Output Parameter

PathofOutput (/PA=.\) ⇧

Range of values: characters
Verzeichnis, in dem die Ergebnisse gespeichert werden. Der Pfad kann relativ angegeben werden (Beispiele: "Ergebnisse", ".\Ergebnisse\" oder "..\Ergebnisse", wobei ".\" das Verzeichnis PathofImages ist) oder absolut (Beispiel: "D:\Urlaub\Ergebnisse").

LeftRightOutput (/LR=2) ⇧

Range of values [0,1,2]
Writing two left-right-separated images:
0: No files for separate images are generated.
1: Two files for the left and right images are written into the path, given with the PathofOutput, one <name>_cl.* and one <name>_cr.*, with "_cl" means correct left and "_cr" means correct right.
2: The files will be tored into the subdirectories \cl and \cr of PathofOutput. Attention: If you have set AviOutput = 9, alle single frames will be stored also into the same subdirectory \clr.
3: As 1 with identifiers "_OS" for oculus sinister and "_OD" for oculus dexter.

StereoOutput (/SO=2) ⇧

Range of values [0,1,2]
Writing one side-by-side stereo image:
0: No side-by-side mounted image file is generated.
1: A side-by-side mounted image file <name>_cs.* is written into the directory PathofOutput. Here "_cs" means correct stereo.
2: The file is stored into the subdirectory \cs of PathofOutput.
Note: Clicking the AV checkbox to the right of the combo box will additionally generate an SBS video (without audio), which contains all result images as a picture show. Please do not forget to set the parameter VideoFramerate, e.g. for a 5-second standing time VideoFramerate = 0.2fps. For a MP4 video at least 12 frames should be used, otherwise the video will not play correctly in some players.

AnaglyphOutput (/AO=2) ⇧

Range of values [0,1,2]
Stores as anaglyph images:
0: No anaglyph image file is generated.
1: An anaglyph image file <name>_ca.* is written into the directory PathofOutput. Here "_ca" means correct anaglyph.
2: The files is stored into subdirectory \ca of PathofOutput.

ExtraOutput (/EX=0) ⇧

Range of values [0,5]
Additional image output for special purposes. All additional images will be stored into the subdirectory .\ce_xxx\. xxx is a layout identificator.
0: No additional output images.
1: This writes a 1:2 squeezed side-by-side (SBS-H) mounted image with the ending "-lrq". The subdirectory identificator is "_lrq".
2: This writes a top-and-bottom (TAB) mounted image (left one on the top) with the ending "-ab". The subdirectory identificator is "_ab".
3: This writes a 1:2 squeezed top-and-bottom (TAB-H) mounted image with the ending "-abq". The subdirectory identificator is "_abq".
4: An interlaced image will be created (first line left, e.g for ACER 5740DG) with the ending "-eo". The subdirectory identificator is "_eo".
5: An MPO image will be created (for compatibility with JPG-Subsampling 4:2:0). The subdirectory identificator is "_mpo".
6: An L-R-L image is created with the extension -lrl. The identifier for the subfolder is "lrl".
7: An LR/RL image is created with the extension -lrrl. The identifier for the subfolder is "lrrl".
8: Not implemented.
9: If there is an RGB image and a depth mask, a ChromaDepth image is created in the "cd" subfolder. The identifier for the ChromaDepth image is also "_cd".
Note-1: Some stereo viewer software (e.g. the Stereoscopic Player) evaluate the layout identificator automatically to display the images. For that application, the minus sign ("-") must be used in the filenames instead of the (general with Cosima used) underline ("_").
Note-2: Clicking the AV checkbox to the right of the combo box will additionally generate an video (without audio), which contains all result images as a picture show. Please do not forget to set the parameter VideoFramerate, e.g. for a 5-second standing time VideoFramerate = 0.2fps. For a MP4 video at least 12 frames should be used, otherwise the video will not play correctly in some players.

Outputformat (/OF=0) ⇧

Range of values [0-9]
Setting the graphic format of the output images:
0: Resulting images will be written using the original file format.
1: Resulting images will be written using the jpg format.
2: Resulting images will be written using the tga format.
3: Resulting images will be written using the bmp format.
4: Resulting images will be written using the png format.
5: Resulting images will be written using the tif format (RGB8, 24bit/pixel).
6: Resulting images will be written using the tif format (RGB8A, 32bit/pixel).
7: Resulting images will be written using the tif format (RGB16, 48bit/pixel).
8: Resulting images will be written using the tif format (CMYK, 32bit/pixel) without applying any color profile.
9: Resulting images will be written using the tif format (CMYK, 32bit/pixel) with the application of a color space transformation according to sRGB to CMYK_Profile.

Attention: With the Outputformat = 7 (48Bit/Pixel) there are the following limitations at the moment:
- CroppInput != 2, no scaling of the smaller input image
- only LeftRightOutput, StereoOutput and ExtraOutput = 2, for the other output options (e.g. anaglyph) partly fallback to Outputformat = 5 (24Bit/Pixel)
- only InterpolationMode = 2, bi-cubic interpolation
- only SharpenIntensity = 0, no post sharpening
- only RotMirOutput = 0, no mirroring or rotation of the output image
- only GammaCorrection = 0, no gamma correction with the interpolation
- only AutoGhostbust = 0, no cross-over compensation
- only ExtColorManagement = 0, no apply of external color profiles (e.g. Infitec)
- only CaptionText empty, no caption
- only ImageText empty, no text inside of the image
- only InterlacedMode = 0, no interlaced input images

CMYK_Profile (/CM=) ⇧

Range of values: characters
If you choose Outputformat = 9, a further inquiry window opens to choose a CMYK color profile.
The following Adobe CMYK color profiles are delivered:
CoatedFOGRA27.icc
CoatedFOGRA39.icc
CoatedGRACoL2006.icc
JapanColor2001Coated.icc
JapanColor2001Uncoated.icc
JapanColor2002Newspaper.icc
JapanColor2003WebCoated.icc
JapanWebCoated.icc
UncoatedFOGRA29.icc
USWebCoatedSWOP.icc
USWebUncoated.icc
WebCoatedFOGRA28.icc
WebCoatedSWOP2006Grade3.icc
WebCoatedSWOP2006Grade5.icc
Attention: Please read the Color Profile EULA.pdf in the cmyk directory

CompressBitmap (/CO=1) ⇧

Range of values [0,1]
This activates for tif- and png-output files a lossless run length type compression (like as zipped):
0: Writes uncompressed files.
1: Writes compressed files.

JPG_Quality (/JQ=100) ⇧

Range of values [1, 100]
Parameter to set the write quality of jpg images. For this conversion the color subsampling is always switched off.

SharpenIntensity (/IS=0) ⇧

Range of values [0, 100] percent
Intensity for post-sharpening using unsharp masking:
0: Post-sharpening deactivated.
unequal 0: Parameter to set the strongness of the post-sharpening.
The unsharp masking method usually knows three parameters: The intensity can be set here between 0% and 100%. The possibility to set even higher values (e.g. 200%) was deliberately omitted. The radius is calculated automatically depending on the resolution of the image size (e.g. 2 pixels for FullHD). In order not to increase image noise, dark areas are sharpened less or not at all.

Attention: If an image is heavily downscaled and with InterpolationMode = 5 a Gaussian filter is therefore activated, then the image is automatically sharpened with SharpenIntensity = 50%. If a manual value has been selected, this value is of course decisive.

IdentOutputDir (/ID=) ⇧

Range of values: characters
Using this parameter, the names of the resulting directories can be additionally modified.The text given here will be added to the standard names cl, cr, cs and ca. Example: With IdentOutputDir = '_1920' you get the directory names cl_1920, cr_1920, cs_1920 and ca_1920.

StampFiles (/ST=1) ⇧

Range of values [0,1,2,3]
Enables the integration of additional information into the filenames.
0: No change of the filename.
1: Integration of the deviation as a 3-digit promille value into the filenames.
2: Date and time using the format yyyy.mm.dd_hh.mm.ss are included into the filename. This enables the run of batchjobs with different parameter settings on the same images.
3: Both deviation and date /time are included into the file names.
Attention: The reference of the relative deviation is dependent on the parameters Projection Ratio. The deviation can refer to the image width or the width of the projected image.
If it refers to the width of the projection image, one of the following identifiers is added: "(4;3)" for an aspect ratio of 4:3, "(16;10)" for 16:10, "(16;9)" for 16:9 and "(Kino)" for 2048:1080.

DeviationThresh (/DT=999) ⇧

Range of values [0:999]
Deviation tolerance limit in per mille to sort out images with too much depth: Images with a higher deviation than given with DeviationThresh are created in the same directory, but with the prefix "zz_", so they land at the end of the image list.

CroppOutput (/CR=2) ⇧

Range of values [0-4]
Method for output cropping:
0: The output image remains completely unframed. Attention: In this case no proper stereo window is achieved.
1: The output image will be framed with a not necessarily rectangular frame. The passepartout shape is taken from the input image.
Hint: When deshaking videos, CroppOutput is forced to be 1, to enable the setting of (internal) mask information.
2: The output image will be rectangularly cut, the result is edgeless. Some image content and all passepartout information are lost. Image will be cropped to match proportions of OutputHeight and OutputWidth when both are set.
3: The output image will be rectangularly fit in a box OutputHeight times OutputWidth without loosing any image content. With portrait format images, the value OutputHeight is matched, for landscape format images, the value of OutputWidth is matched - in each case the other value will be smaller than set.
4: Cutting to a fixed output ratio, which is set by the parameters OutputHeight and OutputWidth, e.g. OutputHeight = 9 and OutputWidth = 16 give a ratio of OutputWidth/OutputHeight = 16:9.
5: Similar as 4 with always keeping the longer image edge, so OutputWidth/OutputHeight = 16/9 and OutputWidth/OutputHeight = 9/16 will produce the same result. Attention: With CroppOutput = 4 or 5, you may change the image ratio still afterwards, when opening the Cosima Viewer and using there the key commands 1/2/3/4/5 (in this order you get: 1:1, 5:4, 3:2, 16:9, 2:1). With the keys Shift 2/Shift 3/Shift 4/Shift 5 you get 4:5, 2:3, 9:16 and 1:2.

OutputHeight (/OH=0) ⇧

Range of values [0,16384] pixel
Wanted height of output image. If set not equal 0, the output images will be scaled to this pixel height. The output width will be scaled proportional.

OutputWidth (/OW=0) ⇧

Range of values [0,16384] pixel
Wanted width of output image. If set not equal 0, the output images will be scaled to this pixel width. The output height will be scaled proportional.

Both values = 0 (default): With images there is no scaling, due to the correction process the output resolution is some pixels less than the input resolution .
With videos the output resolution is identical to the input resolution (there is a little cropping with videos).

Only one of the two values is not equal to zero, the other one is zero: With images, there is a scaling to the given edge, due to the correction process, the height/aspect ratio is somewhat different to the input ratio.
With videos there is a scaling to the given edge, the height/aspect ratio is identical to the input ratio.

Both values set not equal to 0: Dependent on the value of the parameter CroppOutput, there is a scaling and cropping to the given values:
With CroppOutput = 2 an image or frame with the resolution OutputWidth x OutputHeight and maximum image content will be created.
With CroppOutput = 3, OutputHeight and OutputWidth will set the values for a box, into which the image/frame will fit.
With CroppOutput = 4, the values of OutputHeight and OutputWidth indicates only the ratio of the image height to the image width.

Attention: The maximum reachable image size is somewhat below 16384x16384 pixel and depends on the image correction values!

RotMirOutputLeft (/ML=0) ⇧

Range of values [0-7]
Rotates and mirrors the left output. With this parameter it is possible to rotate and mirror the left output image:
0: No input rotation/mirroring at all.
1: Rotation by +90 degree (no mirroring).
2: Rotation +180 degree (no mirroring).
3: Rotation +270 degree (no mirroring).
4: Horizontal mirroring (left-right).
5: Rotation +90 degree and horizontal mirroring.
6: Vertical mirroring (same as rotation +180 degree and horizontal mirror).
7: Rotation by +270 degree and horizontal mirroring.

RotMirOutputRight (/MR=0) ⇧

Range of values [0-7]
Rotates and mirrors the right output. With this parameter it is possible to rotate and mirror the right output image.
0: No input rotation/mirroring at all.
1: Rotation by +90 degree (no mirroring).
2: Rotation +180 degree (no mirroring).
3: Rotation +270 degree (no mirroring).
4: Horizontal mirroring (left-right).
5: Rotation +90 degree and horizontal mirroring.
6: Vertical mirroring (same as rotation +180 degree and horizontal mirror).
7: Rotation by +270 degree and horizontal mirroring.

Examples:

Halfmirror-separated TFTs (CoBox): RotMirOutputLeft = 0, RotMirOutputRight = 4; top-and bottom or View-Magic (upper = right): RotMirOutputLeft = 3, RotMirOutputRight = 3;

InterpolationMode (/IP=5) ⇧

Range of values [0-5]
Controls method of image interpolation:
0: There will be no interpolation, the nearest point is selected.
1: Two-dimensional linear interpolation.
2: Two-dimensional cubic interpolation, theoretically somewhat better than bilinear.
3: Gauss filter, use it specially for downscaling images (smoothing).
4: Automatic choose of the best method. If the ratio output-image-size/input-image-size is smaller than 2/3 (strong downsizing the images) the Gauss filter (InterpolationMode = 3) is used, otherwise the bicubic interpolation (InterpolationMode = 2) method is choosen.
5: As 4, but including additionally post sharpening, when for downscaling the Gauss filter is activ (SharpenIntensity = 40% is used for that, if not otherwise manually set).

GammaCorrection (/GA=1.0) ⇧

Range of values [0.0, 10.0]
Power factor for gamma correction. Interpolation methods always need a gamma correction in the interpolation area. Unfortunately this reduces the dynamic of color representation. Though, with photos a value of 1.0 (no correction at all) is recommended (according to Prof. Helmut Dersch) - with graphics 2.4 has been proven to be adequate.

InverseOutput (/IO=0) ⇧

Range of values [0,1]
Changes left and right output image. This Parameter makes only sense, if StereoOutput = 1 is set.
0: Writes sidecorrect mounted stereo images.
1: Writes inverse mounted stereo images (left/right exchanged). Should be set with "*.jps."-files.

FillHeight (/FH=0) ⇧

Range of values [0,16384] pixel
Filling with frame to a given the image height. If this value is unequal 0, a frame is added at the upper and lower border of the image with resulting an image height of exactly FillHeight pixel. In that case, the values of the parameters FrameTop and FrameBottom are ignored.

FillWidth (/FW=0) ⇧

Range of values [0,16384] pixel
Filling with frame to a given the image width. If this value is unequal 0, a frame is added at the left and right border of the image with resulting an image width of exactly FillWidth pixel. In that case, the value of the parameter FrameLeftRight is ignored. For the side-by-side mounted stereo image, the complete width is equal = 2*FillWidth + FrameCentral pixels.

FrameTop (/FT=0) ⇧

Range of values [0,1024] pixel
Frame broadness at the top. A frame with FrameTop pixel is added at the upper image border. (Is only evaluated for FillHeight = 0.)

FrameBottom (/FM=0) ⇧

Range of values [0,1024] pixel
Frame broadness at the bottom. A frame with FrameBottom pixel is added at the lower image border. (Is only evaluated for FillHeight = 0.)

FrameLeftRight (/FL=0) ⇧

Range of values [0,1024] pixel
Frame broadness left and right. A frame with FrameLeftRight pixel is added at the left and right image border. (Is only evaluated for FillWidth = 0.)

FrameCentral (/FC=0) ⇧

Range of values [-1024,1024] pixel
Frame broadness inbetween the two images. To the stereo output image (only StereoOutput) a central stripe can be added. FrameCentral is the width of this additional stripe in pixel.

FrameFloating (/FA=0) ⇧

Range of values [-1024,1024] pixel
Additional shift of an existing frame by the given pixel value. Either with FillWidth or with FrameLeftRight a frame must be created with at least a resulting width of FrameFloating/2 pixel. Examples:
FrameLeftRight = 20 and FrameFloating = 0 creates on the left and on the right hand side a frame with each 20 pixel.
FrameLeftRight = 20 and FrameFloating = -10 creates inwards a frame with 15 pixel and outwards with 25 pixel (shift of 10 pixel).
FrameLeftRight = 20 and FrameFloating = -20 creates inwards a frame with 10 pixel and outwards with 30 pixel (shift of 20 pixel).
FrameLeftRight = 20 and FrameFloating = -40 creates inwards a frame with 0 pixel and outwards with 40 pixel (shift of 40 pixel).
In all cases a floating and symmetric stereo window will be created in front of the screen.

AutoFrameFloating (/AR=0) ⇧

Range of values [0-6]
Selects the method for automatically setting floating stereo windows.
0: No floating stereo window are created.
1: The (unchanged) images are laterally shifted against each other by adding frames until the given far point disparity is reached.
This method is not available, if far point framing (EstimateWindow = 4/5) is choosen!
2: Similar to 1, but the image size is automatically reduced to the extent, that the sum from the frame width and the usable image content corresponds exactly to OutputWidth. The width of the usable image content allone is therefore less than OutputWidth.
This method is not available, if far point framing (EstimateWindow = 4/5) is choosen!
3: On the left and on the right edge, disturbing image parts are covered with curtains to avoid violations of the stereo window. The image content is varied, the image width remains constant. Stereo window violations at the bottom and top of the image will be assigned both to the left and right hand side of the image. The floating stereo windows are thus always symmetrical.
This method is not available, if near point framing (EstimateWindow = 2/3) is choosen!
4: Similar to 3, with the inner edges will be smoothed by a transparent gradient covering a range of about 1 percent of the image width.
This method is not available, if near point framing (EstimateWindow = 2/3) is choosen!
5: Similar to 3, but stereo window violations at the top and at the bottom will be ignored (for asymmetrically floating stereo windows).
6: Similar to 4, but stereo window violations at the top and at the bottom will be ignored (for asymmetrically floating stereo windows).
Attention: The far point disparity for AutoFrameFloating = 1,2 is to be given in the small box next to the selection window of AutoFrameFloating (Parameter FloatingTarget). For AutoFrameFloating = 3,4, an additional space between the near point and the curtains can be specified, again using the parameter Floating Target. The unit for the parameter FloatingTarget is specified with the parameter WinOffReference.

FloatingTarget (/FE=33) ⇧

Range of values [0,99]
Extended setting for parameter AutoFrameFloating.
AutoFrameFloating = 1,2: Farpoint disparity limit
AutoFrameFloating = 3,4: Extra disparity between nearpoint and curtain
Attention: The unit for the parameter FloatingTarget is specified with the parameter WinOffReference.

FrameColorRed (/FR=0) ⇧

Range of values [0,255]
Red color value of the frames and the central stripe.

FrameColorGreen (/FG=0) ⇧

Range of values [0,255]
Green color value of the frames and the central stripe.

FrameColorBlue (/FB=0) ⇧

Range of values [0,255]
Blue color value of the frame and the central stripe.

ThinBlackLine (/TB=0) ⇧

Range of values [0,255]
Number of pixels for the thickness of a black line between image and frames. Preferred for making of stereo cards (cs-image). The black lines are created with cl, cr, cs, ce-lrq, ce-ab, ce-abq and ce-mpo, but not with ca- and ce-eo images.