Where I can understand why my picture has been removed?

https://e621.net/post/deleted_index?user_id=101192

This is right on your profile.

elsera said:
Where I can understand why my picture has been removed?

user:elsera status:deleted will list your deleted posts and you can see the reason at the top of the page

Might I hijack the thread, seems better to me than making a new one.
I had two images deleted recently (not that deletions are important, but rather why there were posted to begin with)

post #1009795 and post #1009796
Now I took the images directly from source (you get to it after 5 or so clicks on pixiv)
So my question is, why are the images different (and hence got posted)

Is my url#http://i1.pixiv.net/img-original/img/xxxx/xx/xx/xx/xx/x/xxxxxxx_px.jpg or .png not the true source?

DelurC said:
Might I hijack the thread, seems better to me than making a new one.
I had two images deleted recently (not that deletions are important, but rather why there were posted to begin with)

post #1009795 and post #1009796
Now I took the images directly from source (you get to it after 5 or so clicks on pixiv)
So my question is, why are the images different (and hence got posted)

Is my url#http://i1.pixiv.net/img-original/img/xxxx/xx/xx/xx/xx/x/xxxxxxx_px.jpg or .png not the true source?

my guess is that you or previous poster saving them on computer altered something in the file that didnt have visual change in image

Mutisija said:
they were dupes. you reposted same images that were already posted here months ago.

Please...(You didn't even read what I wrote, my question was what is the source of Pixiv images not why were they deleted!)
Try posting 2 same images and observe what happens.

Mutisija said:
my guess is that you or previous poster saving them on computer altered something in the file that didn't have visual change in image

Simply downloading the image doesn't change it.
Besides, If I compare the two images, the one that remained is actually inferior. The deleted one has 96dpi while the currently remaining one is 72dpi. Resolution is the same and format, but 72dpi one is 24KBs larger.
(Higher file size =/= better)

DelurC said:
Please...(You didn't even read what I wrote, my question was what is the source of Pixiv images not why were they deleted!)
Try posting 2 same images and observe what happens.

i realized seconds after posting that i misunderstood your comment

DelurC said:
Simply downloading the image doesn't change it.
Besides, If I compare the two images, the one that remained is actually inferior. The deleted one has 96dpi while the currently remaining one is 72dpi. Resolution is the same and format, but 72dpi one is 24KBs larger.
(Higher file size =/= better)

whatever but your posts were not deleted because of file size difference, they were deleted because you posted dupes of older posts. if the images are visually exactly same, then the older one is kept. the dpi doesnt matter here because its for printing. it would be unfair to delete post from someone and decrease their upload limit just because someone else posted a dupe of it much later.

DelurC said:
Simply downloading the image doesn't change it.

It actually can. If you have metered connection and/or certain software, they can reduse the images quality for bandwidth purposes. Alternatively if you have connection that drops packages, it can actually alter the image during download if it didn't corrupt it.

In any case, the full original image was removed in favor of pixiv generated one. To me that doesn't sound good at all.

post #1009795 had http://i1.pixiv.net/img-original/img/2013/07/25/02/06/06/37298512_p10.jpg as source. This is the original image which can't be accessed with direct URL (gives 403), but you have to press the fullscreen icon thingy on the left side of image on site.

The post replacing that was post #744750, which has the same MD5 as http://i1.pixiv.net/c/1200x1200/img-master/img/2013/07/25/02/06/06/37298512_p10_master1200.jpg which is Pixiv generated preview image.

They are visually almost identical, but the original is less bloated. Many boorus have strict rules that they will not accept preview images from pixiv and they need to be replaced with original image.

post #1009795 had http://i1.pixiv.net/img-original/img/2013/07/25/02/06/06/37298512_p10.jpg as source. This is the original image which can't be accessed with direct URL (gives 403)

It can be, you just need to insert Pixiv as referer.

Mutisija said:
deleted because you posted dupes of older posts.

They were not dupes, they were better and source images.

Mutisija said:
1. If the images are visually exactly same, then the older one is kept. 2. the dpi doesn't matter here because its for printing. 3. it would be unfair to delete post from someone and decrease their upload limit just because someone else posted a dupe of it much later.

1. Again wrong, if you look around you can find a hundreds thousands of examples where your theory holds no water.
2. dpi or ppi is not for printing only.
3. Dupes can not be posted, only better or worse images can. Also, its unfair reducing my upload count for source images, don't you think that?

DelurC said:
They were not dupes, they were better and source images.

1. Again wrong, if you look around you can find a hundreds thousands of examples where your theory holds no water.
2. dpi or ppi is not for printing only.
3. Dupes can not be posted, only better or worse images can. Also, its unfair reducing my upload count for source images, don't you think that?

They were dupes if they were deleted for being dupes. The admins know what they are doing, Ratte even saw through this , which I was so damn sure it was a dupe that I asked Ratte for them to point it out for me.

Siral_Exan said:
They were dupes if they were deleted for being dupes. The admins know what they are doing, Ratte even saw through this , which I was so damn sure it was a dupe that I asked Ratte for them to point it out for me.

I don't get it. How is it not a dupe? I've got the two images on different tags and swapping between them makes no difference that I can tell.

The filesize is different but I can't see anything else changed.

DelurC said:
They were not dupes, they were better and source images.

We usually keep the source images.
Some of the ones you posted might've got deleted out of habit, because just about everything you had posted up until now were inferior versions.

Clawdragons said:
I don't get it. How is it not a dupe? I've got the two images on different tags and swapping between them makes no difference that I can tell.

The filesize is different but I can't see anything else changed.

??? do you have monitor issues or? because the other one has much darker shadows than the other one.

Ok so, the bottom line is.
The source I provided IS the Pixiv source, and deletions shouldn't be a common occurrence.

P.S. What about those two?

Mutisija said:
??? do you have monitor issues or? because the other one has much darker shadows than the other one.

You say one has darker shadows, so I downloaded both images and compared various parts of the image to each other. Same Hue, same RGB, same Luminosity. Literally the exact same in every way. I checked quite a few areas too. Not a single pixel difference that I could find.

So yeah not a monitor issue.

I also tried a different browser, so not a browser issue. Dunno why it would be but obviously something screwy is going on.

Are you just messing with me?

Dunno about the shading, but post #548550 is noticeably lighter overall.

I was able to see it at a glance, though it's not a major difference. (The color tool that I'm using shows average RGB values of 92/108/107 for post #994787, and 88/109/108 for post #548550. It also claims that the latter has less unique colors, but an another program reports the color counts as equal. Not sure which is correct...)

Genjar said:
Dunno about the shading, but post #548550 is noticeably lighter overall.

I was able to see it at a glance, though it's not a major difference. (The color tool that I'm using shows average RGB values of 92/108/107 for post #994787, and 88/109/108 for post #548550. It also claims that the latter has less unique colors, but an another program reports the color counts as equal. Not sure which is correct...)

o_O they look the same to me. wait...what? here on the site they look the same. like a 100% exact copy. but once downloaded... 548550 is actually a good bit brighter than 994787.

the darker 994787 and the brighter 548550

wth...are you messing with me, visipics? those are clearly 2 different pics. why are you, on any and all filter settings, saying they are duplicates and have the same brightness?

these are some strange pics. o_O

They have always looked different to me just by browser comparison, which is why they're both on the site despite the newer one getting flagged occasionally.

treos said:
the darker 994787 and the brighter 548550

That is NOT how they look on the site. Both are the darker version.

And again I downloaded them both and they were both the darker version.

Something is going on with those two pictures. Different users are seeing them differently.

http://i.imgur.com/M0oNii2.png

This is how the two look to me.

If I save both of them and look at them in PS or SAI, they both look like the darker one, but if I simply copy the images from my browser and paste them in PS or SAI, they look different. However, if I open them as files with a normal image viewer, they look different.

If I convert the darker version to CMYK, then save as a highest-quality jpg, then save that jpg as a png, I get something that is akin to the lighter version. Running both through an EXIF viewer show both as RGB (not CMYK) hence the saving from png to jpg back to png (png cannot save as CMYK, only RGB). I'm wondering, then, if this explains the lighter version? I'm not very sure considering the lighter image is older and the source is no longer available for a better comparison.

Just something to put out there.

By the way, here is a direct comparison made from the copypasting of the images into SAI: https://i.imgur.com/j8dfAoZ.png

Best way to verify if there are some differences is to put each version on its own layer, and set the top one as Difference. If it's all black, then it means there's no difference whatsoever, and that's what I get.

One of them seems to have been crushed/optimized for the size, would be the single reason to why it's seen as a different but same picture.

And I just opened them with the Windows Image Viewer and now I can clearly see one is lighter than the other. Okay there is some black magic going on here

Ratte said:
If I save both of them and look at them in PS or SAI, they both look like the darker one, but if I simply copy the images from my browser and paste them in PS or SAI, they look different. However, if I open them as files with a normal image viewer, they look different.

If I convert the darker version to CMYK, then save as a highest-quality jpg, then save that jpg as a png, I get something that is akin to the lighter version. Running both through an EXIF viewer show both as RGB (not CMYK) hence the saving from png to jpg back to png (png cannot save as CMYK, only RGB). I'm wondering, then, if this explains the lighter version? I'm not very sure considering the lighter image is older and the source is no longer available for a better comparison.

Just something to put out there.

By the way, here is a direct comparison made from the copypasting of the images into SAI: https://i.imgur.com/j8dfAoZ.png

Wha... huh?

I decided to try this too and... If I open them in Windows Photo Viewer they are distinct.

That's screwy.

That said, copypasting them and downloading them makes no difference to me.

Clawdragons said:
Wha... huh?

I decided to try this too and... If I open them in Windows Photo Viewer they are distinct.

That's screwy.

That said, copypasting them and downloading them makes no difference to me.

Yes, if I open them up in a viewer they are different, but if I open them in an editor like SAI or PS, they're identical. If I take the darker version and change RGB to CMYK, the colors are the same as the lighter version. Since png can't save CMYK, this can be circumvented by saving as anything else, reopening the image, and resaving as png, lol.

EDIT: Opening both images in Firefox shows them both identical darker versions while Chrome/Chromium show two distinct versions. This may explain why people are seeing different things at least in their browsers since browsers will have different color management. This just leaves me with more questions regarding how they're viewed in an image viewer versus an image editing program.

Ratte said:
Yes, if I open them up in a viewer they are different, but if I open them in an editor like SAI or PS, they're identical. If I take the darker version and change RGB to CMYK, the colors are the same as the lighter version. Since png can't save CMYK, this can be circumvented by saving as anything else, reopening the image, and resaving as png, lol.

EDIT: Opening both images in Firefox shows them both identical darker versions while Chrome/Chromium show two distinct versions. This may explain why people are seeing different things at least in their browsers since browsers will have different color management. This just leaves me with more questions regarding how they're viewed in an image viewer versus an image editing program.

Sorry, perhaps I wasn't clear.

I understood what you were saying happened. I was saying that the same thing was happening for me - that it depended what program I opened it with, for how it looked. I wasn't confused about what happened for you, I was confirming that it was happening that way for me as well.

Again, agreeing with you: I just tried Chrome, my experience agreed with yours, different versions. While on Firefox (which is what I usually use), both are the darker version.

At least there's some consistency.

Ratte said:
Yes, if I open them up in a viewer they are different, but if I open them in an editor like SAI or PS, they're identical. If I take the darker version and change RGB to CMYK, the colors are the same as the lighter version. Since png can't save CMYK, this can be circumvented by saving as anything else, reopening the image, and resaving as png, lol.

EDIT: Opening both images in Firefox shows them both identical darker versions while Chrome/Chromium show two distinct versions. This may explain why people are seeing different things at least in their browsers since browsers will have different color management. This just leaves me with more questions regarding how they're viewed in an image viewer versus an image editing program.

what a strange pair of pics we've found. o_O well, i have no clue as to what going on with them past what i pointed out earlier though it would be interesting to see this mystery cleared up.

You guys just discovered the beauty of exif data and treat it like the first humans who saw fire. Firefox usually ignores exif and just displays everything as rgb because your device is all but guaranteed to use a rgb display. On the other hand a lot of specialized image viewers read and understand the exif data and try to follow it to emulate a more "true" render of the presented data.
Most notably for this effect are DPI and color range, although color range is debatable because rgb displays can't properly display the cmyk color range.

NotMeNotYou said:
You guys just discovered the beauty of exif data and treat it like the first humans who saw fire. Firefox usually ignores exif and just displays everything as rgb because your device is all but guaranteed to use a rgb display. On the other hand a lot of specialized image viewers read and understand the exif data and try to follow it to emulate a more "true" render of the presented data.
Most notably for this effect are DPI and color range, although color range is debatable because rgb displays can't properly display the cmyk color range.

When I put both in an EXIF data reader they both said it's RGB. This is in Opera.

Darker: http://puu.sh/rux8H/0fed0ce068.png
Lighter: http://puu.sh/ruxar/c7a15990bc.png

It wouldn't matter because png is not compliant with CMYK, so if they are both png (which they are) CMYK would not show up anyway unless it was saved as something else before being saved as png. EXIF data would not have really done much here.

NotMeNotYou said:
treat it like the first humans who saw fire.

The majority of e621 users in this thread (myself included)

Clawdragons said:
The majority of e621 users in this thread (myself included)

I expected this

oh great, what have i gone and discovered this time?

sorcery! even when downloaded and viewed in Windows Photo Viewer they are 100% identical!

post #979991 Size: 866x1069 (1022.2 KB)

post #991164 Size: 866x1069 (644.9 KB)

both are also .png

no difference in brightness or anything this time so how are these not dupes?

there also seems to be some poorly done edits/cropping on some other pics mixed in the search results of this artist such as...

post #985085 and post #979989 not the white edge that was removed.

*facepalm* did i seriously just do that? my stupid mouse opened the same tab twice when i middle clicked the thumbnail once.

treos said:
sorcery! even when downloaded and viewed in Windows Photo Viewer they are 100% identical!

post #979991 Size: 866x1069 (1022.2 KB)

post #991164 Size: 866x1069 (644.9 KB)

I zoomed in on them. One of them clearly has some compression. Definitely different levels of quality.

EXIF data and embedded color profiles can inflate filesize. It won't be anything visible to a viewer. That's the case of post #979991 and post #991164. The former is larger because it has an embedded color profile (and generally more EXIF data) while the latter does not. The duplicate-checking system relies on a piece of data you can't actually see but if the image is the same with all the same data (again, data you can't actually see) it would match and the duplicate would be rejected. Both of the images have a different MD5 hash so the site would not have been able to recognize them as duplicates.

Further, the newer one is better quality.

Ratte said:
EXIF data and embedded color profiles can inflate filesize. It won't be anything visible to a viewer. That's the case of post #979991 and post #991164. The former is larger because it has an embedded color profile (and generally more EXIF data) while the latter does not. The duplicate-checking system relies on a piece of data you can't actually see but if the image is the same with all the same data (again, data you can't actually see) it would match and the duplicate would be rejected. Both of the images have a different MD5 hash so the site would not have been able to recognize them as duplicates.

Further, the newer one is better quality.

:/ well that certainly makes sorting out potential dupes more annoying.