Adobe Acrobat 9 query

Question

Hi,
I have downloaded the trial of Acrobat 9 to test.
I have a folder of 3600+ greyscale jpg images which I converted to pdf using PDF Toolbox and Acrobat 7. The resulting files were 6MB over a full DVD. So I opened one of the larger files (50MB+) in Acrobat 7 and used the reduce file size function. This halved the file size.
Out of curiosity I ran all the files through Acrobat 9 using batch processing and the reduce file size function. The resulting pdf's amount to 250MB! I open some of the files and there is little difference between the large (acrobat 7) files and the smaller ones.
I also opened one of the files (50KB) and extracted the 2 pages as jpg images anly to find that each jpg was around 500KB.
How can the pdf be 50KB if the 2 images it contains are 500KB?
Also, does the standard version have the batch processing function like the Pro version?
Thanks

Answer 1

If there is no OCR involved, I'm inclined to wonder why you would want to contain each individual image in a PDF anyway. What functionality is PDF bringing to the table for you?
Is it simply a matter of providing a browsing mechanism? If so, wouldn't it be simpler to just use a browser application? For example:
1. Scan the images as grayscale TIFFs.
As others have explained, you are misreading your results if you think JPEG yields "better image quality" than TIFF. This is what's going on: JPEG is lossy compression. It throws away data by evaluating adjacent pixels. When it finds a range of contiguous pixels of similar value, it sets them all to the same value. Then it uses data compression to store the ranges of values.
So the reason you think JPEG is giving you "better" images is: contiguous areas of similar but different relatively light grayscale values are getting converted to white or light grays. In other words, the JPEG lossy conversion is effectively clipping and posterizing the light end of the grayscale range. The "improvement" you are seeing is just a happy accident, and is far less than the improvement you could likely gain by a better workflow that gives you better control.
To improve a batch of similar-quality grayscale scans of what should be line-art originals (that's what you have here), you need to do two things: Maximize the dynamic range, and then adjust the gamma (midpoint) to improve the contrast. In other words, you want the lightest areas to go all the way to white, the darkest areas to go all the way to black, and gray value that best describes the "edge" between black and white to go to 50%. Once you have done that, you can then convert the grayscale images to 1-bit images, in which all pixels are one of two values: White or black. The end result will be more legible and the data size will be much smaller. Then you can use lossless compression to further reduce the file size.
Depending upon the sophisication of the scanning sofware (and the undestanding of the user), all that
can be accomplished in the original scan with optimized settings. If the scanning software is crude, then the adjustment can be accomplished instead by batch-processing the images with Photoshop, applying a curve to the original grayscale TIFFs, optionally resampling, and then saving to the desired format.
2. Now that the images are optimized and reasonably-sized files, you just need to deliver them in a browsable form. That brings me back to my original question. If no OCR is involved, then how are the images identified and organized? If they are to be identified by their file names, then I wonder if Acrobat is the best mechanism, since it's not really adding any value. Sure, you can open a folder full of individual images with Acrobat, converting each to individual PDFs, or combining them into a single PDF and/or further combining them into a PDF package. Doing so, Acrobat can give you a "table of contents" by generating bookmarks for each image.
But if no OCR is involved, how is the dumb list of bookmarks any better than simply browsing the file names on disk? Wouldn't you be better off with one of many image browsing/cataloging applications? For one example, consider these simple steps in FileMaker Pro:
1. New database. Two fields: A text field named "FileName" and a container field named "Image".
2. ImportRecords>From Folder. The import dialog lets you map the image files into the container field
as a link, and the file name into the text field.
3. Bind the database as a standalone runtime application.
That's it. You're basically done. Now consider the versatility you have gained by using a proper database as a browser, instead of PDFs:
- The title field is searchable any way the user wants. Rather than just being able to do dumb linear searches, he can do more intelligent searches using logical operators. (And, Or, Not, Greater Than, Less Than, etc.)
- The found records can be sorted any way the the user wants.
- You can add additional fields to add value to the solution. For example, add one or two or a whole series of fields that you or the end user can use to categorize the records. For example, suppose part of that un-OCR-able hand-written image is a signature. The user could, at his choice and convenience while browsing the material, key the signature names into a "Signer" field. Doing so, he would be continually improving the usablilty of the solution. Eventually, he would be able to do Finds and Sorts on the signature names. And so on.
Another alternative would be to use an HTML form as the browser. That could also be delivered on the disk along with the images, would be launchable in the user's web browser, and could also be customized for additional functionality.
I'm not knocking Acrobat or PDF. But if PDF isn't really being used to any value (such as its ability to quickly generate a full-text index from OCR content), and if all you are doing is browsing a bunch of images, you might want to consider other delivery mechanisms that yield more versatile and higher-quality results than what you have so far described.
JET