Wikipedia:Featured picture candidates/Image:Quality comaprison jpg vs saveforweb.jpg

JPG Quality Comparison Chart[edit]

Decided to expand my horizons a bit here and try something new - a chart! I've had this idea on my mind for a while and it's finally done - actually took a surprisingly long time.

I didn't use one of the "standard" images as I felt this one was much better. It has both a nicely graduated sky (an area which typically suffers most from compression) and high detail areas such as the grass, sand and planks.

And yes I know it has a spelling mistake but does it warrant the 11mb upload time? Anyway I think its a good enc image with plenty of value. Perhaps a animation might be good too, but I'll see what you think...

Appears in JPEG and Image compression

Support Self Nom. --Fir0002 12:42, 1 October 2007 (UTC)[reply]
Support--Mbz1 13:01, 1 October 2007 (UTC)[reply]
Comment I see some jpeg artifacts. ^8thstar 16:48, 1 October 2007 (UTC)[reply]
Support This is a great diagram, with the spelling fixed. From nom, "..but does it warrant the 11mb upload time?" -- But yet you would consider making this into an animation, which is exponentially larger? ;) Puddyglum 17:02, 1 October 2007 (UTC)[reply]
Support - Interesting and informative. I fixed the filename. --Sean 17:04, 1 October 2007 (UTC)[reply]
Oppose In order to view this comparison properly, the entire image would have to be downloaded. As it appears on this page (and would appear on the front page), it just looks like a small graph with a bunch of tiny thumbnails... not 'featured picture' material, in my opinion. The fact that the entire image itself has been saved as a jpeg doesn't truly give a good comparison either. And you don't really explain what 'saved for web jpg' and 'saved as jpg' mean, and why there is a difference, and you don't state what program you used to generate the thumbnails. This is also Original Research. Rawr 17:21, 1 October 2007 (UTC)[reply]

Why does it not give a good comparison? I saved it as a 12 quality jpg (hence the 11mb file size). I could have saved as a PNG but that would have been ridiculous. As the chart shows there is very minimal quality loss at jpg 12. Ok well if you were confused by it I apologise. "Saved for jpg" means that the image has been saved directly as a jpg with a standard compression on a quality scale of 0-12. "Saved for web" means it has gone through an optimization feature found in many image editing programs - in this case it was done in Photoshop. "Save for web" optimization is made to reduce file size compared to a normal "save as jpg". As can be seen by the graph this is generally the case, although mid way this is not true. Furthermore the images in the chart allow you to compare the resulting quality from the two compression techniques. I'm sorry you haven't found it informative because personally I found it so. --Fir0002 23:19, 1 October 2007 (UTC)[reply]

Oppose not all that aesthetically appealing, Original Research, and lots of elements are awkward, like the stetched fonts, the Excel-default colors in the graph, the empty space in the bottom row, etc. Debivort 18:26, 1 October 2007 (UTC)[reply]

What are "stetched fonts"? It is unfortunate that 13 isn't divisible by two - but then it is a prime so there wasn't much else I could do... --Fir0002 23:19, 1 October 2007 (UTC)[reply]

The labels on all the bottom two rows look vertically stretched. Debivort 04:18, 2 October 2007 (UTC)[reply]

Oppose, mostly as Rawr - I think that 'Save' and 'Save for web' are simply two different methods that Photoshop offers to select image quality. I'm not aware that there's an encyclopedically-interesting difference between them; so half the information on the image is redundant, unless you can explain one. Otherwise, I don't particularly see the advantage of arranging the data in one 11mb image, as opposed to a series of individual images, per JPEG#Photographs. TSP 19:09, 1 October 2007 (UTC)[reply]

Well the main information, the exponential file size growth of the compression, is still there, you just get the added element of being able to see whether it is worth optimizing for web. I fail to see how that could detract from the value of the chart. --Fir0002 23:19, 1 October 2007 (UTC)[reply]

~~Oppose~~ First the subject is dubious even more for a FP (bad quality jpeg as a FP ?). Second it tooks me too long to download the image. I think most of the shots could go away because a step by step increase of quality doesn t indicate much more than a selection of 3 or 4 shots. I don't want to download a 11 Mo file for that kind of result. Ksempac 19:54, 1 October 2007 (UTC)[reply]

I'm sorry but opposing based on it being an 11mb file is not valid. Please read the evaluation criteria for a FPC. Furthermore to complain about the quality of the jpg's is ridiculous - that's the whole point! --Fir0002 23:19, 1 October 2007 (UTC)[reply]

Not a valid oppose rationale. Please give another one, referencing the featured picture criteria. MER-C 07:16, 2 October 2007 (UTC)[reply]

Oppose. Too much information. The graph shouldn't be in there, and there should be fewer images overall. — BRIAN0918 • 2007-10-01 20:07Z

Possibly their are too many images, but I think the graph is the key element. Without it the pictures would be meaningless. As explained the chart shows both the exponential growth of a jpg file as quality improves, and shows where a "saved for web" optimization procedure is better than a standard compression in terms of file size. The rest of the images allow people to judge whether the file size difference is worth the quality differences. --Fir0002 23:19, 1 October 2007 (UTC)[reply]

Oppose Fairly interesting information but unappealling presentation. Is the SFW optimization algorithm common across photo editors or is it unique to photoshop? It is a useful contribution to the articles it is in, but in my opinion is not a FP. Probably your only submission I haven't loved. :) Meniscus 04:01, 2 October 2007 (UTC)[reply]
Oppose Simply too much info crammed into a space that won't fit on any screen. The graph is the main thing - skip all the small images. Instead, enlarge a small portion of each of the six big images to clearly show the artifacts. Also, mention the uncompressed file size somewhere. Then, I might consider supporting. --Janke | Talk 08:29, 2 October 2007 (UTC)[reply]
Oppose - Per above and all oppose --Childzy ¤ Talk 09:41, 3 October 2007 (UTC)[reply]
Comment. I have a few concerns with the image:
1. The concept of "save for web" is not encyclopaedic. Presumably this is a photoshop-specific concept, and so isn't really useful as a comparison. Is the same true for the compression numbers of 1, 2, 3 etc? Are they comparable with other software? If so, I would much rather a version without the "save to web".
2. The graph doesn't really belong. As soon as you add a graph like that into the image, it becomes less useful, as you're forcing a certain layout. It would be better having a graph in a separate image, then articles can lay them out as they want.
3. Too many images: Without zooming in, I'm not sure I can even see a difference between quality 12 and 100. It would be better, imho, to have just 4 or maybe 5 images, to the make the different quality more apparent.
4. And speaking of zooming in, it would actually be better to show part of each image magnified, so you can really see the artefacting. You could blow up part of the spinifex, where it's most likely to be visible. Stevage 05:20, 4 October 2007 (UTC)[reply]
Oppose — This image is useful only in Adobe-related articles, as the "save for web" feature used here is only available in their products. It's a program-specific function, thus it means nothing in the general definition of JPEG compression. A reader would have absolutely no idea what the difference is between save for web and standard JPEG compression by looking at this picture. They aren't told that it goes through a special filtering algorithm to optimize the image before being stored as JPEG, what options were used (other than compression percentage), nor what the technical difference is (ie. save for web removes EXIF data). I also find it highly amusing that a lossy format was used to display an image about lossy compression (regardless of how minor the loss is by using maximum quality JPEG). ♠ SG →Talk 20:38, 5 October 2007 (UTC)[reply]

OpposeOnly useful for Adobe, to big, taken the above statements into account. Neozoon 22:32, 7 October 2007 (UTC)[reply]

Not promoted MER-C 03:54, 8 October 2007 (UTC)[reply]