What went wrong with C1 20?
So I've been giving v20 a try, mostly because I wanted to try out the "new" Dynamic Range and Noise Reduction features.
After 4 days I have to say I'm quite shocked! Somehow they made the new version worse than the old version was:
- There are bugs with hardware acceleration, leading to artifacts in the editor and corrupting files during export and finally crashes that require a system reboot to fix.
- The noise reduction tool is not very useful for modern cameras with a very high frequency noise, causing artifacts in the finale image unless you reduce the details by a lot.
- The dynamic range tool does not bring any actual improvements, simply working with levels and curves bring the exact same results and it's not even quicker.
- The new color editor tool is a gimmick and far too limited in reality and gives no advantage compared to the previous version. It looks nice in the videos but it's super clunky and tedious, imprecise and is very limited in its range and results...
I'm glad I didn't upgrade and the way it looks v12 is not going to be replaced any time soon.
And now that they totally dismantled the support system for Capture One users vs PhaseOne shooters I don't see any reason to continue purchasing anything from PhaseOne any more. This has been very frustrating and disappointing and I can't help but feel very empathetic to anyone who spent their hard earned money on the "upgrade" - even when it was -30% cheaper.
After 4 days I have to say I'm quite shocked! Somehow they made the new version worse than the old version was:
- There are bugs with hardware acceleration, leading to artifacts in the editor and corrupting files during export and finally crashes that require a system reboot to fix.
- The noise reduction tool is not very useful for modern cameras with a very high frequency noise, causing artifacts in the finale image unless you reduce the details by a lot.
- The dynamic range tool does not bring any actual improvements, simply working with levels and curves bring the exact same results and it's not even quicker.
- The new color editor tool is a gimmick and far too limited in reality and gives no advantage compared to the previous version. It looks nice in the videos but it's super clunky and tedious, imprecise and is very limited in its range and results...
I'm glad I didn't upgrade and the way it looks v12 is not going to be replaced any time soon.
And now that they totally dismantled the support system for Capture One users vs PhaseOne shooters I don't see any reason to continue purchasing anything from PhaseOne any more. This has been very frustrating and disappointing and I can't help but feel very empathetic to anyone who spent their hard earned money on the "upgrade" - even when it was -30% cheaper.
0
-
Well, your experience is quite different to mine.
Interesting.0 -
SFA wrote:
Well, your experience is quite different to mine.
Interesting.
Don't you think the new support system regarding Capture One is atrocious?
Do you mind if I ask which camera you're using?0 -
C-M-B wrote:
SFA wrote:
Well, your experience is quite different to mine.
Interesting.
Don't you think the new support system regarding Capture One is atrocious?
Do you mind if I ask which camera you're using?
Re: the support system .. not used it much so far as I would guess the timing of its introduction and the challenges they are facing with, if nothing else, the order and activation process may just be stressing things to the point where it could go into perpetual overload and I have no reason right now to add to that.
However I did communicate about one apparent anomaly about that service implementation and received a 'stock' reply after a few days. As I have some past experience on both supplier and consumer side of software to run such systems and have from time to time managed such operations, I thought would respond with some suggestions. I'll see if there is any follow on to that.
Of course if the preference for the modern age is to run support mainly through social media ... then I am lost! 😉
The information that is appearing in the FAQ section as starting to look rather useful and relevant on a daily basis. That is encouraging but it may appear to be a bit frustrating trying to find the information if one is sort of 'old school' and used to Contents pages and Index listings. Especially when it is a work in progress, or seems to be, and therefore not yet complete. Or at least that is the impression it leaves with me for now.
I suspect it needs a little time to bed down but fear we the users will not be keen to allow that time given the usual chaos of updates from all directions at this time of year. Far too much potential for disruption.
As for my cameras - a couple of older Canons, mostly - 1D3 and a 600D used mainly for some manual only lenses I have.
However I have a collection of photos from various sources that I keep for testing and out of interest. Or I will download form a decent RAW file providing source if something seems specific to a camera. Today that led me to discover some rather interesting images from a Ricoh GR III .
So have sample from Nikon, Sony, Fuji mainly. Plus Canon of course and Phase to see what extremely high resolution looks like.
I'm also still using Windows 7 so that may be a factor influencing my perception especially in the area of stability.
In testing I had a tricky to spot problem with a high res Fuji file that turned out to be a problem with the GPU kernel build activity. Somewhat obscure and only really at all obvious when viewing fine detail and 100% or 200%. Getting a working kernel fixed that one.
I have seen, sometimes, the NR challenge you mention but it was rather specific to the content of the image (with the sample files I had available) and the details setting in the NR tool for those images did not seem to be adding anything more by cranking it up. It was what I would consider to be a rather tricky file due to content but the results comparing V12 and V20 side by side were significantly better an more flexible for V20. At least they were on my screen.
That has been true for all the images I have looked at provided they actually had some noise to manipulate!
I would happily take a look at any problem images you can point me to.
Have just been looking at an EOS R file shot at a floodlit soccer game using ISO6400 and have not spotted any artifacts with both luminance and detail set to max. Though there is little point in that because the setting counter each other on this image.
HDR looks like it will take some getting used to and the effect seems to vary a lot depending in the image and starting exposure level. My impression is that I can get results I prefer but at the moment, for many candidate files, it seems to be taking me more time.
That said, historically I have not used the Auto adjustments feature much other than individually for Levels but trying the functionality in V20 is leading me to reconsider that.
The basic colour editor is, of course, optional but now available in layers which in many situations for many people might suit well. I like and enjoy it and for small tweaks in certain types of subject I think it will be very useful. Rebalancing skin tones in a head and shoulder portrait? Probably not. But tweaking a landscape to post on social media ...? Sure. And no doubt also for things somewhere between the two once fully familiar with it.
It seems to work pretty well on my Notebook trackpad. I full admit that is not necessarily the final word about it!
Grant0 -
I'm currently using an Olympus E-M1X (and before that I was testing v. 20 with a Canon M6 Mk II and a Canon 1D x) and - sorry, C-M-B - I don't recognise any of the problems you're perceiving, least of all NR issues: in fact it's the significantly-improved high ISO NR that has drawn me back to Capture One.
And as a bird/wildlife photographer, I'm obsessive about fine detail: I can't and won't use a converter whose NR kills detail.
That said, my workflow might be different to other people's: I never change the NR from their default 50/50/50 settings, and apply no sharpening to the overall image, but instead "paint in" sharpening where I want it, along with as much "Clarity" - "Natural" mode, Clarity slider usually left alone but Structure applied as high as I need to go - to make the details pop.
I also prefer the new HDR implementation, and the Basic Colo(u)r Editor is wonderfully fast, intuitive and efficient.
And it's 100% reliable so far on my Win 8.1 box.
In fact I'll go as far as to say that - for the first time ever - I'm seeing, in Capture One 20, a Raw converter as a "one stop" workflow solution. I'm simply not seeing the need to further post-process images once they're out of Capture One 20: I'm even OK with the watermarking solution, now that I'm using a PNG rather than a text watermark.0 -
Keith Reeder wrote:
I'm currently using an Olympus E-M1X (and before that I was testing v. 20 with a Canon M6 Mk II and a Canon 1D x) and - sorry, C-M-B - I don't recognise any of the problems you're perceiving, least of all NR issues: in fact it's the significantly-improved high ISO NR that has drawn me back to Capture One.
And as a bird/wildlife photographer, I'm obsessive about fine detail: I can't and won't use a converter whose NR kills detail.
That said, my workflow might be different to other people's: I never change the NR from their default 50/50/50 settings, and apply no sharpening to the overall image, but instead "paint in" sharpening where I want it, along with as much "Clarity" - "Natural" mode, Clarity slider usually left alone but Structure applied as high as I need to go - to make the details pop.
I also prefer the new HDR implementation, and the Basic Colo(u)r Editor is wonderfully fast, intuitive and efficient.
And it's 100% reliable so far on my Win 8.1 box.
In fact I'll go as far as to say that - for the first time ever - I'm seeing, in Capture One 20, a Raw converter as a "one stop" workflow solution. I'm simply not seeing the need to further post-process images once they're out of Capture One 20: I'm even OK with the watermarking solution, now that I'm using a PNG rather than a text watermark.
So you don't have any of these strange artifacts showing up when using noise reduction?
I'm glad it works for you!
Painting in sharpening is okay for wildlife or portraits but doesn't work for fashion, landscape, architecture or interior photography. You might argue that interior photography and landscapes don't require NR as they would always be made at low ISO settings on a tripod - but recently I had to do an interior shoot for a client with only a short windows of 15minutes and I wasn't able to use a tripod most of the time.
So high ISO was necessary and with artifacts showing up during noise reduction...well that's not going to work.0 -
I don't see why you couldn't paint in sharpness with the other four genres.
Use a monopod?0 -
C-M-B wrote:
So you don't have any of these strange artifacts showing up when using noise reduction?
These artifacts are extremely hardware and driver dependent. Please create a support-ticket, if you haven't done so already.0 -
C-M-B wrote:
- The noise reduction tool is not very useful for modern cameras with a very high frequency noise, causing artifacts in the finale image unless you reduce the details by a lot.
There will always be a limit as what can be recovered. Compared to V12, the new noise reduction is a huge step forward. Especially recovering and retaining small luminance- and color detail.C-M-B wrote:
- The dynamic range tool does not bring any actual improvements, simply working with levels and curves bring the exact same results and it's not even quicker.
Levels/Curves are quite different. The new tool only affect a subset of the histogram, with a non-linear roll-off. I like to use the Blacks for pulling the deeps shadows after having recovered with the Shadows-slider in examples.C-M-B wrote:
- The new color editor tool is a gimmick and far too limited in reality and gives no advantage compared to the previous version. It looks nice in the videos but it's super clunky and tedious, imprecise and is very limited in its range and results...
That is the point of a Basic color editor and it's new Direct Color Editor cursor. Get basic things done in hurry.
If you have more advanced and precise application, then Advanced Color Editor is the answer.0 -
C-M-B wrote:
So you don't have any of these strange artifacts showing up when using noise reduction?
Nope - and in fact I don't even know what you mean by "strange artefacts".Painting in sharpening is okay for wildlife or portraits but doesn't work for fashion, landscape, architecture or interior photography.
Can't see why it wouldn't work.So high ISO was necessary and with artifacts showing up during noise reduction...well that's not going to work.
What's "high" ISO to you?
Which camera?0 -
Christian Gruner wrote:
Again: it seems that this is only true for certain noise-types. Not all noise is equal.C-M-B wrote:
- The noise reduction tool is not very useful for modern cameras with a very high frequency noise, causing artifacts in the finale image unless you reduce the details by a lot.
There will always be a limit as what can be recovered. Compared to V12, the new noise reduction is a huge step forward. Especially recovering and retaining small luminance- and color detail.
A bit different,but combined they have a very similar effect. It's certainly notC-M-B wrote:
- The dynamic range tool does not bring any actual improvements, simply working with levels and curves bring the exact same results and it's not even quicker.
Levels/Curves are quite different. The new tool only affect a subset of the histogram, with a non-linear roll-off. I like to use the Blacks for pulling the deeps shadows after having recovered with the Shadows-slider in examples.
C-M-B wrote:
- The new color editor tool is a gimmick and far too limited in reality and gives no advantage compared to the previous version. It looks nice in the videos but it's super clunky and tedious, imprecise and is very limited in its range and results...
That is the point of a Basic color editor and it's new Direct Color Editor cursor. Get basic things done in hurry.
If you have more advanced and precise application, then Advanced Color Editor is the answer.
Quick, sure - but it's much too imprecise and clumsy to get things done in a hurry, especially with a tablet input. (which bring me to another point : they still have not fixed the tablet input issues)0 -
Keith Reeder wrote:
C-M-B wrote:
So you don't have any of these strange artifacts showing up when using noise reduction?
Nope - and in fact I don't even know what you mean by "strange artefacts".
Here's a comparison between a 5D II and a GFX 50s at the same ISO.
https://i.imgur.com/00wh8o2.jpg
Where do you paint in sharpness in an architecture image or a landscape image that has no sky?Painting in sharpening is okay for wildlife or portraits but doesn't work for fashion, landscape, architecture or interior photography.
Can't see why it wouldn't work.
So high ISO was necessary and with artifacts showing up during noise reduction...well that's not going to work.
What's "high" ISO to you?
Which camera?
ISO 1000+ on a Fujifilm GFX 50s.0 -
C-M-B wrote:
A bit different,but combined they have a very similar effect. It's certainly not
If we are to get a bit more technical about it, import or photograph a B/W gradient in CO.
Observe the histogram while you do your adjustments, and you'll how different they are!
Basically the new sliders have very specific, non-linear limitations to their luma ranges, the Levels tool affect the whole histogram.0 -
C-M-B wrote:
Keith Reeder wrote:
C-M-B wrote:
So you don't have any of these strange artifacts showing up when using noise reduction?
Nope - and in fact I don't even know what you mean by "strange artefacts".
Here's a comparison between a 5D II and a GFX 50s at the same ISO.
https://i.imgur.com/00wh8o2.jpg
Where do you paint in sharpness in an architecture image or a landscape image that has no sky?Painting in sharpening is okay for wildlife or portraits but doesn't work for fashion, landscape, architecture or interior photography.
Can't see why it wouldn't work.
So high ISO was necessary and with artifacts showing up during noise reduction...well that's not going to work.
What's "high" ISO to you?
Which camera?
ISO 1000+ on a Fujifilm GFX 50s.
That artefacts only show up when you put everything up to 100%
I share my magic settings with you:
Luma:70
Detail:0
Color:40
Single:27
Sharpening:
strength: 291
radius: 0,3
threshold: 0
Halo: 00 -
C-M-B wrote:
Where do you paint in sharpness in an architecture image or a landscape image that has no sky?
Wherever you need the sharpening? So not on OOF areas, distant features, etc.
You seem to be trying to make architecture and landscape photography into unique use-cases as far as workflow and processing are concerned, and they really aren't...0 -
Keith Reeder wrote:
C-M-B wrote:
Where do you paint in sharpness in an architecture image or a landscape image that has no sky?
Wherever you need the sharpening? So not on OOF areas, distant features, etc.
You seem to be trying to make architecture and landscape photography into unique use-cases as far as workflow and processing are concerned, and they really aren't...
I pretty much always need sharpening, that's the key to getting detail and getting images that look great and snappy even with heavy crops.
IMO a lot of camera and software developers intentionally produce softer images to soften the grain - but it comes at the cost of details.0 -
Tom-D wrote:
That artefacts only show up when you put everything up to 100%
I share my magic settings with you:
Luma:70
Detail:0
Color:40
Single:27
Sharpening:
strength: 291
radius: 0,3
threshold: 0
Halo: 0
I don't think I would like to have images with these settings compared to mine. I'd rather have more grain (and maybe no noise reduction at all) than images with no real detail...0 -
I've been looking at some more high ISO Sony images today. From an RX10 MkIV
My conclusion, across a number of subject types and light sources is that the default settings C1 offers are pretty good for all of the types but can be specifically enhanced for most situations.
However, for a given ISO there are no guaranteed magic numbers - at least not with the images samples I have obtained.
Much depends on the subject and how far one can (or is prepared to) go with the NR vs regular editing settings to get a result.
Now that concept in itself is not a great revelation but I was surprised at what I found for the sets of images I have available from 3 or 4 different Sony models. The content and the type of shot in general seem often to demand some very individual adjustments to several parameters (however slight the adjustment might be) to make the very best of a shot.
However the key to the artefacts problem seems to be to avoid seeking to force too much detail. Go with NR details level setting first and then take care with Clarity and Sharpening and anything else that seeks to work through some sort of Contrast based tweaking.
Using layers as a matter of course is probably going to be a good habit.
That is more or less what I had anticipated but the significance of combined small adjustments across different tools is often more sensitive than I was expecting - at least with the sample files I have available.
Grant0 -
I think this all comes back to what others have already criticized: a lack of documentation/manual for Capture One. And I'm not talking about a simple description of the tool (as can be accessed by clicking on the "?" - but a real documentation showing the practical use of it.
If users (including myself) knew about the limitations and special requirements of new tools there would be much less confusion, disappointment and anger all around!
Right now I feel like PhaseOne is turning its users into beta-testers and quasi-supporters, which allowed them to cripple their own support service for the software they ask an extraordinary amount of money.
A few years ago I defended the price of C1 and one of my arguments was the excellent support and feedback from support. Now that's gone I'm kind of left wondering what's really left to justify it (especially when it comes to upgrades such as this).0 -
Something like this?
https://blog.captureone.com/clean-sky-bugs-bees-spot-removal-tool/0 -
C-M-B wrote:
Right now I feel like PhaseOne is turning its users into beta-testers and quasi-supporters
Ooh - and you were doing so well in avoiding the "we're all just beta-testers now" cliché...
😉
Phase One provides an absolute ton of "how to..." guidance - I can't think of anyone else who even gets close.
Remember that there is also "The Image Quality Professor" - which Grant has pointed at - and Paul Steunebrink, both of whom provide excellent tutorials.
Phase One does use video a lot. While I accept that not everyone is a fan of video as a medium for "training" (I'm not a fan myself) it exists; and along with the perfectly usable written Help guides that are available, I really don't buy the idea that Phase One isn't doing its bit here, at all.
I'd further argue that for a lot of the information that you want relating to:the limitations and special requirements of new tools
normal "Help files" aren't the place for it anyway.
Furthermore, I really don't think Capture One is that hard to figure out just through using it. I've yet to find anything I've wanted to do in Capture One that I couldn't figure out by just trying - with reference to F1 as needed.0 -
Keith Reeder wrote:
C-M-B wrote:
Right now I feel like PhaseOne is turning its users into beta-testers and quasi-supporters
Ooh - and you were doing so well in avoiding the "we're all just beta-testers now" cliché...
😉
Phase One provides an absolute ton of "how to..." guidance - I can't think of anyone else who even gets close.
Remember that there is also "The Image Quality Professor" - which Grant has pointed at - and Paul Steunebrink, both of whom provide excellent tutorials.
Phase One does use video a lot. While I accept that not everyone is a fan of video as a medium for "training" (I'm not a fan myself) it exists; and along with the perfectly usable written Help guides that are available, I really don't buy the idea that Phase One isn't doing its bit here, at all.
I'd further argue that for a lot of the information that you want relating to:the limitations and special requirements of new tools
normal "Help files" aren't the place for it anyway.
Furthermore, I really don't think Capture One is that hard to figure out just through using it. I've yet to find anything I've wanted to do in Capture One that I couldn't figure out by just trying - with reference to F1 as needed.
See the thing is if they actually fixed the bugs that have been present since V11 or sometimes V9 it wouldn't have to be a clichée. Not my fault! And now they messed up the whole support system too so yeah, that's mostly a bad beta tester experience.
C1 is not hard to figure out but it has many quirks and unexpected results to regular, normal, expected inputs.0 -
SFA wrote:
Something like this?
https://blog.captureone.com/clean-sky-bugs-bees-spot-removal-tool/
Hahaha yeah for V7. No.0 -
C-M-B wrote:
SFA wrote:
Something like this?
https://blog.captureone.com/clean-sky-bugs-bees-spot-removal-tool/
Hahaha yeah for V7. No.
What has changed that makes this redundant?0 -
I am also very unhappy with the v20 changes.
On the color editor, why can't I change hue still oly +-30? I want to change blue to red or something similar!
Why can't I generate a mask out of a selected color?
The new noise reduction is really strange with higher level noise (ISO >1000 on my Z7) you need to reduce the details to close to Zero when you want to reduce noise completely. DXO Photolab and Topaz Noise AI are still ahead and needed at least for >2000 ISO for me.0 -
Franz1 wrote:
On the color editor, why can't I change hue still oly +-30? I want to change blue to red or something similar!
You can, but only 30 degrees round the colour wheel at a time. In the Advanced tab of the colour editor, create a new layer, and brush a rough mask over the blue object. (I'm looking at an image that includes a blue plastic chair.) Use the picker and click on the blue object. Pull the hue slider all the way to the right. It will get as far as a mid purple colour. Click in the same place again, and pull the slider all the way to the right again. And again. The third time got my dark blue chair to a slightly maroonish red. A fourth go, if the slider is pulled all the way to the right (30) again, got me past red to orange. About 10 to 15 got me to a good red. It's much quicker and easier to do than to describe.Why can't I generate a mask out of a selected color?
You can do that too. On the background layer, go to the Advanced tab of the colour editor again. Click on the blue object (my blue chair again). Check the box for view selected colour range. In my image that shows that I have included a light blue object (my grand-daughter's head-dress in the school nativity play!) as well as the dark blue chair: so fine tune the selected colours by dragging the edges of the wedge until only what you want is included. When you are satisfied, uncheck show selected colour range again. Click the ... on the colour editor tool, and choose Create Masked Layer From Selection.
Ian0 -
Franz1 wrote:
I am also very unhappy with the v20 changes.
On the color editor, why can't I change hue still oly +-30? I want to change blue to red or something similar!
Why can't I generate a mask out of a selected color?
The new noise reduction is really strange with higher level noise (ISO >1000 on my Z7) you need to reduce the details to close to Zero when you want to reduce noise completely. DXO Photolab and Topaz Noise AI are still ahead and needed at least for >2000 ISO for me.
Topaz looks very Artifical sometimes and does also produce some artefacts sometimes.
And it does take a lot of time. 20 seconds with a RTX5000 per image. (Batch with photoshop, 10 seconds load image and ai denoise and 10 seconds denoise a 24 mp image.)
So, denoising 1200 wedding images take a lot of time and power.
CO does only take a split second with almost the same quality.
For advanced denoising i reactivated neat image, together with CO it does a really good job and does only take 1 second for a 15.5 mp image. And it does a difference between a iso 100 and iso 6400 image automatically. So you dont loose detail in iso 100 and get nice denoise at iso 6400. Ai denoise does eat up skin and clothes detail
I did sone testing with DXO and besides the name "Prime" its not very better but slower.0 -
Ok, understood that I can change color more than +-30 on hue buy multiplying work.
I also understoot that I can make a mask from a color region, will test that if it helps on the rare cases I need to change colors.
On DXO, the reason why I use DXO for 90% of the pictures is that it has a lot of automatic tools to help.
On the noise reduction, I just use standard for every photo and sort my folder then by ISO and apply prime to photos >=ISO 1000. Only photos >ISO6400 I sometimes (if I need them at all) I correct with Topaz.
The individual settings for noise reduction in normal or Pro mode I leave to DXO and I Never changed so far!
The same is with correcting horizont or vertical lines. I can set that to Auto and DXO will correct the level of the picture automatically.
In C1 there are some automatic corrections for lightning and exposure/HDR but you cannot influence the and therefore have to correct always manually.
In general I edit only 3+ star pictures which are maybe 10-20% of the pictures I take but even then 80% of them I can handle with PL3 automatically and the rest I do with C1 in detail.
A view I finalize in Luminar4, mainly if I need to correct faces (eyes, lips, skin).
As you may see the pattern I miss a bit the ability of C1 to look at the picture and base some decisions (what is an Eye, What is an face, what is skin, what is a horizontal/vertical line, what is noise or bokeh) to help to make changes on the different parts.
p.s.: Here is a comparison of what I could get out of C1 v12 (right side) and Topaz AI (left side): It's Nikon Z7 ISO 6400 natual light indoor shot.
https://photos.smugmug.com/photos/i-4GTZKkX/0/95953919/O/i-4GTZKkX.png0 -
Franz1 wrote:
Ok, understood that I can change color more than +-30 on hue buy multiplying work.
I also understoot that I can make a mask from a color region, will test that if it helps on the rare cases I need to change colors.
On DXO, the reason why I use DXO for 90% of the pictures is that it has a lot of automatic tools to help.
On the noise reduction, I just use standard for every photo and sort my folder then by ISO and apply prime to photos >=ISO 1000. Only photos >ISO6400 I sometimes (if I need them at all) I correct with Topaz.
The individual settings for noise reduction in normal or Pro mode I leave to DXO and I Never changed so far!
The same is with correcting horizont or vertical lines. I can set that to Auto and DXO will correct the level of the picture automatically.
In C1 there are some automatic corrections for lightning and exposure/HDR but you cannot influence the and therefore have to correct always manually.
In general I edit only 3+ star pictures which are maybe 10-20% of the pictures I take but even then 80% of them I can handle with PL3 automatically and the rest I do with C1 in detail.
A view I finalize in Luminar4, mainly if I need to correct faces (eyes, lips, skin).
As you may see the pattern I miss a bit the ability of C1 to look at the picture and base some decisions (what is an Eye, What is an face, what is skin, what is a horizontal/vertical line, what is noise or bokeh) to help to make changes on the different parts.
p.s.: Here is a comparison of what I could get out of C1 v12 (right side) and Topaz AI (left side): It's Nikon Z7 ISO 6400 natual light indoor shot.
https://photos.smugmug.com/photos/i-4GTZKkX/0/95953919/O/i-4GTZKkX.png
That looks like much too great a difference in sharpness alone.
Can you share the file (or something similar for which you have a comparison) so that I can compare what you have from C1 to what I see on my system?
Grant0 -
I need to do some tests with new 2020 engine but when using NR slider on higher ISO's (~>=4000) the noise get well managed if you turn slider to 80+ but artifacts start as well and therefore you need to turn details down to reduce artifacts.
This is a bad implementation of the interface because you always need to tweak two sliders.
I checked then with ISO 64 and 400 images and found out that actually you need in some cases to adjust the noise color to reduce fringing!
Also noise reduction is doing something (set to 50) even if there is no noise to remove! Also details are on 50 and it seems to sharpen picture a bit (which I usually do not expect from the noise reduction settings.
The noise tool should not have any impact on RAW conversion aside of noise reduction.
Especially with new Z lenses I do not want that basically an additional sharpening happens on already super sharp images.0 -
Your previously posted smugmug side by side example image suggested that that the C1 rendition was completely soft - no sharpness at all.
So what is different between that example and your more recent testing?
As I understand the NR tool the initial presentation of the image with sliders set to 50 represents (or is intended to represent, what is considered to be the best overall result for a typical file for that camera combination using the recorded ISO.
Of course it may not be to everyone's taste and it may not work sometime from image to image - partly depending on content. However all I can say is that so far in my observations across a range of files from different cameras, the results, as presented and with no changes from me, look pretty good and I only see artefacts when I push the images a lot for detail retention in dark areas where the resolution of the detail content at that point in the image is marginal. Effectively trying to use NR to create detail where none really exists. In which case it would likely be better to apply it on a layer and unmask the problem patches - assuming it was worth setting the correction values to the level that induces the artefacts in the first place.
For other people things may be different, but that has been my observation.
One option, presumably, would be to set different starting values and save those as the preferred defaults?
And of course there may be some difference of opinion about the best initial settings to be used although I think the only way to start a discussion with the C1 team about that observation would be to start a Support Case.
HTH.
Grant0
Post is closed for comments.
Comments
35 comments