Using both Capture One and Lightroom

[Robert Ardill]

I used C1 years ago and changed to LR mainly for the integration with Photoshop and the asset management features.  I recently changed over from Canon to Sony and this has given me the impetus to look at other raw converters again.

I must say that comparing one converter to another is very confusing - and I would very much welcome your own experience with these converters.  Also, I am only comparing the raw conversion features here, not other features like slideshow, maps, books etc as Lightroom is clearly way ahead there.

The first thing I would say is that both have strengths and weaknesses but that these can be quite easily equalized in Photoshop (or with plug-ins). This is fine for me as I always finish off my images in Photoshop, but I've found that there is nevertheless an advantage in using one converter over the other for specific types of images.

So first of all resolution.  I can see absolutely nothing between the two (at least not on Canon 1DsIII and Sony A7RII images).

Sharpening: There's nothing between them IMO. Both use an Unsharp Mask type sharpening (in spite of the claims that LR uses deconvolution when the Details slider is swung one way rather than the other) and they are both OK if that is the only option. However there's no doubt that deconvolution sharpening in Focus Magic or Topaz is significantly better for capture sharpening.

Noise Reduction: again, nothing between them IMO.  Fine for color noise reduction, but I wouldn't use either for serious luminance noise reduction.

Tonal adjustments: On the whole apples for apples here. However if an HDR-look is wanted then the C1 HDR sliders are very good.  On the other hand, LR's Shadow and Highlights sliders gives a more natural (less HDR-ish) look, which in general I prefer.  Having said that, these effects can be achieved using the Levels or Curves tabs, so having the HDR sliders is very nice.

Clarity: Nothing much between them. However C1 has the Structure slider which is clarity at the micro level and it is very good at bringing out small detail (and also noise).

Color Balance: C1 is better here as it gives control over shadows, mid tones and highlights, while LR only gives overall control plus shadow control (in the Camera Calibration tab).  C1 also gives full control whereas LR only gives tint control (plus Red/Green/Blue primary adjustments which are only really useful to correct camera calibration, not to adjust the image hues).

Color Editor: Again C1 has a definite edge as it is makes it much easier to target color ranges (whereas LR only allows changes to specific hues, like yellow and aqua).  The Advanced tab of the C1 Color Editor is particularly powerful.

Local Adjustments: C1 has a major advantage over LR here as far more local adjustments are possible and the use of layers is really good.

Lens Correction: LR is better IMO, partly because it has support for far more lenses and very quickly after they are released, and also because the distortion corrections are more powerful.

Effects: LR has quite a big plus here with the Dehaze slider which can rescue images that would otherwise be trashed.

Camera calibration: In principle I like C1's use of ICC profiles but in practice I think LR's use of DCP profiles which can be quickly and easily modified for different lighting conditions is better.

So these would be my impressions on a detailed level, not taking into account entirely missing modules in C1, and C1's poor integration with Photoshop (no smart objects for one major thing).  But if I had to make a choice between the two, purely for raw conversion and assuming as much as possible to be done in the raw converter, then I would pick C1 because of the color editor and local adjustments.

Having said that, what I've found is that the two converters handle different images quite differently. This is something that has made it very difficult for me to decide which one to opt for.  I've found, again and again, that I prefer the way that one converter renders a particular image and I've found it difficult to achieve the same look with the other converter.  For example, I find that photos of buildings where there is a big dynamic range tend to come out much better in C1; on the other hand, pale seascapes look more natural to me in LR.  Also, creative effects are sometimes much more easily achieved in one converter than in the other (in spite of C1's more advanced color editor) and this seems to be quite image-specific.

If the integration between the two programs had been better (and better integration between C1 and PS), I wouldn't have hesitated to use the two, depending on the image.  But unfortunately the integration is really very poor and I particularly dislike the way that C1 does not keep the sidecar files with the image and doesn't put the adjustments into the DNG files.

So I'll cough up the $50 for the Capture One Pro (for Sony), but I wouldn't pay the full price for Capture One Pro, even though I have a huge number of Canon files that I still need to process.  If Phase One improves the integration with Adobe apps then I will buy the full version because it has some really nice features and does a better job on some images than does Lightroom (and vice-versa).

It would be really good if raw converters became a bit like plug-ins ... in the sense that one could use one converter for a particular type of image for which it was particularly strong, and another converter for another image for which that converter was particularly strong (perhaps using DxO for images needing serious lens corrections, for example, when the normal everyday editor was C1, say).

I would certainly very much welcome a different perspective as I'm still somewhat conflicted, to be honest .

Robert

[Bart_van_der_Wolf]

So first of all resolution.  I can see absolutely nothing between the two (at least not on Canon 1DsIII and Sony A7RII images).

Hi Robert,

I beg to differ. As attached (I've shared this before in this post), a comparison of the same Raw file from my 1Ds Mark III, 400% zoomed. BTW Capture One since version 7 (version 8 and 9 have the same demosaicing resolution on this file). The resolution difference is not huge, after all it's the same file, but there is more resolution in the C1 version 7 and later and it's better behaved (reduced aliasing artifacts). False color artifacts can be suppressed in both converters, but that has not been applied to the example (which was only white balanced and exposure matched) yet.

Sharpening: There's nothing between them IMO. Both use an Unsharp Mask type sharpening (in spite of the claims that LR uses deconvolution when the Details slider is swung one way rather than the other) and they are both OK if that is the only option. However there's no doubt that deconvolution sharpening in Focus Magic or Topaz is significantly better for capture sharpening.

The deconvolution in Lightroom is not particularly good, and pales in comparison with what e.g. RawTherapee has to offer (Richardson-Lucy deconvolution). I do hope that Capture One, who have already improved resampling quality and very useful layers functionality, have deconvolution sharpening on their to-do list. They seem to be more quality focused than Adobe has demonstrated lately, so I'm hopeful. Adobe seems to be more focused on mobile.

Noise Reduction: again, nothing between them IMO.  Fine for color noise reduction, but I wouldn't use either for serious luminance noise reduction.

Capture One's single pixel noise reduction can be useful on long exposure time shots. I use very little noise reduction (shoot mostly at ISO 100), but when needed there are better dedicated tools available (Topaz Denoise, NeatImage, and Noiseware).

Tonal adjustments: On the whole apples for apples here. However if an HDR-look is wanted then the C1 HDR sliders are very good.  On the other hand, LR's Shadow and Highlights sliders gives a more natural (less HDR-ish) look, which in general I prefer.  Having said that, these effects can be achieved using the Levels or Curves tabs, so having the HDR sliders is very nice.

I do not like the highlight tone compression that Lightroom applies by default. It kills the brilliance in images, and requires a heavy minus Highlight control adjustment to restore a more normal look.
Capture One has additional tools for shadows and highlights besides the HDR panel controls. The adjustment layers can use, amongst others, Exposure to locally brighten/darken, and the Luminance curves controls allow to adjust globally (or locally with adjustment layers). Even the revamped brightness and contrast sliders are more intelligent than they were, and are now better at avoiding color changes since C1 version 9.

Clarity: Nothing much between them. However C1 has the Structure slider which is clarity at the micro level and it is very good at bringing out small detail (and also noise).

And it can additionally be applied locally with adjustment layers in C1. But tools like Topaz Clarity and Detail allow more control and higher quality results.

Color Balance: C1 is better here as it gives control over shadows, mid tones and highlights, while LR only gives overall control plus shadow control (in the Camera Calibration tab).  C1 also gives full control whereas LR only gives tint control (plus Red/Green/Blue primary adjustments which are only really useful to correct camera calibration, not to adjust the image hues).

Color Editor: Again C1 has a definite edge as it is makes it much easier to target color ranges (whereas LR only allows changes to specific hues, like yellow and aqua).  The Advanced tab of the C1 Color Editor is particularly powerful.

Yes, very powerful and high quality Color control. Additionally useful is the ability to create an adjustment mask based on a color editor selection. That is a huge timesaver for difficult selections (e.g. blue sky through branches or a selection in a variety of vegetation types).

Local Adjustments: C1 has a major advantage over LR here as far more local adjustments are possible and the use of layers is really good.

Yes, super powerful and very useful, and it means that less work needs to be done in Photoshop.

Lens Correction: LR is better IMO, partly because it has support for far more lenses and very quickly after they are released, and also because the distortion corrections are more powerful.

That's possible but I rarely use lens corrections. They can be useful on architecture shots, but when e.g. stitching is required the the most important geometrical distortion corrections are automatically applied then, and with a higher resampling quality. Distortion correction usually means a loss of resolution, and without deconvolution sharpening that may be a high price to pay.

Effects: LR has quite a big plus here with the Dehaze slider which can rescue images that would otherwise be trashed.

It's an interesting effect and it can be applied quickly, but there are other solutions available when we also use Photoshop Topaz Clarity in our workflow. I have not used LR Dehaze enough, but I think it tends to also affect saturation, so it should be used with restraint.

Camera calibration: In principle I like C1's use of ICC profiles but in practice I think LR's use of DCP profiles which can be quickly and easily modified for different lighting conditions is better.

The automatic interpolation of profiling between tungsten and daylight illuminant settings is nice. Capture One has its roots in Professional photography, which used to pay more attention to (artificial) lighting, thus a better controlled starting point, with dedicated profiles. Capture One does offer a choice from several profiles as a starting point, and I believe they go further than setting the color temperature/tint, so choosing the closer profile as a basis for further corrections may help. Capture One Pro can also create a tweaked ICC profile with the Color Editor quite easily, and that adds a lot of flexibility.

Sorry for the long post, but there's a lot to be discussed.

Cheers,
Bart

[Robert Ardill]

I beg to differ. As attached (I've shared this before in this post), a comparison of the same Raw file from my 1Ds Mark III, 400% zoomed. BTW Capture One since version 7 (version 8 and 9 have the same demosaicing resolution on this file). The resolution difference is not huge, after all it's the same file, but there is more resolution in the C1 version 7 and later and it's better behaved (reduced aliasing artifacts). False color artifacts can be suppressed in both converters, but that has not been applied to the example (which was only white balanced and exposure matched) yet.

That's interesting ... and there are clear differences.  It would be interesting to try this file on Lightroom 6/CC (which I could do if you send me the file).

I only did a visual check on landscape photographs and I must say that I couldn't see any differences at high frequencies.  But I'm not so sure at lower frequencies.  I remember looking at a C1-rendered photo and thinking that it looked definitely sharper than the LR version, but when I examined the detail I couldn't see any difference.  So it may be that C1 also does a better job at MTF 100 to 70, say.  I don't know when I'll be able to do this, but I will at some stage run some tests.  It isn't immediately obvious from your test image, especially as it cycles quite quickly, but lower-frequency contrast on the C1 v 7 image does seem better.

The deconvolution in Lightroom is not particularly good, and pales in comparison with what e.g. RawTherapee has to offer (Richardson-Lucy deconvolution). I do hope that Capture One, who have already improved resampling quality and very useful layers functionality, have deconvolution sharpening on their to-do list. They seem to be more quality focused than Adobe has demonstrated lately, so I'm hopeful. Adobe seems to be more focused on mobile.

Do you really think that Lightroom does deconvolution sharpening?  I can't see it.  All I've found on the web regarding deconvolution sharpening in LR is that it kicks in when the Detail slider is over 50% "AFIK".

If I put the Detail slider to 100% and get the best sharpening I can (on the image below it works out around Amount 40, Radius 1.8 ) I get the effect shown below (LR sharpened image on the right).  If I then bring the Detail slider back to 0 and increase it in increments of 10 all that I can see is that the haloing and noise increases linearly.  I don't see the sudden jump to artifacts that you get in FM for example.  It just looks to me like USM.

Unsharpened image on the left, FM sharpened in the middle, LR 'deconvolution-sharpened' on the right:

I do not like the highlight tone compression that Lightroom applies by default. It kills the brilliance in images, and requires a heavy minus Highlight control adjustment to restore a more normal look.

Interesting observation, although I'm not sure I agree (or understand).  If you look at the image below, I've shown a scene with bright highlights. Both histograms are pretty much at the same position at the highlight and shadow ends. The C1 histogram is overall a bit flatter.  There is a bit more contrast in the light blue sky in the LR version.  The main difference I see though is that the LR shadows are opened up a bit more (which is probably why the C1 histogram is flatter).  All settings at default/standard.

It's an interesting effect and it can be applied quickly, but there are other solutions available when we also use Photoshop Topaz Clarity in our workflow. I have not used LR Dehaze enough, but I think it tends to also affect saturation, so it should be used with restraint.

Yes, I did make the point that a lot of the differences between the two can be evened out once we go to Photoshop (or equivalent).  The same goes for the local adjustments in C1 ... they are way better than LR's, but they can be done just as well or better in Photoshop (although not in as unified a way: Photoshop requires many layers to do what can be done in one layer in C1, since many adjustments can be applied to one layer in C1 (I'm assuming non-destructive editing)).

The automatic interpolation of profiling between tungsten and daylight illuminant settings is nice. Capture One has its roots in Professional photography, which used to pay more attention to (artificial) lighting, thus a better controlled starting point, with dedicated profiles. Capture One does offer a choice from several profiles as a starting point, and I believe they go further than setting the color temperature/tint, so choosing the closer profile as a basis for further corrections may help. Capture One Pro can also create a tweaked ICC profile with the Color Editor quite easily, and that adds a lot of flexibility.

The use of ICC profiles in C1 was one of the things that really interested me a great deal.  However it does seem that making a really good profile is not such an easy job, based on feedback from another topic I started here last week.  Personally I don't see why that should be the case (after all, we make profiles for printers as a matter of course and printers are far more complex than cameras) - so I will give this a go when I have the time.

Sorry for the long post ...

On the contrary ... thank you for taking the time and for your valuable input!

Cheers,

Robert

[Robert Ardill]

Well, here's a pretty significant resolution test:



The C1 image is way better from MTF 100 right to MTF 40 which explains the apparent sharpness of the C1 images and the more contrasty look compared to LR images.  The LR image is a little better at Nyquist but that could well be artifacts as the curve gets a bit jagged.

So is this because of a better demosaicing algorithm, or does C1 apply some sharpening?  What's interesting (to me, that is) is the flatness of the C1 curve at MTF100.  That doesn't look like sharpening.

It's apparent, looking at the edge profiles, that the C1 black point is darker and the white point lighter ... but adjusting the black and white points on the LR image makes very little difference to the MTF curve.

Cheers

Robert

[Bart_van_der_Wolf]

Do you really think that Lightroom does deconvolution sharpening?  I can't see it.  All I've found on the web regarding deconvolution sharpening in LR is that it kicks in when the Detail slider is over 50% "AFIK".

Yes, as confirmed by Eric Chan.

If I put the Detail slider to 100% and get the best sharpening I can (on the image below it works out around Amount 40, Radius 1.8 ) I get the effect shown below (LR sharpened image on the right).  If I then bring the Detail slider back to 0 and increase it in increments of 10 all that I can see is that the haloing and noise increases linearly.  I don't see the sudden jump to artifacts that you get in FM for example.  It just looks to me like USM.

The Detail slider sort of blends between USM at 0 and 'deconvolution' at 100. But the deconvolution is a quick and dirty kind of implementation, and I get serious artifacts above a detail , say, 50.

Interesting observation, although I'm not sure I agree (or understand).  If you look at the image below, I've shown a scene with bright highlights. Both histograms are pretty much at the same position at the highlight and shadow ends. The C1 histogram is overall a bit flatter. There is a bit more contrast in the light blue sky in the LR version.  The main difference I see though is that the LR shadows are opened up a bit more (which is probably why the C1 histogram is flatter).  All settings at default/standard.

Make sure you start with a linear tonecurve in C1, not a film curve (which has a non-linear toe and shoulder, and boosts exposure by about a stop, suggesting clipping when the file is ETTR).

Cheers,
Bart

[Robert Ardill]

Make sure you start with a linear tonecurve in C1, not a film curve (which has a non-linear toe and shoulder, and boosts exposure by about a stop, suggesting clipping when the file is ETTR).

Cheers,
Bart

Yes ... I should have seen that.  I will have also to redo the MTF test as obviously the tone curve will affect it.  Unfortunately I now only have Capture One Pro (for Sony) as my trial has expired and the image I used was a 1DsIII image.

Here's the image with the C1 tone curve set to linear.  I adjusted the white point using the Levels adjustment and then applied a tone curve to taste. As you can see, there is some more detail in the highlights.  However getting a similar level of detail using LR is quite easy.



  How would you typically develop a photo using a linear tone curve as the starting point?

Cheers

Robert

[Bart_van_der_Wolf]

Yes ... I should have seen that.  I will have also to redo the MTF test as obviously the tone curve will affect it.  Unfortunately I now only have Capture One Pro (for Sony) as my trial has expired and the image I used was a 1DsIII image.

Maybe when an update is released you'll get another trial period.

Here's the image with the C1 tone curve set to linear.  I adjusted the white point using the Levels adjustment and then applied a tone curve to taste. As you can see, there is some more detail in the highlights.  However getting a similar level of detail using LR is quite easy.

Yes, but LR starts with a default that I do not like. I need to reduce the Highlight control to  between -100 to -50 to get a more reasonable highlight tonal rendering. Without that, e.g. bright clouds look dull blobs of white, with the -100 Highlights there is cloud formation structure and a more subtle 3D rendering.

How would you typically develop a photo using a linear tone curve as the starting point?

I have C1 set up to start by default with a Style. That style consists of an ICC profile, and a linear curve. The style also has a slight (+5) Contrast boost, zero Brightness, and a slight (+5) Saturation boost (positive values are actually Vibrance, negative values are Saturation), but that's for the profile of my camera and the common contrast of my subject scenes. Other cameras and subjects may benefit from other settings. Also, cameras with a large DR might benefit from other settings than low DR cameras. One can also apply a curve by default should one wish. The style also sets noise reduction to zero. I also have the sharpening set for a value that allows me to better judge sharpness when I zoom the image to 100%, but I have it disabled in the output recipe (so I can use e.g. FocusMagic in Photoshop to do a better job until C1 gets deconvolution sharpening).

I then apply White balance and I adjust Exposure ('Auto' often gives you quite reasonable settings) to almost saturate the histogram highlights (clipping warning is on, the sensitivity is a preset). A reflected white is usually not boosted all the way, but closer to 235 (a Color readout an be added to the image, and then I can scroll my mouse wheel while hovering over the exposure slider till the correct value is displayed). Specular whites and reflected lightsources or white clouds may be pushed towards but just shy of really clipping. Sometimes really clipping some specularity produces a better image (one can always locally reduce the exposure of specular highlights with an adjustment layer).

The shadows are either handled by a levels or better a (luminosity) curves adjustment.

I verify that Chromatic Aberration correction indeed improves the image, or that the lens profile (if any) alone is better. Also verify that Distortion correction and Hide Distorted Areas doesn't crop the image (I have that set in the default style).

Then creative adjustments to tonality, color (Color temperature/tint for a change in mood, Color balance between shadows midtones and highlights, Color Editor for specific adjustments like differentiation in foliage colors or sky color via a Color Editor selection and applied as layer adjustment), vignetting, and other local adjustments can be made.

I also take LCC shots (if I suspect that sensor dust can become a major postprocessing burden), which can then be applied to multiple images taken around the same time. This can also be useful if a Tilt Shift lens is used to adjust for off-center vignetting correction.

Proofing is set for the output profile in the output recipe which is one of the first things I select, if already known.

That's basically it, but depending on subject I tend to run through all controls to see if something can be improved. Also keystone corrections get used a lot with suitable subjects (architecture, reproductions under difficult angles), although one can also use it to play with perspective a bit, e.g. (de)emphasizing foreground in nature shots with few vertical features that could get too much perspective. Sometimes a minor Color noise correction of 1 can produce a much easier to process image in Photoshop, if I want to save some time instead of using Topaz Denoise for higher ISO shots.

Cheers,
Bart

[Robert Ardill]

Yes, but LR starts with a default that I do not like. I need to reduce the Highlight control to  between -100 to -50 to get a more reasonable highlight tonal rendering. Without that, e.g. bright clouds look dull blobs of white, with the -100 Highlights there is cloud formation structure and a more subtle 3D rendering.

That's because you are exposing quite strongly to the right.  I've stopped doing that, partly because at ISO 100, especially with the newer cameras, the shadows are excellent and I see no real advantage in exposing too far to the right because of the risk of one or more of the colors getting clipped.  Still, there's no doubt that using a linear curve I could safely expose one stop further than I currently do, which is significant!

... and a slight (+5) Saturation boost (positive values are actually Vibrance, negative values are Saturation)...

I don't get that.  As far as I can see, a negative saturation desaturates.  Is there a setting for this? It would be very useful!

 Specular whites and reflected lightsources or white clouds may be pushed towards but just shy of really clipping. Sometimes really clipping some specularity produces a better image (one can always locally reduce the exposure of specular highlights with an adjustment layer).

You could set the top Levels slider to 250, say, which would prevent clipping.

Many thanks Bart for taking the time to write this ... it's certainly been very useful to me and I'm sure it will be to others too!

=======================================================================================

I've done a test with my A7RII to compare LR with C1 and here are the results (the shot wasn't taken very carefully). LR on the left, C1 on the right:



Not so good this time!  I've sent the results to Phase One to get their comments.  I was very careful to have like for like settings in both converters, sharpening off, luminance noise reduction off etc.  I just white-balance and adjusted the white point.

I also tried the test with Standard and Linear curves, but the differences are insignificant.

Cheers

Robert

[Bart_van_der_Wolf]

I don't get that.  As far as I can see, a negative saturation desaturates.  Is there a setting for this? It would be very useful!

The saturation slider has a dual functionality. Negative values desaturate (like a regular Saturation slider), but positive values behave like Vibrance.

Cheers,
Bart

[Robert Ardill]

The saturation slider has a dual functionality. Negative values desaturate (like a regular Saturation slider), but positive values behave like Vibrance.

Cheers,
Bart

Yes, I can see that the saturation slider in C1 is smart in that it limits the increase in saturation of already saturated colors, allowing less saturated colors to catch up, so to speak.  Which I guess is what 'vibrance' is.  Having both Saturation and Vibrance in LR is an advantage as
- reducing saturation and increasing vibrance equalises the less saturated and more saturated colors
- reducing vibrance and increasing saturation boosts the already saturated colors and desaturates the less saturated colors

(just found this out - generally I don't mess around too much with either ... perhaps just a little extra vibrance).

Robert

[tho_mas]

Having both Saturation and Vibrance in LR is an advantage

in C1:
- the "saturation" silder in the exposure-tool of C1 as explained by Bart
- "saturation" in the Basic Color Editor likewise applies "smart saturation"
- "saturation" in the Advanced Color Editor acts like conventionell saturation
- color edits in both the Basic and the Advanced Color Editor alter the actual input- ("camera"-) profile. Increasing saturation in the Advanced Color Editor is effectively expanding the input profile (which prevents clipping in the source profile "by design").

[Bart_van_der_Wolf]

I've done a test with my A7RII to compare LR with C1 and here are the results (the shot wasn't taken very carefully). LR on the left, C1 on the right:

Hi Robert,

I noticed 2 things.

First, the slant is very minor The ISO organization recommends a slope of 5-6 degrees off horizontal/vertical, your shot is 1.43 degrees. This means that you only have 7 full 40 pixel phase rotations over the crop height, which is a bit low for statistical purposes. My resolution target also has slanted edges, and they are designed at 5.71 degrees, so slightly rotated camera or target shots will fall near the sweet spot, and would have given some 29 rotations over the crop height.

Second, the edge profile doesn't normalize to 0.0 to 1.0, which (if I remember correctly) would normally be the case. I would also expect a 10-90% rise of 1.8 - 2.0 pixels for a good well focused lens at f/5.6, so both of the numbers that you found are relatively low with only slightly more diffraction from the f/7.1 to explain it. Something is not what I'd expect.

I was very careful to have like for like settings in both converters, sharpening off, luminance noise reduction off etc.  I just white-balance and adjusted the white point.

I also tried the test with Standard and Linear curves, but the differences are insignificant.

Maybe the brightness (which seems a bit low) and Color balance (the crop looks a bit blue) could still play a role. I usually try to get the paper white brightness at something like 235, and verify that medium gray lands around 120 -130 which would indicate a normal contrast.

Just some things to verify.

Cheers,
Bart

[Robert Ardill]

Hi Robert,

I noticed 2 things.

First, the slant is very minor The ISO organization recommends a slope of 5-6 degrees off horizontal/vertical, your shot is 1.43 degrees. This means that you only have 7 full 40 pixel phase rotations over the crop height, which is a bit low for statistical purposes. My resolution target also has slanted edges, and they are designed at 5.71 degrees, so slightly rotated camera or target shots will fall near the sweet spot, and would have given some 29 rotations over the crop height.

Second, the edge profile doesn't normalize to 0.0 to 1.0, which (if I remember correctly) would normally be the case. I would also expect a 10-90% rise of 1.8 - 2.0 pixels for a good well focused lens at f/5.6, so both of the numbers that you found are relatively low with only slightly more diffraction from the f/7.1 to explain it. Something is not what I'd expect.

Maybe the brightness (which seems a bit low) and Color balance (the crop looks a bit blue) could still play a role. I usually try to get the paper white brightness at something like 235, and verify that medium gray lands around 120 -130 which would indicate a normal contrast.

Just some things to verify.

Cheers,
Bart

Hi Bart ... yes I must have had a setting wrong somewhere (in Imatest, most likely).  I've redone the test and here are the results:



I wouldn't be overconcerned about the actual 10-90% rise or the MTF values as my chart isn't the best, nor is my lighting. Also, Imatest tells me that they are working on a solution for sensors over 24MPx as the current setup isn't good enough (and that would be using their best quality charts, really good lighting and super-careful focusing).

It's really the comparative figures that I am interested in.  C1 does much better this time fortunately (seeing as how I've just purchased C1 I wasn't happy at all with the figures I got in the first test).  On the MTF, LR is a bit better above MTF60 and C1 significantly better below MTF40.

The analysis is very sensitive to the tone curve, which could explain some of the differences. For example, increasing the contrast a bit on both gives these results:



Now C1 is significantly better from MTF80 down and has  equalized on the 10-90% edge rise. 

One difference between this test and the one I did on the 1DsIII is that the chart contrast is much higher on this test (I picked a square that was probably only 10:1).  Which has reminded me that I did not change the contrast ratio in Imatest to 20:1, which is the contrast on this image.  This further improves the results:



Now C1 is leading the all way ... which is the result I wanted (and I hope I haven't strayed away from scientific rigor too much ). With a Focus Magic sharpening of 1/100 it looks like this, which is pretty outstanding (so my photograph couldn't have been so bad after all ).  The MTF80 of 3012 lw/ph is really good for such a low sharpening. A nice thumbs-up for FM!



My turn to apologize for the long post ... but these tests are not easy to do, as you know.

Cheers,

Robert

[tho_mas]

now that you’ve discovered C1 is producing finer details and offers some real benefits it is maybe time to simply work with the software and get accustomed to it.
Beside all that graphs and pixel peeping comparisons the operability of a software also plays a role (at least for me).
One strength (beside the aspects listed above) is user customization … namely the option to arrange the tools within the interface by your liking („workspaces“) and to assign a wide selection of keyboard-shortcuts. An „intelligent“ user interface together with appropriate shortcuts is a real timesaver … and makes working with the software a real joy.

As far as „camera profiles“ go personally I think „color accuracy“ of captures is totally overrated by many people on this forum (unless you really do serious reproduction work). Then again - as you already know - it is pretty easy to roughly adjust a given profile and save it as a new „camera profile“. In addition you can use all the profiles provided by Capture One for your shots and select one that you like better. In case of your Sony A7R2 it makes sense to try the camera profiles of the Phase One IQ3-100MP (basically the same Sony CMOS sensor with very similar color response … at least if you look at the camera profiles for both models). For the IQ3-100MP there are profiles for „Flash, „Outdoor Daylight“, „Portrait“, „Tungsten“ and „Flash - Flat Art Reproduction“. These profiles may give a better starting point as the „generic“ (or „standard“) SonyA7R2 profile. Personally I do use the IQ3-100MP profiles for my A7R2 captures as a starting point …
You can even rename the IQ3-100MP profiles in a way so that they show up in the menu of the Sony A7R2.
How to…:
- On Mac go to the C1 application -> right click on application -> show package content -> Contents -> Framworks -> AppCore.framework -> Versions -> A -> Resources -> Profiles -> Input.
- Search for the „PhaseOneIQ3100MP“ profiles (see attachment 1).
- Copy (!!!) the profiles to the Desktop.
- Rename the file names from „PhaseOneIQ3100MP“ to „SonyA7RM2“ (see attachment 2 … I’ve also added „CC“ as a suffix so that I quickly recognize that these are the renamed IQ3-100MP profiles). Carefully replicate the naming conventions (camera name and hyphen without blank spaces in between!). Actually you should also rename the internal name of the profile (you could do this in Color Sync -> „desc“ attribute -> ASCII-Name) so that other softwares read the names correctly, too (but for the work with C1 renaming the actual file name will do…).
- Move the renamed profiles to the ColorSync folder of your user domain. Then the profiles show up in C1 as „A7RM2“ profiles (see attachment 3 - note: the prefix of the camera name is not displayed within the camera-specific menu).

[Robert Ardill]

now that you’ve discovered C1 is producing finer details and offers some real benefits it is maybe time to simply work with the software and get accustomed to it.
Beside all that graphs and pixel peeping comparisons the operability of a software also plays a role (at least for me).
One strength (beside the aspects listed above) is user customization … namely the option to arrange the tools within the interface by your liking („workspaces“) and to assign a wide selection of keyboard-shortcuts. An „intelligent“ user interface together with appropriate shortcuts is a real timesaver … and makes working with the software a real joy.

As far as „camera profiles“ go personally I think „color accuracy“ of captures is totally overrated by many people on this forum (unless you really do serious reproduction work). Then again - as you already know - it is pretty easy to roughly adjust a given profile and save it as a new „camera profile“. In addition you can use all the profiles provided by Capture One for your shots and select one that you like better. In case of your Sony A7R2 it makes sense to try the camera profiles of the Phase One IQ3-100MP (basically the same Sony CMOS sensor with very similar color response … at least if you look at the camera profiles for both models). For the IQ3-100MP there are profiles for „Flash, „Outdoor Daylight“, „Portrait“, „Tungsten“ and „Flash - Flat Art Reproduction“. These profiles may give a better starting point as the „generic“ (or „standard“) SonyA7R2 profile. Personally I do use the IQ3-100MP profiles for my A7R2 captures as a starting point …
You can even rename the IQ3-100MP profiles in a way so that they show up in the menu of the Sony A7R2.
How to…:
- On Mac go to the C1 application -> right click on application -> show package content -> Contents -> Framworks -> AppCore.framework -> Versions -> A -> Resources -> Profiles -> Input.
- Search for the „PhaseOneIQ3100MP“ profiles (see attachment 1).
- Copy (!!!) the profiles to the Desktop.
- Rename the file names from „PhaseOneIQ3100MP“ to „SonyA7RM2“ (see attachment 2 … I’ve also added „CC“ as a suffix so that I quickly recognize that these are the renamed IQ3-100MP profiles). Carefully replicate the naming conventions (camera name and hyphen without blank spaces in between!). Actually you should also rename the internal name of the profile (you could do this in Color Sync -> „desc“ attribute -> ASCII-Name) so that other softwares read the names correctly, too (but for the work with C1 renaming the actual file name will do…).
- Move the renamed profiles to the ColorSync folder of your user domain. Then the profiles show up in C1 as „A7RM2“ profiles (see attachment 3 - note: the prefix of the camera name is not displayed within the camera-specific menu).

Thanks for your suggestion ... I'll have a look at the IQ3 profile. 

I agree that the customisation of C1 is very good - I have it set up just as I want it and it was very straightforward to do.

Of course there is a lot more to it than just resolution differences ... as I mentioned earlier, I find that LR does a better job with some images while C1 is better at others.  With more time using C1 I hope to be able to get a feel for which images are best handled in C1 and which are best done in LR.

I also hope that Phase One will improve their interface with LR and Photoshop ... to my mind they can only benefit by making it easier for people like me to use their software. I would have used C1 a long time ago if it had better integration with the Adobe apps ... and if C1 had the kind of plug-in support that LR has.

Cheers

Robert

[Robert Ardill]

I received a reply from Phase One regarding the differences in resolution between the LR and C1 test images.  The technician said that "This test is actually a test of base level sharpening by a program".  He then continued to say that the best way to decide on which program to use is to develop an image in both programs, doing the best that one can, and then decide which one prefers.

There was no claim that C1 produces an inherently more detailed image due to its superior demosaicing algorithm.

I've been on a photoshoot for the last 10 days and I've tried LR and C1 on a number of images that would be fairly representative, for me.  I've found that some images look better out-of-the-box in C1 and others look better in LR.  I have always been able to tweak the other program to give me the preferred look quite easily.  But I have to say, perhaps because I am less familiar with C1, that I find LR easier and quicker.  The thing that makes the biggest difference to the image look is the camera profile and tone curve ... and I think LR has a very simple, effective and quick way of making new camera profiles ... and tweaking them to one's taste.  Of course, C1 can do this too, but I think the automation of dng profile creation for LR is very hard to beat.

As I always go through Photoshop, the C1 color editor advantages over LR are not significant for me.  Equally, the advantages/disadvantages of one raw converter over the other in terms of noise reduction and sharpening are not relevant as I do these from PS.

As LR has so many advantages in terms of integration with PS (and support for plug-ins), web support, Publish Services, geocoding etc, and it is so awkward to work with both C1 and LR, it's pretty much a no-brainer for me.

The reason for this post was a tutorial by Michael Reichman A Workflow Combining Capture One and Lightroom in which he said that: "The major steps which I take in Capture One are to choose the appropriate camera profile, do a white balance, and then set black point and white point. These are the critical steps that need to be done in raw mode. I then export the file to Lightroom for further processing.". Admittedly this was written in 2009 and he may have changed his views since then.

So, I thought that if he was right and that C1's raw conversion was significantly superior to LR's, that it would be worth the nuisance of using both.  My conclusion at this stage is that the two programs are neck-and-neck when it comes to raw conversion (at least with Sony A7/A7RII) and any slight differences in resolution (and one really needs to pixel-peep to see them) are most likely due to different base sharpening applied by the raw converters ... and these can be easily corrected by a good sharpening tool like Focus Magic.

Of course this may not be true for other cameras ... and from Bart's tests it would seem that C1 has an advantage for the 1Ds3.  I have a lot of 1Ds3 images ...

Cheers

Robert

[ErikKaffehr]

Hi,

My issue with LR is that it introduces some more aliasing artefacts than C1 on A7rII images and for that part also P45+ images. Other than that I don't know if I find C1 superior.

What I would do is use LR, but if I had issue develop in RawTherapee and reimport the TIFF.

Best regards
Erik

I received a reply from Phase One regarding the differences in resolution between the LR and C1 test images.  The technician said that "This test is actually a test of base level sharpening by a program".  He then continued to say that the best way to decide on which program to use is to develop an image in both programs, doing the best that one can, and then decide which one prefers.

There was no claim that C1 produces an inherently more detailed image due to its superior demosaicing algorithm.

I've been on a photoshoot for the last 10 days and I've tried LR and C1 on a number of images that would be fairly representative, for me.  I've found that some images look better out-of-the-box in C1 and others look better in LR.  I have always been able to tweak the other program to give me the preferred look quite easily.  But I have to say, perhaps because I am less familiar with C1, that I find LR easier and quicker.  The thing that makes the biggest difference to the image look is the camera profile and tone curve ... and I think LR has a very simple, effective and quick way of making new camera profiles ... and tweaking them to one's taste.  Of course, C1 can do this too, but I think the automation of dng profile creation for LR is very hard to beat.

As I always go through Photoshop, the C1 color editor advantages over LR are not significant for me.  Equally, the advantages/disadvantages of one raw converter over the other in terms of noise reduction and sharpening are not relevant as I do these from PS.

As LR has so many advantages in terms of integration with PS (and support for plug-ins), web support, Publish Services, geocoding etc, and it is so awkward to work with both C1 and LR, it's pretty much a no-brainer for me.

The reason for this post was a tutorial by Michael Reichman A Workflow Combining Capture One and Lightroom in which he said that: "The major steps which I take in Capture One are to choose the appropriate camera profile, do a white balance, and then set black point and white point. These are the critical steps that need to be done in raw mode. I then export the file to Lightroom for further processing.". Admittedly this was written in 2009 and he may have changed his views since then.

So, I thought that if he was right and that C1's raw conversion was significantly superior to LR's, that it would be worth the nuisance of using both.  My conclusion at this stage is that the two programs are neck-and-neck when it comes to raw conversion (at least with Sony A7/A7RII) and any slight differences in resolution (and one really needs to pixel-peep to see them) are most likely due to different base sharpening applied by the raw converters ... and these can be easily corrected by a good sharpening tool like Focus Magic.

Of course this may not be true for other cameras ... and from Bart's tests it would seem that C1 has an advantage for the 1Ds3.  I have a lot of 1Ds3 images ...

Cheers

Robert

[Robert Ardill]

Hi,

My issue with LR is that it introduces some more aliasing artefacts than C1 on A7rII images and for that part also P45+ images. Other than that I don't know if I find C1 superior.

What I would do is use LR, but if I had issue develop in RawTherapee and reimport the TIFF.

Best regards
Erik

Thanks Eric ... that sounds like a good plan.  So far I haven't had any problems with aliasing artifacts with the A7RII, but when I do I'll give the image a go in RT.

Cheers

Robert

[ErikKaffehr]

Hi Robert,

You could check this for instance: http://forum.luminous-landscape.com/index.php?topic=104708.0

But also check this response: http://forum.luminous-landscape.com/index.php?topic=104708.msg860933#msg860933

Another example is this discussion (with Phase One P45+ images):
http://forum.luminous-landscape.com/index.php?topic=94812.0

With the A7rII (and the P45+) I often see extra contour artefacts like in the attachment below, check along the antenna and under the text 911. These are not ringing artefacts from sharpening but seem to arise in demosaic. So, would I make a large print from this image I would convert it in another converter and do the rest of the processing in LR or Photoshop. My favourite sharpening tool is Focus Magic, but I use LR built in tools as long as practical.

Printing at A2 (16"x23") which is my normal print size I simply wouldn't care, but for a 30"x40" print I would make the extra effort.

Best regards
Erik

Thanks Eric ... that sounds like a good plan.  So far I haven't had any problems with aliasing artifacts with the A7RII, but when I do I'll give the image a go in RT.

Cheers

Robert

[Robert Ardill]

Hi Robert,

You could check this for instance: http://forum.luminous-landscape.com/index.php?topic=104708.0

But also check this response: http://forum.luminous-landscape.com/index.php?topic=104708.msg860933#msg860933

Another example is this discussion (with Phase One P45+ images):
http://forum.luminous-landscape.com/index.php?topic=94812.0

With the A7rII (and the P45+) I often see extra contour artefacts like in the attachment below, check along the antenna and under the text 911. These are not ringing artefacts from sharpening but seem to arise in demosaic. So, would I make a large print from this image I would convert it in another converter and do the rest of the processing in LR or Photoshop. My favourite sharpening tool is Focus Magic, but I use LR built in tools as long as practical.

Printing at A2 (16"x23") which is my normal print size I simply wouldn't care, but for a 30"x40" print I would make the extra effort.

Best regards
Erik

Thanks Eric ... I'll have a good look when I get home tonight (I'm travelling at the moment).  It sounds interesting ... do you think the artifacts are removed by some converters, or created by poor demosaicing in others?

Cheers

Robert