Learning to use RAW files

Ab S wrote:

..regarding NR I fully agree: DxO DeepPrime or Topaz. And I also agree that in PL you have it built in..

It's not just that it's built in, but is done as part of the demosaicing process. Even if you set DeepPRIME to 0, you still get a better result than if you don't use DeepPRIME at all. And dxo isca very good raw developer, arguably the very best available. In contrast, Topaz DeNoise is a very poor raw developer, so it's best applied to files that have already been demosaiced in a proper raw developer. .

So you end up using multiple products with transfer files in between, and after all that complexity, you end up with worse quality images than you'd have got effortlessly, in a single step, in PhotoLab.

So I only use DeNoise AI to add a final polish to very high ISO images that have already been processed in PhotoLab. But that's seldom needed.

JeremyB2 wrote:

I also need to look into perhaps buying ViewPoint to help slim the 'fatties' at the outer edges of photos containing people. I wonder if it will become available at a discounted price around November?

Viewpoint is included with PL5 so no need to buy it separately.

Michael Fritzen wrote:

Hi,

in fact I notice in C1 the distortion towards the sides which you mentioned.

It seems to me though that this is something coming from the lens already, even more so with the f/2.8 aperture not covering entirely the focus differences from centre to the sides.

The lens needs corrections. DXO PL5 does a much better job applying those lens corrections. Lens corrections have always been a DXO forte'.

Guy Parsons wrote:

Digital Nigel wrote:

And this is how DxO fixes the fattened people at the edges:

DxO Viewpoint is built into PhotoLab, and allows the wide angle effect of fattened people at the edges to be automatically fixed

Viewpoint is a separate program that installs as both a plugin for programs and also available as a standalone program for jpegs and 8/16 bit tiffs.

Just now I opened Viewpoint as standalone and it wanted to update to V3.3. I don't get any auto update messages for Viewpoint when opening Photolab. User manual at https://download-center.dxo.com/Support/docs/ViewPoint_v3/user-guide/VP3_manual_EN.pdf

PL5 gets updates all the time. When it's updated the latest Viewpoint lens corrections are included.

tbcass wrote:

JeremyB2 wrote:

I also need to look into perhaps buying ViewPoint to help slim the 'fatties' at the outer edges of photos containing people. I wonder if it will become available at a discounted price around November?

Viewpoint is included with PL5 so no need to buy it separately.

Unfortunately, it's not. You do need the licence.

tbcass wrote:

Guy Parsons wrote:

Digital Nigel wrote:

And this is how DxO fixes the fattened people at the edges:

DxO Viewpoint is built into PhotoLab, and allows the wide angle effect of fattened people at the edges to be automatically fixed

Viewpoint is a separate program that installs as both a plugin for programs and also available as a standalone program for jpegs and 8/16 bit tiffs.

Just now I opened Viewpoint as standalone and it wanted to update to V3.3. I don't get any auto update messages for Viewpoint when opening Photolab. User manual at https://download-center.dxo.com/Support/docs/ViewPoint_v3/user-guide/VP3_manual_EN.pdf

PL5 gets updates all the time. When it's updated the latest Viewpoint lens corrections are included.

The lens corrections aren't done by ViewPoint. They're part of the core PhotoLab raw processor.

Digital Nigel wrote:

AlwynS wrote:

Digital Nigel wrote:

AlwynS wrote:

Digital Nigel wrote:

Ab S wrote:

Digital Nigel wrote:

Ab S wrote:

tbcass wrote:

Ab S wrote:

If Capture One Express for Sony is still free of charge than that is a perfect way to start with RAW processing. However, I read at DPR that C1 stopped with this free version. But I would say just give it a try.

At this time I would not recommend a full C1 to start with; DxO PL-5 with DeepPrime NR would be a better choice for non-professional use.

Why wouldn't a professional use it? Yes, Photoshop is preferable for publishing but most professionals are not in the publishing industry.

C1 has the tethering option; in C1 you see real time the effect of NR (okay, that would also be possible with DxO non-AI options)..

C1's NR is really inadequate; I'm often shocked at the poor quality images it produces. I couldn't tolerate using C1 for that reason, quite apart from the clunky UI.

Tom, I am sorry to say so but I am really afraid that you did not master to use C1..

I'm sure you're right.

C1 seemed to take much more effort to use, and produced worse quality results, so I saw no point in wasting time mastering it. PhotoLab just seemed a much better solution: much easier to use, highly productive workflow, better quality results. So why would I want to master C1?

Should a person not have bought any of these programmes yet, I consider C1 to be an excellent FREE piece of software as an entrée to RAW processing. I found it very easy to get used to, the workflow is quite simple and quick and I find the results very good in by far the majority of cases. I am curious as to your implied comment that PL-5's workflow is better: seeing as I am not that familiar with PL-5, could you expand on in which way you found it superior to C1? I really find C1 very straight forward for my use, so I find it interesting that you would consider PL-5 superior in that regard.

PL includes a number of best-of-class capabilities that can be invoked automatically, leaving very little additional editing effort required. For example, try this raw file in C1:

https://www.dropbox.com/s/3l718w6w38ogaeb/Ship%27s%20crew%20farewell%20parade%20RX603262.ARW?dl=0

This is what I got from it simply by opening the raw file in PL5, and exporting the processed result to a JPEG. Simply opening it causes my default preset to be applied, and I can then do further tweaks as needed. I have all my commonly used tools arranged in a logical order in my custom palette, so if I do want to adjust any setting it's really quick and easy. I normally do make a few tweaks, but in this case I deliberately didn't make any manual adjustments at all. So this is the zero-click option:

My default preset applied automatically, with no further tweaks or adjustments

My guess is that you will find it much harder to edit this in C1, and simply won't be able to properly replicate this result, however much you try.

In principle: human nature being what it is, I would be quite surprised if most people would not find the programme that they happened to start with and are familiar with the easiest to use and the other daunting. Any relatively powerful piece of software has a learning curve and unless there is a compelling reason to change, not something to be taken on lightly.

Yes, agreed. But knowing the poor quality results that C1 produces, I saw no point in making that effort.

I tried the trial version of PL-5 earlier this year and I did not like it... most likely because I was proficient with C1 and comfortable with its UI and workflow. I accept that PL-5 could give better noise reduction, upscaling (not something I do) and sharpening.

Yes, PL5 can do upscaling, but it's not very good at it. If you want the highest quality, you need to use something like the Topaz apps. For example, here's a raw file I took last month of a distant polar bear:

https://www.dropbox.com/s/sgg4rssk9asnz7g/Distant%20polar%20bear%2C%20Eolusneset%2C%20Svalbard%20R1001925.ARW?dl=0

I took a small (2.31mp) crop of the image, and upsized it 2x (9.63mp) using both bicubic sharper in PL5 and Topaz Sharpen and Gigapixel AI:

Upsized 2x in PL5

Upsized using the Topaz apps

I suggest comparing these at 100%

But because C1 gives me results that I am happy with 99% (an arbitrary number... ) of the time and changing would cost me US$200+, so, just as you could not be bothered to change to C1, so I could not be bothered to change to PL-5. FOR ME, the improvements that I could potentially get in a relatively small number of cases did not justify the price plus going through the learning curve.

I found it the prefect solution to a non-existent problem

Non-existent for you perhaps, but C1 simply wouldn't be able to produce results that I find acceptable, even despite the extra effort those inadequate results require.

Thanks Nigel, appreciate the very good response. For me, the takeaways from this are:

1. As expected, noise reduction is way, way better than C1. This is not something I normally use much so not such a big deal for me. However, I agree that this would absolutely be a critical differentiator if you expect to/need to do significant noise reduction.

Yes, I apply DeepPRIME to every image I process.

2. No click edit is good. Closest C1 can come is two clicks with a style that you define and save for future use. Again on the personal needs/preference thing: I tend to treat each image individually by choice. If I do have a burst that has very similar/identical characteristics and which I want to treat similarly, I copy and apply settings from an edited image in the burst.

Yes, I do the same. With PL5, you can copy all the efit settings, then paste some or all of the settings to one or a range of images. So I start with the default preset, make a few adjustments as required, then paste some or all of them to all the similar images. I then process them in the background, while moving on to the next group of images. It's a highly productive workflow.

And then back to the personal requirements thing: I totally agree that PL is superior in certain respects, particularly noise reduction. At this time, based on my photography and what I want to do, those features, while significant, are not sufficiently important in my use case for me to pay $200 and go through the learning to make the change. Maybe if it comes up on a special at 50% off I might be tempted.

That should happen in just over three months, in the usual BF sale. It will apply to the next version, PL6.

But at that time I would have to do a careful review/comparison of for instance individual Topaz tools vs. DxO.

I use both, as they have different strengths. I also use Luminar Neo, though I currently treat that more as a toy, and Affinity Photo.

I would be interested in seeing the results that C1 produced in each of these cases. Almost certainly, poor NR is not the only problem. For example, I'm pretty sure there would be at least two other significant problems in its rendering of the first image.

I did a quick conversion of your RAW in C1 without taking a lot of effort to get it to look similar to your processing. An immediate difference that jumps out is the wider view that the PL conversion provided.

For images like this, PL is undoubtedly superior.

AlwynS wrote:

Digital Nigel wrote:

AlwynS wrote:

Digital Nigel wrote:

AlwynS wrote:

Digital Nigel wrote:

Ab S wrote:

Digital Nigel wrote:

Ab S wrote:

tbcass wrote:

Ab S wrote:

If Capture One Express for Sony is still free of charge than that is a perfect way to start with RAW processing. However, I read at DPR that C1 stopped with this free version. But I would say just give it a try.

At this time I would not recommend a full C1 to start with; DxO PL-5 with DeepPrime NR would be a better choice for non-professional use.

Why wouldn't a professional use it? Yes, Photoshop is preferable for publishing but most professionals are not in the publishing industry.

C1 has the tethering option; in C1 you see real time the effect of NR (okay, that would also be possible with DxO non-AI options)..

C1's NR is really inadequate; I'm often shocked at the poor quality images it produces. I couldn't tolerate using C1 for that reason, quite apart from the clunky UI.

Tom, I am sorry to say so but I am really afraid that you did not master to use C1..

I'm sure you're right.

C1 seemed to take much more effort to use, and produced worse quality results, so I saw no point in wasting time mastering it. PhotoLab just seemed a much better solution: much easier to use, highly productive workflow, better quality results. So why would I want to master C1?

Should a person not have bought any of these programmes yet, I consider C1 to be an excellent FREE piece of software as an entrée to RAW processing. I found it very easy to get used to, the workflow is quite simple and quick and I find the results very good in by far the majority of cases. I am curious as to your implied comment that PL-5's workflow is better: seeing as I am not that familiar with PL-5, could you expand on in which way you found it superior to C1? I really find C1 very straight forward for my use, so I find it interesting that you would consider PL-5 superior in that regard.

PL includes a number of best-of-class capabilities that can be invoked automatically, leaving very little additional editing effort required. For example, try this raw file in C1:

https://www.dropbox.com/s/3l718w6w38ogaeb/Ship%27s%20crew%20farewell%20parade%20RX603262.ARW?dl=0

This is what I got from it simply by opening the raw file in PL5, and exporting the processed result to a JPEG. Simply opening it causes my default preset to be applied, and I can then do further tweaks as needed. I have all my commonly used tools arranged in a logical order in my custom palette, so if I do want to adjust any setting it's really quick and easy. I normally do make a few tweaks, but in this case I deliberately didn't make any manual adjustments at all. So this is the zero-click option:

My guess is that you will find it much harder to edit this in C1, and simply won't be able to properly replicate this result, however much you try.

In principle: human nature being what it is, I would be quite surprised if most people would not find the programme that they happened to start with and are familiar with the easiest to use and the other daunting. Any relatively powerful piece of software has a learning curve and unless there is a compelling reason to change, not something to be taken on lightly.

Yes, agreed. But knowing the poor quality results that C1 produces, I saw no point in making that effort.

I tried the trial version of PL-5 earlier this year and I did not like it... most likely because I was proficient with C1 and comfortable with its UI and workflow. I accept that PL-5 could give better noise reduction, upscaling (not something I do) and sharpening.

Yes, PL5 can do upscaling, but it's not very good at it. If you want the highest quality, you need to use something like the Topaz apps. For example, here's a raw file I took last month of a distant polar bear:

https://www.dropbox.com/s/sgg4rssk9asnz7g/Distant%20polar%20bear%2C%20Eolusneset%2C%20Svalbard%20R1001925.ARW?dl=0

I took a small (2.31mp) crop of the image, and upsized it 2x (9.63mp) using both bicubic sharper in PL5 and Topaz Sharpen and Gigapixel AI:

I suggest comparing these at 100%

But because C1 gives me results that I am happy with 99% (an arbitrary number... ) of the time and changing would cost me US$200+, so, just as you could not be bothered to change to C1, so I could not be bothered to change to PL-5. FOR ME, the improvements that I could potentially get in a relatively small number of cases did not justify the price plus going through the learning curve.

I found it the prefect solution to a non-existent problem

Non-existent for you perhaps, but C1 simply wouldn't be able to produce results that I find acceptable, even despite the extra effort those inadequate results require.

Thanks Nigel, appreciate the very good response. For me, the takeaways from this are:

1. As expected, noise reduction is way, way better than C1. This is not something I normally use much so not such a big deal for me. However, I agree that this would absolutely be a critical differentiator if you expect to/need to do significant noise reduction.

Yes, I apply DeepPRIME to every image I process.

2. No click edit is good. Closest C1 can come is two clicks with a style that you define and save for future use. Again on the personal needs/preference thing: I tend to treat each image individually by choice. If I do have a burst that has very similar/identical characteristics and which I want to treat similarly, I copy and apply settings from an edited image in the burst.

Yes, I do the same. With PL5, you can copy all the efit settings, then paste some or all of the settings to one or a range of images. So I start with the default preset, make a few adjustments as required, then paste some or all of them to all the similar images. I then process them in the background, while moving on to the next group of images. It's a highly productive workflow.

And then back to the personal requirements thing: I totally agree that PL is superior in certain respects, particularly noise reduction. At this time, based on my photography and what I want to do, those features, while significant, are not sufficiently important in my use case for me to pay $200 and go through the learning to make the change. Maybe if it comes up on a special at 50% off I might be tempted.

That should happen in just over three months, in the usual BF sale. It will apply to the next version, PL6.

But at that time I would have to do a careful review/comparison of for instance individual Topaz tools vs. DxO.

I use both, as they have different strengths. I also use Luminar Neo, though I currently treat that more as a toy, and Affinity Photo.

I would be interested in seeing the results that C1 produced in each of these cases. Almost certainly, poor NR is not the only problem. For example, I'm pretty sure there would be at least two other significant problems in its rendering of the first image.

I did a quick conversion of your RAW in C1 without taking a lot of effort to get it to look similar to your processing. An immediate difference that jumps out is the wider view that the PL conversion provided.

For images like this, PL is undoubtedly superior.

And just for perspective: quickly grabbing some recent images is an indication of what is more representative of the type of photography I do:

From this you can see why FOR ME, for instance the absolute best in noise reduction and capturing the most possible pixels on the edges of the frame are not significant concerns. For what I do, I find the free C1 to be quite satisfactory.

As always: YMMV

AlwynS wrote:

Digital Nigel wrote:

AlwynS wrote:

Digital Nigel wrote:

AlwynS wrote:

Digital Nigel wrote:

Ab S wrote:

Digital Nigel wrote:

Ab S wrote:

tbcass wrote:

Ab S wrote:

If Capture One Express for Sony is still free of charge than that is a perfect way to start with RAW processing. However, I read at DPR that C1 stopped with this free version. But I would say just give it a try.

At this time I would not recommend a full C1 to start with; DxO PL-5 with DeepPrime NR would be a better choice for non-professional use.

Why wouldn't a professional use it? Yes, Photoshop is preferable for publishing but most professionals are not in the publishing industry.

C1 has the tethering option; in C1 you see real time the effect of NR (okay, that would also be possible with DxO non-AI options)..

C1's NR is really inadequate; I'm often shocked at the poor quality images it produces. I couldn't tolerate using C1 for that reason, quite apart from the clunky UI.

Tom, I am sorry to say so but I am really afraid that you did not master to use C1..

I'm sure you're right.

C1 seemed to take much more effort to use, and produced worse quality results, so I saw no point in wasting time mastering it. PhotoLab just seemed a much better solution: much easier to use, highly productive workflow, better quality results. So why would I want to master C1?

Should a person not have bought any of these programmes yet, I consider C1 to be an excellent FREE piece of software as an entrée to RAW processing. I found it very easy to get used to, the workflow is quite simple and quick and I find the results very good in by far the majority of cases. I am curious as to your implied comment that PL-5's workflow is better: seeing as I am not that familiar with PL-5, could you expand on in which way you found it superior to C1? I really find C1 very straight forward for my use, so I find it interesting that you would consider PL-5 superior in that regard.

PL includes a number of best-of-class capabilities that can be invoked automatically, leaving very little additional editing effort required. For example, try this raw file in C1:

https://www.dropbox.com/s/3l718w6w38ogaeb/Ship%27s%20crew%20farewell%20parade%20RX603262.ARW?dl=0

This is what I got from it simply by opening the raw file in PL5, and exporting the processed result to a JPEG. Simply opening it causes my default preset to be applied, and I can then do further tweaks as needed. I have all my commonly used tools arranged in a logical order in my custom palette, so if I do want to adjust any setting it's really quick and easy. I normally do make a few tweaks, but in this case I deliberately didn't make any manual adjustments at all. So this is the zero-click option:

My default preset applied automatically, with no further tweaks or adjustments

My guess is that you will find it much harder to edit this in C1, and simply won't be able to properly replicate this result, however much you try.

In principle: human nature being what it is, I would be quite surprised if most people would not find the programme that they happened to start with and are familiar with the easiest to use and the other daunting. Any relatively powerful piece of software has a learning curve and unless there is a compelling reason to change, not something to be taken on lightly.

Yes, agreed. But knowing the poor quality results that C1 produces, I saw no point in making that effort.

I tried the trial version of PL-5 earlier this year and I did not like it... most likely because I was proficient with C1 and comfortable with its UI and workflow. I accept that PL-5 could give better noise reduction, upscaling (not something I do) and sharpening.

Yes, PL5 can do upscaling, but it's not very good at it. If you want the highest quality, you need to use something like the Topaz apps. For example, here's a raw file I took last month of a distant polar bear:

https://www.dropbox.com/s/sgg4rssk9asnz7g/Distant%20polar%20bear%2C%20Eolusneset%2C%20Svalbard%20R1001925.ARW?dl=0

I took a small (2.31mp) crop of the image, and upsized it 2x (9.63mp) using both bicubic sharper in PL5 and Topaz Sharpen and Gigapixel AI:

Upsized 2x in PL5

Upsized using the Topaz apps

I suggest comparing these at 100%

But because C1 gives me results that I am happy with 99% (an arbitrary number... ) of the time and changing would cost me US$200+, so, just as you could not be bothered to change to C1, so I could not be bothered to change to PL-5. FOR ME, the improvements that I could potentially get in a relatively small number of cases did not justify the price plus going through the learning curve.

I found it the prefect solution to a non-existent problem

Non-existent for you perhaps, but C1 simply wouldn't be able to produce results that I find acceptable, even despite the extra effort those inadequate results require.

Thanks Nigel, appreciate the very good response. For me, the takeaways from this are:

1. As expected, noise reduction is way, way better than C1. This is not something I normally use much so not such a big deal for me. However, I agree that this would absolutely be a critical differentiator if you expect to/need to do significant noise reduction.

Yes, I apply DeepPRIME to every image I process.

2. No click edit is good. Closest C1 can come is two clicks with a style that you define and save for future use. Again on the personal needs/preference thing: I tend to treat each image individually by choice. If I do have a burst that has very similar/identical characteristics and which I want to treat similarly, I copy and apply settings from an edited image in the burst.

Yes, I do the same. With PL5, you can copy all the efit settings, then paste some or all of the settings to one or a range of images. So I start with the default preset, make a few adjustments as required, then paste some or all of them to all the similar images. I then process them in the background, while moving on to the next group of images. It's a highly productive workflow.

And then back to the personal requirements thing: I totally agree that PL is superior in certain respects, particularly noise reduction. At this time, based on my photography and what I want to do, those features, while significant, are not sufficiently important in my use case for me to pay $200 and go through the learning to make the change. Maybe if it comes up on a special at 50% off I might be tempted.

That should happen in just over three months, in the usual BF sale. It will apply to the next version, PL6.

But at that time I would have to do a careful review/comparison of for instance individual Topaz tools vs. DxO.

I use both, as they have different strengths. I also use Luminar Neo, though I currently treat that more as a toy, and Affinity Photo.

I would be interested in seeing the results that C1 produced in each of these cases. Almost certainly, poor NR is not the only problem. For example, I'm pretty sure there would be at least two other significant problems in its rendering of the first image.

I did a quick conversion of your RAW in C1 without taking a lot of effort to get it to look similar to your processing. An immediate difference that jumps out is the wider view that the PL conversion provided.

For images like this, PL is undoubtedly superior.

As expected, the volume deformation is also present. But, slightly surprisingly, the C1 image is very slightly wider than the OOC JPEG, but noticeably narrower than PL5.

AlwynS wrote:

And just for perspective: quickly grabbing some recent images is an indication of what is more representative of the type of photography I do:

From this you can see why FOR ME, for instance the absolute best in noise reduction and capturing the most possible pixels on the edges of the frame are not significant concerns. For what I do, I find the free C1 to be quite satisfactory.

As always: YMMV

Agreed, the particular strengths of PL5 are less relevant for (excellent)  images like those. However, I think it might still produce slightly better, sharper results  than C1 even in these cases, because of the superior lens correction. The DxO lens corrections include custom sharpening for each lens.

Digital Nigel wrote:

tbcass wrote:

PL5 gets updates all the time. When it's updated the latest Viewpoint lens corrections are included.

The lens corrections aren't done by ViewPoint. They're part of the core PhotoLab raw processor.

Thank you for the information.

Digital Nigel wrote:

tbcass wrote:

JeremyB2 wrote:

I also need to look into perhaps buying ViewPoint to help slim the 'fatties' at the outer edges of photos containing people. I wonder if it will become available at a discounted price around November?

Viewpoint is included with PL5 so no need to buy it separately.

Unfortunately, it's not. You do need the licence.

I see it cost $79. Does it go on sale during Black Friday because I'm not sure it would be worth it to me?

Your point about intuitive is right on. We aren’t born with any innate ability to use software or a camera UI. We learn through practice. Once we master something anything similar will seem “intuitive”.

tbcass wrote:

Digital Nigel wrote:

tbcass wrote:

JeremyB2 wrote:

I also need to look into perhaps buying ViewPoint to help slim the 'fatties' at the outer edges of photos containing people. I wonder if it will become available at a discounted price around November?

Viewpoint is included with PL5 so no need to buy it separately.

Unfortunately, it's not. You do need the licence.

I see it cost $79. Does it go on sale during Black Friday because I'm not sure it would be worth it to me?

Yes, DxO normally offers the discount on its whole product line, and not just the newly updated products.

sybersitizen wrote:

tbcass wrote:

Viewpoint is included with PL5 so no need to buy it separately.

That's incorrect. ViewPoint has always been a different product that must be purchased separately. The last version of DxO's converter/editor that included a subset (but not all) of what ViewPoint does was Optics Pro 9, which I still keep on my system alongside PL5.

https://support.dxo.com/hc/en-us/articles/360048490491-Why-do-I-need-DxO-ViewPoint-as-a-standalone-program-if-I-already-have-DxO-PhotoLab-

Quite a lot of DOP 9 users were pretty annoyed when DxO took fairly basic, essential features out of DOP 10 and put them into a separately purchased product. People wouldn't have objected if they got a discount on ViewPoint, but I don't recall that there was one.

tbcass wrote:

Digital Nigel wrote:

tbcass wrote:

JeremyB2 wrote:

I also need to look into perhaps buying ViewPoint to help slim the 'fatties' at the outer edges of photos containing people. I wonder if it will become available at a discounted price around November?

Viewpoint is included with PL5 so no need to buy it separately.

Unfortunately, it's not. You do need the licence.

I see it cost $79. Does it go on sale during Black Friday because I'm not sure it would be worth it to me?

For the last two years at Black Friday, everything at DxO was 50% off. So $79 becomes $39.50 if they indeed repeat what happened before.

sybersitizen wrote:

JeremyB2 wrote:

I also need to look into perhaps buying ViewPoint to help slim the 'fatties' at the outer edges of photos containing people.

Would you like to know how to closely simulate that particular ViewPoint correction with other software? I recently described the method here:

https://www.dpreview.com/forums/post/66321772

Thanks for that.