N
N
Nostromo2015-06-05 14:40:28
JPEG
Nostromo, 2015-06-05 14:40:28

How to compress images without quality loss?

In which program is it better to compress a jpg image without quality loss. After photographing objects, 1 photo weighs more than 5 mb, when compressed it comes from 700-900 kb, although the image size is 1280 * 800 .
I found an example where the image weighs 174 kb, although the quality remains www.orangecountyresorthotel.com/rooms.asp#tab6 .
I would be grateful for a detailed answer.
Thank you.

Answer the question

In order to leave comments, you need to log in

3 answer(s)
D
Denis Ineshin, 2015-06-05
@Armashko

JPEG is always lossy compression. Another thing is that it is almost invisible in the photographs. For example, in Photoshop, save for web, select 81% optimized, you should get what you need.
At one time, when I was making up websites, I experimented with compression quality, in some illustrations it was possible to achieve acceptable visible quality even with 61% compression. But all this is done by eye. So if you want to do it automatically, then you don’t need to set less than 81.

D
Deerenaros, 2015-06-05
@Deerenaros

The given photos have not very high resolution - 1280x795. At the same time, there is practically no noise in the photo.
In general, compression itself occurs approximately as follows - a certain source, giving out information, is subject to repetition. Moreover, repetitions are usually different - some sequences are repeated more often than others: by encoding frequent sequences with shorter sequences and rare sequences with longer ones, you can get a gain in the final size.
Sources are different. In addition, there are different receivers. Some, like our eyes and brains, have severe limitations. So, for example, two adjacent pixels that have close values ​​​​(that is, encoding similar shades) will be processed by the brain so that the shades will not differ. That is, you can make these pixels the same - there will be no difference for the brain, but for the same sequences it will increase, allowing you to compress information more efficiently.
If you take a close look at the images, you will notice a few interesting things - the image itself is monochromatic, rare contrasting borders separate almost monochromatic areas. This compresses perfectly and png.
So we came to the problem - firstly, noise makes compression very inefficient - they must be removed. Retouching to help you. If the image is very multi-colored, it is worth removing unnecessary unnecessary colors - by cropping or applying filters. In addition, some very natural moments, such as pimples and blackheads on faces, folds on clothes and in the interior, a sharp background, not only do they give a not very aesthetic appearance to the object of photography, but also complicate the perception of the composition. That is, removing unnecessary objects will not only improve compression, but also greatly contribute to perception. Finally, color correction - if it is important to get the minimum image size, greatly increase the contrast and dynamic range - not the best idea, but you can make a compromise - expanding the range to reduce contrast and vice versa.

R
Roman Mirilaczvili, 2015-07-07
@2ord

Read the article What About mozjpeg? and draw a conclusion.

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question