Most of the time, I use JPGs for images; but, occasionally (uisually with family images) I use very high resolution (600-1200).
Therefore, the files are usually pretty large when saved to TIFF (which I understand is the best format for saving images).
I also make JPG copies; but even they are very large.

I hope there is a very aggressive compression program and/or method to reduce the size of the images so that I can email them.



Who is Participating?
Jose ParrotConnect With a Mentor Graphics ExpertCommented:

An uncompressed bitmap of a image with 32 bits per pixel has the highest size and no loss of quality. An image of 1000 x 1000 pixels waste as much as 32/8 x 1000 x 1000 bytes, or 4 million bytes, almost 4 MB.
BMP can be 24 bpp (bits per pixel), 8 bpp or monochromatic 8 bpp (256 gray levels from all wahite to all black) and has no compression and, of course, no loss.

GIF compression has a good compression algorithm, but uses only 8 bits, but has no loss in quality in its 8 bpp.

JPG has one of the highest compression algorithms, but at a price: the higher the compression, the lower the quality. That is, if the quality is set to 100%, compression is with no loss. If the quality is 80% there are losses of around 20%.
By "loss" we understand the number of pixels that are retrieved from the file with different color regarding the original image. Actually the quality is poor in JPG, but the loss is well distributed through the whole image and in most cases are more than acceptable, if the image has much details and colors which should not be exact to the observer. Example: if a tree image with 100 green levels is stored with half of this number, very few people will notice that and the quality, despite the 50% loss, is still acceptable. More noticiable loss is a color group of 5 or 6 green levels when reduced to a single color in squares of 4 or 5 pixles. Graphics images, like bar graphs, or geometric shapes are not suitable to JPG compression, as the losses are very noticiable. For these images, GIF, TIF or PNG are the best choice.

Conclusion: the highest agressive compression is a JPG with 10% quality. Not interesting... the file is very small but the image is absolutely unacceptable.

Solution: if the images aren't to be printed or their sizes can be small, use JPG 80% for photos and GIF for sharp geometric shapes. And, before send by email, make them 512 x 384 or even 320 x 200 that are good sizes ina screen. To reduce sizes with very few quality loss, use "half size" resizing command twice, set to no dithering and interpolation in the free tool IrfanView (http://www.irfanview.com).

The best compression methods for images are fractal compression, a method that involves creating a mixture of fractals that approximates the appearance of the original image, and wavelet compression, which involves converting the image data into a collection of mathematical formulas that can be used to regenerate the original pixels.

Both of these can produce better compression than JPEG under some conditions, but the hitch is, nobody actually USES these algorithms, except in some highly specialized applications, and the software to view them isn't free.

JPEG really is the best option for emailing photos. Follow Jose's suggestion, and reduce the resolution of your images.
Another option is to upload them to a web site and email links to the pictures instead of sending the actual images as attachments.


Whoops, I sort of misspoke. By "best compression method" I really meant "smallest compressed file," not that it's actually the best approach.

Fractal compression sucks in practice, because it's incredibly slow to compress, requires lots of fiddling about to get good results, and is patented out the wazoo so nobody can use it for free.

Wavelet compression is used in the JPEG 2000 specification.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.