First, what is a JPEG? Basically, it is a common lossy compressed image file format developed by the Joint Photographic Experts Group that allows you to work with image files that are much smaller in size. There are a variety of compression ratios and quality settings from which to choose depending upon your needs. It is claimed that at its lowest compression ratio, there is little or no perceptible difference in image quality.
Here is what I found on Brittanica.com:
The JPEG standard works by averaging color variation and discarding what the human eye cannot see, a process known as “lossy” compression. Depending on the level of compression, it is possible to compress an image by a factor of 100 to 1, though there may be some loss of quality at the compression limits. Factors of 20 to 1 are easily done, however, and the loss of quality is practically invisible to the human eye. In comparison, the loss-less graphics interchange format (GIF) only compresses by about 4 to 1.Of course, RAW files are just unprocessed data captured by the camera's sensor without any assignment of color, contrast, saturation, sharpness, etc. You can't actually view a RAW file so the camera manufacturers embed a JPEG inside the RAW file otherwise you wouldn't be able to view it on your LCD. TIFF files are also typically files that do not lose any image quality as well.
Today's digital sensors are so good that the associated JPEG files they produce are exponentially better than in the past, in my opinion. Of course, the camera's image processing engine has a lot to do with that as well. Typically, I still shoot RAW but I don't hesitate shooting JPEGs if more convenient unless the situation is outside of what I would call everyday or normal photography. That is situations of really high contrast, strange mixed lighting conditions and at very high ISOs. In other words, the exception to what most people's photography normally would be.
Photographing in JPEG still has its drawbacks, such as a lesser ability to recover highlights, a lesser bit depth and file compression. An 8-bit JPEG file can restrict your ability to make significant adjustments to exposure, white balance, saturation, etc. due to fewer available colors which can cause "banding" in areas such as a plain blue sky. Compression artifacts and loss of fine detail have been a worry in the past as well. But in my experience I've found JPEG files from my camera very malleable and able to take some pretty significant adjustments if necessary. In fact, whenever I send a file off to a pro lab to large prints, I send a JPEG version instead of a TIFF or RAW version of the file. I never could tell any difference.
Here is my analogy of JPEGs versus RAW files. I think of JPEGs as you going to a bakery and buying a cake already made. The baker chose what ingredients to include in what quantities, what flavors to add and what kind of icing to use. RAW files are like buying all the ingredients for making a cake but you make it at home and have a full range of choices on exactly how you want your cake to be made. Almost.
That analogy doesn't fully work as you get to choose a number of in-camera settings to determine how you want your "cake" (JPEG) to be made. You can choose color balance, sharpness, contrast, camera profile/film simulation, etc. So you do have some say in how your JPEG files are "baked." I guess a better analogy would be you telling the baker how you want your cake made and then the baker making it to your specifications. But I digress.
If you do everything you can to get the JPEG right in the camera, my sense is that you can successfully use JPEGs most of the time for most of your photography with no penalty. Many will disagree but notice I put some heavy qualifiers in that statement. Getting it right in the camera and the qualifier "most" are crucial.
The minimum compression for JPEGs at their highest quality is currently 25% (if I'm wrong please let me know). That means, even when saving at the best and highest quality JPEG setting, you are losing 25% of the information in your already restricted file. So, how many times you can open your JPEG, edit it, then re-save it before you see banding, compression artifacts and start to lose fine details. I set up some experiments with a couple of different files to see just how many times I could re-save them and still retain high quality.
I chose two different files for my experiments. The first file, the Grand Canyon, contains some very fine detail both in the rocks and in the distant trees on the far rim. After opening, editing and re-saving it several times, I expect to see a loss of detail in both areas. The question becomes how many times can I re-save the file and retain the fine detail?
The second, the entrance to the famous Cadillac Ranch in Amarillo, Texas, has a plain blue sky with a variety of shades from light to dark also with some fine detail in the distant background. After opening, editing and re-saving it several times, I expect to see some banding occur in the blue sky and a loss of detail. Again, the question is how many times can I do that.
Both files are 36mp from my Nikon D800E, which has a unique anti-aliasing filter system to keep fine detail intact that a normal anti-aliasing filter would slightly smear. If you want to better understand this system, look here. I exported both files full size out of Lightroom Classic as JPEGs using a setting of 80, which is my normal setting. I then opened each file in Photoshop and created a duplicate file which would be the ones I would repeatedly open and re-save. Each time I opened a file I did a minor manipulation (increased the luminance by 2, then the next time I opened it I reduced the luminance by 2), then re-saved the JPEG files at maximum quality. I wasn't trying to manipulate the files extensively to cause artifacts, but test how many re-saves could be accomplished before producing artifacts. In other words, a best case scenario.
Between each cycle of opening/saving, I compared the original with the duplicate. It was after repeating this process 10 times that I started to very faintly see some artifacts in the plain blue sky of the Cadillac Ranch image. But I saw no loss of detail in either file, even after opening/re-saving 10 times. I didn't go farther with the experiment looking for the "loss of detail" threshold as I thought is was unlikely that an individual would open, edit and re-save a JPEG image file as many as 10 times to begin with.
In the blue sky images posted below I'm hoping you can see the slight artifacts I see when looking at them on my monitor. You may not, depending upon Google's compression algorithms as well as potential variables. To make the sky artifacts potentially easier to see, I enhanced them in the second image using the texture and clarity sliders.
In these two examples, the answer I found to my question is about 10 times. Other image files from other cameras using slightly different methodologies may differ and I don't have enough data to extrapolate these results to the general image population. It is clear to me that artifacts appear in the clear blue sky. That being said, I did not see any degradation of fine detail in either file. I looked and looked and, even after 10 re-saves, the fine detail held up.
![]() |
This is the same as the above only I enhanced the artifacts to make them easier to see. (click to enlarge) |
![]() |
100% crop of original file showing fine detail on the horizon (click to enlarge) |
![]() |
100% crop of JPEG file that had been opened, edited and re-saved 10 times. In looking at the files on my computer monitor, I cannot notice any loss of fine detail. (click to enlarge) |
Join me over at Instagram @dennisamook or my website, www.dennismook.com.
Thanks for looking. Enjoy!
Dennis A. Mook
All content on this blog is © 2013-2020 Dennis A. Mook. All Rights Reserved. Feel free to point to this blog from your website with full attribution. Permission may be granted for commercial use. Please contact Mr. Mook to discuss permission to reproduce the blog posts and/or images.
No comments:
Post a Comment