Ad
Ad
Ad
Pages: « 1 2 [3] 4 »   Bottom of Page
Print
Author Topic: CS3 ability to open jpegs in Raw  (Read 21759 times)
jbrembat
Full Member
***
Offline Offline

Posts: 177


« Reply #40 on: May 24, 2007, 07:41:11 PM »
ReplyReply

Quote
The highlight clipping in the highlights around the lights is not significant, since they contain no important image elements and may be allowed to blow out. The problem is that the blue channel is completely lacking in data in the midtones and shadows. It might possible to reconstruct the blue channel via calculations from the red and green channel, but Photoshop can not make something from nothing.
The photo was not supposed to be white balanced. The whole point of the exercise was to emulate what would happen with an in camera JPEG when exposing a tungsten illuminated scene with daylight white balance. The test with a linear TRC and 16 bit conversion was to track down the source of the data loss in the JPEG.

The illumination is not far from the Planckian locus as shown in this screen capture in ACR. I preset the white balance with a reading from a white card.

[attachment=2553:attachment]

You completed the exercise by showing that a decent white balance was possible from the shot with the wrong color balance.

Bill
[{POST_SNAPBACK}][/a]

Ok we have a completely different vision of what is color and what an histogram represents.
Let me try to show my point of view:
This is a better balanced result of Rodney photo.
[a href=\"http://img508.imageshack.us/img508/3801/balanjy0.jpg]http://img508.imageshack.us/img508/3801/balanjy0.jpg[/url]

You can think the result is good,"since they contain no important image elements and may be allowed to blow out". NOT FOR ME.

The resulting histograms:
http://img455.imageshack.us/img455/8541/balanhzq9.jpg

Quite different from the original!

Quote
I preset the white balance with a reading from a white card.

so
Quote
The illumination is not far from the Planckian locus

That is not true.

Jacopo
Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #41 on: May 24, 2007, 07:59:50 PM »
ReplyReply

Quote
That is not true.
[a href=\"index.php?act=findpost&pid=119480\"][{POST_SNAPBACK}][/a]

Really? I was under the impression that the color temperature slider adjusted temperature along the Planckian locus and tint at right angles to the locus. What is your line of reasoning?  
Logged
jbrembat
Full Member
***
Offline Offline

Posts: 177


« Reply #42 on: May 25, 2007, 03:26:43 AM »
ReplyReply

Quote
Really? I was under the impression that the color temperature slider adjusted temperature along the Planckian locus and tint at right angles to the locus. What is your line of reasoning?
I don't know if this is the behavior, I believe you.

I prefer to speak about color temperature and correlated color temperature.
If you look at the following PhotoPedia article, you can see the planckian locus.
Your photo has a yellow color cast but a magenta color cast too. So you can see that magentas are far away from planckian locus.

But you said that a WB produced using a gray card is performed along the blackbody curve. I said that this kind of WB is able to perform well also if the illuminant  is at a "correlated color temperature".

Jacopo
Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #43 on: May 25, 2007, 06:21:26 AM »
ReplyReply

Quote
I don't know if this is the behavior, I believe you.

I prefer to speak about color temperature and correlated color temperature.
If you look at the following PhotoPedia article, you can see the planckian locus.
Your photo has a yellow color cast but a magenta color cast too. So you can see that magentas are far away from planckian locus.

But you said that a WB produced using a gray card is performed along the blackbody curve. I said that this kind of WB is able to perform well also if the illuminant  is at a "correlated color temperature".

Jacopo
[{POST_SNAPBACK}][/a]


You didn't give a reference to the PhotoPedia article, but this [a href=\"http://en.wikipedia.org/wiki/Color_temperature]Wikipedia[/url] article discusses  correlated color temperature. WB using a gray card does not necessarily produce values along the Planckian locus, but since my picture was taken under incandescent illumination, I would expect the WB to lie on this locus. In the ACR screen shot the WB readings are 2000 for temperature (this adjustment works in the blue-yellow axis, presumably moving along the Planckian locus), whereas the tint is -2 (this is in the green-magenta axis,  which is at right angles to the Planckian locus and parallel to the lines of correlated color temperature as shown in the Wikipedia article). I would consider a tint of -2 very close to the Planckian locus.

Perhaps you are referring to the picture simulating the daylight exposure of a tungsten illuminated scene.
Logged
jbrembat
Full Member
***
Offline Offline

Posts: 177


« Reply #44 on: May 25, 2007, 07:00:15 AM »
ReplyReply

Sorry, I missed the reference http://www.photoresampling.com/photopedia/...lor_Temperature

Quote
but since my picture was taken under incandescent illumination, I would expect ...

Perhaps the lights were not exactly "incandescent". (tungsten).
I don't  know the scale for the  -2 value, but :

as explained in the referenced article, the CCT values ar not on perpendiculars from planckian locus in CIE_xy space

as you can see magentas are not so near to planckian locus


Jacopo
Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #45 on: May 25, 2007, 07:58:34 AM »
ReplyReply

Quote
Sorry, I missed the reference http://www.photoresampling.com/photopedia/...lor_Temperature
Perhaps the lights were not exactly "incandescent". (tungsten).
I don't  know the scale for the  -2 value, but :

as explained in the referenced article, the CCT values ar not on perpendiculars from planckian locus in CIE_xy space

as you can see magentas are not so near to planckian locus
Jacopo
[a href=\"index.php?act=findpost&pid=119529\"][{POST_SNAPBACK}][/a]

The CIE XY space is not perceptually uniform, and your article states "correlated color temperature is the color temperature of the point on the blackbody locus that is closest in appearance (chromaticity-wise) to the chromaticity of interest. Technically, this means the point on the blackbody locus nearest the point representing the chromaticity of interest when the blackbody locus is plotted not on the CIE-xy chromaticity diagram but rather on the CIE-uv space chromaticity diagram"

Therefore, for this type of plotting you should be using a perceptually uniform space such as CIE-uv or CIE LAB, not CIE-XY. If you want to find the shortest distance from a point to a line, you drop a perpendicular from the point to the line. With CIE-xy the lines of correlated color temperature are not perpendicular to the Planckian locus, but  with a perceptually uniform space, I would think that they should be. If you want to nit pick, you should have your ducks lined up in order.

A tint of -2 in the magenta direction is very close to the Planckian locus. Of course, magenta is not on the Planckian locus, which runs in the yellow-blue direction. Really, what point are you trying to make?
« Last Edit: May 25, 2007, 08:03:25 AM by bjanes » Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #46 on: May 25, 2007, 08:20:11 AM »
ReplyReply

Quote
Going back to the original point of this set of posts. We have a baked JPEG that's pretty butt ugly for a number of reasons and a DNG shot under similar conditions which seem to be more pliable in being corrected in Lightroom. Remember, the initial point was, trying to adjust a JPEG is a lot harder to do than a Raw (duh) but also, is it 'better' to attempt to fix this kind of file in a raw converter like LR or use Photoshop? This is a severely ugly image but one I suspect isn't that rare. We have two possible tools, Photoshop and LR/ACR, at least in the context of this series of posts. Its possible neither can fix this mess, the image is too far gone? Does anyone have a clear idea which tool would be better at bringing this JPEG back into a somewhat acceptable color apparence? Does the WB tool in LR really help or is it too far gone? Is converting a baked gamma encoded JPEG into a linear color space just so we can use LR's tools the right solution?

I don't expect a single image will answer these questions but they are questions I think need to be asked.
[a href=\"index.php?act=findpost&pid=119477\"][{POST_SNAPBACK}][/a]

Andrew,

I don't think that Photoshop and LR/ACR corrections are mutually exclusive. As this discussion has shown, LR/ACR WB correction can work well when the original color information of the JPEG rendering has not been clipped, distorted, or lost. The in camera JPEG rendering takes considerable liberties with the data as pointed out in the Bruce Fraser quote, and the LR/ACR correction may be inaccurate. One could correct as best as possible in LR/ACR and finish up the correction in PS.

Converting from a narrow gamma encoded space to a wide linear space and then back to the original space does entail loss of data. If you work only in Photoshop, no data are lost from conversions. By way of analogy, considerable data are lost when one converts to LAB for corrections and then back to RGB, especially when working in an 8 bit environment. According to Dan Margulis, the losses are not perceptually evident, but according to my memory, you don't necessarily agree with him.  

Bill
Logged
digitaldog
Sr. Member
****
Online Online

Posts: 8972



WWW
« Reply #47 on: May 25, 2007, 08:23:11 AM »
ReplyReply

Quote
Really? I was under the impression that the color temperature slider adjusted temperature along the Planckian locus and tint at right angles to the locus. What is your line of reasoning? 
[a href=\"index.php?act=findpost&pid=119481\"][{POST_SNAPBACK}][/a]

I doubt it does. Its using some correlated color temperature I'm pretty darn sure. We could ask a true color scientist like Karl Lang about this. I think you're correct about the overall angle in color with respect the locus but does it follow it exactly? I would guess not.

Least we forget, the blackbody is based on a theoretical object that doesn't exist. Hence the reason we should use correlated color temperature since so few (any?) natural objects conform to the behavior of this theoretical object (the blackbody).
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
digitaldog
Sr. Member
****
Online Online

Posts: 8972



WWW
« Reply #48 on: May 25, 2007, 08:55:04 AM »
ReplyReply

Quote
Andrew,

I don't think that Photoshop and LR/ACR corrections are mutually exclusive. As this discussion has shown, LR/ACR WB correction can work well when the original color information of the JPEG rendering has not been clipped, distorted, or lost. [a href=\"index.php?act=findpost&pid=119542\"][{POST_SNAPBACK}][/a]

Well does the JPEG I provide fall into that camp, that's what the camera gave me as a JPEG so I'm not sure I'm following you.

Its possible this image is as equally hopeless to repair in LR or Photoshop. And its probably impossible to attempt to match the same corrections given the difference in toolset, let alone how the processing is being applied.

If my point was, its easier to fix a raw like this than a JPEG, nearly everyone here would say Duh! So that's not a useful point to make however, to digress, there is always Dan Margulius who on his color theory list feels the opposite and said this:

Quote
It does, however, beg the question: if saving time is so important that
quality compromises need to be made, why is the raw format being used at all? With rare image-specific exceptions, essentially anybody who is not a beginner will get better final results by shooting JPEG and correcting in Photoshop than an expert can who shoots raw but is not allowed to do any manipulation outside of the acquisition module. And in less time, too. The idea of a raw module is to *empower* the image-manipulation program, not replace it.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #49 on: May 25, 2007, 10:34:08 AM »
ReplyReply

Quote
I doubt it does. Its using some correlated color temperature I'm pretty darn sure. We could ask a true color scientist like Karl Lang about this. I think you're correct about the overall angle in color with respect the locus but does it follow it exactly? I would guess not.

Least we forget, the blackbody is based on a theoretical object that doesn't exist. Hence the reason we should use correlated color temperature since so few (any?) natural objects conform to the behavior of this theoretical object (the blackbody).
[{POST_SNAPBACK}][/a]

I don't know if the ACR temperature slider relates to temperature along the Planckian locus or to CCT, but the fact is that (as you point out in your color management book, p 20) CCT is insufficient to specify the color of an illuminant. For example there are an infinite number of spectra having a CCT of 6500K. In ACR one must use the tint slider in addition to the temperature slider. If the illumination is from a black body radiator, the temperature alone is sufficient; otherwise, one must use the tint along with the temperature.

Incandescent illumination is very close to a black body radiator ([a href=\"http://en.wikipedia.org/wiki/Color_temperature]Wikipedia article[/url]) and I have noted that the the temperature slider will usually achieve WB in ACR when one is working with this type of illumination--only a small amount of tint is needed. This doesn't prove anything, but it suggests to me that the temperature slider is along the Planckian locus.

I do wish that Karl Lange would enter into the discussion.

Bill
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5469


WWW
« Reply #50 on: May 25, 2007, 12:26:20 PM »
ReplyReply

Quote
Incandescent illumination is very close to a black body radiator (Wikipedia article) and I have noted that the the temperature slider will usually achieve WB in ACR when one is working with this type of illumination--only a small amount of tint is needed. This doesn't prove anything, but it suggests to me that the temperature slider is along the Planckian locus.
[a href=\"index.php?act=findpost&pid=119561\"][{POST_SNAPBACK}][/a]

But don't forget to add the the fact that a sensor suffers a LOT from metameric failure...they are designed primarily to provide a spectrally accurate response to "Daylight" and thus have R/G/B filters that do that for continuous light that has some sort of balance in the spectra. There is so little blue light in tungsten that there's no way a sensor will have an accurate response under tungsten-blue cutoff filters are very inefficient and there's very little blue light at 3200K or below.

So, regardless of how this stuff is SUPPOSED to work, all this stuff is really a kludge. It's amazing to me we can anything close to anything to actually work.
Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #51 on: May 25, 2007, 01:20:45 PM »
ReplyReply

Quote
But don't forget to add the the fact that a sensor suffers a LOT from metameric failure...they are designed primarily to provide a spectrally accurate response to "Daylight" and thus have R/G/B filters that do that for continuous light that has some sort of balance in the spectra. There is so little blue light in tungsten that there's no way a sensor will have an accurate response under tungsten-blue cutoff filters are very inefficient and there's very little blue light at 3200K or below.

So, regardless of how this stuff is SUPPOSED to work, all this stuff is really a kludge. It's amazing to me we can anything close to anything to actually work.
[{POST_SNAPBACK}][/a]


The blue output of a 3200K bulb is not that bad as shown in this [a href=\"http://www.gelighting.com/na/business_lighting/education_resources/literature_library/catalogs/downloads/cat_ss_fluorescent_appendix.pdf]SPD 3200K Lamp[/url] graph. The relative energy at 400 nm of the 3200K lamp is half that of daylight. This means that the 8.2 um pixel of a Canon EOS 5D can collect more blue light energy than the 5.5 um Nikon D2X can in full daylight. At base ISO both sensors perform quite well at 3200K, but at high ISO noise becomes a problem with the smaller pixel.

Bill
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5469


WWW
« Reply #52 on: May 25, 2007, 02:51:34 PM »
ReplyReply

Quote
This means that the 8.2 um pixel of a Canon EOS 5D can collect more blue light energy than the 5.5 um Nikon D2X can in full daylight. At base ISO both sensors perform quite well at 3200K, but at high ISO noise becomes a problem with the smaller pixel.
[a href=\"index.php?act=findpost&pid=119601\"][{POST_SNAPBACK}][/a]

Which still ignores the fact that blue separation filters over the site are the least efficient. Take a look at the Wratten filter transmission properties (sorry, I can't find my Kodak Filter Guide at the moment) and I think you'll find that between Red Wratten 29, Green Wratten 61 and a Blue Wratten 47, the blue filter's transmission is way low. Add the filter's low transmission properties to the fact that there is reduced blue light and you'll find there just are not a lot photons hitting a blue sensor when used under tungsten light.
Logged
digitaldog
Sr. Member
****
Online Online

Posts: 8972



WWW
« Reply #53 on: May 25, 2007, 03:42:47 PM »
ReplyReply

Quote
I do wish that Karl Lange would enter into the discussion.

Bill
[a href=\"index.php?act=findpost&pid=119561\"][{POST_SNAPBACK}][/a]


He can be reached at info@lumita.com
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Schewe
Sr. Member
****
Offline Offline

Posts: 5469


WWW
« Reply #54 on: May 25, 2007, 03:46:22 PM »
ReplyReply

Ah...I finally found it. (the book)

A wratten 29 (red) at its primary wavelengths of between 620-700nn is passing about 90% (90.5% at 700nn) of the red light-meaning about 90% efficient. A green 61 at it's peak transmission of 520nn is about 40% (which is one factor of why there are usually two green collection sites) but the blue 47 at peak of 440nn is 50%. The fact that there are two green sites in a Bayer array and a 90% transission efficiency of the red and you see that if you only have 1/2 of the blue light going through a 50% efficient filter and only one site to collect from, the loss of blue data from a digital capture will result in a metameric failure when comparing a capture done under tungsten and one done at D65.

Which is another reason that white balancing using a two axis slider can be problimatic at best. There just isn't a lot of blue data captured with a digital sensor.
Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #55 on: May 25, 2007, 04:31:39 PM »
ReplyReply

Quote
Really? I was under the impression that the color temperature slider adjusted temperature along the Planckian locus and tint at right angles to the locus. What is your line of reasoning? 
[{POST_SNAPBACK}][/a]

Quote
I doubt it does. Its using some correlated color temperature I'm pretty darn sure. We could ask a true color scientist like Karl Lang about this. I think you're correct about the overall angle in color with respect the locus but does it follow it exactly? I would guess not.

Least we forget, the blackbody is based on a theoretical object that doesn't exist. Hence the reason we should use correlated color temperature since so few (any?) natural objects conform to the behavior of this theoretical object (the blackbody).
[a href=\"index.php?act=findpost&pid=119543\"][{POST_SNAPBACK}][/a]

Well, we have the answer straight from the horses mouth in a reply to my question posted on the Adobe Camera Raw Forum:

[a href=\"http://www.adobeforums.com/cgi-bin/webx/.3bb6a85c.3bc4009b/0]Link to thread[/url]

Quote for those who don't use the Adobe Forums:

 
Operation of Temperature and Tint sliders in ACR
Bill Janes - 08:46am May 25, 2007 Pacific
I was under the impression that the temperature slider in ACR adjusted for color temperature along the Planckian locus (blue-yellow) and the tint slider adjusted for situations in which the illumination deviates from that of a black body radiator. Is this correct?


Thomas Knoll - 12:06pm May 25, 07 PST (#1 of 1)
   
Yes.
Logged
digitaldog
Sr. Member
****
Online Online

Posts: 8972



WWW
« Reply #56 on: May 25, 2007, 04:51:06 PM »
ReplyReply

Quote
Well, we have the answer straight from the horses mouth in a reply to my question posted on the Adobe Camera Raw Forum:

Link to thread

Quote for those who don't use the Adobe Forums:

 
Operation of Temperature and Tint sliders in ACR
Bill Janes - 08:46am May 25, 2007 Pacific
I was under the impression that the temperature slider in ACR adjusted for color temperature along the Planckian locus (blue-yellow) and the tint slider adjusted for situations in which the illumination deviates from that of a black body radiator. Is this correct?
Thomas Knoll - 12:06pm May 25, 07 PST (#1 of 1)
   
Yes.
[a href=\"index.php?act=findpost&pid=119635\"][{POST_SNAPBACK}][/a]

I think the question is partially answered. Yes, it sounds like the sliders adjust for color temp along the locus (blue yellow) but does that mean they do this as exactly defined or using CCT? The direction shouldn’t be questionable, of course it moves on that axis. Is it producing those exact SPD's?
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #57 on: May 25, 2007, 06:27:13 PM »
ReplyReply

Quote
Ah...I finally found it. (the book)

A wratten 29 (red) at its primary wavelengths of between 620-700nn is passing about 90% (90.5% at 700nn) of the red light-meaning about 90% efficient. A green 61 at it's peak transmission of 520nn is about 40% (which is one factor of why there are usually two green collection sites) but the blue 47 at peak of 440nn is 50%. The fact that there are two green sites in a Bayer array and a 90% transission efficiency of the red and you see that if you only have 1/2 of the blue light going through a 50% efficient filter and only one site to collect from, the loss of blue data from a digital capture will result in a metameric failure when comparing a capture done under tungsten and one done at D65.

Which is another reason that white balancing using a two axis slider can be problimatic at best. There just isn't a lot of blue data captured with a digital sensor.
[{POST_SNAPBACK}][/a]


We do not really need to consider Wratten filters to answer the question, since the answer is in the raw files. In some experiments with my D70 I photographed a MacBeth Color Checker under 40 watt tungsten illumination (2950 K according to ACR) and daylight and examined the raw files (converted with Iris and with the pixel values multiplied by 8 to convert from the raw 0..4095 to the PS 0..32767 "16 bit" format). For convenience, the results are displayed in Photoshop as 8 bit. Since the sensor is linear, the pixel value is proportional to the luminance seen by the sensor elements.

Here are the results for daylight, showing the pixel values and histogram for the white patch:

[attachment=2556:attachment]

Here is the same for tungsten, ~2950K
[attachment=2557:attachment]

To achieve white balance, one multiplies the red and blue pixel values by a scaling factor as shown in this [a href=\"http://www.pochtar.com/NikonWhiteBalanceCoeffs.htm]table[/url].

For 2950 the RGB multipliers are 1.32, 1.0, and 2.9. In terms of f/stops, the red channel is down 0.4 stops and the blue channel 1.54 stops.

For daylight the RGB multipliers are 2.06, 1.0, and 1.6, or expressed in f/stops the red and blue channels are down 1.04 and 0.68 stops respectively. These factors have nothing to do with the fact that the Bayer array has twice as many green sensors as blue and red sensors: the missing colors is each Bayer pixel are filled in by interpolation during the demosaicing process. I don't think the differences have anything to do with metamerism.

The from the above pixel values, the ratio of light falling on the blue sensor in daylight to that with tungsten 2950K (daylight:tungsten) is 1.68 or about 0.75 f/stops. This is about the same change in ratio that takes place when you change the ISO on the camera from 100 to 160, and this does not affect performance that much. At low ISO we get very good results with the digital sensor at 2950K. The blue sensors have plenty of light to work with, and they are less taxed in this situation, than when base ISO is increased from 100 to 200 with daylight illumination.
« Last Edit: May 25, 2007, 06:52:34 PM by bjanes » Logged
PeterLange
Guest
« Reply #58 on: May 25, 2007, 06:30:01 PM »
ReplyReply

Quote
But don't forget to add the the fact that a sensor suffers a LOT from metameric failure...

WAY OFF.  Here’s another one from Thomas Knoll: "Actually, to create a camera filter set that is "perfect", it is not required to exactly the match the human cone responses (or the XYZ responses). All that is required is the filter responses be some linear combination of the human cone responses. If that is the case, then a simple 3 by 3 matrix can be used in software to recover the exact XYZ values.”

Referring to ACR, matrix-to-matrix interpolation tries to compensate for remaining imperfections: “Both color matrices are going to be similar. The closer the camera's filters are to "perfect", the closer to two matrices are going to be” (again by T.K.). In addition, Calibrating ACR can compensate for unit-to-unit deviations, or, for the surprising finding – just as Barry Clive Pearson states – that the zero state of the calibrate tab may have its own hidden agenda away from best fit matrices and colorimetric rendering. Anyway.

IF your thesis would be at faintest true, we should instantly trash any Raw conversion software which does NOT use Lut type input profiles per illuminant – such as Camera Raw and Lightroom.

Too bad.  And oh, this subject has less to do with whitebalance in terms of linear scaling per R/G/B channel and resulting CIExy moves such as e.g. along the Plankian locus or isotemperature lines thereof (CT slider) and orthogonal CCT lines (Tint). Works nicely with Raw at even light conditions, but does not work very well with output-referred JPEG’s from in-camera conversion where R/G/B curves might be more successful (see Bruce Fraser’s early articles at CreativePro).  And while we are at it, gamma encoding is completely irrelevant here, because the sequence of linear scaling and conversion to different gamma even back to 1.0 is commutative.

Cheers! Peter

--
Logged
bjanes
Sr. Member
****
Online Online

Posts: 2781



« Reply #59 on: May 25, 2007, 06:44:20 PM »
ReplyReply

Quote
I think the question is partially answered. Yes, it sounds like the sliders adjust for color temp along the locus (blue yellow) but does that mean they do this as exactly defined or using CCT? The direction shouldn’t be questionable, of course it moves on that axis. Is it producing those exact SPD's?
[a href=\"index.php?act=findpost&pid=119636\"][{POST_SNAPBACK}][/a]

Maybe you have some fine points for clarification, but the sliders operate exactly as I postulated to the previous poster in this thread. Mr. Knoll's answer is good enough for me.
Logged
Pages: « 1 2 [3] 4 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad