TheDman Loc: USA

Correct. And the bigger the adjustment, the more lossy each save will be.

Did my own test. Set up a Photoshop action that did the following:
1. Open file
2. Levels adjustment 0 - 1.20 - 255
3. Save
4. Close
5. Open file
6. Levels adjustment 0 - 0.83 - 255
7. Save
8.Close

Ran that action 50 times, so 100 saves. Here are the results.

Original


100 saves


Original zoom


100 saves zoom

jerryc41 Loc: Catskill Mts of NY

Yes, it was laughable when all three got snoring. Now we have just two.

wdross Loc: Castle Rock, Colorado

This is just what I would expect. Just opening a JPEG, doing a change (even a 100,000 changes), and closing the file will not cause any changes - until the file is saved and stored by the computer. Just closing a file does not save the file. It just saves the instructions as to what you want to do to the file once one hits the save button. When the save button is hit, all those changes, even if it is 100,000 changes, is executed and one new JPEG file is generated. That JPEG file will have thrown out some information due to compression. That information is not recoverable. As more and more saves, not closings, are done, more and more information is thrown out to the point moire patterns, computer generated patterns from the software and algorithms, are generated.

What jerryc41 did would be valid if he opened the JPEG, made changes, closed the file 100 times, and then save the file as one new JPEG, but I do not believe he generated 100 JPEG files. He generated just one new JPEG file with 100 actions done to it. His photograph would look hugely different if he had save his file 100 times generating 100 new JPEGs.

Dngallagher Loc: Wilmington De.

Bill_de Loc: US

Years ago when the best HTML editor was Notepad I helped a friend build a commercial website. I was new to digital photography and the camera I used for photographing the products and location were JPG, no choice. As we sent pictures back and forth, each of us making edits, we got to a point where we decided the finished images would be an embarrassment if we used them. That's when I started reading more about digital. We learned on our own, before reading it, that JPGs did indeed deteriorate with multiple edit and saves.

But, as CameraPapi pointed out, that was then and this is now.

I still believe the there is some loss of quality, since every time you save a JPG it tries to compress the file more. It does it by discarding information. The difference today is that we start with more information and compression algorithms have improved.

At a seminar years back Jeff Schewe, a PhotoShop guru of sorts, made a statement that every edit to an image effects image quality. He said individually it is negligible, but with enough edits (many, many) it could be noticeable. His point was that digital is all about numbers and the calculation done by the computer round numbers up and down. Over time those slight 'errors' are multiplied. The three Adobe engineers that were there all agreed with him.

IMHO it isn't a matter of does the image get degraded, but to what extent it occurs. If we can't see it in the end product, why worry about it.

--

cambriaman Loc: Central CA Coast

I think the corruption of image quality occurs in JPEG only when you actually change the image. Opening and saving it does not cause a recompression of the image as I understand it. I have no idea of how much of a change you would need to make, maybe just a minor crop of a few pixels off one edge might do it.

Rongnongno Loc: FL

Actually it does. Opening decompress and show the compression artifacts.

Saving this image re-compresses the image using the 'uncompressed' artifact from the previous compression. This is what I did when I posted my results. If I had used a greater compression ration the results would have been much worse. (cumulative compression comes to mind...)

When you edit an image and save over and over (w/o closing) there is no re-compression. This is what Jerry did and this is where the compression understanding error creeps in.

Dngallagher Loc: Wilmington De.

Visible on screen, easy to toggle between shots to see the results of deterioration due to jpg open, save, close - hardly a myth.

http://www.dpreview.com/forums/post/31964153

Bill_de Loc: US

If you open an image and make no changes there is nothing to save. You can do a save as if you want to change the name. This would create an additional file.

--

les2297 Loc: Central OR

The conclusion is ... you lose a lot of data.
The first pic is 609 KB, 2326 pixels x 1988
The 100 pic is 50.8 KB, 600 pixels x 515
Still works to view and show on your smart phone, but not worth much to make a nice 20 x24 in. enlargement

Rongnongno Loc: FL

Please check your... 'Data'
The pixel size does not change. You seem to be comparing the 'original' to UHH thumbnail.

les2297 Loc: Central OR

downloaded the files.. used properties to gather data (Win 10)


you're right... the second file only loses about 25 KB of data.

jerryc41 Loc: Catskill Mts of NY

As I said, I was not trying to prove anything. I simply presented the results after 100 saves. As I also said, I opened, modified, saved, and closed 100 times.

jerryc41 Loc: Catskill Mts of NY

That is exactly what I did. I selected a picture, modified it, saved it with a different name, and closed it. I repeated that process 100 times, and I have 100 very similar images. I renamed the first one "First Picture" and the last on "Picture #100."

I didn't do this to generate controversy, just to show what happened after I made 100 changes and saves. If you use a different procedure, you would probably get different results.

wdross Loc: Castle Rock, Colorado

A word of warning for all the UHHs that think the software and algorithms have changed that much for JPEG compression. Do your own test of opening a JPEG, modifying that JPEG, and then saving it as a new JPEG at least 10 times. It is my guess that you will find, with one's normal modifications to photographs, they will not want to venture much pass opening and saving the image pass three new JPEGs. I would suggest one decide at what point they want to stop.