If I have an image that's 1500x1500 pixels and then I resize it to 750x750 but after this do it yet again to 600x600 would that be the same as resizing the 1500x1500 image to 600x600? Does it matter whether I resize the larger one if all I do is reduce the current images size? Would taking the 1500x1500 and resizing it to 600x600 be better than resizing the 750x750 to 600x600?

Never tried it but it would seem that if you went from 1500 to 600 it should be the same as 1500 to 750 to 600. The end result shouldn't change I'd think, but again I've not tried it .


Ever hear of this?




http://www.dogpile.com/info.dogpl/clickit/search?r_aid=223A2D28FEAC4CECA8BBDED0DFD1F030&r_eop=3&r_sacop=2&r_spf=0&r_cop=main-title&r_snpp=1&r_spp=1&r_wsm=0&qqn=pN-.iVDj&r_coid=239138&rawto=http://www.imaging-resource.com/SOFT/GF/GF.HTM

Resizing 1500x1500 to 600x600 will produce much better quality because each time an image is resized, some pixels will be discarded and depending on the change of resolution that occurs, those pixels could be significant.

It's totally dependent on the Mipmapping algorithm used by the program in question.

The two images if done with a proper algorithm should be exactly the same since the algorithm should be able to split it's overall sum down into factors. The problem is however that for it to know what factors should factor to it would need 'Benchmarks' where set sizes have the specific factor predefined.

The likelihood is that most programs have predefined sizes you can shift too. What you could try is comparing the same image after it's been done by both.

[Edit]
I had to hunt around the internet for a picture big enough to do this test with since my images and even clipart CD's didn't size up to 1500x1500

I eventually found: http://mikal.org/photo_journal/baytobreakers_2004/ to which I've used a cut version of p1010005.jpg. (I haven't mentioned to the site owner/blogger about using his image, I shall do so after posting although I might be asked to take the image down, but here's the comparison for now)

Since the initial image is 1500 x 1500, I'm going to put it as a link Here! (http://sciforums.stryderunknown.co.uk/images/mipmaptest1.jpg).

But these are the results:

The straight forwards resize from 1500 x 1500 to 600 x 600 weighing in at 380kb:
Http://sciforums.stryderunknown.co.uk/images/600x600_Straight.jpg

The 'Detoured' version, resizing from 1500 x 1500, to 750 x 750 and then to 600 x 600, weighing in at 381kb:
Http://sciforums.stryderunknown.co.uk/images/600x600_detour.jpg

The program used was Photoshop CS2 to trial this.

From the looks of it the 1kb suggests that obviously moving to 750 x 750 (detour) doesn't 'Round' as much as the 600 x 600 (Straight), so you could suggest a more accurate image based on size alone is the detour rather than one resize. However it's up to you to judge the actual images to decide if that 1kb really stands up to anything. Feel free to do a Byte by Byte comparison, the images look pretty much the same to me.

If you dont save the image inbetween it doesnt matter.

If you dont save the image inbetween it doesnt matter.

The following is one where the image is saved in between resizes.
weighing in at 386kb.

Http://sciforums.stryderunknown.co.uk/images/600x600_cutdetour.jpg

Do note I'm using the JPG format at Full Quality (lvl 12), Other formats will obviously give different sizes.

I doubt we need to do any more examples for it, just remember that although there is a spot the difference going on, the changes would of course be extremely subtle.

Ofcourse, the changes are subtle. But if you do this a lot of times you will start to see the image deteriorate. I figured darksidZz meant that.. otherwise i dont really see what the point of the question is. :shrug:

Saving the image in between only take affect when you use the saved version to make the next resize with. Depending on the format you saved the image in there is going to be a huge difference in quality.
As long as you use Photoshops own PSD filetype you won't have any loss in quality, thats why most artists and photographers alway save a version of the original.

The fact that the detoured version is 1kb larger shows that the images are not identical. Probably hard to see, but the file size of the jpeg is largely made up of the number of different colors used inside the image and the level of contrast.
For this 1kb to come in, the difference may be very small and even irrelevant, but it become a larger factor on larger images for huge display or on very delecate artwork with some fine details you don't want them to be messed up after resizing.

Bottomline, I don't know whether it's better to rezise from 1500 straith to 600 or with a detour. Would be nice to see a real in depth benchmark on this one.

Cheers.