Resize methods, which is better

[tommykjensen]

One original photo resized from 3072x2048 to 800 x 533 done two ways.

One with a one step resize in PS CS and the other with the FredMiranda plugin in several steps. I am not telling You which is which.

Please vote which You think is best.

1:
http://photo.klein-jensen.dk/photo.php?n=resizetest1.jpg

2:
http://photo.klein-jensen.dk/photo.php?n=resizetest2.jpg

[ecobo]

I don't see any difference. I've examined the photos very carefuly, but still nothing. Sorry, I can't vote :cry:

[tommykjensen]

Thats my problem too, I simply can't tell the difference. I have added an option for those that can't tell the difference.

[Lagged2Death]

I spent some time peering at these also, but I can't see the difference either. Which may be sort of instructive in and of itself.

I see that #2 is several KB smaller, though, which is a point in its favor.

[Calis]

Nope, I can't see any difference either. I was tempted to download and pixel peep at 400% but thought that if I can't see a diference at 100% "what's the point?"

I certainly won't be mucking around with stepped resizes again - thanks.

Great shot by the way.

[ecobo]

Step-by-step resising worths the time when you interpolate an image to a large size like 50x70 cm (20x28"). In this case the difference is realy noticeable. I use this technique with Photoshop.

[Lagged2Death]

I see I can't change my vote, which I guess is as it should be. But I can amend my answer a little bit.

I think #1 has a tiny, tiny bit more contrast. I noticed this in the highlight in the eyeball, which seemed brighter to me in #1. It can also be seen in the white flecks along the - uh, eyebrow? (Do raptors have eyebrows?) I did try downloading these for viewing at high magnification, (yes, pixel peeping!) and the difference is more pronounced that way, but I wouldn't call it obvious.

All that said, I think they're both better from a technical point of view than any of the pictures I've ever taken, and are for all practical purposes basically identical. The difference isn't that obvious even under completely silly magnification factors; the difference in ordinary viewing - on screen or presumeably in a print - is about nil.

[LightHunter]

What Miranda plug-in did you use?

[tommykjensen]

What Miranda plug-in did you use?



Web Presenter Pro

http://www.fredmiranda.com/shopping/WPpro

[timmyquest]

I also couldnt tell until i looked at the hairs under the neck.

If the one i chose is the one that takes more steps, i wouldnt bother...but i can tell a difference.

[tommykjensen]

Have Scottes been around to vote?

[Scottes]

Have Scottes been around to vote?

Yep, I was one of the first, but kept silent so I could see what people say since I've spoken enough about this.

Alas, now that I'm here I might as well keep yapping...

I think that #1 is better, but very slightly so. That on it's own may not make it justifiable to people who don't have the cash to blow on this little plugin. As I've said pretty much from the beginning, the FM plugin is worth it for convenience if you have the extra cash - *IF* you think it does a better job. In this test and others I think it does.

"But so what - it's so very little of a difference. It's not even noticable."

Yes, I agree that it's barely noticeable. But this is only one step out of several that you may make on an image. If you make 6 changes to an image (resize, sharpen, level adjustment, color adjustment, and so on) and you make each one of those steps a tiny bit better then you've made your entire process much better. Giving each of those 6 steps a measly 3% improvement makes for a 20% improvement overall. Now that's noticable.

[tommykjensen]

Have Scottes been around to vote?

Yep, I was one of the first, but kept silent so I could see what people say since I've spoken enough about this.

Alas, now that I'm here I might as well keep yapping...

I think that #1 is better, but very slightly so. That on it's own may not make it justifiable to people who don't have the cash to blow on this little plugin. As I've said pretty much from the beginning, the FM plugin is worth it for convenience if you have the extra cash - *IF* you think it does a better job. In this test and others I think it does.

Well You and 3 other think #1 is better, I asked the same question at work today 2 was not able to tell the difference 1 said #1 was best, I have asked in a danish forum and so far 3 out of 5 say #1. One of those even said that he think "Your" theory is wrong but still he choose #1 which is the one done with the FM plugin. He sure will be surprised :lol:

"But so what - it's so very little of a difference. It's not even noticable."

Yes, I agree that it's barely noticeable. But this is only one step out of several that you may make on an image. If you make 6 changes to an image (resize, sharpen, level adjustment, color adjustment, and so on) and you make each one of those steps a tiny bit better then you've made your entire process much better. Giving each of those 6 steps a measly 3% improvement makes for a 20% improvement overall. Now that's noticable.

You have a point there.

I have the plugin now, will I use it? Maybe on selected photos.


Thanks for playing along with this little experiment, maybe I will repeat it in a month or two with a new photo :lol: :lol:

[Scottes]

I have the plugin now, will I use it? Maybe on selected photos.

Yep - the ones you really care about get the extra treatment.


Thanks for playing along with this little experiment, maybe I will repeat it in a month or two with a new photo :lol: :lol:

Be mean - make it 3 images. 2 identical, and 1 different. Give a prize to the one who says "#3 looks better, but I can't see the difference between 1 & 2."
:twisted:

[cmM]

Be mean - make it 3 images. 2 identical, and 1 different. Give a prize to the one who says "#3 looks better, but I can't see the difference between 1 & 2."
:twisted:
Yea, and make the B&W too, just to spice things up a little.

I can't tell the difference at 100%

[Scottes]

I can't tell the difference at 100%

I've been a professional pixel-peeper for about 15 years now... I peep so much that I've got calluses on my eyeballs... :)

If I had any clue about composition or color theory I'd be dangerous.

[slejhamer]

Scottes, or TJ, which resampling method was used for the "one step resize in PS CS?"

Supposedly the new "bicubic sharper" does a better job for downsizing than the default plain old bicubic, and the result might be more comparable to that which was achieved by Miranda's method.

[Scottes]

Scottes, or TJ, which resampling method was used for the "one step resize in PS CS?"

Supposedly "bicubic sharper" does a better job for downsizing than the default plain old bicubic, and the result might be more comparable to that which was achieved by Miranda's method.

Bicubic sharper does looks better - if you want to lose control of the sharpening process. I'd rather have as much control of sharpening as possible.

[msol]

Nice experiment! After looking at the two pics (one above the other on screen) I couldn't really decide (although #2 seemed a bit sharper).

So I copied both images into PS, put them in two layers (#2 on top of #1) and then chose blending mode difference on #2. That clearly shows the differences, and these are primarily on the various edges, e.g. the top of the beak. So I zoomed in on that detail (put blending mode back to normal) and repeatedly toggled #2 on and off. That made the difference obvious: #1 looks as if it has been sharpened with too high radius, there is a dark glow around the bird, increasing the darkness of the background, and so creating more contrast with the beak. To me this means that #1 has a "fake" kind of sharpness, so I prefer #2.

I don't know what causes the difference, but this is what I found.

[LightHunter]

I do not dare to say I see adifference but did you try to print both at maximum size?

However I would like to challenge a statement on the FredMiranda web-page for the explanation of the plug-in. Decreasing the image size by 50 % (in this case 1648x2464 pixels) which results according to FredMiramda in an image of 824x1232 is as far as I can still calculate wrong since it is actually only 25% of the image size and not 50%. No?

[vjack]

I voted for #1. Since I am considering buying the FM Web Presenter Pro, I'm eager to learn which was which. If it turns out that the FM plug was used on the second shot instead of the first, I'm not buying it.

[tommykjensen]

I voted for #1. Since I am considering buying the FM Web Presenter Pro, I'm eager to learn which was which. If it turns out that the FM plug was used on the second shot instead of the first, I'm not buying it.


Wow, what page did You find this oldtimer on ;)

I actually can't remember which was done with the FM plugin but I did post it in this thread

choose #1 which is the one done with the FM plugin.

[vjack]

Wow, what page did You find this oldtimer on ;)


I didn't even realize how old it was when I posted. I've been reading about some of the FM plug-ins and was curious about the Web Presenter. I had no idea it was as old as it is. Anyway, this thread was helpful becuase it gave me an example of how it compares to the usual PS method.

[tommykjensen]

I didn't even realize how old it was when I posted. I've been reading about some of the FM plug-ins and was curious about the Web Presenter. I had no idea it was as old as it is. Anyway, this thread was helpful becuase it gave me an example of how it compares to the usual PS method.

I am glad it was helpfull.

I originally said I would use the web presenter occationally to resize but I have found myself using it 99% of the time.

[Robert_Lay]

I vote for #1 as the better image, because:
1) the top of his head seems to have more clarity and contrast.
2) the light bluish portion of his upper beak has more sharpness and contrast.
3) the bright feathers under his eye look less hazy.
4) the highlight on the most upper part of his beak is not as blown out as it is in #2.

[re1ex]

i voted for #2 because the hair under the chin is sharper in the 2nd pic

i saved the images and looked at both of them going back and forth, i saw no difference besides the chin hair.

[vjack]

I am glad it was helpfull.

I originally said I would use the web presenter occationally to resize but I have found myself using it 99% of the time.

I'm somewhat hesitant to buy it because it doesn't look like it has been revised in a few years. I can't help but wonder whether something better has appeared in that time frame or even whether CS2 now does a better job.

[tommykjensen]

I'm somewhat hesitant to buy it because it doesn't look like it has been revised in a few years. I can't help but wonder whether something better has appeared in that time frame or even whether CS2 now does a better job.

Hmm. Maybe an idea to redo this test.... I will see when I can get the time to resize the same photo with CS2.

[Michaelmjc]

I simply dont see any difference at all.

[Radtech1]

I think they were resized equally. They both look exactly the same size to me! :eek:

Rad