Neue Testversion 18.40b9

Hier diskutieren die Betatester von PhotoLine untereinander und mit den Entwicklern
Benutzeravatar
Herbert123
Mitglied
Beiträge: 2140
Registriert: Sa 12 Mai 2012 21:38

Re: Neue Testversion 18.40b9

Beitrag von Herbert123 »

A request for spot channel support, and a possible implementation solution!

Spot channels in Photoshop are solved by adding them to the channel panel. In Photoline spot channels could be implemented by adding support for a new layer type "spot color": in principle the same as a grey scale layer, but with an additional option to set the (named) colour. In Photoline this colour would be used for display.

When the file is saved as a tiff, pdf. or pdf (file formats which support spot channels). this spot colour layer is then converted to an actual spot channel that is compatible for press work.
/*---------------------------------------------*/
System: Win10 64bit - i7 920@3.6Ghz, p6t Deluxe v1, 48gb (6x8gb RipjawsX), Nvidia GTX1080 8GB, Revodrive X2 240gb, e-mu 1820, 2XSamsung SA850 (2560*1440) and 1XHP2408H 1920*1200 portrait
Benutzeravatar
photoken
Mitglied
Beiträge: 2162
Registriert: Sa 28 Sep 2013 01:25

Re: Neue Testversion 18.40b9

Beitrag von photoken »

photoken hat geschrieben:If I have time, I'll do a test of DCB with and without RT's CA correction.
OK, I've done the test and RT's application of CA correction before doing the DCB demosaic does make a difference.

The test image is the same one I used in a previous post -- ISO=80, 28mm (35mm Format equivalent). These images were processed using the RT "Neutral" profile, which means nothing was done to the .RW2 RAW image. The only processing parameter that changed was enabling or disabling the CA correction. In both cases, I used 5 DCB iterations, DCB enhanced mode, and 3 false color suppression steps. These off-axis details are reproduced here at 100%. The results were output as 16-bit TIF images, and cropped in PL.
RT CA comparison.png
Not only is the image sharper with CA enabled, but also there are some jagged edges in the image without CA correction, although the differences are slight.

The CA correction alone does not account for the enhanced sharpness of the RT images shown in my previous post. The RT implementation of sharpening, de-fringing, micro contrast, etc., all add up to the better image quality as opposed to what I can achieve in PL.

What's the bottom line? I think that unless PL can enhance its CA correction as well as its sharpening methods, and provide some form of de-fringing and also micro contrast, there will always be a noticeable difference in quality between PL and RT. However, as I've said in another thread, the PL results are quite good for non-critical work; and I'll be providing an updated PL adjustment layer preset for the LF1 when PL19 (presumably with an updated dcraw having support for the LF1) is released.

These results are valid for my setup, of course, and even though I won't use fewer than 5 DCB iterations that doesn't mean other users with different cameras and lenses wouldn't benefit from the full set of DCB options. All of us will probably use DCB "Enhanced" mode, as well....
Du hast keine ausreichende Berechtigung, um die Dateianhänge dieses Beitrags anzusehen.
Ken
Yes, I think it can be eeeeeasily done....
Just take everything out on Highway 61.
Benutzeravatar
Hoogo
Betatester
Beiträge: 4021
Registriert: So 03 Jul 2005 13:35
Wohnort: Mülheim/Ruhr

Re: Neue Testversion 18.40b9

Beitrag von Hoogo »

I have to put the pictures above each other and add an adjustment for histogram to spot a difference. And even then I can not tell which version looks better, and after clicking an unknown time on the eye I can't tell at which version I'm looking...
----------------
Herr Doktor, ich bin mir ganz sicher, ich habe Atom! /Doctor, doctor, I'm sure, I've got atoms!
Benutzeravatar
photoken
Mitglied
Beiträge: 2162
Registriert: Sa 28 Sep 2013 01:25

Re: Neue Testversion 18.40b9

Beitrag von photoken »

Hoogo hat geschrieben:I have to put the pictures above each other and add an adjustment for histogram to spot a difference. And even then I can not tell which version looks better, and after clicking an unknown time on the eye I can't tell at which version I'm looking...
Yeah, I screwed up -- I thought I was providing 200% images, but something went wrong BKAC*. :oops:
Since I deleted everything before realizing the error (and have no desire to redo the whole thing), you're just gonna have to trust me on this if viewing the provided images at 200% doesn't show the differences....

_________________________________________
*BKAC: Between Keyboard And Chair
Ken
Yes, I think it can be eeeeeasily done....
Just take everything out on Highway 61.
bkh
Betatester
Beiträge: 3674
Registriert: Do 26 Nov 2009 22:59

Re: Neue Testversion 18.40b9

Beitrag von bkh »

photoken hat geschrieben: Since I deleted everything before realizing the error (and have no desire to redo the whole thing), you're just gonna have to trust me on this if viewing the provided images at 200% doesn't show the differences....
No problem scaling it up in PL to see the differences. I've tried to achieve something similar using demosaicing (dcb_dcraw with Q=5) and PL's CA correction and a radius 1,4 colour denoise to clean up colour artefacts:
ca_corr_in_post.png
Btw. I don't see any obvious demosaicing artefacts in either of your versions.

But of course this was an example of strong CA which survives dcb. I guess that the most problematic case is when you just have very slight CA (near the image centre) which gets filtered out by dcb (and probably by a lot of other demosaicing algorithms), and which is then re-introduced (in the opposite direction) if you correct for CA on the image border by just stretching the entire image. Difficult to judge if this is better or worse than losing resolution in the pre-demosaicing scaling.

Cheers

Burkhard.
Du hast keine ausreichende Berechtigung, um die Dateianhänge dieses Beitrags anzusehen.
Benutzeravatar
Hoogo
Betatester
Beiträge: 4021
Registriert: So 03 Jul 2005 13:35
Wohnort: Mülheim/Ruhr

Re: Neue Testversion 18.40b9

Beitrag von Hoogo »

You are seriously discussing CA's that are invisible without help from a histogram correction and zoom? Well, you surely have the wrong lens for that. If you need some pictures with CA's, I can give it to you ;)
----------------
Herr Doktor, ich bin mir ganz sicher, ich habe Atom! /Doctor, doctor, I'm sure, I've got atoms!
Benutzeravatar
photoken
Mitglied
Beiträge: 2162
Registriert: Sa 28 Sep 2013 01:25

Re: Neue Testversion 18.40b9

Beitrag von photoken »

bkh hat geschrieben: Btw. I don't see any obvious demosaicing artefacts in either of your versions.
Yeah, again, I apologize for the poor quality examples. In RT at 200%, there are noticeable artifacts along the edges of the dark red triangle and within the type.
bkh hat geschrieben: Difficult to judge if this is better or worse than losing resolution in the pre-demosaicing scaling.
Yep. The important thing for us to remember is that image quality depends on the total package of adjustments (and their implementation) in the programs. My most recent test had the hypothesis that CA correction before demosaicing was the primary factor in the sharpness of the final image. That proved not to be the case -- although there is a noticeable difference, it's only part of the story.

Also, we shouldn't loose sight of the quality of the image as a whole. It's useful to focus on a small area to examine the effect of a specific adjustment, but it's the overall impression of the image that counts. This was really driven home to me recently as I continued to refine my processing profile for the LF1 in RT. I had one image where a small area showed some colour fringing, so I added the de-fringing adjustment and used varying amounts of de-fringing for the colour spectrum from cyan through blue through magenta and including red. That cured the problem in that image, but I was really surprised when I applied that profile to an image that did not have any noticeable fringing. The whole image suddenly looked "snappier", and had the general impression of being sharper with smoother tonalities. In short, it looks like an image should look from a Leica lens....
Ken
Yes, I think it can be eeeeeasily done....
Just take everything out on Highway 61.
bkh
Betatester
Beiträge: 3674
Registriert: Do 26 Nov 2009 22:59

Re: Neue Testversion 18.40b9

Beitrag von bkh »

photoken hat geschrieben:
bkh hat geschrieben: Btw. I don't see any obvious demosaicing artefacts in either of your versions.
Yeah, again, I apologize for the poor quality examples. In RT at 200%, there are noticeable artifacts along the edges of the dark red triangle and within the type.
Of course, I can see the difference in CAs, even at 100%. But I still can't see anything that looks like a demosaicing artefact, let alone one caused by CAs, even if I look at your sample image at 200% (which should be the same as looking at it in RT). Maybe you can just mark the relevant sections in the sample somehow?

Cheers

Burkhard.
bkh
Betatester
Beiträge: 3674
Registriert: Do 26 Nov 2009 22:59

Re: Neue Testversion 18.40b9

Beitrag von bkh »

Ich habe hier noch ein Problem mit "Vektorebene optimieren" – der Befehl funktioniert für die angehängte Ebene nicht. Hat sie zu viele Punkte? Egal welche Werte ich eingebe, weder die Vorschau noch der Befehl selbst ergeben irgendwelche Änderungen.

Entstanden ist die Vektorebene übrigens aus einer Auswahl mit dem Magnetlasso in einem 24 MPx-Foto.

L.G.

Burkhard.
Du hast keine ausreichende Berechtigung, um die Dateianhänge dieses Beitrags anzusehen.
bkh
Betatester
Beiträge: 3674
Registriert: Do 26 Nov 2009 22:59

Re: Neue Testversion 18.40b9

Beitrag von bkh »

Ich habe hier noch ein Problem mit "Vektorebene optimieren" – der Befehl funktioniert für die angehängte Ebene nicht. Hat sie zu viele Punkte? Egal welche Werte ich eingebe, weder die Vorschau noch der Befehl selbst ergeben irgendwelche Änderungen.

Entstanden ist die Vektorebene übrigens aus einer Auswahl mit dem Magnetlasso in einem 24 MPx-Foto.

L.G.

Burkhard.
Du hast keine ausreichende Berechtigung, um die Dateianhänge dieses Beitrags anzusehen.
Benutzeravatar
Hoogo
Betatester
Beiträge: 4021
Registriert: So 03 Jul 2005 13:35
Wohnort: Mülheim/Ruhr

Re: Neue Testversion 18.40b9

Beitrag von Hoogo »

Vielleicht ist da irgendwo ein einzelner Punkt ausgewählt?
Ich hab das Vektorwerkzeug aktiviert und Strg+A gedrückt. Alle Punkte wurden Rot, und es funktionierte dann auch.
----------------
Herr Doktor, ich bin mir ganz sicher, ich habe Atom! /Doctor, doctor, I'm sure, I've got atoms!
bkh
Betatester
Beiträge: 3674
Registriert: Do 26 Nov 2009 22:59

Re: Neue Testversion 18.40b9

Beitrag von bkh »

Hoogo hat geschrieben:Vielleicht ist da irgendwo ein einzelner Punkt ausgewählt?
Stimmt, danke, das muss es wohl sein – merkwürdig nur, dass das auch dann einen Einfluss hat, wenn z. B. das Ebenenwerkzeug aktiv ist. (Ist das Verhalten eigentlich sinnvoll, wenn nur ein Punkt ausgewählt ist – dann gibt es doch gar nichts zu optimieren, oder?)

L.G.

Burkhard.
Benutzeravatar
OldRadioGuy
Mitglied
Beiträge: 416
Registriert: Fr 24 Apr 2009 19:09
Wohnort: Austin, Texas, USA

Re: Neue Testversion 18.40b9

Beitrag von OldRadioGuy »

Restoration of a customized panel layout does not work correctly

1. Drag the Toolbox panel from the left of the PL Window to the right, docking next to the default panels on the right side of the PL Window.
2. Open the Simple Browse panel and drag to the left, docking it on the left side of the PL window.
3. Go to View -> Panel Layout -> Save Panel Layout and save customization.
4. Go to View -> Panel Layout -> Reset Panel Layout and panel layout is restored to default with Toolbox on left and default panels on the right.
5. Go to View -> Panel Layout -> and select the customization name appearing at the top of the option list.
6. The operation does not restore the customization of the Toolbox nor the Simple Browse panel. In fact, the Toolbox is not switched on again, requiring going to Panel Layout to turn it on.

Bob
PhotoLine 24.xx |DxO PureRAW 3 |Various Third-Party Plugins | macOS 14.11 | Apple M2 Max | 64 GB Memory | E-M1markIII.
Benutzeravatar
photoken
Mitglied
Beiträge: 2162
Registriert: Sa 28 Sep 2013 01:25

Re: Neue Testversion 18.40b9

Beitrag von photoken »

bkh hat geschrieben:But I still can't see anything that looks like a demosaicing artefact, let alone one caused by CAs, even if I look at your sample image at 200% (which should be the same as looking at it in RT).
"Artifacts" might have bee the wrong word for me to use. What I was attempting to describe was the jagged outer edge of the red triangle that's most noticeable on the left of the triangle. Instead of being a smooth "line" (blurred), it's composed of steps or blocks of colour. Same is true of the type.
Ken
Yes, I think it can be eeeeeasily done....
Just take everything out on Highway 61.
Benutzeravatar
Herbert123
Mitglied
Beiträge: 2140
Registriert: Sa 12 Mai 2012 21:38

Re: Neue Testversion 18.40b9

Beitrag von Herbert123 »

A small request! Would it be possible to include a ppi resolution setting for the the "new document" dialog and "new layer" dialog?

I ask because there is a ppi option in the document settings panel, but not in the dialog. It would also be very handy to be able to set the dimensions AND ppi for new layers.

Btw, instead of "DPI" the resolution units in the layer and document dialogs and panels shouldn't it be "PPI"? Because the bitmap resolution is meant, not the dpi (dots) on paper?
/*---------------------------------------------*/
System: Win10 64bit - i7 920@3.6Ghz, p6t Deluxe v1, 48gb (6x8gb RipjawsX), Nvidia GTX1080 8GB, Revodrive X2 240gb, e-mu 1820, 2XSamsung SA850 (2560*1440) and 1XHP2408H 1920*1200 portrait