Skip to main content

⚠ī¸ Please note that this topic or post has been archived. The information contained here may no longer be accurate or up-to-date. ⚠ī¸

Anybody has experience with this weird preview problem?

photo by FA
Hi all,

Please see the image

http://i63.tinypic.com/10fx7pt.jpg

Has anybody experienced with this problem? This happens after creating a preview and if I see the original RAW file on my iMac with Preview (mac app) it has this issue but if I see it on my iPad there is no issue. Regenerating the preview doesn't work as well.

Any ideas?

Cheers,

Fatih
0

Comments

17 comments

Date Votes
  • John Doe
    If you see the same artefact with Preview.app, then the RAW file is probably corrupted. I would guess what you see on your iPad is the embedded JPEG thumbnail?
    0
  • photo by FA
    [quote="John Doe" wrote:
    If you see the same artefact with Preview.app, then the RAW file is probably corrupted. I would guess what you see on your iPad is the embedded JPEG thumbnail?


    But before generating the preview there isn't such a problem.

    See the picture below, this is still on the SD card, reading with Preview.app doesn't create a problem as well

    http://i65.tinypic.com/hs03fd.jpg

    Once I create the preview, CO corrupts the image. In fact if I re-import the image there is no problem.
    0
  • John Doe
    You said "if I see the original RAW file on my iMac with Preview (mac app) it has this issue ", so I don't understand. Please clarify.
    0
  • photo by FA
    [quote="John Doe" wrote:
    You said "if I see the original RAW file on my iMac with Preview (mac app) it has this issue ", so I don't understand. Please clarify.


    Sorry for the confusion 😊

    Steps I take;

    1) Insert the SD card
    2) Copy the images from SD card to a folder on my Desktop. Reason is I GPS tag them.
    3) Import the images to the CO with the import tool, I use referenced file system. CO imports the images on the desktop to appropriate location on my NAS drive based on my settings
    4) CO starts to create the previews.
    5) Some of the files randomly creates this problem. When I go back to the both files on my desktop and on CO referenced file system, I see them with Preview.app as corrupted.

    If I go to the very original file in the SD card, I see exactly the same image as not corrupted. If I import that image to the CO separately, CO imports without any corruption.

    Hope this helps.

    Cheers,

    Fatih
    0
  • photo by FA
    Well apparanrently this is OpenCL problem and according to Tech Support my GPU is not compatible with CO, reminding that it is the latest iMac 5K and yet it is not supported. Some funky business 😊

    So all the lines on the release notes about supporting iMac 5K basically doesnt mean anything now ☚ī¸
    0
  • George Barron
    [quote="fatihayoglu" wrote:
    Well apparanrently this is OpenCL problem and according to Tech Support my GPU is not compatible with CO, reminding that it is the latest iMac 5K and yet it is not supported. Some funky business 😊

    So all the lines on the release notes about supporting iMac 5K basically doesnt mean anything now ☚ī¸


    I've got a new iMac 27" 5K. I have not had this problem. I have also heard conflicting things about the GPU support. Some say the GPU is not compatible but I'm pretty sure a PO employee said this was not true in this very forum.

    I will say that I cannot tell any performance difference with OpenCL off or on.
    0
  • photo by FA
    I've thought theyve worked on iMac5k's and make them compatible but the TS sent me a list of compatible GPUs and sadly on that list there is no iMac5k GPU. Just to tedt this, I have turned off OpenCL settings and imported the exact same 500+ pictures and none of them created this problem. So from my experience I have to say that PO didnt do a good job on here.

    PS my GPU is AMD Radeon R9 M395X. What is yours, as apparently R9 290 and R9 290X are compatible.
    0
  • Eric Nepean
    [quote="fatihayoglu" wrote:
    I've thought theyve worked on iMac5k's and make them compatible but the TS sent me a list of compatible GPUs and sadly on that list there is no iMac5k GPU. Just to tedt this, I have turned off OpenCL settings and imported the exact same 500+ pictures and none of them created this problem. So from my experience I have to say that PO didnt do a good job on here.

    PS my GPU is AMD Radeon R9 M395X. What is yours, as apparently R9 290 and R9 290X are compatible.


    Here is a conversation on this user forum with Christian Gruner of Capture One 6 weeks ago where he tells me the complete opposite of what Tech Support has told you:



    🙄 🤓

    It would be very interesting to our user community if you could post the comments from Tech Support to you.
    0
  • SFA
    I would have thought this sort of problem has more to do with the drivers for the GPU card installed than a generic "compatibility" problem. Especially since, as reported, it is not consistent for every image or even, if I read things correctly, the same image file when processed more than once.

    So the theory of compatibility for OpenCL (for example) should be consistent BUT there is nothing to stop external influences causing consistency challenges either directly or indirectly.
    0
  • photo by FA
    Here is the link of supported GPUs by CO9

    https://www.phaseone.com/en/Search/Arti ... nguageid=1

    This is provided by TS;


    Hi Faith,

    This is not a bug, but rather a compatibility issue with the graphics card. More on this here: https://www.phaseone.com/en/Search/Arti ... nguageid=1

    Kind regards,

    Phase One Technical Support
    0
  • photo by FA
    So sadly the top of the line GPU which is R9 395X is not supported but mid level R9290 is supported.
    0
  • SFA
    [quote="fatihayoglu" wrote:
    So sadly the top of the line GPU which is R9 395X is not supported but mid level R9290 is supported.


    I have always assumed that the list provided was intended to provide a general list at the start (back in C1 V7 and V8 days when everyone was new to the ways of OpenCLPU card vendors) and more recently to list every card that had been specifically tried because there had been some problems reported.

    There are far too many cards from far too many board manufacturers to be able to explicitly test all of them even once.

    In the case of a PC it might be possible to take a machine to a lab and run tests on every card ever produced but even than there are so many possible hardware combinations that the effort is unlikely to be proof of 100% successful working in all situations. Even if such proof existed it could be invalidated by the next driver code change.

    For those Macs where the hardware spec. is not upgradable it would require every developer making use of OpenCL (or whatever feature they might have a specific use for) to obtain and test every possible combination of configuration and then retest after every software update.

    Apple may have enough resources to do that (although when some users have issues with OS updates one wonders whether they actually do it ....) and Microsoft ought to be in that position with their own brand hardware. However I think the rest of the world cannot be expected to keep track of everything all the time based on a guess that someone somewhere may have a problem with a product component that produces unexpected results.

    Having glanced at a review of the hardware type you mention and noting that the problem does not always appear consistently, have you checked for possible GPU temperature issues? The apparent randomness of the problem that arises does tend to suggest that the cause may related to something that is not consistent - like the unit's temperature for example.

    I could be way off target with that but it is one of many things that may be worth considering as potential sources of the problem.


    Grant
    0
  • photo by FA
    Hi Grant,

    While I agree with you about the fact that software developers cannot test every work on every single piece of hardware, there that's why they use universal codes. I am quite sure that PO hasn't tested CO on every model of hard disk from every single manufacturer, every model of processor form every single manufacturer, every model of GPU from every single manufacturer or every single model of RAM from every single manufacturer but then CO still works. So I am not quite sure why CO is not compatible with AMD R9 395X GPU.

    About the temperature problem, my iMac is pretty clean so I don't think the possible dust inside makes the fans less efficient and also I don't think the temperature increases too much as I have a temperature gauge and it is stable.

    Cheers,

    Fatih
    0
  • SFA
    Hi Fatih,

    On the temperature thing - in the article I read it was not so much a question of blocked cooling as much as it was that the GPU naturally tended to run rather hot. I can't vouch for that report of course and it sounds like you have already monitored your system and seen no problem.

    As for universal codes - if something is a frozen development there ought to be no problems but for anything that is still evolving and open, trying to offer a performance edge or adapt to new connected hardware there will always be opportunities for problems to creep in somewhere.

    Of course that can be frustrating but it seems to me tat the only low risk option would be to avoid pushing the performance envelope both as developer and consumer.

    People may not want to accept that limit.

    That said, using C1 9.3 to investigate some 40+ Mb Canon 5D4 files the on screen responses I have seen for my Windows notebook have been great and this from a specification which is now about 4 years old in terms of CPU and running a very lowly old Nvidia GPU.

    In fact my early impression is that, for processing, the practical response time from input to display is not different to what I see with files from older cameras that are only about one third the size. The stop watch may tell a different story but I doubt it would make much difference to my personal working speed.

    Of course for output processing the result may be different but that is something I have yet to try.

    One possible marginal benefit (should be unnecessary) from running Windows is that there is freedom to go outside Windows support and update software directly from the vendor to see if a problem can be resolved. As I understand it that is not so possible in Mac world. That may result in corrections or workarounds taking much longer than might be hoped.


    Grant
    0
  • Eric Nepean
    It's kind of interesting to see that for COP 9, the only suppported GPU for Mac is Intel HD Graphics 520, no GPU cards are listed. There are GPU cards listed for Mac, but only for COP8 and earlier.
    The situation is worse than it was at the beginning of the summer.
    0
  • photo by FA
    Unfortuantely Eric, now I am really dissappinted with PO. The reason is, while we were complaining about the performance of CO9 on iMac 5k computers, all we were hearing was the GPU is not strong enough for this screem but in reality the GPU wasnt supported by PO at all. Maybe if it is fully supported, the performance would be much better.
    0
  • Eric Nepean
    Fatih, I feel the same as you.

    I work in the telecom industry developing SW. When you have a diversity of platforms and complex features, (as we do, but also PO) the way you maintain quality and reliability is by defining test cases that span your feature set (you can't test everything, but you test "landmark" cases), and these are run on key platforms and configurations.

    It's not an easy task to set up such a test plan, and it requires discipline in execution and identifying and resolving failures.

    Once you have it, there is no confusion about supported configurations and you have strong confidence that they work. I don't perceive that confidence and clarity here.
    0

Post is closed for comments.