• Welcome to the Lightroom Queen Forums! We're a friendly bunch, so please feel free to register and join in the conversation. If you're not familiar with forums, you'll find step by step instructions on how to post your first thread under Help at the bottom of the page. You're also welcome to download our free Lightroom Quick Start eBooks and explore our other FAQ resources.
  • Stop struggling with Lightroom! There's no need to spend hours hunting for the answers to your Lightroom Classic questions. All the information you need is in Adobe Lightroom Classic - The Missing FAQ!

    To help you get started, there's a series of easy tutorials to guide you through a simple workflow. As you grow in confidence, the book switches to a conversational FAQ format, so you can quickly find answers to advanced questions. And better still, the eBooks are updated for every release, so it's always up to date.

Library module working with 32bit TIFF files in Lightroom Classic

rip

New Member
Joined
Aug 23, 2023
Messages
8
Lightroom Version Number
12.5
Operating System
  1. Windows 10
There must be something I don’t understand about working with 32bit TIFF files in Lightroom.

If I open a DNG in Photoshop, save it to a 32bit TIFF without any adjustments, and then open it in Lightroom without any adjustments, the histogram goes from this:
orig.png

To this:
32 bit.png

The image appears overexposed and has lost detail.

How can I open a 32bit TIFF without shifting the colors?


Thanks
 
Where did the DNG come from? Straight out of a camera as DNG? converted by LrC on Import? Result of an Enhance operation? Somtehing else?

The reason I ask is that DNG's can be similar to a RAW file or may have already been de-mosauiced (converted to pixels). If the later, then the de-mosaicing (sometimes called rendering) process made decisions about how to convert the RAW data into colored pixels. This includes the applicaiton of a profile, presets, and other conversion settings and could explain what you are seeing.
 
Thanks. In my case, the originals are camera RAW files converted by Lightroom to DNG.

I'd like to understand if the original DNG files have been converted to RGB on import, and if so, what the bit depth is.

I'd also like to understand if the DNG files created by Lightroom Denoise have been converted to RGB, and if so, what the bit depth is.
 
If Lightroom has already converted the RAW files to DNG RGB, that might explain why opening them in Photoshop and saving them to 32bit TIFF destroys detail and overexposes the images.
 
Someone else needs to verify this but I do not believe Lightroom can handle 32 bit TIFF files.


Sent from my iPhone using Tapatalk
 
I think that max for DN G is 16 bit but when converting from a camera RAW file I don't think it will upscale the bit depth if the camera RAW file was less than 16.

There seem to be 3 flavors of DNG as explained by https://cloudinary.com/guides/image-formats/dng-format-how-it-works-and-pros-cons-to-know-about. As follows

There are three types of DNG files: In-camera DNG, Converted DNG (raw), and Converted DNG (linear).

In-camera DNG

In-camera DNG files are generated by cameras that have adopted the DNG format as their raw image format. For example, when you take a raw photo with a Leica M9 or a Pentax K5, the file created is an in-camera DNG. The advantage of this is that the DNG files created in-camera are universally compatible with raw-processing software, and there is no need to convert these files to another format for editing or archiving.

Converted DNG (RAW)

Converted DNG (raw) files are created when a raw file from a camera is converted into a DNG file. When creating a converted DNG, you typically have the option to save the original raw image data as is, preserving all the image information that the camera wrote into the original raw file.

Converted DNG (Linear)

Converted DNG (linear) files are DNG files with some of the rawness processed out of the file. These files can be useful for enhanced compatibility with different software. It’s also possible to convert a JPEG or TIFF file to a DNG, which will produce a Linear DNG file. While a Linear DNG does not contain the same amount of data as a raw DNG, it is still a versatile file type that can be used in a variety of applications.

What is not clear (and I haven't been able to find any info on) is how to determine which of these 3 any particular DNG file is.
 
There must be something I don’t understand about working with 32bit TIFF files in Lightroom.

If I open a DNG in Photoshop, save it to a 32bit TIFF without any adjustments, and then open it in Lightroom without any adjustments, the histogram goes from this:
View attachment 23285

To this:
View attachment 23286

The image appears overexposed and has lost detail.

How can I open a 32bit TIFF without shifting the colors?


Thanks
I am curious to know what camera the original file came from. You mention opening a DNG in PS and then trying to save it as a 32-bit tiff, but was the original file 32-bit?

--Ken
 
When Lightroom converts a raw image to DNG, that DNG is still a raw image.
When Lightroom creates a Denoised DNG, that DNG is a linear RGB image.
When looking at the bit depth, make sure you do not compare apples and oranges. Image data can be stored as an integer, or as floating point. https://www.rawdigger.com/usermanual/floating-point
 
When its true, that 8 and 16 bit tif-files are integer files and 32-bit tif-files are floating point files, than you see on the file size of a picture what is it: 16 bit dopples the file size and 32 bit makes the file 4 times larger. I think that only a hdr process makes 32 bit files, for example Photoshop does it, when you store a hdr file without tone-mapping. Such a file can imported in LR and you can do the tone mapping in this program. The sliders in the development module have a larger range as usual. I did that the whole time making HDR-files in PS with a batch script "Batch HDR" and import them for development in LR. I used Ps because the Hdr Process in PS was robuster and more failure tolerant (handheld photos!) than the merging in LR. Now i use LR, which is gone better and faster to merge 5 bracketed raw files from my Panasonic G9.
 
I am curious to know what camera the original file came from. You mention opening a DNG in PS and then trying to save it as a 32-bit tiff, but was the original file 32-bit?

--Ken

I do not believe any camera records RAW files in other than 12 or 14 Bit values. JPEGs from the camera are going to be 8 Bit lossy compressed. Opening either in Photoshop, you have the option to process and save as 32bit. That only means that the mathematic results of any develop adjustment are saved to 32 bits. Just as a 12 or 14 Bit RAW date is converted to RGB and saved in a 16 bit pixel value.


Sent from my iPad using Tapatalk
 
I do not believe any camera records RAW files in other than 12 or 14 Bit values. JPEGs from the camera are going to be 8 Bit lossy compressed. Opening either in Photoshop, you have the option to process and save as 32bit. That only means that the mathematic results of any develop adjustment are saved to 32 bits. Just as a 12 or 14 Bit RAW date is converted to RGB and saved in a 16 bit pixel value.


Sent from my iPad using Tapatalk
Is there additional value in working with a 32-bit file as opposed to one that has been initially converted to 16-bit? And, is there value in going from 16-bit to 32-bit? If the file started at 32-bit, I could understand trying to stay there, but it is not clear to me of the benefit of converting the file?

--Ken
 
When Lightroom converts a raw image to DNG, that DNG is still a raw image.
When Lightroom creates a Denoised DNG, that DNG is a linear RGB image.

In Adobe's terminology, "raw" includes linear-raw DNG images produced by Denoise. In general, Adobe's meaning of "raw" includes:

- proprietary manufacturer formats such as NEF or CR3, whether they include mosaic or demosaiced sensor data (e.g. Canon sRAW).

- DNG files that include mosaic and demosaiced (linear raw) sensor data.

The DNG specification uses the term "linear raw" to refer to demosaiced sensor data. The DNG tag PhotometricInterpretation can have the values "CFA (Color Filter Array)" or "LinearRaw".

Eric Chan, the lead developer of Denoise and Camera Raw, wrote in Denoise Demystified:
After multiplying and adding up a gazillion numbers, your computer will produce a new raw file in the Digital Negative (DNG) format that contains your denoised photo. As with previous Enhance features, any adjustments you made to the source photo will automatically be carried over to the enhanced DNG. You can edit this DNG just like any other raw photo, applying your favorite presets and custom tweaks. [My emphasis]

LR's Develop uses "raw" to refer to any image containing mosaic or demosaiced sensor data, regardless of file format. Develop treats all such images identically -- e.g. you can apply camera profiles (and enhanced profiles based on camera profiles) to any such raw image; the Temp slider uses pseudo-Kelvin values rather than -100 .. + 100; and Lens Corrections will allow raw-only lens profiles to be selected.

Similarly, in LR Preferences, "Raw Defaults" apply to proprietary formats and DNGs containing mosaiced and demosaiced camera data. E.g. if you import a linear-raw DNG produced by Denoise, Raw Defaults will be applied to it.

LR isn't completely consistent, of course. The filter and smart-collection criterion File Type uses "Raw" to mean just manufacturer-proprietary formats, and it won't match mosaic or linear-raw DNGs.

Yes, I'm a pedant and proud of it.
 
Last edited:
Is there additional value in working with a 32-bit file as opposed to one that has been initially converted to 16-bit? And, is there value in going from 16-bit to 32-bit? If the file started at 32-bit, I could understand trying to stay there, but it is not clear to me of the benefit of converting the file?
That is my question as well @Replytoken. I can see 32 bit TIFF useful if you a constructing graphic art on a blank canvas in PS. However, I don't see the benefit of upping the bits over the original photo. @rip is your TIFF from a photo or graphic created in PS?
 
@Califdan In-camera DNG ... Converted DNG (RAW) .. Converted DNG (Linear) ... What is not clear (and I haven't been able to find any info on) is how to determine which of these 3 any particular DNG file is.
In general, you have to use Exiftool. The DNG panel is pretty much useless for this.

You can tell which files were converted by Adobe DNG Converter. They will have this field:

Code:
[XMP-xmpMM] History Parameters: converted from image/x-sony-arw to image/dng, saved to new location

which will also tell you the original file type.

You can determine whether LR's Enhance generated the linear DNG by looking at these fields:

XMP:EnhanceDenoiseAlreadyApplied
XMP:EnhanceDetailsAlreadyApplied
XMP:EnhanceSuperResolutionAlreadyApplied

or by looking for a sub-IFD with Subfile Type of Enhanced Image Data.

Any Filter will tell you the enhance type:

1716158185721.png


Of course, you could also look at the file name, provided you didn't rename the file, or you could enable Automatically Add Keywords To Enhanced Images (which is a poor design for image-processing metadata, given the already existing mechanisms for filters and smart collections).
 
Is there additional value in working with a 32-bit file as opposed to one that has been initially converted to 16-bit? And, is there value in going from 16-bit to 32-bit? If the file started at 32-bit, I could understand trying to stay there, but it is not clear to me of the benefit of converting the file?
It depends

Just going to a larger bit depth does not add anything to the image except file size. as an example let's say your have pixel that is some color represented by a number (or 3 numbers if RGB). That number is stored in a bucket that can hold a certain number of digits (actually bits) - this is the bit depth. Making the bucket bigger does not alter the numeric value of the original number. So, as an analogy, let's say I had 3 pixels in my image that in decimal had values of 27, 28, and 29 respecrively in a scheme that used 2 decimal digits per pixel. But now I convert it to a scheme that has 3 decimal digits per pixel. these 3 pixels would then have values 270, 280 and 290 respectiverly. But no pixel in the image could have a value between 271-279, or 281-289 inclusive. So I have not gained any additional colors or details in the image.

But, now that I have 3 decimal digits, I have extra room between 270 and 280 which I didn't have between 27 and 28. This means that if do editing on the image I can have pixels with values bewtwen 270 and 280 which could not have been created if I were still using 2 digits. In other words when using 2 digits I was limited to 100 different colors but once I went to 3 digits I could have 1000 different colors but right after the conversion would actaully have no more than 100.

So, if an image started at 32 bit but during the workflow this was reduced to 16 bit, many pixels that had been different colors may now be the same color (i.e. you have less shades of red). Re-expanding it to 32 bits will not bring back those lost colors but will alllow you to manually create those lost colors.

Hope that makes sense.
 
But, now that I have 3 decimal digits, I have extra room between 270 and 280 which I didn't have between 27 and 28. This means that if do editing on the image I can have pixels with values bewtwen 270 and 280 which could not have been created if I were still using 2 digits. In other words when using 2 digits I was limited to 100 different colors but once I went to 3 digits I could have 1000 different colors but right after the conversion would actaully have no more than 100.

So, if an image started at 32 bit but during the workflow this was reduced to 16 bit, many pixels that had been different colors may now be the same color (i.e. you have less shades of red). Re-expanding it to 32 bits will not bring back those lost colors but will alllow you to manually create those lost colors.
This is exactly what I was wondering and suspected, especially what you covered in your second paragraph.

Thanks,

--Ken
 
That is my question as well @Replytoken. I can see 32 bit TIFF useful if you a constructing graphic art on a blank canvas in PS. However, I don't see the benefit of upping the bits over the original photo. @rip is your TIFF from a photo or graphic created in PS?

I believe my first encounter with 32 Bit files was with HDR images. 32bit colors have more color detail than 16 bit color which has more color detail than 8 bit color.

The 10 in HDR-10 monitors represents to bits of color per channel or 30-bit color . These are more than the human eye can resolve. But this explains (to me) why you might want to output a 32-bit color TIFF file.

The standard number of bits is a byte is 8 for personal computer architecture. So a factor of 8 (16 or 32) is used to store a 16 or 32 bit RGB color pixel. HDR-10 (30 bit) wastes the extra two bits in a 32 bit architecture.

When making color computations (Photoshop edits) you arrive at more accurate color results if you allow for 32 bit color as opposed to 16 or 8. When saving the edited result, you can save at 16 bit you map to 16 bit colors and adjust all of the colors between 16 and 32 bit to a 16 bit result. We’ve been making these same color truncations when we have been saving our 16 bit files as 8 bit JPEG.


Sent from my iPad using Tapatalk
 
Back
Top