Open graeme-winter opened 7 years ago
Final wall clock times attached
X is # frames Y is time to convert / frame
Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
cbf2nexus -o l-cyst_01.h5 -c zlib l-cyst_01_0*cbf
DATASET "data" {
DATATYPE H5T_STD_I32LE
DATASPACE SIMPLE { ( 1, 1679, 1475 ) / ( H5S_UNLIMITED, 1679, 1475 ) }
DATA {
(0,0,0): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,19): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,37): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0,
(0,0,55): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,
(0,0,73): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,91): 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,109): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,
(0,0,127): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,145): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,163): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,181): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,0,199): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
....
(0,1678,1270): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1287): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1304): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1321): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1338): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1355): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1372): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1389): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1406): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1423): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1440): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
(0,1678,1457): 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
(0,1678,1474): 0
}
ATTRIBUTE "signal" {
DATATYPE H5T_STD_I32LE
DATASPACE SCALAR
DATA {
(0): 1
}
}
}
minicbf2nexus appears to work as expected i.e. including more frames makes the h5 files larger. Does also appear to exhibit increasing time per frame for including extra data....
minicbf2nexus worked well (though did slow down in the compression from ~ 0.2 seconds / image early on to 2.4 seconds / image at the end. Resulting data file works correctly in DIALS.
For information - timing seems better behaved when using bslz4 compression => issue probably not in minicbf2nexus code per se.
Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter notifications@github.com wrote:
Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2 .
Dear Herbert
First hundred of these should work fine
https://sandbox.zenodo.org/record/51405/files/l-cyst_01.tar.gz
(data small - whole run here < 200MB)
Best wishes Graeme
On 26 Oct 2016, at 15:34, Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
Could you give me access to a few of those cbf's?
Regards, Herbert
On Wed, Oct 26, 2016 at 10:17 AM, Graeme Winter <graeme.winter@gmail.com mailto:graeme.winter@gmail.com> wrote: Dear Herbert
As follows:
/home/gw56/svn/cbflib_build/bin/cbf2nexus -c zlib -o test.h5 l-cyst_01_000*cbf
Best wishes
Graeme
On 26 Oct 2016, at 14:48, Herbert J. Bernstein <yayahjb@gmail.com mailto:yayahjb@gmail.com> wrote:
Dear Graeme,
Exactly how did you run it?
Regards, Herbert
On Wed, Oct 26, 2016 at 7:32 AM, Graeme Winter <graeme.winter@gmail.com mailto:graeme.winter@gmail.com> wrote: Dear Herbert
I appreciate this however the first index into this array ranges from 0 to 0 i.e. there is only one slice there.
Best wishes Graeme
On Wed, Oct 26, 2016 at 11:58 AM Herbert J. Bernstein <yayahjb@gmail.com mailto:yayahjb@gmail.com> wrote: Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter <notifications@github.com mailto:notifications@github.com> wrote: Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2.
Dear Herbert
I appreciate this however the first index into this array ranges from 0 to 0 i.e. there is only one slice there.
Best wishes Graeme
On Wed, Oct 26, 2016 at 11:58 AM Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter notifications@github.com wrote:
Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2 .
Dear Graeme,
Thank you. I'll look into it.
Regards, Herbert
On Wed, Oct 26, 2016 at 4:16 AM, Graeme Winter notifications@github.com wrote:
Using cbf2nexus to make an NXmx file from a few hundred CBF files I find that the initial file compression tasks are very fast:
Time to convert 'l-cyst_01_00005.cbf': 0.050s Time to read 'l-cyst_01_00006.cbf': 0.000s Time to convert 'l-cyst_01_00006.cbf': 0.070s Time to read 'l-cyst_01_00007.cbf': 0.000s Time to convert 'l-cyst_01_00007.cbf': 0.060s Time to read 'l-cyst_01_00008.cbf': 0.000s Time to convert 'l-cyst_01_00008.cbf': 0.070s Time to read 'l-cyst_01_00009.cbf': 0.000s Time to convert 'l-cyst_01_00009.cbf': 0.070s Time to read 'l-cyst_01_00010.cbf': 0.000s Time to convert 'l-cyst_01_00010.cbf': 0.050s Time to read 'l-cyst_01_00011.cbf': 0.010s Time to convert 'l-cyst_01_00011.cbf': 0.060s Time to read 'l-cyst_01_00012.cbf': 0.000s Time to convert 'l-cyst_01_00012.cbf': 0.080s Time to read 'l-cyst_01_00013.cbf': 0.000s Time to convert 'l-cyst_01_00013.cbf': 0.060s Time to read 'l-cyst_01_00014.cbf': 0.000s Time to convert 'l-cyst_01_00014.cbf': 0.070s Time to read 'l-cyst_01_00015.cbf': 0.000s Time to convert 'l-cyst_01_00015.cbf': 0.070s Time to read 'l-cyst_01_00016.cbf': 0.010s Time to convert 'l-cyst_01_00016.cbf': 0.060s Time to read 'l-cyst_01_00017.cbf': 0.000s Time to convert 'l-cyst_01_00017.cbf': 0.070s Time to read 'l-cyst_01_00018.cbf': 0.000s
However after a few hundred images these slow down a great deal:
Time to convert 'l-cyst_01_00686.cbf': 1.690s Time to read 'l-cyst_01_00687.cbf': 0.000s Time to convert 'l-cyst_01_00687.cbf': 1.690s Time to read 'l-cyst_01_00688.cbf': 0.010s Time to convert 'l-cyst_01_00688.cbf': 1.700s Time to read 'l-cyst_01_00689.cbf': 0.000s Time to convert 'l-cyst_01_00689.cbf': 1.690s Time to read 'l-cyst_01_00690.cbf': 0.010s Time to convert 'l-cyst_01_00690.cbf': 1.740s Time to read 'l-cyst_01_00691.cbf': 0.000s Time to convert 'l-cyst_01_00691.cbf': 1.720s Time to read 'l-cyst_01_00692.cbf': 0.000s Time to convert 'l-cyst_01_00692.cbf': 1.720s Time to read 'l-cyst_01_00693.cbf': 0.000s Time to convert 'l-cyst_01_00693.cbf': 1.720s Time to read 'l-cyst_01_00694.cbf': 0.000s Time to convert 'l-cyst_01_00694.cbf': 1.740s Time to read 'l-cyst_01_00695.cbf': 0.000s Time to convert 'l-cyst_01_00695.cbf': 1.750s Time to read 'l-cyst_01_00696.cbf': 0.000s Time to convert 'l-cyst_01_00696.cbf': 1.740s Time to read 'l-cyst_01_00697.cbf': 0.000s Time to convert 'l-cyst_01_00697.cbf': 1.750s Time to read 'l-cyst_01_00698.cbf': 0.010s Time to convert 'l-cyst_01_00698.cbf': 1.750s Time to read 'l-cyst_01_00699.cbf': 0.000s Time to convert 'l-cyst_01_00699.cbf': 1.760s Time to read 'l-cyst_01_00700.cbf': 0.000s Time to convert 'l-cyst_01_00700.cbf': 1.760s Time to read 'l-cyst_01_00701.cbf': 0.000s
Thought it may be worth checking in the code to see if there is some loop which scales as #frames squared or similar?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAbBktc8cMkdUl_cc3D6jamT9QfmAks5q3wx1gaJpZM4Kg4X2 .
Thanks.
On Wed, Oct 26, 2016 at 10:37 AM, Graeme Winter graeme.winter@gmail.com wrote:
Dear Herbert
First hundred of these should work fine
https://sandbox.zenodo.org/record/51405/files/l-cyst_01.tar.gz
(data small - whole run here < 200MB)
Best wishes Graeme
On 26 Oct 2016, at 15:34, Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
Could you give me access to a few of those cbf's?
Regards, Herbert
On Wed, Oct 26, 2016 at 10:17 AM, Graeme Winter graeme.winter@gmail.com wrote:
Dear Herbert
As follows:
/home/gw56/svn/cbflib_build/bin/cbf2nexus -c zlib -o test.h5 l-cyst_01_000*cbf
Best wishes
Graeme
On 26 Oct 2016, at 14:48, Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
Exactly how did you run it?
Regards, Herbert
On Wed, Oct 26, 2016 at 7:32 AM, Graeme Winter graeme.winter@gmail.com wrote:
Dear Herbert
I appreciate this however the first index into this array ranges from 0 to 0 i.e. there is only one slice there.
Best wishes Graeme
On Wed, Oct 26, 2016 at 11:58 AM Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter < notifications@github.com> wrote:
Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2 .
Dear Graeme,
Could you give me access to a few of those cbf's?
Regards, Herbert
On Wed, Oct 26, 2016 at 10:17 AM, Graeme Winter graeme.winter@gmail.com wrote:
Dear Herbert
As follows:
/home/gw56/svn/cbflib_build/bin/cbf2nexus -c zlib -o test.h5 l-cyst_01_000*cbf
Best wishes
Graeme
On 26 Oct 2016, at 14:48, Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
Exactly how did you run it?
Regards, Herbert
On Wed, Oct 26, 2016 at 7:32 AM, Graeme Winter graeme.winter@gmail.com wrote:
Dear Herbert
I appreciate this however the first index into this array ranges from 0 to 0 i.e. there is only one slice there.
Best wishes Graeme
On Wed, Oct 26, 2016 at 11:58 AM Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter <notifications@github.com
wrote:
Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2 .
Dear Graeme,
Exactly how did you run it?
Regards, Herbert
On Wed, Oct 26, 2016 at 7:32 AM, Graeme Winter graeme.winter@gmail.com wrote:
Dear Herbert
I appreciate this however the first index into this array ranges from 0 to 0 i.e. there is only one slice there.
Best wishes Graeme
On Wed, Oct 26, 2016 at 11:58 AM Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter notifications@github.com wrote:
Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2 .
Dear Herbert
As follows:
/home/gw56/svn/cbflib_build/bin/cbf2nexus -c zlib -o test.h5 l-cyst_01_000*cbf
Best wishes
Graeme
On 26 Oct 2016, at 14:48, Herbert J. Bernstein yayahjb@gmail.com wrote:
Dear Graeme,
Exactly how did you run it?
Regards, Herbert
On Wed, Oct 26, 2016 at 7:32 AM, Graeme Winter <graeme.winter@gmail.com mailto:graeme.winter@gmail.com> wrote: Dear Herbert
I appreciate this however the first index into this array ranges from 0 to 0 i.e. there is only one slice there.
Best wishes Graeme
On Wed, Oct 26, 2016 at 11:58 AM Herbert J. Bernstein <yayahjb@gmail.com mailto:yayahjb@gmail.com> wrote: Dear Graeme,
The way the NXmx structure works is that each image is a slice of a larger array.
Regards, Herbert
On Wed, Oct 26, 2016 at 5:45 AM, Graeme Winter <notifications@github.com mailto:notifications@github.com> wrote: Finally also interesting, it did not work right, final HDF5 file contains only one image... (which is more of a thing) - perhaps I did not use it right?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256299559, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAdjfho72__AHeAjPQOuwBuFe-EJrks5q3yE1gaJpZM4Kg4X2.
Dear Herbert,
Also with minicbf2nexus I don't think compression filters != zlib are working right:
[gw56@cs03r-sc-serv-16 INS1_1]$ du -hs *h5
112G INS1_1_bslz4.h5
8.4G INS1_1.h5
112G INS1_1_lz4-2.h5
from
/home/gw56/svn/cbflib_build/bin/minicbf2nexus -C config -c bslz4 -o INS1_1_bslz4.h5 INS1_1_*cbf
/home/gw56/svn/cbflib_build/bin/minicbf2nexus -C config -c 'lz4**2' -o INS1_1_lz4-2.h5 INS1_1_*cbf
i.e. zlib filter gives small data set (original 4800 images 28GB in CBF format - gzip compressed these are around 4GB) - 112 GB feels like raw uncompressed int32's - built using cmake on RHEL6... Not in the office for a few days but will take a look when I am back in.
Best wishes Graeme
That usually means the dynamic load of filters failed, I'll try to track it down.
On Thu, Oct 27, 2016 at 3:37 AM, Graeme Winter notifications@github.com wrote:
Dear Herbert,
Also with minicbf2nexus I don't think compression filters != zlib are working right:
[gw56@cs03r-sc-serv-16 INS1_1]$ du -hs *h5 112G INS1_1_bslz4.h5 8.4G INS1_1.h5 112G INS1_1_lz4-2.h5
from
/home/gw56/svn/cbflib_build/bin/minicbf2nexus -C config -c bslz4 -o INS1_1_bslz4.h5 INS11*cbf
/home/gw56/svn/cbflib_build/bin/minicbf2nexus -C config -c 'lz4*_2' -o INS1_1_lz4-2.h5 INS1_1__cbf
i.e. zlib filter gives small data set (original 4800 images 28GB in CBF format - gzip compressed these are around 4GB) - 112 GB feels like raw uncompressed int32's - built using cmake on RHEL6... Not in the office for a few days but will take a look when I am back in.
Best wishes Graeme
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/yayahjb/cbflib/issues/1#issuecomment-256569199, or mute the thread https://github.com/notifications/unsubscribe-auth/AEPiAfFdQDfOJzr3_pNnzp2UZS33XY6Dks5q4FS1gaJpZM4Kg4X2 .
The timing problem appears to be an inefficiently written axis search. Reworking it to use a hash table. -- HJB
Using cbf2nexus to make an NXmx file from a few hundred CBF files I find that the initial file compression tasks are very fast:
Time to convert 'l-cyst_01_00005.cbf': 0.050s Time to read 'l-cyst_01_00006.cbf': 0.000s Time to convert 'l-cyst_01_00006.cbf': 0.070s Time to read 'l-cyst_01_00007.cbf': 0.000s Time to convert 'l-cyst_01_00007.cbf': 0.060s Time to read 'l-cyst_01_00008.cbf': 0.000s Time to convert 'l-cyst_01_00008.cbf': 0.070s Time to read 'l-cyst_01_00009.cbf': 0.000s Time to convert 'l-cyst_01_00009.cbf': 0.070s Time to read 'l-cyst_01_00010.cbf': 0.000s Time to convert 'l-cyst_01_00010.cbf': 0.050s Time to read 'l-cyst_01_00011.cbf': 0.010s Time to convert 'l-cyst_01_00011.cbf': 0.060s Time to read 'l-cyst_01_00012.cbf': 0.000s Time to convert 'l-cyst_01_00012.cbf': 0.080s Time to read 'l-cyst_01_00013.cbf': 0.000s Time to convert 'l-cyst_01_00013.cbf': 0.060s Time to read 'l-cyst_01_00014.cbf': 0.000s Time to convert 'l-cyst_01_00014.cbf': 0.070s Time to read 'l-cyst_01_00015.cbf': 0.000s Time to convert 'l-cyst_01_00015.cbf': 0.070s Time to read 'l-cyst_01_00016.cbf': 0.010s Time to convert 'l-cyst_01_00016.cbf': 0.060s Time to read 'l-cyst_01_00017.cbf': 0.000s Time to convert 'l-cyst_01_00017.cbf': 0.070s Time to read 'l-cyst_01_00018.cbf': 0.000s
However after a few hundred images these slow down a great deal:
Time to convert 'l-cyst_01_00686.cbf': 1.690s Time to read 'l-cyst_01_00687.cbf': 0.000s Time to convert 'l-cyst_01_00687.cbf': 1.690s Time to read 'l-cyst_01_00688.cbf': 0.010s Time to convert 'l-cyst_01_00688.cbf': 1.700s Time to read 'l-cyst_01_00689.cbf': 0.000s Time to convert 'l-cyst_01_00689.cbf': 1.690s Time to read 'l-cyst_01_00690.cbf': 0.010s Time to convert 'l-cyst_01_00690.cbf': 1.740s Time to read 'l-cyst_01_00691.cbf': 0.000s Time to convert 'l-cyst_01_00691.cbf': 1.720s Time to read 'l-cyst_01_00692.cbf': 0.000s Time to convert 'l-cyst_01_00692.cbf': 1.720s Time to read 'l-cyst_01_00693.cbf': 0.000s Time to convert 'l-cyst_01_00693.cbf': 1.720s Time to read 'l-cyst_01_00694.cbf': 0.000s Time to convert 'l-cyst_01_00694.cbf': 1.740s Time to read 'l-cyst_01_00695.cbf': 0.000s Time to convert 'l-cyst_01_00695.cbf': 1.750s Time to read 'l-cyst_01_00696.cbf': 0.000s Time to convert 'l-cyst_01_00696.cbf': 1.740s Time to read 'l-cyst_01_00697.cbf': 0.000s Time to convert 'l-cyst_01_00697.cbf': 1.750s Time to read 'l-cyst_01_00698.cbf': 0.010s Time to convert 'l-cyst_01_00698.cbf': 1.750s Time to read 'l-cyst_01_00699.cbf': 0.000s Time to convert 'l-cyst_01_00699.cbf': 1.760s Time to read 'l-cyst_01_00700.cbf': 0.000s Time to convert 'l-cyst_01_00700.cbf': 1.760s Time to read 'l-cyst_01_00701.cbf': 0.000s
Thought it may be worth checking in the code to see if there is some loop which scales as #frames squared or similar?