What is your workflow for upressing, sharpening and printing?

I don't print from Lightroom, but it has some algorithms derived from the folk at Pixel Genius which includes master printer and well published guru Jeff Schewe.

The advice is to up res goat least the next highest size of the "natural" resolution of the print head. For Epson it's 360 and Canon it's 300. So if one is below that number, up res to that number. If one is above, go at least to the next multiple, i.e., 720 for Epson and 600 for Canon. Still, it's recommended to use 1200 or even 2400 to get the finest detail out of the file and this is especially valuable in gloss paper where it can show.

I saw Jeff Schewe's examples of printing at either 360 or 300 on the Epson and Canon respectively or their highest resolutions. There was a much cleaner appearance of detail especially in diagonals.

But now I have discovered Nik Sharpening and getting to wonder how best to take advantage of the newest tools for preparing a picture for printing.

Asher

Hi, Asher,

Asher Kelman said:

I don't print from Lightroom, but it has some algorithms derived from the folk at Pixel Genius which includes master printer and well published guru Jeff Schewe.

The advice is to up res goat least the next highest size of the "natural" resolution of the print head. For Epson it's 360 and Canon it's 300. So if one is below that number, up res to that number.

Click to expand...

That makes sense, but we have to be careful in implementing that. Of course, it matters what software one is using as to how one steers this odd boat.

Suppose I have an image that, after cropping, has pixel dimensions 3000 px × 2400 px. Suppose I want to print that at a size of 10" × 8", from my image editor. To follow the guidelines you mention, I should uprez it to 3600 px × 2880 px.

Now to do this in the image editor I use, I don't set anything to 360 px/inch. I set the resize dialog to 3600 px wide (and tell the machine to hold the aspect ratio).

When I do that, I have the image in memory with dimensions of 3600 px × 2880 px, and then maybe I do some other things to it.

It might be that in some editing software (perhaps with the P-word), to do this this one sets something to 360 px/inch. But not mine.

Actually, I almost never print out of my editing software. I print with Qimage. In the example above, I load the image (3000 px × 2400 px) into the print cue, tell the machine that I will be using my Epson R1900, tell it that I want the finished size to be 10" × 8", and turn it loose. It mediates between the pixel dimensions of the loaded image and what it knows the printer would best like to see, resizing (perhaps "uprezing") "on the fly", to suit.


But I never tell it what pixel dimensions the image, as sent across the interface to the printer driver, should have. Nor do I get any indication of what pixel dimensions the program chose.

Best regards,

Doug

Doug,

I think you are using Picture Perfect, perhaps? In Photoshop, however, changing pixels resolution really is done when you agree to those conditions and press, OK".

Asher

Hi, Asher,

Asher Kelman said:

Doug,

I think you are using Picture Perfect, perhaps?

Click to expand...

Picture Publisher 10.

In Photoshop, however, changing pixels resolution really is done when you agree to those conditions and press, OK".

Click to expand...

How handy. There is so much to be said for an established religion.

Is this done with the Image Size dialog? If so, I can't figure out how to work it in the context under discussion.

I have loaded an image with pixel dimensions 3000 px × 2400 px. I want to resize it so its resolution will be 360 px/inch. Of course that has no meaning unless I tell what dimensions I want to print it (an image file by itself has no resolution.).

So we'll say 10" x 8". I guess I enter that in the Document Size panel, is that right?

Then what do I do?

Thanks.

Best regards,

Doug

I have been reading up on the new Qimage software, Qimage ultimate. It appears that the last year or so, a new company and greatly expanded team of engineers has completely rewritten the software and it seems very sophisticated in applying sharpening according to print size, paper and the printer used. It claims to assess the image itself for what needs sharpening in "deep sharpening" and so far, just from the videos, I'm very impressed. At $69 it wont break the bank and it is able to send work directly to the printer or print to files to be given to the print shop.

Good thing is that when one wants a different size output it goes back to the source file and redoes the calculations, all in the b.g. It seeks to match the print result to the particular monitor one is using.

Seems pretty impressive.....just for Mac users, it has to be installed in Parallels, but it does require a windows shell to work in! I'll have to find my PC software, LOL!

Which of the Windows versions should I use?

Asher

Hi, Asher,

Asher Kelman said:

I have been reading up on the new Qimage software, Qimage ultimate. It appears that the last year or so, a new company and greatly expanded team of engineers has completely rewritten the software . . .

Click to expand...

I had not realized that. I know that the publisher is still (apparently) ddisoftware under Mike Chaney.

The new version seems very capable.

It still does not have one feature I write to Mike every year (I think for about ten years now) to plead for: a "Compact Centered" layout mode in which multiple prints on a page are tight together but the cluster is centered in the page. If for example I print 3.5" × 5" prints "four up", this makes for the easiest way to separate them with a trimmer.

As a work-around, I actually use the "Template Centered" mode to get that result, but it has a lot of other special properties that make the work more difficult (it is more difficult under that mode to get the orientation of the images in their "frames" what you want than in all the other layout modes, because this mode tries to outguess what I have in mind).

When I complain about those difficulties, Mike tells me, "That behavior comes along with the Template Centered mode, which you should not be using for the work you describe."

Indeed. I should be using the Compact Centered mode. But it doesn't exist.

The alternative is to use, for example, the Compact layout mode, in which the multiple prints on a page are indeed tight together but the cluster is up against two of the printer margins. That means trimming the prints out involves taking a small strip (the width of the printer margins) off two of the edges of the whole array, which my trimmer does not do gladly. (Yes, I can use a paper clamp bar.)

So we use the Template Centered mode and live with it.

Best regards,

Doug

I wonder why you don't print from Lightroom.

I print at home on an Epson using Lightroom. Lightroom's native print res is 240ppi but I set it to 360ppi which tells my printer's drivers to upscale the res and not me upscaling it. (I started doing this after a conversation with Cem where he told me that doing it through Lightroom's print module meant the printer driver itself would do the upscaling and it would do the best job)

As for sharpening, I do regular low- sharpening in Lightroom during post-processing, then in Photoshop, I do very selective higher sharpening using the high-pass filter and masking to get only extra sharpening in areas where it is important (therefore making sure I don't get any halos or any soft bokeh being sharpened, then allow Lightroom to do standard sharpening via the print module when I print.

I don't print gloss as it doesn't suit the type of images I do, and I don't allow the printer driver to enhance anything. I print mostly with Epson Ultra Premium Luster paper and have been very happy with the results. I do however have an older camera, and usually print on 13 x 19 with a inch border which leaves the actual print at 11 x 17 which is why detail may look perfectly fine. The fact that neither of you use Lightroom is worrisome to me, as it seems perhaps that you believe the results through Lightroom are not good?

As for print multiples, I don't use that, but, again in Lightroom, you could create a custom template and just save it - then you'd have it exactly the way you wanted it, no?

Maggie Terlecki said:

I wonder why you don't print from Lightroom.

I print at home on an Epson using Lightroom. Lightroom's native print res is 240ppi but I set it to 360ppi which tells my printer's drivers to upscale the res and not me upscaling it. (I started doing this after a conversation with Cem where he told me that doing it through Lightroom's print module meant the printer driver itself would do the upscaling and it would do the best job)

Click to expand...

The print driver is pretty dumb compared to the algorithms in Lightroom to enlarge a print. Jeff Shewe advises against allowing Mac OS or Windows OS print drivers to create size for you.

In Lightroom hold the option key down, ("Alt" in a PC), and then adjust the amount of sharpening and Lightroom (really Adobe Camera RAW) will sharpen just the essential features and not the b.g. or just the eyes if you wish. Qimage hs a similar recognition process called "deep sharpening" which is simply brilliant, to the extent I can judge from their samples.

The advantage of the latest, totally rebuilt from scratch, QImage is that each print is recorded with the details you used and if you change size in a future print, it will bring up that setting and you can alter it to your new needs and it will go back to the original image and then make decision based on the size and type of paper and the color space. For example, Hahnemülle 100% cotton rag baryta paper can hold a much high gamut of blues and greens than adobe RGB. You can send that to the printer with QImage, and that's 30" increase in gamut. Qimage knows all these things.

From what I've seen, the latest Qimage may give sharper results than Photoshop. But there are many personal workarounds and if the prints are fine, then no need to change!

Asher

Hi, Asher,

Asher Kelman said:

I don't print from Lightroom, but it has some algorithms derived from the folk at Pixel Genius which includes master printer and well published guru Jeff Schewe.

The advice is to up res goat least the next highest size of the "natural" resolution of the print head. For Epson it's 360 . . .

Click to expand...

I know this can be a confusing matter, but somehow I thought the "native" resolution of the Epson Stylus Photo R1900 (except in Draft mode) is 720 px/inch (in both directions).

With the R1900 selected (and not set to Draft mode), Qimage Ultimate shows the printer resolution as 720 px/inch.

With the R1900 selected (and not set to Draft mode), Qimage Ultimate offers a choice of output resolutions: 720, 360, 240, and 180 px/inch. The default is 720.

What am I missing here?

Thanks.

Best regards,

Doug

Hi, Asher,

Asher Kelman said:

The advice is to up res goat least the next highest size of the "natural" resolution of the print head. For Epson it's 360 and Canon it's 300. So if one is below that number, up res to that number. If one is above, go at least to the next multiple, i.e., 720 for Epson and 600 for Canon. Still, it's recommended to use 1200 or even 2400 to get the finest detail out of the file and this is especially valuable in gloss paper where it can show.

Click to expand...

I'm having some trouble understanding this.

If the native resolution of the printer is 720 px/inch, what do we gain by uprezing the image so that, at the size we will print it, its resolution is 1200 px/inch or 2400 px/inch.

I assume that, having done so, the printer driver will have to downrez it to deliver it to the "engine".

What am I missing here?

Best regards,

Doug

Doug Kerr said:

Hi, Asher,



I'm having some trouble understanding this.

If the native resolution of the printer is 720 px/inch, what do we gain by uprezing the image so that, at the size we will print it, its resolution is 1200 px/inch or 2400 px/inch.

I assume that, having done so, the printer driver will have to downrez it to deliver it to the "engine".

What am I missing here?

Best regards,

Doug

Click to expand...

Doug,

As I have started to face this issue, I have been studying as many good sources as I can find and spoken to an expert photographer and print technician mavin.

While your Epson printer can manage any input resolution, it has to fudge to print that out as it only knows how to perfectly print at its set dpi print levels. The number of pixels should fit exactly at least one of the print options designed by the MFR for that printer. For Epson 360, 720 and so forth. The printer welcomes that and no interpolations ions are needed! For Canon it would be 300, 600 or 1200 dpi. So one needs sufficient pixels in that dimension for the picture size specified, 300, 600 or 1200 pixels per inch.

If, for example, one has an image with 230 pixels/inch at the desired print size, then the number of pixels has to increased to at least 360 pixels/inc. Do not leave it to be grabbed by some print driver of the Mac or Windows OS!! We must, instead one of the sophisitcated algorithms using Lightroom, bicublic smoother ot the OnOne software option, Perfect Resize, I think it's called, (it's "modern", but still long-toothed version of "Genuine Fractals", especially great for iphone pictures to be printed 6x7, the worse the picture, the better it performs ).

Or one can use Qimage and have that optimized automatically for choice of paper, your monitor and the print size you desire on this occasion. In addition, one gets documentation of all the settings for next time!

If however, one has more than the native minimum printer specification, 360 pixels dpi for the Epson, then one is recommended to upress to the next level, 760 dpi. In any case, except for roughe/matt papers, modern semi and gloss papers can provide more detail at twice the minimum native print resolution and the best papers can show improvements at 1200 dpi setting. This is obvious especially in diagonal lines and the finest details.

Asher

Hi, Asher,

Asher Kelman said:

If however, one has more than the native minimum printer specification, 360 pixels dpi for the Epson, then one is recommended to upress to the next level, 760 dpi. In any case, except for roughe/matt papers, modern semi and gloss papers can provide more detail at twice the minimum native print resolution and the best papers can show improvements at 1200 dpi setting. This is obvious especially in diagonal lines and the finest details.

Click to expand...

If I was using Qimage, I wonder how I could force it to send the image to the printer driver at, say, 1200 px/inch?

Would I need to uprez the image file so that it would have a resultion of 1200 px/inch at the intended print size?

Thanks.

Best regards,

Doug

Let me see if I understand the recommendation:

• We will assume for this exercise that the "native resolution" of the printer of interest is 360 px/in (which is said to be so for typical Epson ink jet printers).

• We will assume that we will print from our image editor.

• If, considering the size to which we wish to print the image, the current image pixel dimensions constitute a resolution less than 360 px/in, then we should uprez the image to pixel dimensions that, considering the size to which we wish to print the image, constitute a resolution of 360 px/in, and then print from our image editor with that image resident.

• If, considering the size to which we wish to print the image, the current image pixel dimensions constitute a resolution greater than 360 px/in, but less than 720 px/in, then we should uprez the image to pixel dimensions that, considering the size to which we wish to print the image, constitute a resolution of 720 px/in, and then print from our image editor with that image resident.

• But, actually, in many cases, we should uprez the image to pixel dimensions that, considering the size to which we wish to print the image, constitute a resolution of 1200 px/in or so, and then print from our image editor with that image resident.

Is that it?

Best regards,

Doug

Doug Kerr said:

Let me see if I understand the recommendation:

• We will assume for this exercise that the "native resolution" of the printer of interest is 360 px/in (which is said to be so for typical Epson ink jet printers).

• We will assume that we will print from our image editor.

• If, considering the size to which we wish to print the image, the current image pixel dimensions constitute a resolution less than 360 px/in, then we should uprez the image to pixel dimensions that, considering the size to which we wish to print the image, constitute a resolution of 360 px/in, and then print from our image editor with that image resident.

• If, considering the size to which we wish to print the image, the current image pixel dimensions constitute a resolution greater than 360 px/in, but less than 720 px/in, then we should uprez the image to pixel dimensions that, considering the size to which we wish to print the image, constitute a resolution of 720 px/in, and then print from our image editor with that image resident.

• But, actually, in many cases, we should uprez the image to pixel dimensions that, considering the size to which we wish to print the image, constitute a resolution of 1200 px/in or so, and then print from our image editor with that image resident.

Is that it?

Best regards,

Doug

Click to expand...

Exactly, Doug, my good friend!

The exception would be that some images do not have data that would show at 1200 dpi better than 600 dpi. Also the paper has to be at least semi glossy to benefit from the highest dpi settings.

Asher

Hi, Asher,

Asher Kelman said:

Exactly, Doug, my good friend!

The exception would be that some images do not have data that would show at 1200 dpi better than 600 dpi. Also the paper has to be at least semi glossy to benefit from the highest dpi settings.

Click to expand...

I'm having trouble understanding how an image that, at the intended print size, had pixel dimensions corresponding to, say 900 px/in would acquire "more detail" if uprezed to 1200 px/in.

Of course if we send the uprezed image to an Epson R1900 printer chain, the driver will first downrez it to 720 px/in.

I'm still missing some mystic ingredient of comprehension in all this. I'm sure if I had done a lot of work with printing, I would have a better idea of the realities.

Best regards,

Doug

I have absolutely no idea how the various algorithms available in Qimage and Qimage Ultimate for mediating between an image presented for printing and the chosen print chain work.

But I would certainly bet a few bucks that:

• if we start with an image with pixel dimensions of 4000 px × 5000 px

• have in mind printing this at a size of 8" × 10" (at which size the image has a resultion of 500 px/in)

• will print it on an Epson SP R1900, with native resolution 720 px/in

the algorithm does not, in effect, start by uprezing the image to any size other than 5760 px × 7200 px (which, at a print size of 8" × 10" would constitute a resolution of 720 px/in).

And I greatly suspect that:

• if we start with an image with pixel dimensions of 4000 px × 5000 px

• have in mind printing this at a size of 4" × 5" (at which size the image has a resultion of 1000 px/in)

• will print it on an Epson SP R1900, with native resolution 720 px/in

the algorithm does not, in effect, start by uprezing the image to 5760 px × 7200 px (which, at a print size of 4" × 5" would constitute a resolution of 1440 px/in).

Of course, have I lost bets before.

Best regards,

Doug

Hi, Asher,

Asher Kelman said:

The exception would be that some images do not have data that would show at 1200 dpi better than 600 dpi. Also the paper has to be at least semi glossy to benefit from the highest dpi settings.

Click to expand...

Let us consider an image that would benefit from uprezing to 1200 px/in before being printed.

Suppose that we plan to print at a size of 8" × 10".

What would be credible pixel dimensions for an image that might benefit from uprezing to 1200 px/in?

You say "Also the paper has to be at least semi glossy to benefit from the highest dpi settings."

What settings are you speaking of? Are you speaking of a px/in setting in the image editor when preparing to uprez the image? Or are you speaking of a px/in setting in the printer driver? (An Epson SP R1900 can be set to a resultion of 180 px/in or 720 px/in, so far as I know.)

Thanks.

Best regards,

Doug

My guess is that the rationale for perhaps uprezing the image to an integral multiple of the printer native resolution is this.

Suppose we have an image whose pixel dimensions, at the intended print size, correspond to a resolution of, say, 987 px/in.

We plan to print this on a printer chain with a native resulution of 720 px/in.

If we send the image as is (987 px/in), the printer driver will have to downrez it to a pixel count corresponding to a resolution of 720 px/in. It probably does this in a fairly clumsy way.

If we uprez the image to dimensions that amount to 1440 px/in, the printer driver will have to downrez it to 720 px/in, but it can do that in a trivial way and still do "as well as is possible") (I think).

Now would this be better than our first downrezing the image (in a sophisticated way) to pixel dimensions that amount to 720 px/in at the intended print size? Dunno. I can't at the moment see why.

(I'm waiting for somebody to say, "The reason you can't understand that maneuver in the meeting is that you didn't know that Smith is both the corporate treasurer and the vice president of IT.")

Best regards,

Doug

Of course what I tend to overlook with regard to different strategies of "uprezing" an image is that there are two different considerations (at least!):

• How much information is there in the modified image? It can of course never be more than there was in the original image (if we forget about pattern recognition and such).

• How "nice" does the modified image look?

And it is of course the latter on which we hope to do the best job.

So things that seem counter-intuitive from an "information content" basis may in fact make sense from the "how nice does it look" basis.

Best regards,

Doug

The ink jet head is made to pass over and over the paper repeatedly until the finest detail possible is achieved. With 300 dpi, it's less challenging. With 1200 dpi, the Canon print head, for example has to be controlled with the finest stepping motor instructions and then the jets have to be fired with extraordinary judiciousness! Specifying 1200 dpi means that the printing of that feature will take much longer, as the positional restraints are much more strict. However, this is the last stage in getting that fine detail. First one needs a software to calculate the right number and nature of pixels to allow a feature to be optimally drawn. That's where upressing with Lightroom or Photoshop or On One's version of Genuine Fractals can be expected to do better than some OS print driver, or have a program like Nik or QImage ultimate do it with knowledge of the intended paper type and size and desired resolution. Without a very accurate file representing the finest detail in the format expected by the printer, the printer cannot be expected to work at its optimum quality level. There's no point in accurately printing at the highest resolution a fuzzy image with no detail or with jagged diagonals. That's where the best programs separate themselves in actually transforming the details seen on a small image file to a larger version in the print, but without added artifact.

So, for me, my current effort now is to see what Qimage Ultimate can do for me and compare it with my usual workflow in Photoshop CC.

Asher

Hi, Asher,

Asher Kelman said:

The ink jet head is made to pass over and over the paper repeatedly until the finest detail possible is achieved. With 300 dpi, it's less challenging. With 1200 dpi, the Canon print head, for example has to be controlled with the finest stepping motor instructions and then the jets have to be fired with extraordinary judiciousness!

My problem is that I don't know how we would tell an Epson SP R1900 print chain to "operate at 1200 px/in."

If:

• I have an image that I have uprezed to have dimensions of 4800 px × 6000 px, which would be a resolution of 1200 px/in if printed to a size of 4" × 5", and

• I tell my image editor to print it at a size of 4" × 5" on my Epson SP R1900

does the program expect that the Epson printer driver, seeing the combination of:

• Pixel dimensions of 4800 px × 6000 px, and

• Print size of 4" × 5"

deduce that I have in mind a rendering at 1200 px/in and make arrangements for the printer machine proper to operate on that basis?

My understanding is that in such a case, the printer driver will downsize the image to 2880 px × 3600 px (which would be a resultion of 720 px/in at the commanded print size) and send it to the printer on that basis.

Now, regarding the nozzle pitch in the two directions, now we are speaking of dots (which is what the nozzles make, not pixels).

The dot pitch will of course be much finer than 720 dots/inch; many dots are used to make each pixel. How much finer the dot grid is than the pixel grid will depend on the quality mode chosen for the job.

But of course my understanding of this may be wholly off base.

I will ponder this further.

Best regards,

Doug

Click to expand...

By the way, do not think that the value of the resolution indicator(s) in the file (or saved with the file with the image in memory in the image editor) tells the printing system anything, or tells us anything about the image itself.

I can have in memory two copies of the same image, both with pixel dimensions of 2880 px × 3600 px (which would be 720 px/in if printed at a size of 4" × 5"), one of which showed a resolution of 50 px/in and one a resultion of 2000 px/in. Both actual images are identical, pixel-for-pixel.

I'm fairly sure that if I told PS to print them both to my Epson SP R 1900 at a size of 4" × 5" (lets say in each case with me manually setting the printer to its "highest quality" mode, the results would be indistinguishable.

So how would I make the printer "operate at 1200 px/in"?

But of course I could be wrong.

Best regards,

Doug

I just conducted a primitive test of some of the matters I have been discussing.

I started with an image with pixel dimensions of 2880 px × 3600 px. This would imply a resolution of 720 px/in if printed at a size of 4" × 5".

I made three versions of the image file:

• Original pixel dimensions (2880 px × 3600 ps), resolution indicator(s) set to 50 px/in.

• Original pixel dimensions (2880 px × 3600 ps), resolution indicator(s) set to 2000 px/in.


• Uprezed (in Picture Publisher 10) to pixel dimensions 4800 px × 6000 px (which would imply a resolution of 1200 px/in if printed to a size of 4" × 5"), resolution indicator(s) set to 1200 px/in.

I then loaded all three into PS CS5 and asked it to print all three to my Epson SP R1900 at a size of 4" × 5". In each case, I set the printer chain to its highest quality option.

The time required to print all three was essentially identical.

A quick examination shows no difference between the three prints.

A quick examination of a detailed area (the model's hair) with a loupe shows no perceptible difference between the three prints. Under a comparison microscope (I don't have one) there might be perceptible differences.

I do not attempt to draw any conclusions from this admittedly-primitive test program.

Best regards,

Doug

Hi, Asher,

So let me get this clear. Suppose we had an image with existing dimensions 2880 px × 3600 px. We had in mind printing it out of Photoshop on an Epson SP R1900 at a size of 4" × 5" (at which size its current pixel dimensions would correspond to a resolution of 720 px/in).

We believed that it was the kind of image that would benefit from "printing it at 1200 px/in".

How would we do that. Would we:

• uprez it to dimensions of 4800 px × 6000 px (which would, at the planned print size, correspond to a resolution of 1200 px/in (using whatever sophisticated uprezing tool we think to be the best for that).

• load it into Photoshop and tell Photoshop to print it to the R1900 at a size of 4" × 5". (We would presumably, on the printer driver control panel, set the printer system to the best available "quality".)

Or do we have to do something beyond that to instruct the printer system what we have in mind?

Best regards,

Doug

Hi, Asher,

Where can I find the writing by Jeff Schewe (or perhaps it was a writing based on his) from which the recommendations you mention in the original message were drawn? I would like to read it in context.

Thanks.

Best regards,

Doug

Doug Kerr said:

Hi, Asher,

Where can I find the writing by Jeff Schewe (or perhaps it was a writing based on his) from which the recommendations you mention in the original message were drawn? I would like to read it in context.

Click to expand...

Doug,

Here's a start!

http://www.digitalphotopro.com/technique/workflow/the-right-resolution.html?start=1#.VNHDVfTF8nV

Hi, Asher,

Asher Kelman said:

Here's a start!

http://www.digitalphotopro.com/technique/workflow/the-right-resolution.html?start=1#.VNHDVfTF8nV

Click to expand...

Thanks so much.

Best regards,

Doug

Hi, Asher,

Asher Kelman said:

Here's a start!

http://www.digitalphotopro.com/technique/workflow/the-right-resolution.html?start=1#.VNHDVfTF8nV

Click to expand...

Interesting article. Thanks for the cite.

Here is a summary of my interpretation of the recommendations in the article. In certain specific areas, I will use the discussion that pertains to typical Epson printers. It should be obvious how to extend the principles to other brands of printer.

1. If the pixel dimensions of the image, considered at the size of the intended print, constitute a resolution less than the "native" resolution of the printer, upsize the image to pixel dimensions that, considered at the size of the intended print, constitute a resolution equal to the "native" resolution of the printer.

2. In this regard, if the printer is a contemporary Epson "Professional" printer whose "basic" native resultion is 360 px/in, and the original pixel dimensions of the image, considered at the size of the intended print, constitute a resolution less than 360 px/in, then set the printer for the basic mode and upsize the image to dimensions that, considered at the size of the intended print, constitute a resolution of 360 px/in.

3. Again considering an Epson "professional" printer which offers a "fine detail" mode with a native resolution of 720 px/inch, if the original resolution of the image, considered at the size of the intended print, is greater than 360 px/in but less than 720 px/in, then set the printer for the fine detail mode and upsize the image to dimensions that, considered at the size of the intended print, constitute a resolution of 720 px/in.

4. If the original resolution of the image, considered at the size of the intended print, constitute a resolution greater than the (highest available) native resolution of the printer, do not downsize it to match that native resolution.* You may wish to upsize it to the next integral multiple of that native resultion. **



Best regards,

Doug

Thanks, Doug,

There's another point to consider. In looking at a postacard sized version of one's inage, the dark and white area are closer together and the image might seem wonderfully contrasty as a whole. The same file, simple accurately upressed to say poster size, might seem much less bold as it has a greater area on which gradients can be spread. So it could be that, "to get the same impression or experience", the print needs to have adjustments to account for the increased area. This is independent of viewing angle is we assume that all "Art" in a gallery are stared at and examined at about 6-18 prior to actual purchase.

Mike doesn't explain his algorithms, but Qimage. Lightroom, (I believe) and Nik do sharpening based on the dpi and size of the print in some cases, (Nik for one), as well as viewing distance. Choice of paper comes into this as a rough art paper, for example, cannot show the detail of a high gloss Hahnemühle 100% cotton Baryta gloss paper.

I am still learning what is possible. So no one should be shy in adding their own experience and point of view.

Asher

Hi, Asher,

Asher Kelman said:

Thanks, Doug,

There's another point to consider. In looking at a postacard sized version of one's inage, the dark and white area are closer together and the image might seem wonderfully contrasty as a whole. The same file, simple accurately upressed to say poster size, might seem much less bold as it has a greater area on which gradients can be spread. So it could be that, "to get the same impression or experience", the print needs to have adjustments to account for the increased area. This is independent of viewing angle is we assume that all "Art" in a gallery are stared at and examined at about 6-18 prior to actual purchase.

Click to expand...

Yes, these are all subtle-sounding matters than can be of great significance. It is a very complex matter, overall.

Mike doesn't explain his algorithms, but Qimage. Lightroom, (I believe) and Nik do sharpening based on the dpi and size of the print in some cases, (Nik for one), as well as viewing distance. Choice of paper comes into this as a rough art paper, for example, cannot show the detail of a high gloss Hahnemühle 100% cotton Baryta gloss paper.

I am still learning what is possible. So no one should be shy in adding their own experience and point of view.

Click to expand...

Indeed.

Thanks.

Best regards,

Doug