Printing Article part V - Imageprint by Dan Wells

[IPDOUGLAS]

I am sure many of us have been enjoying the series of Printing articles by Dan Wells now up to part V.  I certainly have.   I am a bit disappointed in part V as it is so software specific on Imageprint.   Also I fundamentally disagree with some of what Dan has written, firstly he states that Imageprint at $399 is the most expensive software most will have purchased - does he forget how much the perpetual licences for Adobe software cost?  I certainly paid over £600 (yes £s not $s) 15 years ago!  Secondly he states that updates every few years at 1/3 of purchase price are cheap?  I suppose all this is relative to an individual?

More fundamentally, since I am highly unlikely to need a RIP or shell out as much as $399 for one with others being cheaper I feel his article series at this point in its lifecycle might have covered a few RIPs rather than one specific brand.  Perhaps he will?  I know a few seem to like QImage at a cost of peanuts in comparison and I am sure there are others that might be considered.

I do enjoy articles that go in-depth to the Art of Printing and that includes Dans so do keep this interesting and valuable work up, great stuff!

[Dan Wells]

I've looked a little at Qimage, and it has two major issues compared to IP (it is cheaper, and as such, worth looking at - I just haven't gotten around to a thorough test, but I mean to). One is that the Mac version is a bit of a stepchild compared to PC - for many years, it didn't even have one, now it has a feature-incomplete third-party port. That port is likely to go away with Apple Silicon Macs (I have NO inside information here, just guessing because the port is a bit half-hearted).

The second differenceis that ImagePrint is an enormously research-based program. John Pannozzo of ColorByte has been involved in inkjet printing from the very, very beginning, and one reason ImagePrint is somewhat expensive is that he's forever trying to squeeze the most out of the hardware. Qimage is mostly a collection of off the shelf tech (not entirely, they've done some work in resizing).

That doesn't mean it's not valuable, and, as I've said, I mean to take a closer look.

As for the price of ImagePrint, yes , the old Adobe perpetual licenses were massively expensive (and came with ~30% ANNUAL upgrade costs, with huge penalties for missing one - skipping one version meant that the upgrade went from 30% to 50-60%, as I recall, and skipping two meant no upgrade at all?). I didn't mention them because that hasn't been the case for years.

As a percentage, ImagePrint's upgrade cost IS very low - the relatively expensive Capture One offers a choice of subscription or 30%+ annual upgrades. Most other raw converters charge roughly 50% for annual upgrades (on a much lower baseline). Adobe is subscription-only, which you can think of as 100% annual upgrades (although they lowered their baseline cost significantly when they did that). 30% or so every two or three years is much lower than any of those.

Dan

[digitaldog]

The Mac version of Qimage is excellent, the resampling and sharpening is well worth a download of a demo for people to actually test.
The combo of Lightroom and Qimage (which has a plug in for LR) is compelling and price affordable.
As for IP, I've been a past user, beginning in the days when it ran under OS9. And it actually took days to get it properly installed! Major PITA at the time.

[Dan Wells]

ImagePrint's setup and installation is MUCH improved in the recent versions.

[digitaldog]

ImagePrint's setup and installation is MUCH improved in the recent versions.

Compared to an OS9 install, yes. Utter nightmare.

[Rand47]

. . .
As for IP, I've been a past user, beginning in the days when it ran under OS9. And it actually took days to get it properly installed! Major PITA at the time.

Andrew,

How do IP's ICC profiles compare to good custom "my printer specific" profiles?  Say, those compared to some "dog"gone expert? 

Rand

[digitaldog]

Andrew,

How do IP's ICC profiles compare to good custom "my printer specific" profiles?  Say, those compared to some "dog"gone expert? 

Rand

Been year's but mine were better and theirs had issues with blues. Maybe John fixed this issue; he was certainly informed at the time.

[Dan Wells]

The current IP profiles are the best I've seen (including handmade ones with a lot of patches).

Disclaimer: I haven't worked with Digitaldog profiles.

[Binartem]

Dan,

We welcome your future review of Qimage Ultimate and Qimage One. Once you fully review Qimage we are confident that you will see the benefits.

Just to address the points that you mentioned about Qimage One specifically.

Functionality -

Qimage on Windows has over 20 years of development and features invested in it, including working with both photographers and printer manufacturers to discover how to optimize both quality and consistency across all printer models. It has been around since the first consumer digital cameras such as the Kodak DC40 (1995) were released and there was a need to print to some of the earliest consumer inkjet photo printers.

When the opportunity arose to develop a macOS version of Qimage, the decision was made to concentrate only on the printing features otherwise the port would have taken many years to complete. Since its release in January 2018 we have continued to add new features to make Qimage One even more useful. We would not categorize this as “half-hearted”, rather a reasonable business decision to deliver a useful cross-platform software application within a reasonable time frame.

Development -

Binartem, Inc., which develops Qimage One, is solely owned by Mike Chaney (author of Qimage Ultimate) and Andrew Wilford (a cross-platform developer with 25 years experience). They are close friends and neighbors and continue to work closely together to develop Qimage One. This is not a port which has been farmed out to some unknown third party. It benefits from all of the core algorithms and image processing that are present in Qimage Ultimate.

Apple Silicon -

In case you are not aware, Apple have already announced a transition for apps written for macOS to support their new chips:

https://www.apple.com/newsroom/2020/06/apple-announces-mac-transition-to-apple-silicon/

So the issue of Apple silicon is not relevant and has no bearing on the continued development and support of Qimage One. We have continued to stay aware of the changes from Apple, such as when we updated Qimage One to support our customers through the challenging transition to macOS Catalina.

If one were to speculate that this transition by Apple would result in one application to “go away” then one would reasonably expect it to affect all existing macOS apps, including the version of ImagePrint for macOS. Yet this concern was not raised in your review of ImagePrint.

We can assure you that Qimage One will continue to be developed on both macOS and Windows.

Mike Chaney
Andrew Wilford

[Dan Wells]

This is GREAT news - many developers I have talked to are worried about the transition (and I think we're going to lose quite a bit of software that gained half-hearted Mac versions when the Mac went Intel). The fact that the QImage folks are not worried, and are committed to the Mac is wonderful news. The decision to concentrate on core features, as Mike explained it, also makes sense. I've also talked to John at ColorByte (ImagePrint) about this, and he was very reassuring as well - we shouldn't be losing core print applications. I'm not a developer myself, so I don't know how to evaluate the degree of difficulty.

I'm VERY glad to hear of the relationship between Qimage Ultimate (Windows) and QImage One (Mac). Since they are different companies, I didn't realize the developers were friends and neighbors. It looks (from the information available on the web) like a typical third-party port - I'm glad to hear that it is really a first-party situation. This reduces my concerns about the Mac tremendously (and moves QImage up in the review queue)...

Dan

[Binartem]

Sounds good.  I did want to clarify one point on the topic of porting.  When people think of a port, they imagine giving a third party some Windows code, they run it through some "universal translater", put on a few bandaids, and just hope it runs well on a Mac.  We took the time to develop the Mac version on the Mac, using Mac tools, from the ground up.  This allows us to get the same quality results while also optimizing for macOS and its features.

[Rand47]

The current IP profiles are the best I've seen (including handmade ones with a lot of patches).

Disclaimer: I haven't worked with Digitaldog profiles.

So, if I understand correctly, what you’re saying is that generic profiles that are not machine specific are as good as or better than custom, machine specific profiles created by competent color management experts?  I don’t get it. 

Rand

[Dan Wells]

The porting process Mike specifically says they didn't do is what I thought might have happened to QImage until he stepped in here to clarify the actual process. The fact that the Mac version was made with such care, and with Mac tools bodes REALLY well for QImage on Apple Silicon.

There are two kinds of applications that may not do as well on Apple Silicon - one is the halfhearted Windows ports translated by third parties (many of which are games). The second is really complex software with older, weird code bases. Much of the Adobe Creative Suite is a bear to do anything with - I have heard rumors that there is some assembly language code in Photoshop, or was until relatively recently. If there is any left, it will have to be ported to ARM by hand, almost by definition.

[Dan Wells]

Replying to Rand, the ImagePrint Black profiles are very, very good - MUCH better than any paper manufacturer profiles I am aware of, and better than anything I ever made by hand, using very large targets created by experts (I used the really big Bill Atkinson targets in the 4000 patch range) scanned on an i1 Pro.

I haven't used a digitaldog profile, or any profile tweaked after scanning by an expert. Those might be better, but ColorByte's profile library is a huge advantage. If you print on only one or two papers, really top professionally made profiles are certainly an option. If you print on a bunch of papers, ColorByte having profiles for most things (and being willing to make what they don't have) is a massive advantage.

Today's printers have built-in linearization, and are a lot closer machine-to-machine than they used to be. 10-15 years ago, any acceptable profile had to be built on the individual printer. Today, not only are individual printers of the same model much closer than they've been, printers of closely related models (width variants that use the same inks and heads) are also extremely close.

[Rand47]

Replying to Rand, the ImagePrint Black profiles are very, very good - MUCH better than any paper manufacturer profiles I am aware of, and better than anything I ever made by hand, using very large targets created by experts (I used the really big Bill Atkinson targets in the 4000 patch range) scanned on an i1 Pro.

I haven't used a digitaldog profile, or any profile tweaked after scanning by an expert. Those might be better, but ColorByte's profile library is a huge advantage. If you print on only one or two papers, really top professionally made profiles are certainly an option. If you print on a bunch of papers, ColorByte having profiles for most things (and being willing to make what they don't have) is a massive advantage.

Today's printers have built-in linearization, and are a lot closer machine-to-machine than they used to be. 10-15 years ago, any acceptable profile had to be built on the individual printer. Today, not only are individual printers of the same model much closer than they've been, printers of closely related models (width variants that use the same inks and heads) are also extremely close.

Dan,

Thanks for the response / perspective.  I’d be interested in Andrew’s take on this. 

Rand

[Binartem]

I find that a lot more people are choosing to create their own profiles because they like to be able to use third party papers or even third party inks.  For years I've been using my relatively inexpensive spectro (about $300) with Argyll (free) which allows you to print as many test patches as you like.  Test prints are a perfect match across different printer models and paper (within the limitations of the paper and ink) and match the profiled monitor.  They also matched prints printed by a local printing company on the same printer/paper using ImagePrint Black.  Not sure how to quantify something that is "better than" a perfect match.  I guess it comes down to where you want to spend your money and time.  I find more and more people being willing to take the step of learning how to create their own printer profiles.  The price of IP Black leaves a lot of leeway to buy another solution plus your own profiling tools to get profiles for your printer, paper, and ink, and still have a lot of money left over.

[JRSmit]

Replying to Rand, the ImagePrint Black profiles are very, very good - MUCH better than any paper manufacturer profiles I am aware of, and better than anything I ever made by hand, using very large targets created by experts (I used the really big Bill Atkinson targets in the 4000 patch range) scanned on an i1 Pro.

I haven't used a digitaldog profile, or any profile tweaked after scanning by an expert. Those might be better, but ColorByte's profile library is a huge advantage. If you print on only one or two papers, really top professionally made profiles are certainly an option. If you print on a bunch of papers, ColorByte having profiles for most things (and being willing to make what they don't have) is a massive advantage.

Today's printers have built-in linearization, and are a lot closer machine-to-machine than they used to be. 10-15 years ago, any acceptable profile had to be built on the individual printer. Today, not only are individual printers of the same model much closer than they've been, printers of closely related models (width variants that use the same inks and heads) are also extremely close.

Would be interested to see some elaboration on why the IP profiles are “very very good” , what specifically makes them so much better. Profiles and the printed results can be measured, so perhaps some substantiation can be provided?

[Ernst Dinkla]

No mention so far of Imageprint's license limitations on printer sizes, two versions for Imageprint R.E.D. for example, which in essence uses the same OS drivers as QU does. Qimage Ultimate license has no limitations in that sense.

Met vriendelijke groet, Ernst Dinkla

https://www.pigment-print.com/spectralplots/spectrumviz_1.htm

[JRSmit]

The porting process Mike specifically says they didn't do is what I thought might have happened to QImage until he stepped in here to clarify the actual process. The fact that the Mac version was made with such care, and with Mac tools bodes REALLY well for QImage on Apple Silicon.

There are two kinds of applications that may not do as well on Apple Silicon - one is the halfhearted Windows ports translated by third parties (many of which are games). The second is really complex software with older, weird code bases. Much of the Adobe Creative Suite is a bear to do anything with - I have heard rumors that there is some assembly language code in Photoshop, or was until relatively recently. If there is any left, it will have to be ported to ARM by hand, almost by definition.

So? What are you implying?

[digitaldog]

Dan,

Thanks for the response / perspective.  I’d be interested in Andrew’s take on this. 

Rand

One can build custom ICC profiles through ImagePrint (at least in the old days, I could).
One CAN compare IP's canned profiles to a custom profile and that's what should be done.
This has very little to do with 'number of patches'; more isn't necessarily better. It has a lot to do with the color engine used to build the profile. In the days I compared profiles thorough IP, John's in-house profile package had an issue rendering saturated blues (they would shift magenta) which is a common issue for some profilers and one of the warts in Lab color space. My profiles didn't exhibit this through IP nor anything else, I'm using the X-rite color engine. It's quite easy to see this if still producing such a color shift.
Someone with IP who wishes to know if their profiles are the 'best' or not therefore needs to build a custom profile through IP using whatever specified software product, sure, specify number of patches and Spectrophotometer, settings used to build a Perceptual table, then output a number of good color reference images with RelCol and Perceptual tables, and share the results. One could also compare the soft proofing tables of the two, less an issue but to some, pertinent.
Then there are colorimetric tests one can preform on profiles where one measures output and compares the dE to a reference for all tables. That takes a lot more work and some specialized software (like ColorThink Pro) but to some, useful for actual reporting of profile qualities. But at the very least, an Apples to Apples visual comparison using good color reference images, at best, the entire Roman 16 suite and a few others if possible.