[3dem] Tomography, Multiposition Acquisition (Tomo5) - Defocus tracking fail

[Jonathan Bouvette]

Hi Julien,

I have made a mistake in my calculation before (my calculator was in
radians...). The predicted z offset is roughly sin(tiltangle)*distance from
the tilt axis (stage y). You may calculate based on your coordinates for
the tilt series you acquired. The other thing that may be happening is if
your holder is not eucentric throughout the tilt series but usually we
don't observe a very nice slope like yours in the defocus plot.

Best,
Jonathan

Le jeu. 7 nov. 2024 à 10:00, Maufront Julien <julien.maufront at curie.fr> a
écrit :

> Hi Jonathan,
>
> Thank you for your quick answer.
>
> Actually I already sent a message to TFS like a month ago... In principle
> it has been sent to "a specialist".
> But I am still waiting for feedback after repeated requests..
>
> You've been faster !
>
> Thank you and have a nice day.
>
> Best regards,
>
> *Julien Maufront*
> Research Engineer  –  Cryo-Electron Microscopy Platform
> ______
>
> *Institut Curie  – Research Center*
> *Physics of Cells and Cancer (UMR 168)*
> Molecular Microscopy of Membranes
> *+33 1 56 24 67 82*
>
>
>
> ------------------------------
> *De :* Jonathan Bouvette <jonathan.bouvette at gmail.com>
> *Envoyé :* jeudi 7 novembre 2024 15:08
> *À :* Maufront Julien <julien.maufront at curie.fr>
> *Cc :* 3dem at ncmir.ucsd.edu <3dem at ncmir.ucsd.edu>
> *Objet :* Re: [3dem] Tomography, Multiposition Acquisition (Tomo5) -
> Defocus tracking fail
>
> Hi Julien,
>
> This is something I observed in SerialEM before while doing off-axis
> tomography. You are right that the defocus compensation is inverted from
> that it should be. Tomo5 limits the image-shift distance to 5um, at 60
> degrees, without any correction, the defocus offset should be 1.5um and you
> see about double of that indicating that the correction is inverted. I
> would bring this up to TFS applications so that they can fix the
> configurations for you as I wouldn't know how to do that in Tomo5. I think
> you have enough data to support your case here.
>
> Best,
> Jonathan
>
> Le jeu. 7 nov. 2024 à 08:35, Maufront Julien <julien.maufront at curie.fr> a
> écrit :
>
>
>
> *Julien Maufront*
> Research Engineer  –  Cryo-Electron Microscopy Platform
> ______
>
> *Institut Curie  – Research Center*
> *Physics of Cells and Cancer (UMR 168)*
> Molecular Microscopy of Membranes
> *+33 1 56 24 67 82*
>
>
>
> ------------------------------
> *De :* Maufront Julien
> *Envoyé :* jeudi 7 novembre 2024 14:32
> *À :* 3dem at ncmir.ucsd.edu <3dem at ncmir.ucsd.edu>
> *Objet :* [3dem] Tomography, Multiposition Acquisition (Tomo5) - Defocus
> tracking fail
>
> Dear 3dem community,
>
> During an application session on our Glacios with a ThermoFisher Engineer
> we had the opportunity to try the multiposition acquisition scheme '("*Off-axis
> Multi-shot acquisition*") within *Tomo5*.
> The software worked well for *tracking *positions, even far from the tilt
> axis.
>
> But when I went to check the average *Defocus* of each tilt image among
> the * full tilt series*, the result was disappointing.
>  I observed a huge *drift from the Target defocus* while going to
> negative tilt angles or positive tilt angles.
> Here is an example of the measured defocus in regards of tilt angle values
> within a Tilt Serie being acquired at off-axis position:
>
>
> So it seems that the z displacement of the off-axis position is not well
> taken into account for target defocus compensation.
> Is it an expected behaviour ?
>
>
> On the other side I noted that the *tilt angles* registered (from -60 to
> +60) are inverted regarding the *global convention* (Cf:
> https://urldefense.com/v3/__https://wiki.dynamo.biozentrum.unibas.ch/w/index.php/Tilt_angle_convention__;!!Mih3wA!CxyOfV4yLyn6Yz23qX1HBVsVdedenaesHeWlZ_GT6E4gRBpirWIhxAn1cwJHXyl-q-VU3jIVvpNrbMiGirBbrUPtoenuyQ$ 
> <https://urldefense.com/v3/__https:/wiki.dynamo.biozentrum.unibas.ch/w/index.php/Tilt_angle_convention__;!!HLrAl2XzZ3iCLg!EtUHtX4pBonU0MJaEHxIHwMRqPioLC-0rGu3X0vcdN0BN8CDvbS0Hb6WbucrNiGuxtVYE2m2vpfeKNAEXhHNbn_aVdbfAahNks3bVg$>).
> My tilt image being tagged -60 correspond to +60 according to the
> convention and so on.
> Indeed when I tested the defocus from left side and right side of a tilt
> image it shown an inconsistent behaviour in regards with the convention.
>
> In the meantime, the* tilt axis angle* registered in the mdoc files show
> a suspicious negative sign: *-90.30*.
> If then we consider the tilt axis angle to be rather* +89.70* (inverted
> orientation of the tilt axis), the registered tilt angles become consistent.
>
> So here I am wondering if the disappointing defocus behaviour during
> off-axis tilt serie acquisition could be due to a wrong recorded
> orientation of the tilt axis.
> Does it make sense ? Does anyone here had already faced such situation ?
>
>
> Thank you for your help.
>
>
> *Julien Maufront*
> Research Engineer  –  Cryo-Electron Microscopy Platform
> ______
>
> *Institut Curie  – Research Center*
> *Physics of Cells and Cancer (UMR 168)*
> Molecular Microscopy of Membranes
> *+33 1 56 24 67 82*
>
>
>
> _______________________________________________
> 3dem mailing list
> 3dem at ncmir.ucsd.edu
> https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20241107/2282d8f3/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 87048 bytes
Desc: not available
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20241107/2282d8f3/attachment-0001.png>