Thanks for your analysis and question.
Yes, in this case, with a very small sample image for testing convenience, the difference is by 2kB. In real cases, we have experienced changes from 160kB to 220kB. Everything in the PDF file is the same, except the images, that are no more indexed, thus their stream are dramatically bigger.
In the linked example, PDF object 23 0 obj has passed from 7968 to 10175, thus a 20% more. In a file with many indexed images, this is too heavy. I linked an image to show this difference in the PDF image dictionary.
On my side, the point is: what changed the image treatment and how to enable their indexed mode?
For sure not all images can be indexed, but, when possible, it would be better to continue to convert them as indexed, as done in the past.
I’m asking if the new behavior can be moved to the old one via some PdfSaveOptions property or if you can kindly investigate to explain the change and, if possible, to give the previous good conversion behavior.
Might be anything missing or not clear, please let us know.
Thank you for help.
Differences.png (28.8 KB)