No…There isn’t a specific “setup” for the D810. The Z8 uses a new, different, compressed file format that isn’t fully supported yet…
I’m sorry, your initial response implied there was something distinctly different about the D810 which created an “issue”. THAT’s all I was commenting on. There is no special issue with the D810 within the scope of this discussion.
Good day sir. You’re talking in circles accomplished nothing!
Do you have any RAW drone images that do not load with correct color and/or do not save as a valid DNG? If so please send us example images and I can work on adding support for them.
When we say we support a camera or update a camera profile, we mean that you should be able to open the image and save it out as a RAW that opens in other applications. There are still ongoing corrections surrounding color for various cameras but those are handled on a case by case basis.
I can make a note of this. Makes sense to keep the selection.
Adam - I do not typically shoot in RAW, nor a need to save as a DNG. I was asking as Topaz mentioned the DJI Mavic 3 in it’s update notes. I own an Autel Lite+ and most of its pictures come out well as JPEG’s straight from the drone. I do run Topaz Photo and Video AI to remove any excess noise and such - with good results. The Autel has "digital zoom; unlike the DJI Mavic 3 with optical zoom capabilities. We all know digital zoom produces less quality than optical zoom and if I use more than 4x (I can go to 16x) - the noise / grain starts becoming a major factor and even Photo AI does not do well in removing / adjusting it.
I would like to have Photo AI be able to enhance all my photo’s when I check off using Enhancement versus it working only on the initial photo I select and chose Enhance. Otherwise, the AI does what it does and may not produce the output I want without having to go thru every single photo and click on Enhance. I do understand that the AI is supposed to check each photo and apply only the corrections it sees fit to apply. So, if that is the case, then why have all these extra features you can individually select & apply, yet only for 1 out XXXX photos you’re running thru the program.
I’m getting a checkerboard pattern on export. It ONLY happens when Enhance is enabled. Happens on original RAW and on DNG/TIFFs exported from other applications.
PhotoAI 1.3.8 (and earlier versions also)
Macbook M1 Pro running Ventura 13.4
Here’s a crop showing the issue:
Here’s the original RAW:
IMG_7059.CR2 (27.8 MB)
Thank you for your support and develloping with endurance
In a variaty of cases denoise Normal gives better results than Normal v2. In some other cases Normal v2 is just the best choice. Is it possible to leave all 4 choices? I think all of them good basic adjustments to work with.
I tested in Windows 10 now. Same issue here. But it happens with files exported from other apps and files devellopedi n photoai. No difference. These pattern are very agressive in dark areas.
Denoise normal and Enhance - These pattern are very agressive
Without Enhance - pattern not visable
Denoise Strong and Enhance - These pattern are visable but not as much as with normal denoise
Still doesn’t work with Z8 NEF files I just get a white screen!
If you search for Nikon Z8 on this site you will see various threads all with the same reply.
And please don’t duplicate posts over the forum.
Strong v2 denoise looks good overall however I found out that it creates heavy artifacts with some really high ISO images. I’ll add some comparisons with DeNoise 2.2.12 which had the best low light denoise IMO but also added some artifacts in the background. However these artifacts from String v2 are much severe.
An example of a bobcat in very low light (left Strong v2, right DeNoise 2.2.12)
Notice how much better the chest fur looks however the background especially where the out of focus leg is or the area around the head where the fur goes into the background is prone to heavy artifacts.
In other cases it creates less artifacts but also maintains a lower amount of details as can be seen in the snow leopards face here (left Strong v2, right DeNoise 2.2.12)
The artifacts where the tongue starts are gone with Strong v2 but the face overall maintains less details however that it leaves a fine noise adds to the overall look of the image.
I get another strong artifact with a low light portrait of a moaning tiger which leaves a pattern way stronger than the original noise (left Strong v2, right DeNoise 2.2.12)
Theres a strong pattern in the area with the least amount of light in the image.
With not so high ISO values but still high it looks better and is less prone to oversharpen artifacts than the mentioned DeNoise version hovewer adds some in at different location as example where the fur from the front leg lay over the other leg.
However if we look closer at the nose we see again it lets it get more blurry and preserves less details.
That are some of my findings so far. It’s good progress and pretty capable, I limited my examples more to what could need improvements compared to an older DeNoise version. Could you add the models of DeNoise 2.2.12 as another choice of denoise models? They’re still superior for ultra low light shots and a shame that the new models aren’t near there yet. Which is also why I find the price a bit steep as an owner of the single standalone apps of the older version.
An addition for the completeness:
For not that severe noise covered images it looks better already. Here an example of play fighting lynxes where Strong 2 (left side) does a better job in every part of the image while DeNoise with sharpen set to 2 still has oversharpen artifacts.
There is no need for any camera profile. What I do with my DNG/RAW is that I have edited the photo (correct color, edit contrast, exposure, etc., choose the correct camera/lens profile, correct the vignetting, etc.) and I have pass it out of the container into a TIFF file and let Photo AI do it’s thing. You only want Photo AI to do the noise and sharpening.
You’re 1 year behind Topaz Denoise. Did you know that 3.7.2 has been deployed since February 13? With many bug fixes and other improvements since your version dated February 2022.
I know but I’m using the older version on purpose as the low light denoise got worse with high iso images. If you look at the example images you’ll notice that this old version does some things better than the current new denoise version in photo ai. This is the reason why I rolled back to it and use it as a comparison.
It looks like doing a right click on the preview or clicking the 3 dots on the thumbnail list and selecting Apply Settings to All is what you are looking for. That way you can turn on enhance for one and it will apply it to everything without you have to manually toggle them.
Would like to adjust face selection to add or subtract areas.
Just a bit of feedback:
- I love the new feature (from 1.3.7, I believe) that allows us to scroll thru the bulk load images with the right and left arrow keys on the keyboard. It has been working perfectly, and it has eased my workflow.
- It appears that Face Recovery is much better at discerning between freckles and noise. Freckles are rendered better. They are not eliminated or freakishly sharpened.
- Next: it would help my workflow if the pixel dimension that I input on the preference screen would apply to both enlarging and shrinking the overall dimensions of the original image. NB: I do not use the upscale feature, because I require a specific dimension on the long side.
Is there any way of testing one of your photos at high iso? I’d like to test it in Topaz Denoise v3.7.2 just to compare with the same settings.
Been using Topaz Denoise for a while and just upgraded to Photo AI. Unfortunately when trying to install it I get the message “The older version of Topaz Photo AI canno be removed. Contact your technical support group.” I think it might be conflicting with the free trial version, but I have deleted all folders and registry entries containing “Topaz” from my computer and am still having this issue. Any ideas what could be causing it?