Hi Ed,
unwarping is indeed very important. Most (CCD) detectors do it before
the data leave the detector housing, that's why it is usually unnoticed.
If you only want to get cell parameters, I'd pull some lines in the adxv
display. I find it very helpful to actually see what you are doing and
often use its help for tricky cases.
Regards,
Tim
On 09/29/2015 11:29 PM, Edward A. Berry wrote:
> Thanks to all who answered, on and offline.
> And apologies for troubling the board with beginner questions.
> I will see if I can get things working now, and if not take
> Harry Powell's kind offer and send some images for expert help.
>
> One clarificatin- "imgtomar" is a misnomer for that xcalibur command-
> with the "img" argument it is saving the images in native oxford format.
> But it does some unbending, or warping they call it, I suppose to correct
> for imperfection in the optical fiber transfer to ccd. Maybe also
> correcting some nonlinearity in the response or something like
> the "corner correction" needed for some quantum detectors.
> Anyway, the oxford people felt this was important for getting
> good results from their machine with mosflm. Probably not so
> important for getting approximate cell parameters which is
> what I want now.
> eab
>
> On 09/29/2015 05:04 PM, Harry Powell wrote:
>> Hi Ed
>>
>> (*) You can edit the phi angles in the “Images” pane - left-click
>> twice (i.e. click-pause-click, *not* double-click) on the values you
>> want to change - the field changes colour and you can enter the values
>> you want. They will be propagated through all subsequent images in the
>> sector.
>>
>> This is documented in the online tutorial (immediately before Section
>> 4 - Image Display.
>>
>> (*) I’d need to see the images to work out why the spot finding is not
>> working as expected - can you send me a couple off-line?
>>
>> (*) If you go to the “History” pane, then select the “Log” tab at the
>> bottom, a scrollable light-green on dark-green display is visible, and
>> you can search through this for the reasons why the spots have been
>> rejected.
>>
>> (*) If you use the “three images” icon, you select the images (and
>> find spots on them) by double-clicking the circular icon next to the
>> image - BUT it’s much more convenient to type the image number(s) in
>> the entry box with the automatically chosen images.
>>
>> (*) You shouldn’t need to convert the Excalibur images to Mar format -
>> unless Oxford (Agilent, Rigaku…) have changed the format they write
>> their native images in. We used the Mar converted format for about a
>> month or so when the detectors were first marketed (in 2003 IIFRC),
>> but Mathias Meyer spent a couple of days with me then to make sure we
>> could read the native images properly.
>>
>> (*) The image display can be a little confusing - it uses the internal
>> Mosflm frame of reference, which may not be what you see when looking
>> at the physical machine - but that shouldn’t be a problem, because we
>> take that into account when doing the calculations.
>>
>> HTH
>>
>> Harry
>> --
>> Dr Harry Powell, MRC Laboratory of Molecular Biology, Francis Crick
>> Avenue, Cambridge Biomedical Campus, Cambridge CB2 0QH
>> Chairman of International Union of Crystallography Commission on
>> Crystallographic Computing
>> Chairman of European Crystallographic Association SIG9
>> (Crystallographic Computing)
>>
>>> On 29 Sep 2015, at 21:39, Edward A. Berry <[log in to unmask]> wrote:
>>>
>>> I'm trying to index some weak data with mosflm, mainly just to get
>>> cell parameters.
>>> The automatic procedure fails, so I'm trying to modify things
>>> manually, and having trouble.
>>>
>>> I run imosflm (7.2.1 on Fedora 14 ( is that a problem?))
>>> File menu: load images, select the first image, OK, they all 10 get
>>> listed in the window
>>> (the first image has wrong phi angles in header, I can't find how to
>>> enter phi angles manually or exclude it so rename and start again
>>> without it)
>>> click index in the left toolbar. first and last are selected for spot
>>> picking but autoindexing fails.
>>> Looking at spots (picked) many don't correspond to visible spots and
>>> the strongest spots are not picked.
>>> But quite a few spots are correctly picked, so it is interpreting the
>>> image (not a byte-swap problem)
>>>
>>> I see how to adjust the resolution (in the graphics window) and
>>> threshold etc in the main window.
>>> Change these and click the find spots icon in the appropriate image
>>> (one of the two already selected)
>>> and the number of spots changes in the expected direction (at least
>>> for threshold and spot separation)
>>> But many spots are not picked and I don't know why. There used to be
>>> a list of which spots were rejected for what reason. Is that still
>>> available?
>>>
>>> So I want to add more images besides the two initial frames. I click
>>> the 3-images icon in main window and all frames are displayed, with
>>> check boxes. But they cannot be checked! OK, because spots have not
>>> been picked for those images. But when I click the find-spots icons
>>> in those lines, nothing happens! All the frames can be read, because
>>> I can step through them in the image window.
>>>
>>> I'm sure I'm missing some pretty obvious things, but I need help to
>>> go on.
>>>
>>> The images are from oxford xcalibur, exported with dc imgtomar img 50 1
>>> in mosfilm I would give "detector oxford" But I assume that gets
>>> detected automagically now.
>>> For example does mosflm know the phi axis is vertical as the image is
>>> displayed in the image window?
>>> But first I need to get spot picking working.
>>>
>>>
>>> Ed
>>
--
--
Paul Scherrer Institut
Tim Gruene
- persoenlich -
OFLC/107
CH-5232 Villigen PSI
phone: +41 (0)56 310 5754
GPG Key ID = A46BEE1A
|