ACR 6.3 Bug - Introducing Noise

I just upgrading from CS4 to CS5 the other day for PS, including 6.3 of ACR.  I've had something happen in the workflow of the images I'm working on, but you can duplicate it below (or tell me if you can't).
In PS, create a 200x200 pixel image.  Fill with white.  Save as a .tiff .  Close the file.
Open Bridge.  Find file.  Open in Camera Raw.  No adjustments.  Click open, to open it in PS.
Back in PS, make a levels adjustment layer.  Drag the black point all the way to the right.  Check out the random grey noise like pixels that have suddenly appeared everywhere.  Try it with default settings in ACR, change anything you want.  I can't get them not to appear.
I don't think this is just me.  Anyone else want to verify?

function(){return A.apply(null,[this].concat($A(arguments)))}
sjlocke215 wrote:
I would prefer not to add another two steps to my workflow by processing this or that way to get around it.
I agree that the dithering regardless of settings is a bug - the Camera Raw color transform should respect the Color Settings choice you have made in the Color Settings preferences panel of Photoshop, and it does not:
Are you opening your images into Photoshop?
If you set the converter to output 16 bit data, then just work in Photoshop with the 16 bit data, then Save As or Save For Web & Devices there aren't really extra steps...  And all your processing will be in the more accurate 16 bits/channel color depth, which yields better quality images overall.
I would also suggest converting to the ProPhoto RGB color space to avoid channel clipping due to gamut issues.
It's pretty clear that because of quirks in Camera Raw, extra steps may be warranted to get the best possible image quality.
-Noel
P.S.,  There was another thread where this was discussed extensively, mid to late last year.

Similar Messages

  • N80 Gallery Bug introduced in firmare 5.0719.0.2

    In case you guys from Nokia didnt notice it yet ...
    There was a new bug introduced into the Gallery application of the Nokia N80 with Firmware 5.0719.0.2.
    The bug looks as following:
    If you open the Picture Gallery there simply wont be any pictures or videos shown (from both internal memory and memory card). All one can see are the two weblinks in the gallery.
    Only thing that help to bring the Gallery back to normal condition is to reboot the phone.
    Several people who have upgraded to 5.0719.0.2 have this problem. And it definitely was not there before in any other Firmware Version 3.x or 4.x.
    Is Nokia already aware of this bug and will fix it in a future firmware release? It really annoys me a lot. So much that i already think of buying a new phone from another manufacturer.
    Besides ... Why doesnt Nokia release the 5.0725.0.1 firmware for European N80s? Maybe the bug is already corrected in this newer firmware...

    Hey
    To tell the truth this is one of the most common most annoying and of course older bug from N73, N80, N91.
    Later nokia introduced new gallery in N93, N93i and N91 8GB.
    I discussed a tip here:/discussions/board/message?board.id=smartphones&me​ssage.id=54784&page=1
    It would just **bleep** me off.Message Edited by korngear on 04-Oct-200704:31 PM

  • [svn] 3149: Fix bugs introduced by the display object sharing optimization.

    Revision: 3149
    Author: [email protected]
    Date: 2008-09-08 16:58:35 -0700 (Mon, 08 Sep 2008)
    Log Message:
    Fix bugs introduced by the display object sharing optimization. This checkin should fix ordering problems when dynamically adding/removing graphic elements, and ordering problems when changing graphic element properties that require a display object (rotation, alpha, filters, etc).
    Bugs:
    MXMLG-219: BitmapGraphic component content always rendered from 0, 0 origin.
    MXMLG-220: setting visible property of Rect and Ellipse does nothing.
    MXMLG-222: setting the alpha of a rectangle changes z-order when there are 3 or more rectangles
    SDK-16754: Z-order incorrect when a Group is present
    QA: Yes - the test files from these bugs (or something equivalent) should be added to our standard test suite
    Reviewer: Deepa
    Ticket Links:
    http://bugs.adobe.com/jira/browse/MXMLG-219
    http://bugs.adobe.com/jira/browse/MXMLG-220
    http://bugs.adobe.com/jira/browse/MXMLG-222
    http://bugs.adobe.com/jira/browse/SDK-16754
    Modified Paths:
    flex/sdk/trunk/frameworks/projects/flex4/src/flex/core/Group.as
    flex/sdk/trunk/frameworks/projects/flex4/src/flex/graphics/BitmapGraphic.as
    flex/sdk/trunk/frameworks/projects/flex4/src/flex/graphics/graphicsClasses/GraphicElement .as

    This is the great thread ;-)
    My favorite threads are below.
    Munky posted great reply :8}
    SQL Query Help
    order in a connect by

  • [svn] 1388: Fixing a bug introduced for DefineFont3 equality testing.

    Revision: 1388
    Author: [email protected]
    Date: 2008-04-24 10:19:44 -0700 (Thu, 24 Apr 2008)
    Log Message:
    Fixing a bug introduced for DefineFont3 equality testing. DefineFontAlignZones already checks its DefineFont3 for equality and so this goes back to the old behavior and avoid an infinite equality check.
    Doc: No
    QA: Yes, also please add a test case that makes use of two, duplicated embedded fonts with advanced anti-aliasing to the automated mxunit testsuite.
    Bug: SDK-15378
    Ticket Links:
    http://bugs.adobe.com/jira/browse/SDK-15378
    Modified Paths:
    flex/sdk/trunk/modules/swfutils/src/java/flash/swf/tags/DefineFont3.java

    I just posted something about this too, and while I feel that its wrong to "bump" an old post, this issue has still stayed unresolved and my frustration is peaking. In my thread my problem is i can't even use the KXdrivers on an x64 operating system, so this things an expensi've paper weight. Great.
    Creative makes good sound cards, NO DOUBT about that. Their customer service on the other hand however, is absolute trash. I have read that "statement" as well, and it seems an easy way out to just say "yeah we addressed it, so please, shut up about it already". And I agree, it is NOT a fix, for all of us. If one would take the time to read the forum, its not only ABIT motherboards. Even in that one thread, one if the first couple of replies a guy says "but I have a gigabyte motherboard". Surprise surprise, he was ignored by the mod who posted that (she answered some other poster's question with "read the instructions please" . Please creative, give us some faith back in your company by helping out the many of us with this problem. Only then will I reccomend and buy your products.

  • H.264 Video Encoding Bug Introduced?

    I'm not sure if the issue is related to a fix for a previous bug, or just a new bug introduced for this player.
    My original problem was with the reporting of Camera.activityLevel. It only reports 0 and 100 when encoding with H.264 in FP11+ release, Sorenson reports the results properly. I was told it was fixed in this beta, which it does appear to be
    See this post: http://forums.adobe.com/message/3999479
    However I now see a new problem in this beta.  My application allows me to change the video settings on my H.264 encoded stream (quality, resolution, fps, etc) during the broadcast. And it works perfectly in FP11+. However, with the beta it now appears to cause problems and be broken.
    Changing settings on an existing stream (it looks to be when changing from 320x240 to 640x480) causes streaks of red, green, and blue vertical lines over the  video image and the aspect ratio is wrong. The image appears squished - everything short and fat like mabe it's increaing the width value, but not the height. I'm testing the beta on Windows 32bit IE7 with a Logitech 9000 and on a different machine using Windows 32bit IE8 with a Logitech C910 with the same results.
    The camera.width & camera.height properties to seem to report their values properly, but the stream as viewed on the server is definitely not at the correct resolution.
    Can anyone verify this or advise? This version looks like it will break my app if released so thankfully it's still a beta.
    Thank you!

    OK, in FP11 Beta 2 it appears you have fixed this bug (aspect ratio messed up when switching from 320x240 to 640x480), but now you've reintroduced the Camera.activityLevel bug again. It only reports 0 and 100, this was fixed in Beta 1.
    To summarize:
    FP 11 release
    Camera.activityLevel - broken, only reports 0 and 100
    Changing camera size from 320x240 to 640x480 - works
    FP 11.2 beta 1
    Camera.activityLevel - fixed
    Changing camera size from 320x240 to 640x480 - broken, turns video from color to black and white and has a strange aspect ratio
    FP 11.2 beta 2
    Camera.activityLevel - broken, only reports 0 and 100
    Changing camera size from 320x240 to 640x480 - works
    Hopefully you'll have everything working in beta 3.
    Do I need to submit a bug report for the Camera.activityLevel bug in FP 11.2 Beta 2?

  • [svn] 3780: Fix for bug introduced yesterday when adding clipContent to FxContainer and FxDataContainer .

    Revision: 3780
    Author: [email protected]
    Date: 2008-10-21 10:51:40 -0700 (Tue, 21 Oct 2008)
    Log Message:
    Fix for bug introduced yesterday when adding clipContent to FxContainer and FxDataContainer. The default value needs to be true.
    Also fixing a bug in SystemManager since application isn't always an IRawChildrenContainer anymore. This was previously fixed and got clobbered in a merge.
    QE Notes: None
    Doc Notes: None
    Bugs: SDK-17423, SDK-17697
    Reviewer: Hans
    Ticket Links:
    http://bugs.adobe.com/jira/browse/SDK-17423
    http://bugs.adobe.com/jira/browse/SDK-17697
    Modified Paths:
    flex/sdk/trunk/frameworks/projects/flex4/src/mx/components/FxContainer.as
    flex/sdk/trunk/frameworks/projects/flex4/src/mx/components/FxDataContainer.as
    flex/sdk/trunk/frameworks/projects/framework/src/mx/managers/SystemManager.as

    The network admin solved the problem:
    “When I recreated the User Profile Service, I did not enable “NetBIOS Name” on the service, so all user accounts did not have the proper domain, which is why none of the targeting worked.
    All users were listed as corp\username instead of rcd\username.”, Network Admin.
    Further information: http://www.windowsclever.com/sharepoint/netbios-name-is-different-domain-name-user-profie-service-sharepoint-2010.html

  • Another bug introduced by 6u10 ?

    Here is another possible bug introduced by 6u10. Looks like that the feature of ‘setLookAndFeel’ under 6u10 breaks Mouse Listener.
    See the short test program below, which works fine under 6u7 but fails under 6u10.
    Here are the test steps :
    1.     Click on ‘menu’ but don’t click on any menu items.
    2.     Move mouse to the white area (JPanel)
    3.     Click mouse once. Nothing happens but it should draw a red dot. The mouse listener does not recognize the first click. The subsequent clicks are fine.
    If you perform this test under 6u7, it works fine ( recognize the first mouse click) . If you take out setLookAndFeel, it works fine also. Anyone knows a workaround.
    import javax.swing.*;
    import java.awt.*;
    import java.awt.event.*;      
    public class MouseClickTest  extends JFrame
         static int sxw=800;
         static int sxh=800;
         int mx, my;
         JPanel jpl;
         public MouseClickTest()
              init();
         private void init()
              jpl =new JPanel()
                   public void paintComponent(Graphics g)
                        Graphics2D g2=(Graphics2D) g;
                        g2.setColor(Color.white);
                        g2.fillRect(0,0,800,800);
                        if ( mx > 0 && my > 0 )
                             g2.setColor(Color.red);
                             g2.fillOval(mx,my, 20,20);
              jpl.setPreferredSize(new Dimension(sxw, sxh));
              jpl.addMouseListener(new MouseAdapter(){
              public void mousePressed(MouseEvent e)
                   mx=e.getX();
                   my=e.getY();
                   jpl.repaint();
              JMenuBar mb = new JMenuBar();
                      JMenu menu = new JMenu("menu");
                      for(int i=0; i<5; i++)
                   menu.add(new JMenuItem("Item "+i));
                      mb.add(menu);
              setJMenuBar(mb);
              getContentPane().setLayout(new BorderLayout());
              getContentPane().add(jpl, BorderLayout.CENTER);
              setSize(800,800);
              pack();
              setVisible(true);
               addWindowListener(new WindowAdapter() {
                     public void windowClosing(WindowEvent ev) {
                              System.exit(0);
         public static void main(String[] args)
              try {
                   //this is ok.
                   //UIManager.setLookAndFeel(UIManager.getCrossPlatformLookAndFeelClassName());
                   //these are bad
                   //UIManager.setLookAndFeel(UIManager.getSystemLookAndFeelClassName());
                             UIManager.setLookAndFeel("com.sun.java.swing.plaf.windows.WindowsLookAndFeel");
                   //UIManager.setLookAndFeel("com.sun.java.swing.plaf.nimbus.NimbusLookAndFeel");
                  catch (Exception e) {     }
              SwingUtilities.invokeLater(new Runnable()
                       public void run()
                        new MouseClickTest();
    }

    I can confirm that the behavior is different in 6u10. If a JMenu or JPopupMenu is displayed, and the user clicks on some other part of the GUI to hide the popup, that part of the GUI doesn't actually receive (or react to) a MouseEvent. If you click on an unselected tab, for example, that tab doesn't become the active tab, whereas in 1.5 and 1.4 (and according to the OP, 1.6 pre 6u10), it did.
    Is this really a "bug"? I suppose it is. Native Windows appears to send mouse events to items when a popup menu is dismissed with a click. But why does your application rely on this behavior - that is, why is this change so serious to you?
    Oh, and XOR painting isn't broken in 6u10, it's just unusably slow. :)

  • New bug introduced with Keynote 5.2 - Set timing and order of each build has stopped working properly on Smart Builds

    Hi there,
    We're heavy users of Keynote and have noticed a new bug has recently been introduced into something that was stable. I think it has emerged in Keynote 5.2, could have been Keynote 5.1 also.
    It's in the area of Set Timing and Order of each build when used on a Smartbuild. You can no longer set the Start Build field for each instance within the Smart build when the instances are interspersed between other items. The Start build field is just blank and unselectable.
    To see this :
    1. Create a bullet point list with 4 lines
    2. Set build to Appear
    3. Set "Set Timing and Order of each Build" to on
    4. Create a Smart Build - set to flip
    5. On the Smart Build Set "Set Timing and Order of each Build" to on
    5. re-order the items in your Build Order so that they alternate between the bullet point and the smart build flip
    At this point, you used to be able to set the bullets so that they Started on Click and then set the smart builds to Start with previous. This gave you a nice and very useful effect where you could click to reveal a bullet point and at the same time flip in a corresponding image. We've used this loads of times in our company presentations.
    With a recent keynote update ( I think 5.2 maybe, could have been 5.1) this functinality has been lost. have tried it on Lion and Mountain Lion and it's not an OS issue, its a Keynote issue.
    Could anyone from Apple acknowledge that they know the issue and suggest if it will get fixed? Right now we're hunting round for an old Keynote 9 DVD so we can install a pre build 5.2 version.
    Thanks

    Hi glennelliott72,
    I've noticed the bug you describe and to me it's very irritating.
    I also found some differences in the build menu between the last release and the previous one, and between the Mac OS X and iOS versions. For instance, the blast build doesn't exist anymore in Keynote 5.2 but it's still in the version for iOS 6.
    What is your experience about?
    I currently use OX Lion.
    Thank you

  • Changing Brush size in ACR 7.2 - bug oder feature?

    The update to ACR 7.2 changed the way the adjustment of the brush size is behaving. Before one pressed the right mouse button and moved left or right to change the size. Now it's not just left to right but right to left or up and down depending on where the tiny mouse indicatior on the outmost circle is placed. The has as an effect that it's not immediately clear in which direction the mouse has to be moved to change the size. Also the brush circle is move to the side while adjusting and is not congruent with the brushcircle after releasing the mouse botton.
    Any ideas if this is a feature or a bug?

    The ACR7.3 update did not fix the brush resize problem on my system either - also Win7x64 Ultimate, Ps 13.0.1 (currently I use only the x64 install but the problem exists with ACR hosted in the 32-bit install as well) and ACR 7.3.0.71.
    Another annoying thing I've noticed with 7.3 (maybe it was there before 7.3 but I never noticed it) is that even if you set the Feather slider to zero pixels, as soon as you set the brush size to 10 pixels or more, the brush is shown as being feathered. Maybe the 10 pixel cutoff is related to my screen size and resolution: 1920x1200 and 88ppi (NEC MultiSync LCD 2690 WUXi2) but it is still disconcerting at any size.
    The unexpected movement of the centre of the brush when you right click reported by the OP, another annoying characteristic, is always towards the centre point of your screen. If you are left of centre, it moves right and vice versa. If you are above centre it moves down and vice versa. Combinations move diagonally towards the centre.
    Eric Chan seemed to acknowledge the mouse right-click brush resize and brush centre movement problems with ACR 7.2 in his post #1. Looks like he's forgotten us or has been too busy with other matters. Either way , it would be nice to hear from him again with an update.

  • RAW conversion bug with Noise Reduction

    Hello,
    I have found a serious bug in the RAW conversion when noise reduction is applied. When converting from two types of Canon RAW files (a CRW from a Powershot G6 and a CR2 from a 20d) I found that if you apply Noise Reduction to a RAW file on very low settings (the default setting in the NR function will produce this reliably) single-pixel lines appear at regular intervals throughout the image. Here is an example:
    You can see several lines in this image:
    http://farm1.static.flickr.com/140/3821480263171e76604b.jpg
    A 100% detail of which is here:
    http://farm1.static.flickr.com/179/382148021af6586d27eo.jpg
    Has anyone else had this problem? Can someone from the Aperture dev team fix this?
    -Steve G

    Well I find this filter is quite good in 'masking' block artifact that codec like xvid, or other low compression codec have. I only apply it if I find the block artifact is too much and I find this filter is less offending to my eyes than the block artifact.
    In manual it said that if you have noisy video and want to lower the size then you can use this filter. It also blur the video a bit. But I suspect it is more than blur as I try gaussian blur in time line and the result is not as good. You can see the result as well. There is the tab between source and target and you can compare the result by togling between source and target tab.
    BTW, anyone with 1 core, dual, or quad core, can you tried to encode with it? Just cancel it after few minutes as I want to see what is your processor utilization with this filter on. Also you can see how long does it take to process this video from the 'estimation time left'.

  • Bugs introduced in EA releases

    Hello guys,
    I just want to mention few bugs which were introduced in your last three EA releases and which are really annoying...
    1) When you have multiple table/view tabs open and browsing amongst those tabs, it happens the grid is reloaded. It looks like the database is queried again. I noticed this happens only once for that particular table or view. This is making me crazy when there is some complicated query behind that view or when you have filter set since this will clear the filter and reloads the table without it.
    2) Almost everytime when I have some table open and switch between Columns tab to Data tab I get warning message "Connection is currently busy. Try again?" If I click abort it continues to load the table...
    3) If there are connection problems, or simply when the connection is lost and you want to open a table from the table tree, SQL Developer get stuck and you have to shoot it down via task manager. I bet this has to do something with the Reconnect feature. Guys if it doesn't work, don't put it there. Current implementaion of reconnect is useless anyway.
    4) When opening new tables, I very often get "Data stream already closed" error message. Then you have to refresh the table...
    I don't have time to investigate how exactly reproduce these bugs, you should do it. But I think, anyone who is working with latest release of SQL Developer, has to find these bugs as well.

    Unfortunatelly I have to say all these bugs are present in the production release as well...

  • ACR 7.3 Bug on Adjustment Brush tool - changing brush size problem.

    I update photoshop and using ACR 7.3 and now having big problem when I use Adjustment Brush took. Phe problem is when try to change brush size with holding right mouse button and moving the mouse. Behavior completely changed and its impossible to predict if I move mouse right what will happen.
    Please check solution for it. old behaviour was great. I think this is not new feature, its bug.
    Thanks in advance.

    No need.
    Already bugged and acknowledged: http://feedback.photoshop.com/photoshop_family/topics/bug_crash_adjustment_brush.

  • Another ACR 4.4 bug?

    I noticed this post over on FM. An ACR/PS user encountered a weird issue where an ACR-converted image in PS went completely black when the mode was changed to LAB inside PS. The issue was resolved when he went back to v4.3.
    Has anyone else seen this or can they reproduce it if they have 4.4 installed?
    The problem may have been related to the way the user installed/de-installed 4.4 and the situation is maybe a bit unique but if this is really a result of 4.4, I think it implies other problems than just EXIF data.
    I am holding off on updating but if there is anyone who has already installed 4.4 and would like to try to reproduce the anomaly, it might be helpful for the community.
    Link:
    http://www.fredmiranda.com/forum/topic/628466

    Ah, okay. Whew. I was able to revert to ACR 4.3.1 by doing the following (and it fixed the bug):
    - Go to http://www.adobe.com/support/downloads/detail.jsp?ftpID=3822 and download the .8bi file (e.g., download to the desktop)
    - Make sure your Bridge and PS programs are closed
    - Copy the downloaded .8bi file to:
    \Program Files\Common Files\Adobe\Plug-Ins\CS3\File Formats
    - Restart bridge, open any file in RAW, and you'll see that the version at the top of the window says 4.3.1. Open the image
    - In PS3, Lab color now works properly.
    Best of luck to others with this issue.

  • Another bug introduced in production release - table filter

    Hello guys,
    check this scenario:
    1) Open some table
    2) Use the column filter to filter the table
    3) Try to use the column filter on same column again without clearing all filters.
    The table filter will display only "Remove ('xyz')" message, where xyz is current value used to filter the column. It is not possible to delete the value and enter new value. You have to clear the filter first and then enter new value...
    Sadly, I have to say, with every new release you are making SQL Developer more and more unusable :(

    My bad.
    Logged Bug 13788551 - regression: have to remove quickfilter before entering new one.
    -Raghu

  • Data-corrupting JPEG bug introduced in Lr 3.4, CR 6.4

    FYI in case you are unaware.....
    I noticed this on the Rob Galbraith web-site ( http://www.robgalbraith.com/bins/content_page.asp?cid=7-11532-11591 ) which is reporting LR 3.4 & CR 6.4 corruption of JPEG metadata in certain (rare) situations.....
    "To be susceptible to this problem, the JPEG must contain "an unusually large block of private camera data," says Tom Hogarty, product manager for Lightroom. So far, Adobe has found only one camera that creates a JPEG that can trigger the bug: the HP Photosmart R607.
    Hogarty nevertheless suggests that anyone using the affected software to update the metadata in JPEGs to stop doing so until a fix is available. He indicates that will come in Lightroom 3.4.1 and Camera Raw 6.4.1, which Adobe intends to make available by May 27.
    More information is a Lightroom Journal post."

    As for the preference file, will I lose anything by letting LR rebuild it?  Do I just need to jot down my main preferences or also the view options to recreate it?
    Thank you both for your help
    It affects the general preferences (not the presets, not the catalog settings), and I believe also the view options, but I'm not sure about that.
    Beat

Maybe you are looking for

  • T60 Laptop Doesn't Display On-Screen Volume After Hotkey Features Integration Update

    I have a ThinkPad T60 (2008-HC6) laptop running Windows 8 Pro (32-bit). Today (1-24-2013), I installed the "Hotkey Features Integration for Windows 8" update that was released on January 18, 2013. After reboot, the on-screen indicator that pops up wh

  • Is is possible to install a second (SSD) drive in a DV6

    Hi folks, I was wondering if it is possible to install a second SSD hard drive in my HP DV6 i7? The current hard drive is 500GB but if there was space I was wondering about the possibility of installing a small SSD that would contain the OS to speed

  • Expand the Document Window's Page Name Area

    In FW CS4, the area for the page name is too small.  It truncates my page names so that I can't tell them apart like this: Is there any way to expand that are where the page name shows up on the right of the document window toolbar? Thanks!

  • BackgroundImage bug?

    I'm having an issue with the backgroundImage style of a canvas and would appreciate any help. I have a tilelist with a selection of png images listed and when one is selected from the list it is set as the backgroundImage of a larger canvas on the pa

  • SOAP Async Sender Adapter - Many logons

    Hello, I have a SOAP asynchronis sender adapter that is being used to recieve invoices.  We are noticing that the suppliers who are invoicing us are calling the web service with all their invoices at the same time. (i.e. 2000 invoices in a matter of