DxO Legacy DOP and current PL long-term compatibility

[Digital Nigel]

I've been a long-time DxO user, starting with DxO Optics Pro 7 in late 2011. I've purchased every version since then, and am currently evaluating PhotoLab 9.

I thought it might be amusing to take an old image I'd originally processed in DOP7, and re-process it in PL9. I first processed it using exactly the same settings I'd used with DOP7 to see if PL9 produced the same result, and then re-edit it to my taste today. In neither case did I do any perspective adjustment or cropping.


Processed in DOP7 in 2012 (resized to 3000 wide)

I then opened it in PL9, and made no changes to any settings (so it used the same dop sidecar file as DOP7 had produced):


Processed in 2025 in PL9, using the original DOP7 settings (full-size)

They're very similar, but not quite identical. I think PL9 did a slightly better job, even with the same settings. Nevertheless, it's nice that PL9 is pretty much fully compatible with its predecessor from 14 years earlier. But it's also a lesson that simply using a newer version doesn't give you much improvement -- you do need to actually take advantage of the new features.

I then created a virtual copy, and edited it as I would today. In fact, I didn't use any PL9-specific settings, such as AI masks, so I could have done it in PL8, too:


Processed in PL9 as I would today

 

[Richard James]

Nigel,

I've been a long-time DxO user, starting with DxO Optics Pro 7 in late 2011.

Optics Pro v1 or 2....

I thought it might be amusing to take an old image I'd originally processed in DOP7, and re-process it in PL9.

I then opened it in PL9, and made no changes to any settings (so it used the same dop sidecar file as DOP7 had produced):

They're very similar, but not quite identical.

I would expect that.

Nevertheless, it's nice that PL9 is pretty much fully compatible with its predecessor from 14 years earlier.

I am aware of at least one incompatibility. The syntax of one line in the DOP file was changed at some time. A ' was replaced by " (or vice versa). It is editable, but a PITA if you need to do many images.

But it's also a lesson that simply using a newer version doesn't give you much improvement -- you do need to actually take advantage of the new features.

Yes, It would be unusual if new features/settings were automatically implemented.

What is more problematic is being able to go back and use an older version on a file processed in a newer version. That would require ignoring (skipping) unrecognized settings.

Related, there is a big comparability problem with the NIK Collection. You can't go back and change a previously edited file with the current version or reinstall a previous version (the installer for earlier versions removed other versions).

I have been asked to demonstrate how I created an image... I have to recreate it and "guess" the setting that I used.

Richard

 

[CAcreeks]

Much better colors, especially in the sky. Did you use camera profile or film emulation?

Everything is skinnier in the newer PhotoLab. Can you explain that?

 

[Digital Nigel]

Much better colors, especially in the sky. Did you use camera profile or film emulation?

That’s partly because I now use Leica M colours. The original rendering just used the camera’s native colours. I also used ClearView Plus, which wasn’t available in 2012, and reduced highlights more (that was available in 2012, but I’d not used it).

Everything is skinnier in the newer PhotoLab. Can you explain that?

Not everything, just things at the edges. That’s because I enabled Volume Deformation correction in the new version. I don’t think that was available in 2012, so I didn’t use it in the first reprocessing.

 

[CAcreeks]

Much better colors, especially in the sky. Did you use camera profile or film emulation?

That’s partly because I now use Leica M colours. The original rendering just used the camera’s native colours. I also used ClearView Plus, which wasn’t available in 2012, and reduced highlights more (that was available in 2012, but I’d not used it).

Did you try out lots of camera profiles before you selected Leica M? It is quite a bit darker than Fujifilm camera profiles. Lately I've found myself selecting DxO Neutral quite often. Even for portraits, I often prefer it to Astia film emulation.

Everything is skinnier in the newer PhotoLab. Can you explain that?

Not everything, just things at the edges. That’s because I enabled Volume Deformation correction in the new version. I don’t think that was available in 2012, so I didn’t use it in the first reprocessing.

Thanks for that. I did not notice lack of center skinniness.

Looking forward to your assessment of PhotoLab 9.

 

[Digital Nigel]

Much better colors, especially in the sky. Did you use camera profile or film emulation?

That’s partly because I now use Leica M colours. The original rendering just used the camera’s native colours. I also used ClearView Plus, which wasn’t available in 2012, and reduced highlights more (that was available in 2012, but I’d not used it).

Did you try out lots of camera profiles before you selected Leica M? It is quite a bit darker than Fujifilm camera profiles.

Yes, I tried lots of them. Initially I chose a different Leica model, but someone here (I forgot who) suggested the change.

Lately I've found myself selecting DxO Neutral quite often. Even for portraits, I often prefer it to Astia film emulation.

Everything is skinnier in the newer PhotoLab. Can you explain that?

Not everything, just things at the edges. That’s because I enabled Volume Deformation correction in the new version. I don’t think that was available in 2012, so I didn’t use it in the first reprocessing.

Thanks for that. I did not notice lack of center skinniness.

Looking forward to your assessment of PhotoLab 9.

I like the AI masks, but if I use more than a couple in one image, the export fails with an Internal errors (probably a catch-all message), and I have to re-start PL9 (it doesn’t crash, but once it gets the error, I know all subsequent exports will also fail). So, for the moment at least, I’m only using them when the other masks won’t do the job. That really needs fixing urgently. I also find they’re not so good at recognising things like faces.

I also like the way that sub-masks can be combined or subtracted and how NR and sharpness settings can vary across local adjustments. You can now see auto-masks properly. So, apart from the extreme fragility of the AI masks, it’s a big improvement in local adjustments.

Something else I’m glad they’ve finally improved is the way you can now have both prefixes and suffixes in exported file names, and the lengths can be much greater.

You can now display the whole image using DeepPRIME, but I find that this is fragile as well, so I’ve disabled the option.

Performance seems OK, but I’ve not yet benchmarked it against PL8. It looks like exports are slower to start, but then run at the normal speed.

I hope the dreaded Internal errors are caused by fixable bugs, and not the inevitable result of my GPU only having 8GB. If the latter, I won’t get much benefit from an upgrade until I get my next PC.