Hi,
Yes, ctffind3 crashes if the images size is smaller than half the window
size in ctffind3. You can run by decreasing the ctffind3 window size, or
alterlatively you may get the same results if you omit the windowing (i.e.
put a -1 in the window size).
HTH, S
> Hi Michael,
>
> Yes, I did mean the “Estimate CTF on window size".
>
> I left the "Box size" option at the tutorial default, 512 pix.
>
> Thanks,
>
> David
>
> On 07/01/2014 18:05, Saur, Michael wrote:
>> Hi David,
>>
>> just to be on the same page. You really mean the “Estimate CTF on
>> window size" option inside Relion and not the box size that is used by
>> ctffind3, don't you?
>>
>> Cheers,
>>
>> Michael
>>
>>
>>
>> ------------------------------------------------------------
>>     Dr. Michael Saur
>> Institute of Zoology, AG Markl
>> Müllerweg 6, Room 01-232
>>     55118 Mainz
>>     Germany
>>     Phone: +49 6131 - 39 23091
>> E-mail: [log in to unmask] <mailto:[log in to unmask]>
>>
>> ------------------------------------------------------------
>>
>> On 07. 01. 2014, at 18:09, David Gallagher <[log in to unmask]
>> <mailto:[log in to unmask]>> wrote:
>>
>>> Dear Sjors and other Relion users,
>>>
>>> We've been trying to run through the Relion tutorial using the
>>> clusters we have here at the MBU and have a run into a problem.
>>>  ctffind3.exe is raising a "segfault" with the default settings and
>>> the demo dataset.  All the parameters were initially left as default
>>> - this resulted in the below error.  When we decrease the CTF
>>> estimate window size from the default (950 pix) to 400 pix, then the
>>> job will run (it will not run using a window size of 560 pix).
>>>
>>> We see errors in /var/log/messages like:
>>>
>>> Jan  7 12:49:08 zeta kernel: ctffind3.exe[20596]: segfault at
>>> fffffffe0f7d68e0 rip 0000000000406294 rsp 00007fff5e07fa90 error 4
>>>
>>> This is on both CentOS 5.6 and CentOS 6.4, running on AMD x86_64
>>> hardware.  Total available RAM is 56 GB on the CentOS 5 system and 64
>>> GB on the CentOS 6.
>>>
>>> We're using the 130307 version of ctffind3.exe (i.e. the latest
>>> available) downloaded from http://grigoriefflab.janelia.org/ctf
>>>
>>> Do you have any idea what might be causing this problem and how we
>>> might be able to get around it?
>>>
>>>
>>> Many thanks,
>>>
>>> David
>>
>
>
> --
> *David Gallagher *
> PhD Student
> Bacterial Complex I Group
>
> MRC Mitochondrial Biology Unit
> Wellcome Trust / MRC Building
> Hills Road
> Cambridge
> CB2 0XY
>
> Email: [log in to unmask] <mailto:[log in to unmask]>
> Phone: 01223 252913
>


-- 
Sjors Scheres
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061
http://www2.mrc-lmb.cam.ac.uk/groups/scheres