update210926

README.md

@@ -36,35 +36,35 @@ The present version can also be compiled with Visual Studio 2008 and CUDA Toolki
 A test dataset in TIFF format is available from the binary folder. 
 
 <UL>
-  <LI>Build-210621. Frame loss detection routines were updated to deal with multiple frame loss. A bugfix in sample center calculation in the zoomed reconstruction. 
-  <LI>Build-201127. Updates for 'syncreadout' mode of the Hamamatsu camera and also for several other minor functional improvements. 
-  <LI>Build-191001. A bugfix in the CUDA memory freeing routine. 
+  <LI>Build-210621. Frame loss detection routines were updated to deal with multiple frame loss. A bugfix in sample center calculation in the zoomed reconstruction.</LI> 
+  <LI>Build-201127. Updates for 'syncreadout' mode of the Hamamatsu camera and also for several other minor functional improvements.</LI> 
+  <LI>Build-191001. A bugfix in the CUDA memory freeing routine.</LI> 
   <LI>Build-190729. Default memory usage in the Computing config dialog is now set to 40%. This minimizes memory paging and boosts the batch recon calculation. 
     <UL>
     <LI>Tau (Quadro P4000, 1792 cores, 1.48 GHz) = 0.161 nsec (1.22 sec for a 2048x2048 slice from 1800 projections)</LI>
-    </UL>
-  <LI>Build-190713. A bugfix in multithreading. Some improved GUI functions were implemented. CUDA stream is enabled by default. The source code can now be compiled also by using Visual Studio 2008 and CUDA Toolkit 5.5.  
-  <LI>Build-190530. A checkbox to enablie/disable CUDA stream pipelining was implemented in the Tomography-Computing config dialog.  
+    </UL></LI>
+  <LI>Build-190713. A bugfix in multithreading. Some improved GUI functions were implemented. CUDA stream is enabled by default. The source code can now be compiled also by using Visual Studio 2008 and CUDA Toolkit 5.5.</LI>  
+  <LI>Build-190530. A checkbox to enablie/disable CUDA stream pipelining was implemented in the Tomography-Computing config dialog.</LI>  
   <LI>Build-190118. CUDA routines were revised by using the 'stream' pipelining to parallelize the calculation. Performances are: 
     <UL>
     <LI>Tau (Quadro K5000, 1536 cores, 706 MHz) = 0.130 nsec (0.98 sec for a 2048x2048 slice from 1800 projections)</LI>
     <LI>Tau (Core i5-4670 (x64), 4 threads, 3.4 GHz) = 0.38 nsec (2.9 sec for a 2048x2048 tomogram from 1800 projections)</LI>
-    </UL>
-  <LI>Build-190111. On-the-fly reconstruction with which you can adjust the rotation axis in real time was implemented. If your PC has a sufficient computing power (over 1500 cores of GPU or 8 cores of AVX2-compatible CPU preferably Skylake or later), the axis position can be adjusted in real time using mouse wheel + shift key. Procedures for the data transfer between CPU and GPU were revised to accelerate the calculation.
-   <LI>Build-190102. We moved to Visual Studio 2017 and CUDA Toolkit 10.0 in this release. CUDA executable on the x86 platform became unavailable. ATI executable is not compiled at present. The x86/x64 reconstruction kernels were revised by using the AVX2 instruction set. CUDA routines were revised to be compatible with the new CUDA Toolkit. 
-  <LI>Build-181222. Axis-scan images can be generated by specifying different center values with the same slice position. An alternative robust algorithm for least-square 3D fitting is now implemented. A bugfix in multithreading. 
-  <LI>Build-180614. Bugfix of the pause/resume function in the queueing dialog.  
-  <LI>Build-180505. Reslicing routine was revised to minimize memory usage and to allow users to specify the reslicing direction with a vector. Polygon lasso function was implemented for defining ROI of statistical analysis.  
-  <LI>v5.4.1 (27 Jan 2018). Image analysis routines were updated. No revision were made in the offset CT kernel. Errors observed for some offset-CT data taken at SPring-8 were due to the 'conv.bat' files. Its workaround is described in the FAQ section below. Executable of this release is available only for x64 platforms because this is a very minor revision. 
-  <LI>v5.4.0 (3 May 2017). Image analysis routines were updated. The offset CT routine is now under revision. 
-  <LI>v5.3.0 (6 Jan 2017). Routines for <a href="https://github.com/mizutanilab/BluetoothDials">our dial input device</a> were implemented. 
-  <LI>v5.2.0 (13 Nov 2016). Frame selection dialog was implemented. Routines for the sinogram generation were revised.
+    </UL></LI>
+  <LI>Build-190111. On-the-fly reconstruction with which you can adjust the rotation axis in real time was implemented. If your PC has a sufficient computing power (over 1500 cores of GPU or 8 cores of AVX2-compatible CPU preferably Skylake or later), the axis position can be adjusted in real time using mouse wheel + shift key. Procedures for the data transfer between CPU and GPU were revised to accelerate the calculation.</LI>
+   <LI>Build-190102. We moved to Visual Studio 2017 and CUDA Toolkit 10.0 in this release. CUDA executable on the x86 platform became unavailable. ATI executable is not compiled at present. The x86/x64 reconstruction kernels were revised by using the AVX2 instruction set. CUDA routines were revised to be compatible with the new CUDA Toolkit. </LI>
+  <LI>Build-181222. Axis-scan images can be generated by specifying different center values with the same slice position. An alternative robust algorithm for least-square 3D fitting is now implemented. A bugfix in multithreading.</LI> 
+  <LI>Build-180614. Bugfix of the pause/resume function in the queueing dialog.</LI>  
+  <LI>Build-180505. Reslicing routine was revised to minimize memory usage and to allow users to specify the reslicing direction with a vector. Polygon lasso function was implemented for defining ROI of statistical analysis.</LI>  
+  <LI>v5.4.1 (27 Jan 2018). Image analysis routines were updated. No revision were made in the offset CT kernel. Errors observed for some offset-CT data taken at SPring-8 were due to the 'conv.bat' files. Its workaround is described in the FAQ section below. Executable of this release is available only for x64 platforms because this is a very minor revision.</LI> 
+  <LI>v5.4.0 (3 May 2017). Image analysis routines were updated. The offset CT routine is now under revision.</LI> 
+  <LI>v5.3.0 (6 Jan 2017). Routines for <a href="https://github.com/mizutanilab/BluetoothDials">our dial input device</a> were implemented.</LI> 
+  <LI>v5.2.0 (13 Nov 2016). Frame selection dialog was implemented. Routines for the sinogram generation were revised.</LI>
   <LI>v5.1.1 (released 18 Sep 2016). Reconstruction kernels for x86/x64 processors were updated using assembly code. The kernels are accelerated with vectorized instructions. Performances are:
     <UL>
     <LI>Tau (Quadro K4200, 1344 cores, 780 MHz) = 0.27 nsec (2.1 sec for a 2048x2048 tomogram from 1800 projections),</LI>
     <LI>Tau (Core i5-4670 (x64), 4 threads, 3.4 GHz) = 0.34 nsec (2.6 sec for a 2048x2048 tomogram from 1800 projections),</LI>
     <LI>Tau (Core i5-4670 (x86), 4 threads, 3.4 GHz) = 0.34 nsec (2.5 sec for a 2048x2048 tomogram from 1800 projections).</LI>
-    </UL>
+    </UL></LI>
   <LI>v5.1.0 (released 12 Sep 2016). OpenCL routines for ATI processors were updated. Performances are:
     <UL>
     <LI>Tau (Quadro K5000, 1536 cores, 706 MHz) = 0.23 nsec (1.75 sec for a 2048x2048 tomogram from 1800 projections),</LI>
@@ -74,12 +74,12 @@ A test dataset in TIFF format is available from the binary folder.
     <LI>Tau (Core i5-4670 (x86), 4 threads, 3.4 GHz) = 0.91 nsec (6.9 sec for a 2048x2048 tomogram from 1800 projections),</LI>
     <LI>Tau (Xeon E5-2609 (x64), 4 threads, 2.4 GHz) = 0.93 nsec (7.0 sec for a 2048x2048 tomogram from 1800 projections).</LI>
     </UL>  
-Here, tau is the time constant required for one pixel projection. For example, one tomogram of 2000 x 2000 pixels from 1800 projection frames can be reconstructed in tau x 2000 x 2000 x 1800 sec.  
-  <LI>v5.0.1 (released 5 Aug 2016). Reconstruction kernels were revised in order to improve sin/cos func precision in the reconstruction calculation. There would be no obvious difference in the appearance of reconstructed images, though pixel values are different from those of previous versions. The previous execulables are also still provided in the binary folder. Fourier domain plot for the resolution estimation can be generated from the "Analysis - Resolution plot" menu, without using the spread sheet software.
-  <LI>v4.9.0 (released 21 Jul 2016). APS-ANL data in HDF5 format are now supported. This page was moved here from our laboratory site.
-  <LI>v4.7.0 (released 13 Nov 2015). A dedicated routine for resolution estimation plot (J. Microsc. 261, 57-66, 2016) was implemented. CSV files now can be generated from 'Analysis==>Resolution plot' menu. ATI processors are not supported in this release.
-  <LI>v4.5.0 (released 6 Jan 2015). Update to support recent output.log format. Fourier transformations now can be generated from Tomography menu. A faster HIS-file reading routine was also implemented. A problem with the tilt angle direction of x64 reconstruction routine was fixed. User interfaces were updated.
-  <LI>v4.0.2 (released 22 Oct 2013). Back projection routines running on CUDA processors were updated to support the CUDA 5.5 computing environment. The CUDA version for x64 platforms was also included from this release. Queues from multiple RecView instances are now executed sequentially. 
+Here, tau is the time constant required for one pixel projection. For example, one tomogram of 2000 x 2000 pixels from 1800 projection frames can be reconstructed in tau x 2000 x 2000 x 1800 sec.</LI>  
+  <LI>v5.0.1 (released 5 Aug 2016). Reconstruction kernels were revised in order to improve sin/cos func precision in the reconstruction calculation. There would be no obvious difference in the appearance of reconstructed images, though pixel values are different from those of previous versions. The previous execulables are also still provided in the binary folder. Fourier domain plot for the resolution estimation can be generated from the "Analysis - Resolution plot" menu, without using the spread sheet software.</LI>
+  <LI>v4.9.0 (released 21 Jul 2016). APS-ANL data in HDF5 format are now supported. This page was moved here from our laboratory site.</LI>
+  <LI>v4.7.0 (released 13 Nov 2015). A dedicated routine for resolution estimation plot (J. Microsc. 261, 57-66, 2016) was implemented. CSV files now can be generated from 'Analysis==>Resolution plot' menu. ATI processors are not supported in this release.</LI>
+  <LI>v4.5.0 (released 6 Jan 2015). Update to support recent output.log format. Fourier transformations now can be generated from Tomography menu. A faster HIS-file reading routine was also implemented. A problem with the tilt angle direction of x64 reconstruction routine was fixed. User interfaces were updated.</LI>
+  <LI>v4.0.2 (released 22 Oct 2013). Back projection routines running on CUDA processors were updated to support the CUDA 5.5 computing environment. The CUDA version for x64 platforms was also included from this release. Queues from multiple RecView instances are now executed sequentially.</LI> 
   <LI>v3.0.0 (released 27 May 2010). Back projection routines running on Tesla, GeForce, Quadro processors were updated to support the CUDA 3.0 computing environment. The OpenCL 1.0 routines for ATI Radeon, FireStream or FirePro processors were also implemented. Performances are:
     <UL>
     <LI>Tau (448 cores, 607 MHz) = 0.0813 nsec (23.4 sec for a 8000x8000 tomogram from 4500 projections) by using GeForce GTX470 with 1.28 GB of GDDR5 memory,</LI>
@@ -88,16 +88,16 @@ Here, tau is the time constant required for one pixel projection. For example, o
     <LI>Tau (4 threads, 3.16 GHz) = 0.825 nsec (5.94 sec for a 2000x2000 tomogram from 1800 projections) by using Xeon E5410 (x86),</LI>
     <LI>Tau (4 threads, 3.16 GHz) = 0.808 nsec (5.82 sec for a 2000x2000 tomogram from 1800 projections) by using Xeon E5410 (x64).</LI>
     </UL>
-Here, tau is the time constant required for one pixel projection. For example, one tomogram of 2000 x 2000 pixels from 1800 projection frames can be reconstructed in tau x 2000 x 2000 x 1800 sec. Drift correction options were also provided in this revision.
+Here, tau is the time constant required for one pixel projection. For example, one tomogram of 2000 x 2000 pixels from 1800 projection frames can be reconstructed in tau x 2000 x 2000 x 1800 sec. Drift correction options were also provided in this revision.</LI>
   <LI>v2.0.5 (released 9 Feb 2009). Back-projection routines running on Tesla, GeForce or Quadro processors supporting the CUDA programming environment were implemented. The x86 and x64 assembler routines were also revised. Performances are:
     <UL>
     <LI>Tau (112 cores, 500 MHz) = 0.314 nsec (2.28 sec for a 2000x2000 tomogram from 1800 projections) using a Quadro FX 3700,</LI>
     <LI>Tau (3.16 GHz) = 0.633 nsec (4.56 sec for a 2000x2000 tomogram from 1800 projections) using Core2Duo (x86, 2 threads), or tau = 1.336 nsec (x86, single thread),</LI>
     <LI>Tau (3.16 GHz) = 0.802 nsec (5.77 sec for a 2000x2000 tomogram from 1800 projections) by using Xeon E5410 (x86, 4 threads), or tau = 1.724 nsec (x86, single thread),</LI>
     <LI>Tau (3.16 GHz) = 0.768 nsec (5.53 sec for a 2000x2000 tomogram from 1800 projections) by using Xeon E5410 (x64, 4 threads), or tau = 1.651 nsec (x64, single thread).</LI>
-    </UL>
-  <LI>v2.0.0 (released 21 Jan 2009). x64 platform is now supported. The x64 version can generate larger tomograms upto 10<sup>6</sup> x 10<sup>6</sup> pixels. Performance: tau(3.16 GHz) = 0.789 nsec using Xeon E5410 (x64, 4 threads). Performace of x86 version: tau(3.16 GHz) = 0.849 nsec using Core2Duo (x86, 2 threads), or tau = 1.395 nsec (x86, single thread).
-  <LI>v1.0.4 (released 1 Apr 2008). Multithreaded reconstruction routine was implemented. This function is partially written in x86 machine language, allowing faster execution of the reconstruction calculation.
+    </UL></LI>
+  <LI>v2.0.0 (released 21 Jan 2009). x64 platform is now supported. The x64 version can generate larger tomograms upto 10<sup>6</sup> x 10<sup>6</sup> pixels. Performance: tau(3.16 GHz) = 0.789 nsec using Xeon E5410 (x64, 4 threads). Performace of x86 version: tau(3.16 GHz) = 0.849 nsec using Core2Duo (x86, 2 threads), or tau = 1.395 nsec (x86, single thread).</LI>
+  <LI>v1.0.4 (released 1 Apr 2008). Multithreaded reconstruction routine was implemented. This function is partially written in x86 machine language, allowing faster execution of the reconstruction calculation.</LI>
   <LI>v1.0.0 (released 6 Mar 2008).</LI>
 </UL>