Hi all, I have just been accepted at SS (thanks to the people who offer advice on the critique forum, I've learnt a lot from the advice offered there over the last few months).

I'm guessing the 7/10 that were accepted are already on sale, although they don't show up in my portfolia or keyword searches yet, the site is reporting that I have one d/l already ?

I downsized all my pics for teh application to just over 4mp, is this the best way to go from now on or is there an advantage in trying to get larger image sizes accepted?

TIA

Sponner.

On a few occasions, resizing can be useful to reduce noise from very high iso.

Otherwise, I don't downsize for SS: I suspect I could miss some sales by limiting available resolution.

fyi, I can see your 7 images...if you still don't see them, go to your portfolio and change the view to "newest first" and you should see them.

cheers, I can see them now just ook a few days to get indexed I suppose.

I used to and like just about everyone else, it wasn't worth the time. I know some people think that selling for a lower price, they should sell smaller images. Maybe the buyers look at the sizes, so it's counter productive? I don't know, but I know I'm lazy, so I don't bother anymore.

Last place I used to downsize before upload was StockXpert and they are gone.

Answer: Get Irfanview, it's free. There's a batch conversion tool that will resize and also do other things if you want, but just reducing the size is easy. You set the longest side dimension and click... bing, it's done. I'd say use the automatic rename function at the same time, also included, so you know they are the reduced files. Something like <filename>-SS.jpg for example and you know it's a SS file. Easy.

http://www.irfanview.com/

Get the software and get the plugins, it's small, and It's Free! It works.

Presumed you are using Photoshop: How exactly do you downsize images?

(Sorry if I translated some words wrong here, but I am not using an English version of Photoshop. Hope you understand what I mean.)
I use: Image, then Imagesize and in Documentsize I choose Percent.
Then 90 or 95 percent and Bicubic sharper for reduction.

For I am not very impressed by the differences, I am asking myself if there are better ways to downsize.
(Google is not my best friend here� )
Perhaps someone here can help me out with a good suggestion?

I am one of the people who recommended downsizing the images for the initial application. That's how I was accepted originally after the 3rd try - several years ago now. The reviewers are more picky on the initial application. I agree that you shouldn't have to downsize in most cases. Congratulations on getting accepted.

After accepted, I submit the images without downsizing them, and the cases I get a refusal caused by noise or sharpness, I resubmit that ones downsized. This way I get accepted about 70% of refused files.

I downsize to the minimum accepted, so I put directly the number of pixels of the largest side. 2400*1600 is the minimum, so only if putting 2400 the other side has less than 1600 I put a higher value. You can induce if the total is 4Mb multiplying (approx, because the system is not decimal, but binaric, and 1MB is not 1000*1000 but 1024*1024).

If you use Photoshop I guess you have also Bridge. There, Tools/Photoshop/Image processor (or similar, I neither use English version) lets you to resize a batch (and apply actions if you want).

Thanks very much for the replies, but downsizing on itself is not my problem. I know several ways to do so. Sorry if I have been misunderstood by being not clear enough.
And I have Irfan view and Faststone viewer installed on my computer already.
(Faststone is great for a quick first selection of Raw files)

My problem is that I find the quality of the downsized image nearly as �worse� as the original (when viewed at 100 percent).

Scott Kelby gives a solution in his book too with the motto: Everything Looks Good When It Is Small.
But the price you pay: you are throwing lots of information from an image away by downsizing this way and I was asking myself if there is perhaps a better way to do it without losing so much info.

Quote from: fxegs on February 23, 2011, 14:56

I downsize to the minimum accepted, so I put directly the number of pixels of the largest side. 2400*1600 is the minimum, so only if putting 2400 the other side has less than 1600 I put a higher value. You can induce if the total is 4Mb multiplying (approx, because the system is not decimal, but binaric, and 1MB is not 1000*1000 but 1024*1024).

The bits and bytes doesn't have anything to do with megpixels.

Yes, you're right, I know, sorry about writing fast without taking care.

Yes, it's going to look worse if you view it Full Screen, which is not the same as Actual Size, because now you are enlarging it. You will get a worse looking image. I suspect you aren't viewing them both at 100%?

If viewed at 100% the original will be larger than the new file, but the new smaller file should look just as good or possibly better, because you have smaller pixels. That is unless someone changed the compression when downsizing or something else is modified?

By the way, even though it was off the question you asked, the size for 4MP image that I use is 2460 x 1640 which comes out to... 4,034,400  Some sites list 4MP as 2464x1632 = 4,021,248

If your question is how do you make something smaller and lose no data, that's like saying pour a 1 liter bottle of water into a 750ML jar and still have 1 liter? The fact that you downsize equals less data, you can't have it both ways.

+1
You're loosing information, in fact you're cutting out a great lot of pixels. But for the same reason you loose noise, and unsharpness sensation. Anyway, you should do this only with previously refused files. Do you prefer a minor information file (and for the web and small prints works fine) or nothing?
I know that you only are trying to preserve the maximum quality and information, another way is not to downsize and try to reduze the noise or sharpen the image, but this doesn't work sometimes. But... try, refusals in shutterstock don't matter, you can reupload the files several times.

When downsizing you do this kind of thing �virtual�. A smaller image looks sharper. (And you are loosing info.)
But reviewers look at images at 100 percent or more and then the same problems pop up. Perhaps they let them pass, not for the image is really better, but the usability for buyers is (seems) better?

Even if reviewers look them at 100% they don't see the same after downsizing. Test it yourself. They see a bigger percentage of the image, and things in the pic, smaller. The same problems don't  pop up unleast the image were excessively unsharpness or or had too much noise and it still is visible.

You can have more pixels in an inch (dpi), but how can you have smaller pixels?

Of course you haven't smaller pixels, and neither you have more pixels in an inch. You have the same pixels in an inch and they (every one) sizes the same. When you downsize the image, you loose pixels, you delete them. If your image is 12Mp in the original and you downsize it to 4Mp, you loose 8Mp! 66% of the information. If you view the image 100% before and after, the second time you view all 66% smaller (at 100%, yes, for seeing the same size than before, you have to view it 300% this time, in this case) and you view in your screen 66% more of the image than before.

There is no way to preserve information, you are deleting it! You have to choose. Anyway, for downsizing, Photoshop recommends to use the Bicubic sharper option in the case 'Resample Image'.

But you should downsize only in a few cases. The result is not so bad: 2464*1632 as RacePhoto suggest, is an image at 300ppi resulting in 8.213*5.44 inches. 300ppi is for industrial prints. For a best quality printing in a home printer you have 220ppi , 11.2*7.418 inches, or a good / medium quality printing 150ppi , 16.427*10.88 inches. For the web, which uses 72ppi, you have 34.222*22.667 inches!. I think web usage is the more common one (I guess).

Only in a very very few cases one client will need an extra-large file for a perfect print at a great size. For all the others, a 4Mp file works. And a downsized 12Mp to 4Mp file is not worse than an original 4Mp image.

OK, sorry for my mistakes. English is not an easy language. (Perhaps no language is...)
But I learn better every day....

So the conclusion is: there is no better way to downsize images.
Perhaps I have to downsize more to see better results. 10 percent is propably not enough to see much difference.

Others may be of a different opinion, but the last step I do is the downsizing from the TIF but save the full size TIF for future use. I'd hate to work on the smaller image, when the big one is easier to see, and I'd hate to save it small and find out, I wanted to fix something later. So I do everything full size, save the TIF, then reduce the size and save as a jpg.

+1

I would not downsize.  Remember that they have on-demand as well as subscription.  On demand is size related.  I make an extra 50-60 a month with OD's.  And if a designer purchases a sub plan and is looking for high res content, you're hosed.  And finally......Futurama Bender thinking here.....IF Shutterstock changes their model for whatever reason to be more like IS or Fotolia and less of SUB site, OR if they get purchased by another site that uses a different model that sells imagery by scalable factors (i.e. files sizes) you would be extremely limited in your RPD.  If you have all high resolution images, they would theoretically be transferred to that new site and you're good to go.

Just some food for thought.

James Bond.

Mantis, we're not talking about downsize as a rule, but in very specific circumstances, in which the alternative is downsize or nothing. The post talks about how to downsize in those cases and if it is possible to save quality or information in the process.