What is your System for managing/storing your files?

[Carey Brown]

This is an important point. Whatever you use it is important to use a system that will be of use if your software or system simply disappears.

YES : i did make a yearly list of all folder + caption, when i was shooting news. I am kind of retired at the moment.

This is one reason to embed the EXIF meta-data back into the image file itself. Where ever the file goes then so does the data.

--
Photos at http://inasphere.com

 

[Michael Fryd]

...

For a unique ID most people go with a time stamp of when the photo was taken: YYYYMMDD-HHMMSS(+optional description).

...

I find there is a very useful advantage of using a sequence number (YYYYMMDD-SEQN) rather than a timestamp in my filenames. When discussing images from a shoot with a client, we can refer to images by simply the sequence number. My client can tell me that they want prints from images -023, -049, and -050.

I find that being able to refer to images by a short simple number makes communication easier. Sure, we could use 6 digit time stamps, but I think it's easier to use smaller numbers.

Of course, this is simply my personal preference.

This sounds like an additional and -as you say- useful way to relate to a client from a photoshoot. It requires an agreement and understanding of that procedure in your images taken (before the photoshoot?). The question I have, is, is this common for professional photographers to do or do most clients just want images that they have in their mind catogorised differently? And is this sequence number an automatic feature in the EXIF data? Or where?

I use Photo Mechanic to ingest the images. For location shoots, it ingests from the card. For studio shoots, the camera sends every image to my computer via FTP, and Photo Mechanic ingests it from the FTP folder.

As part of the ingest, Photo Mechanic renames the images to YYYYMMDD-SEQN, and adds any metadata that is common to the whole shoot. Photo Mechanic keeps track of the next sequence number to be used. Photo Mechanic can also take a folder of existing images, sort them by capture time, and rename them to my standard.

Once I have the image, I can do a first pass through the images, and mark the ones the client shouldn't see (For instance shots of grey cards). Photo Mechanic then produces a website of watermarked images, which I upload for the client to peruse.

If I didn't want the client to know that some images had been removed. I can renumber the images I want to show the client.

For simple studio shoots (like a headshot), the client can immediately review the images with me using Photo Mechanic and choose the one they want.

I specifically use time instead of sequence because I'm often shooting with more than one camera and would like the images in chronological order regardless of which camera I took them with.

When I am shooting with multiple cameras I ingest all the files into Photo Mechanic, click on "sort by Capture time", and then rename them as YYYYMMDD-SEQN.

However, if your tools don't have that functionality, I can understand why you may prefer to name according to time rather than a sequence number.

.

I also take a photo of a clock with each camera. The clock needs to have a second hand. Often I use my smart watch or smart phone. Both of those are synced to atomic time, and will be correct. By comparing the capture time in the EXIF data to the time shown on the clock, I can tell how far off the camera's clock was.

I pull up all the images from each camera, and go to that camera's picture of the clock. If the capture time in the file doesn't match the time shown on the clock, I type in the time shown on the clock, and Photo Mechanic will compute the appropriate offset, and correct the capture time for those images.

If I do this for all the cameras, I know that the capture time in every file is within a second of matching my reference clock.

Once corrected, I have Photo Mechanic sort all of the images by capture time. I can be confident they are now in chronological order, even if the clocks in the various cameras were wildly wrong.

 

[grsnovi]

I don't see enough of a difference to have bothered reprocessing.

 

[Ryan Schwartz]

On my Mac, I have begun to organize files by date. In my Photos folder, I have a Year folder, then the photos are stored in subfolders by month and day. Any RAW photos I want made into JPEGs, I have a separate Export folder within the month/day folder.

I just let Lightroom take care of it when I import photos off the camera and tell it to use an Export folder for edited JPEGs.

 

[Guy Parsons]

....

I tried keeping files at home and office in synch and failed. Someone more diligent might be able to do it. Keeping a safe deposit box synced just ain't gonna happen. The friend thing could work with some sort of auto-synch. That could work for an office also, now that I think about it. Auto is the important part, so that you never are exposed to losing days of work.

Here's a strategy that works for some people:

Have a master copy and two backup copies. Let's call the copies "Blue" and "Green"

Put the Blue copy offsite. Have the Green copy locally connected with a background process that periodically updates the backup to match the master.

Every so often, take the local backup and swap it with the offsite backup. Your background process should be smart enough to update the previously offsite backup to match the master.

Perhaps you swap offsite backups weekly. The local backup protects you from a failure on the master copy. If your building burns down, you lose at most the changes since the last swap.

Well, that could work, so long as you're willing to potentially lose a week's worth of work. Keep in mind, what you're working on right now is the most inconvenient to lose.

You only lose a week worth of work if something takes out both your master copy and the local backup.

If my house burns down, losing a week's worth of changes would be the least of my concerns.

If the master copy fails, I lose only the changes since the last auto-sync to the local backup.

It's incredibly easy to keep my Backblaze backup current.

Houses do burn. Mine did. I had lots of worries, but I had no worries about my 8-TB of image files.

No burn yet but we do love in an officially designated bushfire zone. Multiple backups and one of them is an all computers backup which all photos and my wife's rather extensive family research all fit nicely on a 5TB USB HDD that is with us if we leave the house for any extended time. Just bought a second 5TB HDD so now will have two of those "grab and run" drives.

I know two people that lost their lives' work to fire and flood. Stuff happens. Oh, and HDs fail without fire and flood, for those that only have one copy.

The only failure I've had was a SATA SSD system drive in a laptop, luckily the 30 minute interval SyncBack software backup to USB HDD had all the good stuff, so new drive, reinstall Windows and needed software and then put relevant folders back. Nice clean and tidy again and nothing lost after about 2 hours total to rebuild.

As for HDD, lots of them all sizes used for many years and never has one has failed. One gave errors when mounted sideways so moved it to less important tasks and mounted horizontally and no more problems. Another started reporting errors in the NAS, let it keep messaging for about 2 years before I replaced it. Reformatted completely (long every sector test) from NAS format to NTFS and so far no errors in a PC.

From my experience I trust HDD way more than I trust SSD. The HDD do have a life of something like 3 to 5 years of operational time but if kept off or not spinning for a lot of their time then that lifetime can be many years indeed. I only power up computers when I need them, always shut down and off at the wall plug otherwise.

From my earlier life as a repair and maintenance tech on big room filling computers, the thing that makes a bad computer is heat. Keep them cool and check all fans are clean and working and keep an eye on internal temperatures and they last "forever".

 

[Michael Fryd]

....

I tried keeping files at home and office in synch and failed. Someone more diligent might be able to do it. Keeping a safe deposit box synced just ain't gonna happen. The friend thing could work with some sort of auto-synch. That could work for an office also, now that I think about it. Auto is the important part, so that you never are exposed to losing days of work.

Here's a strategy that works for some people:

Have a master copy and two backup copies. Let's call the copies "Blue" and "Green"

Put the Blue copy offsite. Have the Green copy locally connected with a background process that periodically updates the backup to match the master.

Every so often, take the local backup and swap it with the offsite backup. Your background process should be smart enough to update the previously offsite backup to match the master.

Perhaps you swap offsite backups weekly. The local backup protects you from a failure on the master copy. If your building burns down, you lose at most the changes since the last swap.

Well, that could work, so long as you're willing to potentially lose a week's worth of work. Keep in mind, what you're working on right now is the most inconvenient to lose.

You only lose a week worth of work if something takes out both your master copy and the local backup.

If my house burns down, losing a week's worth of changes would be the least of my concerns.

If the master copy fails, I lose only the changes since the last auto-sync to the local backup.

It's incredibly easy to keep my Backblaze backup current.

Houses do burn. Mine did. I had lots of worries, but I had no worries about my 8-TB of image files.

No burn yet but we do love in an officially designated bushfire zone. Multiple backups and one of them is an all computers backup which all photos and my wife's rather extensive family research all fit nicely on a 5TB USB HDD that is with us if we leave the house for any extended time. Just bought a second 5TB HDD so now will have two of those "grab and run" drives.

I know two people that lost their lives' work to fire and flood. Stuff happens. Oh, and HDs fail without fire and flood, for those that only have one copy.

The only failure I've had was a SATA SSD system drive in a laptop, luckily the 30 minute interval SyncBack software backup to USB HDD had all the good stuff, so new drive, reinstall Windows and needed software and then put relevant folders back. Nice clean and tidy again and nothing lost after about 2 hours total to rebuild.

As for HDD, lots of them all sizes used for many years and never has one has failed. One gave errors when mounted sideways so moved it to less important tasks and mounted horizontally and no more problems. Another started reporting errors in the NAS, let it keep messaging for about 2 years before I replaced it. Reformatted completely (long every sector test) from NAS format to NTFS and so far no errors in a PC.

From my experience I trust HDD way more than I trust SSD. The HDD do have a life of something like 3 to 5 years of operational time but if kept off or not spinning for a lot of their time then that lifetime can be many years indeed. I only power up computers when I need them, always shut down and off at the wall plug otherwise.

From my earlier life as a repair and maintenance tech on big room filling computers, the thing that makes a bad computer is heat. Keep them cool and check all fans are clean and working and keep an eye on internal temperatures and they last "forever".

I agree that HDD are more reliable than SSD.

If you have never experienced an HDD failure, then you are very lucky. Statistically, we expect that there should be some people who experience fewer than average failures, and some that see more.

However the fact that you have so far seen fewer than average failures does not mean that your luck will continue.

 

[Guy Parsons]

I agree that HDD are more reliable than SSD.

My SATA SSD died dead one day no warnings, nothing. It was just dead as a dodo when I tried to start the laptop. Less than a year old so a warranty replacement fixed that. So far the current NVMe M2 type SSD in a couple of (no, make that 3) laptops have caused no problems.

If you have never experienced an HDD failure, then you are very lucky. Statistically, we expect that there should be some people who experience fewer than average failures, and some that see more.

Never counted them but hard drives from maybe 1990 at first 40MB HDD to modern 8TB HDD in my collection, many others of many sizes in between, so they don't get used until they die but they get used until they are not big enough then the data gets moved to the next biggest ones. Most are backup copies of the maybe 4 or 5 computers that we might use in the house. Multiple backups of same data in many cases.

So drives never wear out, they just get sidelined because of limited size. They all still work when I play with them years later.

However the fact that you have so far seen fewer than average failures does not mean that your luck will continue.

Sure, I know that anything can happen at any time. That was my solid belief from my many years of mainframe computer experience.

As time moves on though, the gear gets more reliable and those disaster events come way less often, but they still may happen at any time so be prepared and have proper backups.

My usual care with cameras and computers does help prolong their life.

So far my luck has been good and hope it lasts more years to come. But if luck runs out then I am ready to deal with it.

 

[jjbinx]

Windows 11 PC

A single folder for new raw files - from the latest shoot(s) - images for review and culling and images for detailed processing in Lightroom. I synch this folder with Lr after copying new files to it. I use Windows File Explorer to copy files from memory card.

An internal HDD folder called Photo Archive with sub-folders for each year.

After processing in Lr I use Lr to tag the image files with Keywords and GPS data if I have it and then move the keeper images to the current Year folder.

I have several hundred keywords and use Lightroom’s relational database query facility to find and sort. Eg I can find/sort on date, camera, lens, location, and multiple keywords.

I have several external USB HDDs to hold a backup of all my raw files. As each fills I buy another and eventually get a bigger drive as needed. I also keep a backup of my Lr catalogue file.

Every few weeks I do a manual backup (Windows Explorer copy) of the current Year folder and Lr .cat.

Whenever I exit Lr I select Backup to make sure the current .cat file is copied to a second internal HDD.

I keep my ext HDDs de-powered and disconnected until I want to do a backup.
I keep them in a padded storage bag which I take to a (trusted) off-site location when we go on a vacation. I’m going to (soon) make a backup of the backups for permanent off-site storage.

Very much a manual process.

jj

 

[dcstep]

..

I know two people that lost their lives' work to fire and flood. Stuff happens. Oh, and HDs fail without fire and flood, for those that only have one copy.

The only failure I've had was a SATA SSD system drive in a laptop, luckily the 30 minute interval SyncBack software backup to USB HDD had all the good stuff, so new drive, reinstall Windows and needed software and then put relevant folders back. Nice clean and tidy again and nothing lost after about 2 hours total to rebuild.

As for HDD, lots of them all sizes used for many years and never has one has failed. One gave errors when mounted sideways so moved it to less important tasks and mounted horizontally and no more problems. Another started reporting errors in the NAS, let it keep messaging for about 2 years before I replaced it. Reformatted completely (long every sector test) from NAS format to NTFS and so far no errors in a PC.

From my experience I trust HDD way more than I trust SSD. The HDD do have a life of something like 3 to 5 years of operational time but if kept off or not spinning for a lot of their time then that lifetime can be many years indeed. I only power up computers when I need them, always shut down and off at the wall plug otherwise.

...

You might be interested in Backblaze's periodic reports of the HDD/SSD failures, by brand and model. It shows that ALL drives fail, just at different rates. SSD is actually more reliable, but they haven't been in BB's network as long. Send me a message if you can't find it and I'll send a link. I think it's public, but it might only be for subscribers.

 

[J A C S]

..

I know two people that lost their lives' work to fire and flood. Stuff happens. Oh, and HDs fail without fire and flood, for those that only have one copy.

The only failure I've had was a SATA SSD system drive in a laptop, luckily the 30 minute interval SyncBack software backup to USB HDD had all the good stuff, so new drive, reinstall Windows and needed software and then put relevant folders back. Nice clean and tidy again and nothing lost after about 2 hours total to rebuild.

As for HDD, lots of them all sizes used for many years and never has one has failed. One gave errors when mounted sideways so moved it to less important tasks and mounted horizontally and no more problems. Another started reporting errors in the NAS, let it keep messaging for about 2 years before I replaced it. Reformatted completely (long every sector test) from NAS format to NTFS and so far no errors in a PC.

From my experience I trust HDD way more than I trust SSD. The HDD do have a life of something like 3 to 5 years of operational time but if kept off or not spinning for a lot of their time then that lifetime can be many years indeed. I only power up computers when I need them, always shut down and off at the wall plug otherwise.

...

You might be interested in Backblaze's periodic reports of the HDD/SSD failures, by brand and model. It shows that ALL drives fail, just at different rates. SSD is actually more reliable, but they haven't been in BB's network as long. Send me a message if you can't find it and I'll send a link. I think it's public, but it might only be for subscribers.

Everything fails over time; we fail, too. In my use, the portable HDD I use are rarely powered, and those reports seems to be about frequently used drives. I never had a drive failing on me, and I have had personal computers and peripherals during the last 30+ years, at least 3-4 computers at any given time (office, personal, family members). I had two SD cards failing, however, a major brand.

That does not mean keeping one backup only but one does not need to be too paranoid.

 

[tbcass]

dcstep wrote:...

You might be interested in Backblaze's periodic reports of the HDD/SSD failures, by brand and model. It shows that ALL drives fail, just at different rates. SSD is actually more reliable, but they haven't been in BB's network as long. Send me a message if you can't find it and I'll send a link. I think it's public, but it might only be for subscribers.

Everything fails over time; we fail, too. In my use, the portable HDD I use are rarely powered, and those reports seems to be about frequently used drives. I never had a drive failing on me, and I have had personal computers and peripherals during the last 30+ years, at least 3-4 computers at any given time (office, personal, family members). I had two SD cards failing, however, a major brand.

That does not mean keeping one backup only but one does not need to be too paranoid.

I had a hard drive fail on me one time. It was a long time ago, maybe 20-25 years, and it was soon after I bought the computer so it was apparently faulty from the factory. It was replaced at no cost.

 

[Guy Parsons]

dcstep wrote:...

You might be interested in Backblaze's periodic reports of the HDD/SSD failures, by brand and model. It shows that ALL drives fail, just at different rates. SSD is actually more reliable, but they haven't been in BB's network as long. Send me a message if you can't find it and I'll send a link. I think it's public, but it might only be for subscribers.

Everything fails over time; we fail, too. In my use, the portable HDD I use are rarely powered, and those reports seems to be about frequently used drives. I never had a drive failing on me, and I have had personal computers and peripherals during the last 30+ years, at least 3-4 computers at any given time (office, personal, family members). I had two SD cards failing, however, a major brand.

That does not mean keeping one backup only but one does not need to be too paranoid.

I had a hard drive fail on me one time. It was a long time ago, maybe 20-25 years, and it was soon after I bought the computer so it was apparently faulty from the factory. It was replaced at no cost.

My only HDD issue was many years back. Bought a 3.5" drive, got it home and plugged it in and no noise, it would not rotate. So unplugged it, held it horizontally in my hand and then gave it a quick wrist flick to cause the HDD to flick CW then CCW to unstick the disk. Plugged in again and it then spun up and gave no problems at all for the next number of years that I used it.

My HDD spend more time resting with power off than they do working so they last "forever" as far as I am concerned.

...............

I have seen the Backblaze statistics on failures and was aware of which brands seemed to be better/worse than others, but they use drives in a very different way to me, so interesting but does not affect me or what I buy.

 

[Michael Fryd]

dcstep wrote:...

...

My only HDD issue was many years back. Bought a 3.5" drive, got it home and plugged it in and no noise, it would not rotate. So unplugged it, held it horizontally in my hand and then gave it a quick wrist flick to cause the HDD to flick CW then CCW to unstick the disk. Plugged in again and it then spun up and gave no problems at all for the next number of years that I used it.

My HDD spend more time resting with power off than they do working so they last "forever" as far as I am concerned.

...............

I have seen the Backblaze statistics on failures and was aware of which brands seemed to be better/worse than others, but they use drives in a very different way to me, so interesting but does not affect me or what I buy.

Your position seems to be that your usage pattern should result in a lower usage rate than Backblaze sees. You may be correct. I suspect that some parts of the drive degrade with usage, and some with time.

I remember a time where a large manufacturer had made a bad choice of drive lubrication. If a drive was left powered down, the lubrication tended to harden. Leave th drive off for too long, and the drive's motor might not be able to get the drive spinning again.

But let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan. If your archive is large enough to span three drives, there's a 15% (about 1 in 6) chance that you will one of those three drives fail during a ten year period.

The question is how valuable is your image archive. Without backups, you have a 1 in 6 chance that during the next decade you will lose some of your images.

If your drive failure late is only 0.25% per year, you have only a 1 in 12 chance of losing a drive. If your failure rate is closer to 1.5% per year, then you have a 45% chance that one of your three drives will fail in the next decade.

Obviously, your personal odds of failure may be better or worse. Do you keep your drives in environmentally controlled storage? Are your drives filled with helium rather than air (helium is better at escaping)? Are your drives subject to vibration during storage?

Many people will never see a drive failure. Most people will. It can be dangerous to look at the experience of only a few people and assume that's the general case.

 

[J A C S]

Putting my pedantic hat on:

But let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan.

Close enough but it is actually 1-(1-0.005)^10 = 4.88...%

If your archive is large enough to span three drives, there's a 15% (about 1 in 6) chance that you will one of those three drives fail during a ten year period.

1-(1-0.005)^30 = 13.96....%

(assuming you meant "at least one"). You count the chance of survival in each case, and those are independent events.

The question is how valuable is your image archive. Without backups, you have a 1 in 6 chance that during the next decade you will lose some of your images.

If your drive failure late is only 0.25% per year, you have only a 1 in 12 chance of losing a drive. If your failure rate is closer to 1.5% per year, then you have a 45% chance that one of your three drives will fail in the next decade.

More like 1-(1-0.015)^30 = 36.45...%.

 

[Guy Parsons]

dcstep wrote:...

...

My only HDD issue was many years back. Bought a 3.5" drive, got it home and plugged it in and no noise, it would not rotate. So unplugged it, held it horizontally in my hand and then gave it a quick wrist flick to cause the HDD to flick CW then CCW to unstick the disk. Plugged in again and it then spun up and gave no problems at all for the next number of years that I used it.

My HDD spend more time resting with power off than they do working so they last "forever" as far as I am concerned.

...............

I have seen the Backblaze statistics on failures and was aware of which brands seemed to be better/worse than others, but they use drives in a very different way to me, so interesting but does not affect me or what I buy.

Your position seems to be that your usage pattern should result in a lower usage rate than Backblaze sees. You may be correct. I suspect that some parts of the drive degrade with usage, and some with time.

I remember a time where a large manufacturer had made a bad choice of drive lubrication. If a drive was left powered down, the lubrication tended to harden. Leave th drive off for too long, and the drive's motor might not be able to get the drive spinning again.

But let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan. If your archive is large enough to span three drives, there's a 15% (about 1 in 6) chance that you will one of those three drives fail during a ten year period.

The question is how valuable is your image archive. Without backups, you have a 1 in 6 chance that during the next decade you will lose some of your images.

If your drive failure late is only 0.25% per year, you have only a 1 in 12 chance of losing a drive. If your failure rate is closer to 1.5% per year, then you have a 45% chance that one of your three drives will fail in the next decade.

Obviously, your personal odds of failure may be better or worse. Do you keep your drives in environmentally controlled storage? Are your drives filled with helium rather than air (helium is better at escaping)? Are your drives subject to vibration during storage?

Many people will never see a drive failure. Most people will. It can be dangerous to look at the experience of only a few people and assume that's the general case.

In my case each computer has a locally attached USB HDD (may be permanently plugged in or only occasionally plugged in, it varies computer to computer. That drive handles the regular 30 minute SyncBack backup updates.

Friday and Saturday the NAS is automatically turned on and used for hourly backing up all computers that happen to be on at the time. The NAS is my ultimate backup store of all data, many old versions of downloads etc. whereas the USB HDD are tidier versions of backups.

More often than once a week the 5TB "grab and run" drive visits all computers and gets updated.

The collections tend to be tidied up on the NAS then it is backed up to its 8TB HDD on maybe a monthly basis.

When some collection of images or raw files is deemed tidy enough it then gets backed up to a set of 3 separate HDD each for jpeg and for raw files.

It's all on HDD and nothing that I want to keep is on any cloud service.

In all that I've never had a HDD failure (only one SSD failure) and the only data disasters are of my own causing.

My backup paranoia stems from my years as a main-frame computer tech, saw too many distasters that were made worse by faulty or incompetent backup methods.

Plus of course as collections expand the backups move to bigger drives and the wear starts afresh on new drives. The old drives become "project" drives where they get used and not clutter up one of the computers.

 

[Carey Brown]

Putting my pedantic hat on:

But let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan.

Close enough but it is actually 1-(1-0.005)^10 = 4.88...%

So, "^" apparently doesn't mean "raise to the power of".

What does it mean?

 

[tbcass]

I don't see enough of a difference to have bothered reprocessing.

I find the difference quite noticeable. It took less than a minute so for me the improvement was well worth while. I will add I enjoy RAW editing.

--
Tom

 

[J A C S]

Putting my pedantic hat on:

But let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan.

Close enough but it is actually 1-(1-0.005)^10 = 4.88...%

So, "^" apparently doesn't mean "raise to the power of".

It does. You compute the probability of one drive surviving ten years, it is 0.0995 multiplied by itself ten times.

If you copy and paste the whole expression in google, it will recognize and compute it.

BTW, assuming a failure rate of 0.5% per year, if you use two drives to back up the same data, the chances of both failing (and you notice that at the end of the year, no sooner), is 0.0025%.

 

[Michael Fryd]

Putting my pedantic hat on:

But let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan.

Close enough but it is actually 1-(1-0.005)^10 = 4.88...%

If your archive is large enough to span three drives, there's a 15% (about 1 in 6) chance that you will one of those three drives fail during a ten year period.

1-(1-0.005)^30 = 13.96....%

(assuming you meant "at least one"). You count the chance of survival in each case, and those are independent events.

The question is how valuable is your image archive. Without backups, you have a 1 in 6 chance that during the next decade you will lose some of your images.

If your drive failure late is only 0.25% per year, you have only a 1 in 12 chance of losing a drive. If your failure rate is closer to 1.5% per year, then you have a 45% chance that one of your three drives will fail in the next decade.

More like 1-(1-0.015)^30 = 36.45...%.

Thank you for the corrections.

.

As you pointed out in another post, if we assume that the drive failures are completely independent, then having a second copy dramatically decreases the chance of losing data.

Of course, there are some hazards that can cause the loss of drive data that are not independent. For instance, someone can break into your home and take your computer, along with any nearby drives. A house fire, roof leak, power surge, flood, etc., are all examples of hazards that can affect multiple drives.

Perhaps the annual chance of your house flooding and taking out a hard drive stored in your basement is 1%. If you store ten hard drives in your basement, you still have a 1% chance that a flood will take out all your basement drives.

Correlated failures are one of the reasons I strongly suggest that one doesn't keep all copies of your data in the same RAID enclosure. There are hazards that can take out all the drives in the enclosure.

My goal is to minimize the likelihood that a single hazard can take out all copies of my data. That's one of the reasons I keep some copies offsite, and some copies on different types of media (you can find writeable BluRay discs that hold 100GB of data). When using the same types of media, I like to use media from different manufacturers.

But then I tend to be a little crazy about this.

 

[Cherry_F]

[../..] let's assume that your drives have a 0.5% of a drive failure each year. That's a 5% chance a drive will fail over a ten year lifespan. If your archive is large enough to span three drives, there's a 15% (about 1 in 6) chance that you will one of those three drives fail during a ten year period.

The question is how valuable is your image archive.

Absolutely!

I have two backups and my life expectancy makes me believe at least one of them will stay operational before I leave this world.

In addition, I'm confident my images won't be valuable to everyone else.

___
Photography is so easy, that's what makes it highly difficult - Robert Delpire