What's new

Jpeg compression


{OZ}

Well-Known Member
Messages
72
Likes
0
I read Erik mentioned about Boxtop ProJpeg plugin. It really does produce much smaller jpg files compared to photoshop given the same compression levels. nice plugin.

There is a program called IrfanView. I've been using it for years. It's one of those "must have" programs. It's an image viewer, converter etc. and its free... by all means free... no ads no nothing.

anyway, where I'm going is that I made a comparison between the jpeg file sizes and quality.

I used the same image, reduced size, compression 60

Here is what I got..

First image is Photoshop 6.0 Save for web.

Fire size 46,893 bytes
 

kiwi

Guru
Messages
593
Likes
0
Hmmm ya makes ya wonder why Adobe cannot match that compression :)


Stu.
 

Erik

Guru
Messages
1,534
Likes
2
I like your approach OZ. Trying things out to verify yourself.

A really difficult test for any jpg are hard edges. JPG is made for photographs, but when you for example fill some selections (no feather ot Anti Alias) with contrasting colours and then try to save as JPG, you will see the "steam" coming out of the edges.

Stu: I'm told that Image readyJPG should be better,...? But since I have BoxtopPro, I always use it. It also has more settings. I could impress you (try to at least) wiyh a lot of technical mumbojumbo like Huffman etc, but I never looked it up myself)
 

sfm

Guru
Messages
1,163
Likes
4
Hey OZ I just took a look at the program that I use in PS to reduce the size of images.......... Altamira Print Pro (www.atamira-group.com) I think that is their web site
anyway I made it the same as yours
the file size is 45,716 bytes
sfm
 

Tron

Guru
Messages
1,176
Likes
9
Hey oz good work there in researching. I read what you said and the went to the Infranview site and downloaded it.
Looks like a pretty good program
Thanks for the info

:perfect:
 

vogonpoet

Well-Known Member
Messages
98
Likes
0
I think its worth mentioning the monitors colour settings also will affect how a graphic is viewed :)

And as we all know they can vary greatly heh
 

sPECtre

Guru
Messages
879
Likes
1
JPEG 2000 and SuperPNG free plug-ins

http://www.fnordware.com/

Beware! do not expect all the images to work with all the extra functions in all browsers (16 bits, JPEG transparency...) they rather provide a very good solution to send light files with less degradation than the "old" [honesty] JPGs or the "regular" PNG (if one sends the plug-in along the files to allow the "destination" user to open the files)
 

Nuke_m

Member
Messages
10
Likes
0
Rick said:
Hey oz good work there in researching. I read what you said and the went to the Infranview site and downloaded it.
Looks like a pretty good program
:perfect:
Beware:
v3.80 is bugged, use v3.75

E-Mail he sent me=
 

MindBender

Guru
Messages
611
Likes
0
PS7.01 save for web... minor tweaking of settings. roughly 24k. It did start out at 46k by default though. You have to play with the optimize settings in "save for web" to get anything out of it.
 

Erik

Guru
Messages
1,534
Likes
2
Did you start from an original in TIFF or PSD, or from one of the jpg's from this thread?
 

MindBender

Guru
Messages
611
Likes
0
Well, it won't make any difference as far as JPEGS are concerned because they don't use a visual compression method. ie, a .gif does compression mostly through number of colors, so if you started with a .gif you could boil it down further, but a jpeg shouldn't make a diff. I took a screenshot of the first tiger image which produces a PICT file (basically the same size and content as a TIFF) and I worked with that. If someone want's to post an uncompressed TIFF file that we could all use as a starting point that might be a good exercise and produce some good comparative data.

I don't have time right now, but if no one has posted by later, I'll track something down. I would suggest a photo AND something like a logo so we can see the differences in compression quality.
 

Erik

Guru
Messages
1,534
Likes
2
Perhaps you did not get my point...your image is less sharp, and the reason of that probably lies in the fact that you had to start from an already lossy pic that will inevitably deteriorate further when it is jaypegged for a second time.
As for this kind of tests: sorry, but I'm not that motivated. I use my proJPG filter and the Irfanview batch for thumbs etc, or Fireworks, and I'm satisfied with these.
And satisfaction makes me lazy... [sleepy]
 

Top