[SOLVED] Easiest way to join 2 .jpg's...

jk_1966

Reputable
May 28, 2021
51
1
4,535
Hello,
Which is the easiest way to do the above with minimum quality loss? I have a picture cut in 2 separate .jpg files (name them, upper half & lower half)
So far use Windows Paint like this:
  1. I open both files with Paint and Zoom Out once on both.
  2. On the upper half I drag the down border of the picture to extend the its length to about the double
  3. On the down half file I choose Select All and then Copy
  4. I paste it on the extended space.
At the joined photo I noticed that the quality of the lower half is not so good compared to the upper half....
What can I do to improve it? Do you see something wrong with my "technique" ?
 
Windows Paint is a poor tool for this.

paint.net is almost certainly better, and also free.

Or GIMP. But the learning curve can be steep.
 
Solution
Save your "original" jpegs as lossless, such as tif or bmp.....if possible in Paint. I don't use it.

Do your editing and fiddling with the tifs or bmps rather than with jpegs.

When done with the editing and fiddling, possibly resave the tif or bmp as jpeg if you must, but preferably leave it as a tif or bmp for the least quality loss.

At small sizes, you may not be able to tell the difference.
 
  • Like
Reactions: jk_1966
Irfanview
Go to image->create panorama image->add and sort your images and you are done.
When you go to file-> save as... to save the final pic you get a quality settings window displayed alongside the save dialog so you can choose the quality of the final picture.
A lot of trial and error is needed to get the best quality/size ratio.
 
I have a picture cut in 2 separate .jpg files
Under certain conditions, you can fully lossless combine two files using the hex-editor (you need to add RST markers). The necessary conditions are:
  1. Both images are the same width.
  2. The height of the first image is a multiple of MCU height (8 in common case).
  3. Both images have baseline Huffman encoding.
  4. Both images use the same Huffman and Quantization tables.