Testing photohosting for digital wear (21 pieces)
2017-10-30 16:43. Disclaimer: This analysis was conducted based on the model of user behavior using specialized photo hosting sites. In the case of universal mail hosting and social networks, the user has a different model of behavior and there will be separate studies about it.
Last time I described the phenomenon and introduced the term " digital depreciation ", and the next morning, after publication, I found out that on the same day a review with a reference to a similar topic was published on mobile-review. The review tells the story of a man who for several years uploaded his photos to Google Photo and eventually lost the source.
')
“This is a sign,” I thought, and decided to analyze popular photo hosting sites for digital wear of photos.
For this, I selected the wear criteria and modeled the real situation from the life cycle of digital photos.
Experimental conditions: the user went to a public event (in my case, the All-Russian Lumberjack 2017 championship), when he returned home he threw the photos onto the Internet (using a browser) and then tried the source. I tried in the most usual way - the disc was colored, accidentally erased the folder, deliberately cleaned the disc, being sure that everything was saved on the Internet, etc. And then, time later, the user realized that he did not have the original and went to download what he had "on the Internet."
But he forgot that any services at any time can pass your photos through a meat grinder or close at all.
Let's see what was at the beginning, what was left at the end and how big is the wear.
Go!
As a base of photo hosting sites I used the collective unconscious - I found an article and took those sites that are in the first 10 thousand on Alexa. (Then it turned out that the data on Wikipedia were outdated, but the list was good.) I added services from similar Russian-language Wikipedia articles and a couple of my head to the list.
As a result, turned out sites that meet the criterion of "where the average user is likely to upload their photos."
At first it turned out 20 sites, but then I added Tumbler and it turned out a beautiful even number.
Neither one nor the other is important in this experiment.
And in general, our test user is a dolt out, he uploaded a picture not to make a backup, but to ... in general, for which you now upload pictures to the Internet, so you downloaded it.
As a decent dolt, he preferred not to register, and if he registered where, he forgot the password safely.
In short, a classic kettle.
It was checked how much initial information was lost.
The results are tabulated:
FILE NAME - knowing the original file name, you can try to search it on the disk - suddenly he was lying around somewhere. Changing the number in the file name, you can try to find the disk other photos from the same event.
MD5-HASH FILE - whether the contents of the file coincides with all the junk
FILE SIZE - some kind, and the quality criterion for pictures whose size is in pixels
EXIF (cleared) - useful technical information. The elements' FileName ', FileDateTime', 'SectionsFound', 'COMPUTED', 'FileSize' were deleted from the array that the php-function exif_read_data produced. The result is serialized and hashed. “No” means that there is NO COMPLETE match, but part of the data can be saved (date, camera model, etc.)
DATE SHOOTING - since the shooting is reportage with us, it is very important to know when it happened. This information was extracted from exif data.
WIDTH OF THE PICTURE - if otresayzili, then how bad
ALEXA RANK - how popular is the service
The result was three groups of sites:
EXCELLENT - zero or almost zero digital wear photos. Given the complete preservation of the contents, I decided not to insist on preserving the name of the source file.
GOOD - the photo got minimal digital wear. The original resolution was retained, but the file was recompressed. Here and there exif is preserved.
THE REST - not that the picture turns into stuffing, which cannot be turned back, but there are significant losses. Surprisingly, a third of these services retained the original exif, and more than half the shooting date.
Links to specific pages on photohosting will be after the table. You can check it yourself.
Additional notes on various photo hosting sites:
https://pinterest.com/ - surprisingly careful handling of materials, given that it is also a social network. Requires registration. I downloaded the picture from the viewing page.
https://www.flickr.com/ - Flickr he is Flickr. Requires registration. Download from the watch page. There are settings for downloading files and saving exif.
http://www.photobucket.com/ - was once popular, but now it has become rigidly monetized, demanding money for the possibility of embedding images on other sites and very roughly hung with advertising. It's bad to be like this.
http://www.fotolog.com/ - Requires registration. Downloading from the public part. Quality is worse than photobucket.com. What does he even do in this review?
https://imgur.com/ - does not require registration. Downloaded from the page viewand then deleted the file . Previously, it could be stored without compression, but then changed the rules .
https://photos.google.com/ - Requires registration. Download from the watch page. You need to select the correct file storage option.
https://fotki.yandex.ru/ - Requires registration. Download from the watch page. You can prohibit downloading the original
https://my.mail.ru/ - Requires registration. Download from the watch page. It used to be called Photo@Mail.Ru, but was later transferred to My World and, as they say on Wikipedia, " The originals of the photos were deleted without warning the users. The compressed copies of the photos were transferred to My World. "
https://500px.com/ - Requires registration. You can only download the page, and it does not matter that you are the owner.
http://www.fotki.com/ - Requires registration (some 30-day trial is announced). A variant of the file that is included in the table has been downloaded from the owner’s account. In the public part of the picture is available in width of 1400 pixels.
radikal.ru - The old man was surprised. Really surprised! Registration is not required. Filled with selected fill option without recompression. Then deleted.
https://www.smugmug.com/ - Requires registration (some 14 day trial). The original was downloaded from the admin. In the public part, the quality is worse.
https://unsplash.com/ - Requires registration. Downloading from the public part.
http://tinypic.com/ - Does not require registration (it is, but did not register). Downloading from the public part. It is hung with advertising.
http://photo.qip.ru/ - Requires registration. Downloading from the public part. At the level of photos, there is some sort of option “Protecting originals of photos” in which even the logged in authors cannot be downloaded. It should be noted that the "original" is a very strange name for a file in which the pixels are 4 times smaller than the original, and the byte is 14 (!) Times.
https://pixabay.com/ - Requires registration. Downloaded logged in from the public part. It's funny, but in the end the file did not pass some moderation. This crap with the date is also when they have only DateTimeOriginal saved from the three fields (DateTime and DateTimeDigitized are not). In spite of the not very hard chewing of the photo - FU SUCH BE!
https://www.shutterfly.com/ - Requires registration. And in general, this is not a photo hosting service, but a printing service. Download from admin. Very respectful attitude!
https://www.snapfish.com/ - Requires registration. Another print service. When registering you can not insert data from the buffer. Download from admin. From the public part with a lower resolution. In general, a very strange service - they saved the file name, but they reshoot. In exif saved a lot of data, then I will look why it did not coincide with the original.
https://www.deviantart.com/ - Requires registration. Unexpectedly good result. Downloading from the public part. From the admin part can not be downloaded. There is a ban on downloading the original. That is, you can not download it yourself without allowing others.
https://www.tumblr.com/ - Requires registration. Downloading from the public part. From the admin part - no. The expected result for the social network.
https://commons.wikimedia.org - Requires registration. It would be strange if they did not keep the original, but it is strange that they renamed the file.
With this analysis of popular photo hosts, I wanted to show how digital wear occurs in reality, when a person loses the ability to get the original photo and remains so that he once uploaded to the Internet.
In practice, it turns out the same way as with paper photos - once you had a negative, and the photos were printed for the whole class, but after years there were only a couple of photos left, one rumpled and the other cut.
What happened to me in the experiment is pure wear, albeit digital.
This text has not only demonstrated how theoretically immortal digital photos turn into dust, but also how and by what criteria to choose a photo hosting site.
So, if you feel sorry for releasing photos in good quality into the world, use the services from the group “OTHER”.
In all other cases, choose services from the “EXCELLENT” group, but remember that not only can you download the originals.
And, in any case, keep in mind that any services at any time can pass your photos through a meat grinder or close at all.
* * *
The people, and throw in the comments of the information on the closed photo sharing sites that were previously cool.
Picas
Panoramo
… who else?
Last time I described the phenomenon and introduced the term " digital depreciation ", and the next morning, after publication, I found out that on the same day a review with a reference to a similar topic was published on mobile-review. The review tells the story of a man who for several years uploaded his photos to Google Photo and eventually lost the source.
')
“This is a sign,” I thought, and decided to analyze popular photo hosting sites for digital wear of photos.
For this, I selected the wear criteria and modeled the real situation from the life cycle of digital photos.
Experimental conditions: the user went to a public event (in my case, the All-Russian Lumberjack 2017 championship), when he returned home he threw the photos onto the Internet (using a browser) and then tried the source. I tried in the most usual way - the disc was colored, accidentally erased the folder, deliberately cleaned the disc, being sure that everything was saved on the Internet, etc. And then, time later, the user realized that he did not have the original and went to download what he had "on the Internet."
But he forgot that any services at any time can pass your photos through a meat grinder or close at all.
Let's see what was at the beginning, what was left at the end and how big is the wear.
Go!
1. Selection of photo hosting sites
As a base of photo hosting sites I used the collective unconscious - I found an article and took those sites that are in the first 10 thousand on Alexa. (Then it turned out that the data on Wikipedia were outdated, but the list was good.) I added services from similar Russian-language Wikipedia articles and a couple of my head to the list.
As a result, turned out sites that meet the criterion of "where the average user is likely to upload their photos."
At first it turned out 20 sites, but then I added Tumbler and it turned out a beautiful even number.
2. What was not counted
- The total volume and maximum size of the downloaded file was not checked.
- Ignored the license under which the photo is loaded (yes, I did it consciously and I will not recall the photo from Wikipedia - you can insert it into the article about Ilona Bronevitskaya).
- Shelf life and cost are also not considered.
Neither one nor the other is important in this experiment.
And in general, our test user is a dolt out, he uploaded a picture not to make a backup, but to ... in general, for which you now upload pictures to the Internet, so you downloaded it.
As a decent dolt, he preferred not to register, and if he registered where, he forgot the password safely.
In short, a classic kettle.
3. What counted
It was checked how much initial information was lost.
The results are tabulated:
FILE NAME - knowing the original file name, you can try to search it on the disk - suddenly he was lying around somewhere. Changing the number in the file name, you can try to find the disk other photos from the same event.
MD5-HASH FILE - whether the contents of the file coincides with all the junk
FILE SIZE - some kind, and the quality criterion for pictures whose size is in pixels
EXIF (cleared) - useful technical information. The elements' FileName ', FileDateTime', 'SectionsFound', 'COMPUTED', 'FileSize' were deleted from the array that the php-function exif_read_data produced. The result is serialized and hashed. “No” means that there is NO COMPLETE match, but part of the data can be saved (date, camera model, etc.)
DATE SHOOTING - since the shooting is reportage with us, it is very important to know when it happened. This information was extracted from exif data.
WIDTH OF THE PICTURE - if otresayzili, then how bad
ALEXA RANK - how popular is the service
4. Results
The result was three groups of sites:
EXCELLENT - zero or almost zero digital wear photos. Given the complete preservation of the contents, I decided not to insist on preserving the name of the source file.
GOOD - the photo got minimal digital wear. The original resolution was retained, but the file was recompressed. Here and there exif is preserved.
THE REST - not that the picture turns into stuffing, which cannot be turned back, but there are significant losses. Surprisingly, a third of these services retained the original exif, and more than half the shooting date.
Links to specific pages on photohosting will be after the table. You can check it yourself.
Additional notes on various photo hosting sites:
https://pinterest.com/ - surprisingly careful handling of materials, given that it is also a social network. Requires registration. I downloaded the picture from the viewing page.
https://www.flickr.com/ - Flickr he is Flickr. Requires registration. Download from the watch page. There are settings for downloading files and saving exif.
http://www.photobucket.com/ - was once popular, but now it has become rigidly monetized, demanding money for the possibility of embedding images on other sites and very roughly hung with advertising. It's bad to be like this.
http://www.fotolog.com/ - Requires registration. Downloading from the public part. Quality is worse than photobucket.com. What does he even do in this review?
https://imgur.com/ - does not require registration. Downloaded from the page view
https://photos.google.com/ - Requires registration. Download from the watch page. You need to select the correct file storage option.
https://fotki.yandex.ru/ - Requires registration. Download from the watch page. You can prohibit downloading the original
https://my.mail.ru/ - Requires registration. Download from the watch page. It used to be called Photo@Mail.Ru, but was later transferred to My World and, as they say on Wikipedia, " The originals of the photos were deleted without warning the users. The compressed copies of the photos were transferred to My World. "
https://500px.com/ - Requires registration. You can only download the page, and it does not matter that you are the owner.
http://www.fotki.com/ - Requires registration (some 30-day trial is announced). A variant of the file that is included in the table has been downloaded from the owner’s account. In the public part of the picture is available in width of 1400 pixels.
radikal.ru - The old man was surprised. Really surprised! Registration is not required. Filled with selected fill option without recompression. Then deleted.
https://www.smugmug.com/ - Requires registration (some 14 day trial). The original was downloaded from the admin. In the public part, the quality is worse.
https://unsplash.com/ - Requires registration. Downloading from the public part.
http://tinypic.com/ - Does not require registration (it is, but did not register). Downloading from the public part. It is hung with advertising.
http://photo.qip.ru/ - Requires registration. Downloading from the public part. At the level of photos, there is some sort of option “Protecting originals of photos” in which even the logged in authors cannot be downloaded. It should be noted that the "original" is a very strange name for a file in which the pixels are 4 times smaller than the original, and the byte is 14 (!) Times.
https://pixabay.com/ - Requires registration. Downloaded logged in from the public part. It's funny, but in the end the file did not pass some moderation. This crap with the date is also when they have only DateTimeOriginal saved from the three fields (DateTime and DateTimeDigitized are not). In spite of the not very hard chewing of the photo - FU SUCH BE!
https://www.shutterfly.com/ - Requires registration. And in general, this is not a photo hosting service, but a printing service. Download from admin. Very respectful attitude!
https://www.snapfish.com/ - Requires registration. Another print service. When registering you can not insert data from the buffer. Download from admin. From the public part with a lower resolution. In general, a very strange service - they saved the file name, but they reshoot. In exif saved a lot of data, then I will look why it did not coincide with the original.
https://www.deviantart.com/ - Requires registration. Unexpectedly good result. Downloading from the public part. From the admin part can not be downloaded. There is a ban on downloading the original. That is, you can not download it yourself without allowing others.
https://www.tumblr.com/ - Requires registration. Downloading from the public part. From the admin part - no. The expected result for the social network.
https://commons.wikimedia.org - Requires registration. It would be strange if they did not keep the original, but it is strange that they renamed the file.
5. Conclusions
With this analysis of popular photo hosts, I wanted to show how digital wear occurs in reality, when a person loses the ability to get the original photo and remains so that he once uploaded to the Internet.
In practice, it turns out the same way as with paper photos - once you had a negative, and the photos were printed for the whole class, but after years there were only a couple of photos left, one rumpled and the other cut.
What happened to me in the experiment is pure wear, albeit digital.
6. Another conclusions
This text has not only demonstrated how theoretically immortal digital photos turn into dust, but also how and by what criteria to choose a photo hosting site.
So, if you feel sorry for releasing photos in good quality into the world, use the services from the group “OTHER”.
In all other cases, choose services from the “EXCELLENT” group, but remember that not only can you download the originals.
And, in any case, keep in mind that any services at any time can pass your photos through a meat grinder or close at all.
* * *
The people, and throw in the comments of the information on the closed photo sharing sites that were previously cool.
Picas
Panoramo
… who else?
Source: https://habr.com/ru/post/341238/
All Articles