CTFTILT seg-faults on certain 3710x3838 size images

Submitted by dlyumkis on Sat, 2015-09-12 18:22

Forums

ctffind & ctftilt

Hi,

I am getting a seg-fault on certain micrographs from a series of tilted images collected on K2. All micrographs are 3710x3838 from K2 and tilted. Certain micrographs work, however others seg-fault. I have not been able to systematically track down why some work and others do not. Image mins, maxs, means, and stdevs are all similar. Nominal defocus values are standard (1-2 microns). Tilt angles are all 30° for this set.

I am using the latest version of CTFTilt.

For the images that did NOT work: I did a prime number decomposition for the image size (3710x3838), which provides ([2,5,7,53] and [2,19,101]). I realize this isn’t ideal for FFTs. I clipped the micrograph to 3072x3072 (prime numbers [2,3] only, 2**10*3). This fixes the issue.

While I can clip the images for the purpose of CTF estimation, this isn't necessarily ideal. Any ideas of how to best proceed?

Dmitry

dlyumkis Sat, 2015-09-19 15:49

After trouble-shooting this for a bit, it seems that setting the field size to 256 (instead of 512, which is what we've been using) solves this problem. CTFTILT now works on all micrographs and does not seg-fault. A field size of 512 still seg-faults on some images.