Quote from: Uncle Pete on December 01, 2022, 19:11
What if the training use is a one time payment or maybe we get nothing at all, for the training. But mostly who says we would be paid for every use of every image that was used to train the AI. Why?

And that's assuming that we get any image selected to be part of the dataset. 360 million images, no one will need to license all of them.

It's one use, from a giant, low cost, subscription, to train the AI. What if they paid us the dime standard, which is supposed to be the minimum. That's it. One dime, one use.

"But mostly who says we would be paid for every use of every image that was used to train the AI. Why?"

Shutterstock says that, citing: "Contributors whose content was used to train either model will be compensated for the role their IP played in the development of the original models, as well as through royalty payments tied to future generative licensing activity." or "Earnings from datasets and downloads of AI-generated content produced with integrated technology on our platform are pooled in a collective fund and will be distributed every 6 months." Why? It is obvious. All the other stock agencies are forbidding AI-generated images because they cannot say the training images were properly sourced and properly compensate them. SS can do that as it know the source and authors of the images. It provides them itself.

There are two different situations here. One is training the AI for other projects. This will be probably one time payment. The second one is AI generation based on our images. And this is "generative licensing activity" which should lead to royalty payments. I understand that we are going to be paid for every generaion of AI image where the AI was trained by our image.

"360 million images, no one will need to license all of them."

The quality of AI is extremely dependent on the size of training dataset. Extremely. It is not only about the quality of the algorithm but also about the quality (size) of dataset. How would you make better AI. You will put more data to that (in our case, more images). Of course, not all of them have to be chosen, but this is how the AI works. More data = usually better AI. 

Quote from: k_t_g on May 07, 2021, 04:58
Ugh! why do you guys have to "fix" things if it ain't broke?

I really cannot agree with you. Using FTP means that something is broken very badly. This is something which really needs to be fixed.

For me and my girlfriend the SFTP is also not working. Not able to login at all.

For me, the sales are not so bad (usually relatively lot of sales but at the minimum price - this December is an exception with more high-priced sales, approx. 30 % of SS earnings). But the problem with reviews is real.

I have got an answer from BlackBox and you have been right.

"You can send images to agencies, that are working us, but not videos when those videos are on Blackbox.
If you have trouble keeping Blackbox apart from other agencies then Blackbox might not be the right solution for you.

The best way is to send all video to Blackbox which distributes to multiple storefronts while you are free to upload your images to any storefront you like without worrying about duplication.

Also, remember that we are expanding our partnership with other big agencies as well."

So it seems that everything is fine if there are different media types even though they are taken from the same place with the same composition.

There is also one question. What to do with images and videos from the same location? Is it possible to send a video from one location to BlackBox and to send photographs from the same location to others agencies manually (because BlackBox does not support images)? In such a case, there can be video on the BlackBox's account (for example on Shutterstock) and exactly the same composition on the photograph (different media) from contributor's account. That would be weird. Is it againt some rules? How to solve that?

In my case, the number of sales is relatively fine at 123RF but the total earnings are horrible as almost all the sales are at $0,22. And one cannot improve the level as they are not reviewing the files :-(

The same here. It seems that they review images randomly in my case. Videos have not been reviewed since February, it is very similar with editorials. Normal images have not been reviews sice August as well. The only thing that is sometimes reviewed, are vectors. Although, it is weird here as well. Sometimes, they are accepted even before I finish the submission process (images seem to be processed in the step 3 or 4 in the workflow but when the website redirect itself to the step 5, they are already accepted). If this situation does not happen, they are never going to be reviewed.

Since the beginning in 2007 (photo+illustration+video)

SS: 50 %
FT: 19 %
DT: 16 %
123RF: 9 %
BigStock: 6 %
Pond5: 0.3 %

Quote from: Pauws99 on April 28, 2018, 10:23

Quote from: BalkanskiMacak on April 24, 2018, 14:43
Terrible month on 123RF as well, plus the fact that the review of editorial pictures has stopped almost 3 months ago... I'm starting to wonder if there's still somebody behind the website, or if rats are deserting the ship...

I didn't realise my editorials are sitting in pending ;-). Don't suppose its lost me many $$. Gone from a reasonable middle tier earner to also ran for me. Though thats pretty true of all "middle tier" really.

Videos are even worse, they have not been reviewed for almost half of the year now.

For me, DT was dead for a long time. This month is surprisingly good so far.

I can upload there but the files are not reviewed. I have video clips since April 2016 there and still have not been reviewed (stuck in "Pending" stage).

Quote from: Shelma1 on February 24, 2015, 19:38
I have no idea how many times the image will be resold, and neither do you, Milan. That's the problem in a nutshell. Sometimes relatively old images of mine catch on later for some reason and do sell hundreds or thousands of times. And though four out of eleven of your chosen images hadn't sold before, you had no choice about which images Google would choose. You didn't even know it was Google when you were required to decide whether to opt out or in.

As for my images being "so beautiful, universal and suitable for so much different advertising topics that 10000 advertisers would use this each image in 10000 different advertising campaigns," I do have images that have sold thousands of times. I got paid for each sale.

Am I "naive?" I've been on the "buyer" side of advertising for more than 30 years. I've seen ad agencies pay tens of thousands (or hundreds of thousands) of dollars for one image, many many times. I see art buyers pay hundreds for images from SS and Getty on a daily basis. That's their sole job...to buy art. They make six-figure salaries buying art, BTW (though I'm not sure how much longer that job will exist).

You got $22 for a full year of usage—who knows how many times—for 11 images. That's less than one ED sale for one image on Shutterstock.

I agree, we have not idea. Therefore I do not think that this deal is exteremely bad for sure as well as extremely good. We do not know. Yes, some images can be used many and many times. But the problem is somewhere else. It will not certainly apply for all of them. Some of these images will be surely very successful but the others will not. So, in the average, it does not need to be so extremely bad. And do not forget, if the image will be successful af the one year period, the EL licence will be paid (and one already has $2 for non-successful images). If the sub sale on DT is $0,35 and I got $22, it is 63 sub sales. Until now, these my 11 images had 6 sales from all posible companies and individuals an all DT channels. And now, there is a question. Can they manage to sell them 63 times in one year on this Google/DT channel only if they were sold only 6 times until now on all DT channels (all DT customers)? Let's admit that some of these Google/DT customers could buy the image somewhere else if there is not such G/DT deal. If only 6 customers bought my images on DT until now, how many of these 6 sales used that image in Google advertisment? Maybe 1? I do not think that suddenly these images would be used 63 times more often in Google ads than now. But maybe... they will. I do not know.

I agree that paying for each use is better, but on the other hand, I do not think that this deal is such an evil one. I only oppose the extreme negative opinions on this deal. But time will show. Maybe, it will be really disaster.

Quote from: Shelma1 on February 24, 2015, 12:29
And who are you, exactly, person who hides behind anonymity yet feels free to dis other people's work?

I really have not say that your work is not good! Have I? I only asked, if you really think that your image is so beautiful, universal and suitable for so much different advertising topics that 10000 advertisers would use this each image in 10000 different advertising campaigns. Do you really think that Google has so many advertisers in one suitable topic that each of them would use exactly this one your image? If there is such demand for this image, would not you think that you would already sell this image several thousand times? If I count my 11 picked images, they were sold 6 times (8×0, 1×1, 1×2 and 1×3). Do you really think that now, as a miracle, will they be used 110000 times instead of 6? Each one of them 10000 times? 10000 times, if they had only 0.5 sales on average until now? Do be so naive.

Maybe it is not great deal, we do not know. On the other hand, almost everybody, who is opted-in, says, that mostly non-selling images were picked. So, maybe my image will be used 5 times or more and it will be not good for me, maybe they will never use it and I will get $2 for nothing (and the image would not be deleted after 4 years of no-sale). Maybe it will continue as EL (after year) and more money will come. I really do not know the numbers and how many times will be used these image on average, but I am certainly sure that it will not be 10000 times per image on average.

In 2014, FT makes 18% of the money that SS made. In 2/15, it is 67 %. So, FT is improving for me. 

Quote from: Shelma1 on February 24, 2015, 11:29To each his own, I suppose. I don't mind the SS/Facebook deal, where the images are also tiny but we're paid for each use. I have a problem with gigantic Google paying me $2 and then making my image available 10,000 times. Clearly, it's possible to make deals where artists are compensated for each use, because SS did that. I also see that SS and Getty are selling our work--the same work--for hundreds of dollars for each use while DT sells it for two bucks for thousands of uses. SS is looking in the right direction, IMO. Their prices are going up and they're aiming for larger sales. I've decided I'd rather look in that direction too.

Do you really think that your images are so super-brilliant that they will be used 10000 times during this one year (or less)? For your information, 11 of my 813 images were selected in this deal and 8 of them were never sold in DT (and 4 of them were also never sold in any other agency). As it was already said, there is probability that many of these selected images will be never used (and replaced) and some may by used many times and in the average, it could be not so bad deal. Who knows.

593 online images (on average) earns me $104 monthly. So... it is about 570 images in my case. Not very good :-(

The same here for a long time. I have AR 60-85% on other agencies, but less than 5% at FT. They reject EVERYTHING.

I think Sigma is better. Although it is no perfect lens, it has not any significant disadvantage (a little bit of CA, but sharpness and contrast is good, it has almost everything slightly above average and it is very usable even wide open). Canon is a mixed bag. It has some better characteristics like better contrast with low F numbers but with the most wide open aperture, Sigma is slighlty more user friendly (Canon has small focus shift). Also CA is more visible. I will buy Sigma (it is more fool-proof :-)) but it does not mean that Canon is not good.

I have tried also Canon 50mm F1.2L and I think it is one of the worst "L" glasses. Of course, it hase very nice apeture, gorgeous bokeh and very nice contrast but it also suffer form enormous CA and the focus shift is unacceptable for me.