Pages: 1 ... 3 4 [5] 6 7   Go Down

Author Topic: OSX Colour Management Problems  (Read 99780 times)

Serge Cashman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 200
OSX Colour Management Problems
« Reply #80 on: March 04, 2007, 10:07:55 pm »

Quote
Otherwise, no, you can't really make it work without an engineering background.

Yes, thanks, that's what I meant. I'll try that utility - but for practical purposes the no-LUT-adjustments profiles on Bruce Lindbloom's site are probably easier.
Logged

orangekay

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 65
OSX Colour Management Problems
« Reply #81 on: March 05, 2007, 03:41:43 am »

Quote
Yes, thanks, that's what I meant. I'll try that utility - but for practical purposes the no-LUT-adjustments profiles on Bruce Lindbloom's site are probably easier.
[a href=\"index.php?act=findpost&pid=104695\"][{POST_SNAPBACK}][/a]

It accomplishes the same thing in fewer clicks, but all roads lead to Rome.
Logged

Andrew Fee

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
    • http://
OSX Colour Management Problems
« Reply #82 on: March 07, 2007, 05:42:20 am »

My ColorEyes Display Pro box has just turned up along with the DTP94 sensor, and I've done a quick calibration. Now, this is 10am and it's quite bright even with my blinds down, so I'll have to wait until dark to do a proper comparison, but the profile is quite noticeably different from the one my Spyder2 created.

For a start, with my backlight set to 14/16, the DTP94 is reading 197.8cd/m2 whereas the Spyder read 133.4cd/m2. (which calibrated to around "120cd/m2" at D65) I suspect that the DTP94 is far more accurate here, as 14/16 did seem a lot brighter than I would have preferred to be using.

Gamma seems like it's probably more accurate too, as I'm seeing more shadow detail now. (still using 2.2 as my target)

In ColorSync, the gamut is slightly smaller now and colours seem more accurate - blues seem a lot more blue than they did before. (though I've not done any specific tests yet)

I think I might have to contact Colorvision to see what they say about the significant difference in brightness measured with my Spyder2 and the DTP94.
Logged

Andrew Fee

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
    • http://
OSX Colour Management Problems
« Reply #83 on: March 07, 2007, 08:08:28 am »

As an update: seeing as this is a laptop, I've been able to find a dark room for it where I could calibrate the display again, so I've now created new profiles with both the Spyder2 sensor and the DTP94. I also opened up the Spyder's LCD filter attachment to make sure it was clean beforehand. (which it was)

With the DTP94, I went with 11/16 brightness on the display, which was 129.9cd/m2, and 115cd/m2 when calibrated to 2.2 Gamma, D65 white point.

With the Spyder, this reads as 97.3cd/m2 and 85cd/m2 when calibrated to the same target.

With fresh profiles created in a dark room within 20 minutes of each other, the difference in gamut was even greater than before. Here's the two profiles for comparison:

http://sr-388.net/images/misc/colourproblems/Profiles.zip

The greyscale is noticeably better with the DTP94 but then it is supposed to be the better sensor after all. Colours seem more natural too.
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #84 on: March 07, 2007, 10:09:33 am »

Hi Andrew,

Thanks for the update.

I can see that your new profile done with the DTP94 shows a smaller gamut for your monitor in the blues, bringing it closer to the sRGB profile.

This would explain why blues look bluer, because, presumably, color-managed applications are now doing less correction to blue values in a sRGB image that fall outside your monitor's measured color gamut.

I'm sorry I haven't yet posted my detailed explanation -- will get to that ASAP.

Also, two questions:

1.) Is it generally always recommended to calibrate in utter darkness (yes, I do find more accurate profiles generated this way)?

2.) What is this LCD filter you speak of? My Pantone EyeOne Display 2 did not come with it. Is it an optical filter? I'm still trying to figure out why my EyeOne Display 2 keeps measuring my monitor has having a gamut that extends in the blues well past the gamut of sRGB... it's very suspicious.

I honestly don't have the money to invest in a DTP94, and was hoping that the Gretagmacbeth/Pantone EyeOne would be good enough from previous reviews I'd read.

Thanks in advance for any help.
Cheers,
Rishi
Logged

tale

  • Newbie
  • *
  • Offline Offline
  • Posts: 8
OSX Colour Management Problems
« Reply #85 on: March 07, 2007, 10:28:20 am »

I do now know what the problem on my system is:

Today I "calibrated" the display with the System Preferences Utility, adjusting apple logos to match the interlaced background and set the gamma value to 2.2. And look - Photoshop/Lightroom behave now *exactly* as they should! Either my colorimeter is not working correctly (but it does when running windows) or my Spyder2 does not like the MacBook display (perhaps it is an older hardware revision?) or the Spyder2 software does not like the Macbook. Either way - I will try the ColorEyes Software trial, if that does not solve my problem it's trying another piece of hardware or keep using the Apple-"calibration" tool...

sarangiman: This could be the issue why the Spyder2 fails. But it would mean that other colorimeters would fail, too, but as other Macbook (Pro) users state, it works on their systems. Probably only an older version of the Spyder2 has these problems?

Thanks for all your help!

Manuel
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #86 on: March 07, 2007, 10:46:49 am »

Andrew,

Using your DTP94 profile for my MacBook Pro display, you're right, the change in pure blues for color-managed applications is 'less offensive'... which is expected given that the gamut of blues lies *less far* outside the sRGB (i.e. this is more *reasonable* to me) gamut in your DTP94 profile in comparison to your previous Spyder (and my i1) profile.

Also, using your DTP94 profile, greyscale test patterns show less color casts than when using my i1 profile... which is generally a good sign. Using my i1 profile I tend to get some random yellows & pinks in greyscale test patterns... (though very subtle).

Hmm... I'm wondering now if it's the software or the hardware. I will try the ColorEyes demo and then post results.

I definitely like your DTP94 profile over my i1 profile, Andrew.

Rishi
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #87 on: March 08, 2007, 12:44:43 am »

Andrew,

I tried the ColorEyes software with my i1 colorimeter... gave me pretty much the same profile as the i1 software gave me.

Blues are still pretty offensive in color-managed applications.

BTW, for 'Black Point Target' in the ColorEyes software, I selected 'Min. Luminance'... but then didn't know what to do with the 'Precalibration' button... I went straight to profiling. Is this wrong? What did you set your Black Point Target as?

Thanks,
Rishi
Logged

Andrew Fee

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
    • http://
OSX Colour Management Problems
« Reply #88 on: March 08, 2007, 01:53:18 am »

Quote
Andrew,

I tried the ColorEyes software with my i1 colorimeter... gave me pretty much the same profile as the i1 software gave me.

Blues are still pretty offensive in color-managed applications.

BTW, for 'Black Point Target' in the ColorEyes software, I selected 'Min. Luminance'... but then didn't know what to do with the 'Precalibration' button... I went straight to profiling. Is this wrong? What did you set your Black Point Target as?

Thanks,
Rishi
[a href=\"index.php?act=findpost&pid=105412\"][{POST_SNAPBACK}][/a]
I'm now using maximum brightness for white and minimum for black. (previously I was using 120cd/m2 white)

Make sure you have the correct type of monitor set (LCD Brightness only for an Apple display) and when you run the precalibration adjust the backlight until Y is near your target. (preferably one notch above it, as calibration tends to lower brightness a bit from my experience)
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
OSX Colour Management Problems
« Reply #89 on: March 08, 2007, 02:20:30 am »

I don't know how severe this purple blue's issue is. I don't know if it's subtle or pronounced but I suspect it's a chromatic adaptation transform error ocurring probably caused by the overly bright luminance settings indicated here and how the colormeter conveys this to the software in arriving at the final luminance targets in the calibration/profile. The blue channel greatly influences brightness on displays and having luminance settings of over 120cd/m2 to start out with may be beyond the hardware and software to cope with.

A while back I made a chart of what chromatic adaptation transfrom errors show up as in CM previews using old versions of Apple and Supercal eyeball calibrators whose colorant descriptors chosen within the software had incorrect color temp XYZ formulas. A lot of calibration software sometime before 2004 were claiming great improvements in this area.

Old versions of EyeOneDisplay's i1Match software (pre-3.0) indicated an improvement to its chromatic transform formula in version 3.01. I saw this improvement after upgrading because with the old version I was getting the same hue/saturation errors but more pronounced than described in this thread as I was getting with the eyeball calibrators.

Here's the chart, see if it's similar to what you're describing here.
Logged

Andrew Fee

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
    • http://
OSX Colour Management Problems
« Reply #90 on: March 08, 2007, 03:19:26 am »

Quote
I don't know how severe this purple blue's issue is. I don't know if it's subtle or pronounced but I suspect it's a chromatic adaptation transform error ocurring probably caused by the overly bright luminance settings indicated here and how the colormeter conveys this to the software in arriving at the final luminance targets in the calibration/profile. The blue channel greatly influences brightness on displays and having luminance settings of over 120cd/m2 to start out with may be beyond the hardware and software to cope with.

A while back I made a chart of what chromatic adaptation transfrom errors show up as in CM previews using old versions of Apple and Supercal eyeball calibrators whose colorant descriptors chosen within the software had incorrect color temp XYZ formulas. A lot of calibration software sometime before 2004 were claiming great improvements in this area.

Old versions of EyeOneDisplay's i1Match software (pre-3.0) indicated an improvement to its chromatic transform formula in version 3.01. I saw this improvement after upgrading because with the old version I was getting the same hue/saturation errors but more pronounced than described in this thread as I was getting with the eyeball calibrators.

Here's the chart, see if it's similar to what you're describing here.
[a href=\"index.php?act=findpost&pid=105427\"][{POST_SNAPBACK}][/a]
That is pretty much exactly the issue I'm having, and I am now calibrating with the backlight set to 116cd/m2. (the next step down is 98cd/m2, and I was told that I should be calibrating within 100-120cd/m2)

The profiles that I'm getting from the DTP94 seem improved over the Spyder2, but still have the same problem.

In Firefox (on the left) your image looks as I'm assuming it is supposed to - a light blue mortar. In Safari, it takes on a purple tint, which is worse than this photo shows. (at least how it looks on my display that is)


(the banding is caused by me using the camera to downsample the image)

So now that we seem to have identified the problem, is there any solution? I see this with either Colorvision's Spyder2PRO software and the Spyder2 sensor, or ColorEyes Display Pro with the Spyder2 or the DTP94.
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #91 on: March 08, 2007, 02:41:15 pm »

I second everything Andrew says regarding this image of the mortar and pestle.

I still get this effect with blues when I calibrate at 90 cd/m2 (recommended for laptop LCDs as per the GMB EyeMatch 3.6 software).

Also looking for a solution...
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
OSX Colour Management Problems
« Reply #92 on: March 08, 2007, 04:05:45 pm »

I regretfully have to inform you that the photo on the right is spot on to the original as seen in nonCM PictureViewer and PS on my system. It's just a bit darker looking which in RGB land will increase perceived saturation just a bit. The left image is totally off with the lower left iiyama looking cyanish blue and the two side by side bottom center blue patches extremely desaturated and cyanish. If I assign Andrew's 1-ColorLCD profile to this image, it shows pretty much the same thing just darker as it is in Andrew's photo. Very accurate capture of a display, BTW.

Dang! Apple sure can make their display WP so freakishly neutral. I've been trying for years to get my CRT to display off-white like Andrew's MacBook and my just received 2004 iMac, my first LCD equipped computer. The EyeOne calibrated 6500K on my CRT is noticeably navy bluish tinted with a tinge of cyan. My iMac's white has no detectable color. So much for colormeters measuring accurate color temp.

None of the color patches or pestle and mortars should look blazingly oversaturated or glowing but the reds shouldn't look desaturated either. The mortars should look natural/real looking, typical of most digicam shots.
Logged

Andrew Fee

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
    • http://
OSX Colour Management Problems
« Reply #93 on: March 08, 2007, 06:08:29 pm »

Unfortunately I'm not sure how good my captures are if they aren't showing the problem. Perhaps my camera isn't reproducing the purple very well. (or maybe my eyes are wrong!) It could just be that whatever is causing this tint is causing it to look like the photograph I have taken is showing the issue when it's not. (or at least it isn't showing it as well as I had hoped)

If I take your sample image and load it up in Photoshop, everything other than the upper-right mortars are quite heavily purple-tinted. If I assign it my display's profile rather than sRGB they look natural - particularly the upper-left ones, but that's not how programs like Aperture, Lightroom and Safari show the image. (none of which have any options to get it looking like this) There are also quite a lot of artefacts (JPEG compression?) showing when they're tinted like this which disappear in an unmanaged browser like Firefox or with the image having my display profile assigned to it. (which I was told essentially disables colour management)

This is of course assuming that they're supposed to be blue.

I'm still pretty convinced that the problem lies with Intel machines, so it wouldn't show on your iMac if that's the case. I don't think there's a problem with the profile itself, just how the system is using it.
Logged

jackbingham

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.integrated-color.com
OSX Colour Management Problems
« Reply #94 on: March 08, 2007, 07:01:54 pm »

So much for colormeters measuring accurate color temp.

Lets be fair. This may not be the sensor at all. The software could easily be to blame for these results
Logged
Jack Bingham
Integrated Color Corp Maker

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
OSX Colour Management Problems
« Reply #95 on: March 08, 2007, 08:58:43 pm »

"This is of course assuming that they're supposed to be blue."...They're not suppose to look blue.
 
The left top and bottom and right bottom mortars should have varying degrees of violet=(lighter shade of purple) except the upper right one which should look cyanish baby blue. If the other three are richer looking violet/purple, that's not a that big of a problem and is more an issue with luminance/backlighting/gamma differences between displays. What is a problem is hue shifts which this chart was made to emulate thus you see the violet/purple and baby blue caused by inaccurate display profiles.

Comparing the chart on my AppleCaL calibrated iMac and my i1CaL CRT, the iMac is a tad richer and a bit more contrasty but the hues are spot on. The chart has an embedded Apple-sRGB profile which the file was converted to from my i1 CRT profile in PS.

Andrew said..."Heavily purple tinted"... ONLY in Photoshop? or in all CM apps mentioned. In PS Color Settings what is listed right after MonitorRGB-xxx in the RGB Working Space dropdown menu? It should be your custom made profile.

Just a note,viewing this chart in Apple's Preview app on my iMac-(OS 10.3.9) and changing display profiles in System Pref/Display requires I quit and relaunch Preview before the CM preview updates to the changed system profile. The same might be required on your end with your other CM apps. Not sure. I'm new to OS X. And just to be clear-clicking on the upper left X in the window pane doesn't quit the program. You have to keyboard enter command/Q. Of all the things Apple should copy from Window's GUI. Sheesh!

I transferred a number of display profiles from my OS 9 system to the iMac to use with the Colorsync Utility gamut analyzer. I compared your 1-ColorLCD profile to the iMac's factory profile and took a screenshot-see below. The inner small plot is the 1-ColorLCD.

I think you may either have a corrupt, very inaccurate or "misreferenced by CM app" profile. I switched to several display profiles on the iMac including my i1CRT, AppleCaL's, SuperCaL's, canned LCD's, GenericRGB and my i1 2000 Pismo Powerbook profile and none made my midgray desktop so yellow except the 1-ColorLCD. MacBook displays can't be that bad.

I get the compression artifact effect with each changed zoom in Apple's Preview app on the iMac, but it takes a few seconds for these artifacts to go away as the preview seems to update to each changed zoom. It's weird looking and nothing on my OS 9 system behaves this way. Maybe something is messing with the image caching scheme within your video system.
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #96 on: March 08, 2007, 09:58:50 pm »

Andrew,

How can it be a problem with the 'Intel machines' if, when you go back to Display Preferences and apply the original 'Color LCD' ICC profile supplied by Apple as your monitor profile, blues no longer shift to violet in color-managed (CM) applications (make sure that when you switch profiles, you actually click on the image window in Adobe Photoshop so that Photoshop uses the new monitor profile)?

I think there's a serious problem with the monitor profiles being generated by our colorimeters/software. I don't have these problems when profiling my Sony CRT; only my MacBook Pro LCD. Blues remain blue in CM applications on my Sony CRT (running off my same MacBook Pro, of course).

What's also really bothersome is the fact that pure blue (0,0,255) paint dumps in images in different color spaces look very different in CM applications while using my i1-calibrated monitor profile.

Take a look here:



The square on the left is blue (0,0,255) paint dumped on a new sRGB image in PS. The square on the right is blue (0,0,255) paint dumped on a new ProPhoto RGB image in PS. My working space is ProPhoto RGB; but that is irrelevant, because the same exact effect is seen even when my working space is sRGB, aRGB, etc. The laptop screen has been tilted downward for emphasis on the difference in colors. The square on the left is literally *purple*!

Here's another shot taken with the screen at a more reasonable angle:



Here the difference is harder to see but, basically, the square on the left should be lighter and more violet than the square on the left (this is what my eyes see when looking at my MBP screen).

As soon as I soft-proof through my monitor profile (essentially viewing the image as in a nonCM applications), both squares look the square on the right... i.e. pure blue.

Andrew, why don't you give this a shot in Photoshop? That is, creating some blue squares with different color profiles in PS, and seeing how these colors show up in CM vs. nonCM applications?

My explanation:

What's happening is simply that the monitor profile is telling Photoshop to throw in some red. That's why when you convert to the monitor profile, (0,0,255) goes to (101,0,255).

Initially I thought this had something to do with the fact that the monitor profile shows the gamut of the monitor extending, in the blues, well outside the gamut of sRGB, as can be seen here:

http://web.mac.com/rishisanyal/iWeb/Homepa...Comparisons.jpg

Why did I think this? Because, take a look at the gamut of the original 'Color LCD' monitor profile provided by Apple against the sRGB color space. Its blues fall within sRGB, and, so, when I have this 'Color LCD' profile set as my monitor profile, all blues look pure blue... in both CM and nonCM applications.

Application of my i1-generated ICC profile (as well as Andrew's Spyder & DTP94 generated profiles) is what goes and mucks everything up.

Why? Perhaps because pure blue in the sRGB color space lies 'further inside' the i1-generated monitor profile (not the case with the original Apple-provided 'Color LCD' profile), and so color-management is doing something screwy to blues that fall 'further inside' the blue gamut of the monitor profile. Not the case with the blue square generated in the ProPhoto RGB color space, which has a blue gamut well outside of the i1-generated monitor profile.

But then I generated pure blue (0,0,255) squares in every single profile space available on my computer, and the ONLY ones that looked pure blue (in CM applications) were the ones generated in:

     1.) ProPhoto RGB
     2.) ROMM-RGB
     3.) NTSC (1953)
     4.) all monitor profiles generated by Andrew & I (obviously, since this is like turning off color-management)

I am frying my brain trying to figure out what the he$$ is going on here, but, so far, not a clue. It IS pretty obvious to me, though, that the monitor profile is at fault. Especially because I get the SAME BEHAVIOR as explained above when I switch over to my Windows operating system on the MacBook Pro, using the same i1-generated monitor profiles.
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #97 on: March 08, 2007, 11:04:52 pm »

By the way, I also calibrated an Apple Cinema Display (ACD) using my i1.

Of course, what I love about all this calibration business is: even my 16-bit negative film scans with very subtle colors all look the same across my laptop LCD, ACD, and Sony CRT -- after hardware profiling. Before hardware profiling, oranges turned to yellows, reds, etc. It was just insane.

However, this issue with the blues is also insanity.

Here are some observations that should help someone (hopefully) figure this all out:

After calibrating my Sony CRT, MBP LCD, & ACD using my i1 colorimeter + ColorMatch 3.6 software, I created a pure blue (0,0,255) square in the sRGB color space in Photoshop. I then converted this blue square to each of the monitor profiles, much like what Photoshop is doing on-the-fly when each respective monitor is being used as the main display. Here is what happens to the RGB values:

Conversion to MBP LCD profile: (0,0,255) --> (101,0,255)
Conversion to ACD LCD profile: (0,0,255) --> (55,0,249)
Conversion to Sony CRT profile: (0,0,255) --> (16,0,255)

As you can see, the violet hues being introduced to blues for Andrew and I are coming from the introduction of reds as Photoshop converts to our monitor profiles on-the-fly. This violet hue is much less noticeable on the ACD, and almost completely unnoticeable on the Sony CRT. Makes sense, given the values above, right? On the MBP, a 0% red in a pure blue image is being changed to 40% red before output to monitor, resulting in the hue shift to violet. On the ACD, a 0% red in a pure blue image is being changed 21% red before output to monitor, resulting in a slight hue shift to violet. On my Sony CRT, 0% red in a pure blue image is being changed to 6% red before output to monitor, essentially allowing pure blues to remain pure blue on the monitor.

So, this has got to be a problem with the profiles our colorimeter/software packages are generating.

Any ideas, anyone?

I read something somewhere about a UV filter to be placed on the colorimeter. Anyone know anything about that?

Thanks in advance,
Rishi
Logged

sarangiman

  • Newbie
  • *
  • Offline Offline
  • Posts: 30
OSX Colour Management Problems
« Reply #98 on: March 08, 2007, 11:08:52 pm »

BTW, far as I remember, the UV filter was for the spectrophotometer version of the Eye-One, which I believe also measured reflective targets.

Don't see why a colorimeter would need a UV filter... so the problem must lie elsewhere.
Logged

Andrew Fee

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
    • http://
OSX Colour Management Problems
« Reply #99 on: March 09, 2007, 03:58:22 am »

Quote
So much for colormeters measuring accurate color temp.

Lets be fair. This may not be the sensor at all. The software could easily be to blame for these results
[a href=\"index.php?act=findpost&pid=105548\"][{POST_SNAPBACK}][/a]
If I have two colorimeters now that produce roughly the same results, I don't think it's the hardware, or the software creating the profiles at fault at all - I never did think that would be the case. (though I am surprised at the differences in luminance I'm seeing between the Spyder2 and the DTP94)

Quote
They're not suppose to look blue.
 
The left top and bottom and right bottom mortars should have varying degrees of violet=(lighter shade of purple) except the upper right one which should look cyanish baby blue. If the other three are richer looking violet/purple, that's not a that big of a problem and is more an issue with luminance/backlighting/gamma differences between displays. What is a problem is hue shifts which this chart was made to emulate thus you see the violet/purple and baby blue caused by inaccurate display profiles.[a href=\"index.php?act=findpost&pid=105565\"][{POST_SNAPBACK}][/a]
If that's the case, then things are looking fine then, as that's how they look here.



As a test, I set up a variety of objects I found arond the house with a lot of colour (particularly shades of blue) set white balance manually with my expodisc and took a photo - it looks almost perfect in Aperture, and in unmanaged applications it's a bit desaturated on some colours. So I guess colour management is doing its job properly, but for whatever reason if I actually send it an image that wasn't a photograph, such as a test pattern, things don't look right.

Most of this came about with me trying to get accurate captures of displays, but it seems now that the issue is more to do with the camera capturing the image from the display, rather than colour management being the problem as it seems I can get "real-world" colour to match up very well. (though it is strange that the unmanaged look of these particular images is closer than the managed one)

My concern now is this: how do I ensure that anyone viewing my photographs is seeing them as they are supposed to look? Is it even possible seeing as the majority of people use an unmanaged browser for the internet?
Logged
Pages: 1 ... 3 4 [5] 6 7   Go Up