MicrostockGroup
Agency Based Discussion => Alamy.com => Topic started by: ovbg on March 31, 2007, 09:03
-
Hello, I have PS CS1, and was wondering about the best upsizing method for Alamy. Would I be correct in using the standard menu option: "Image/Image Size", constrain proportons and Bicubic, and set the size to say 110%?
Thnaks.
-
Yes, 10% at a time, and make sure you use "Bicubic Smoother" in the drop down menus at the bottom.
I recommend Genuine Fractals, but making an action to increase the size 10% at a time, using "Bicubic Smoother"
might be just the ticket as long as you don't upsize it too much ...maybe 2 times or 3.
-
Actually, you don't need to resize in steps. Since resizing involves interpolation, and interpolation involves error, you're compounding the interpolation error by adding in multiple steps. That is, rather than adding a lot of error in once, you're adding a bit of error at each step, which then becomes compounded with the error introduced in the next step. In the end, you find you've increased the error more by using a stepped algorhythm. It's not typically a big difference, but in significant uprezzing, it might be visible.
Nonetheless, there likely isn't much of a difference between the two techniques in most situations. So, doing the uprez in one step will save time.
I use Qimage for increasing file size. It works very well, but others like Bob use Genuine Fractals to great advantage.
One last thing--it's vital that you not uprez a sharpened image. It will look absolutely terrible.
Just my 2 cents worth--for what it's worth.
Yes, 10% at a time, and make sure you use "Bicubic Smoother" in the drop down menus at the bottom.
I recommend Genuine Fractals, but making an action to increase the size 10% at a time, using "Bicubic Smoother"
might be just the ticket as long as you don't upsize it too much ...maybe 2 times or 3.
-
oops--hit the wrong stupid button.
-
On my few experiences with 50-100% upsizing, I asked a friend who has Genuine Fractals to use it so I could compared with other cheaper software. I was highly disappointed with it. Maybe there are special adjustments that she doesn't master, but, comparing with other software, GF didn't bring as much detail.
In these experiments, originals were only 640x480 or 1280x960 pix images, so maybe results are much better with larger images.
If I need to make small upsizing (10-15%), I use PSP's "Smart size" (I don't use PS) with no visible loss of quality.
Regards,
Adelaide
-
I agree with the Professor ;D
I've upsized using bicubic interpolation and I've upsized using ACR - both are acceptable for Alamy. I also have Genuine Fractals. You do get better detail using Genuine Fractals and I use it more frequently than any other method. There is another program I have heard of people using called 'Blow Up' that is supposed to get better results than Genuine Fractals but I have not tried it. I've also heard the Fred Miranda plug-in is on par with Genuine Fractals but have not used it either.
If you are using bicubic interpolation for Alamy - you want to (roughly) have a long edge of 5025 - that's usually my starting point.
Good Luck!
-
Thanks for your help guys.
:)
-
The following images are excerpts from a 640x480 image upsized in 100% with GF, Blow Up and a $20 software - not shown in this order. I'll let you enjoy them and make your own judgements (and preferences regarding the resulting texture may vary) before unveiling which one is which.
The original image is a JPEG from my old Finepix 2650 camera, saved to TIFF and then resized. The montages were saved in minimum compression.
Regards,
Adelaide
PS: I can't upload both in the same message due to size limitations.
-
The other comparison file.
-
oops--hit the wrong stupid button.
At least you can find the "stupid button". I cant find any on my keyboard.
I dont do Alamy yet but know of people you just use the bicubic in one step and have no problem with acceptances.
-
Hmmm. Nobody interested in the software comparison?
Regards,
Adelaide
-
Hmmm. Nobody interested in the software comparison?
Regards,
Adelaide
I couldnt see any major difference. they are different but quality seems ok in all. i am not a pixel peeper though.