failed issue

Reply #8 on: October 26, 2020 at 10:56 am

Thanks.  Yeah, next time you get the error, then go and check the log again.  As you can see there’s dates and times in each line so you can look at the dates that are relevant for the time period when the error happens.

It’s not so much that it’s (necessarily) a CUPS problem but rather that there are so many different components that have to talk to each other and any kind of error at that point often results to the “filter failed” error.  Clearly the error reporting should be improved.  The error_log contains often detailed information what’s going on.  No use to guess what the issue might be (and there may be more than one issue), will just have to wait until it happens again.

[Mark D Segal Silver Member Posts: 951]

Re: Filter Failed issue

Reply #9 on: October 26, 2020 at 11:36 am

Glenn Kroeger wrote:

It’s hard for me to fathom how CUPS could be impacted by the choice of color management. I would think Lightroom/PS/EPL do the color conversion if requested, so the only thing CUPS needs to know is whether or not to use ColorSync. It’s hard to see how that binary software switch could be so problematic.

Epson really needs to be the driving force behind diagnosing and fixing this issue. They have software engineers, presumably debugging hooks into CUPS, and most importantly, a vested interest in people being able to use printers!

I am sure Mark has better connections in Epson, but my conversations with tech support suggest that they haven’t a clue about how serious this issue is,  or how to begin diagnosing the problem… they just throw suggested fixes against the wall and hope something sticks.

Glenn, much of colour management is actually handled by the CMS (colour management system) which is part of MacOS. As I mentioned in my article, in the case of MacOS, photo editing applications that have a printing function need to add a special interface for ICC-based colour management to work with the MacOS printing system (CUPS). These algorithms need to talk to each other before the data emerges from the “Back Ends” of CUPS to reach the printer driver. As for the printer drivers, and as I mentioned in the article, the “Filter Failed” problem affects many major printer brands, not just Epson, so there is no particular reason for Epson to be the stand-alone “driving force” behind diagnosing and fixing the issue. You are right they have a vested interest in people being able to print in a trouble-free manner and I have good reason to believe they have taken a more serious interest in this issue behind the scenes than perhaps those other parties in the industry who really do have this problem under their management and have apparently done nothing to resolve it. I also have good reason to know that if Epson thought they could have fixed this problem they would have done so long ago. As for tech support “throw(ing) suggested fixes against the wall” and hoping something sticks”, I think this is unfair to tech support. They are only doing what they are trained to do and this reflects the best that the engineers in the companies know what to instruct them. The problem is that “nobody knows but their hairdresser”. And the fact is (because I’ve been there, dun dat”), if you follow their advice doing a bunch of stripping out and resetting, that fix can and does temporarily work, as I also mentioned in the article, only for the failure problem to return later on. We don’t know what-all gets changed-up in these resets and why they get derailed after further use. If we did, the problem would likely be easy to pinpoint and resolve. To use an unpleasant contemporary analogy, this problem strikes me a bit like the interval in COVID-19 between the time you get infected and the time you develop symptoms and get tested – the disease is there, lurking but nobody knows it, they can’t pinpoint how they got it, and it’s waiting to pounce.

I think perhaps I should clarify that my purpose in publishing this article is not to assign blame, but at the same time to identify the paths where logic takes one, hoping more adequate information can be developed from there. I had three objectives in mind: I wanted to (1) elicit useful information from anyone who has any substantive insights into what these root causes may be, (2) collect additional experience from affected people in an effort to further refine and improve our insights into the paths where the problem may lie and (3) to incite Apple Computer and Adobe Inc., in particular, to start taking this issue seriously. It can all be avoided buying a Windows computer, but that has other issues………..

[Glenn Kroeger Participant Posts: 15]

Re: Filter Failed issue

Reply #10 on: October 26, 2020 at 11:48 am

Mark:

Thanks for these details. I have no way of knowing about other printer makes, but I assumed from your article that the problem wasn’t limited to Epson. Still, Epson should be pushing Adobe and Apple as hard as possible given the stakes involved.

As for “stripping out and resetting”, I have had no luck. And, as I mentioned, I have done this multiple times on three separate MacBook Pros running both Catalina and Mojave and two different versions of Lightroom (2019 and 2020). I guess I could do it more, but I have read that insanity is defined as doing the same thing over and over again and expecting different results. As a coder, I find this particularly true of software.

There are some other possible solutions than buying a Windows machine. Today I am going to try using my High Sierra/Lightroom 2018 system in a virtual machine to see if that works. One could perhaps also run Windows in a virtual machine. A final question is whether available RIPS avoid this problem by avoiding all or part of the CUPS pipeline.

Glenn

[Mark D Segal]

Re: Filter Failed issue

Reply #11 on: October 26, 2020 at 12:06 pm

I like that definition of insanity! So true. And yes, the workaround I use when I get a Filter Fail is to resort to my 10-year old MacBook Pro on EL Capitan using 2018 versions of Lr and PS. That always works, much as it’s a nuisance to have to send the files over there just for printing. I can’t answer your problem about RIPS because I have no experience using a RIP and I have not seen any reports from those who have, so I just don’t know. I think it is also the case that there are different implementations of CUPS, so that may also come into play. There are RIPS that replace the printer driver, but whether they also work around CUPS is also something I don’t know. Perhaps someone who does could chime in here and advise.

 

• This reply was modified 3 years, 10 months ago by Mark D Segal.

[Author]

Posts