Recommended workflow to retain maximum quality of an image for printing

[Muzician]

Starting with a .jpg taken by a cousin with an iPhone of a painting done by a relative. 

Properties of the original .jpg I was sent per Gimp are: 

42" x 56" 
72 ppi
3.7 mb
RGB

It's a pic in sharp focus of the painting in a frame. The plan is to crop out just the painting, straighten it and bring it down to the original size of 23.5" x 29.25" and have it printed with some border around it for framing.

What do you recommend as the optimal workflow to go from the original to a format to have a print made of it and what method of printing do you recommend? Office Depot does prints that size as posters but they use a variety of inkjet printer and I wonder if there's something better that an actual print shop might use. The goal is to retain as much of the sharpness and detail as possible.

What DPI should I use for best results? I gather most places like .pdf as a file to print from? Is there a particular stock you recommend?

Thanks!

[Ofnuts]

See here for a background understanding of the size of images.

If you say your image is 42"x56" @72DPI, it means it is 3024*4032 pixels, and printed at 23.5"x29.5" image this will be around 130DPI (possibly around 100DPI depending on how wide the frame is in the original photo. Not very good (that would be 300DPI or better) but not bad either for a painting (which people are not going to pixel-peep).

Typical workflow is is to make the image exactly rectangular (the slightest angle in the camera makes it trapezoidal):

• Starting with the initial photo (frame included)
  
• If the colors don't look exactly right, and if there is something in the picture that is naturally gray/white, insert a color correction step here.
  
• With the crop tool, crop the image to the smallest rectangle that still contains the whole painting (this can leave some bits of the frame).
  
• Start the perspective tool, and set it to Direction: Corrective and the Clipping: Clip
  
• Drag the four handles in the corners to the 4 corners of the painting, and apply the transform. This will put just the painting in the canvas
  
• If there are optical distortions in the picture, there can be small bits of the frame left on the sides, it yours to decide to crop the picture a bit more at that point or to keep these or to fix the problem by other means (cloning....)
  
• Compute the aspect ratio (width/height) of the picture you obtained with the AR of the original picture (0.8) and scale the image vertically or horizontally to restore the aspect ratio.
  

Otherwise, all images taken with a smartphone are in sharp focus, because the smartphone lenses have a very small focal length and this gives them a very large depth of field (advanced smartphones use post-processing to obtain the same background blur (aka "bokeh") that you have with full-size cameras). But they are also over-processed, to look right on the phone screen, and not in print (color balance, etc...). You "sharp focus" could be my "sharpening artifacts", not too visible on a smartphone but an eyesore once you see them.

[Muzician]

(12-19-2022, 09:07 AM)Ofnuts Wrote: See here for a background understanding of the size of images.

If you say your image is 42"x56" @72DPI, it means it is 3024*4032 pixels, and printed at 23.5"x29.5" image this will be around 130DPI (possibly around 100DPI depending on how wide the frame is in the original photo. Not very good (that would be 300DPI or better) but not bad either for a painting (which people are not going to pixel-peep).

Typical workflow is is to make the image exactly rectangular (the slightest angle in the camera makes it trapezoidal):

• Starting with the initial photo (frame included)
  
• If the colors don't look exactly right, and if there is something in the picture that is naturally gray/white, insert a color correction step here.
  
• With the crop tool, crop the image to the smallest rectangle that still contains the whole painting (this can leave some bits of the frame).
  
• Start the perspective tool, and set it to Direction: Corrective and the Clipping: Clip
  
• Drag the four handles in the corners to the 4 corners of the painting, and apply the transform. This will put just the painting in the canvas
  
• If there are optical distortions in the picture, there can be small bits of the frame left on the sides, it yours to decide to crop the picture a bit more at that point or to keep these or to fix the problem by other means (cloning....)
  
• Compute the aspect ratio (width/height) of the picture you obtained with the AR of the original picture (0.8) and scale the image vertically or horizontally to restore the aspect ratio.
  

Otherwise, all images taken with a smartphone are in sharp focus, because the smartphone lenses have a very small focal length and this gives them a very large depth of field (advanced smartphones use post-processing to obtain the same background blur (aka "bokeh") that you have with full-size cameras). But they are also over-processed, to look right on the phone screen, and not in print (color balance, etc...). You "sharp focus" could be my "sharpening artifacts", not too visible on a smartphone but an eyesore once you see them.

Thanks! You feel this will result in the least amount of degradation to the original .jpg image? I'm of course working off a copy of it. I gather it would have been better if she'd taken the pic as a .tiff file but she's kind of a technophobe and it was something of an ordeal for her to take the pic in the first place and don't want to ask her to re-do the photos and I don't know if she'd be able to figure out how to change the file type.

I'm using a 4k TV as my monitor. I haven't seen and am not in proximity to the original so can't compare but even if the image is a bit over-sharpened from what I've seen of how this Office Depot poster printer performs it's going to soften the image a bit anyway.

[Muzician]

(12-19-2022, 09:07 AM)Ofnuts Wrote: Typical workflow is is to make the image exactly rectangular (the slightest angle in the camera makes it trapezoidal):

• Starting with the initial photo (frame included)
  
• If the colors don't look exactly right, and if there is something in the picture that is naturally gray/white, insert a color correction step here.
  
• With the crop tool, crop the image to the smallest rectangle that still contains the whole painting (this can leave some bits of the frame).
  
• Start the perspective tool, and set it to Direction: Corrective and the Clipping: Clip
  
• Drag the four handles in the corners to the 4 corners of the painting, and apply the transform. This will put just the painting in the canvas
  
• If there are optical distortions in the picture, there can be small bits of the frame left on the sides, it yours to decide to crop the picture a bit more at that point or to keep these or to fix the problem by other means (cloning....)
  
• Compute the aspect ratio (width/height) of the picture you obtained with the AR of the original picture (0.8) and scale the image vertically or horizontally to restore the aspect ratio.
  

Following these directions what ends up happening is parts of the canvas get chopped off around the edges.

[Ofnuts]

(12-19-2022, 05:14 PM)Muzician Wrote:

(12-19-2022, 09:07 AM)Ofnuts Wrote: See here for a background understanding of the size of images.

If you say your image is 42"x56" @72DPI, it means it is 3024*4032 pixels, and printed at 23.5"x29.5" image this will be around 130DPI (possibly around 100DPI depending on how wide the frame is in the original photo. Not very good (that would be 300DPI or better) but not bad either for a painting (which people are not going to pixel-peep).

Typical workflow is is to make the image exactly rectangular (the slightest angle in the camera makes it trapezoidal):

• Starting with the initial photo (frame included)
  
• If the colors don't look exactly right, and if there is something in the picture that is naturally gray/white, insert a color correction step here.
  
• With the crop tool, crop the image to the smallest rectangle that still contains the whole painting (this can leave some bits of the frame).
  
• Start the perspective tool, and set it to Direction: Corrective and the Clipping: Clip
  
• Drag the four handles in the corners to the 4 corners of the painting, and apply the transform. This will put just the painting in the canvas
  
• If there are optical distortions in the picture, there can be small bits of the frame left on the sides, it yours to decide to crop the picture a bit more at that point or to keep these or to fix the problem by other means (cloning....)
  
• Compute the aspect ratio (width/height) of the picture you obtained with the AR of the original picture (0.8) and scale the image vertically or horizontally to restore the aspect ratio.
  

Otherwise, all images taken with a smartphone are in sharp focus, because the smartphone lenses have a very small focal length and this gives them a very large depth of field (advanced smartphones use post-processing to obtain the same background blur (aka "bokeh") that you have with full-size cameras). But they are also over-processed, to look right on the phone screen, and not in print (color balance, etc...). You "sharp focus" could be my "sharpening artifacts", not too visible on a smartphone but an eyesore once you see them.

Thanks! You feel this will result in the least amount of degradation to the original .jpg image? I'm of course working off a copy of it. I gather it would have been better if she'd taken the pic as a .tiff file but she's kind of a technophobe and it was something of an ordeal for her to take the pic in the first place and don't want to ask her to re-do the photos and I don't know if she'd be able to figure out how to change the file type.

I'm using a 4k TV as my monitor. I haven't seen and am not in proximity to the original so can't compare but even if the image is a bit over-sharpened from what I've seen of how this Office Depot poster printer performs it's going to soften the image a bit anyway.

TIFF wouldn't make a difference, the degradation is in the heavy-handed processing.

[Ofnuts]

(12-19-2022, 06:40 PM)Muzician Wrote:

(12-19-2022, 09:07 AM)Ofnuts Wrote: Typical workflow is is to make the image exactly rectangular (the slightest angle in the camera makes it trapezoidal):

• Starting with the initial photo (frame included)
  
• If the colors don't look exactly right, and if there is something in the picture that is naturally gray/white, insert a color correction step here.
  
• With the crop tool, crop the image to the smallest rectangle that still contains the whole painting (this can leave some bits of the frame).
  
• Start the perspective tool, and set it to Direction: Corrective and the Clipping: Clip
  
• Drag the four handles in the corners to the 4 corners of the painting, and apply the transform. This will put just the painting in the canvas
  
• If there are optical distortions in the picture, there can be small bits of the frame left on the sides, it yours to decide to crop the picture a bit more at that point or to keep these or to fix the problem by other means (cloning....)
  
• Compute the aspect ratio (width/height) of the picture you obtained with the AR of the original picture (0.8) and scale the image vertically or horizontally to restore the aspect ratio.
  

Following these directions what ends up happening is parts of the canvas get chopped off around the edges.

Screenshot?

Or use Clipping:Adjust and after the transform is applied Image ➤ Fit canvas to layers. But then you will have to crop.

[Muzician]

(12-20-2022, 12:50 AM)Ofnuts Wrote:

(12-19-2022, 06:40 PM)Muzician Wrote: Following these directions what ends up happening is parts of the canvas get chopped off around the edges.

Screenshot?

Or use Clipping:Adjust and after the transform is applied Image ➤ Fit canvas to layers. But then you will have to crop.

Showing the cropped image with perspective tool selected and the settings:






Handles at corners of canvas:







Results after perspective transform applied:

[Krikor]

Quote:"With the crop tool, crop the image to the smallest rectangle that still contains the whole painting (this can leave some bits of the frame)."

I believe that this point should be better observed.

After the handles are dragged to the corners of the image, they will be aligned to the initial rectangle defined by the crop tool.

[Ofnuts]

In the preliminary crop, set the Delete cropped pixels option to that the result is the size of the canvas.  Then

   

   

   

[Muzician]

(12-20-2022, 08:46 AM)Ofnuts Wrote: In the preliminary crop, set the Delete cropped pixels option to that the result is the size of the canvas.  Then

That "Delete cropped pixels" did the trick. Thanks!