Ctf estimation with phase plate images

Submitted by vkumar on Tue, 2016-05-31 04:52

Forums

ctffind & ctftilt

Hi,
We recently collected some phase plate images with slight defocus and in the process of estimating the CTF.

What sort of amplitude contrast does one uses with these images and is it fine to use standard CTFFIND settings (meaning the default kV, ac but changing the range of focus) with the phase shift estimation set to yes.

Has the estimation of CTF from movie frames (as implemented in CTFFIND4 new version) been tried with phase plate images.

thanks

Vinoth

Alexis Tue, 2016-05-31 09:20

Hi Vinoth,

Even though arithmetically the amplitude contrast factor and the phase shift from the plate have the same effect, I like keeping those two things separate conceptually. Indeed with a Volta phase plate, the phase shift due to the plate will be variable over time, while the amplitude contrast (due to aperture contrast and/or energy filtering contrast) will be constant to a very good first approximation. Therefore, for cryo my standard recommendation is to keep amplitude contrast constant at, say, 7%, and to let the phase plate phase shift vary.

I don't have much experience processing VPP data yet. The only dataset I really played with was recorded near focus, such that essentially no Thon rings were visible. In this case, I did use the option to estimate CTF from the frames, but it was difficult to judge whether this had "worked"/helped. It certainly gave sensible-looking results, which let the final reconstruction go to a good resolution, so I suspect nothing major went wrong.

Perhaps other ctffind users will comment with their experience with VPP data.

Cheers,
Alexis