nci / drishti Goto Github PK

Completely new to 3D volumetric data presentation, so bear with me. Running current version of Drishti (v2.6.3) on Mac OS X (2.6 GHz Intel Core i5 processor, 16 GB 1600 MHz DDR3 memory, Intel Iris 1536 MB graphics). However, any time I click on a button in Drishti, the entire application crashes. Are there additional requirements for software I need to download? Or is this because I have not loaded anything into Drishti Import first? Also, what would be the recommended Texture Size? Many thanks, Waheed

If you add a clip plane and then apply any keyframe animation to the scene then even after you delete all keyframes you can no longer position or delete the existing clip plane using viewport interaction. Or at least no way I could find.

Hi, just wondering if you have any plans to support CentOS / Ubuntu etc. for this code?

Cheers,
MD

When using Drishti in the past, there was an occasion that the window size only showed up at the lower left quarter corner. Set it as the low resolution in Finder/Info resolved the issue. However, it came back again.
Don't know what to do to use the full window. Can anyone help?

Hello,

I am creating figures which distinguish cortical and trabecular bone in rodent femurs.
I have no problem importing 2 datasets, i.e. see figure 1.

Figure 1 (image is cropped to show inside cortical shaft)

When I add a third region which is cortical bone covering the rest of the trabecular bone, there initially doesn't seem to be a problem, when in lowres mode (figure 2).

Figure (2). Green is the trabecular region, orange is the cortical region (shown in figure 1) and yellow/brown is the other cortical region (third volume).

The problem comes when I toggle to the hires mode. I get this:

Do you have any idea what is going wrong? (could it be issues with my PC?)

Kind regards,

Jonathan Williams (PhD Student)

Hi Ajay,

I've been using the new version of Paint, and intermittently getting this error message when using the 3D preview window. I can't see a pattern in when it happens, but it requires restarting the program otherwise it just keeps popping up.

Many thanks,
Tom

When I import timeseries in DrishtiImport and then save them the created volume files contain the data of the first time step.

When attempting to MOP CARVE, I receive an "ERROR MOP" EmptySpaceSkipping not switched on.

I have the update on. Any ideas how this can be fixed?

Thanks

Import works with netCDF-3, but not netCDF-4
This is surmountable by converting netCFD4 back to netCDF3; e.g. via nccopy:
nccopy -k classic foo4c.nc foo3.nc

But, would be nice to be able to import netCFD4 files

I followed Robert Asher's YouTube tutorial, linked from the GitHub Drishti home page. He illustrates downloading and importing a Mongolian Gerbil from DigiMorph. When I tried this with version 2.6.6, I got multiple errors for each of the 1394 images:

TIFFReadDirectory Warning

C:/gerbil/8bitTIFF/merio0001.tif: Unknown field with tag 50838 (0xc696) encountered

Since the YouTube video was based on version 2.6.4, I tried that and was able to convert the images without incident. A google search suggests many other tools show these warnings. It would nice if the importer could be run in a quiet mode, or could combine all the warnings and show them in a single dialog window or log file, instead of confronting the user with thousands of sequential warnings.

The command for selecting points (shift + left click) while in Hires mode is not working. Instead of making a point this command changes the text in Hires on the bottom left from "step 1" to "step 2"

1 I have a problem with building with qmake on linux (it complains about backslashes in paths, I've replaced all the backslashed in paths with slashes but it still complains).
2 cmake is more cross-platform tool (I mean it has support of lot of compilers)

Hi Ajay,

I'm having issues overlaying two samples on top of each other. Reslicing to create roughly aligned samples isn't sufficient for our lab's purposes...any recommendations for getting the samples to lay exactly on top of each other?

Hello !

I wanted to tell you guys that everytime I exit Drishti Paint 2.6.2, there is a pop up window telling me there's an error to the save file.

Title : Save to file error
Message : Unable to save to file .qglviewer.xml: access refused [traducted from french Accès Refusé].

It is not affecting our work/curves/fibers saves, although, but I wanted to mention it :).

Thanks!

Can anyone help explain why transfer function goes into 1D when I load a volume?
Sometimes it stays in 2D.
But when its in 1D, there no way to change it back to 2D

Hi,
I'm trying to use 2.6 version but I have some issues. Drishti import only saves Filename.pvl.nc.00X files without the header file Filename.pvl.nc. In addition to that, I can't open older files since it gives me this error "not a valid preprocessed volume" and unexpectedly quits.
I'm using a MacBook Pro 15" Retina Late 2013 with OSX 10.10.4

Best Regards,
Marco

When loading more than one volume (I tried 2), the voxel size scaling entered in the volume information window/dialog for volume 2 still applies to volume 1, which is clearly visible when changing this to something "unusual" (say 10 1 1) and observing the change in the main window.
Tested using drishti on Windows 10, version 2.6.3

Hello,

First I am very pleased to see the VR version of Drishti being on Github. Thanks for that.
Saw your VR demo@Neubias.
Wanted to give it a try now, but get a shader error when starting the program: before switching the VR mode on.
It disallows to render any volume. When loading another volume the error is rethrown.
Error is: Error in fragment shader
ERROR: 0:66: '=' : cannot convert from 'highp float' to 'highp int'.
Closing this shows: drishtiVR: cannot create ee shader.

Attached the full error trace.
error.txt

Thanks again for the VR source code. The VR mode seems to work, but with this error, I am unable to load a volume. Hope you know how to fix this.

Kind Regards

Michiel Van Dyck

Using single volumes is not an issue, however, when trying to load multiple volumes, file names are grayed out and cannot be selected.

Our lab recently got a new Windows machine to handle µCT volume rendering and segmentation. Trying to use Drishti Import (2.6.5) I'm running into a problem I've never encountered before: the slices load into Import successfully, but "save as" only writes the .pvl.nc.001 data file, not the .pvl.nc header. As far as I can tell, the renderer can't read the data file without the header.
Tested with .tiff directory, .dicom directory and .img files, all return the same result.

Hello,

I am trying to segment blood vessels in a contrast enhanced CT scan. I'm still new to Drishti, so I just may not have watched the right tutorial yet, but it seems that the tools that are best at segmenting thin, winding structures (i.e., the fibers tool) can't work with branching structures. Is there a 'best tool' to use for this type of strucutre? Or a trick to potentially make it easier?

Thanks in advance!
Collin

I'm on Arch Linux and I'm trying to compile the recent version of drishti. However I encountered some problems.
I had to install the files /usr/include/ncvalues.h and /usr/include/netcdfcpp.h (Ifound them via google here: ftp://ftp.geomar.de/users/astrehz/include/ ). Also I had to install the packages netcdf-cxx and netcdf. And I had to generate thefollowing link:
ln /usr/lib/libnetcdf_c++4.so /usr/lib/libnetcdf_c++.so
But I still get some errors with undefined references:

networkobject.cpp:(.text+0xb98d): undefined reference to `NcError::~NcError()'
networkobject.cpp:(.text+0xbb5e): undefined reference to `NcVar::get(short*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xbcdc): undefined reference to `NcVar::get(signed char*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xbda2): undefined reference to `NcVar::get(int*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xbe66): undefined reference to `NcVar::get(signed char*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xbf1e): undefined reference to `NcVar::get(float*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xbfdb): undefined reference to `NcVar::get(int*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xc08c): undefined reference to `NcVar::get(float*, long, long, long, long, long) const'
networkobject.cpp:(.text+0xc112): undefined reference to `NcError::~NcError()'
networkobject.cpp:(.text+0xc129): undefined reference to `NcFile::~NcFile()'
networkobject.cpp:(.text+0xc210): undefined reference to `NcFile::get_var(char const*) const'
networkobject.cpp:(.text+0xc22b): undefined reference to `NcFile::get_var(char const*) const'
networkobject.cpp:(.text+0xc246): undefined reference to `NcFile::get_var(char const*) const'
collect2: error: ld returned 1 exit status
Makefile:926: recipe for target '../bin/drishti' failed
make: *** [../bin/drishti] Error 1

I guess the problem is that I need an older version of netcdf, which would explain the two missing files, but I'm not really sure. What are my next steps? I really need this program. If I can't get it to work I will have to use wine or windows.

Hi I was using drishtiimport in another computer, recently I changed it and when I try to use the program it does not work, When it supposed to generate the histogram, it turns off...
I tried with old versions but it happened the same..

Hi everybody!

I'm a new Master's Degree Candidate in a Paleontology Lab. We're studying fossils stuck in sandstone/schist.

We are slowly taking the CTScan methods to see what's inside our nodules, after a good mechanical prep. The CTScan used for my issue is the Bruker SkySkan 1173. The scans are looking good on CTAn and CTVox.

I just downloaded Drishti 1.0 on the lab computer and on mine to try it out after such a request from my director. So far, I imported a standard image directory (the .bmp reconstructions by NRecon of the scans), played a bit with contrasts, and converted the file into a .raw and then into a pvl.nc file.

When I try to load this pvl.nc file in Render or Prithvi (just to see if it loads), the program crashes (Drishti stopped working). It does that on these 2 systems -
#1 : Windows 7 i7-4770k computer with 16Gb RAM, GTX 770 GPU and 1tb hdd
#2 : Windows 10 i7-4790k computer with 16 Gb RAM and 2xGTX970 GPUs and 4Tb HDD.

I would like to know what is the error I am going through ^^ if you have any questions, if I missed something, feel free to ask! Thanks for the support.

Drishti is a powerful and easy-use software. I just wandering could it be possible to open raw rotation CT dataset (image series with rotation rather than sliced)?

Hello everyone,

When importing data from a .raw file the slices in Z and X direction look as expected, but the y-slices behave strangely (see attached images of 4 consecutive Y-slices). Every second slice looks as if the byte order is wrong. I used data from completely different CT-setups. The used .raw files are either 'uint16 little endian' or '32-bit float' format and I am using Drishti for Windows.
Despite this, I am able to create a .pvl.nc file and the rendered volume looks totally fine.
As I have no idea what the problem could be, I'd appreciate every contribution.

Regards,

Frank

P.S.: I just downloaded AMIDE to check if the slice problem also occurs there, but it looks normal.

After I start DrishtiVR (0.4), open a reasonable sized volume (1.4 GB) and click on VR mode, SteamVR starts but nothing happens.

Hi I am new to Drishti, when I load a file into the window the Transfer process editor looks different than your tutorials... I can't switch from 1 to 2 d and i dont have an adjustable spine as described in your videos.

My email is
[email protected]

Hi! I have the latest version of Dristhi but I cannot open the DICOM files with Dristhi import, the program closes Did you know what's happen?
Thank you!

Drishti 2.6.4 .. latest Mac version correct?
Catalina MacOS.
Get the following error when trying to save an imported volume as either jpeg files or dicom directory ...
"The save file operation failed. The save file operation failed to connect to the open and save panel service".

Hi,

I'm having an issue where when I take the *.mask.pvl.nc file I've made with Drishti Paint and try to load it into Drishti Render, it gives me a message "Some problem with pvl.nc files" and displays a volume that is garbled (it's different every time).

I've tried starting from scratch but it didn't help.

I'm using Drishti 2.6.5 on Windows.

Any help would be greatly appreciated!

• I download the Linux release and unzip it
• In ~/Downloads/drishti-2.3.2.linux/drishti I issue qmake && make.
• The make process does not finish, with the error message

boundingbox.cpp: In member function ‘void BoundingBox::draw()’:
boundingbox.cpp:293:23: error: ‘class MyManipulatedFrame’ has no member named ‘grabsMouse’
make: *** [boundingbox.o] Error 1

i want to insert some scalebars(10, 50, 100 micron scale).
i can't find it in Command Help.
please tell me how.

Hi,

I can't load my 3D volume on my laptop without getting the shader failed error below:

`-------------Error----------------

Fragment shader failed to compile with the following errors:
ERROR: 0:85: error(#132) Syntax error: ")" parse error
ERROR: error(#273) 1 compilation errors. No code generated

-----------Shader---------------

1 : #extension GL_ARB_texture_rectangle : enable
2 : uniform sampler2D lutTex;
3 : uniform sampler2DRect dragTex;
4 : uniform int gridx;
5 : uniform int gridy;
6 : uniform int gridz;
7 : uniform int ncols;
8 : uniform int llod;
9 : uniform int lgridx;
10 : uniform int lgridy;
11 : uniform int lgridz;
12 : uniform int lncols;
13 : uniform bool opshader;
14 : uniform float tfSet;
15 : void main(void)
16 : {
17 : vec2 tc = gl_TexCoord[0].xy;
18 : int lcol = int(tc.x)/lgridx;
19 : int lrow = int(tc.y)/lgridy;
20 : float x = tc.x - float(lcollgridx);
21 : float y = tc.y - float(lrowlgridy);
22 : float z = float(lrowlncols + lcol);
23 : vec3 pos = vec3(x,y,z);
24 : bvec3 pless = lessThan(pos, vec3(1.5,1.5,1.5));
25 : bvec3 pgret = greaterThan(pos, vec3(float(lgridx)-2.5,float(lgridy)-2.5,float(lgridz)-2.5));
26 : if (any(pless) || any(pgret))
27 : { gl_FragColor = vec4(0.0,0.0,0.0,0.0); return; }
28 : float xO = xfloat(llod);
29 : float yO = yfloat(llod);
30 : float zO = zfloat(llod);
31 : int alod = llod-1;
32 : vec4 fcolor = vec4(0.0,0.0,0.0,0.0);
33 : for(int xa=-alod; xa<=alod; xa++)
34 : for(int ya=-alod; ya<=alod; ya++)
35 : for(int za=-alod; za<=alod; za++)
36 : {
37 : float x = xO + xa;
38 : float y = yO + ya;
39 : float z = zO + za;
40 : int row = int(z)/ncols;
41 : int col = int(z) - row*ncols;
42 : row *= gridy;
43 : col = gridx;
44 : tc = vec2(float(col)+x, float(row)+y);
45 : vec2 vg;
46 : vg.x = texture2DRect(dragTex, tc.xy).x;
47 : vec3 sample1, sample2;
48 : float x1 = max(0.0, float(x-1.0));
49 : sample1.x = texture2DRect(dragTex, vec2(float(col)+x1, float(row)+y)).x;
50 : x1 = min(float(gridx-1), float(x+1.0));
51 : sample2.x = texture2DRect(dragTex, vec2(float(col)+x1, float(row)+y)).x;
52 : float y1 = max(0.0, float(y-1.0));
53 : sample1.y = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y1)).x;
54 : y1 = min(float(gridy-1), float(y+1.0));
55 : sample2.y = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y1)).x;
56 : int z1 = int(max(0.0, float(z-1.0)));
57 : row = z1/ncols;
58 : col = z1 - rowncols;
59 : row *= gridy;
60 : col = gridx;
61 : sample1.z = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y)).x;
62 : z1 = int(min(float(gridz-1), float(z+1.0)));
63 : row = z1/ncols;
64 : col = z1 - rowncols;
65 : row *= gridy;
66 : col = gridx;
67 : sample2.z = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y)).x;
68 : vg.y = distance(sample1, sample2);
69 : float val = vg.x;
70 : float grad = vg.y0.125;
71 : vg.y = tfSet + grad;
72 : if (opshader)
73 : {
74 : float alpha = texture2D(lutTex, vg.xy).a;
75 : fcolor += vec4(val, grad, alpha, 1.0);
76 : }
77 : else
78 : {
79 : vec4 clr = texture2D(lutTex, vg.xy);
80 : clr.rgb = clr.a;
81 : clr.a = step(0.01, clr.a);
82 : fcolor += clr;
83 : }
84 : }
85 : gl_FragColor = fcolor/pow(2.0(float)alod+1, 3.0);
86 : }
87 : `

I have been able to open the same volume on another PC. I'm running drishti 2.6.5 on a windows 10 laptop with integrated amd radeon r6 graphics card. I've updated the driver but I still get the same error.

Thanks,
Nathan

Dear Maintainer,

This is to notify you on behalf of the FSF Licensing team that licensing terms need be added to Drishti software for it to be included in FSD (Free Software Directory).

Thanks!

ResliceVolume seems to be missing from the clipping plane command options in V2.6.4 for Windows (10) (see screenshot).

Was following instructions on the instructional video at [https://www.youtube.com/watch?v=zW77274p3Yg], but there seems to be no option to realign.

Tried to find an older version (e.g. 2.5.1) that might have retained this function, but not sure where to download it from.

Hello,
i animated a little rotation movie but i am not able to safe it because the "safe movie" button is not clickable.
What could be the reason for that ?

greets
Stephan

The low res mode is working fine. However, when I transition to high res mode, the software cannot convert the low res images into high res.

When I switch to hires mode with a volume loaded, the background displays artefacts in the form of lines or an inverted version of the volume (depending on window size). It doesn't seem to matter what volume I load.

These artefacts change in appearance as I resize the window, and occasionally I get the normal background back again certain window sizes.

I've tried adjusting the texture memory size, and setting background colour, but neither of these help. If you had any suggestions that would be much appreciated.

I'm running El Capitain (10.11.3) on a non-retina iMac

Many thanks.

I have a dicom directory that opens fine on Drishti 2.6.4 on Mac .... but then issue #50 comes into play.
But Drishti 2.7 under Windows fails to read the dicom directory, crashes.
Ideas?

Hi,

I'm having trouble using the carve function.
The instructions on this page describe how to turn it on in low-res mode: http://www.scribd.com/doc/191007517/The-Basics-of-Drishti-A-Free-To-Download-Volume-Exploration-Presentation-Tool
I'm not sure if these instructions are for Drishti 2.6 though...

I can get it to work in high-res mode by pressing space bar then typing "mop carve", but it's slow, and if I switch back to low-res the carving I've done disappears.

Many thanks,
Tom.

Thanks for providing thorough Tutorials and Sample Dataset. However, the file block_flow.pvl.nc incorrectly reports

<voxeltype>unsigned short</voxeltype>

However, this should be

<voxeltype>unsigned char</voxeltype>

To match the provided image image block_flow.pvl.nc.001, as can be determined by both the file size and first byte of block_flow.pvl.nc.001 in order to fit the format described here and here.

Hi Ajay,

Is there anyway to slice processed volumes (pvl.nc files) into .dcm files?

I have microCT scans of embryos (DICOM directories) that have the stage/sample bed in the raw data.

I want to use the Pydpiper package (https://github.com/Mouse-Imaging-Centre/pydpiper) to automatically identify volume differences between different microCT scans but it requires:

1. the sample bed to be removed from these scans.
2. These files to be in MINC (.mnc) format.

Currently, I'm using Drishti's --> Data Ops --> Reslice to remove the stage from the volume and then DrishtiPaint to slice the image in .png files (all default settings/parameters).
From there, I can create a .tiff stack using FIJI (https://fiji.sc/), change the file extension to .tiff and then use C3D toolkit within a linux virtual machine (https://github.com/CobraLab/MINC-VM) to convert the image into .mnc.
These files become unreadable, however, when they are run through the Pydpiper software.

If I can directly save processed volumes as DICOM images, I can just use dcm2mnc (http://bic-mni.github.io/man-pages/man/dcm2mnc.html) to convert these files into .mnc files and ensure that the conversion is not changing the resolution when I'm doing the analysis.

Kind Regards and thank you for reading this
Kyle Drover
(PhD candidate - Arkell group)

Hi I am trying to import a dental dicom CBCT into Drishti and it says that no dicom data can be found? I am running a Mac Air OS 10.9.5 1.7 Ghz intel core i7 with 8 GB 1600 mHz DDR3. I am sure there is dicom data in the containers and then the application crashes?... Please help I need to use your app for a presentation asap...

Thank you
Amarjit Rihal
[email protected]

Hi,
I'm new to 3D modelling and working on my master's thesis with this. I'm trying to calculate the volumetric information for my rootsystem-model, which is displayed well in Drishti. Whenever I try to measure the volume or the surface area, the program instantly crashes and the window is closed. Otherwise, it works fine. I'm currently using v.2.6.5 and tryed the same with v.2.6.4 as well.
Might there be a a memory problem? Or is there another way to obtain the volumetric information besides 'Functions>Measurements>Volume'?
Any help would be welcome

Hi,
Each time I try to import the paint tags from Paint to Drishti Renderer the tags do not appear with seperate segmented colours or seperate transfer functions.

Any ideas?
Thanks,
Cat.

Recently I saw that drishti 2.6.5 has been released. When I try to load a volume dataset it gives me an error and closes. I've been using 2.6.4 so far, and never had this issue with that version.

Steps to reproduce:

1. open drishti
2. file / Load Volume / Load 1 Volume
3. select a drishti volume file and press "open".

Here's the contents of the error message:

-------------Error----------------

WARNING: 0:37: 'assign' : implicit type conversion allowed from GLSL 1.20 
WARNING: 0:38: 'assign' : implicit type conversion allowed from GLSL 1.20 
WARNING: 0:39: 'assign' : implicit type conversion allowed from GLSL 1.20 
ERROR: 0:85: ')' : syntax error syntax error


-----------Shader---------------

1 : #extension GL_ARB_texture_rectangle : enable
2 : uniform sampler2D lutTex;
3 : uniform sampler2DRect dragTex;
4 : uniform int gridx;
5 : uniform int gridy;
6 : uniform int gridz;
7 : uniform int ncols;
8 : uniform int llod;
9 : uniform int lgridx;
10 : uniform int lgridy;
11 : uniform int lgridz;
12 : uniform int lncols;
13 : uniform bool opshader;
14 : uniform float tfSet;
15 : void main(void)
16 : {
17 :   vec2 tc = gl_TexCoord[0].xy;
18 :   int lcol = int(tc.x)/lgridx;
19 :   int lrow = int(tc.y)/lgridy;
20 :   float x = tc.x - float(lcol*lgridx);
21 :   float y = tc.y - float(lrow*lgridy);
22 :   float z = float(lrow*lncols + lcol);
23 :   vec3 pos = vec3(x,y,z);
24 :   bvec3 pless = lessThan(pos, vec3(1.5,1.5,1.5));
25 :   bvec3 pgret = greaterThan(pos, vec3(float(lgridx)-2.5,float(lgridy)-2.5,float(lgridz)-2.5));
26 :   if (any(pless) || any(pgret)) 
27 :     { gl_FragColor = vec4(0.0,0.0,0.0,0.0); return; }
28 :   float xO = x*float(llod);
29 :   float yO = y*float(llod);
30 :   float zO = z*float(llod);
31 : int alod = llod-1;
32 : vec4 fcolor = vec4(0.0,0.0,0.0,0.0);
33 : for(int xa=-alod; xa<=alod; xa++)
34 : for(int ya=-alod; ya<=alod; ya++)
35 : for(int za=-alod; za<=alod; za++)
36 : {
37 :   float x = xO + xa;
38 :   float y = yO + ya;
39 :   float z = zO + za;
40 :   int row = int(z)/ncols;
41 :   int col = int(z) - row*ncols;
42 :   row *= gridy;
43 :   col *= gridx;
44 :   tc = vec2(float(col)+x, float(row)+y);
45 :   vec2 vg;
46 :   vg.x = texture2DRect(dragTex, tc.xy).x;
47 :   vec3 sample1, sample2;
48 :   float x1 = max(0.0, float(x-1.0));
49 :   sample1.x = texture2DRect(dragTex, vec2(float(col)+x1, float(row)+y)).x;
50 :   x1 = min(float(gridx-1), float(x+1.0));
51 :   sample2.x = texture2DRect(dragTex, vec2(float(col)+x1, float(row)+y)).x;
52 :   float y1 = max(0.0, float(y-1.0));
53 :   sample1.y = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y1)).x;
54 :   y1 = min(float(gridy-1), float(y+1.0));
55 :   sample2.y = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y1)).x;
56 :   int z1 = int(max(0.0, float(z-1.0)));
57 :   row = z1/ncols;
58 :   col = z1 - row*ncols;
59 :   row *= gridy;
60 :   col *= gridx;
61 :   sample1.z = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y)).x;
62 :   z1 = int(min(float(gridz-1), float(z+1.0)));
63 :   row = z1/ncols;
64 :   col = z1 - row*ncols;
65 :   row *= gridy;
66 :   col *= gridx;
67 :   sample2.z = texture2DRect(dragTex, vec2(float(col)+x, float(row)+y)).x;
68 :   vg.y = distance(sample1, sample2);
69 : float val = vg.x;
70 : float grad = vg.y*0.125;
71 : vg.y = tfSet + grad;
72 :  if (opshader)
73 :    {
74 :      float alpha = texture2D(lutTex, vg.xy).a;
75 :      fcolor += vec4(val, grad, alpha, 1.0);
76 :    }
77 :  else
78 :    {
79 :      vec4 clr = texture2D(lutTex, vg.xy);
80 :      clr.rgb *= clr.a;
81 :      clr.a = step(0.01, clr.a);
82 :      fcolor += clr;
83 :    }
84 : }
85 :  gl_FragColor = fcolor/pow(2.0*(float)alod+1, 3.0);
86 : }
87 : 

The file was produced with the import tool that is distributed with 2.6.5. I verified that the same file loads without problems in 2.6.4.

I'm running windows 7 on a laptop using on board graphics (Intel HD graphics 5500).

If I can do anything else to help resolving this issuse I'd be happy to do so.

I had a session of Drishti lock up, the main display blanking out and the app no longer responsive to input. (I suspect this was because the graphics configuration was invoking an underpowered graphics card rather than the NVIDIA GeForce GTX 1080 that was also installed.)

Unfortunately, when I resorted to shutting down the process from the Task Manager, it seems to have left things in an unstable state. Trying to launch again, I get a series of popups, "Invalid imageBuffer" and "Invalid lowresBuffer" alternating. The main screen opens, then one more pair of warnings, and it shuts down.

Is there any cleanup I can do, such as with regedit, to clear this? This is Windows 10.

Hi Ajay,
Thank you very much for Drishti.
In 2.6.6, the Scalebar dialog asks for "Number of voxels". Here (attached) I've put 1000. The Scalebar returns with 1000 um. If the voxel size does not = 1 um, either the scalebar length displayed or the units displayed cannot be correct.
Should the dialog instead ask "Number of voxel units?"
For example, this volume has a voxel size of 3.24 um. I would like to place a scalebar of 1 mm, but I'm not sure if the length is correct, or the units (or neither?), or maybe I'm simply not understanding something correctly.
Thank you,
Matt