NavList:

Hi Alan

Frank and Don have explained it properly.  The JPEG image may be saved at levels, trading image quality for size.  The sharper and crisper the image quality (like crisp readable text versus fuzzy edges) the larger the file size.

Given a fixed jpeg, wrapping in a PDF container just adds bytes. There is no savings when the fixed JPEG is wrapped in a PDF You use a PDF when you want to control the entire package presentation, text, images, pagination & etc.  Its fixed in presentation and therefore much harder (for the average user) to manipulate or change.

When Kermit, who I have lots of respect and admiration for, saved the scanned image as a PDF, it MUST have been at lower quality JPEG.  The sum of the low quality JPEG plus the PDF wrapper bytes was less than a high quality JPEG image alone.  His data point is correct but not in detail.  If we were to unwrap his PDF, to extract the JPEG from it (which can be done, albeit by non standard users), we would find that the JPEG is smaller in size and therefore, lower in quality.  Think of it as a false horizon under the moon.  It can trick even the best.

Its a judgement call as to what quality you want.  Consider this.  The latest iPad, with the so called retina display, has pixels smaller than the naked eye can resolve.  The crowd is wow'd.  Text is super crisp.  But the much cheaper Kindle has perfectly readable text.  Which one is good for your purposes?  I'd say, judgement call.

I've been around the software & computer field since 1969.  So this is not just my opinion, but rather a technical assessment. 

There are no sides to take here.

Brad