DXO photolab!

PhotoLab takes about 30-50% longer to export a RAW file to which PRIME noise reduction - and nothing else - has been applied. Ouch.

--
If you think digital is hard, try slide film.
I was a slide shooter for many years; if digital hadn't been available when I got back into photography, I'd still be. :-)
DXO has always been slow with PRIME. It reflects the processing going on.
Yes. This is a screenshot of Process Explorer's CPU graph right after DxO PhotoLab processed a few images with PRIME NR. It shows how intense the processing is, and also how effectively multithreaded DxO is.

5d58bd79b5bb4a12a2fbfc4794b3ae27.jpg

Adding more CPU cores/threads should speed things up quite a bit.

Good news for the Intel Skylake-X and AMD Threadripper users, I think.
 


Adding more CPU cores/threads should speed things up quite a bit.

Good news for the Intel Skylake-X and AMD Threadripper users, I think.
Is this a situation where having an i7 CPU instead of an i5 in my iMac would be a benefit?
 
You will need to create your own panels then drag individual adjustments into them. Then hide unwanted default panels (which can't be edited).

In DxO Optics Pro 11, instead of default panels I have customised: WB/Colour, Tone/Contrast, Geometry/Crop, Noise/Clearview, Rendering.
 
Adding more CPU cores/threads should speed things up quite a bit.

Good news for the Intel Skylake-X and AMD Threadripper users, I think.
Is this a situation where having an i7 CPU instead of an i5 in my iMac would be a benefit?
I know virtually nothing about Macs or macOS, but as I understand the Process Explorer graph, if this particular i7 has hyperthreading it should be faster than a similar i5 of the same generation and clock rate that lacks hyperthreading.
 
I can only conclude I'll need to watch some tutorials to understand what it can do and what not.
Come on - I've been using Nik software for 3 years and understand what it can do.
I did not mean to imply any criticism of your explanation at all, just of my limited understanding after playing with PhotoLab for just an hour. I have used the Nik plugins in PSE, just never this kind of local adjustment tool (PSE still seems vastly superior for doing any of the local edits that interest me personally).
In contrast, the third image shows the 'display mask' from Nik Collection (Color Efex Pro 4). Notice how it is a totally different display from that provided by Photolab.

It is this display that is missing from Photolab.

a52497d1013c463883187126a9bcb23e.jpg.png
Very clear; thanks!

--
Mark
 
Worth taking a look at this video.

 
Worth taking a look at this video.

Thanks! I managed to get PhotoLab do something like that once out of the three times I tried, but it's a beginning.
 
It's gotten a lot slower. Let's hope they improve performance on the next point update. Soon.
Sounds like a Mac issue. Both versions run at about the same speed on my 4-year-old PC, under 30 seconds for a Canon 5D3 raw to 16-bit TIFF. If anything, PhotoLab might be slightly faster, but I'd have to do additional testing to be certain.
That's good to hear. Sad, though, that DxO seems to have fallen down on the Mac version. Seems they're giving all their love to Windows. Also, I remember how they claimed OP11 was faster than OP10, but on my Mac it wasn't. I sure hope they'll get PL at least to OP11 parity soon with an update.
 
They must sell a different version in Germany. I see no change in Photolab compared to V11 except a new box in the upper right. They must have misplaced the font library as it has a different appearance. Up comes a wheel and a text box which feels way different then the rest of the software, like some kind of hack. This is the result of year long development? Makes Lightroom shine again.
 
The DxO Photolab's implementation of the control points it's louzy, to say the least. e.g. if you push the exposure way up or way down it acts almost exclusively to the place where your center point lays and not inside the complete circle. Besides the control points (up to now very far from the efficiency of the ones used by Nik) I have not seen any implementation of the Nik plug-ins. Hoping for the future and crossing our fingers!

Claudio
Claudio,

I noticed that "spotlighting" effect as well. I have confirmed with DxO Support that there is no aggregate mask depiction in PL yet. To DxO's credit, I received a very prompt reply even though I haven't actually bought the software. The support person indicated that my query would be forwarded to the developers upon close of the ticket. I replied stating the importance of this issue and asked that the developers read this thread.

Lloyd
 
I can only conclude I'll need to watch some tutorials to understand what it can do and what not.
Come on - I've been using Nik software for 3 years and understand what it can do.
I did not mean to imply any criticism of your explanation at all,
No worries. In any event, this one falls squarely on me.

I'm getting a bit too involved with on-line stuff since the Adobe announcement :( Partly as a result, I've become less careful in reading comments and posts. When I first read your response to me, I thought you were telling me to go watch some videos to improve my understanding. People have been throwing those things around a lot over the past week. Hence, my mis-guided reaction.

Clearly, your comment was quite different than I had thought and the tone of my response was inappropriate. I apologize fully for that :( I wish you all the best
 
I have to say that the exposure of the fur is changed, inside and FAR outside the circles. Not by pixel peeping, it clear and obvious to me the fur's exposure goes up, not as much at the tongue, but way too much, and way too far outside the circles. I'd consider this implementation unusable if for every control point, 10 negative points need to be set up.
 
Last edited:
It does not indicate the area being effected , but as far as I can remember NIK didn't either.
It sure did, has an option to show the mask, and the intensity of the mask as it feathered out. It was critical to use to place appropriate negative points.
 
I have to say that the exposure of the fur is changed, inside and FAR outside the circles. I'd consider this implementation unusable if for every control point, 10 negative points need to be set up.
Having seen your second post, my comments may be more appropriate directed to the larger thread audience.

Different tools with different uses. My example wasn't meant to show off Nik capabilities - I just threw some points on the image to show the difference in mask display. And don't forget - my third image is showing the mask color, not the actual color of the dog's fur.

The image I created had several 'positive' points and one 'negative' one (on the tongue). I would have added some negative ones on the fur is I were doing this to create an image. The dog's fur was not 100% black but contained areas of white and some whiter toned pixels. Since Nik's masking is based around tonality not just geometry, the effect can reach out to areas you may not expect - hence the need for negative control points. There can be both a positive and negative to the method.

The number of points needed depends on the nature of the image and the impact of the adjustment. Sometimes, moving a point just slightly leads to a very different mask. But, I would tend to use 10-15 points for most images to define the masking I wanted to use.

When I use the Nik Interface, I often work (at least in the final stages) with the mask display enabled as I add points. I will know where I want the effect to appear and will just add control points (and adjust their size) until I get the mask that I want. This is one reason why people are concerned by the absence of the mask display in Photolab - you lose that workflow and are adding circles without any feedback of the effect of your placement.

In the original Nik interface, you can add points very easily, using the 'opt' key (with a mouse drag) to copy an existing point to a new location. Then, you can group them to adjust them as a unit. It looks like the DxO interface assumes that all positive points are grouped until you define a new mask. The 'opt' key option works but, I couldn't find a way to delete a point that I didn't need or was badly placed. The original interface also allows for the control points to be applied with multiple adjustments (e.g. glow effects, , not just the few shown in the DxO list.

It definitely takes time to develop a feel for the different approach of the Control points. If it works for you, that's great. if it doesn't, that's great too - there are lots of other adjustments that can do the same thing.

Hopefully, DxO will expand the utility when they get past Version 1.0.

--
Nick
 
Last edited:
My experience with the nik plugins was that the mask size was effected by the size of the control point. For example, if changing the color of a sky, you'd need a bunch of linked control points. In the 2nd picture, the control point on a red bow was increasing exposure all the way to the end of the image of dark fur.
 
My experience with the nik plugins was that the mask size was effected by the size of the control point. For example, if changing the color of a sky, you'd need a bunch of linked control points. In the 2nd picture, the control point on a red bow was increasing exposure all the way to the end of the image of dark fur.
Yes, I agree with both of your comments. When I tried it again just now, I needed to add about 10 negative control points to prevent lightening of the fur and other dark areas. I also did increase exposure to maximum which would accentuate the longer range effect.
 
Could be the LR replacement I'm looking for!
I thought you needed DAM functionality?

Ian
This doesn't have DAM?
I thought you had looked at DXO and said it was no good because it did not have local edits? Didn't you notice their was no DAM?
No DAM is one reason I adopted DxO years ago. I didn't want my image management tied to any one RAW processor; I see that as an entirely different
Interesting! I never thought of it this way (longtime LR user). Can you please tell me how your workflow is ? Step by step? Importing raw files from sd card to your main hard drive...and then....?

thank you!
Lightroom (and others) DAM looked to me like a tool to lock in users to one company's software. But I can't deny it was effective at doing that for Adobe. ;-)

After an hour or so playing with DxO PhotoLab, I decided to go ahead and buy it; $69 for the upgrade from OP 10 (I skipped OP 11).
 
The DxO Photolab's implementation of the control points it's louzy, to say the least. e.g. if you push the exposure way up or way down it acts almost exclusively to the place where your center point lays and not inside the complete circle. Besides the control points (up to now very far from the efficiency of the ones used by Nik) I have not seen any implementation of the Nik plug-ins. Hoping for the future and crossing our fingers!

Claudio
Claudio,

I noticed that "spotlighting" effect as well. I have confirmed with DxO Support that there is no aggregate mask depiction in PL yet. To DxO's credit, I received a very prompt reply even though I haven't actually bought the software. The support person indicated that my query would be forwarded to the developers upon close of the ticket. I replied stating the importance of this issue and asked that the developers read this thread.

Lloyd
Many thanks Lloyd!

Claudio
 
PhotoLab takes about 30-50% longer to export a RAW file to which PRIME noise reduction - and nothing else - has been applied. Ouch.

--
that is the deal breaker. Prime in OpticsPro already slow

even slower? now more than 1 minute a pic?
Don't jump to conclusions, try it out for yourself. For me there was just a 1 second difference between exporting from OP11 and from PhotoLab, with Prime Noise Reduction activated. Took about 25 seconds.
 

Keyboard shortcuts

Back
Top