Constructing a linear color gradient from a string

I am using the following string value for a linear color gardient.
String value = "from 0% 0% to 100% 0%, white, yellow, black 60%, red, green";
LinearGradient lg2 = LinearGradient.valueOf(value);I am expecting it to use the following list of color-stops:
white 0%, yellow 30%, black 60%, red 80%, green 100%However, I am getting the following at runtime:
white 0%, yellow 25%, black 50%, red 75%, green 100%It is not placing the yellow color in the middle of 0% and 60%, and at the same time, it is modifying the "black 60%" to "black 50%".
It seems to be a bug. I tried to debug the com.sun.javafx.scene.paint.GradientUtils class. It seems that the second last for loop in "public Stop[] parseStops(boolean proportional, double length)" method is the culprit.
Can someone guide me in the right directtion as whther to take it as a bug or it should be interpreted differently?
Thanks
Kishori
Edited by: user6588303 on Oct 15, 2012 2:56 PM

Thanks for your response. Yes. I have tried specifying positions for all color-stops and the runtime will use them correctly. The problem occurs only when you leave the positions for some and not others.

Similar Messages

Getting Color object from String

hi
i need to get a Color object from a string provided at runtime.ie for the string "RED" i need to get a object of Color.RED.I have already tried to use decode() from Color class but it throws numberformat exception,can someone guide me.

Color.decode(String) will decode octal and hexidecimal representations of colors e.g.
Color poo = Color.decode("0x215DB8"); // hex description of color
works.
If you want to use RED, BLUE ... read them at run into a String or whatever and then test this String to get what color the user wanted e.g.
read from text file at run time that button colour should be ORANGE.
In your code, put
String colourStr; // read what colour the user specified into this var
Color their_color = null;
if(colour.equals("ORANGE"))
their_color = new Color(Color.orange);
else if(colour.equals("BLACK"))
// you can guess what goes here.. and so on.
It often nice to make them specify the hex value of the colour because there are more of these than the predefined colours (shock horror).
Anyway,
seeya, Edd.

Color gradient in a non linear geomerty

Hi all,
I try to got an color gradient from the outer side to the inner side of this geometry
the problem : The bandwidth of the gradient should be more or less the same in all areas of the green geometry(means the whole spectrum of the colors)
And the gradient has to go perpendicular from the outer to the inner side.
I ask a similar question already some weeks ago but this looks at least for me different
Could anyone help me with this problem ?
Thx in advance
Mark

Mylenium,
thank you for the answer. But unforunately I'm a newby. Could you go into more details please ?
Mark
PS:
The problem I always got with the gradient tool is the following
The gradient starts at a certain point and uses a hard coded distance (e.g. 50 pixels from this point to this direction like a vector)
This makes it impossible to change the geometry with the full gradient
For example :Filling this rectangle with a gradient and I’m defining a vector mask to it
Afterwards I changed the vector mask to this
The gradient is cut instead of changing, compressing it.
Is there no way to define a starting point (edge, path), a direction and a target edge, path ?
Because then, after changing one of the edges, the gradient has to be recalculated.

How to create a node-typed XmlValue from a String with Java?

Hi all,
It's already been a while since I'm trying to get how to construct a node-typed XmlValue from a String value but I keep getting different exceptions no matter what method I try.
I started with the first way that came to my mind:
XmlValue value = new XmlValue("<element/>"); //Even though the String represents a valid xml node, calling value.isString() returns true.
then I thought ok, so I should maybe use the typed constructor:
XmlValue value = new XmlValue(XmlValue.NODE, "<element/>"); //_It does not work either, calling value.isNull() returns true.
I also tried with the rest of types (XmlValue.DOCUMENT, XmlValue.DOCUMENT_NODE, XmlValue.ELEMENT_NODE.....) and it did not work either with any of these.
Fiinally I saw that you can construct a new XmlValue from a XmlDocument but I cannot get a XmlDocument unless I phisically create one, which obviously wont be a very good idea.
So I'm definitely stuck with these, I would really appreciate if someone could tell me how to do this...
Thanks in advance,
Pablo Pareja
PS: I'm using Berkeley dbxml 2.5.13

Hi Pablo,
        XmlDocument doc = manager.createDocument();
        doc.setContent("<bla/>");
        XmlValue val = new XmlValue(doc);
        System.out.println(val.getType() == XmlValue.NODE);
        System.out.println(val.asString());Try this. It doesn't create physically (that is in the container) a document. The output of this piece of program:
true
<bla/>Hope it helps
Vyacheslav

Converting a gradient from a spot color to a process build, hides vector elements

I have been using Illustrator for many years and consider myself very familiar with it, but recently ran into a strange issue I have never seen before.
The background element is a radial gradient with points of 0% to 50% of a spot color. The file came to me as one single layer, with the gradient in question above a vector rectangle. The rectangle shows up fine, however, until I convert the gradient from a spot to a process build at which point the box is no longer visible. The initial issue I found was that the box is beneath the gradient, but my question is why would it be visible as long as the gradient uses a spot and then disappear behind it when converted to a process build?

Version is CS5.
Part of what made the issue so confusing is that there are no effects applied. Applying the overprint feature or multiply effect don't effect the outcome either. It's only something that can be fixed by moving the gradient to a layer behind the element which is disappearing. My question is less about how to fix it (I've done that and moved on) and more about why this would happen - it seems to be a strange anomoly.
In response to the replies involving the method of color conversion:
I have used all of the methods mentioned and the end result is the same. It is starting to look to me like this is a fluke that happened with this particular file and that it is not something that can be duplicated intentionally.

Weird and unwanted color gradient after Photomerge

Hi all,
I got an annoying problem with a weird color gradient that appears where it shouldn't. My context is, that I'm producing underwater photomosaics. I use Photomerge for merging, which works great for the often uniform seabed sediment images.
[merge1.jpg]
The example is 65 images, btw. After placing the images manually within Photomerge, all looks great and ready for a perfect result after rendering (see above). However, when doing so, the final merge from within Photomerge (by clicking the OK button) creates perfect smooth blending, but a weird color gradient crossing the whole mosaic. this might originate in the fact, that the individual images also are not perfect, due to the uneven lighting situation underwater. In case all images are a bit reddish at the top and yellowish at the bottom, the final gradient over the full mosaic shows that as well. So to me it looks as if Photoshop just tries to be too smart and thinks, reapplying that pattern to the whole would be a great thing to do, which it isn't of course.
[merge2.jpg]
I discussed this "bug" with (indeed very helpfull) Adobe staff and they recommended, to create the merge with the "Blend Images Together" option in the "Load" dialog unchecked. This creates a mult-layer document after rendering, with the images being placed properly but not blended.
[merge3.jpg]
From here on, Adobe staff recommended, I shall use the "Auto Blend Layers" function (File menu), but the result is identical as image 2. Ok, then, they recommended, to uncheck "Seemless Tones and Colors" in the "Auto Blend Layers" dialog. Indeed, this gets rid of the gradient, ...
[merge4.jpg]
... but shows very hard edges of the images, which get very prominent after optimizing levels:
[merge5.jpg]
Was that understandable? Would anyone have experienced (and solved) the same issue? Or would anyone have an idea, how this could be handled in a way, that my resulting mosaic looks as perfect with respect to the original colors as in the initial merge window before rendering it (image 1)? Maybe it's a specific Color Setting (Edit menu)? I tried a lot so far but nothing helped and I might not have tried the right thing.
The problem appears, btw, in CS3 to CS6!
Any help appreciated, more information on request! Many thanks,
g

H there,
@TLL: you're right with what you said about the subjective element. Getting spoiled is easy, but staying ambitious is good as well. I do these mosaics since 2003, when Photomerge was part of Elements only. Then, in the first years with CS and CS2, I never had this bug with the color gradient being applied to the final merge. This started with, I guess, CS3 and I think it's due to enhancements Adobe made to the programm (making it more "intelligent"). Unfortunately, sometimes enhancements have dark sides and cause unexpected issues elsewhere. And since we are a minority (making large merges), these do not get fixed, since the majority, stitching what, 20 image panoramas of their home and garden do not suffer. Their images have strong contours and contrasts and the issue gets effective only on uniform images like underwater sediment or, in your case, landscapes shot from above. Going back to CS or CS2 is no option, since only the newer versions give me the processing power I need, also is the blending much smoother.
I played around with gradient masks but I haven't got the right touch yet. Even if I get rid of the color differences, there will always be a brightness gradient, since borders of the images are always darker. And then, the weird gradient is a bright / dark one, not a red / yellow one. Not much better. Also, each batch job has to work for thousands of images, so has to be a compromise. I just cannot spend more than a day for a 1000 image mosaic, or a week for 6000 images, time is money on a ship (and elsewhere).
I just think the solution would be, splitting the "Seemless tones and colors" function up into "Seemless Blending" (doing soft blending but not changing the colors) and "Seemless Color Transitions" (just dealing with the colors). And making both optional by check boxes...
So far,
g

Premiere Pro CC-MediaEncoder : Setting sequence composite linear color ?

Hello, hello,
Someone would tell me what is this option?
Setting sequence> composite linear color (or GPU acceleration rendering quality max neccessary)?
I wonder if the fact that my media encoder takes me 6 H encoding to make a video of 17mn which has no effect does not come from there?
Since I do not know what is this option.
Ahmotiond
Bonjour, bonjour,
Quelqu'un serait-il me dire à quoi sert cette option ?
Reglage de sequence> composite en couleur linéaire (accéleration GPU ou qualité de rendu max, néccessaire) ??
Je me demande si le fait que mon média encodeur me tape des 6 H d'encodage pour encoder une vidéo de 17mn qui ne comporte pas d'effet ne viendrait pas de là ?
Vu que je ne sais pas à quoi sert cette option.
Ahmotiond

You are not missing anything. This is how it works now in CC.

Flex/AS3 Best way to construct a derived class instance from an existing base class instance?

What is the best way to handle the instantiation of a derived class from an existing base class.
I have a base class which is being created via remote_object [RemoteClass alias] from the server.   I have other specialized classes that are derived from this baseclass, but serialization with the server always happens with the base class.     The base class has meta data that defines what the derived class is, for example
[RemoteClass (alias="com.myco...')]
public Class Base
     public var derivedType:String;
     public function Base()
public Class Derived extends Base
     public "some other data"
     public function Derived()
In my Cairgorm command which retrieves this object from ther server I want to do this:
public function result (event: Object):void
    var baseInstance:Base = event.result;
     if (baseInstance.derivedType = "derived")
          var derivedInstance:Derived = new Derived( baseInstance );
What is the most efficient way of doing this?   It appears to me that doing a deep-copy/clone and instantiation of the derived class is pretty inefficient as far as memory allocation and data movement via the copy.

Thanks for the assistance. Let me try to clarify.
MY UI requires a number of composite classes.    The individual components of the composite classes are being transfered to/from the server at different times depending upone which component has changed state.    The construction of the composite classes from the base class happens in my clients business logic.
Composition happens in a derived class; but server syncronization happens using the base class.    When I recieve the object from Blazeds through the remote object event, it is in the form of the base class. I then need to instantiate the derived class and copy the elements of the base class into it (for later composite construction).   And likewise when sending the base class back to the server, I need to upcast the derived class to its base class.   But in this case just a mere upcast does not work. I actually need to create a new base class and copy the attrbutes into it. I believe this is limitation of how remoting works on Flex/AS3.
My question is, what is the best way to turn my base class into it's derived class so further composite construction can take place.   The way I am currently doing it is to create a load method on the base class, that takes the base class as on argument. The load function, copies all of the instance attribute references from the base class to the target class.
public Class Base
     public function Base()
     public function load(fromClass:Base)
    { // copy the references for all of the instance attributes from the fromClass to this class }
Then, after I recieve the base class from the server.   I create a new derived class and pass the base class into the load function like this:
            for (var i:int=0; i < event.result.length; i++) {
                var derived:Derived = new Derived();
                derived.load(event.result[i]);
The drawbacks of this approach is that it now requires 2 extra instance creations per object serialization.   One on recieving the object from the server and one sending it to the server.    I assume copying references are pretty efficient. But, there is probably some GC issues.     The worst of it is in code maintenance.   The load function now has to be manually maintained and kept in sync with the server class.
It would be interesting to hear how others have solved this problem.      The server side is an existing application with around 2M LOC, so changing the code on the server is a non-starter.
Thanks for your help.

How to remove HTML tags from a String ?

Hello,
How can I remove all HTML Tags from a String ?
Would you please to give me a simple example ?
Best regards,
Eric

Here's some code I cooked up. I have created an object that processes code so that it can be incorporated directly into a project. There is some redundancy so that the it can be used in more than one way. Depending on your situation you might have to make the condition statement a little more sophisticated to catch stray ">" tags.
I have also included a Tester application.
//This removes Html tags from a String either by submitting the String during construction and then
// calling getProcessedString() or
// by simply calling " stringwithoutTags=removeHtmlTags(stringWithTagsSubmission); "
//Note: This code assumes that all"<" tags are accompanied by a ">" tag in the proper order.
public class HtmlTagRemover
     private String stringSubmission,processedString,stringBeingProcessed;
     private int indexOfTagStart,indexOfTagEnd;
     public HtmlTagRemover()
     public HtmlTagRemover(String s)
          removeHtmlTags(s);
     public String removeHtmlTags(String s)
          stringSubmission=s;
          stringBeingProcessed=stringSubmission;
          removeNextTag();
          return processedString;
     private void removeNextTag()
          checkForNextTag();
          while((!(indexOfTagStart==-1||indexOfTagEnd==-1))&<indexOfTagEnd)
               removeTag();
               checkForNextTag();
          processedString=stringBeingProcessed;
     private void checkForNextTag()
          indexOfTagStart=stringBeingProcessed.indexOf("<");
          indexOfTagEnd=stringBeingProcessed.indexOf(">");
     private void removeTag()
          StringBuffer sb=new StringBuffer("");
          sb.append(stringBeingProcessed);
          sb.delete(indexOfTagStart,indexOfTagEnd+1);
          stringBeingProcessed=sb.toString();
     public String getProcessedString()
          return processedString;
     public String getLastStringSubmission()
          return stringSubmission;
public class HtmlRemovalTester
     static void main(String[] args)
          String output;
          HtmlTagRemover h=new HtmlTagRemover();
          output="The processed String: "+h.removeHtmlTags("<Html tag>This is a test<another Html tag> string<yet another Html tag>.");
          output=output+"\n"+" The original string:"+h.getLastStringSubmission();
          System.out.print(output);

Color gradients not displaying correctly

I'm having an issue with my color gradients in type and objects not displaying correctly in the content viewer. For the last app I created, I made sure the colors were RGB and that seemed to fix it. With the current app I'm working on, that dosen't seem to fix the issue. Actually, when I test in the viewer on my iPad, it shows the correct gradient in the split second that it's loading the update but then once it's loaded, it goes to a gray white gradient (the correct gradient is from a yellow to a gray; attached a pict for referrence).
Anyone know why this is happening? Thanks in advance!
UPDATE: It seems to view fine if I preview the page via the foilo overlay but when I preview the folio, that's when it's no viewing correctly.

The Color Picker is unreliable for choosing colors based on what you see because it is not color managed while Color panel and Swatches are.
In your image, the color that you see in the Color Picker is how your monitor displays R=0 G=255 B=0. The color that you see in the Color Panel represents the same values in the file but the color management sends different values to the video card and thus to the monitor to display a color with appearance in the color space selected for the document. In your image this is a CMYK color space which you can check from the Show popup menu at the bottom left of the document window. To check the different values the video card is using, make a screen capture, save the image, open it in Illustrator and measure the color with the eyedropper. The color translation from the values in the file to the video card allows simulation of color spaces other than your monitor's. This simulation is accurate if you have a monitor profile installed in your system that correctly describes how your monitor displays colors. Such profile is created by using a color measuring device and profiling software. If your monitor profile (color space) installed in the system is the same as the color space of the document then there will be no translation (simulation) which in effects shuts down the color management - then the colors in the Color Picker will be the same as in the document. Many people do not bother to get an accurate profile of their monitor and use a generic sRGB profile for that. In such case working in RGB color mode using the sRGB color space (default in the Color Settings) will give the same colors in the Color Picker and the document. This usually is a sign of using a monitor that incorrectly displays colors.

Copy and Paste gradients from Indesign to Illustrator???

Hello,
for some reason when I copy and paste an object with a gradient from indesign to Illustrator, the object becomes colorless....What happened to the gradient? I can copy and paste a solid colored object from Indesign to illustrator (and for some reason then, when I copy and paste an object with a gradient from illustrator to Indesign, it acts like a "placed" image....what's that about????)
Can anyone tell me the best way to transfer a gradient swatch from Indesign to Illustrator, without having to regreate it? And why when I paste objects from Illustrator to Indesign, Indesig treats it as a "placed" image (meaning that once you double click on it, the hand pops up allowing you to move within the bounding box, like with an image)
Thanks...
Thank you very much

Hi Jacob,
Thank you for a quick answer. Your suggestion is correct, but saving image to a file is an additional step.
Since I do this task "thousands a day", this is unacceptable
Any other suggestion?
Regards,
Jarda

A Mercury linear color delimma... bad transitions vs. bad animation

Am I missing something here? I'm writing this as sort of a PSA... if you aren't aware of all sides of this issue, I hope you read the second half because you may not realize it's affecting you, too.
When I enable CUDA or Maximum Render Quality, I basically can't use any one-sided transitions. They look wrong and "pop" at the ends.
Apparently, Adobe forgot to rewrite any of their transitions for a linear color space, so Mercury+CUDA basically "breaks" all the transitions that ship with CS5, causing them to render differently than they do in their originally intended gamma of 1.8. I've heard they've added a linear-compatible transition into CS5.5 called "Film dissolve", but there aren't any in CS5. This seems like a pretty huge oversight... I'm surprised they still haven't fixed it in a patch to the original CS5.
So why not just turn off Maximum Render Quality and CUDA during output?
Well, there's another big problem.
If you do any kind of stills animation a la "the Ken Burns Effect" in Premiere, you'll quickly find Premiere leaves a "pixel shimmer" or "ripple" effect across the animation because it is using substandard methods to antialias motion. Looks like a cheap NLE from 2002.
So, you're left to choose... do you want bad one-sided dissolves or bad animation? Personally, I believe a good NLE should not leave identifiable footprints in the edit. If you happen to have have *both* in a Premiere CS5 timeline and it goes out on TV, I'll be able to "out" your NLE as Premiere because no matter which output setting you choose, one or the other problem will show up.
SOLUTION 1: Use AE. The simplest solution is probably to avoid Maximum Render quality and avoid keyframing any Motion parameters. Do it all in After Effects.
SOLUTION 2: Use CS4. It may be possible to edit your sequence with all the benefits of MPE and CUDA, and then send your timeline back to Premiere CS4 and have Premiere/Encoder CS4 render it at max quality. I haven't tested this yet, but I'm intrigued by this possibility. I'll try it soon and report back.
Hopefully I am missing something here. I'm an Avid/FCP/AE guy who is still relatively new to Premiere.
P.S. A second issue is lower thirds and alpha graphics. I bet you made all of yours in After Effects with a 1.8 gamma, didn't you? As far as I know, Premiere does not ship with any video effects that easily control the gamma curve of the alpha channel, so there's no quick fix for all of your currently rendered transparent graphics.

I know if I use the MPE in GPU mode transitions between two PIPs will get a funny black border aroud the PIP. CUDA can do some funny things when transitions and filters that do not have CUDA support are used with GPU enabled. I can only hope Adobe is getting ready to switch to Open CL or Open GL because there are a few glitches with Nvidia's CUDA technology. For Quality I almost think CS 4 was better but you don't get the same amount of realtime. I hope PP CS 6.0 is true broadcast quality instead of just realtime previews. Edius can do it with out the use of GPU acceleration. I think CUDA is a step in the right direction but I also think it needs a bit more time to mature.

How can I export formatted text from a string indicator?

Does someone know how I can export formatted text (i.e., parts of the text have different formatting, such as color, fontsize, etc.) from a string indicator? Using copy/paste does not work, as it only exports unformatted plain text.

Hello Sparti,
Thank you for your suggestions, they are all very useful, and I plan to use the HTML feature under Report Generation to export the formatted text from Labview. However, I am still not sure how I can extract the formatted text from a *string indicator* and transfer it to one of those VIs, so that it can be exported to other applications. Let me give some more specific info on what I am trying to achieve: I am monitoring the communication between two pieces of equipment. A string indicator shows all the data flow, with different colors for data coming from different instruments. I managed to do that by using a property node and playing with the selection and font color properties. Now, if you just wire the output of the string indicator, the formatting is gone and all you get is just plain black text (for instance, try to programmatically transfer the formatted text from one string indicator to a different string indicator and you will see that the formatting is not preserved). Even if you try the "brute force" method of manually selecting and copying the text in the indicator and pasting it to Word, LV does not export the formatting. But, if you paste *within* LV (for example, paste it to a string constant in your diagram), then it works. To extract the formatted string from the indicator, I also tried to use a property node, but without success. I am trying to avoid duplicating part of my code to generate the same color-coding scheme on a report. It would be way easier to be able to transfer the formatted text from the string indicator. This is particularly annoying, because the information is there, stored in the data structure associated with the string indicator. But how can I put my hands on it? Any ideas?

Un-smooth color gradient

I'm a new mac user, and had used the macbook pro 2.4G with LED backlight for one more month. I find the color gradient is not smooth in second primary color to primary color. Please check the PNG files as bellow links.
http://img.photobucket.com/albums/v102/teazeb/colorgradient001.png
http://img.photobucket.com/albums/v102/teazeb/colorgradient002.png
http://img.photobucket.com/albums/v102/teazeb/colorgradient003.png
http://img.photobucket.com/albums/v102/teazeb/colorgradient004.png
http://img.photobucket.com/albums/v102/teazeb/colorgradient005.png
In those patterns you will find many vertical lines. Those will make you see some photos with contour or noise by Macbook Pro. I have double check many other PC notebooks. They don't have this issue. It is hard to believe the high-end notebook has such kind of color performance. The color performance is too poor to use the "Pro" title. I feel so disappointed to spend a huge amount of money on the Macbook pro( Poor?).

Mylenium,
thank you for the answer. But unforunately I'm a newby. Could you go into more details please ?
Mark
PS:
The problem I always got with the gradient tool is the following
The gradient starts at a certain point and uses a hard coded distance (e.g. 50 pixels from this point to this direction like a vector)
This makes it impossible to change the geometry with the full gradient
For example :Filling this rectangle with a gradient and I’m defining a vector mask to it
Afterwards I changed the vector mask to this
The gradient is cut instead of changing, compressing it.
Is there no way to define a starting point (edge, path), a direction and a target edge, path ?
Because then, after changing one of the edges, the gradient has to be recalculated.

SampleImage() in 4-color gradient issue

hi there
I am trying a technique for logo/element removal that I saw in a post from a couple years back, where a tracked, solid "patch" is placed over the unwanted element, and a 4 color gradient effect is applied to the patch, which then uses the sampleImage() method to grab local colors around the element.
I've gotten it working fine with the below code, except for one problem: when I parent my null (acting as a "color picker" position) to the original tracking info, the color no longer samples properly. Unparenting the null makes the sampling work fine, but then, I have to hand key the null, which is a pain and doesn't produce great results.
myPoint = thisComp.layer("Null 1").transform.position;
target = thisComp.layer("myAwesomeFootage.avi")
target.sampleImage(target.fromWorld(myPoint), radius = [1, 1], postEffect = true, t = time)
Any tips on getting around this parenting problem?
thanks!
ML

You need to add an extra toWorld() layer space transform to convert the layer coordinates to comp/ world coordinates.
Mylenium

Constructing a linear color gradient from a string

Similar Messages

Maybe you are looking for