Skip navigation
Tsnmrz
Currently Being Moderated

how to change "real pixels" to "relative resolution" in control pannel

Mar 30, 2012 4:10 AM

Hello,

 

Is there any way to change the "real pixels" (PPP 589,45x799...) info in the control pannel to "relative resolution" (ppp 300). I would like to obtain the resolution an image has at its size in Illustrator.

 

Excuse my English, Many thanks.

 
Replies
  • Currently Being Moderated
    Mar 30, 2012 5:21 AM   in reply to Tsnmrz

    Tsnmrz,

     

    This depends on your version. In the latest one(s) the Document Info palette/panel should show it when you have the flyout options Objects and Selection only ticked, the latter so you get the information about the selected image only.

     
    |
    Mark as:
  • Currently Being Moderated
    Mar 30, 2012 8:16 AM   in reply to Tsnmrz

    Tsnmrz,

     

    Resolución: 325,667 por 259,716 pixeles por pulgada means:

     

    Resolution: 325,667 PPI and 259,716 PPI.

     

    So the image is 325.667 PPI in the X/W direction and 257.716 PPI in the Y/H direction.

     

    In other words, the proportions have been changed from something where the resolution has been the same in both directions.

     
    |
    Mark as:
  • Currently Being Moderated
    Mar 30, 2012 2:38 PM   in reply to Tsnmrz

    Tsnmrz,

     

    The control panel information seems rather strange ideed.

     

    I wonder whether something is faulty, possibly the preferences.

     

    You may try to  Move the folder

     
    |
    Mark as:
  • Currently Being Moderated
    Mar 30, 2012 9:57 PM   in reply to Tsnmrz

    Tsnmrz wrote:

    ...

    The 1st one is 1in x 1in, and the second one 1in x 2in.

    For the 1st one i get in the control pannel: PPP: 5, which is obious because you can count by yourself.

    For the second one i get in the control pannel: PPP: 2,5 x 10, and I´m affraid to don´t understand the relationship or the formula i have to apply to this one to get the 5 PPP resolution.

     

    Thanks in advance.

    ppi.jpg

    I hope this image helps.

    I made it by creating a 1 inch square with a gradient fill. Then I chose Object > Rasterize and entered 5 ppi for Other. Then duplicated the object and stretched to 2 inches wide.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 2, 2012 5:40 AM   in reply to Tsnmrz

    Rotate the 5 x 10 pixel image -90° (as its thumbnail is displayed in the Links palette) and the display will change to "PPI: 5" in the Control Panel, and "Resolution: 5 by 5 pixels per inch" in the Document Info palette.

     

    Don't ask me to explain it. I think you've found a bug.

     

    JET

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 2, 2012 5:52 AM   in reply to Tsnmrz

    You are welcome, Tsnmrz.

     

    I am afraid I am unable to see the issue because I am still with 10.

     

    It seems that you may have to rely on the Document Info which seems to show the right figures:

     

    Pixel size: 977 x 872

    Point size: 216 x 241.741

     

    PPI/PPP: 325,667 x 259,716

     

    You get this by dividing the pixel size(s) by the point size(s) and multiply by 72 (72 points to an inch).

     

    Edit: Hi James.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 2, 2012 9:56 PM   in reply to JETalmage

    JETalmage wrote:

     

    Rotate the 5 x 10 pixel image -90° (as its thumbnail is displayed in the Links palette) and the display will change to "PPI: 5" in the Control Panel, and "Resolution: 5 by 5 pixels per inch" in the Document Info palette.

     

    Don't ask me to explain it. I think you've found a bug.

     

    JET

    I believe it is not a bug but behavior. The resolution of raster images rotated or transformed differently than their original placement is reported by calculating the number of pixels per sides of the rectangular bounding box encompassing the rotated/transformed image as ppi of the corresponding original size of the sides. For example, create a rectangle 1 inch wide by 2 inches high,  rasterize the object at 5 ppi, rotate 90 degrees and you will get the same result as Tsnmrz. What is happening is the original height of 2 inches occupied with 10 pixels, after the rotation is is now occupied with 5 pixels which results in reported resolution of 2.5 ppi.So, the reported resolution of transformed images is about a relative change of  the original and not the ppi of the actual changed size. The actual ppi will be defined when the image is rasterized and until then the ppi are nothing but a relation to the original image.

     

    To fix the reported resolution of a transformed image, rasterize it. But this is not a good solution if the image's sides are not straight - for example skewed or with tilted rotation.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 3:19 AM   in reply to Tsnmrz

    Tsnmrz,

     

    I wonder what it says when you rotate it by a non right angle, such as 30 or 45 degrees, or whichever angle may apply, 13.571 or whatever.

     

    I am afraid it will be discomforting.

     

    In any case, unless the purpose is to confuse the user, one constant PPI should apply to the image at whichever angle so it is possible to know the actual size and to decide whether the resolution is adequate for the print in question.

     

    A bug is a bug is a bug. And this seems to be one.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 6:56 AM   in reply to Jacob Bugge

    Jacob Bugge wrote:

     

    ...

    In any case, unless the purpose is to confuse the user, one constant PPI should apply to the image at whichever angle so it is possible to know the actual size and to decide whether the resolution is adequate for the print in question.

     

    .....

    The problem with achieving this is the fact that raster images can be manipulated to certain extent as vector graphics and this makes it impossible to define one constant ppi of the current size. For example take the skew tool and make an extreme skew of the raster image. Its new effective resolution should be the ppi per the sides of the bounding box you get when you rasterize the image. But at what ppi this bounding box should be rasterized? Because this is not defined yet by the user, Illustrator maps the number of the original pixels per side to the size of the new dimensions.

    And to complicate the matter more, try making an extreme deformation of the image with tools like envelope distort so it has a very different size and shape than the original. Even after expanding the image, Illustrator does not report any ppi change until you use transform tools that change shape and orientation.

    So, regarding the reported ppi, I think  the programmers tried to put some sort of logic apparently based on the underlining math which may not be the best logic on the user interface, but it is behaving consistently with certain logic which apparently is not a bug.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 7:46 AM   in reply to Tsnmrz

    You are welcome, Tsnmrz.

     

    I suggested the Document Info because it seems to show consistent numbers.

     

    You may try to see whether it works for arbitrary rotations.

     

    If it does, or if it does for your purposes, you may (have to) rely on that and forget about the Control Panel.

     

     

    emil,

     

    Because this is not defined yet by the user, Illustrator maps the number of the original pixels per side to the size of the new dimensions.

     

    But in the case shown in this thread it seems to swap sides instead of keeping the original sides.

     

    To bug (us) or not to bug (us).

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 9:12 AM   in reply to Jacob Bugge

    Jacob Bugge wrote:

     

    ...

    But in the case shown in this thread it seems to swap sides instead of keeping the original sides.

     

    To bug (us) or not to bug (us).

    I guess I'm very bad explaining it with words, so I made this image to illustrate the logic

     

    Capture.JPG

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 10:06 AM   in reply to emil emil

    I thought so, Emil. So it is completely independent on the original sides of the image, comparing the current W and H values of the Bounding Box of the rotated image to the W and H values of the original image (itself equalling that of the original Bounding Box).

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 10:53 AM   in reply to Jacob Bugge

    Yeah, I'm pretty sure that's the way they made it and is intended to work but it is another question how smart this is and if any user will find the ppi information of objects transformed like that useful and not confusing. In fact in this example, the information is outright misleading for the purpose of finding out the needed ppi of such objects, at the end there are actually 10 real pixels from the raster image available for the width but with the method Illustrator uses they are reported as 5 and for the height it does the opposite.

     

    edit: I don't know how this can be solved but I think it will be better off if for raster objects that are transformed in such way, instead of giving this confusing and misleading ppi information to dim the ppi information or to say something like "ppi not applicable - freely transformed object" because at the end the users have to figure out the available ppi in another way.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 12:03 PM   in reply to Tsnmrz
    For the 1st one i get in the control pannel: PPP: 5, which is obious because you can count by yourself.

    For the second one i get in the control pannel: PPP: 2,5 x 10, and I´m affraid to don´t understand the relationship or the formula i have to apply to this one to get the 5 PPP resolution.

     

    PPP: 5 = 5 * 5 in a perfectly square inch, since your second rectangle is not a square, Illustrator is doing you a favor by giving you a somewhat not so friendly measurements display of the width * height as if it was a square inch...but remember you have 2 square inches.

     

    here's the formula

    resolutionFormula.png

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 12:37 PM   in reply to CarlosCanto

    Carlos,

    at then end of your formula it says:

    W=10 px

    H= 2.5 px

     

     

    the problem we are trying to explain is why Illustrator says:

    W= 2.5 px

    H= 10 px

     

    Capture2.JPG

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 4:13 PM   in reply to Tsnmrz

    Bugs are not limited to programming errors. Something that acts completely unexpectedly by users can also be a bug.

     

    I see no value whatsoever in this behavior. It is ridiculously counter-intuitive. The resolution of this image is not 2.5 PIXELS per inch by 10 PIXELS per inch, just because it is rotated. That is absurd. Such a value would suggest non-square pixels, which this clearly is not.

     

    It's a bug. And just more AI interface slopiness.

     

    JET

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 4:24 PM   in reply to emil emil

    I've could have done it backwards, after all, I was multiplying a 5 by 5 square by a unit value.

    resolutionFormula3.png

    I think there's a Mathematical explanation for having 2.5 x 10...I need try to come up with that....

     

    in the mean time, look at how my values are shown....try to explain that

    resolutionFormula2.gif

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 6:08 PM   in reply to Tsnmrz

    ...I think we're getting somewhere...

     

    from Control Pannel: (PPP 589,45x799...) instead relative resolution (ppp 300)

    .02.png


    is the image above scaled down?, click on "Embeded" in the control panel, then on "Link Information", if it shows other than 1400%, then the image is scaled.

     

    this is what I got in your sample file

    resolutionFormula4.gif

     

     

    then I scaled the image 200%, and the scaling shows 2880%, good.

    resolutionFormula5.gif

     

     

    based on that this formula should do the trick

    resolutionFormula6.png

     

    Message was edited by: CarlosCanto

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 6:35 PM   in reply to JETalmage

    Well, while I agree with you that this behavior is completely unexpected by users and is confusing I would consider a programing bug something that is not working as intended by the programmers. In this case I believe this behavior was intended and is not accidental or random. My reasons are based on the fact that Illustrator is capable of knowing only 3 pieces of information about its raster images and that's not enough to behave as users would expect.

    1. the number of pixels per side (pixel count) of the original unmodified raster image and their ppi
    2. the physical size of the of the original unmodified raster image as specified in the file info or assumed as 72 dpi
    3. the physical size of the modified image in Illustrator

     

    With these 3 pieces of information it can't do much. All it can do is to take the original pixel count per side, figure out which is the corresponding side of the modified raster object for which it uses a bounding box, and then  reports the original number of pixels per side as if mapped on the new size. Illustrator does not have a way to count the actual eventual pixels that will be added or removed  by transformations and deformations because it is not a raster program and doesn't have a canvas grid to count the number of pixels in any direction. All it has is the info from the original file. And until the user rasterizes, save as raster file, or print the file, there is no new pixels in the Illustrator's universe.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 6:27 PM   in reply to CarlosCanto

    CarlosCanto wrote:

    ...

     

    in the mean time, look at how my values are shown....try to explain that

    resolutionFormula2.gif

    Did you rotate the image before that? The problem in question occurs only if the image is not aligned with its original bounding box like when you rotate it.

    I will get the same thing on my panel as yours if I create the image like that, or place it (import). Also at any time after rotated, transformed, deformed, etc. and reported "weirdly" if I choose rasterize it will be reported as expected.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 3, 2012 11:57 PM   in reply to emil emil

    I see what you're saying, and the more I get into it, the more confused I get. As you said, Illustrator is reporting based on the original unmodified raster. I'm not sure about the scaling anymore, I tried with different images of my own and the 1400% was nowhere to be found, I don't know what the scaling percentage means...anyway the formula above should get the ppi right which is for the original post 291 ppi. If the OP says it should be 300, then the image is scaled....

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 4, 2012 2:52 AM   in reply to Tsnmrz

    Obviously, I agree with James on this.

     

    The question seems to be whether Tsnmrz, and others, can use my suggestion of using the Document Info panel, which seems to be consistent in connexion with right angle rotations, in the general case of arbitrary rotation angles, as a general temporary solution for the number of years that it takes to unbug this particular one.

     

    Whay say you, Emil, Carlos, James, does the Document Info panel behave?

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 4, 2012 6:38 AM   in reply to CarlosCanto

    For raster objects, Illustrator assumes  100% scale at 72 ppi unless the raster object originated from a file with info (meta data) saying otherwise.  For example if you rasterize a vector object in Illustrator at 72 ppi  it will be 100% scale above 72 ppi it will be less than 100% and below 72 ppi it will be more than 100%. With placed files, they could be at 100% with any ppi if this is specified so in the file info.

     
    |
    Mark as:
  • Currently Being Moderated
    Apr 4, 2012 7:59 AM   in reply to Jacob Bugge

    Jacob Bugge wrote:

     

    ....

     

    The question seems to be whether Tsnmrz, and others, can use my suggestion of using the Document Info panel, which seems to be consistent in connexion with right angle rotations, in the general case of arbitrary rotation angles, ...

     

    The ppi  info is most useful if the only transformation is  proportional scale. It is also easily understandable if the image is not scaled proportionally (stretched).

    In the first row of the image below I put examples of object transformations that can have useful ppi info.

    In the second row are examples of object transformations and deformations where the ppi info is not much of a help with deciding if the available pixels of the original raster image are enough - relying on common sense will help much more.

    In both rows Illustrator applies the same rule for transformed objects - in the reported ppi, it takes the original number of pixels of a side and uses that number as the pixels of the new corresponding side on the bounding box. However for the deformed objects as in the case of the Twirl, Crystallize tools, and Envelope distort, Illustrator doesn’t report any change in ppi.  This is because objects manipulated with tools like Twirl and Crystallize are rasterized on the fly after each manipulation using the original ppi of the object, and with Envelope distort after expanding.

     

    And this is definitely not a bug by any common definition. May be poor programing design, poor user interface, solution, method, etc, but it is working as intended. Better solution will not come with some sort of debugging but entirely new program design that expands the raster capabilities of Illustrator. For the rotation only the solution can be achieved by making the vector objects more intelligent. For example in inDesign the rotation and other transformations like Scale and Skew are separate extensions to the object description which allows resetting the transformations at any time to the original state of the object and also treating a raster object as not being modified if only rotated. In Illustrator any transformation is baked into the transform values (the objects are dumb) and this doesn't allow to treat the rotated raster objects as not being modified.

    111.jpg

     
    |
    Mark as:

More Like This

  • Retrieving data ...

Bookmarked By (0)

Answers + Points = Status

  • 10 points awarded for Correct Answers
  • 5 points awarded for Helpful Answers
  • 10,000+ points
  • 1,001-10,000 points
  • 501-1,000 points
  • 5-500 points