What are the essential adjustments that SHOULD be done in the raw processor?

[Robert Ardill]

Hi,

I'm a Lightroom user normally, but having changed from Canon to Sony recently I decided to have a look at Capture One as CO is the free raw converter for Sony cameras. Also, because so many people say that CO is WAY better than LR at raw conversion.

I've done some tests between LR and CO, doing nothing but white balance, white point, black point and color noise reduction.  Comparing a number of images I see no advantage to either detail-wise.  There is some color difference and slight contrast difference which is better with one converter on some images and better with the other converter on other images, but the differences are easily adjusted in the raw converters themselves, or in post-processing.

With some structure added in CO there is a marked improvement detail-wise, but this is easily compensated on the LR tiff with Topaz Detail3 (for example).

Which leads me to my question.  Does anyone know what are the adjustments that absolutely should be done in the raw converter, and which are just as well done in post processing? 

My assumption has been that the only adjustments that should be done in the raw converter are white balance and white/black point, and that everything else (noise reduction, sharpening, color adjustments, chromatic aberration, color fringing etc.,) can all be left to post-processing.  I have no other reasons to believe that this is correct except that my tests seem to bare out the assumption.  For example, opening an image in ACR with tonal adjustments and comparing it to the same image that has been opened without tonal adjustments (save for white point / black point) but then processed using the Camera Raw filter with exactly the same settings, shows only very slight differences, only just visible at 1:1 (very slight sharpness difference, slight contrast difference, slight color differences).

This isn't a Capture One v Lightroom question as clearly each has strengths. If my assumption that only basic tonal adjustments should be done in the raw converter is wrong then I would probably use CO for the raw conversion and other basic adjustments, retaining Lightroom for all the other features.  On the other hand, if the bulk of adjustments can be left to post-processing, then I would stick to LR as I can do pretty much everything I want with LR + PS (and plug-ins).

I appreciate your advice!

Robert

Robert

[Jack Hogan]

Hi Robert,

If by raw conversion you mean just opening the raw file into a 16-bit TIFF, imho the most critical parameters are EC (easy as pie), white balance (most can do a decent job), CA corrections (not hard but some do it better than others), choice of color profile (fiendishly hard) and demosaicing algorithm (some give you options, some don't).  Every raw converter knows (or should know) how to subtract the correct black level from the raw data.  And you don't need to touch the 'white point' before rendering. Some purport to perform magic under the hood (noise reduction etc.), but 99% of that stuff can just as well be performed on the 16-bit TIFF.

Jack

[Mark D Segal]

Hi,

I'm a Lightroom user normally, but having changed from Canon to Sony recently I decided to have a look at Capture One as CO is the free raw converter for Sony cameras. Also, because so many people say that CO is WAY better than LR at raw conversion.

I've done some tests between LR and CO, doing nothing but white balance, white point, black point and color noise reduction.  Comparing a number of images I see no advantage to either detail-wise.  There is some color difference and slight contrast difference which is better with one converter on some images and better with the other converter on other images, but the differences are easily adjusted in the raw converters themselves, or in post-processing.

With some structure added in CO there is a marked improvement detail-wise, but this is easily compensated on the LR tiff with Topaz Detail3 (for example).

Which leads me to my question.  Does anyone know what are the adjustments that absolutely should be done in the raw converter, and which are just as well done in post processing? 

My assumption has been that the only adjustments that should be done in the raw converter are white balance and white/black point, and that everything else (noise reduction, sharpening, color adjustments, chromatic aberration, color fringing etc.,) can all be left to post-processing.  I have no other reasons to believe that this is correct except that my tests seem to bare out the assumption.  For example, opening an image in ACR with tonal adjustments and comparing it to the same image that has been opened without tonal adjustments (save for white point / black point) but then processed using the Camera Raw filter with exactly the same settings, shows only very slight differences, only just visible at 1:1 (very slight sharpness difference, slight contrast difference, slight color differences).

This isn't a Capture One v Lightroom question as clearly each has strengths. If my assumption that only basic tonal adjustments should be done in the raw converter is wrong then I would probably use CO for the raw conversion and other basic adjustments, retaining Lightroom for all the other features.  On the other hand, if the bulk of adjustments can be left to post-processing, then I would stick to LR as I can do pretty much everything I want with LR + PS (and plug-ins).

I appreciate your advice!

Robert

Robert

Hi Robert,

First let's get some commercial stuff out of the way. The version of C1 that Sony advertises as "free" is a very basic application called "C1 Express for Sony". It isn't nearly as full-featured as other version of C1 or LR. The next step up in C1 is a $50 upgrade of C1 Express for Sony which gives you a few more features. Beyond that you pay another $299 for the fully featured C1 application. None of these applications - the three versions of C1 or LR yet provide support for the new Sony A6300 - which I discovered much to my regret after buying the camera. No regret buying the camera - it's great - but to see the raw files, heaven forbid, one must download the Sony Image Converter application which is nothing much to write home about - but at least it demosaics the files, lets you enlarge the raw files on display and has a range of very basic adjustments.

I believe both C1 and LR are fine raw converter applications. As far as I'm concerned the choice depends on work flow preferences, and those preferences are often determined by what one knows how to use best. Personally, I find I can do a lot more much better and faster in Lightroom than I can in Photoshop, but having reasonably good integration between Photoshop and Lightroom provides with considerable ease that huge pack of tools for certain things LR can't do. That said, LR has become such a fully-featured capable application for a high percentage of what many users need, that for a proficient Lightroom user recourse to Photoshop may well be limited. Most of my work, for example, starts and finishes within LR - from camera to print. This provides for a huge saving of hard drive space and is so convenient. (For example, yesterday I converted a Sony raw file of 24 MB into a TIFF for further work in PS. By the time I finished and saved the file with the Layers, it was 499 MB. If I could have done the same adjustments in LR (a) it would have taken less time, (b) the file size would have remained 24 MB, and (c) everything is undoable thereafter without the need for saving space-hogging Layers. Doing as much as possible in the raw converter means getting to know the application well enough to extract maximum performance from it, and in so doing provides excellent results.

[Tim Lookingbill]

All things being the same in the workflow you describe which of the two is the easiest and fastest to deal with seeing you seem to nail the intended look of the image during capture.

Just out of curiosity could you post an image unedited out of the newer camera and post the same image but with the minimal edits you think all it needs. What is your idea of finish is what I'm getting at.

I've been using an old 2006 DSLR for years shooting Raw by exposing to preserve highlights and I have to apply a lot more than minor adjustments on every one of them, but I don't know if in your situation it's on account of the camera and/or the newer editing software.

[Robert Ardill]

Hi Robert,

If by raw conversion you mean just opening the raw file into a 16-bit TIFF, imho the most critical parameters are EC (easy as pie), white balance (most can do a decent job), CA corrections (not hard but some do it better than others), choice of color profile (fiendishly hard) and demosaicing algorithm (some give you options, some don't).  Every raw converter knows (or should know) how to subtract the correct black level from the raw data.  And you don't need to touch the 'white point' before rendering. Some purport to perform magic under the hood (noise reduction etc.), but 99% of that stuff can just as well be performed on the 16-bit TIFF.

Jack

So really what you're saying is that all you really need to tell the raw converter is:
- white balance
- CA correction (could this not be done as well in post-processing?)
- color profile (do you mean the camera profile, or the tiff image profile, or both?)
- demosaicing algorithm, assuming the raw converter gives you this choice (which neither LR nor CO one do)

Then the rest can just as well be done in post-processing (for example in Photoshop). Correct?

Robert

[Robert Ardill]

First let's get some commercial stuff out of the way. The version of C1 that Sony advertises as "free" is a very basic application called "C1 Express for Sony". It isn't nearly as full-featured as other version of C1 or LR. The next step up in C1 is a $50 upgrade of C1 Express for Sony which gives you a few more features. Beyond that you pay another $299 for the fully featured C1 application. None of these applications - the three versions of C1 or LR yet provide support for the new Sony A6300 - which I discovered much to my regret after buying the camera. No regret buying the camera - it's great - but to see the raw files, heaven forbid, one must download the Sony Image Converter application which is nothing much to write home about - but at least it demosaics the files, lets you enlarge the raw files on display and has a range of very basic adjustments.

Yes, I used Capture One Pro 9, full version, for the tests.  I don't think you're right about the Sony versions though. Capture One Express for Sony is free and is very limited. The $50 upgrade gives you Capture One Pro 9 (for Sony) which has all of the features of Capture One Pro 9, but only supports Sony cameras.

"The feature sets of the two upgrade options are the same, and the only differences are the supported camera brands and the price. Capture One Pro (for Sony) only offers file support for selected Sony cameras while Capture One Pro 9 offers file support for more than 400 leading cameras."

I believe both C1 and LR are fine raw converter applications. As far as I'm concerned the choice depends on work flow preferences, and those preferences are often determined by what one knows how to use best. Personally, I find I can do a lot more much better and faster in Lightroom than I can in Photoshop, but having reasonably good integration between Photoshop and Lightroom provides with considerable ease that huge pack of tools for certain things LR can't do. That said, LR has become such a fully-featured capable application for a high percentage of what many users need, that for a proficient Lightroom user recourse to Photoshop may well be limited. Most of my work, for example, starts and finishes within LR - from camera to print. This provides for a huge saving of hard drive space and is so convenient. (For example, yesterday I converted a Sony raw file of 24 MB into a TIFF for further work in PS. By the time I finished and saved the file with the Layers, it was 499 MB. If I could have done the same adjustments in LR (a) it would have taken less time, (b) the file size would have remained 24 MB, and (c) everything is undoable thereafter without the need for saving space-hogging Layers. Doing as much as possible in the raw converter means getting to know the application well enough to extract maximum performance from it, and in so doing provides excellent results.

Yes, as I said, I'm not trying to measure one raw converter against another, simply to try to find out what can be safely left to post processing and what should be done prior to demosaicing.

My own rather limited tests seem to indicate that you can leave most everything to post processing (should you prefer to do things in post-processing, obviously).  Jack reckons white balance and CA. Obviously you need to specify what output color profile to use.  I would have thought that you would be well advised to set the black and white points first, but as one would be going from 12 or 14 bits to 16 bits perhaps that isn't necessary either.

Of course I do accept that there are benefits for doing a lot more (or everything) in the raw converter of choice, but that isn't always possible.  For example, in my case I do absolutely want to use Lightroom for all of the image organisation, selection, previewing, web output, geo-coding etc., etc., because it does it so well; but if I find that C1 is really much better at the raw conversion and decide to use C1 for that purpose then I will have no choice but to render to tiff because the integration with LR (and Photoshop) is so poor (shockingly so in my opinion).

Cheers,

Robert

[Mark D Segal]

Your not saying anything very different from what I said about the available versions for C1. But I do think (from the little I could make out in their description) the full application has yet more editing capabilities than the 50 dollar upgrade of the version for Sony. Exactly what I don't know.

If you find using C1 is so much better than LR, why not just do as much of your image editing as possible in C1 and only convert it to TIFF if there is something you need that C1 can't do?

[bjanes]

Which leads me to my question.  Does anyone know what are the adjustments that absolutely should be done in the raw converter, and which are just as well done in post processing? 

My assumption has been that the only adjustments that should be done in the raw converter are white balance and white/black point, and that everything else (noise reduction, sharpening, color adjustments, chromatic aberration, color fringing etc.,) can all be left to post-processing.

I think your assumptions are correct, but I would add that those operations that are best performed on linear scene referred data are best done in the raw converter. Certainly this applies to white balance. The author of Noiseware once stated that NR is best done in the raw converter as the noise characteristics are changed by gamma encoding, exposure adjustments and a tone curve. One can perform inverse gamma encoding later on, but changes related to a tonal adjustments would be undocumented in most cases.

Regards,

Bill

[Robert Ardill]

Just out of curiosity could you post an image unedited out of the newer camera and post the same image but with the minimal edits you think all it needs. What is your idea of finish is what I'm getting at.

Hi tim,

Here is an image which is pretty out-of-kilter, mostly very underexposed but slightly overexposed for the lamp.  In this case I would do more than just basic black/white point and white balance, as I show in the second image:



However, if I only do the most basic tone correction and then post-correct (using the Camera Raw filter in Photoshop) then I get this (the post-processed version is on top with the pre-processed version below it for comparison):



As you can see there is little visual difference although the pre-processed version histogram does look a bit better (the shadows need to be pushed up a bit, which would be easy enough to correct).

Robert

[Robert Ardill]

If you find using C1 is so much better than LR, why not just do as much of your image editing as possible in C1 and only convert it to TIFF if there is something you need that C1 can't do?

I don't know if C1 is better or not ... it really depends on whether it's important to do things like color adjustments, adding structure, things like that, before demosaic.  If it doesn't matter then C1 boils down to whether or not the raw conversion is better than LR's (with just basic tonal adjustment).  From my tests I would say that it isn't for sharpness (but it's as good) and it's image-dependent for things like color fidelity. So it wouldn't be worth the trouble to use both packages for me.

There is a whole lot that C1 doesn't do well for me (or at least that LR does much better, IMO). I could list plenty, but just for one thing alone, take Photoshop integration.  So C1 is not an option for me on its own. It would only be useful just for the raw conversion and I would then always have to go to tiff (which is OK for me because I do that anyway ... disk is very cheap nowadays, so I'm not worried about big files).

To try to clarify, it's like this:
- if it is important to make as many adjustments as possible pre-demosaic, then I would use C1 as the raw converter (even though it's a pain using both LR and C1)
- if only basic tonal adjustments are required pre-demosaic then I would stick with LR (because I can make all the adjustments I want either in LR or in Photoshop and plug-ins).

[Robert Ardill]

I think your assumptions are correct, but I would add that those operations that are best performed on linear scene referred data are best done in the raw converter. Certainly this applies to white balance. The author of Noiseware once stated that NR is best done in the raw converter as the noise characteristics are changed by gamma encoding, exposure adjustments and a tone curve. One can perform inverse gamma encoding later on, but changes related to a tonal adjustments would be undocumented in most cases.

That's the crux of it really.  It seems to me that at this stage things like noise reduction and deblur /sharpening are not done very well in either LR or C1, so we pretty much have to do them post-processing if we want the best results.  So the answer may be that at this point it's down to whether we think the raw conversion in C1 is better than the raw conversion in LR (or vice versa); however that could well change (and hopefully will), when the raw converters get decent deconvolution and noise reduction features.

Would you agree?

Robert

[bjanes]

That's the crux of it really.  It seems to me that at this stage things like noise reduction and deblur /sharpening are not done very well in either LR or C1, so we pretty much have to do them post-processing if we want the best results.  So the answer may be that at this point it's down to whether we think the raw conversion in C1 is better than the raw conversion in LR (or vice versa); however that could well change (and hopefully will), when the raw converters get decent deconvolution and noise reduction features.

Would you agree?

Robert

Robert,

Yes, I agree mostly. However, since I started using the Nikon D800e, my need for noise reduction is much less than previously. Even at ISO 3200, properly exposed images need little NR. I do have Noiseware, one of the better reviewed NR programs, but I use it infrequently. However, LR/ACR could benefit from better deconvolution/sharpening. It is not that bad, but for my more important images I bring them into Photoshop and sharpen on a layer with the luminosity blend mode using Focus Magic or Topaz detail. I hope Bart van der Wolf will enter into the discussion, since he is well regarded in this area.

Bill

[Tim Lookingbill]

Thanks for the image samples, Robert. Those do look finished. Not much else you can do to improve them other than the fireplace shadow definition and lift.

So which is easier to accomplish the same between C1 and ACR.

I'm also a bit perplexed by the ACR filter in photoshop in why it must be used over just launching ACR from Bridge.

[Robert Ardill]

Yes, I agree mostly. However, since I started using the Nikon D800e, my need for noise reduction is much less than previously. Even at ISO 3200, properly exposed images need little NR. I do have Noiseware, one of the better reviewed NR programs, but I use it infrequently. However, LR/ACR could benefit from better deconvolution/sharpening. It is not that bad, but for my more important images I bring them into Photoshop and sharpen on a layer with the luminosity blend mode using Focus Magic or Topaz detail. I hope Bart van der Wolf will enter into the discussion, since he is well regarded in this area.

Hi Bill ... for sure the new cameras are hugely improved noise-wise. I have a Sony A7RII and the dynamic range and very low noise at high ISO is fantastic.  In the image I posted above there is essentially no noise even though the exposure is boosted by over 3EV on an ISO 250 image.  Still, there will be times when some noise reduction is going to be needed, especially color noise reduction.  ACR/LR isn't bad - but it's a pity there isn't masking as in Sharpening.

I hope Bart joins it too -- especially as he uses C1, if I recall.

Cheers

Robert

[Robert Ardill]

Thanks for the image samples, Robert. Those do look finished. Not much else you can do to improve them other than the fireplace shadow definition and lift.

So which is easier to accomplish the same between C1 and ACR.

I'm also a bit perplexed by the ACR filter in photoshop in why it must be used over just launching ACR from Bridge.

I find LR/ACR easier, but that's no doubt because I'm so used to it. C1 is very nice though, once you get over the rather confusing workspace, and it as features that are missing in LR, for example adjustment layers.

I just used the ACR filter because I wanted to apply the adjustments to the tiff file.  I could have saved the tiff and opened in using Bridge just as well.

Robert

[Robert Ardill]

Here is a crop of an image comparing C1 and LR with only the following adjustments:
- white balance on the same spot
- exposure
- color noise reduction
- all other adjustments off or zeroed



To my eye there is nothing at all between them detail-wise. (You can right-click and download the image if you're interested).

However ... there are huge color differences.  Look at the neon sign for example, the wall behind the neon sign, the shop sign, the yellows, the reds. I don't know which one is the more accurate although I would guess that C1 is.

In C1 I used the SonyA7RMII STANDARD profile, in LR I used a DNG profile for the camera.

So I'll need to do some further checking.

Robert

[stamper]

Here is a crop of an image comparing C1 and LR with only the following adjustments:
- white balance on the same spot
- exposure
- color noise reduction
- all other adjustments off or zeroed



To my eye there is nothing at all between them detail-wise. (You can right-click and download the image if you're interested).

However ... there are huge color differences.  Look at the neon sign for example, the wall behind the neon sign, the shop sign, the yellows, the reds. I don't know which one is the more accurate although I would guess that C1 is.

In C1 I used the SonyA7RMII STANDARD profile, in LR I used a DNG profile for the camera.

So I'll need to do some further checking.

Robert

Robert,where are you going with this thread? You wrote this in your first post.

This isn't a Capture One v Lightroom question as clearly each has strengths

[Mark D Segal]

I really wonder how much mileage there is trying to assess the quality differences between raw converters by looking at out of the box colour differences between the originating renderings, when you consider that neither is likely colour accurate and most of these differences are anyhow adjustable. I think the far more important factor to look for is whether either converter produces artifacts once the image is magnified to more or less replicate the size at which it would be printed. As for what adjustments to do where, previous advice from experts who tested these alternatives rather carefully suggest that the cleanest editing is performed on the raw data in the raw converter application to the extent it allows.

[Robert Ardill]

Robert,where are you going with this thread? You wrote this in your first post.

This isn't a Capture One v Lightroom question as clearly each has strengths

Quite true 

What I'm looking at is the effect of applying the very basic adjustments to the two converters and then seeing the result.  If, for example, I found that certain colors were incorrectly rendered in one or other converter, with these basic adjustments only, then I would conclude that it would be advantageous to apply some further basic color corrections pre-rendering.

Robert

[stamper]

Quite true 

What I'm looking at is the effect of applying the very basic adjustments to the two converters and then seeing the result.  If, for example, I found that certain colors were incorrectly rendered in one or other converter, with these basic adjustments only, then I would conclude that it would be advantageous to apply some further basic color corrections pre-rendering.

Robert

How do you judge that .... colors were incorrectly rendered? Surely you aren't trying to remember the colors of the scene you shot? The colors are subjective and trying for "reality" is futile and ultimately time wasting.