I've been experimenting with trying to get the optimum print settings wrt detail/speed, for my printer and came across this web site http://www.inkjetart.com/tips/ppi/index.html (and many others).
 
I would like to suggest that you print out the two pdf files linked from the inkjetart site, and examine the print with a high power magnifying glass. I am particularly interested in the results for Epson printers. Do not print from Qimage or anything other than the free pdf viewer print function. (If you really must use Qimage, then you will need to turn off sharpening, resizing and most everything else - but for your own personal amusement you could leave those settings as normal, but I'm not interested in those results at this time). For my 4000 printer, 180PPI and 360PPI were both excellent, and 240PPI not too bad, others had the expected artefacts and everything above 400PPI gave a pink tone (this was using 2880*1440dpi print settings.) I am particularly interested in the results for the 2200 series printers.

Best wishes,

Ray

Hi Terry,

I guess you are not interested, then? Worried what you may see?

Best wishes,

Ray

Terry.

That all sounds very interesting and informative, but, does it have any relevance to 'print to file' for lab printing?

I usually upload my prints to Fugi for printing.  I used to use ProAm who required 402dpi, fugi don't seem to mind.

Jeff

Terry,
Sorry for butting in, but!
This has been one of my pet irritations for years.
We would have debates at our club meetings week after week and the folks that cut their teeth on Photo Shop just never understood that ppi means pixels per inch, and without inches specified, you don't have ppi.
As I have seen, Photo Shop opens images with an arbitrary selection of a ppi. Lately they tell me the default is 180. It used to be 72, and the initial print size showed in Photo Shop as 42 x 28 inches (something like that depending on the image resolution divided by 72.
This was not a real print size,,,, just something to reflect the arithmetic.
Then, based on the print ideas of the '80s where 300 DPI was the goal, they worked their images into a size they wanted and interpolated the ppi tp 300 and then later, 360.
Interpolation is the very essence of digital printing.
Interpolation is inescapable (except in an extremely rare perfect match). Interpolation is the reason why Qimage makes the absolute best prints.
The interpolators in Qimage were written by Mike Chaney to be better that the run of the mill interpolators like Bicubic.

In simplified terms, interpolation will add pixels in the correct proportion, correct colors, to bring the number of pixels to up to match the native input of the printer.
In a sense it is "faking" the image for the sake of being able to choose the size print we want.
So we don't want to fake too much as that is not going to improve our picture after a certain point is reached.
Since Qimage has the best set of interpolators and Qimage's algorithms utilize them to perfection, I submit to you that more than  *ONE* interpolation is  degrading your image. So why do it twice.

It seems that old batch of seaweed I must have 300 or 360 PPI washes up on the shore from time to time.
Qimage takes your image (and remember what I pointed out earlier, there is no ppi until we know the i,) and depending on the size of your intended print, now knows the ppi and will interpolate to the native input resolution of *your* printer.
I have yet to see the instance where (making your own prints) anything is gained by interpolating to 300 and then redoing the interpolation at print time.
Let's take an extreme example of a shot you took of a Heron at a distance, and after cropping you find that your 11 x 14 print is showing that it will be printed at 105 ppi.
You will be amazed at the print Qimage will make for you because of the quality of the Fusion interpolator, plus the one time interpolation done at printing time.
If you choose to get that image to 300 ppi first because there was a note in a bottle that washed ashore with that old seaweed, then it will be interpolated again at print time.
That's fake pixels on top of fake pixels.

Mike Chaney is the recognized printing expert and digital image expert on the planet.
He sells Qimage not to compete with CS4 etc, but to give all digital photo producers the best prints that can be made. That's why there are so many professionals, and amateurs alike who print with Qimage regardless whether they start their processing in CS4 or Paint Shop Pro etc.
The best kept secret, it seems, is that Qimage will allow a user to process his work (Raw shots as an example) better and easier than Brand X at 750.00 a pop.
That's why people like Terry and Fred and many other converts are trying to tell anyone who wakes up in 2010, to try Qimage for other things besides printing.

Sorry for the long dissertation. It's just a passion with me, and I love my Qimage.
I took 71 raw shots yesterday at the waterfront in Sarasota, Fl.
I have a problem with Qimage. (I am tongue in cheek serious)
I usually use the blue filename as an indicator that I already checked and adjusted that image if needed. If a filename was still black, it meant that I hadn't looked at it yet
Lately, Qimage is decoding my Raw shots so perfectly that they don't need any fill  or exposure adjustment at all, and I find that I am giving a -1 or a +1 fill just to turn the filename blue.
I have to talk to Mike about that!  

When I was a youngster, my dad would take my brother and me to the Polo Grounds in New York City to see the Giants play baseball. Their star player was Mel Ott.
He was a terrific ball player, but better yet was the light he invented.  
I got an OTT LITE finally. I usually depend on the Sarasota sunshine for my natural lighting to judge prints, but this thing is wonderful!
Why didn't you tell me it was this good.?

Fred

It is quite funny, if it wasn't so sad. It was a simple request. I was looking for folk to actually try something for themselves, a simple printing exercise, to take a few minutes. I was interested in a number of aspects, depending on the results. I posted here, because I thought there may be folk interested in the mechanics of printing, interested in seeing how this stuff works for themselves. I mentioned Qimage, since, unless you set most things off, it will try and compensate? for some artefacts that occur.

Instead of actually trying the simple tests, in the manner I suggested, we get the usual load of unrelated twaddle about how wonderful Qimage, Mike, father Christmas and all are. What a load of ***! Just try printing the images, report what you find, or stfu. Don't hijack the thread.

On the page referenced, there are tif files for the separate patterns, which you may have been tempted to print in Qimage, and there is plenty of software that can convert pdf's to a format that Qimage can handle, that is why I suggested not using Qimage, since you are unlikely to know how the file conversion works, never mind what artefacts Qimage interpolation introduces, unless you turn it all off.

Not everything is about 'how wonderful Qimage/Mike is'. Open your minds, think outside the box. A number of folk use other software too, for specific purposes. This section is headed printers. It is not about Qimage, it may not be even about printing onto normal media, maybe not even printing at all, as you know it, - but thanks to some, it is likely to  never be about anything on this topic, thanks to the way this has been sent off course, by folk who think they have the answer to everything.

I hope I have irritated you guys in the way you have irritated me, by your hijacking of the thread.

Best wishes,

Ray

What he said, plus...
Ray, I have no idea where you get the idea that you have been thwarted in some way.
No one is stopping anyone from doing the tests you suggest, nor has anything been said referring to your tests.
As Terry said, we are Qimage enthusiasts and after so many years of experimenting and comparing, we KNOW that Qimage does the best job.
Right on target! Why reinvent the wheel?
Artifacts? Qimage offers an artifact free Hybrid SE, and an anti aliasing slider when downsizing to eliminate artifacts, if you so desire.
Oh what's the use.
You feel threatened, and I don't know why.
Have a fine day!
Fred

Wow, I'm going to address the original question and ignore the rest.  I'm not able to test an Epson as I have an HP and a Canon at the moment and no (working) Epsons.  I'm also not able to test HP/Canon properly because no driver-resolution (600 PPI) test is provided at that link: quite an oversight I should think given that all Canon and HP printers run at either 600 or 1200 PPI.  I can provide some insight into the methods, however, and I have run very similar tests in the past so let me give you my findings on this subject.  I did extensive testing on this years ago and have had a similar test online for many years.  Just click the "Quality Challenge" link from the Qimage web page and the test charts are at the top.

The bottom line is that all printers produce the best quality at their "native" setting.  For Epsons (since you asked), this is typically 720 PPI except for wide format printers that run at 360 PPI unless you specify "Finest Detail" in the driver, at which time you'll get the full 720 PPI.  Manufacturers specify this native resolution because their halftoning algorithms work based on the native resolution.  For example, if you pass 720 PPI data to an Epson printer that prints at 5760 x 2880 DPI, the halftoning algorithms take your 720 PPI input data and map that to an array of 8 x 4 (32 total) printer dots.  Manufacturers usually have baseline halftoning algorithms that work on the native PPI, meaning that for an Epson, whatever you send it is going to end up at 720 PPI before the dots get placed on the page anyway!  This is why sending non-integer-multiples give drivers a problem.  If you send 300 PPI to a 720 PPI driver, or even something more odd, like 518 PPI, the driver has trouble because it's interpolation algorithms are very basic.  In fact, sending 518 PPI can be worse than sending 360 PPI!  For Epsons, if you send 360 PPI, it is almost as good as 720 PPI but not quite, on most papers.  If the printer/paper couldn't handle any detail at all past 360 PPI, the driver wouldn't request 720!

That link in your original post is simply showing that drivers do best at the native resolution reported by the driver (the 720 you see above the preview page in Qimage) or simple fractions thereof (1/2, 1/3, 1/4, etc.).  So an Epson runs well at 720, 360, 240, and so on.  Canon's run well at 600, 300, 200 and so on.  Why pick the largest: the "native" resolution, for Epsons, usually 720?  Because while there will generally be no real artifacts at 360, there is some resolvable detail that the printer can produce between 360 and 720 PPI.  So we run it at max (720) to get the most detail possible?  There is no down side, and no discernible artifacts caused by printing at 720 versus 360.

Mike

Thanks Terry,

Did you try printing the pdf's from the acrobat reader print function? iirc you have a r1800, which will give higher resolution than my 4000. Among other things, At this time I  am interested in what happens when it attempts to print beyond its native resolution. On close inspection, I get the pink tone that I mentioned starting at 380, but at 720 upwards there is hardly any black ink at all. I would not expect that to happen with your printer with these charts - maybe beyond 1440lpi?

Thanks Mike for your explanation. I'd seen it before, or similar. It is why I bought your software a few years ago, having tested it out for myself with your demos.

Best wishes,

Ray