MicrostockGroup Sponsors
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - cascoly
Pages: [1] 2 3 4 5 6 ... 170
1
« on: July 04, 2025, 14:50 »
They should be ordered to pay an extended license to photographers who have intellectual property for each image ingested by deep learning.
Better still... they should pay a PERPETUAL recurring revenue license. Because after all - this is not a 'one-use' case - they plan on RESELLING YOUR ASSETS via their tools FOREVER... So - for every generation based in any part on YOUR asset - you should get a micropayment for EACH & EVERY generation.
you continue to ignore that there are 2 distinct questions here - the first concerns use of scraped images to train the AI and this is the question addressed in these lawsuits. the second is the actual generation of images from the resulting dataset, which does not store any actual images. you continue to show your ignorance of how AI works - each image generated looks at millions+ of possibilities so any payment based on usage would be vanish ing small. but more importantly, there is no way to track who created the original images used to train the dataset.
2
« on: June 27, 2025, 19:57 »
Congratulations all, don't spend it all in one spot. You must feel great taking a 20% pay cut from last year. Next year maybe is will also be cut by 20%. I really don't understand the logic of selling yourself short. I'd have more dignity collecting empties for the refund deposit.
in a declining/ transforming industry $X income will always be better than zero income. we're dancing with elephants in a commodity-based industry. take what's available or continue to whinge about how much your beauties should be earning! my income has been steady for the past few years, mostly due to the new possibilities for earnings that others decline. fine by me!
3
« on: June 27, 2025, 19:47 »
The only way I can give away my images is through my website, to do marketing FOR ME, not for agencies.
So, this year too: opt-out! Not a single image of mine in the Adobe Stock Free collection, if you want them, you pay! 
the point is NO ONE WANTS these images! i gladly take $4 of passive income for older images that havent sold.
and nothing prevents you from private sales - likely a very different audience
Many photos that I don't sell on Adobe bring me tens of $ on Shutterstock or Istock.
y so why not get $ from AS which otherwise would get nothing? it's unlikely to affect your sales elsewhere since they're buying elsewhere but one's buying on AS. how many users shop around when man agency offers images for pennies? these variousm dataset payments have kept my total income where it was 3+ YEARS AGO.
4
« on: June 25, 2025, 17:10 »
The only way I can give away my images is through my website, to do marketing FOR ME, not for agencies.
So, this year too: opt-out! Not a single image of mine in the Adobe Stock Free collection, if you want them, you pay! 
the point is NO ONE WANTS these images! i gladly take $4 of passive income for older images that havent sold. and nothing prevents you from private sales - likely a very different audience
5
« on: June 25, 2025, 17:07 »
I got nothing. Never happened before. I hope you got what you wanted.
i got over $300 for images that had almost no sales - those who opt out because they think their non-sellers will suddenly make more than $4 are, at best, unrealistic
6
« on: June 21, 2025, 15:19 »
yes, canva was originally 3rd tier but they really expanded their internet presence - unfortunately for artists, their 'review' process automatically rejects almost all uploads INSTANTLY but won't respond to emails
however, their monthly share distribution often outpaces AS for me, and i have only about 20% on Canva vs AS
7
« on: June 02, 2025, 12:47 »
The Internet is not the place to disclose your personal data. A lot is known about me, I shoot videos and a few photos. This information is enough for questions and answers. I did not suggest that you publish the address of your YouTube channel here (although usually the owners of these channels like to do this). You were asked a simple question, to indicate approximate figures. Especially since you yourself raised this question.
LOL - from anyone else this would be a reasonable request (and of course others are free to decline), but your response is despicable hypocrisy - you constantly demand stats while refusing t even show your portofolio and earlier you told us ... The forum was created to share information.
guess it's a one way street
8
« on: June 02, 2025, 12:38 »
...
9
« on: May 29, 2025, 18:39 »
+about $300
10
« on: May 20, 2025, 13:54 »
i got mine too, i dont know if i'm happy or sad with IT, 80% less of what i was recevied previous year!!! what happend with canva???
same decrease here - i stopped bothering to submit when they began rejecting almost everything (accepted at other agencies) and gave no actual reasons
11
« on: April 14, 2025, 13:41 »
For anyone who has had files removed and can't work out why, please can you check to see if they were duplicates? Do you still have the same files in your portfolio, with file numbers that are different to the ones that have been moved to the Not Accepted tab?
i've had about 300 removed - it's not worth my time to track them down - a much bigger proble is AS doing bulk rejections of batch es accepted elsewhere
12
« on: April 13, 2025, 14:40 »
Cheers i might give them a look. My workflow is similar to yours with a master XLSM of keywords as well i guess.
I'm having some luck creating Gems from Gemini (because its free) and trained it with the AS and other docs and some other criteria. Its single image currently but i could Python batch it if needed.
looks like you still need to transfer results to images manually? for 100 images that eats time. my captions xlsx contains more than keywords -- for each image, i store description and keywords. key to my system is renaming all images to YYMMDD-{seq}, so can quickly find any image by date. details are at https://cascoly-images.com/building-a-microstock-tracking-system-part-1VM edits the meta and provides a csv of all images
13
« on: April 12, 2025, 15:39 »
I found one that was a lake removed for "incompatible with terms" - not sure what terms it might be incompatible with. I might have a typo for "Sierra" that is "Seirra" - maybe that is a bad word in some language? But a search for "seirra" reveals 17 results for seirra in images but over 3000 in premium. It is still visible if I search for it on Adobe but not under seirra. ...
not that it's an excuse, but it's possible it was rejected since 'sierra' is also trademark for a car - for awhile human reviewers at SS rejected images taken in Newport, OR (a US cigarette) --- as idiotic as thump's cohort removing pix of the 'enola GAY' ( dropped a-bomb on hiroshima) and recently i keep getting some (designated) AI images rejected for "needs a model release" - it may be due to a bug in submittal that doesn't always show a secondary clickbox when 'ai gen' is selected. it says "images are of fictional people"
14
« on: April 12, 2025, 15:14 »
over last month or so i've had < 1% images removed and added > 1000 (w >500 in 'review' )
15
« on: April 12, 2025, 15:08 »
i have thousands of images, many years old, waiting to be meta'd
i've been using visaulmind.ai which does a great job for about 2-3 cents / image. (you can try it for free on 30(?) images) you can append keywords or add to descriptions before submitting (eg adding a byline for editorial, or location when it isn't obvious. but VM has identified the rhine river by a castle in the background or accurately named the interiors of Istanbul mosques. and Even translates from some greek or latin inscriptions on mosaics. it correctly identifies plants & animals. for simitars they actually write different descriptionsi. in a recent example of a poster from last week's protest "only you can prevent fascist liars ", it understood it as a reference to smokey bear!
i can set up 200-300 images in a few minutes & it runs in background - i then select for various post-processing (several thousand) & resizing. i've added >2500 images to SS, >1000 to AS over last 3 months and i can also update my excel captions sheets by appending the resulting csv (these sheets allow me to locate any of 50K archived images by keyword in at most a few minutes)
for free services like chatgpt, you need to consider the time it takes to prompt or upload & you'd still have to add the result to your image meta.i'm not aware of any free services that process in bulk & update the image meta & provide a csv
16
« on: April 07, 2025, 12:49 »
....
17
« on: April 07, 2025, 12:46 »
...
Later, when the large customer base becomes dependent on it, the quality and service/support are reduced and prices are increased. Until eventually a collapse occurs (creative destruction) and the game starts all over again. Currently, we are in the cycle of crap and are drifting towards the big collapse.
no, that's just destruction & decay & may not see a replacement creative destruction is a positive process in which new innovations replace and make obsolete older innovations. -- word processing, transimstors, digital cameras, ...
18
« on: April 06, 2025, 13:19 »
much worse is a bug that rejects groups images saying they need adjusting
but when submitted one by one just get a msg that incorrect keywords ignored but without showing the offenders & images are always accepted
in the last non-buggy version, a msg would appear BEFORE submitting and SHOW the flagged keywords, so one fix allowed the entire group to be submitted successful
19
« on: April 01, 2025, 19:08 »
Olympic, olympics are trademarked by the IOC so you should remove them.
Apart from that, maybe you could add in amphitheatre seeing as you have both UK and American spellings of theatre/theater.
I've had pix taken in Olympic national Park (a public domain name) rejected for this reason (& our state capital is Olympia!) a trademark only covers use that infringes that particular use (eg 2 different uses of trademarked 'Apple' are accepted), but IOC managed to trademark a word that was millennia old and bludgeons anyone trying to use the word in another sense. otoh SS has 300K+ olympic images! so likely a big difference in reviewers
20
« on: March 31, 2025, 20:04 »
...
I have deleted a couple months ago approx. 3k AI images because of the bad quality and started completely again with better generated quality. As you can see my earnings increased over the months one can tell that it worked.
But I think it's rather because of the better quality because several hundreds of the old ones, which I have left are still not downloaded yet. So I would say deleting files will not harm your portfolio ranking.
congrats - your difference here is deleting lower quality images where others are deleting those w no sales
21
« on: March 29, 2025, 22:03 »
dead thread alert!!!!!!!!!!!!
22
« on: March 26, 2025, 14:18 »
In the last few days a lot of sales on SS, but almost all in the range of $0.10-$0.15 (level 5).
In the last 3 days SS sales are 3 times those of Adobe for me.
In this moment Adobe can't grow anymore, SS sells a lot but the RPD is low, high priced sales have disappeared. Not a good situation.
I was hoping that SS would lose market share to Adobe (which has the highest RPD), but it's not happening as I expected.
for me, too, SS sales are 2-3x AS. but 3 days isn't enough to draw any conclusions as the market is random - however, my SS RPD has been around .7 over them last year, while SS is around .45 (showing high priced sales still happen every month)
23
« on: March 24, 2025, 14:50 »
I get the idea of cleaning up a portfolio, but I honestly think it's a waste of time. From what I've read, buyers rarely browse an entire portfoliothey just search for specific clips. Instead of spending time deleting content, why not focus on producing more high-quality clips? That will likely have a much bigger impact on your sales than trying to prune the old ones.
... however my percentage of content sold at least once is only 13.9% i still want to increase this percentage,maybe it doesn't make any difference,but as i said,i think it's worth a try.
...
and I almost forgot,in the week between February 24th and March 2nd only 3 sales!
not everything is worth a try when you're also eliminating the chance of sales what's the difference between an image that sold once a year ago, & one that's never sold? how likely is either type to get a new sale? it's self-defeating to rely on an unproved theories of what a site's algorithm is. and your timeline is unrealistic -- watching day by day is just going to magnify your distress
24
« on: March 24, 2025, 14:48 »
Well, if you do delete so many files, please let us know if it has any effect. ...
whatever happens there'll be no reason to correlate with any previous action much to show it as a cause. what will they use as a baseline ? last week, month? 6 months ago? - the system is volatile & changing. one person's action isn't going to be able to explain anything as there are many independent sources of variation over time.
25
« on: March 24, 2025, 13:35 »
I agree with this theory. Mine that have got stuck in review have location keywords that the bot probably didn't understand and may have flagged for IP concerns. When I re-uploaded them without the location names they were accepted overnight.
The only thing that is for certain is that dumb AI is pre-processing everything.
its stupid, but there are several (unrelated) santiago trademarks for a long time SS wouldnt accept my Newport OR pix because of the same-named cigarette
Pages: [1] 2 3 4 5 6 ... 170
|
Sponsors
Microstock Poll Results
Sponsors
|