Jump to content
You must now use your email address to sign in [click for more info] ×

Equalizer

Members
  • Posts

    25
  • Joined

  • Last visited

Everything posted by Equalizer

  1. It's not quite that simple after all. Other manufacturers manage to do that as well. And in this respect, this would also be expected from Serif. But I also sent that to DXO at the time. In fact, Serif and DXO tested and used my material in a longer process. And the error was reproducible. But they did not find a solution. In an earlier version of Luminar (I think version 4) I had no problems. With Luminar AI they were there again. And also initially in NEO times the purple spots were there again. I had also informed Skylum about the problem. My error was reproduced there as well. But there was no continuous solution until today. By the way, I do not use any software from DXO for this reason. For sharpening and denoising I switched to Topaz. Topaz has the advantage that firstly real DNGs are created. And secondly, Topaz can also process non-RAW images. That is an huge advantage.
  2. The problem with purple spots has been known for a long time. But unfortunately Serif is not working on a solution. And even with Affinity Photo V2 nothing was changed, the same RAW engine, the same error, no improvements. Too bad. Here potential is given away little by little. And if it goes on like this, I'm out of here. Meanwhile, I use Affinity Photo almost not at all. Many functions that other manufacturers have long since AI-supported, are in Affinity Photo only with much effort feasible. On the subject of DXO...the problem is actually quite simple: DXO does not create "real" DNG, i.e. those with real RAW content. These are linear DNGs, which are no longer RAWs, unlike the DNGs created by Adobe's DNG converter.
  3. First of all, thank you very much for all the advice. I tried around a lot myself. I agree that there is a problem with the DNG file created by DXO. This is a linear DNG in which the Bayer interpolation, application of the lens profiles, noise reduction and sharpening have already been carried out and noted in the DNG file. The other areas such as brightness, colors, white balance etc. have not yet been changed and are fully available for editing. And this is where the DNG file generated by DXO differs from a DNG file generated by the Adobe DNG Converter. In this case only the Bayer interpolation is carried out. The rest stays as it is. It is not a linear DNG. And then when I open this DNG file in Affinity Photo (or Luminar AI, which I also use) there are no purple spots in the burned out lights. I conclude that the DNG file generated by DXO does not adhere to the DNG conventions and, depending on the RAW converter, problems then arise. What puzzles me is that in the predecessor of Luminar AI, namely the Luminar 4, none of these problems occur with the DNG file generated by DXO. This leads me to suspect that the RAW engine and the implementation of the recognition of a DNG file by Luminar AI is different from the RAW engine of Luminar 4. And that in turn leads me to conclude that there is also a problem with the serif RAW Engine there. If you activate the Apple Core Engine in Affinity Photo on a MAC computer, the problem does not occur. All in all, I conclude from this that there are RAW converters that don't quite get along with the DNG file that DXO generates, because it deviates from the conventions of a DNG file. This currently applies to the RAW Engine from Affinity Photo (Serif RAW Engine), Luminar AI, RawTherapee and Darktable. The RAW engines of Adobe Photoshop, Luminar 4, Photolemur, Aurora HDR 2019 and Apple Core Engine have no problems. I have already reported the problem to DXO (so far without feedback) and Skylum (Luminar). The people from Skylum now want to take a close look at my DNG files.
  4. Many thanks for your help. But I've already read that, but here, too, there are no solutions that help. The problem only occurs with the DNG or jpg files generated by DXO. For example, if I generate a DNG with the Adobe DNG Converter, there are no problems. Lights are white. I think that the DNG generated by DXO deviates from the norm and therefore the error occurs. Affinity's RAW converter doesn't seem to be able to cope with this either. Programs like Darktable, RawTherapee and Luminar AI also have the problem. Interestingly, however, the error does not occur in Luminar 4, the predecessor of Luminar AI. This could mean that the RAW engine has been changed in Luminar and there is an incompatibility here. It's a shame, because I use both Affinity and Luminar and would have liked to use DXO PureRaw as a preparation for my RAW development.
  5. The DNG files created by DXOPureRaw are incorrectly interpreted in Affinity Photo. This leads to a purple discoloration in the highlights. This problem also occurs in Luminar AI. Interestingly, this bug does not occur in Luminar 4, Photolemur, Aurora HDR and Photoshop CS6. So it's a bug that only occurs in Affinity Photo (and Luminar AI). See sample images. Why is that? Please help. I'm using the latest version of Affinity Photo and Windows 10.
  6. Hi Mark, Loading is only delayed when loading the first image after starting the program. Every further picture after that is loaded quickly and the controllers react immediately.
  7. Hi Walt, thank you for the information. I have now installed the beta version 1.9.1.952. In fact the problem is gone. Very good!!!
  8. Hi Mark, here is the logfile. Log.txt If I deactivate the GPU acceleration, there is no delay in loading the images. There is also no mistake in focus stacking. When editing the pictures you can of course see that the processor takes over the work, the fan turns up significantly. The changes in the picture when a controller is operated are then somewhat more jerky.
  9. With version 1.9 Canon RAW images open more slowly if GPU acceleration is switched on. If a controller is operated after loading an image, it takes time for the image to change. That always happens with the first picture that I load. For the other images, loading takes longer with GPU acceleration too, but the controllers then react faster.
  10. The focus stacking only works correctly if I switch off the GPU acceleration. With GPU acceleration it works a little bit faster, but in the end the last lower third of the picture is pink-black. System: Windows 10, Intel 6700k, Graphic: HD530
  11. I sold Affinity Photo for iPad. I like the app. But: I just edited 3 raw files, saved them on the iPad and in the cloud. The processed files each have a size of approx. 300 Mb. But almost 3 GB of data is also stored on the iPad, which cannot be deleted. Even after I have deleted the pictures on the iPad, the almost 3 GB of data are still available ... Why does this happen? Why can't I delete the data? There is no button in the app to delete this data. Please be sure to correct this bug in the app. Please be sure to add a function in the app to be able to clear the cache. Otherwise the app cannot be used like this.
  12. I have no problems with Canon CR2 Raw files. I see you only have 8 GB of RAM. That is very little in image processing. The error you described seems to be due to something other than the Canon RAW files. Does the error occur with or without GPU acceleration? Are you using an integrated graphic or a separate one. More information could help the developers better understand the problem.
  13. Update to the new beta 834: Plugins do work. Had forgotten to check the box that all, even unknown plugins are allowed. So Luminar 4 and Aurora 2019 work.
  14. My first brief impression with the new Beta 834: The new Beta works not fine with the iGPU Intel HD 530. When opening a picture, again it takes a few seconds longer with GPU acceleration than without to display the picture. The sliders hook again, not working smooth enough as you would expect. In particular, it always takes an eyelash after pressing the slider until it reacts. It works better without GPU acceleration. A focus stack is still not usable. The lower third of the picture remains black. Part of the image is colored yellow or green. My plugin for Luminar 4 and Aurora 2019 does not work. They are not displayed. However, the Nik Collection is displayed and works.
  15. Hello kirkt, Thank you for your help. The lcp file is actually fine because it works fine in Photoshop. I copied it out of the Adobe Camera Raw Profile folder, but there seems to be an error converting it to an xml file. In the meantime I am using the profile of the Tamron model A012. This is the previous model of my lens. That seems to work well, distortion and vignette are calculated out very well. I just thought that with a profile to the Tamron lens A041 (my lens) the results would be even more accurate. But that's okay.
  16. Hi Dan, thank you for your help. The fact is that I couldn't generate the profile. I have to admit that as a Windows user I am used to simply opening a window, then a menu appears and everything works quite simply. But the operation with the Phyton Script overwhelms my knowledge and skills. Perhaps there will be someone who will explain that to me again. My idea was that there might be someone in the forum who might also use the Tamron lens and have an xml profile for Affinity Photo. That's why I asked here in the forum.
  17. Dear fan base, I need an xml lens profile for a Tamron lens. Can someone help me and convert the following profile to an xml file? Unfortunately, this doesn't work on my computer (Windows 10) with the instructions from lensfun and the Phyton3 script. Many Thanks. Canon (TAMRON SP 15-30mm F_2.8 Di VC USD G2 A041E) - RAW.lcp
  18. Uwe, thanks for the answer. I just tested this too. Without OpenCL 6.03 seconds, with Open CL 2.66 seconds. So it's really much faster. However, OpenCL does not work properly with the Intel HD 530, there are still too many bugs that make it impossible to work with hardware acceleration at this time.
  19. Hi Uwe, what exactly has become faster? Can you describe that in more detail? You surely don't mean the complete RAW development... What kind of hardware CPU/GPU do you use?
  20. I agree that it would make more sense to give more information about the new functions of what exactly has been changed in order to better coordinate the tests and give better feedback about the functionality.
  21. Unfortunately no improvement in the errors with built-in Intel HD graphics. My processor has an Intel HD 530. As with Affinity Photo Version 780, opening an image takes too long with activated hardware acceleration. It takes about 3-4 seconds for the image to be displayed in the Develop Persona. When it is displayed, the image is initially greyish, pale and far too dark and cannot be edited. Another 3-4 seconds pass. Only now is the picture displayed normally. And only now can you edit it. When operating the individual controllers in the Develop Persona, the program takes 1-3 seconds to think about each controller until it implements the controller changes. The hardware acceleration is still absolutely unusable. Without hardware acceleration, everything works as usual.
  22. Ich habe die aktuelle Win 10 Version mit allen aktuellen Treibern. Getestet wurde Affinity Photo Beta 1.9.9.780 Beta. Prinzipiell fällt auf, dass sich das Programm bei jeder Aktion eine kurze Pause gönnt, bis es die Aktion umsetzt. Schon beim Öffnen eines Bildes fängt es an. Das Bild wird geladen und dann dauert es auf einmal einige Sekunden (zwischen 2-3 bis zu 7-10 Sekunden) bis ich mit der RAW-Entwicklung starten kann, obwohl mir das Bild bereits angezeigt wird. Bediene ich dann einen Regler in der Develop-Persona, dauert es ca. 1-2 Sekunden, bis die Aktion auch umgesetzt wird. Nicht bei jeder Aktion ist die GPU aktiv: Beim Fokusstacking werden erst einmal die Bilder ausgerichtet (ohne GPU-Last), dann erfolgt das eigentliche Stacking und erst da wird die GPU verwendet. Leider flackert das Bild dabei wild umher, und am Ende fehlt das untere Drittel des Bildes, welches einfach schwarz ist und damit unbrauchbar. Nach meinem ersten kleinen 15-Minuten-Test hinterlässt die Beta sehr gemischte Gefühle. Denn so kann man die Hardware-Beanspruchung nicht nutzen. Wenn die Hardware ausgeschaltet wird, funktioniert alles Beschriebene wie immer.
×
×
  • Create New...

Important Information

Terms of Use | Privacy Policy | Guidelines | We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.