Save RAW as PSD, Tiff, Other?

chuckinsocal · October 29, 2008

Hi Gang,

The title pretty much speaks for itself. After tweaking my NEF file to my liking I've been saving it as a PSD (lossless?) until I'm ready to Save As a JPG for uploading.

Just wondering if there are any advantages to saving it as a Tiff, PNG or some other lossless format.

Your thoughts are appreciated. Thanks.

jfriend · October 29, 2008

chuckinsocal wrote:

Hi Gang,

The title pretty much speaks for itself. After tweaking my NEF file to my liking I've been saving it as a PSD (lossless?) until I'm ready to Save As a JPG for uploading.

Just wondering if there are any advantages to saving it as a Tiff, PNG or some other lossless format.

Your thoughts are appreciated. Thanks.

PSD is a proprietary Adobe format. Many other programs can read it, but you can't always rely on that. PSD files can sometimes be quite large. PSD files support all state in Photoshop (layers, layer sets, hide state, 16 bits, etc...).

TIFF is a standard format and as long as you don't use odd compression settings, it is widely support by lots of programs. It is possible to use compression with TIFF files to reduce their size, but it still isn't a particularly efficient file format.

PNG is a standard format with lossless compression. It is more widely supported today than it was 3-5 years ago, but not as widely supported as TIFF. Of the three, PNG is probably the smallest.

The safest (most interoperable) solution would be TIFF. The smallest solution would be PNG. Nowadays, I don't know of any reason to use PSD over TIFF.

pyry · October 29, 2008

How about DNG? It's an open specification (by Adobe and they've submitted it to ISO for standardization) for a losslessly compressed raw image file with metadata and conversion parameters stored in the file.

The point is you would still have the raw data intact in a non-proprietary format.

I have all of my raw images in dng, as long as the only modifications and tweaks are at the raw converter level - and therefore storable in the dng file. For images with more in depth work in Photoshop including adjustment layers and stuff like that, I save in psd.

chuckinsocal · October 30, 2008

Thanks guys for your comments.

I guess what I was wondering about is picture quality by not losing any pixels.

You guys didn't mention picture quality so I'm assuming all the formats are about the same for retaining pixels and picture quality?

jfriend · October 30, 2008

chuckinsocal wrote:

Thanks guys for your comments.

I guess what I was wondering about is picture quality by not losing any pixels.

You guys didn't mention picture quality so I'm assuming all the formats are about the same for retaining pixels and picture quality?

All three are lossless (either no compression or lossless compression) so there is no change in image quality. You can save and get back the exact same bits you had.

pyry · October 30, 2008

jfriend wrote:

All three are lossless (either no compression or lossless compression) so there is no change in image quality. You can save and get back the exact same bits you had.

True as long as bits per pixel is preserved, this means 16-bits in psd, tiff and others, dng keeps the 10, 12 or 14 from the camera as is (and is only format that isn't de-bayered apart from the original).

rockcanyonphotos · October 30, 2008

Just another point of view and I suppose it depends on your workflow/business model. I shoot almost exclusively sports and....

I have yet to see the value in storing my RAWs in any other lossless format once I have processed them. I don't see any PP software not supporting camera vendors proprietary RAW formats so I can't justify the xtra storage requirements to store yet another uncompressed file. Plus with RAW all my edits are in overlays so I can always go back to them to modify what I have done for different outputs. The only output I use from RAW is JPEG and that is when I am ready to print or upload.

regards, Kevin

pyry · October 30, 2008

rockcanyonphotos wrote:

I don't see any PP software not supporting camera vendors proprietary RAW formats so I can't justify the xtra storage requirements to store yet another uncompressed file.

I've set up Lightroom to do conversion to DNG on the fly as the pictures are downloaded from the card - that way there's no wasted space (the dng files are actually a little bit smaller than cr2).

The main reason I do that is the internal handling of xmp, no databases or sidecar files to worry about.

chuckinsocal · October 30, 2008

Thanks again John and Pyry. I think I've got it now.

Edit: OOps! I just saw two more responses for the first time.

RockCanyon brings up another point I've been wondering about. Is it really necessary to save the file in an intermediate format and have two uncompressed files taking up hard drive space?

My workflow has been to open the RAW file, tweak it in the PS RAW converter, then open again in PS for final adjustments, then save as a PSD until I'm ready to upload. Would I be just as well off saving the file back as an NES then saving as a JPG when I'm ready to upload?

Art Scott · October 30, 2008

chuckinsocal wrote:

Thanks again John and Pyry. I think I've got it now.

Edit: OOps! I just saw two more responses for the first time.

RockCanyon brings up another point I've been wondering about. Is it really necessary to save the file in an intermediate format and have two uncompressed files taking up hard drive space?

My workflow has been to open the RAW file, tweak it in the PS RAW converter, then open again in PS for final adjustments, then save as a PSD until I'm ready to upload. Would I be just as well off saving the file back as an NES then saving as a JPG when I'm ready to upload?

I do pretty much the same except I save as a DNG instead of a PSD....since DNG is open source and at least one camera out there has that as an option as a capture file format.....and I am hoping, but not holding my breath, that others in the near future will see the advantage of it as opposed to all of the propriatary raws we have shoved own our throats...........................

BradfordBenn · November 1, 2008

I really like the DNG option and I have also gone with the convert to DNG on import into Lightroom.

chuckinsocal · November 2, 2008

HI Guys,

I really appreciate all your responses. I've been experimenting for the last few days and here's what I've found (keep in mind that I am a total noobie in working with RAW files):

I have Lightroom 2 and when I use it to import my pics I have them converted to DNG on the fly. No problem there. I'm sure Lightroom is not that hard but, like many Adobe programs, it seems counter intuitive and the learning curve has been steep, but I am getting better at it. As near as I can tell, Lightroom does not have a Save function so am I correct in assuming that I would just close out the DNG file when done tweaking and it will retain all it's corrections next time I open it?

After tweaking in any RAW converter, including Lightroom, I usually like to open in PS for a larger view and do some final adjustments. Some of the tools seem to work better in PS than they do in the RAW converter, Shadows and Highlights for instance. However, PS does not have a Save As option for DNG. It has an option for RAW however when I choose it I get some scary warning that I may lose some changes upon reopening the file (or something like that). So, I Save As a PSD. When doing so, I know by the extension that the file has been processed as opposed to the NEFs and DNGs I haven't gotten to yet or maybe aren't going to "make the cut".

I'm probably making this harder than it needs to be, but I'd love to have a standard workflow that is effective and that I can have confidence in so that I can quit experimenting and do some real work.

As always, your comments are appreciated. Thanks.

pyry · November 2, 2008

chuckinsocal wrote:

I have Lightroom 2 and when I use it to import my pics I have them converted to DNG on the fly. No problem there. I'm sure Lightroom is not that hard but, like many Adobe programs, it seems counter intuitive and the learning curve has been steep, but I am getting better at it. As near as I can tell, Lightroom does not have a Save function so am I correct in assuming that I would just close out the DNG file when done tweaking and it will retain all it's corrections next time I open it?

As long as changes to metadata are saved (this your save button for dng files). You can do that manually or set Lightroom do it automatically in the preferences.

OffTopic · November 2, 2008

Yup, saving your RAW file and saving your edited file are two completely different things. For starters, most edits can't be saved back to a DNG or RAW file and that is why you're getting the warning message when you try to do it.

Just think of your RAW file as your negative. I add my metadata to it and nothing else. My latest conversion settings are reflected in the thumbnail preview, but the file itself isn't changed (other than converting to DNG).

What you're doing now is similar to my current workflow that I’ve developed after years of learning. I can't tell you how I often I want to kick myself for my earlier workflow methods...I had some very poor habits in the beginning which were destructive in nature, and as I learned more about digital processing and RAW in particular, I was really mad that I couldn't go back and undo some of what I had done. Most of this I’ve developed after reading many of Bruce Fraser’s outstanding books in his “Real World” series. I think your optimal workflow really needs to take into account the type of work you shoot. I have to admit that if I were doing a lot of event shooting and processing a high volume of photos, I would probably have a different workflow, but I also probably wouldn’t be as concerned about the long-term implications of a non-destructive workflow.

Here is my somewhat anal-retentive workflow:

I open the file in ACR and do my converting, then click right through to open the file as a smart object in Photoshop. I try to make sure that all my editing is non-destructive, which means I use a lot of layers. Everything I do is done on its own separate layer, which makes it very easy to change, mask, turn on/off, or delete without worrying about how many history states I have. Even something as simple as dust spotting is done on a separate layer. That completed new "master" version is saved as a layered PSD (with Maximize Compatibility turned on so I can view the previews in my other programs, even though it makes the file larger), which makes it easy to go back and make any changes, you know - like when you look at a processed image six months later and ask "what on earth was I thinking?" Or maybe you just want to tweak one little thing without having to redo the entire file.

Now that TIFF files can handle layers (they couldn't always, I believe it was TIF-6 in 1992 that brought layers to TIFF files) you can also save your master as a layered TIFF and a lot of people prefer that TIFF is an openly documented format - which actually is not necessarily true of a layered version. The actual layers implementation itself isn't documented (or not well documented). The increased flexibility of TIFF-6 brings complications that create the same compatibility issues you have with PSDs. I also know there was a bug in LightRoom 1 where saving data back to the XMP file could corrupt the layer structure of a layered TIFF, but I believe that has since been fixed. And I believe you still can't save multitones (duo-, tri- or quadtones) in TIFF.

So if you save layered master files, I'm just not convinced that either one has better archival potential right now. Personal preference and there are good arguments on both sides of the fence. I've just always used PSD because 'back in day' TIFF was really more for standardizing the format of scanned images, so I've been saving in the native format since PS3.0

Layered files, TIFF or PSD, are big files, but I have never regretted saving a layered file while I have often regretted not saving a layered file. If you don't save your layers and you want to change something in the future, you'll be starting all over again from scratch. If you're into more artistic editing, rather than just straight conversions, I think it's even more important.

Even if I decide not to do any additional editing in PS, I still have to open the file and review it at 100%, so once it's saved as a PSD I know that I've completed that quality check.

So like you said, I always know that a PSD file is my "master", as opposed to my "negative" which is my RAW file. That master PSD is then always used to create all of the derivative files I need. Compatibility issues don’t come into play because all of my derivative versions are created in PS from that master. If I need a TIFF file it's easy enough to flatten my PSD, convert it to the required color profile and save it as a (unlayered) TIFF. When I need several different JPEG versions with different color profiles, I have it all set up to run as actions from my master PSD so it takes no time at all.

chuckinsocal · November 2, 2008

Lori, you totally rock

!

You've pretty much validated that I'm at least close to being on the right track with these RAW files.

I know I MUST get in the habit of working with layers. In fact, I was just working on a file and I've gone too far with it and pretty much ruined it. Now I have to go back to the original and start over. No big deal but it would be a lot easier if I could just delete a messed up layer or two and start from there.

I'm not certain what a Smart Object is. Could you expand on that just a little?

I'm used to working with PSDs so I think I'll just stick with those.

I think we are almost neighbors. You're in the OC and I'm right down the road in Carlsbad. If you get down this way let me know and maybe we can shoot some beach scenes or something.

Thanks so much for sharing your wisdom

.

OffTopic · November 3, 2008

chuckinsocal wrote:

I'm not certain what a Smart Object is. Could you expand on that just a little?

I'm used to working with PSDs so I think I'll just stick with those.

I think we are almost neighbors. You're in the OC and I'm right down the road in Carlsbad. If you get down this way let me know and maybe we can shoot some beach scenes or something.

Closer than you think

- I'm in south county (San Juan Capistrano) so I always prefer to head south over north! Would love to get together and shoot, I'm down that way all the time. As much as I love Laguna et al, there are some gorgeous beaches down in SD county!

Smart Object layers were new with CS2; you can transform the file (move it, scale it, warp it, etc.) while maintaining all of the original quality and resolution, because the file isn't rasterized until you output or flatten the Photoshop document. Rendering is done when you commit to a specific resolution. So you can scale the file down to 10 pixels and then bring it back up to original size without causing any damage because it isn't rendered until you flatten it.

When you are in ACR, clicking Open Object opens the file in Photoshop as a Smart Object. If you hold down the Shift key the button turns into Open Image, which opens the file the "old" way...not as a Smart Object. If you hold down the ALT Key, the button changes to "Open Copy", which opens it as a regular image file and does not update the ACR settings to the preview (in Bridge, not sure if that is different in LR) when you want to create several different renderings. I think that is because of the way I have my preferences set, so if your button says Open Image now, hold down the Shift Key and it should change to Open Object.

When you use Smart Objects you can use Smart Filters. You can edit the settings of a Smart Filter at any time without altering the original layer, even after you close and open the file. But Smart Filters use a lot more RAM and create a bigger file, so there is a tradeoff. Personally, the only filters I really use are for sharpening so this doesn't mean much to me.

You can convert a non-smart layer (

) to a smart object by selecting Layer>Smart Objects>Convert to Smart Object. And if you use a lot of Smart Filters, you can keep down your file size by rasterizing a layer when you're committed to it by selecting Layer>Rasterize>Smart Object.

Make sense?

chuckinsocal · November 4, 2008

Cool ... next time you know you're coming down, e-mail me at chuck at socalimages dot com and maybe we can make arrangements to meet up.

HarlanBear · November 10, 2008

OffTopic wrote:

Layered files, TIFF or PSD, are big files, but I have never regretted saving a layered file while I have often regretted not saving a layered file. If you don't save your layers and you want to change something in the future, you'll be starting all over again from scratch. If you're into more artistic editing, rather than just straight conversions, I think it's even more important.

So like you said, I always know that a PSD file is my "master", as opposed to my "negative" which is my RAW file. That master PSD is then always used to create all of the derivative files I need. Compatibility issues don’t come into play because all of my derivative versions are created in PS from that master. If I need a TIFF file it's easy enough to flatten my PSD, convert it to the required color profile and save it as a (unlayered) TIFF. When I need several different JPEG versions with different color profiles, I have it all set up to run as actions from my master PSD so it takes no time at all.

I agree with these statements by Lori 100% and learned the hard way. After all, disc space is relatively cheap, considering how much most of us spend to actually get a picture, and getting cheaper all the time. So how big a deal is storage really?! And since I often have several layers within an image because I experiment with different artistic versions, it is just so much easier to have one image which I can change at will; at least until I decide on the version I want to be the "final master".

Thanks for your explanation, Lori. Makes a lot of sense.

jfriend · November 10, 2008

HarlanBear wrote:

I agree with these statements by Lori 100% and learned the hard way. After all, disc space is relatively cheap, considering how much most of us spend to actually get a picture, and getting cheaper all the time. So how big a deal is storage really?! And since I often have several layers within an image because I experiment with different artistic versions, it is just so much easier to have one image which I can change at will; at least until I decide on the version I want to be the "final master".

Thanks for your explination, Lori. Makes a lot of sense.

Good points. To add a little dollars and sense to this, internal hard drives are around 13 cents/GB now ($99 for a 750GB SATA drive). If a layered PSD or TIFF was 50MB (which is not as big as they can be, but not small either), that would be $0.0067 of storage for that file which is just a little more than half a cent.

speric · January 13, 2009

OffTopic wrote:

Yup, saving your RAW file and saving your edited file are two completely different things. For starters, most edits can't be saved back to a DNG or RAW file and that is why you're getting the warning message when you try to do it.

Just think of your RAW file as your negative. I add my metadata to it and nothing else. My latest conversion settings are reflected in the thumbnail preview, but the file itself isn't changed (other than converting to DNG).

What you're doing now is similar to my current workflow that I’ve developed after years of learning. I can't tell you how I often I want to kick myself for my earlier workflow methods...I had some very poor habits in the beginning which were destructive in nature, and as I learned more about digital processing and RAW in particular, I was really mad that I couldn't go back and undo some of what I had done. Most of this I’ve developed after reading many of Bruce Fraser’s outstanding books in his “Real World” series. I think your optimal workflow really needs to take into account the type of work you shoot. I have to admit that if I were doing a lot of event shooting and processing a high volume of photos, I would probably have a different workflow, but I also probably wouldn’t be as concerned about the long-term implications of a non-destructive workflow.

Here is my somewhat anal-retentive workflow:

I open the file in ACR and do my converting, then click right through to open the file as a smart object in Photoshop. I try to make sure that all my editing is non-destructive, which means I use a lot of layers. Everything I do is done on its own separate layer, which makes it very easy to change, mask, turn on/off, or delete without worrying about how many history states I have. Even something as simple as dust spotting is done on a separate layer. That completed new "master" version is saved as a layered PSD (with Maximize Compatibility turned on so I can view the previews in my other programs, even though it makes the file larger), which makes it easy to go back and make any changes, you know - like when you look at a processed image six months later and ask "what on earth was I thinking?" Or maybe you just want to tweak one little thing without having to redo the entire file.

Now that TIFF files can handle layers (they couldn't always, I believe it was TIF-6 in 1992 that brought layers to TIFF files) you can also save your master as a layered TIFF and a lot of people prefer that TIFF is an openly documented format - which actually is not necessarily true of a layered version. The actual layers implementation itself isn't documented (or not well documented). The increased flexibility of TIFF-6 brings complications that create the same compatibility issues you have with PSDs. I also know there was a bug in LightRoom 1 where saving data back to the XMP file could corrupt the layer structure of a layered TIFF, but I believe that has since been fixed. And I believe you still can't save multitones (duo-, tri- or quadtones) in TIFF.

So if you save layered master files, I'm just not convinced that either one has better archival potential right now. Personal preference and there are good arguments on both sides of the fence. I've just always used PSD because 'back in day' TIFF was really more for standardizing the format of scanned images, so I've been saving in the native format since PS3.0

Layered files, TIFF or PSD, are big files, but I have never regretted saving a layered file while I have often regretted not saving a layered file. If you don't save your layers and you want to change something in the future, you'll be starting all over again from scratch. If you're into more artistic editing, rather than just straight conversions, I think it's even more important.

Even if I decide not to do any additional editing in PS, I still have to open the file and review it at 100%, so once it's saved as a PSD I know that I've completed that quality check.

So like you said, I always know that a PSD file is my "master", as opposed to my "negative" which is my RAW file. That master PSD is then always used to create all of the derivative files I need. Compatibility issues don’t come into play because all of my derivative versions are created in PS from that master. If I need a TIFF file it's easy enough to flatten my PSD, convert it to the required color profile and save it as a (unlayered) TIFF. When I need several different JPEG versions with different color profiles, I have it all set up to run as actions from my master PSD so it takes no time at all.

I like what you've described here. I'm trying to figure out a good system here and have had the experience of "starting over". but i'm still not sure if you convert your RAW files to DNG before working in photoshop and saving edits as a PSD. what format are you saving into PSD from?

arodney · January 13, 2009

Art Scott wrote:

I do pretty much the same except I save as a DNG instead of a PSD....since DNG is open source and at least one camera out there has that as an option as a capture file format.....and I am hoping, but not holding my breath, that others in the near future will see the advantage of it as opposed to all of the propriatary raws we have shoved own our throats...........................

That's a good reason to NOT use PSD (there's really no reason). However, TIFF is also an open format, far more fully supported in other applications and can save all the options PSD can.

For me, DNG on the fly at import of Raw. Rendered images get saved as TIFFs.

arodney · January 13, 2009

OffTopic wrote:

Yup, saving your RAW file and saving your edited file are two completely different things. For starters, most edits can't be saved back to a DNG or RAW file.

At least with Adobe applications, nothing is saved into proprietary Raws. They are read only. DNG files, controlled by Adobe, are read and write (XMP data, the sidecard stuff we no longer have to keep track of, thanks to DNG, are embedded into that container). There's also three possible JPEG previews of the existing rendered data instructions (if updated) that is saved within DNGs.

Metadata is not saved to Raws (that's all in the XMP data in either some kind of sidecard or depending on the product, stored in a database).

More about DNG here:
http://www.ppmag.com/reviews/200709_adobedng.pdf

mercphoto · January 13, 2009

arodney wrote:

That's a good reason to NOT use PSD (there's really no reason).

Ummmm..... if you've done any layering and masking, if you've done any localized edits, if you've applied any filters, if you've used any text layers. Those are all reasons to save a PSD, because saving in any other format will loose all that information and store only the "flattened" image. Thus you cannot open the file again and play with a particular layer. If you edited in Photoshop and if you ever plan to change your edits you're best off saving a PSD file.

arodney · January 13, 2009

mercphoto wrote:

Ummmm..... if you've done any layering and masking, if you've done any localized edits, if you've applied any filters, if you've used any text layers. Those are all reasons to save a PSD, because saving in any other format will loose all that information and store only the "flattened" image. Thus you cannot open the file again and play with a particular layer. If you edited in Photoshop and if you ever plan to change your edits you're best off saving a PSD file.

Absolutely incorrect. TIFF in Photoshop supports saving ALL that data. If you are smart enough to use the maximize backwards compatibility, you'll even get a flattened version embedded for older TIFF readers that don't support layers. With PSD, you also need that option, it will provide roughly the same sized file but virtually no programs outside of Adobe support reading PSDs because to do so, the company has to pay a license to Adobe to support PSD (not with TIFF, its an open format owned by Adobe).

Let me quote my dear friend Jeff Schewe on this to sum it up and save some time:

Wrong...PSD is now a bastardized file format that is NOT a good idea to use. Even the Photoshop engineers will tell you that PSD is no longer the Photoshop "native" file format. It has no advantages and many disadvantages over TIFF.

TIFF is publicly documented, PSD is not. That makes TIFF a preferred file format for the long term conservation of digital files.

TIFF uses ZIP compression for max compression, PSD uses RLE which if you save without the Max compatibility will be a bit smaller, but at the risk of not being able to be used by apps, like Lightroom.

TIFF can save EVERYTHING a PSD can save including layers, paths, channels, transparency, annotations and can go up to 4 GIGS in file size. TIFF can save all the color spaces PSD can. The ONLY thing I can think of that PSD can save that currently TIFF can't save is if you Save out of Camera Raw a cropped PSD, you can uncrop the PSD in Photoshop CS, CS2 or 3. That's one tiny obscure thing that PSD can do that TIFF currently doesn't. How many people even knew that let alone use it?

PSD used to be the preferred file format back before Adobe bastardized it for the Creative Suite. The moment that happened, PSD ceased to be a Photoshop "native" file format. PSB is the new Photoshop "native" file format for images beyond 30,000 pixels. And , at the moment, only Photoshop can open a PSB.

Getting back to the fist point, Adobe can do anything including stopping support for PSD because it's a proprietary file format. TIFF is public, even if it's owned by Adobe (by virtue of the Aldus purchase). Even if Adobe went belly up tomorrow, TIFF would continue.

And, let me be blunt, anybody who thinks PSD is "better" than TIFF is ignorant of the facts. If Adobe would let them, the Photoshop engineers would tell you to quit using PSD. Lightroom for the first beta did NOT support PSD and Hamburg fought tooth and nail to prevent having to accept PSD. He blinked, but you still can't import a PSD without Max compat enabled-which basically makes it a TIFF with a PSD extension.

Look, I'll make it REAL simple...

TIFF = Good
PSD = Bad

Ok?

thayes01 · February 3, 2009

DNG and DxO Optics Pro

BradfordBenn wrote:

I really like the DNG option and I have also gone with the convert to DNG on import into Lightroom.

For those who understand DNG I have a scenario and a question. I am evaluating changing my workflow. I shoot RAW. My plan is to first process my image with DxO Pro for basic color and optics correction. I would have DxO convert the file to DNG after it has been improved by DxO.

Does DNG provide me the ability to reinstate my file back to its original RAW state before DxO applied its improvements or have I in effect created a new baseline for my image because I used DxO Pro to convert it to DNG after it applied color correction?

If the DNG file can be reinstated back (at any time) to the 'RAW' state of the original file, how does one go about reinstating it?

jfriend · February 3, 2009

thayes01 wrote:

For those who understand DNG I have a scenario and a question. I am evaluating changing my workflow. I shoot RAW. My plan is to first process my image with DxO Pro for basic color and optics correction. I would have DxO convert the file to DNG after it has been improved by DxO.

Does DNG provide me the ability to reinstate my file back to its original RAW state before DxO applied its improvements or have I in effect created a new baseline for my image because I used DxO Pro to convert it to DNG after it applied color correction?

If the DNG file can be reinstated back (at any time) to the 'RAW' state of the original file, how does one go about reinstating it?

The kinds of things that DXO is doing to the image with optics correction is creating a whole new image and putting it in a DNG format. Unless DXO includes the original RAW file in the new DNG (which is an option in DNG, but takes a lot of extra disk space), you would not be able to convert the DXO DNG back to the original RAW file.

Here's an interesting thread with more info on that topic.

thayes01 · February 3, 2009

jfriend wrote:

The kinds of things that DXO is doing to the image with optics correction is creating a whole new image and putting it in a DNG format. Unless DXO includes the original RAW file in the new DNG (which is an option in DNG, but takes a lot of extra disk space), you would not be able to convert the DXO DNG back to the original RAW file.

Here's an interesting thread with more info on that topic.

Thanks a bunch John. Appears my best approach is RAW to LR, convert to DNG. LR to DxO Pro. Process as TIF for basic corrections. LR to PS for detailed corrections. LR export to JPG for desired output files.

That article you reference I think supports not using DNG as the "processed" file format. Thanks for the reference.

By the way, I wish to thank you publicly for your numerous contributions to Dgrin, especially in the Customization forum. I have benefited greatly from reading the threads. They've help me get past many a bump in my CSS, HTML and JavaScript endeavors. All the best.

arodney · February 3, 2009

jfriend wrote:

The kinds of things that DXO is doing to the image with optics correction is creating a whole new image and putting it in a DNG format. Unless DXO includes the original RAW file in the new DNG (which is an option in DNG, but takes a lot of extra disk space), you would not be able to convert the DXO DNG back to the original RAW file.

You can't take a rendered image and make it Raw again. Raw is like the individual ingredients of a cake you've yet to mix and bake. Rendered pixels are baked pixels. You can't take the whole and break up the parts again into the ingredients. DxO may build a DNG with Raw plus its rendered corrections but you've drawn a line in the sand, you've now got two iterations (well really one, Raw is Raw).

CynthiaM · February 3, 2009

arodney wrote:

Absolutely incorrect. TIFF in Photoshop supports saving ALL that data. If you are smart enough to use the maximize backwards compatibility, you'll even get a flattened version embedded for older TIFF readers that don't support layers. With PSD, you also need that option, it will provide roughly the same sized file but virtually no programs outside of Adobe support reading PSDs because to do so, the company has to pay a license to Adobe to support PSD (not with TIFF, its an open format owned by Adobe).

Let me quote my dear friend Jeff Schewe on this to sum it up and save some time:

Andrew:

Nice to see you posting here. Always appreciate your viewpoint and expertise.

Up until reading this post, I've always saved as PSD because, frankly, I never saw an explanation like the one you quoted from Jeff Shewe that so clearly explained why I should NOT do this. So what do I do with all of these PSDs? Is there a way to batch convert with all of the layers? Would you recommend that? Should I use the LZW compression?

If you don't mind entertaining another question, I've got one concerning RAW vs DNG. My understanding of DNG is that there is no sidecar file which stores all of your processing so that the actual dng file is being written to which concerns me because if your computer decides to go awry at the moment of writing to the file, the file could be corrupted and no longer readable, whereas with a raw file, it would just be the sidecar that gets corrupted. So how do you protect against this with the dng; downloading a copy to another drive at import? And what do you do with the raw files?

Thanks,

mercphoto · February 3, 2009

CynthiaM wrote:

If you don't mind entertaining another question, I've got one concerning RAW vs DNG. My understanding of DNG is that there is no sidecar file which stores all of your processing so that the actual dng file is being written to which concerns me because if your computer decides to go awry at the moment of writing to the file, the file could be corrupted and no longer readable, whereas with a raw file, it would just be the sidecar that gets corrupted. So how do you protect against this with the dng; downloading a copy to another drive at import? And what do you do with the raw files

Mac OS, and I believe most unix/linux systems, use a journaled file system. Journaled file systems are less likely to have these types of file system errors by design. I've been out of the Windows world for so long I do not know if their file system is like this (though I certainly hope so).

http://en.wikipedia.org/wiki/Journaled_file_system

Personally I like DNG's approach and I despise sidecar files.

arodney · February 3, 2009

CynthiaM wrote:

Is there a way to batch convert with all of the layers? Would you recommend that? Should I use the LZW compression?

You could setup Photoshop's save preferences then use a droplet and do this. Not sure I'd go that route for legacy files. I'm sure I have files going back to Photoshop 1.0.7 days that are in various incarnations of PSD, TIFF and others. I think they are fine although I do recall a few years ago, trying to open a very old legacy document (not sure if it was PSD or not) and not being able to do so. Graphic Converter was able to open the document and I resaved it as a TIFF, confirming Photoshop could open it.

If you don't mind entertaining another question, I've got one concerning RAW vs DNG. My understanding of DNG is that there is no sidecar file which stores all of your processing so that the actual dng file is being written to which concerns me because if your computer decides to go awry at the moment of writing to the file, the file could be corrupted and no longer readable, whereas with a raw file, it would just be the sidecar that gets corrupted. So how do you protect against this with the dng; downloading a copy to another drive at import? And what do you do with the raw files?

Any file could get corrupted at any time. With Raw plus Sidecar, you've got two instead of one. Yes, the sidecar is far less important given the choice of which would go south.

My main drive is a mirror array, I suppose its even possible that as LR writes data to one drive, its already backed up....

I've probably got no less than 5 clones (to be fair, backed up within a month of each other) of my Lightroom disk which contains just my images and LR data (Prefs, library, thubs, saved collections etc).

Save RAW as PSD, Tiff, Other?

Comments