bit Rot examples

[Toronto Photography]

Had a discussion of Bit rot and its effects. Of course some bit rot may just be older USB sticks (as might be the case in this example):


The photo is taken 2008 (thought 2007) of a young lady. Over time 10 years or so we have such a photo. The majority of photos over 90% have no easily visible rotting (Its hard to compare photo to photo since any minor rot may be universal -

another older example stored in the same media 1999 MX700 (1.5 meg)

 

[tko]

Simply finding a bad image doesn't prove bit rot, which is one of those vague and meaningless terms. What you call an ancient scroll that is hard to read? Alphabet rot?

The card has failed or become scrambled or was never good to began with. If it was some type of "bit rot," all your images from that era would shows signs of degradation.

It's a well known fact that some digital mediums do not have a long storage time. CDs, for example. No medium, be it paper, stone, or electronic, lasts forever.

 

[Toronto Photography]

Simply finding a bad image doesn't prove bit rot, which is one of those vague and meaningless terms. What you call an ancient scroll that is hard to read? Alphabet rot?

The card has failed or become scrambled or was never good to began with. If it was some type of "bit rot," all your images from that era would shows signs of degradation.

It's a well known fact that some digital mediums do not have a long storage time. CDs, for example. No medium, be it paper, stone, or electronic, lasts forever.

First I am using Bitrot generically for the phenomena. Short say for "possible" bitrot.

2nd your point is correct its not a proof, nor the only possible cause.

However your counter claim appears weak. If a drive fails you expect lots or all photos to fail. Bit rot in the classical sense has random data missing, and eventually those pieces cause random noticeable damage.

Simple test. Take the same photo make copy 1 and copy 2/. Then make repeatedly company each jpg over and over again the same or x-1 times for the send copy if you want full parity. While both will get worse the effects of that being worse will be significantly different on two identical photos.

Unfortunately the experiment is not documented on line - only single re - savings.

https://www.reddit.com/r/technology/comments/86v2z/see_what_happen_when_saving_600_times_a_jpg_image

 

[MediaArchivist]

Simple test. Take the same photo make copy 1 and copy 2/. Then make repeatedly company each jpg over and over again the same or x-1 times for the send copy if you want full parity.

By copy, do you mean simply copying as with the operating system (e.g. "Duplicate" in the Mac finder)? Or do you mean "save as" from within an image editor (and does it have to be jpg?)?

While both will get worse the effects of that being worse will be significantly different on two identical photos.

I'd like to get more clarity on your test, it seems to me that the result you suggest is either nonexistent or an example of lossy compression.

 

[Sue Anne Rush]

Excellent photographs series - thank you for sharing.

Had a discussion of Bit rot and its effects. Of course some bit rot may just be older USB sticks (as might be the case in this example):


The photo is taken 2008 (thought 2007) of a young lady. Over time 10 years or so we have such a photo. The majority of photos over 90% have no easily visible rotting (Its hard to compare photo to photo since any minor rot may be universal -

another older example stored in the same media 1999 MX700 (1.5 meg)

--
Sue Anne Rush

 

[CAcreeks]

Beautiful, Toronto Photog! It is Art with a capital A. This is the kind of beauty that can't be duplicated with photoshop. Only an old-ish USB stick or SD card is capable of this skill.

Excellent photographs series - thank you for sharing.

Had a discussion of Bit rot and its effects. Of course some bit rot may just be older USB sticks (as might be the case in this example):


The photo is taken 2008 (thought 2007) of a young lady. Over time 10 years or so we have such a photo. The majority of photos over 90% have no easily visible rotting (Its hard to compare photo to photo since any minor rot may be universal -

another older example stored in the same media 1999 MX700 (1.5 meg)

 

[Toronto Photography]

Simple test. Take the same photo make copy 1 and copy 2/. Then make repeatedly company each jpg over and over again the same or x-1 times for the send copy if you want full parity.

By copy, do you mean simply copying as with the operating system (e.g. "Duplicate" in the Mac finder)? Or do you mean "save as" from within an image editor (and does it have to be jpg?)?

While both will get worse the effects of that being worse will be significantly different on two identical photos.

I'd like to get more clarity on your test, it seems to me that the result you suggest is either nonexistent or an example of lossy compression.

Yes .. . all Jpgs by definition have lossy compression. I sent the link, many people have done the single test. Save as is most of them. Not the same as BITROT only that data loss from even the same antilogarithm, because of chaos theory, ends up being different (even thought, it should be the same)

The point is that bitrot is entirely random, and thus its effects like camera failure or SD are not on a set schedule per photo.

JPG compression is done the same each time, and the loss should be the same for the same photo, but its isn't - even thought its very close - due to small insignificant copy errors compounding over time - like the number of words Inuit have for snow going up early.

MOre links

Discussion

Loading…

graphicdesign.stackexchange.com

 

[sybersitizen]

JPG compression is done the same each time, and the loss should be the same for the same photo, but its isn't - even thought its very close - due to small insignificant copy errors compounding over time ...

The difference there is not due to copy errors. It's due to different calculation results when the compression algorithm determines how much data to discard. That has nothing to do with 'bit rot', which is about errors occurring in the recording medium.

 

[mdmarqphoto]

Simple test. Take the same photo make copy 1 and copy 2/. Then make repeatedly company each jpg over and over again the same or x-1 times for the send copy if you want full parity.

By copy, do you mean simply copying as with the operating system (e.g. "Duplicate" in the Mac finder)? Or do you mean "save as" from within an image editor (and does it have to be jpg?)?

While both will get worse the effects of that being worse will be significantly different on two identical photos.

I'd like to get more clarity on your test, it seems to me that the result you suggest is either nonexistent or an example of lossy compression.

Yes .. . all Jpgs by definition have lossy compression. I sent the link, many people have done the single test. Save as is most of them. Not the same as BITROT only that data loss from even the same antilogarithm, because of chaos theory, ends up being different (even thought, it should be the same)

Not really, see below.

The point is that bitrot is entirely random, and thus its effects like camera failure or SD are not on a set schedule per photo.

JPG compression is done the same each time, and the loss should be the same for the same photo, but its isn't - even thought its very close - due to small insignificant copy errors compounding over time - like the number of words Inuit have for snow going up early.

Not exactly. The algorithm may be the same, but the image it's working with is not.

Example:

Open a TIF file in an editor, the image will be rendered in memory with all the original image data. Save that file as a JPG and the algorithm will evaluate the image in memory and compress/save it.

Now open that JPG in an editor and it will be rendered in memory with missing data (as expected due to the JPG compression). Performing a Save at this point will cause the JPG algorithm to be performed on the currently rendered image in memory, which is NOT the original image, so the JPG algorithm will look for ways to compress it as it normally would do. The algorithm has no idea what the original image looked like, only the current image, so it can't match the compression exactly as the first time. That is why a JPG image will degrade over many many new Saves. This isn't bit rot, it's just the JPG algorithm doing what it's designed to do.

MOre links

Discussion

https://graphicdesign.stackexchange...ave-a-jpeg-until-its-quality-noticeably-drops

 

[KAAMBIC]

It looks like art now, you should print and sell it at a gallery.

 

[MediaArchivist]

Unfortunately the experiment is not documented on line - only single re - savings.

https://www.reddit.com/r/technology/comments/86v2z/see_what_happen_when_saving_600_times_a_jpg_image

Key phrase in the link you shared: "increasing compression on each iteration". You are referring to an expected phenomenon that is not bit rot. This is completely explained with (somewhat complex) math.

 

[MediaArchivist]

MOre links

Each of these videos demonstrate successive information loss due to lossy compression. It's called "lossy" compression for a reason. There is nothing unexpected or untoward being demonstrated in these videos.

Discussion

https://graphicdesign.stackexchange...ave-a-jpeg-until-its-quality-noticeably-drops

The title says it all. As an experiment, do the exact same thing but with tiff, psd, bmp, xpm, or png. There will be no drop in quality (assuming a straight "save" or "save as" without resampling).

 

[FrancoD]

The problem is the terminology you use.

Saying bit rot sounds like the data is self ditructing but that is not the case.

Think of your jeans develloping holes. Does that mean that cotton rots all by it self ?

 

[Autonerd]

Well, bit rot is a thing -- it's basically storage media starting to die of old age. Seen it plenty of times. The media slowly starts to get corrupt, and data on those unfortunate parts of the disk are lost or corrupted.

"Rot" is a bad term IMHO since it implies some sort of organic decay. That doesn't happen to data -- a 1 or a 0 will always be a 1 or a 0, unless the media on which it is written goes bad. If you have some form of redundant storage, you won't lose data over time.

Those pics look like classic corrupted JPGs. Seen plenty of those, too -- JPGs that didn't download or transfer correctly.

Moral of the story: Use cloud storage or keep backups and change the drives regularly. Me, I like cloud. Nice and easy.

Aaron

 

[Toronto Photography]

JPG compression is done the same each time, and the loss should be the same for the same photo, but its isn't - even thought its very close - due to small insignificant copy errors compounding over time ...

The difference there is not due to copy errors. It's due to different calculation results when the compression algorithm determines how much data to discard. That has nothing to do with 'bit rot', which is about errors occurring in the recording medium.

Well then I mislabel the term. I am talking about photos that deteriorate over time faster than the recording medium decays. Though I suppose I am talking about that too. How ever you make make the unsubstantiated claim that in somehow all jpgs fail at the same rate.

Please back with data. This is contrary to what I have read, and experienced.

 

[Quarkcharmed]

JPG compression is done the same each time, and the loss should be the same for the same photo, but its isn't - even thought its very close - due to small insignificant copy errors compounding over time ...

The difference there is not due to copy errors. It's due to different calculation results when the compression algorithm determines how much data to discard. That has nothing to do with 'bit rot', which is about errors occurring in the recording medium.

Well then I mislabel the term. I am talking about photos that deteriorate over time faster than the recording medium decays.

No, that simply doesn't happen with digital storages. It doesn't happen with non-digital (analogue) storages either btw.

The information is the medium in certain sense. If the medium is intact, the information is intact. Full stop.

Please back with data. This is contrary to what I have read, and experienced.

There's no need to back that with data - the above statement is in the same league as flat earth conspiracy and alike.

 

[Adrian Tung]

I don't understand how you read this statement:

The difference there is not due to copy errors. It's due to different calculation results when the compression algorithm determines how much data to discard. That has nothing to do with 'bit rot', which is about errors occurring in the recording medium.

... and somehow arrived at this conclusion:

How ever you make make the unsubstantiated claim that in somehow all jpgs fail at the same rate.

 

[sybersitizen]

JPG compression is done the same each time, and the loss should be the same for the same photo, but its isn't - even thought its very close - due to small insignificant copy errors compounding over time ...

The difference there is not due to copy errors. It's due to different calculation results when the compression algorithm determines how much data to discard. That has nothing to do with 'bit rot', which is about errors occurring in the recording medium.

Well then I mislabel the term. I am talking about photos that deteriorate over time faster than the recording medium decays.

The recording medium is the thing that's vulnerable to deterioration over time.

JPEGs that 'deteriorate' with repeated saving, reopening, and resaving cycles do so for a very different reason that has nothing to do with the passage of time. It happens because of the way the compression algorithm works when it's applied recursively.

Though I suppose I am talking about that too. How ever you make make the unsubstantiated claim that in somehow all jpgs fail at the same rate.

That sentence is almost indecipherable, but whatever you intended to say with it is wrong.

Please back with data.

Please communicate something sensible in English, then back that with data.

If you want a reference confirming the existence of JPEG recompression errors, there are hundreds out there - but you can probably get the gist of it from this one:

https://www.impulseadventure.com/photo/jpeg-resaving.html

Then tackle this page, paying particular attention the sections on recompression errors:

https://www.impulseadventure.com/photo/jpeg-compression.html

This is contrary to what I have read, and experienced.

Perhaps you've just misunderstood what you've read and experienced.

 

[Autonerd]

JPEGs that 'deteriorate' with repeated saving, reopening, and resaving cycles do so for a very different reason that has nothing to do with the passage of time. It happens because of the way the compression algorithm works when it's applied recursively.

Erm... I don't think anyone was talking about that (re-compressing images). I think they are talking about files simply going bad over time. (Which is really files going bad because of media going bad.)

Aaron

 

[sybersitizen]

JPEGs that 'deteriorate' with repeated saving, reopening, and resaving cycles do so for a very different reason that has nothing to do with the passage of time. It happens because of the way the compression algorithm works when it's applied recursively.

Erm... I don't think anyone was talking about that (re-compressing images). I think they are talking about files simply going bad over time. (Which is really files going bad because of media going bad.)

Your take regarding what the OP is talking about with me isn't necessary, and your selective removal of the actual context of that discussion is in fact detrimental to the goal of clear communication. Not helping.

You could engage him yourself, though, if you have something to say to him. For example, you might try asking him what this statement of his means:

I am talking about photos that deteriorate over time faster than the recording medium decays.