Apaflo Loc: Anchorage, Alaska

Well I'm glad to see you finally realize that asking questions about things you don't understand is a much better way to discuss this topic. I would imagine, or at least hope, that you realized it is a bit embarrassing to contradict Mad Man Chan, the fellow posting from MIT! Even though I've challenged you multiple times with your crass contradiction of what he says, you have not even hinted once that you read the cited URL, that you understood what it says, or know who is saying it.

We might also note that Jeff Schewe, who once upon a time basically agreed with you, has long since began to understand. "That was then and this is now." is what he has said when asked.

You are also disagreeing with Mike Chaney of DDIsoftware, who says he gets his information from the "Epson Developer's Manual"... but I've never been able to find anything with that title. I have Epson service manuals for a variety of printers, and an ESC/P manual, but none of them specify it in simple enough terms to quote a one liner here. They do say it, but with pages of specifications for nozzle spacing, stepper motor drive frequencies and such!

Anyway, as to your question, the print driver has to be optioned with a parameter for the size of the print. Hence if it is told to make a print that is 14.34 inches wide it does. The first thing it does is multiply that size times the native pixel rate, 14.34 * 360 (for an Epson printer), to get the pixel dimension needed, 5162 (rounded down to an integer size). Whatever the input file dimensions are, the image data is resampled to 5162 pixels per horizontal line. If the original image is from a Nikon D4 that is 4928x3280, it is resampled to 5161x3441. If the original is from a D810 image that was 7360x4912 but has been cropped to a 5:4 aspect ratio 7360x5888, it will be resampled by the print driver to a pixel dimension of 5161x4129 and print at 14.34"x11.47".

It makes exactly zero difference what the Exif tag for resolution is set to on any of the above images! Zero.

However, I would also caution that the print driver might well use either a nearest neighbor algorithm, a bilinear, or a bicubic. For photographs all of those should be avoided. Even worse, after that much of a size adjustment each of those images should be resharpened, which cannot be done if the print driver does the interpolation. Hence the best way is to use an editor to resample the images, to the sizes calculated above, and then resharpen before sending to the print driver. That is also a process that absolutely ignores the Exif data's resolution tag in the original image, but one thing it might do is set that field in the new image created, using the PPI value you use to have it calculate the sizes. If you don't use the editor to calculate the sizes, it might just do that on its own and set the Exif field accordingly. But it is still a case of what the Exif data says is the result of what you've done, not the cause of anything the printer does.

TheDman Loc: USA

That's a whole lot of typing to avoid answering the question. I'll ask it again: how does the print driver know how large or small of an image to make? Should be a very simple, quick answer. The more dodging you do, the weaker you look.

Apaflo Loc: Anchorage, Alaska

"... the print driver has to be optioned with a parameter for the size of the print."

Wasn't 16 words simple enough? Doesn't appear that anything was dodged, until you replied. I think if you would read the previously cited URL from the MIT site it would be clear enough. Here it is again:

http://people.csail.mit.edu/ericchan/dp/Epson3800/faq.html#native_res

Mad Man Chan is pretty definite about it... Probably stronger than me by orders of magnitude, eh? Do you know something he doesn't?

So just what is your excuse now?

TheDman Loc: USA

What is the 'parameter'? You're ducking. Big time.

Let me ask in a way that will force your hand. True or false: If my image were 800 pixels tall by 600 pixels wide and I wanted to print it with a width of one inch, I could merely change the PPI value in the Image Size properties to 600 and click print.

There you go. True or false? I don't want ten paragraphs, I don't want links to crazy Chan's, just a single word. Do you have the guts?

Apaflo Loc: Anchorage, Alaska

The parameter is the physical size you want the image printed at. Just exactly how that is labeled depends on which print driver you are using, but in all the example screen shots you have provided, using Adobe software, it has been labeled as "Width:" and "Height:". It is an editor parameter, not a print driver parameter, and is also not the Exif data tag value from the original image file.

That is an editor parameter. Get that straight. Just because it uses the same words doesn't mean it is the same thing! (In the case of the Width and Height it is also exactly what is passed to the print driver, but the PPI value is not.) If you set the PPI value and tell it to resample, the editor will resample the image data and send a different data set to the printer. If you don't tell it to resample, the original image date is sent to the printer along with the width and height parameters and the print driver will resample the data before it is rasterized and dithered.

Note that not all print drivers are like your print driver. The one that I use looks like this:
Note the first option is "Media Size", and that is uniquely the parameter that sets the size of the image that will be printed. Then note also there is a "Resolution" option, which in this case is set for 1440x1440 DPI. That has absolutely no effect on the size of the print.

Your question of a non-sequitor, and as stated it is utter nonsense. It doesn't mean what you mistakenly think it does!

The answer obviously is that if that is all you do, no it will not, but that is almost certainly not all you are thinking it does. Regardless of that the question here is what does the resolution tag in the image file do, but your question has nothing at all to do with the Exif tag value from the file! You are giving the editor a parameter, which can be used instead of a physical dimension, and will cause the editor software to compute the physical size to supply to the print driver.

Again, that question is immaterial this time, as it has every time you asked it, because you simply do not understand either the question or the software driving the printer.

But since you aren't interested in what Eric Chan says about the software from Adobe that you are using, I'll give you a clue: Eric Chan was the Principal Scientist whose group at Adobe designed the interpolation modules in LR and ACR. You really do want to listen to what he says it does and does not do.

TheDman Loc: USA

A hundred more paragraphs. You know you've lost the debate when you're too terrified to simply answer a question with a true or false. Took several paragraphs to finally spit out an answer:



You're wrong. Again. And Chan would agree with me.

Apaflo Loc: Anchorage, Alaska

Yeah? Quote where he ever has agreed with anything you have said on this topic!

The fact is that you have lost the "debate" when you pose nonsense questions that are non-sequiturs from the start.

Lets not forget that you said you can print 1 pixel 1" wide without any resampling. You claimed a pixel is a "piece of digital data". You claimed that if the image data file you print doesn't change it means the print driver didn't resample the data. You said "No resampling needed. Just change the PPI to 1."

I could look farther back and make that list three or four times as long. You are making assumptions that have nothing to do with fact. You questions are based on these false assumptions, not on real facts. And you learn nothing from one exchange to the next and repeatedly come right back to what has been prove incorrect the first time

You can't find any authoritative credible sources that agree with you. And you totally ignore those that I reference for you, that you apparently don't even both reading. It appears you do this to be disruptive to the threads you post in, to antagonize other posters and confuse those who try to help with valid data.

Explain why I should continue this conversation if you can't discuss anything in some semblance of a rational intelligent way?

tdekany Loc: Oregon

You shouldn't really, what I'd do in your shoes is study his work of art and ask him to help you become a photographer - are you not tired of taking snap shots?

TheDman Loc: USA

I've never spoken with him about this, but I assume he's not impossibly stupid and therefore would answer "True" to my previous question, unlike yourself.




Oh really? Let's see.

"For example, a bitmap image may measure 1,000 × 1,000 pixels, a resolution of 1 megapixel. If it is labeled as 250 PPI, that is an instruction to the printer to print it at a size of 4 × 4 inches. Changing the PPI to 100 in an image editing program would tell the printer to print it at a size of 10×10 inches."
https://en.wikipedia.org/wiki/Dots_per_inch

"That's because image resolution affects only one thing - the size of the image when it's printed. By setting the resolution in Photoshop, we tell the printer, not the screen, how many of the pixels in the image to squeeze into an inch of paper."
http://www.photoshopessentials.com/essentials/the-72-ppi-web-resolution-myth/

"Pixels per inch (ppi) - A measurement of image resolution that defines the size an image will print."
http://graphicssoft.about.com/cs/digitalimaging/f/pixelsprint.htm

"for example, if you want to print a 4-inch-by-6-inch image, and your pixel count is 800 x 1200, you need to divide 800 (the pixel height) by 4 (the photo height) and then divide 1200 (the pixel width) by 6 (the photo width). The PPI will be 200"
http://electronics.howstuffworks.com/cameras-photography/digital/how-to-calculate-image-size-and-ppi-when-printing.htm

"[PPI] is a single number stored as part of the JPG file, and is used my most programs in determining the scale at which to print the image. Here are two pictures. Both are 350x262 pixels, but the ppi settings are different. They probably appear the same size in your browsers. However, printers may render the first more than four times larger than the first. Try downloading each image into your graphics program and printing."
http://users.wfu.edu/matthews/misc/ppi/ppi.html

Yeah boy, can't find Anybody!! LOL! The last one even implores you to download the images and try to print them yourself, something I've asked you to do previously too, but you keep refusing. Wonder why?

Apaflo Loc: Anchorage, Alaska

There were three criteria, and not one of these match all three!

Authoritative means they have to be recognizable experts in this field. Eric Chan and Mike Chaney are examples, which is why I cited them.

Credible means we should believe what they say. If they are not authoritative sources they are not credible.

Agree with you is when what they say is the same thing. That is not just using the same terms to say something entirely different (which is the case with each of your cites).

Your cites fail. Wikipedia is an obvious example, where anyone including you can write what they want there.

But the significant point is that you have claimed all along that changing the Exif resolution tag value is what changes the size, without any resampling of the image data. But as I have said, and I cited the most significant and authoritative source there is (who says exactly the same thing), that when you click on the "Resample" button in Photoshop it does not mean the data is not resampled before printing, it just changes where the data is resampled at. If resampling is disabled in the editor, then the print driver will resample it.

Once again, here is what the Principal Scientist who was responsible for designing the software you are using had to say about this,
And this is what Mike Chaney, owner of DDISoftware and author of QImage said responding to someone who said exactly what you have been saying:

No need to wonder why. I've stated several times that you present a non sequitur. Whether it does this or does that has nothing to do with the statements you've made. Print the two images... but neither of them will be print at the stated PPI. Both will be resampled to the native PPI of the printer you are using. If it is an Epson for example, printing at 360 PPI, the 350x262 image tagged for 72 PPI will be resampled to print at 4.9"x3.6", which will require the printer be fed a 1680x1258 pixel image. The file tagged at 300 PPI will be resampled to print at 1.7"x0.9" and will actually print data that is 420x314 pixels.

TheDman Loc: USA

LOL!! Here we go again. The whole internet is wrong and Apfalo is right. Why does it always come down to this?

Wikipedia and Photoshop Essentials are way credible and authoritative, and each agree with me. Now you're sounding like a conspiracy theorist. They all think Wikipedia is fine until it disagrees with them, then it's "anybody can write whatever they want". Get the tinfoil hats out!

The ultimate point you're trying to make is like saying you cannot code websites in PHP because PHP is written with words, and computers only understand binary code. No, you can totally code websites in PHP. Who cares what happens when it is compiled. Talk about your all-time non-sequitor; that's the very definition of the term!

And regardless of whether or not you think a question is irrelevant, why does that cause you to answer it incorrectly? I asked you to answer true or false: If my image were 800 pixels tall by 600 pixels wide and I wanted to print it with a width of one inch, I could merely change the PPI value in the Image Size properties to 600 and click print. You answered false. You are wrong, in every possible way, and your quote from the Principal Scientist backs me up completely. End of story.

Apaflo Loc: Anchorage, Alaska

1. They are not really credible.
2. They are not close to authoritative.
3. They do not agree with you on the contested issues of substance in this thread.

Your responses are getting less logical and more whimsical. Why should I continue a discussion if you refuse to read references (even those you cite) and do not learn?

TheDman Loc: USA

1. They're way more credible than Apfalo.
1. They're infinitely more authoritative than Apfalo.
3. They completely agree with me on the contested issues of substance in this thread.
4. They completely refute your answer to my true or false question.
5. You're way butthurt about it, and are trying to check out of this thread while saving some sort of face.

See ya.