Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Gio Scope Tool vs The Zone System

Jake Wilganowski said:
Yes 5. But you bring up a very interesting thing regarding the 2 OLPF's. If you had 2 dragons set the exact same but with 2 different OLPF's and pointed them at the same scene, what would you see in Gioscope, would they look different? I have a hunch Phill Holland might be able to give us the answer.
Click to expand...

The GIO Scope should be different then, remeber the GIO scope looks at the raw image, if you have for example a H20 olpf in one camera and a normal olpf in the other. The H2O camera will have a far later burn in the blue channel, etc. Which is right it measures how the "photo buckets" of the sensor is filling up and nothing else, so all filters infront of sensors will be in play, not like the normal histogram that shows whats shown on screen.
 
MichaelJQuill said:
Thanks David. The more I experiment and think about this topic, the more it seems that the practice of 'mapping caucasion skin to chip 11 or 12 aka one stop above middle grey' to be misleading/simplistic.

Doesn't the above totally depend on your environment and the look you're going for? For example, if you are exposing a Caucasian person's face under direct sunlight (i.e., you are using the sun as a 'key light' as opposed to a 'backlight'), shouldn't that be like 2 to 3 stops above middle grey, which would place the key side of the subject's face on chip 13 or 14? If, on the other hand, you are using the sun as a 'backlight,' and 'keying' the subject's face with the ambient skylight, then here you might place the subject's face on chip 12 (one stop above middle grey).

Following the same logic, imagine a Caucasian subject in a dark club environment, where the 'key light' is a candle on the table. Mapping your subject's face to chip 12 would feel too bright for that kind of low light environment. It would seem more natural to map the person's face to one or two stops under middle grey, which would be chip 9 or 10 if using GIO Scope.

So, whether we are applying The Zone System (w/ a spot meter) or Red's GIO Scope, I'm wondering what 'map Caucasion skin to one stop above middle gray to produce proper exposure' really means?

I'm hoping to get a more detailed explanation than, 'It's just a healthy place to set exposure, and you can push or pull the image in post' or 'It's always better to expose to the right and crush your image in post to minimize grain.' Or if these are simply the 'correct' answers, please go into a little more detail :)

Thanks much,
Mike
Click to expand...

Best to go back to the OT: what is the Zone System? The Zone System is not just about exposing middle gray (Zone V). It's about exposing and balancing all the zones, from Zone I to Zone X, and it's about balancing the contrasts within sub-regions to create higher-order balance across the whole image. In this way, a Zone V space can balance (and contrast against) an alternating Zone II/Zone IX space. A Zone V space can transition from Zone IV to Zone VI. Etc.

In the case of a caucasian face under direct sunlight, you have two important questions to ask: (1) how should the exposure of the face define (or be defined by) the overall image, and (2) what technical exposure strategies best accomplish (1). In the case of a caucasian face illuminated by candlelight in a lowlight environment, the two important questions are the same. The GIO scope helps you see the exposure range that your RAW image can comfortably handle (i.e., how much you can pull/push in post). It also helps you anchor that to the look you are trying to create in the image. There are good reasons you might want all the faces in the scenes to fade into the shadows. But if you want the face to look "normal", if you want it to read as "hey, look here! This is what you are supposed to focus on!", then say *that* with the appropriate zone, within the contrast range of the image, not some answer that comes from math about only the subject's face.
 
Thant's right micheal,

It goes back to mapping the middle grey according to the scene and where you (as the image maker) want the the exposure focus to be. There is no specific rule (like the zone system) but more the "place from which to base an exposure decision'. So with the gioscope you can work out the values in such a way that you can map your middle grey higher or lower (by using different exposure methods) to get to where you want the main subject matter to be.

In any case, using any zone system is also very much a process of CAPTURE/PROCESSING, as an example you can then intentionally for creative purposes process the exposed material for a desired result.

It's a very flexible way to go about things. But, it is a capture/process thing, like film.

The one thing the GIOSCOPE gives you is a constant and that constant is that it is mapping middle grey at 10 or 11 based on an exposure rated 800ISO @F8.0 in my tests
 
Thanks for the informative answers, guys. This all makes sense.
 
Last edited:
After reading through Phil's book and comments on middle grAy, I remembered this thread. OK ... shoot me, but I don't like where RED put middle gray default for the dragon epic. I want more dynamic range in my middle grEy area, this is why I shifted my black shading so the gioscope middle grey on my camera lines up to "9" instead of "11". I come at this more from a painting (and old school Californian B&W) viewpoint, where those middle grey/colors are very important to me and I want a little bit of a dynamic range "bump" in that middle grey area. With Phils other thread on weapon dynamic range, it so far looks like RED has spread the middle grey a bit ... so I should be ok (but it's got me worried).
 
PatrickFaith said:
After reading through Phil's book and comments on middle grAy, I remembered this thread. OK ... shoot me, but I don't like where RED put middle gray default for the dragon epic. I want more dynamic range in my middle grEy area, this is why I shifted my black shading so the gioscope middle grey on my camera lines up to "9" instead of "11".
Click to expand...
This is an interesting idea, i think i have seen you mentioning this before. But i have some trouble to understand what the pro's and con's are and how you do that. Aren't you sacrificing DR? And couldn't you achieve the same results by just exposing differently?
 
Maik Müller said:
This is an interesting idea, i think i have seen you mentioning this before. But i have some trouble to understand what the pro's and con's are and how you do that. Aren't you sacrificing DR? And couldn't you achieve the same results by just exposing differently?
Click to expand...
Your right on the DR, and for awhile I did the same with with around 4 nuke nodes, but they where more complex then you'd think (i.e. they had to be multi pixel color nodes with complex alpha masking, a simple color lookup table couldn't do it). I use an "Ansel Adams" like zone system (which is what I grew up on), I like the bottom end of the spectrum linear with deep rich/smooth rembrandt type blacks(zone1-3 with popping highlights in zone 9&10), while the "color" part i like a lot of detail in zones 4-8 (so I like 5 bits here), and I like some compressions in the color space above zone 7 to pure unsaturated whites in zone 10. I like to setup my lighting for this approach, and really can't use more then 10 bits on my display, so I'm not so fanatical as some people on having more then 15 stops of DR. Also Graeme made a modification so when I blow out my whites in zones 9 &10 they don't have a color shift, so I don't mind clipping on the top bit (i do a zone 9-10 adjust anyway to pure white). This also matches up fairly well to 1886 gamma on monitors.
 
How do you change your black balance to remap middle grey?
 
Tommaso Alvisi said:
Patrick, why not use ISO and FLUT or LGG settings in camera which is completely supported by RED SDK?
Click to expand...
How red has implemented those, they don't seem to have the low pass spatial operations in them. I want the "camera" to do low pass spatial operations on zones 1-3 and 9&10, I get good spatial operations on zones 9&10 with the RED OLPF options, and use black shading on zones 1-3(I can't handle the extra time ADD takes, plus i would want ADD as a knob hooked with ISO if I was to use it).

(i.e. I want a frequency filter on both sides of a sensor, on sound it's obvious to put them around 50hz and 12khz during the recording, on a rgb sensor it's awful tricky, and I don't have a monitor in production where I can set it correctly, technically I do have a Ezio 318 on my DIT station, but really i just look at the RED monitors mostly ... and on those I don't notice spatial filter problems until post)
 
VERY interesting stuff Patrick. Thanks for explaining your ideas and intentions. I appreciate that a lot. Have never thought about it that way.
 
GIO Scope, Zone System, Manifold Recording, LUT

GIO Scope, Zone System, Manifold Recording, LUT

OK! I've posted some news about my latest skunkworks project in some other LUT-related threads, but really this is the thread I should have posted to. And to continue the discussion, rather than just posting a link to YouTube (demo video) or DropBox, I'll start by sharing these first few paragraphs of README.txt:

HOW TO USE THE GIO LUT Tool from Manifold Recording

*** This is experimental software and subject to change. ***

But! It is also useful, and it gets better when more people use it
and give feedback. So play with it, be sensible, and let us know what
you like, what you think would make it better (especially if you have
patches to offer!), and whatever else you think we should know about
RED's IPP2, LUTs, CDLs, etc.

The LUTs are in the luts/ subdirectory. Right now there are four
complete GIO cubes:

GIO_Log3G10_601.cube
GIO_Log3G10_709.cube
GIO_Log3G10_max.cube
GIO_Log3G10_min.cube

Each is a 33x33x33 3D LUT that can be loaded into a DSMC2 camera, Red
Cine-X, or Resolve 14.2. The demonstration video
(https://youtu.be/s_Umzcv7RJA) used GIO0_Log3G10_709.cube, but
GIO_Log3G10 cubes (no "0" after GIO) use an updated CDL and more
accurately match the GIO Scope colors. The 601 cube uses a different
blending of R, G, and B values to derive its luminance. The min and
max cubes simply pick the respective min and max of the R, G, and B
values as the luminance. Using the min cube might be helpful in
seeing trouble spots in shadow noise. Using the max cube might be
helpful spotting trouble spots in highlights.

There are also 33 special cubes which return a full intensity value
precisely when R == G == B for a single intensity value. When such a LUT
is applied to an image, everything is black except for those pixels
that fit that singular condition within the cube (one singe entry
in a cube of nearly 36,000 values).

There are two CDL files: GIO_CDL-v0.1.cdl contains the parameters
needed to stretch the Log3G10 IPP2 pipeline to fit the bins of the 3D
LUT. ResetAll.cdl is a convenience for DSMC2 camera operators who
want to cancel out the effects of the GIO CDL.

If you want to play with CDL values, GIO colors, etc., the G-XX-33.cube
files can be helpful to see only the parts of an image that will map to a
specific output level in the LUT, regardless of anything else.

The theory of operation is this: the IPP2 pipeline delivers Log3G10
pixels of RGB (actually REDWideRGB) data to a CDL, which scales them
according to the Slope parameter, adds an offset, and then applies a
Power operator. The result of the CDL then feeds a 3D LUT
implementing the GIO Scope false color mapping. The output of the LUT
can then be transformed by the usual IPP2 output mapping LUTs such as
Medium Contrast, Medium Highlight Rolloff and REC 709 Output gamma.

In a Log3G10 IPP2 pipeline, the setup is as follows [...]

For complete README.txt and relevant files, vist this DropBox directory: https://www.dropbox.com/sh/moy6d0hctjqtt8y/AADQxE7buTKk2Em1ndGlhaNUa?dl=0

You may also wish visit this blog posting (click on image):

 
Hey Michael,

Great work!

I'm a little confused as to why you need the CDL portion; could it not just map the IPP2 (RWG/L3G10) to the GIOscope colours via a 3Dlut? My thinking is you could apply the LUT to a IPP2 image, quickly adjust the ISO/flut to find the "proper" exposure (e.g. make sure the caucasian skin tones fall on chip 12), and then from there you know the starting point is "correct" (even though you may have shot a stop or two lower/higher in order to preserve highlights/shadows).

Sidenote: a few years back (actually, I think it was in GioScope announcement thread), I asked if RED could implement Video and Raw Exposure Check toggles in RCXp (to do a similar thing -- check exposure in post). They didn't respond, heh.
 
Mike P. said:
Hey Michael,

Great work!

I'm a little confused as to why you need the CDL portion; could it not just map the IPP2 (RWG/L3G10) to the GIOscope colours via a 3Dlut? My thinking is you could apply the LUT to a IPP2 image, quickly adjust the ISO/flut to find the "proper" exposure (e.g. make sure the caucasian skin tones fall on chip 12), and then from there you know the starting point is "correct" (even though you may have shot a stop or two lower/higher in order to preserve highlights/shadows). [...]
Click to expand...

I kinda try to answer that in the blog posting, but here's the math. The image pipeline feeds a 3D LUT with data triples that ranges from 0 to 1 for each of the RGB pixels. To simplify things, let's imaging that we're just concerned with perfectly monochromatic data, so let's pretend than our 33x33x33 LUT is actually just a 33-entry LUT. Pixels from zero to 1/32 go into the first (or zeroth) bucket. Pixels from 1/32 to 2/32 goes into the next. And so forth until you get to the last bucket, which catches only pixels that are pure white.

1/32 = 0.03125

Log3G10(0) = 0.09155148771474521, which means that a black pixel in Log3G10 space goes not into the zero bucket, nor the first, nor the second, but the 3rd bucket.

Log3G10(mid-gray = 0.18) = 1/3, which means that GIO Scope 11 goes into the 10th or 11th bucket, depending on how you round. So there are 7-8 buckets available to encode 11 stops. There's no way a LUT can give you distinct colors if it doesn't have enough buckets.

1/64 = 0.015625, Log3G10 skips the first 6 buckets. With a 65x65x65 cube there are 21 buckets between true zero (which Log3G10 only reaches with a negative pixel value of 0.01) and middle gray. When you subtract out the 6 that are missed up front, there are 15 left, which is clearly enough to theoretically encode 11 GIO Scope colors. But also only if the logarithmic math spreads the values widely enough. And I think that's a little dicey.

By themselves, LUTs cannot choose how they space out their buckets to best divide the data. But CDLs can put some powerful shapes on that data before they hit the LUTs, putting the darkest and brightest pixels at the very extremes of the LUT, and sloping the curve so as to optimize how the LOG data is divided to best match the real responsiveness of the GIO Scope curves vs. Log3G10 values. Maybe that's a good approach to use for Davinci Resolve, until they decide to give us access to the CDL part of IPP2 (which they should, anyway).
 
Last edited:
For fun I wrote a program to compare/contrast how CDLs make or break the GIO LUT. Here's 65 buckets with no CDL:

# [scale, offset, power] = 1 0 1.0
gioColors65 = [
[0.29,0,0.36],# UV-1 @ 0
[0.29,0,0.36],# UV-1 @ 1
[0.29,0,0.36],# UV-1 @ 2
[0.29,0,0.36],# UV-1 @ 3
[0.29,0,0.36],# UV-1 @ 4
[0.29,0,0.36],# UV-1 @ 5
[0.29,0,0.36],# UV-1 @ 6
[0, 0.42, 0.51], # MedBlue-5 @ 7
[0, 0.48, 0.55], # Sky-6 @ 8
[0, 0.61, 0.61], # Cyan-7 @ 9
[0.3, 0.51, 0.47], # CyanG-8 @ 10
[0.3, 0.51, 0.47], # CyanG-8 @ 11
[0, 0.51, 0.44], # Aqua-9 @ 12
[0, 0.51, 0.44], # Aqua-9 @ 13
[0, 0.51, 0.44], # Aqua-9 @ 14
[0, 0.51, 0], # Green-10 @ 15
[0, 0.51, 0], # Green-10 @ 16
[0, 0.51, 0], # Green-10 @ 17
[0, 0.51, 0], # Green-10 @ 18
[0.35, 0.48, 0], # GreenY-11 @ 19
[0.35, 0.48, 0], # GreenY-11 @ 20
[0.35, 0.48, 0], # GreenY-11 @ 21
[0.41, 0.45, 0], # YellowG-12 @ 22
[0.41, 0.45, 0], # YellowG-12 @ 23
[0.41, 0.45, 0], # YellowG-12 @ 24
[0.41, 0.45, 0], # YellowG-12 @ 25
[0.45, 0.51, 0], # YYG-13 @ 26
[0.45, 0.51, 0], # YYG-13 @ 27
[0.45, 0.51, 0], # YYG-13 @ 28
[0.45, 0.51, 0], # YYG-13 @ 29
[0.45, 0.51, 0], # YYG-13 @ 30
[0.55, 0.53, 0], # Yellow-14 @ 31
[0.55, 0.53, 0], # Yellow-14 @ 32
[0.55, 0.53, 0], # Yellow-14 @ 33
[0.55, 0.53, 0], # Yellow-14 @ 34
[0.51, 0.37, 0], # Orange-15 @ 35
[0.51, 0.37, 0], # Orange-15 @ 36
[0.51, 0.37, 0], # Orange-15 @ 37
[0.51, 0.37, 0], # Orange-15 @ 38
[0.51, 0, 0], # Red-16 @ 39
[0.51, 0, 0], # Red-16 @ 40
[0.51, 0, 0], # Red-16 @ 41
[0.51, 0, 0], # Red-16 @ 42
[0.34, 0, 0], # Red-16.1 @ 43
[0.34, 0, 0], # Red-16.1 @ 44
[0.34, 0, 0], # Red-16.1 @ 45
[0.34, 0, 0], # Red-16.1 @ 46
[0.34, 0, 0], # Red-16.1 @ 47
[0.23, 0, 0], # Red-16.2 @ 48
[0.23, 0, 0], # Red-16.2 @ 49
[0.23, 0, 0], # Red-16.2 @ 50
[0.23, 0, 0], # Red-16.2 @ 51
[0.1, 0, 0], # Red-16.3 @ 52
[0.1, 0, 0], # Red-16.3 @ 53
[0.1, 0, 0], # Red-16.3 @ 54
[0.1, 0, 0], # Red-16.3 @ 55
[0.05, 0, 0], # Red-16.4 @ 56
[0.05, 0, 0], # Red-16.4 @ 57
[0.05, 0, 0], # Red-16.4 @ 58
[0.05, 0, 0], # Red-16.4 @ 59
[0.05, 0, 0], # Red-16.4 @ 60
[0.03, 0, 0], # Red-16.5 @ 61
[0.03, 0, 0], # Red-16.5 @ 62
[0.03, 0, 0], # Red-16.5 @ 63
[0.03, 0, 0], # Red-16.5 @ 64
]
Click to expand...

Notice first that there's no way to distinguish GIO-2, 3, and 4 from GIO-5. Notice second that GIO-5, 6, and 7 have only a single bucket allocated, which means there will be more than a little color interpolation, blurring the results of the scope. (It would be great if one of the LUT interpolation algorithms was "none" instead of tri-linear, tetrahedral, etc.) Notice third that GIO-16 uses almost a third of all the buckets just for its varying shades of red. GIO-1 and GIO-16 are stealing buckets from all the other colors!

By contrast, here's the latest GIO cube I've built, which sees everything from 10 stops below G-11 to 10 stops above G11. Notice here that in almost all cases there are at least two buckets per color, meaning that values falling between those buckets will be interpolated to precisely their proper GIO color. And of course values less than 10 stops below middle gray are truncated to zero, which means they'll get the pure "ultra-violet" color at the start of the cube.

# [scale, offset, power] = 1.1 -0.10184 0.5
gioColors65 = [
[0.29,0,0.36],# UV-1 @ 0
[0.42,0,0.51], # Magenta-2 @ 1
[0.42,0,0.51], # Magenta-2 @ 2
[0.4,0.39,0.61], # Violet-3 @ 3
[0.4,0.39,0.61], # Violet-3 @ 4
[0, 0.32, 0.46], # DkBlue-4 @ 5
[0, 0.32, 0.46], # DkBlue-4 @ 6
[0, 0.42, 0.51], # MedBlue-5 @ 7
[0, 0.42, 0.51], # MedBlue-5 @ 8
[0, 0.48, 0.55], # Sky-6 @ 9
[0, 0.48, 0.55], # Sky-6 @ 10
[0, 0.48, 0.55], # Sky-6 @ 11
[0, 0.61, 0.61], # Cyan-7 @ 12
[0, 0.61, 0.61], # Cyan-7 @ 13
[0, 0.61, 0.61], # Cyan-7 @ 14
[0, 0.61, 0.61], # Cyan-7 @ 15
[0.3, 0.51, 0.47], # CyanG-8 @ 16
[0.3, 0.51, 0.47], # CyanG-8 @ 17
[0.3, 0.51, 0.47], # CyanG-8 @ 18
[0.3, 0.51, 0.47], # CyanG-8 @ 19
[0, 0.51, 0.44], # Aqua-9 @ 20
[0, 0.51, 0.44], # Aqua-9 @ 21
[0, 0.51, 0.44], # Aqua-9 @ 22
[0, 0.51, 0.44], # Aqua-9 @ 23
[0, 0.51, 0.44], # Aqua-9 @ 24
[0, 0.51, 0], # Green-10 @ 25
[0, 0.51, 0], # Green-10 @ 26
[0, 0.51, 0], # Green-10 @ 27
[0, 0.51, 0], # Green-10 @ 28
[0, 0.51, 0], # Green-10 @ 29
[0.35, 0.48, 0], # GreenY-11 @ 30
[0.35, 0.48, 0], # GreenY-11 @ 31
[0.35, 0.48, 0], # GreenY-11 @ 32
[0.35, 0.48, 0], # GreenY-11 @ 33
[0.41, 0.45, 0], # YellowG-12 @ 34
[0.41, 0.45, 0], # YellowG-12 @ 35
[0.41, 0.45, 0], # YellowG-12 @ 36
[0.41, 0.45, 0], # YellowG-12 @ 37
[0.45, 0.51, 0], # YYG-13 @ 38
[0.45, 0.51, 0], # YYG-13 @ 39
[0.45, 0.51, 0], # YYG-13 @ 40
[0.45, 0.51, 0], # YYG-13 @ 41
[0.55, 0.53, 0], # Yellow-14 @ 42
[0.55, 0.53, 0], # Yellow-14 @ 43
[0.55, 0.53, 0], # Yellow-14 @ 44
[0.51, 0.37, 0], # Orange-15 @ 45
[0.51, 0.37, 0], # Orange-15 @ 46
[0.51, 0.37, 0], # Orange-15 @ 47
[0.51, 0.37, 0], # Orange-15 @ 48
[0.51, 0, 0], # Red-16 @ 49
[0.51, 0, 0], # Red-16 @ 50
[0.51, 0, 0], # Red-16 @ 51
[0.34, 0, 0], # Red-16.1 @ 52
[0.34, 0, 0], # Red-16.1 @ 53
[0.34, 0, 0], # Red-16.1 @ 54
[0.23, 0, 0], # Red-16.2 @ 55
[0.23, 0, 0], # Red-16.2 @ 56
[0.1, 0, 0], # Red-16.3 @ 57
[0.1, 0, 0], # Red-16.3 @ 58
[0.1, 0, 0], # Red-16.3 @ 59
[0.05, 0, 0], # Red-16.4 @ 60
[0.05, 0, 0], # Red-16.4 @ 61
[0.05, 0, 0], # Red-16.4 @ 62
[0.03, 0, 0], # Red-16.5 @ 63
[0.03, 0, 0], # Red-16.5 @ 64
]
Click to expand...
 
Ressurecting this flaming dumpster of a thread to ask a simple question about Gio scope:

Does anyone have a simply communicable process for using Gio scope in their day-to-day that they find really helpful and useful. And if so, what is it?
 
It's works as a raw-sensor false colour, whereas regular false colour (aka video check) is based off the metadata (affected by ISO, curves, LUTs, etc). So if you know how to use false colour, you should be good to go -- just toggle on the chips/stops you want to get critical exposure on (chip 11 being mid-grey)

What confuses me is if it takes the different OLPFs into consideration and if it takes the ISO re-calibration of Helium+ sensors into account. I believe all RED's with gioscope have a raw view of ISO800 (which would too high for dragon with STH), and I'm not sure how it is mapped with Helium onwards (post ISO-recalibration), so that's something to test.

OH, and put it on a button to make it toggleable (so that you can click-check-click back to normal view).
 
GIO Scope makes it somewhat easier to identify not only highlight, shadow, and mid tone exposure values, but also contrast ratios within those ranges. The video exposure false color tool identifies gross exposure levels, but difficult to tell whether you have a 1, 2, 3, or 4-stop contrast range across a subject.
 
What MT is saying is that unless you have previous in-depth experience with exposure bracketing and the Zone System the Gioscope ain't gonna mean much to you. This thread break-down isn't an explanation in using the tool to help you determine proper exposure, it is a break-down on how to use the tool to confirm your lighting ratios from set to camera raw. The reality in 2019 is that all high-end digital cameras capture more dynamic range than film ever did or will. Ansel's ZS was a scientific process in testing film negative for latitude (DR) that was directly related to contrast ratios in the print. If you don't know what contrast ratios you hope to achieve in the print the ZS isn't helpful as a previz tool. The Gioscope is the digital equivalent.

There is a heavy tendency in digital to under-light scenes because contrast can still be applied to a muddy image via final color apps like Resolve. The resulting image is acceptable but definitely not the cleanest. If shot on film these scenes would be unusable because the complete lack of contrast in the negative would equal a complete lack of contrast in the print. There are always compromises.

The Gioscope becomes useful when you know what "print" of your "negative" you hope to achieve. The Gioscope will tell you how much fill needs to be added to your actor's shadow side so his/her features can just be made out but still remain in the toe of the image without crushing to black or becoming infested with noisy pixels.

As always Ansel's The Negative https://www.goodreads.com/book/show/20502.The_Negative is the Good Book for anyone concerned with photographing images.
 
You must log in or register to reply here.
Back
Top