Can renaming multiple times degrade a photo?

[Durant Philosopher]

Is there any chance that I might lose photo quality by renaming photos several different times (JPGs, NEFs or TIFFs) or by adding more keywords (using Adobe Bridge or any other browser) with either individual or batch renaming - without making any edits or other changes to the photo?

Would renaming several times affect the photo's actual appearance in any way? With Adobe Bridge, you don't actually "save" your photos that I am aware of, but perhaps clicking on the "Rename" button is actually the way the changes are saved. Anyway, I know that there is data loss if you save JPGs more than once, so I am just wondering if renaming would also cause data loss and thereby degrade photo quality. Also, what about renaming from the Finder on the Mac? Same question.

I have noticed when I rename RAW NEF files (using Adobe Bridge) that it creates a sidecar, so I'm wondering how the name change is stored for a JPG. Thanks, Linda

[arith]

That is the first time I've heard you lose quality by saving an image; but I thought they were moved around all the time by the computer.

I suppose it could be possible, so you could archive a copy and work on another.

[Colin Southern]

Is there any chance that I might lose photo quality by renaming photos several different times

No.

Would renaming several times affect the photo's actual appearance in any way?

No.

Anyway, I know that there is data loss if you save JPGs more than once

This is something that I hear all the time, but I've never seen any real-world tests to prove it (and I sure would like to). When you save an image in a lossy format like jpeg the first time, a lot of information is "discarded" - in practice these are levels too low for the eye to be able to see (inc a comfortable safety margin), and levels so close together that the eye can't tell them apart (so they're changed to the same value so that they compress better).

Traditional thinking is if you open an image that has been "optimised" for jpeg storage - make some changes - and then resave it then you lose more information, but thinking about it logically, it's going to depend a lot on just what changes were made; if the values that are about to be compressed are already "normalised" for a jpeg then there shouldn't be any further degradation.

The far bigger issue is in trying to make heavy edits to a file that's already saved in a jpeg format (eg trying to lift shadow areas); the problem being that the information that your trying to display just isn't there anymore.

Best practice is to save your images in a non-lossy format (like PSD or TIFF) until you've finished all of your editing and then (and only then) save them as a jpeg if your hard up for storage space or you need to display them online. My suggestion - because storage is cheap these days - is to always keep your originals in some form then you can always regenerate an updated jpeg whenever required.

Hope this helps

[Bill44]

Read an experiment a while back about degradation of JPEG's with saving. From memory it seemed that degradation became noticeable after something like 150-200 saves, but who does that.

[Colin Southern]

From memory it seemed that degradation became noticeable after something like 150-200 saves, but who does that.

Apparantly, the guy doing the test

[wjh31]

If you are simply re-naming it from the OS then i cant see it would make any change to the image itself. If you are opening it in GIMP/PS/other and re-saving it it would be concievable that a re-optimisation might loose data, especially if you cahnge the compression settings on a JPG.

The easy way to check is to rename a file while keeping a copy of the original. Open both files as layers and use difference blending. This will show if there is any difference between them.

[Durant Philosopher]

Thanks Colin. That was a really great explanation of how JPG compression works. So, just to be clear on this, if I am not mistaken, you are saying that just re-saving a JPG (without having made any actual changes to the photo itself (pixels)) would not cause loss of data regardless of how many times you re-save, right? And even if you make changes to the photo, the loss of data during compression would mostly be from actual data that you intended to discard, i.e., when you crop, of course you are throwing away data.

Although I do save my original photos to TIFF before editing them, I like to rename them first. Just wanted to make sure there was no problem with this. Hate the generic camera names.

[Durant Philosopher]

The easy way to check is to rename a file while keeping a copy of the original. Open both files as layers and use difference blending. This will show if there is any difference between them.

Good suggestion, WJH. I may try that, although I'll probably just take Collin's word for it.

Thanks to everyone! Linda

[Dave Humphries]

Hi Linda,

If you are simply re-naming it from the OS then i cant see it would make any change to the image itself.

I agree with Will.

The only way you might 'accidentally' change the image directly from the OS is if you have auto-rotate images switched on (I think its an option somewhere) and it attempts to rotate the image 90 degrees (to portrait orientation) AND the image was saved in a pixel sizes in which either height or width is not integer divisible by 8. But this is getting very technical, and anyway, in my experience, it does warn you in these instances (Vista does).

Beyond that, as Colin says, even within image editors, it does depend what you change.
And even then, it is my belief that repeated saves of an image kept open within the application will not degrade it, e.g. saves as you do more and more things to it.

Damage happens only if you: edit it, save it, close the image/program, then spot a problem, re-open it, edit some more, then re-save, etc. I bet if you did this and compared as Will suggests you'd see some damage, particularly if the second set of changes you made affected the whole image, like a small rotate to straighten a horizon, say. The lower the jpg quality used, the more pronounced I'd expect this to be.

For the record I save at quality 9 (on the scale of 1 - 12) and I do occasionally re-open a jpg and edit and re-save, but no-one has complained so far.

Hope that wasn't too boring,

I also agree with Colin that '8 bit' damage from intensive levels tweaking could be more noticeable and a lossless format is better to save in if you think you may want to re-open and edit further. This pre-supposes you have shot in RAW of course.

[Durant Philosopher]

This was very helpful, Dave, and definitely NOT boring!

Linda

[Colin Southern]

Thanks Colin. That was a really great explanation of how JPG compression works.

Your very welcome Linda

So, just to be clear on this, if I am not mistaken, you are saying that just re-saving a JPG (without having made any actual changes to the photo itself (pixels)) would not cause loss of data regardless of how many times you re-save, right?

I wouldn't expect any "real world" degradation. I'm using the term carefully because I can't rule out perhaps slight rounding errors or the like, but if it were only done a few times then I'd be extremely confident that you wouldn't see any difference (unless you were instructing the program to degrade it each time by selecting a low-quality setting etc).

It's probably a good time to take a step back and just quickly review why we have different file formats ...

- The likes of PSD and TIFF files save image data in a non-lossy format - and potentially in 16 bits (exclusively in my case) which means that the files have the very best chance of surviving any heavy editing. Unfortunately, having all of this "virginal" data available on-cue has a downside; the files are very large (all over 100MB, most over 200MB, some are even over 1GB!). Technically there's no reason why I couldn't put some of those into posts here, but of course we don't yet have gigabit connections to the internet - and thus the sheer size would instantly kill our connections (although give you some great "master copies" of my work!).

- So we have the likes of JPEGs - they do their best to chop out the bits that our eyes can't use (or differentiate between) and squish what remains down into the smallest possible file size so that it transports easily, but at the expense of not being able to do anywhere nearly as much with it later.

A JPEG is a bit like having a can of paint - it's an end result. A PSD or a TIFF is a bit like having a can of paint and the entire factory that made it!

And even if you make changes to the photo, the loss of data during compression would mostly be from actual data that you intended to discard, i.e., when you crop, of course you are throwing away data.

I'm not quite sure what your asking here, but just to expand a little ...

- if you crop an image then the bits you cropped off are discarded (with the resulting size saving).

- the human eye is less sensitive to some colours (it's relatively insensitive to yellow for example) so if you have a photo of a flower and you have pixels with values of 123, 122, 124, 122, 123, 121, 125 a program saving to JPEG might conclude "Hmm - these values are very close together - the human eye can't tell them apart - so lets replace them with 123, 123, 123, 123, 123, 123, 123 - and of course when the compression part of the process kicks in it can essentially store that as 6x{123} rather than 1x {123, 122, 124, 122, 123, 121, 125}. Somewhat over-simplified, but hopefully you get the idea.

- the human eye also has limits on what it can resolve from shadow detail - so if we can't see detail that represents several stops worth then (allowing for an appropriate safety margin) (to allow a little recovery/adjustment, or even differences in monitor calibration) then it may as well not take up space (a bit like a wardrobe filled with things that you'll never use again!).

So my theory is that if you open a jpeg image (ie an image that's already been optimised for size) then compressed structures will losslessly returned to their pre-compressed state (keeping in mind that the loss occurs when the image was "normalised", not when it's compressed; (a subtle but important difference) leaving a bigger internal representation of the file - but - one that still has the "normalised" data; eg pixels of the yellow flower will still be 123, 123, 123, 123, 123, 123, 123. So - if you now go and take the red eye out of the eyes of the person standing in front of the flower garden, when you save the image the yellow pixels are STILL optimised (still 123 etc) thus when the "normalisation" pass examines them it WON'T CHANGE THEM ANY FURTHER and they'll compress just the same - so NO FURTHER DEGRADATION.

On the other hand though - if you make some kind of global correction (say a levels change) then that will change ALL of those values by the same amount (eg 7x {123} might become 7x {137} - changing the data, but all by the same amount - which will all compress (and uncompress) the same way. So no degradation either.

On my 3rd hand though (I have a lot of hands) If you applied, say, a graduated filter across the image then some of those consistant values might then be changed to be all different - with no guarantee that there wouldn't be degradation when they were "re-normalised" again as part of the saving process.

All very technical stuff (and I'm theorising a lot here) (based largely on my knowledge of DVD formats which have the same jpeg roots).

In practice though it's a lot simpler (thank goodness): If you need it to be small, save it as a jpeg. If you need to do (potentially) heavy edits on it, save it as a PSD / TIFF If you need it to be small and still capable of withstanding heavy edits then your out of luck!

Personally - when shooting landscape - I take a number of shots - pick the best one - save all of the "unsuccessful applicants" in a "Bulk DNG" folder for the image - save "The Chosen One" as a PSD - and then save a down-sampled version as a jpeg for my website (pbase.com/cjsouthern <- shameless plug!). That suits my style - if your style is one where most of your images are keepers (eg wedding photography) then I'd suggest (at a minimum) keep your RAW files for ever more (thay're losslessly compressed and still mosaiced and so are "relatively" small (about 25MB each for me) - however you'd have to be pretty keen to keep all of these in a TIFF / PSD format as well as a jpeg format (keeping in mind though that the likes of wedding photos aren't all "individually hand-processed" as you would a feature landscape, so easier not to retain the large intermediate files; just redo an image from the RAW original if/when required).

[Durant Philosopher]

Nice explanation again, Collin. I definitely understand the mechanics of JPG compression better now. I like working with a non-destructive image editing program (like i-Photo or Lightroom), so I don't have to worry about all of this as much. Problem is you basically lose the benefits of non-destructive editing once you convert your RAW file to TIFF. Oh sure, you can always go back to the original RAW file and start over, but meanwhile you have lost all of the other changes you have made in another program. But then again, I guess this doesn't happen very often. I just wish there was more cross-compatibility between programs, but I guess that will never happen unless we do finally accept some universal RAW format, such as DNG, but will probably never happen. Life was so much simpler before I switched to the RAW format. Now there are too many decisions to be made. Guess I'll figure out a realistic digital workflow one of these days.

So, do you still keep your RAW files for your landscape "Chosen Ones" that you convert to PSD or does the PSD replace your original RAW file?

Thanks,

Linda

[Durant Philosopher]

That's interesting, Bill44, because I had always heard that you begin to notice quality loss after only the 2nd or 3rd save. Guess I'll have to test it someday. Linda

[Durant Philosopher]

Sorry, folks. I tried to delete the above duplicate but couldn't get it to go away. Linda

[Colin Southern]

Sorry, folks. I tried to delete the above duplicate but couldn't get it to go away. Linda

No worries Linda - my wand is bigger than your wand, so I zapped it for you

[Colin Southern]

So, do you still keep your RAW files for your landscape "Chosen Ones" that you convert to PSD or does the PSD replace your original RAW file?

Hi Linda,

I use Adobe's DNG converter to pull the images off the card - rename them - and convert them to DNG all in 1 step. The DNGs are kept forever (with the exception of obviously hopelessly unsharp images / misfires etc. There is an option in the DNG converter to embed the original (in my case *.CR2 image), but I don't use it. My philosophy could bite me in the bum if I ever wanted to use something like DxO optics (that won't support DNG as an input file), but I'm not overly concerned about that as I'm pretty much a 100% Adobe shop. I like DNG as the ACR adjustments are stored directly in the DNG file rather than in an intenal database (thay would never get backed up) or in a sidecar file (which would double the number of files).

I don't use *.TIFF at all - PSDs offer the same functionality and are smaller.

With CS3 or CS4 you can also open a RAW file as a smart object (thus retaining the ability to re-edit the RAW and have the changes filter through to your PSD file, but it's not something I currently use; it's very important to get most things looking as close as you can to ideal in ACR before opening in Photoshop "proper" (far less lossy for most things) - but once it's open in Photoshop, relatively minor (or even medium corrections) won't degrade the data to any noticeable degree ("noticeable" being the key word; data loss when image processing is a normal and necessary event).

[Durant Philosopher]

No worries Linda - my wand is bigger than your wand, so I zapped it for you

That's quite funny, Collin. I like your sense of humor. Also checked out your photo gallery yesterday and loved your pictures. The landscapes were gorgeous, and the children's photos were quite nice too. Loved the one with the little girl holding a kitten. I don't normally crop out a portion of anyone's face myself, but it gave quite a nice effect in this case.

[Durant Philosopher]

Hi Collin,

There is an option in the DNG converter to embed the original (in my case *.CR2 image)

I'm not sure I understand how this works. What would be the advantage in doing this?

I will probably convert to DNG at the time of import as well, if I am able to rule out the possibility that I might achieve better raw conversion with Capture NX2. Still need to experiment a bit more. I hate to give up this program after just recently making the purchase, but I also don't really like having to add it to my workflow. Ultimately, I may decide to use it only in very special cases, when I can't seem to achieve the results I am looking for any other way.

Do you know if PS Elements has the same Smart Object feature you mentioned in CS4? Will have to research this.

I don't like the sidecars either, and Bridge creates them when I rename photos. I've noticed that Nikon's ViewNX doesn't use them, so I guess changes are stored directly to the NEF file the same way they are with DNG files. Other programs don't recognize adjustments that are store in the NEF files, but apparently renaming must be stored differently because I opened the renamed files in several different programs and the new name was recognized. So much to figure out still! Uggh!

Thanks for the info,

Linda

[Colin Southern]

Hi Collin,

I'm not sure I understand how this works. What would be the advantage in doing this?

It's just a way to retain the original RAW file without it needing to be kept as a seperate file. A good use for this (if I used it) would have been the ability the extract the original RAW file so that I could have used it in association with the DxO program, whereas as it stands, if I wanted to use DxO I'm out of luck because DxO (for some strange reason) won't accept DNG files, and I ONLY have DNG files. The term "mutually exclusive" comes to mind!

I will probably convert to DNG at the time of import as well, if I am able to rule out the possibility that I might achieve better raw conversion with Capture NX2.

The advantage of using the DNG converter is that you can use it to transfer the files from your card reader - rename them in the process (using some very powerful naming options - very handy) - and converting them to a standardised fomat all in one operation.

The other advantage is (as already mentioned) the elimination of the sidecar files; Adobe do it this way because with a native RAW file they couldn't guarantee that they could write back to the file without corrupting something, since manufacturers don't publish the structures of their files (and they often contain "secret sauce" fields) (so they have to either put the changes into a seperate file, or integrate them into their central database), whereas with DNG they understand the entire structure and can write to it safely (and still import the encrypted "secret sauce" data).

NX2 is positioned as an advanced amateur type package, and whereas I'm sure it's more than adequate for most things, it's not going to compete head on with a top-of-the-line converter like ACR, if all of ACR's controls are made available. The reason I mention the last bit is that PS and PSE both use the same ACR engine, but I understand (from Dave) that when using it with PSE, you only get access to the first 2 tabs.

Do you know if PS Elements has the same Smart Object feature you mentioned in CS4? Will have to research this.

Not sure sorry - Dave might know.

I don't like the sidecars either, and Bridge creates them when I rename photos. I've noticed that Nikon's ViewNX doesn't use them, so I guess changes are stored directly to the NEF file the same way they are with DNG files. Other programs don't recognize adjustments that are store in the NEF files, but apparently renaming must be stored differently because I opened the renamed files in several different programs and the new name was recognized. So much to figure out still! Uggh!

RAW files retain an "internal file name" so that ACR can keep the correct sidecar file associated with it if the original RAW file is renamed.

[Dave Humphries]

Hi Linda, Colin,

I don't know if Elements 8 has Smart Objects, I've just been looking at the website and it isn't mentioned, so very doubtful, I'd say

There are a few things there that may have me upgrading from PSE6 though;
16 bit support (as far as I can tell in my PSE6, it only works for whole image tweaks, as soon as you select a brush applied tool, it wants to convert bit depth to 8 )
Windows 7 supported
Better photo organising is claimed

It's not at all obvious if you look at the feature set pages on Adobe website, where it says it is "US only", but the Adobe UK store does list it for delivery or download*, so I assume it is available now.
(I'm looking at Windows Upgrade version, but I doubt the Mac version will be different as they were released simultaneously in the US - this time)

Bizarrely, when you select the Download option, the price goes UP by £3!
Further investigation reveals this is probably because the Shipping option will add another (minimum) of £5.11 to cover shipping, depending where you are. Oh, and don't forget to allow adding VAT to them all!

Cheers,