RAW or Tiff for high end processing.

[Steven Draper]

My standard traditional workflow for my best images / prints was to take a NEF file and do certain things in Capture NX (white balance, exposure nudge, shadows, highlights, perhaps a curve, select which one of the camera profiles) and then convert to a 16bit Tiff.

Tiff then goes through things like DXO for optical correction, maybe various exposure variations in Photomatix for some kind of blend,  and then into Photoshop for sharpening and all the other creative enhancements, layers, blends etc prior to creating a master file from which various end user files are created.

It seems that today the 'fashion' is to do as much as possible in a non destructive way to the raw file, in many cases ending up with a 'master' raw complete with all the adjustments and the talk of tiff's seems to be less and less.  

Questions??

Is it just simply a case of time (I do a lot of LR only processing - even though I know a longer workflow may tease out more detail - not every image can afford a hours on the screen)
Is there a quality issue, i.e. applying as much change as possible and then rendering.

I'm afraid I'm not a scientific photographer, but I do like to try and understand the scientific principles behind what and I I do certain thing and would really appreciate if someone could explain why working an image completely in RAW is now deemed better than the traditional Raw - convert Tiff using your preferred convertor - and then create and enhance!  

Many Thanks in advance.
Steven

[digitaldog]

Questions??

Is it just simply a case of time (I do a lot of LR only processing - even though I know a longer workflow may tease out more detail - not every image can afford a hours on the screen)
Is there a quality issue, i.e. applying as much change as possible and then rendering.

Do as much work in the Raw converter as you can. Draw a line in the sand when you decide you're going to render pixels for further work (Dxo, Photoshop etc). Now you're done with parametric editing in the Raw converter. You've baked those pixels.

[teddillard]

Do as much work in the Raw converter as you can. Draw a line in the sand when you decide you're going to render pixels for further work (Dxo, Photoshop etc). Now you're done with parametric editing in the Raw converter. You've baked those pixels.

LOVE that analogy...  baked pixels.    

My solution is to do everything in Adobe, processing the file in Camera RAW and then work on them in Photoshop as Smart Objects.  Thus, you can re-bake them pixels to your heart's content.  You can also use many plugins, such as NIK etc, which get used as Smart Filters- also allowing re-baking ad infinitum.  I'll also layer those Smart Object RAW files...  you get huge file "overhead", but amazing results.

[digitaldog]

My solution is to do everything in Adobe, processing the file in Camera RAW and then work on them in Photoshop as Smart Objects.  Thus, you can re-bake them pixels to your heart's content.  You can also use many plugins, such as NIK etc, which get used as Smart Filters- also allowing re-baking ad infinitum.  I'll also layer those Smart Object RAW files...  you get huge file "overhead", but amazing results.

That will work well (despite the overhead). You're still drawing that line in the sand but its more blurred, since you are just updating each time you move back and forth from ACR to PS. What's great however is you can see the Raw plus rendered effects, all in one place and move back to ACR to tweak the Raw (and thus the rendered effects), on the fly.

[Steven Draper]

Many thanks for your thoughts, but I'm still trying to understand why?

Is the reason to follow this approach in order to have maximum convenience for future editing control on the file, ie to go back and make a 'core' tweak to say image WB, or total image quality?

Does the amount of rendering have an effect? ie Does an image that is converted to a TIFF after an initial conversion that then moves through several specialized software applications as each applies a certain image adjustment role experience a negative image quality effect, or is it more about the convenience of making or readjusting the enhancements later?

Steven

[digitaldog]

Rendering is pixel creation (from the Raw + Instructions).

Pixel editing is altering existing pixel values which are not stepless (although far less an issue in 16-bit). Its also far slower. The data path is different.

Raw rendering and subsequent pixel editing are two quite different paths in producing an optimized image.

[teddillard]

Many thanks for your thoughts, but I'm still trying to understand why?

Is the reason to follow this approach in order to have maximum convenience for future editing control on the file, ie to go back and make a 'core' tweak to say image WB, or total image quality?

Does the amount of rendering have an effect? ie Does an image that is converted to a TIFF after an initial conversion that then moves through several specialized software applications as each applies a certain image adjustment role experience a negative image quality effect, or is it more about the convenience of making or readjusting the enhancements later?

Steven

If I may...  I like to describe the RAW processing as the construction of the file, from the original R, G, G and B pixels, or, the "digital latent image".  The chip records the light values in red, green and blue pixels.  (it uses 2 g, actually).  To paraphrase Bruce Radl, of Mosaic Imaging, the processing of that is the process of taking individual red, green and blue pixels and making RGB pixels from it.  You're rendering, or creating, the file for Photoshop to work with.  

In that process you're doing three basic things, (along with a bunch of "extras").  You're setting a black and white point, and mapping your curve between them, or establishing your contrast.  You're setting your color balance, or your neutral point.  Last, you're scaling/sharpening it.  That can be re-done, as it's rendered, repeatedly.  

The RAW file isn't the digital negative, it's the digital latent image- that is, it's the light that fell on the sensor, BEFORE you've made decisions about these three main decisions...  

...hope this helps.  I'm feeling like I've had more lucid moments, but maybe I'll come back and re-edit later, after a good nap.  

[Guillermo Luijk]

Raw rendering and subsequent pixel editing are two quite different paths in producing an optimized image.

Can we be sure the RAW developer will apply all its options (exposure adjustment, contrast, saturation, curves, lens corrections,...) in a single floating point step with no rounding errors till the end? because otherwhise it will be the same as applying all those steps sequentially in PS.

BR

[digitaldog]

Can we be sure the RAW developer will apply all its options (exposure adjustment, contrast, saturation, curves, lens corrections,...) in a single floating point step with no rounding errors till the end? because otherwhise it will be the same as applying all those steps sequentially in PS.

BR

Its not the same as PS. Its still a linear encoded, high gamut processing space. At least with ACR/LR, the order is said to be optimized (not necessarily user order as in PS).

[feppe]

Can we be sure the RAW developer will apply all its options (exposure adjustment, contrast, saturation, curves, lens corrections,...) in a single floating point step with no rounding errors till the end? because otherwhise it will be the same as applying all those steps sequentially in PS.

Perhaps an even more important question is: does it matter? What I'm getting is that even if you do extensive editing on a good-quality 16-bit file, I doubt even 100% visual pixel-peeping will confirm differing results.

[Guillermo Luijk]

Perhaps an even more important question is: does it matter? What I'm getting is that even if you do extensive editing on a good-quality 16-bit file, I doubt even 100% visual pixel-peeping will confirm differing results.

I agree, this is what I think. IMO a 16-bit integer gamma format can per se contain a very high quality and robust image. The only point is that is has to be a high quality image right from the source.

I invite anyone to download this 16-bit TIFF, and start processing it to lift the shadows, or to do anything to it. It can be overexposed by 12EV (+3EV curves are provided) without any posterization, and it was the gamma encoding which made it so robust. With a 16-bit linear DNG we couldn't even dream of producing an image with such a high tonal quality.

[Panopeeper]

I invite anyone to download this 16-bit TIFF, and start processing it to lift the shadows, or to do anything to it. It can be overexposed by 12EV (+3EV curves are provided) without any posterization, and it was the gamma encoding which made it so robust. With a 16-bit linear DNG we couldn't even dream of producing an image with such a high tonal quality.

The above statement may easily be misunderstood as favourism for non-linear encoding in order to preserve higher image quality. That would be not only incorrect, but the opposite is true.

What we have to do here is the bit depth: the 16bits are not enough for that data in linear form. 32bit linear data would be even more robust than 16bit nonlinear.

[Guillermo Luijk]

The above statement may easily be misunderstood as favourism for non-linear encoding in order to preserve higher image quality. That would be not only incorrect, but the opposite is true.

What we have to do here is the bit depth: the 16bits are not enough for that data in linear form. 32bit linear data would be even more robust than 16bit nonlinear.

Right Gabor, but 16-bit is a constraint of this thread: "RAW vs TIFF" (TIFF is assumed to be 16-bit. Maybe tomorrow it is not, but today it is if you want to process your photographs).

So, given we have 16-bit and not more, it's the gamma that allows us to keep much more image information than any 16-bit linear RAW can encode (not more information, but better distributed actually). And if that is true, why caring so much about doing as many operations in the RAW developer as we can? if the 16-bit non-linear TIFF is an appropiate robust container for image edition.

[Panopeeper]

TIFF is assumed to be 16-bit. Maybe tomorrow it is not, but today it is if you want to process your photographs

Click on Image->Mode and seelct 32 bits :-)  (I know, I know)

why caring so much about doing as many operations in the RAW developer as we can? if the 16-bit non-linear TIFF is an appropiate robust container for image edition

For me it is not the question of accuracy (bit depth), but of the most appropriate *stage* of doing certain steps. There are some steps, which are better done in the stage of raw processing. There is a good reason that raw converters are enforcing a certain sequence of operations instead of following the sequence of your selections.

[digitaldog]

because otherwhise it will be the same as applying all those steps sequentially in PS.

The above statement may easily be misunderstood as favourism for non-linear encoding in order to preserve higher image quality. That would be not only incorrect, but the opposite is true.

So its the same as applying all those steps sequentially or its not? Maybe I'm misunderstanding the above two differing points of view?

[digitaldog]

So, given we have 16-bit and not more, it's the gamma that allows us to keep much more image information than any 16-bit linear RAW can encode (not more information, but better distributed actually). And if that is true, why caring so much about doing as many operations in the RAW developer as we can? if the 16-bit non-linear TIFF is an appropiate robust container for image edition.

Its way faster for one. It provides unlimited history for another. As mentioned, at least in Adobe Raw processing, the order is optimized, not user applied.

If you bring a rendered image back into the Adobe product, its going to get converted to linear, ProPhoto Primaries for any editing. You've also got yet another document (a big honking one) to track in addition to the Raw.

[Jack Flesher]

Interesting discussion, though perhaps quickly moved beyond the level of the OP I think

My .02 is I am not sure there is a right or wrong way to look at this, workflows are different and often have different purposes.  Personally, I choose to make all basic "global" adjustments in the raw files, then move to the editor (CS4 in my case) to make "local" adjustments.  And often, those local adjustments include a final "global" curve, hue and/or saturation adjustment on top of them to balance effects from the local edits...

And... What is "accurate" color?  If shooting a product, we can measure and confirm it.  But if shooting a landscape, is it what we really saw or what we remember seeing, or what we wanted to see?

So, in the end I have to agree with the 16-bit TIFF comment -- I can almost always eek more out of a carefully converted raw in the image editor than I ever could directly from the raw itself.  However, it does require first generating an "optimal" file from the raw that will take the further editing.  Thus, my goal for my fine art prints is to use the raw converter to generate the optimal TIFF for further editing in CS, which is usually not the same settings as I would use for getting as optimal a final image as possible out of the raw processor...

PS: And my .02 on files sizes offered only FWIW:  Drives and drive boxes are really cheap right now and setting up a 4 x 1TB RAID 5 array costs half of what a typical high-end DSLR zoom costs, so it should be well within the budget of any serious photographer.  Call it $800 for 3TB of redundant storage, or about 26 cents per Gig. Assume 140MB converted size for  for any current top-end 22+MP DSLRs 16-bit tiff, and you're at around 3-1/2 cents per image to save it that way.

Cheers,

[Steven Draper]

Many thanks to all the replies.

I have to say that I have lots to learn about the processing of files! I think that because I split an exposure into two or three individual files at different exposures to ensure all data is captured and then blend in a second software application I am able to use all the captured data. Some further blending in Photoshop finishes things off. I still find this better in most cases that using Upoint technology, although perhaps with more practice I'll be able to really fine tune my use of that too. The problem with RAW is that Capture NX often provides the best conversion, but I need other software to carry out other tasks.

I'll have to devise some kind of test and put the same file through several workflows!!!!

atb
Steven

[budjames]

Perhaps this is not the correct place for this post, but it's related.

I have read on several forums to save Photoshop files as layered tiffs instead of the native PSD format to ensure future compatibility. The argument to save in layered tiffs as a better "archival" format than PSD is similar to the argument to save original RAWs as DNGs.

Currently, I save PS edited images as PSD files. Currently I'm using PS CS4 and I have had no problems opening PS 5, 6, CS and CS2 PSD files.

Any comments are greatly appreciated.

Cheers.
Bud James
North Wales, PA

[digitaldog]

Any comments are greatly appreciated.

Start here:

http://luminous-landscape.com/forum/index....hl=why+use+tiff?