pancakes

MicrostockGroup Sponsors


Author Topic: Shutterstock Reviewers Beating Me Up.... Anyone Else?  (Read 123122 times)

0 Members and 1 Guest are viewing this topic.

« Reply #25 on: May 17, 2014, 16:39 »
0
...They've certainly upped their standards, but there are enough totally idiotic rejections - such as wrong white balance for pre-sunrise/sunrise/sunset images - that aren't borne out by sales if you talk them into accepting the image....
I'd agree with that.  Frequently, if left to my own devices, would use a warmer light for stuff like candles, time of day etc but, figuring that would lead to white balance rejection, don't do it to the detriment of of the image and (possibly) sales - course, the poxy isolations sell better than any attempt at being creative anyway so probably makes no difference except that they still get accepted.


« Reply #26 on: May 17, 2014, 17:13 »
+4
It seems Shutterstock periodically replace reviewers and every newcomer afflicts contributors for a while during his learning period. I wonder this learning periods costs how much to Shutterstock in total. Yes every undeserved rejection cost more than us to them.

« Reply #27 on: May 17, 2014, 17:38 »
+12
...
I believed the automation theory in the first place - not that they don't have human inspectors but that there is some type of pre-screening (and I'm fairly sure one of their early earnings calls made some reference to automation of the review process).

Go the SEC Filings section of the SS site and pull up Form 424B4 (a prospectus from 2012), search for "review", and you'll find this:  "We also leverage proprietary review technology to pre-filter images and enhance the productivity of our reviewers."

That document is about 18 months old.  I think it's obvious that they've implemented software screens for focus, white balance, noise and "lighting" - that don't work all that well, yet.
« Last Edit: May 17, 2014, 17:44 by stockastic »

« Reply #28 on: May 17, 2014, 19:27 »
+4
...
I believed the automation theory in the first place - not that they don't have human inspectors but that there is some type of pre-screening (and I'm fairly sure one of their early earnings calls made some reference to automation of the review process).

Go the SEC Filings section of the SS site and pull up Form 424B4 (a prospectus from 2012), search for "review", and you'll find this:  "We also leverage proprietary review technology to pre-filter images and enhance the productivity of our reviewers."

That document is about 18 months old.  I think it's obvious that they've implemented software screens for focus, white balance, noise and "lighting" - that don't work all that well, yet.

Good find.  That explains some of why the focus rejections are so prevalent. I assume those are automated rejections then and never see the eyes of a reviewer or the reviewer is too cautious not to override that expensive technology, otherwise why would they need it.

ruxpriencdiam

    This user is banned.
  • Location. Third stone from the sun
« Reply #29 on: May 17, 2014, 19:58 »
-2
...
I believed the automation theory in the first place - not that they don't have human inspectors but that there is some type of pre-screening (and I'm fairly sure one of their early earnings calls made some reference to automation of the review process).

Go the SEC Filings section of the SS site and pull up Form 424B4 (a prospectus from 2012), search for "review", and you'll find this:  "We also leverage proprietary review technology to pre-filter images and enhance the productivity of our reviewers."

That document is about 18 months old.  I think it's obvious that they've implemented software screens for focus, white balance, noise and "lighting" - that don't work all that well, yet.

Good find.  That explains some of why the focus rejections are so prevalent. I assume those are automated rejections then and never see the eyes of a reviewer or the reviewer is too cautious not to override that expensive technology, otherwise why would they need it.
And Ron will tell you no they dont use bots to review in any way, I have brought this up many times over over on the SS forums and Ron always shoots it down, I would post the link addresses but since the search on SS is inop and no longer working it would take to long to do a manual search one by one.

« Reply #30 on: May 17, 2014, 20:09 »
+2
I just love the composition rejection ....I dont do big uploads any more because that seems to trigger a 100% rejection.

gillian vann

  • *Gillian*
« Reply #31 on: May 17, 2014, 21:05 »
+1
It's not just a SS thing. Pretty much all the microstocks seem totally averse to picture elements that make an image interesting. The "golden hour" that photographers love around sunset/sunrise will result in a lower color temp than the reviewers can tolerate. Wrong white balance. Cross lighting or back lighting to show modeling or drama. No. Exposure issues. Full frontal lighting, flattening the subject and reducing tonal values. Oh yes! Scans of 70-year-old negs that would be of great interest to some buyers are out because of (horrors!) film grain or artifacts. And inconsistencies in editorial policies? Don't get me started.

So SS has their product: 35 million bright vacuous happy isolations on white.

Thanks. I feel better.
well that's their "style". Stocksy has a style, and it's all the stuff you mentioned before, even the grain is ok. I have been about 50/50 at SS of late, but i'm shooting for stocksy first and foremost nowadays, so it's little wonder they don't like the few I offer them.  I did have one out of a batch rejected for "noise", even though all the others shot under identical conditions got through. *shrug* that's why we submit to more than one agency. Rejections suck, but they are part of the game.

Ron

« Reply #32 on: May 18, 2014, 00:37 »
+4
...
I believed the automation theory in the first place - not that they don't have human inspectors but that there is some type of pre-screening (and I'm fairly sure one of their early earnings calls made some reference to automation of the review process).

Go the SEC Filings section of the SS site and pull up Form 424B4 (a prospectus from 2012), search for "review", and you'll find this:  "We also leverage proprietary review technology to pre-filter images and enhance the productivity of our reviewers."

That document is about 18 months old.  I think it's obvious that they've implemented software screens for focus, white balance, noise and "lighting" - that don't work all that well, yet.

Good find.  That explains some of why the focus rejections are so prevalent. I assume those are automated rejections then and never see the eyes of a reviewer or the reviewer is too cautious not to override that expensive technology, otherwise why would they need it.
And Ron will tell you no they dont use bots to review in any way, I have brought this up many times over over on the SS forums and Ron always shoots it down, I would post the link addresses but since the search on SS is inop and no longer working it would take to long to do a manual search one by one.
I spoke to SS about this and I am just forwarding on what they told me. I am not shooting anything down, I am just sharing what I know from SS. You and a few others on SS try to gag me every time I dont agree with what you are saying because it doesnt help your agendas. And when I still dont agree after your bullying I am accused of starting arguments. You and a few other try to get some sort of control of the SS forum and bully anyone into shut up and get with the program. I assume you know by now I dont work like that.

Now you can run off to SS forum and spread the word that I attacked you on MSG, as you always do. Forgetting its just another reply where you dragged me into a discussion I had nothing to do with in the first place. 

I apologise for the off topic. Carry on.

Ron

« Reply #33 on: May 18, 2014, 00:47 »
0
How does this explain bots being used?

Quote
Quote from: ruxpriencdiam
Quote from: ruxpriencdiam
Just as thought it is the weekend!

Quote
Poor Lighting--Image has exposure issues and/or incorrect white balance.




Now it has gone from WB to.

Focus--Image is not in focus or the focal point is not appropriate for the composition.



We are lacking in consistency somehow?


If technology is used, it wont be inconsistent on the same image. It would have the same result over and over. But human interference would make something inconsistent.

Beppe Grillo

« Reply #34 on: May 18, 2014, 01:08 »
0
How does this explain bots being used?

Quote
Quote from: ruxpriencdiam
Quote from: ruxpriencdiam
Just as thought it is the weekend!

Quote
Poor Lighting--Image has exposure issues and/or incorrect white balance.




Now it has gone from WB to.

Focus--Image is not in focus or the focal point is not appropriate for the composition.



We are lacking in consistency somehow?


If technology is used, it wont be inconsistent on the same image. It would have the same result over and over. But human interference would make something inconsistent.


As I remember, in the interview that you have had with them some months ago they clearly told you that they use some automatic system to make a pre-review of the images.

From your thread
http://www.microstockgroup.com/shutterstock-com/my-meeting-with-shutterstock/msg352498/

"Reviews:
Not much to say, I asked if they use technology to review the images, yes they do, but a human makes the decision. "
« Last Edit: May 18, 2014, 01:14 by Beppe Grillo »

Ron

« Reply #35 on: May 18, 2014, 01:16 »
0
Yes, I have also explained how they use technology. I have no idea where that quote is, its somewhere on a forum. I think even Scott came in here and explained it.

They open the image automatically for the reviewer at 100% on the screen. To speed things up. The reviewer still makes the decision. How do you explain inconsistency on the same image, if it was 100% automated? Technology is not ambiguous.

Beppe Grillo

« Reply #36 on: May 18, 2014, 03:58 »
+1
Yes, I have also explained how they use technology. I have no idea where that quote is, its somewhere on a forum. I think even Scott came in here and explained it.

They open the image automatically for the reviewer at 100% on the screen. To speed things up. The reviewer still makes the decision. How do you explain inconsistency on the same image, if it was 100% automated? Technology is not ambiguous.

Probably the system proposes to the inspector some  possible "verdicts" and the inspector, hurry to reach his quota of reviewed images for the day, to then go to smoke cigarettes and drink beers with his unemployed friends, pushes hastily one of buttons correspondent to the various rejection reasons offered by the system


(Or do they just use blind people as inspectors??)

« Reply #37 on: May 18, 2014, 06:50 »
0
No rejections here.  Well, a couple for model releases and 3d stuff.

« Reply #38 on: May 18, 2014, 07:11 »
+1
Yes, I have also explained how they use technology. I have no idea where that quote is, its somewhere on a forum. I think even Scott came in here and explained it.

They open the image automatically for the reviewer at 100% on the screen. To speed things up. The reviewer still makes the decision. How do you explain inconsistency on the same image, if it was 100% automated? Technology is not ambiguous.

Probably the system proposes to the inspector some  possible "verdicts" and the inspector, hurry to reach his quota of reviewed images for the day, to then go to smoke cigarettes and drink beers with his unemployed friends, pushes hastily one of buttons correspondent to the various rejection reasons offered by the system


(Or do they just use blind people as inspectors??)

Absolutely no evidence for that.  As Ron said, they use technology for workflow, not for decision making and the killer argument is the apparently inconsistent decisions software will always produce the same outcome based on the same data.

« Reply #39 on: May 18, 2014, 07:27 »
+1
I have always assumed it was people that were at the root of the problem simply because it is so inconsistent. Seriously, if a piece of software was being this erratic, I would scrap it and find a better way.

« Reply #40 on: May 18, 2014, 07:43 »
+1
How does this explain bots being used?

Quote
Quote from: ruxpriencdiam
Quote from: ruxpriencdiam
Just as thought it is the weekend!

Quote
Poor Lighting--Image has exposure issues and/or incorrect white balance.




Now it has gone from WB to.

Focus--Image is not in focus or the focal point is not appropriate for the composition.



We are lacking in consistency somehow?


If technology is used, it wont be inconsistent on the same image. It would have the same result over and over. But human interference would make something inconsistent.


Depends on what the gage R&R study variance is. You are assuming that it performs the same way each time on the same photo (and it should by the way) but we simply do not know for sure without running some statistical testing.  And that ain't gonna happen. We also don't know if 100% of images go through this test (assuming they are still using that proprietary software). Basically, lot's of unknowns.  Maybe they tried it and decided to kill the system, who knows.  I do still get weird rejections there and if I submit for a re-review they almost always get through, which makes me suspect that an upstream filter is being used such as software, but I have no clue really.......and I don't think anyone in this forum really knows for sure.  Your conversations with SS may be accurate and they may have indeed killed the technology or just don't want that cat out of the bag.   

Beppe Grillo

« Reply #41 on: May 18, 2014, 07:56 »
+1
Yes, I have also explained how they use technology. I have no idea where that quote is, its somewhere on a forum. I think even Scott came in here and explained it.

They open the image automatically for the reviewer at 100% on the screen. To speed things up. The reviewer still makes the decision. How do you explain inconsistency on the same image, if it was 100% automated? Technology is not ambiguous.

Probably the system proposes to the inspector some  possible "verdicts" and the inspector, hurry to reach his quota of reviewed images for the day, to then go to smoke cigarettes and drink beers with his unemployed friends, pushes hastily one of buttons correspondent to the various rejection reasons offered by the system


(Or do they just use blind people as inspectors??)

Absolutely no evidence for that.  As Ron said, they use technology for workflow, not for decision making and the killer argument is the apparently inconsistent decisions software will always produce the same outcome based on the same data.

Yes the software will always produce the same outcome based on the same data.
But we don't know what is the response given by the software after having analyzed and image.
Does the software give only one answer, or different possibilities that the inspector should chose then?
If the software gives a choice of different causes for a rejection, letting the task of a  definitive choice to the inspector it is possible that the inspector, in the hurry, choses the wrong answer.

What I mean is that software system or not software system it looks enough obvious that some inspectors do not do their job correctly.
I don't know what are the reasons for this (I can only make suppositions).
One can be a lack of experience or training, because they are new to the job;
The other could be a hurry to finish their daily quota of images to review (if they have one - but in fact I don't know if they work with quotas).



Ron

« Reply #42 on: May 18, 2014, 07:57 »
+1
I agree nobody knows, but some people are all to sure about it, like Ruxperiencediam

« Reply #43 on: May 18, 2014, 10:08 »
+1
I agree nobody knows, but some people are all to sure about it, like Ruxperiencediam

Yea I hear you.

« Reply #44 on: May 18, 2014, 10:43 »
0
I think we can assume that any software used for screening on things like noise, focus and lighting is perpetually 'in development' and is being frequently changed, tweaked and tuned.  As thousands of images are coming in every day, the result of a small change could be major whiplash for some contributors.  The next day might be different.

Obviously we'll never know the details - SS only said enough to impress prospective investors.  As far as I know, no one representing SS has ever - in response to a question - explicitly denied the use of software screening. I think I recall a couple of statements that sort of danced around the subject.
 
« Last Edit: May 18, 2014, 10:47 by stockastic »

« Reply #45 on: May 18, 2014, 11:02 »
0
No rejections here.  Well, a couple for model releases and 3d stuff.

I have also wondered if we all get the same reviewers or if the higher earners get the more experienced reviewers.  I know if I was designing the system this is something I would build in....   

farbled

« Reply #46 on: May 18, 2014, 13:03 »
0
I thought the automation was about making sure the image met the minimum file size and things of that nature. .

Ron

« Reply #47 on: May 18, 2014, 14:09 »
0
File size too low is rejected up front. Those images wont make it to the editor. But I understand what you are saying.

« Reply #48 on: May 18, 2014, 14:37 »
+2
I am guessing that any auto screening lists potential problems for the reviewer who can choose between them. Lazy reviewers just pick on and reject.  I am also guessing that they are constantly tweaking the algorithms as well as possibly changing the thresholds based on keywords (maybe harsher on flowers and cats say).

How sweet would it be if they rejected for focus or shadows and light - and sent it straight to offset.

« Reply #49 on: May 18, 2014, 18:09 »
0
Had a bunch of editorials rejected for "credentials required".  I included a note that the event was public and credentials were not provided but they still rejected the images.  It's not worth fighting them about.  My other sites that take editorials took them all.

I hate this "credentials required" rule they have now.  I've done well in the past with editorials shot at public events.  Now it's almost impossible to get them accepted unless it's something obvious like a parade.


 

Related Topics

  Subject / Started by Replies Last post
25 Replies
10049 Views
Last post April 04, 2015, 16:03
by stuttershock
22 Replies
4991 Views
Last post April 04, 2015, 18:37
by shudderstok
85 Replies
33867 Views
Last post April 04, 2015, 16:02
by stuttershock
10 Replies
5285 Views
Last post June 22, 2015, 14:07
by Freedom
204 Replies
13816 Views
Last post December 13, 2019, 12:46
by Pauws99

Sponsors

Microstock Poll Results