Color shifting with intensity

We have written a CVI routine which converts the bayer color pattern encoded image from a camera to color. It works fairly well; after we do a white balance the RED, BLUE and GREEN LEDs on our instrument cluster we are testing look fine and consistant.
However, there are some LEDs which are yellow-orange. On the screen they look more greenish-yellow. This would not be so much of a problem, however the color seems to shift with intensity and/or time. As we monitor the LEDs for a while they start to shift from greenish-yellow to yellow. As a result, the matching of color test does not work very well as the values change with time. I'm assuming that the LED light output shifts as it heats up.
We also noticed that the brightest spots on
the green leds look yellow. My guess is that the algorithum does not handle intensity differences well and shifts the color.
Any ideas? Anyone else have this problem?

Hello Cabbage,
I'm not 100% certain what's happening here, but I'll make a few guesses for what it's worth. The color shifting over time is probably real. LEDs will change color if the temperature changes. Diode lasers will sometimes have a spec on the datasheet that gives this shift in nm/degC or %/degC. I'm not sure if this is true for LEDs too.
The problem with yellow-orange LEDs looking greenish may be due to the color gains being off. Your Red, Green, and Blue LEDs and the phosphors on monitor are basically monochromatic. The color filter components on the CCD are to a lesser degree. It seems that there might be little coupling beteen the colors when looking at R, G, B LEDs. Even if the color gains are off, red will still look red, green look green,
and blue look blue, however LEDs with a component in multiple bands might not look right.
Most color CCD cameras will provide spectral respose curves in the manual for each color filter and LEDs datasheets should give similar info on their output. It may be helpful to compare the spectra of the LEDs you're measuring with the response of your CCD to see where they fall.
Regards,
Brent R.
Applications Engineer
National Instruments

Similar Messages

  • IBooks author color shift with "Save for web devices..."

    Hi! I'm a web developer and I'm learning to use iBooks. It's a really easy tool, but there is a problem with color management wich is driving me crazy. This happen when I export images from "Save for web devices..." from Photoshop, then drag it to iBooks author.
    As I do for web development, I do this:
    1) Configure photoshop Color Settings to work with "Color Monitor"
    2) Remove any color management from my document
    3) Proof Setup > Monitor RGB
    4) Disable Proof Colors
    5) Use "Save for web devices..." with ICC profile and sRGB unchecked
    The exported JPG looks fine everywhere... Photoshop, Preview, Safari, Chrome, FF... but looks washed out when I drag it to iBooks Author. Why?
    I solved this issue converting to sRGBIEC61966-2.1 profile and checking ICC profile when  "Save for web devices...", but I'm not confortable with this workaround because I'm not sure about what I'm doing exactly, and I am not sure if this images are gonna look good on any device (I only own an iPad 3).
    Somebody know a best method to deal with this?
    Thanks!

    Hi KT. Thanks for your response. I'm already using this formats. Anyway, the problem occurs within iBooks Author interface. I mean: I can see the color shift (or washed out...) at iBooks Author at my computer screen. Of course, when I test it on iPad it also looks bad.
    It's incredibly strange because the JPGs I'm using looks fine everywhere else except iBooks Author.

  • Save to Web Color Shift with sRGB Working Space

    I'm having a problem a lot of other people have, which is the color shift when saving to web or viewing my work in a web browser.
    The solutions I've found in the past have been to simply make sure my working color space in Photoshop is sRGB and that my files are being built with that profile as well. I'm under the impression that sRGB is essentially the same color space as is used in a web browser, and therefore my work should look identical in both Photoshop and a web browser.
    So why am I still getting a color shift? Everything in a browser appears darker and with more contrast, as though the colors have been pumped up a bit.
    My monitor color has been calibrated with a Spyder 3 Pro and I'm using the profile it created for me as my OS X color profile. Thinking the shift may be related to my monitor profile, I tried switching to other profiles, but I still get a color shift in every single one.
    Any ideas? Thanks!

    sRGB based on capabilities and trends of development (circa late 1990s) for commodities CRT computers displays. sRGB per se has nothing to do with web browsers. The reason it has become associated with the web is the assumption that most web users are looking at content using a commodities CRT monitor using Microsoft Windows.
    It should be obvious that the use of a standard of some kind for display is better than none if there is to be any trend towards predictability of image display. Today, in commodities computing, that standard is sRGB: neutral color where R=B=G, and a gray ramp and primaries typical of the electron gun and phosphors used in CRT displays perfected for use in color television. It just so happens that this coding is rather will suited to get good value out of a minimum numbers of bits for color representation (e.g, 8 bits per channel 24 bit color)
    Web browsers in and of themselves don't have a color space. A few, such as Mac Safari, use the ICC color management capabilities of the system, and these display compensated color just like Photoshop does: if the image is tagged with an ICC profile, compensation will be applied in the browser window. But for the vast majority of computers and users, browser color is monitor color, and it's easier and most reliable for designers to just assume sRGB because it's the space that's the best approximation of the internet at large.
    You work usually will not look identical in Photoshop and a web browser (assuming that the browser does not or cannot do color management) because Photoshop is always soft proofing the working space onto the display. It is this quality that leads you to see parity between photoshop and the web browser when you choose you monitor profile as the working space: the soft proof produces no compensation.
    But while this fixes up your local experience it is -not- the solution because this projects your monitor characteristics into the image when you share it. For example, if your monitor has a lumpy non-linear tonal response and you correct image shadows or saturation for a certain look the image data will code your monitor response as well as your intentions for the image appearance. Others will see the combination of your correction for your monitor response mapped over their monitor response, in fact they will see an inversion of your monitor's response coded in the image, and to whatever extent your monitor deviates from the norm is the extent which at the very least others displays of the images you correct will deviate from your intentioned look.
    The answer to your problem--and this may surprise you--is to tolerate and appreciate the difference in display between photoshop's rendition of color and your browsers. Photoshop is showing your image more truly--within certain significant local limits--in terms of your working space than your web browser. By choosing an appropriate working space, and for the web it should be sRGB, and by profiling your monitor as you have, you are ensuring your images are normalized to current standards for web presentation. Others may still see crap, but it's most likely their problem and if they don't want to do anything about it that's their problem, not yours.
    There are a wide variety of reasons why a properly calibrated monitor will not match a working space which the soft proof cannot overcome. A monitor gamut larger than sRGB is not one of them. Photoshop will properly display sRGB data if the display profile properly characterizes the larger gamut display. This leads to questions about the effects of monitor quality on workflow. Maybe the most important of these is "How do I know my Spyder has properly calibrated my display!!!" But this is a topic for a whole 'nother thread.
    Where I run into trouble is when trying to make visual color comparisons during web design between my photoshop website mockup and a published site. I get confused and try to make color picker style matches by eye rather than by the numbers because I forget my web browser is not color managed. Sometimes I forget how all the parts play together and get confused and disoriented, then I make compounding mistakes. Then I start blaming my tools. Pity me.

  • Color Shift with ProPhot RGB in CS4

    When I had CS3 I would be able to open a RAW image in Photoshop with the ProPhoto RGB color space. I would then change the color space as needed.
    In CS4, I am using the 64-bit version, and there is a color shift when I open a RAW image in Photoshop with ProPhoto color space. The color is desaturated.
    There is no color sift when I open the same image using Adobe RGB.
    Why is this happening?
    Lynnea

    There is a color shift when the Photoshop color space is set as ProPhoto (Edit>Color Settings) and the RAW color space to open in Photoshop is set as ProPhoto.
    There is no color shift if RAW color space is Adobe RGB and the Photoshop is set with ProPhoto (Edit>Color Settings).
    But, when I save the image, the color space in file information as seen in Bridge is Adobe RGB.
    Could this be connected to GPU and "Move Color Matching to the GPU?"
    I have seen comments concerning GPU and color sifts, but I don't really understand them.
    It is frustrating since I do calibrate my monitor and have seen some of the images in both CS3 and now CS4 and thus can see the difference.
    LYnnea

  • MacBook Pro, dual Monitor color shift with Photoshop CS4, OS 10.5.5

    I'm observing a strange behavior: With PS CS3 i could just drag an image from
    the laptop monitor to the external Eizo and the color would adjust instantly, with
    consistent color (All monitors calibrated, no settings have changed). The same applies to Lightroom
    2.0 too. Now when i do the same move with the same image opened in PS CS4 the color does not
    adjust to match but stays the same and the result is a oversaturated image on the external monitor.
    Any thoughts? Thanks, Mark

    It should adjust to the display the same as CS3.
    Have you tried turning off OpenGL image display?
    (it could be a bug, or a driver issue)

  • Aperture Colour/ Color Shift with Lion

    Hi I have just added Lion to my least important computer and in Aperture when I double click on a thumb and it goes into full photo mode it is fine for a millisecond and then you can see the Colours shift to a dull flat image as you can see from images. Not a massive shift but a shift nonetheless.
    If you just view them with "Quick Preview" On they are fine but when the image is rendered the colours shift.
    Anyone else got this or is it just me again and if so any solution?

    Well,
    same old story...with every OS upgrade a new or revised RawCameraBundle file is provided and this, at least in my system, is causing the colour shift.
    I had this problem since the very first serious Aperture 3 ugrades, my Canon 30D images were horribly magenta-shifted, no way to change them back properly...well, maybe with an insane amount of extra work.
    So I keep the "good" RawCameraBundle file in a folder (it's version 3.0.1) and at every OS upgrade I delete the new RawCameraBundle file, put back the new one, restart and everything's fine again.
    The last upgrade for me was OS 10.6.8, it installed the RawCameraBundle 3.7.1 (RawCameraBundle file is located in Library/CoreServices folder)...I decided anyway to open Aperture...I displayed an image from Canon 30D....for a while it was good then suddenly (when reprocess starts with the new engine) skin tones were shifted to magenta...ok then, deleted RawCameraBundle, put back the old one, restart, launch Aperture 3, the image was good.
    Honestly I don't know how far I can go with this, I mean I don't know if Lion will allow that or if the RawCameraBundle file has been changed again.
    Maybe you could check the Lion version of the RawCameraBundle file, I bet that is not 3.7.1 anymore.
    Maybe is this causing your colour shift problem within Aperture 3, maybe not...You can try...if you have the old RawCameraBundle file; I think you may retrieve it easily from a Time Machine backup.
    Bye!
    Alberto

  • Color shift with second monitor using develop module

    LR 4.1 RC2, OSX 10.6.8 - when I'm trying to use the develop module with my second monitor, the image initially looks accurate. After a second or two and before any edits have been applied, the images desaturate and look very flat. Images viewed with the second monitor in Library module look great. I can't figure out what is happening and am wondering if any others are experiencing this. Thanks.

    Will be fixed in 4.1 final.
    Sent from phone.

  • Third party profiles results in color shifts

    Using LR on a calibrated iMac and printing to an Epson Pro 3800, I'm now getting easily noticeable color ***** in the printed output, when using 3rd party paper profiles (e.g. Hahnemuhle Fine Art Pearl or Pixel Genius Epson Exhibition Fiber profile), but if I specify an Epson profile, the output is a perfect screen match!? I specified the Epson Glossy Photo profile in LR and printed on their new Exhibition Fiber paper, and the print was perfect. Of interest here, is that print Preview showed the color shift with the other profiles and printed the same, which suggests Print Preview is an accurate rendition of the final output to paper (and yes ... no color control was specified in the driver).
    This was not the case prior to the last LR update, as the 3rd party profiles the job accurately!?

    "Also, I've emailed the correct people at Epson regarding the issue with the info sheet inside the paper. From what I know and the discussions I've had with the product manager for the 3800 and the EFP, the info sheet is WRONG. But, due to the holidays, I don't expect a response till mid-week next week. "
    I'd be curious to hear back what the response is. I hate ambiguities. There is one other thing that is bothering me about your theory, and that is I just came back from mounting an exhibit of my photos, printed using LR specifying the HFA3800FineArePearlPK.icc profile printing to their Fine Art Pearl paper. The results on paper matches the current LR screen images. This I did I think prior to applying 1.3.1 (a few weeks past). Those images in LR look the same to me, yet the profiles are now visibly altering the tonality. I will re-calibrate, but my confidence level is in the sewer.
    btw I'd love to see what PS does with the profile, but I can't install CS3 trial. It says CS3 is already installed, but it isn't.
    "System check: Adobe Photoshop CS3 cannot be installed because it conflicts with:
    Adobe Photoshop CS3"
    I had installed it a while back, but it got uninstalled along the way. Any tips around this. My gut feeling is CS3 won't have any problems with the profiles.

  • Color shift from Photoshop to InDesign?

    I'm using Adobe's Creative Cloud family of apps, which have been color synchronized with Adobe Bridge using the North American General Purpose 2 color setting. All of the files I'm working with are set to CMYK. My Photoshop files are set to 8 bits/channel.
    I just placed several .psd files in an InDesign document and there is a rather large color shift, with the placed image in InDesign being much darker than the Photoshop image. In addition, and even with InDesign's Display Performance set to "High Quality Display", the placed image is not as crisp as the Photoshop image.
    So, my questions are:
    Why is there a large color shift between both apps?
    Why is the placed image not as clean as the Photoshop image?
    What can I do to fix both issues?
    Thanks in advance for your help. 

    First and foremost, do you have a high-quality monitor, and is it properly calibrated and profiled (using an instrument)? Without that, judging color on-screen is like picking the best picture at the TV dealer.
    When you place an image in ID you are looking at a screen preview, not the pixels you see in Photoshop. View > Display Quality > High Quality Display will get you the best quality preview, but it's still just a preview. Turning on Overprint Preview will give you the most accurate color rendition for CMYK.

  • Solution to Gamma/Color Shift Problems with Apple Codecs - Adobe BUG

    Hi....
    In AfterEffects when choosing an output module and that module be ProRes or 10 Bit Uncompressed,
    when trying to color manage the file, there is a fine line called
    'This Codec Does Not Support Output Profile'
    You can find it int he Color Management Pane of the output module....
    Funny thing is - That Profile Embedding IS support by:#
    Apple's Compressor and THE CODEC 4444 AND UnCompressed
    Episodes Encoder
    See following Screen Shots:
    First one from Apple's Compressor...
    Look at that little beautiful Color Option... Adobe did you MISS that one, perhaps...
    You must have - because your APPS are NOT including that when encoding or should I say
    butchering movie files....
    Here adobes famous but wrong dialog!!!!
    And in this next screen shot you can easily see that adobe just strips the color profile out of the file...
    AND THAT IS WHY THERE IS A COLOR SHIFT...........
    left image = Apple's Compressor Encoding (Colors match of course) - Right Adobe's BRUTAL encoding (Colors WONT match...)
    Now, in after effects you wont notice the shift. But ALL mac and PC users using quicktime to watch YOUR movies
    are going to be looking at a movie with a different look than you thought it WOULD be... Courtesy of adobe ;-)
    Jokes aside, Adobe.... It seems as if this one is on you to fix... Apple's Compressor has NO problems. And other encoders are also preserving those values.
    Now... Let us assume that I encode in compressor... Looks are matching 100% that of the original source. All it takes to mess that up is ONE
    roundtrip to after effects, premiere pro or Media Encoder. That is if I wanna maintain my initial choice of transcodec - in my case 4444 or 10bit Uncomp.
    I could of course export to animation and then use compressor AGAIN to transcode to my format of choice... But in the long run I think that will become tedious.
    Perhaps you could assign a few engineers on this one and in the future spend MORE time on living up to the name PRO(The most misused word in the software business) instead of wasting time generating new Application icons for your apps....
    Thanks for reading and for a PROMPT and near-future BUG FIX

    Mylenium wrote:
    It's not really "embedding" anything, it merely sets the QT color space flags which have been there forever, but are mostly unused. Specific to ProRes it also toggles different internal optimizations in the encoding. That's also documented somewhere on the QT developer pages. I'm not sure I share your views otherwise. The point really is, that Apple developed ProRes just as much as a closed workflow within their products as Avid did with their CoDecs. Therefore the widespread use of the CoDecs is merely a byproduct of the use of the products, but not per se an intended workflow outside of those and, which is the more important part, also not explicable as a causal circumstance of the quality. If I may be so bold: What you are saying sounds like everyone would be using the Animation CoDec because it is so good, when the simple fact is that there are no decent free cross-platform alternatives and people are willing to accept limitations because of that. Don't get me wrong, ProRes is well designed and pretty robust, but the world doesn't come to an end without it and outside the Mac world it is difficult to deal with. Hell, it's even difficult to deal with between different versions of FCP occasionally...
    Mylenium
    I have been testing codecs for a long long looooooong time. My criterion(s) are:
    1) Image Quality and that through multiple encoding generations and color grading
    THEEEN
    a looooong way down the road
    2) Performance
    BTW - I find that working with files less compressed and thus HUGE is much better. Hard drives are so cheap that one can build himself an 8 Disk RAID0 and that takes care of everything. In PPRO I can cut real time in FulRes with an uncompressed codec with the processors maxing at 3%  while the H264 editing sequence will stutter like a drunken wasp and maxing the CPUs at 63%.... Go figure...
    ProRes 4444's impeccable imageQuality:performance ratio, is unbeaten.
    Especially on H264 material like that of the Canon5D. You might say if the source is 4:2:0 why bother upping it to 4444... Well in theory perhaps
    it is overkill. But rather than reading in theoretical white papers I conduct my own tests. And the end results are (REALITY) and way different from what theory states. TO anyone reading this. DONT believe ANYTHING you read about codec nor what I am saying.... DO YOUR OWN TESTING and judge for yourself. Problem is that you have to do 100's and 100's of transcodings with ONE source to various destinations with lost of grading and transcoding generation. THEN you have to keep track of your tests and THEN compare them. It takes WEEKs. But it is well worth while it. In my case... I start with a Canon5DMKII H264 movies. Transcode it to PRORES4444 then I do LOTS of compositing and re-encoding then lots of color grading etc etc etc. Then I export back to H264 - and the end result is STUNNING has NO artifacts and is adorable to look at. Not so with at least 6 other codecs claiming to be top-notch....
    Sure Mylenium, the world would NOT end with PR4444 not in it... BUT the WORLD has ALREADY decided that it WANTED ProRes4444 in it. It is growing and that every day..........
    Had the codec not been so darn excellent - it would NOT have been used. Period.

  • My display color shifts to a blue tint when opening photoshop and bridge with Yosemite

    The display on my mac book pro color shifts to a blue tint when opening Photoshop CS6.  Any suggestions on how to maintain the current color profile?  My photography work is dependent on proper display calibration.  Thank you

    First open the Universal Access preference pane and in the 'seeing' tab, make sure that the slider for 'enhance contrast' is all the way to the left. Next, open the Displays preference panel and click on the 'color' tab. The click the 'calibrate' button. This will take you through a color/brightness/contrast set up procedure that may eliminate the problem.
    There are two options in the calibration process, standard, which is the default option and has fewer adjustment screens, and expert, which is selected by checking the box for it on the first calibration screen and offers more adjustments. You can't do any harm with either since any settings you create can be discarded, or profile(s) you save can be deleted or ignored if necessary.
    This won't necessarily resolve the problem, but if not, then it would indicate that there's an issue with the hardware - possibly just the connector on the back of the system, which needs to be fastened really tightly. If you can't readily resolve this, since your system is so new, I'd then suggest you call Apple.

  • Pentax K-x RAW color shift problem with Aperture 3

    I have seen a few posts similar to what my problem is but not exactly.I have two Pentax K-200D's and a new Pentax K-x. I shoot RAW with all of them and the K-200D's work great. The RAW files however from the K-x are doing something strange. Sometimes (more and more though) the files are being imported into Aperture and there is a radical color shift after they finish loading. ALL of the preview thumbnails look fine in the camera and the Aperture import window. However, after importing them and processing is complete, some of the shots have a magenta color shift, but most that have this problem shift radically to a green tint. Everybody looks like they are in "The Matrix."
    I do not have any brightness correction on the camera turned on as other posts imply might be the problem. I have also remove the Aperture RAW 3.2 and tried 3.0 and 3.1 with no solution. 3.0 doesn't recognize the files at all and 3.1 does the same thing as 3.2. Here are two examples of UNEDITED RAW files:
    Frame that looks fine in RAW
    http://gallery.me.com/mattcline#100470/IMGP5982&bgcolor=black
    The next frame taken with the same settings on the camera:
    http://gallery.me.com/mattcline#100470/IMGP5983&bgcolor=black
    I do have one possible pattern. I had the camera set on rapid fire capture. It appears to me that the first frame I take when the shutter button is depressed looks and translates fine. All the subsequent frames take in rapid fire under the same shutter release have the green shift when taking shots in relative shade. Things taken in bright sunlight have the magenta shift.
    Any ideas on what is going on or how to fix this?
    Thanks.

    kb8wfh wrote:
    I just downloaded Apple's RAW 3.3 and it still made no difference.
    Going back over the files, I found that going back and forth between the two, see no difference in the RAW processing headings. I did see however that in the meta data that the two pictures I have in the example of those listed above have one difference. There is a setting on the Exposure Bias for the one that was fine was -0.3 ev and the one that has the green contrast blast has an Exposure Bias of 0.3 ev. So the one that looks right is slightly negative. Doesn't seem like that should be a big difference, but apparently it is.
    I did take some pictures with the RAW+JPG setting and I had the same problem with the RAW files. Some were ok, some were green, others were magenta shifted. However, all the JPG files looked just fine.
    Switching from Aperture isn;t really an option for me. I have been using it for a long time and it generally works very well...perfectly with my K-200D's. It's just some bug with the RAW decoder for the K-x.
    It's a known bug and I have been in touch with Apple staff over it, including engineers. It's not just the K-x. Other brands (Canon) and models suffer from the RAW processor gone wrong.
    Right now there is no fix. At the pace Apple works at, there may not be one for quite some time. Use JPEG, use iPhoto, or move to Lightroom.
    Another suggestion is to phone customer service and complain that their "Pro" program is not working as expected. I spent hours with an engineer going over the issue including a screen share session, only to be told to use JPEG.
    This was 4 months ago.

  • Edit With - Photoshop color shift

    When using Edit With -> Photoshop with my 12 and 14-bit NEF files, the resulting file (after editing in PS) has a color shift. Same happens when I export a Version as 16-bit TIFF or PSD and then re-import it into Aperture.
    I have tried it with 8-bit JPGs from camera, and the shift does not occur. Has anyone come across this problem and have a remedy? Your help would be greatly appreciated guys, thanks!

    Ah, thanks for the tip. I've changed my Photoshop color settings to warn me when opening files with embedded color profiles. I simply choose to work in the embedded profile and upon saving and returning to Aperture, no longer notice any color shift. Thanks again!

  • Color shift when printing with LR 3.6

    LR 3.6, win 7, screen Nec PA 241 W, Epson R 3000. When printing with colors managed by LR and paper/R 3000 profiles, I get prints with a strong color shift (printer pilot "mode" : desactivated) . How can I correct it ?

    You answered your own question: CS5.  There have been a slew of reports of workflows spun upside down when upgrading to CS5.  It may require updating your system software and/or Adobe applications, along with any printer driver updates that may be lurking out there somewhere.  The twist to your situation is ( and is very common ) that everything was working before and now it does not.  Or there is a certain print disorder of some sort.  Very troubling.  I wish I had a smoking gun answer for you, but I don't.  It sounds like you have a very viable workflow in place that worked before and damn well should perform for you now.  I'd stick with what works and dump what doesn't ( if you can ).

  • Color shift when filling with pattern

    Hello!
    I use the 'define pattern' and 'fill with pattern' feature in photoshop all the time to prepare graphics for print / production in my business. However, I have noticed a small but problematic color shift going from the main image (defined pattern) to the page (filled with pattern). You can see it in the image included here. The main image is vibrant, and lighter and the page behind it is dull and darker. It will print this way too. The main image prints great, and gets approved. But the page layout doesn't match. Both documents are set to sRGB color space when I check the Edit -> Color Settings option. Any ideas how to correct this?  Thanks!!

    Hello, in your screenshot, the first tab in the background has no profile that's what (#) means in the title of the tab, untitled one and two have the same profile as the working space (which tab is visible?), while the one in the foreground as a different profile than the working space that's what (*) means in the title bar.
    So, that tells me they do not have the same profile.
    Also, is there noise or very small textures in the circles? You might want to look at them all with the same magnification.
    Finally, a smart object can be of a different color space than the document that holds it.

Maybe you are looking for