Behavior of AUTO TONE Develop

[Paul_DS256]

Lightroom Version Number

LR Classic 9.4

Operating System

1. Windows 10

I normally put any noisy image through Topaz Denoise before making any adjustments in LR. However, I miss-stepped today so saw the AUTO TONE on the NEF before I reset to go to Topaz. When I came back and tried to AUTO TONE the TIFF file, the image was not as exposed as the AUTO TONE on the NEF. When I manually set the TIFF file to the adjustment values on the NEF AUTO TONED, the image still did not come back. Attached is the progression of images.

I have a number of questions:

• Why is there a difference between the AUTO of NEF and TIFF? I see the color profile for the NEF is Adobe Color and only Color for the TIFF.
• Should I be adjusting the NEF before sending it to TOPAZ Denoise? Is the conversion from RAW to TIFF losing some information in the image?
• Should I be able to bring the TIFF file up to the same presentation as the NEF using the same settings? I think the answer is no but I'm trying to understand what's happening.

Thanks

 

[johnrellis]

Why is there a difference between the AUTO of NEF and TIFF?

An educated guess: The Auto algorithm is a machine-learned algorithm trained on a large set of photos manually adjusted by expert LR users. The set may have included both raws and non-raws, since in general Develop settings work differently on raws and non-raws (to the chagrin of many users who copy the settings from a raw to a TIFF created from the raw). Machine-learned algorithms can be very sensitive to the training sets, so if the non-raws in the set didn't include the same set of visual images as the raws, then the algorithm will sometimes produce much different visual results for a given raw/non-raw pair.

 

[johnrellis]

I see the color profile for the NEF is Adobe Color and only Color for the TIFF.

Adobe Color is the default color profile for raws, and Color is the default profile for non-raws.

 

[johnrellis]

Should I be able to bring the TIFF file up to the same presentation as the NEF using the same settings?

As I mentioned above, some Develop settings work differently on raws than on non-raws, and you can often notice differences when copying them from a raw to a TIFF exported from the raw. Usually the differences are small to modest but sometimes are quote noticeable. Typically you'll notice modest differences in Shadows and Highlights when set to larger values. And large changes in Temp and Tint don't copy well (the Temp/Tint sliders have completely different ranges in raws and non-raws, and LR tries half-heartedly to translate between them).

Here's an example:

There are significant differences in white balance and modest differences in shadows and highlights (easier to see on a full screen). Here's the Photoshop difference:

 

[Paul_DS256]

Thanks John.

So, given the potential differences, do you think one should apply some basic adjustments to an image before sending it to Topaz?

 

[johnrellis]

do you think one should apply some basic adjustments to an image before sending it to Topaz?

Maybe someone else can make a recommendation -- I don't have much experience with external editors.

 

[Paul_DS256]

I just had a though that the 'denoising' process in Topaz may be responsible.

 

[clee01l]

Thanks John.

So, given the potential differences, do you think one should apply some basic adjustments to an image before sending it to Topaz?

My approach to developing photos is to do as much as I can in Lightroom Classic . IF I feel additional processing is needed, then I send it to an external editor. So I will apply NR to the images before I send it to deNoise. I have heard that the recommendation would be to not apply NR before sending an image to deNoise Or Sharpening etc. But any Raw image imported into Lightroom as some NR and some sharpening applied. I've tried sending images to Topaz with and without applying the Lightroom equivalent tool to the image and found that the returned image showed little difference. Perhaps this is because the image once it reaches Topaz would require less of the Topaz process if some of the process has already been applied in the TIFF sent to Topaz.

 

[Paul_DS256]

My approach to developing photos is to do as much as I can in Lightroom Classic . IF I feel additional processing is needed, then I send it to an external editor.

Thanks for the insight Cletus. I'm going to play around with this a bit.
I developed the practice of sending the unmodified image to Topaz upon recommendations from others as well as Topaz's recommendation DeNoise AI Frequently Asked Questions

 

[clee01l]

Thanks for the insight Cletus. I'm going to play around with this a bit.
I developed the practice of sending the unmodified image to Topaz upon recommendations from others as well as Topaz's recommendation DeNoise AI Frequently Asked Questions

Try this test. Take an image and process it completely using LrC. If it still needs work, send it to Topaz twice. Once with the Lightroom adjustments and a second time with the LrC adjustments removed and compare the results. Then post both results back to this thread.


Sent from my iPad using Tapatalk

 

[Zenon]

I use Topaz from time to time. I first apply Auto using Adobe Colour colour profile as it seems work best. I have Color Fidelity profiles (for Canon) but Auto saturates them a bit more. After Auto I select a colour profile. Then I send to Topaz if it needs to but I do avoid it if possible. Before I do that I apply a preset that takes Sharpening and NR to 0.

One that note. Adobe. Where is my NR Sensei I requested over a year ago?

 

[Zenon]

I should say I make sure the open with LR edits is checked as it exports to Topaz.

 

[Paul_DS256]

Thanks Zenon and Cletus. I'll work on Cletu's suggestion of a comparison.

Just one note this is around Topaz Denoise only and when to apply it. I also had Topaz Sharpen which I will use at the end of my processing.

 

[Paul_DS256]

Try this test. Take an image and process it completely using LrC. If it still needs work, send it to Topaz twice. Once with the Lightroom adjustments and a second time with the LrC adjustments removed and compare the results. Then post both results back to this thread.

Here you go. Attached are the steps and images in JPG

• The Adjustments before denoising seems to make the image noiser
• The results comparing denoising before and after adjustment shows Denoise did a good job.

So, it sounds like, contrary to popular guidance, do the adjustments in LR before sending to Denoise.