I'll preface this by saying that you still haven't provided enough information about the before and after files for me to be certain of any answer. Ideally, showing the composition of layers and their properties would help a lot.
You also haven't expressed your understanding of
jpeg compression so I don't know if my answer is repeating information you already know. Nonetheless, I hope it's useful:
I'm just after seeing your answer to a separate question on
files, it appears that you're likely familiar with everything I describe below, which means I may have guessed the missing details of your question wrong. I'll leave it incase it's useful for anybody else as it's still relevant to the current details should anyone be searching for the same question.
I have a hunch that this is down to confusion over how various files (
) store their information.
Jpegs shrink the filesize of images by lossy compression, meaning that if you have an image that's
10 pixels by
10 pixels, it can save it in a filesize smaller than it would have typically taken to store what those
100 pixels were, but that the
100 pixels would't be identical to the originals. Hence, lossy. It's advantage is that we can't, typically, perceieve that they're different with our eyes. The more compression you use, the more likely we are to perceive it.
PSDs also use compression to shrink filesize but, conversely, it's lossless - meaning that whatever tricks it uses to shrink the filesize of
10 pixels by
10 pixels, the resulting
100 pixels will be identical to the original
This has an advantage in image data integrity. You will likely be editing the images and want to know each save and open of the file is exactly as you left it.
The disadvantage is that this type of compression doesn't produce as drastic a reduction in file sizes as
Saving JPEGs as PSDs
If you have a
PSD open in
10 pixels by
10 pixels, you can bring a
JPEG into the software by any means (placing, copying and pasting, etc.) but once it's saved, the
100 pixels of
JPEG information are saved using the
PSD compression technique instead. They are now just
100 pixels, not a
JPEGs impression of
100 pixels. The filesize will likely go up.
But how does the filesize stay down on massive layers of pure colour or large areas of colour?
This is one area that
PSD compression is good at. Shrinking the filesize dramatically on areas of the same colour. So blank layers and white layers of any dimensions will have a very small file size. Introduce a photograph or drawing as another layer (or in place of that layer) and the
PSD can no longer see lots of the same colour next to each other and can't hope to keep the file size nearly as small when it tries to compress.
It's worth knowing that for speed of editing and previewing, PhotoShop also saves a hidden preview layer of each layer in your
PSD at 1:1 pixels. This little known piece of information answers a lot of confusion about ballooning filesizes and frustrations as to how
smart objects and copies of
smart objects can see file sizes increase against the belief of how they're expected to work. I digress, but it's related to the subject and useful to know.
If the information here was all already known to you or is way off the problem, I'd suggest editing the question to include a lot more information about the before and after files.