PPI on-screen and its affect on DPI on print

[ncJohn]

25 minutes ago, Medical Officer Bones said:

Well, you did mention that there's a whole lot of noise relating to ppi and dpi on the web. I just wanted to clarify things a bit more.

By the way, you wrote that you have a Canon printer? If it is a newer model (last few years) it may very well support 600ppi prints as well at high quality settings on quality glossy paper. In that case you might want to feed it 600ppi images instead of 300ppi ones. That is another point I made: know (the hardware limits of) you output device.

Apparently it's hard to know whether the Pro-100 supports 300 or 600. The actual driver only specifies fast, standard, or high quality (or, if you go into the custom dialog, it's standard, high, or fine). And a search online turns up a lot of people who say it's 300 and a lot of people who says it's 600. But I didn't see any who say how they know that. But thanks for the idea; maybe somebody who knows will speak up.

[Alfred]

4 hours ago, toltec said:

So, based on a 300 pixels per inch image, an inch could be anything from just over one inch, to just under seven inches. 

Which is why pixel density on a computer screen is sometimes referred to as pixels (or, more commonly, dots!) per ‘logical’ inch, and why there is a distinction in web design between CSS pixels and device pixels. 

[ncJohn]

14 hours ago, αℓƒяє∂ said:

This pretty much sums it all up for me.

[R C-R]

14 hours ago, αℓƒяє∂ said:

Which is why pixel density on a computer screen is sometimes referred to as pixels (or, more commonly, dots!) per ‘logical’ inch...

Apparently this "logical" inch comes from how Microsoft decided fonts should be measured:

Quote

This scaling factor is described as 96 dots per inch (DPI). The term dots derives from printing, where physical dots of ink are put onto paper. For computer displays, it would be more accurate to say 96 pixels per logical inch, but the term DPI has stuck. 

Maybe Apple did something similar but a (very) quick search on the web turned up this, so that is as far as I took it.

[lepr]

.

[Wosven]

On 7/24/2018 at 6:28 PM, toltec said:

I agree totally, the problem is that new users (new to photo imaging or printing) have so many difficulties because of the DPI and PPI settings. I think using inches (especially in Europe where we discarded them decades ago) is confusing. Especially when an inch (on a monitor/phone is never an inch. Surely there’s a better way to explain this for new users?

You need too think of Pixels as a screen propertie, like tiles of different sizes. My 24 inches screen 's got bigger tiles (pixels) than my phone's screen. Imagine if I did the same with real floor tiles… With 10×10 cm tiles, or larger ones of 30×30 cm…
When talking about print dots, it's not square tiles but rounds, like with some old printers that spit ink dot on a paper.

PPI => for screen (virtual) mesurements
DPI => for print ((paper) physical) mesurements

When printing, the logical assumption is 1 pixel = 1 dot, and the optimal density is 300 DPI. So we try to get/create/resize images to 300 ppi => 300 dpi.

 

If a client gives you a logo in low resolution, you'll have to import the image in AD and redraw it, and when finished, you'll modify the "DPI" (but on screen it's PPI) to be able to export your logo in hight definition. This way, you don't need to think about its resolution later, you can import and use the exported files in your document for printing and get a good result.

When working with vectors, you don't mind since you can modify DPI value and it will translate in pixels depending of your exports size as images (JPG, PNG, TIF…), but if you work with rasterised/pixels images, you'll need to think at the beginning which density of pixel you need for your final result.
That's when you need to think: "how many pixels by inch I need at least for my final result?"
And create a file with enough or more pixels that needed… Since more is not a problem.

[toltec]

1 hour ago, Wosven said:

 

PPI => for screen (virtual) mesurements
DPI => for print ((paper) physical) mesurements

When printing, the logical assumption is 1 pixel = 1 dot, and the optimal density is 300 DPI. So we try to get/create/resize images to 300 ppi => 300 dpi.

So

PPI is not actually a precise or fixed measurement because the inches are only “virtual” and are are in fact of a variable length between about 1/2 of an inch and 10 inches. So the I of PPI means nothing.

DPI is incorrect because actually, we don’t mean dots, in litho we actually mean lines per inch. i.e. 150 lines per inch. So the D of DPI means nothing.

Then, in LPI, the L of lines is not correct because we simulate an old fashioned half tone screen with dots. i.e. 2540 imagesetter dots to create a 150 LPI halftone screen “dot”. This  halftone “dot” varies in size because it is simulating shades of grey. So the L of LPI is no better than the “P” or the “D”. At least the inch is an inch (geographical variations excepted). However, because the imagesetter dots (physical dots) can’t vary in size, we use varying quantities of physical dot to simulate a varying size halftone dot which is simulating shades of grey.

Even “halftone” is a bit daft because there is actually only one tone (black) or two if you count “no-tone”as white, with these two “tones”  simulating an infinitive numbers of greys. 

Who came up with all this?

 

Normally, I would blame Microsoft but I don’t think they were around in the 1800s when all this halftoning started.

[R C-R]

52 minutes ago, toltec said:

PPI is not actually a precise or fixed measurement because the inches are only “virtual” and are are in fact of a variable length between about 1/2 of an inch and 10 inches. So the I of PPI means nothing.

Again, the "i" of PPI means just one thing when you are talking about the pixels of the screen itself, a physical object. It is not the same thing as the "i" in the context of the virtual pixels of whatever that screen is displaying, or of the pixels encoded into a digital graphics file, or of the dots or the lines of a printout of that file. 

1 hour ago, toltec said:

Who came up with all this?

It all came about because over time a lot of different people adapted existing terminology they were familiar with to new things that did not exist when that terminology was invented. There is nothing unusual about that. There is no lead in "leading," a term we use for the space between lines of text because lead was used for that. We don't keep uppercase & lowercase letters in different cases but we still use those terms derived from that use. We talk about "raster images" even though that term came from the raster lines of black & white CRT TV displays & doesn't really apply very well, even to the addressable pixels of modern flat panel color displays. "Computer" as a device was adapted from its use to mean people who computed the results of calculations by hand or with the help of a mechanical machine.  

It is just the way things are. Collectively, we adapt existing or coin new terminology for convenience, or for brevity, or sometimes even just for amusement. There is nothing that can be done about that, other than to be aware that the same term often will mean different things in different contexts.

[Alfred]

6 minutes ago, R C-R said:

or for brevity, or sometimes even just for amusement

I’m a little surprised that you didn’t take the opportunity to say “or for brevity, or sometimes even just for levity”.

 

[Alfred]

3 hours ago, Wosven said:

PPI => for screen (virtual) mesurements
DPI => for print ((paper) physical) mesurements

When printing, the logical assumption is 1 pixel = 1 dot, and the optimal density is 300 DPI. So we try to get/create/resize images to 300 ppi => 300 dpi.

I like your floor tiles analogy, but I think the above PPI/DPI description is the wrong way around. On a computer screen, one dot (or one triad of dots, if it’s a colour screen) corresponds to one pixel in the image, but for printing on paper there are often multiple dots (sometimes of different sizes to achieve tonal variations) corresponding to each pixel in the image. So it follows that it doesn’t matter whether we say DPI or PPI for images on a computer display, but it matters a lot when outputting to other devices.

[toltec]

Personally. I think we should use PPI for printing, which is technically correct. Using 2 pixels as the ideal quantity to produce a "halftone dot", as in LPI.

For screens, maybe, PQ for Pixel Quantity. 

But nobody ever listens, sigh! 

 

 

[R C-R]

1 hour ago, toltec said:

Personally. I think we should use PPI for printing, which is technically correct.

How is it any more technically correct than DPI? Neither pixel nor dot has a simple, unambiguous meaning that applies to all printing methods.

Maybe you know of others but the only one I can think of where dot would have an unambiguous meaning is for old school dot matrix impact printers, the ones that drove a column of pins against a typewriter ribbon to print ribbon-colored dots on paper, & made an annoyingly loud screeching noise as they printed. For them, dot & pixel were effectively the same thing, unless maybe you consider the 'high resolution' modes some of them supported that made two passes with the paper advanced by ½ of the pin pitch to crudely print at a higher pixel than dot resolution.

In ancient times I had one of those once "standard" printers & a G-Wiz adaptor hooked up to my Commodore C=64 & later C=128 computers. It was not much more than a glorified typewriter.

[Wosven]

Really interesting, @toltec, Thoses notions depend of our type of work, and we all need tricks to use today's work with pixels and vectors with old notions… I need someday to update my knowledge about lithography — mine are outdated, a mix of when I was studying and turn around old  posters from Toulouse-Lautrec's period — and few pics I saw here and there when you talk about it.    

@Alfred :  

Quote

So it follows that it doesn’t matter whether we say DPI or PPI for images on a computer display, but it matters a lot when outputting to other devices. 

Perhaps I didn't stydied enough about pixels, but for me they're squares on my screen, and round on e-ink display. And round when printed

And those yellows icons on my desktop are folders and not suitcases! (I'm not sure if it comes from using Mac… but people tend to use the word "suitcase", and it confuses me when they put files in their suitcase on their desktop!)

 @toltec please, no PQ… I won't be able to stay serious if you use our crude abreviation for "papier-toilette"… if I dare spell it here: "papier-cul" (last work pronouced as the letter "Q" in French).

[Alfred]

6 minutes ago, Wosven said:

Perhaps I didn't stydied enough about pixels, but for me they're squares on my screen, and round on e-ink display. And round when printed

I can't begin to tell what shape they are on my screen. Your eyesight must be much better than mine!

Quote

And those yellows icons on my desktop are folders and not suitcases! (I'm not sure if it comes from using Mac… but people tend to use the word "suitcase", and it confuses me when they put files in their suitcase on their desktop!)

Hmm ... that definitely sounds like a (very strange) Mac thing.

Quote

 @toltec please, no PQ… I won't be able to stay serious if you use our crude abreviation for "papier-toilette"… if I dare spell it here: "papier-cul" (last work pronouced as the letter "Q" in French).

Why am I reminded of Marcel Duchamp's « L.H.O.O.Q. » ?

 

[Wosven]

1 minute ago, αℓƒяє∂ said:

I can't begin to tell what shape they are on my screen. Your eyesight must be much better than mine!

I'm not sure too looking at this modern screen, but in the old times it seemed those 15 inches monitors had square pixels we could count on the screen

 

Quote

Hmm ... that definitely sounds like a (very strange) Mac thing.

I suppose it's from the compressed fonts folders we called "valises de polices" (suitcase of fonts) that contained differents size of bitmap fonts.

 

And for Marchel Duchamp, yes, he added some humour in Art and it's a nice heritage

[R C-R]

32 minutes ago, Wosven said:

I'm not sure too looking at this modern screen, but in the old times it seemed those 15 inches monitors had square pixels we could count on the screen

Those pixels probably were not square. Check out the Wikipedia Pixel article for some examples of what sub-pixel screen arrays really look like.

[Wosven]

@R C-R  "Les pixels sont approximativement rectangulaires, parfois carrés. Leur dimension peut être changée en réglant l'écran ou la carte graphique. " ( The pixels are approximately rectangular, sometimes square. Their size can be changed by adjusting the screen or graphics card. — https://fr.wikipedia.org/wiki/Pixel)
"Pixels on computer monitors are normally "square" (that is, have equal horizontal and vertical sampling pitch); pixels in other systems are often "rectangular"" (https://en.wikipedia.org/wiki/Pixel)

Thanks for the info, but that's not the point, more about how to understand and use DPI/PPI while creating images that we usually later need to print, avoiding usual mistake like creating a low resolution image when we need it a lot more bigger.

How to get a basic understanding that can help in a fast way to find the canvas size needed.

 

But the link is interesting in a technical way, the same as the way e-ink is displayed. I used to spend a lot of time on another forum for mobile readers, and the various discussions about e-ink diplays and new technologies created and tested for a e-ink color display — seems none were good enough for real production — were interesting readings  too. The next reading is how eyes distinguish colors and how some people can perceive more colors than other ones.

Next step we will all use a microscope to look at our screens