Re: Upgrading to Windows 10 and new custom computer for confocal

Posted by George McNamara on
URL: http://confocal-microscopy-list.275.s1.nabble.com/Spinning-disk-or-new-point-scanning-confocals-for-live-imaging-of-3D-engineered-tissues-tp7590297p7590331.html

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hi Arvydas,

A couple of quick thoughts:

1. clone the current drive (probably an HDD) onto a same or larger
capacity SATA-6 SSD (i.e. Silicon Power), unplug the old drive, plug in
the new drive, upgrade that to Windows 10.

2. Can at your and your computer geek(s) leisure, build or buy a new
computer, pop the new Win10 SSD into that. If you go with a new PC with
an NVMe SSD as C: you can (try to) clone the SATA-6 SSD contents onto
that. New PC would also need to get whatever special cards Zeiss uses to
drive your LSM780 (and hopefully have win10 drivers).

Notes:

* in our (work ratio: my computer geek's 99% perspiration, my 1%
inspiration) experience, sometimes "takes" Win 10 a couple of tries to
install onto a drive.

* the ASUS Hyper M.2 PCIe card + (I suggest four 2Tb Silicon Power) NVMe
combination (~$1050) is not expected to be usable as a boot drive. Also
needs BIOS that supports x16 --> x4x4x4x4 bifurcation (5 year old PC
BIOS probably not, but worth testing). One NVMe SSD in the top position
of the Hyper M.2 card (top = away from PCIe slot) can be recognized in
conventional x16 slot (other NVMe's would not be seen), so $57 Hyper +
$244 (Silicon Power 2TB NVMe) could be useful in older PCs.

happy holidays,

George


On 12/26/2019 11:28 AM, Arvydas Matiukas wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> Post images on http://www.imgur.com and include the link in your posting.
> *****
>
> Hello List,
>
> The HP 820 workstation that controls LSM780 is over 5 years old and no longer supported under service contract, and Windows 7 support ends in January 2020. Please share your latest experience of:
> 1) upgrading/not upgrading to Windows 10
> 2) upgrading to a (cheaper) new custom built computer.
>
> One computer geek explained me that termination of Microscoft support for current OS is not a big deal - businesses still are using computers running on Windows XP and even 2000. It is a good and up-to-date antivirus software that provides the protection.
>
> Regarding computer upgrade I  looked up what imaging workstations HP is offering. HP Z2 Tower G4 Workstation (i5 8500 6 cores 3GHz/Nvidia Quadro P620 4GB DDR5/8GB DDR4/256GB PCIe SSD/DVD writer/500W power/supply/Windows 10 Pro 64  at $1099 seems like a real bargain and can be easily upgraded with 16GB RAM and 2TB HDD.
> To build custom computer from parts (especially high end ones as George McNamara suggested in earlier posts) would cost even more though it will run much faster as well that may be important if ,e.g, online deconvolution  is/will be performed.
>
> Any ideas, feedback, suggestions are appreciated.
>
>
> Merry Christmas (2nd day European Tradition) and Happy New Year,
> Arvydas
>
> +++++++++++++++++++++++++++++
> Arvydas Matiukas, Ph.D.
> Manager of NRB Shared  Research Equipment
> Director of Advanced Microscopy Core
> SUNY Upstate Medical University
> Neuroscience & Physiology Dept
>
> Email: [hidden email]
>
>
>>>> Francesco Pasqualini <[hidden email]> 12/16/2019 4:40 AM >>>
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.umn.edu_cgi-2Dbin_wa-3FA0-3Dconfocalmicroscopy&d=DwIBaQ&c=ogn2iPkgF7TkVSicOVBfKg&r=DtXPX1Vw9uh8rqlCEC9BTBr2oYBP4dEG1ecsgK6l-4k&m=5O_UqzQCsaTecA6gtZOvYZ0L387YferRuYhXvzpJlyA&s=AJjv8jedDfsaIj6mFFgKL-FINkuWS7WNz9z1rbZvgjU&e=
>
> Post images on https://urldefense.proofpoint.com/v2/url?u=http-3A__www.imgur.com&d=DwIBaQ&c=ogn2iPkgF7TkVSicOVBfKg&r=DtXPX1Vw9uh8rqlCEC9BTBr2oYBP4dEG1ecsgK6l-4k&m=5O_UqzQCsaTecA6gtZOvYZ0L387YferRuYhXvzpJlyA&s=mggAMhnKbnPe2ZdbN-8CjPUuE8RUMwNtFynTTmgt63Q&e=  and include the link in your posting.
>
> *****
>
> George,
> thanks for the note re: AausFP1. I will definitively check it out.
>
> In a spectrum that goes from (slowest speed / highest 3D SNR) to (highest
> speed / lowest 3D SNR), my understanding is that MPEF < Scanning confocal <
> Spinning confocal. That is, one has to trade 3D SNR for speed and I need to
> be able to hit 100s fps in certain important applications. Am I wrong? (I
> understand AasusFP1 increased brightness with help SNR and potentially
> enable faster recordings, but I am not sure how linear that
> relationship would be given photobleaching...)
>
> Another option would be (lattice) light-sheet microscopy but I have not
> found an implementation that works well on multi-well plates. The possible
> exceptions are the 3i implementations and Luxendo QuVi but they are pretty
> untested at this point. Does anyone have any experience with these
> platforms?
>
> Thanks,
> Francesco
>
>
> On Mon, Dec 16, 2019 at 3:57 AM George McNamara <[hidden email]>
> wrote:
>
>> Hi Francesco,
>>
>> since >50 um thick, think about multiphoton excitation fluorescence
>> microscopy ... second harmonic generation (some types of collagen).
>>
>> Fiber lasers that output single wavelength, typically ~100 femtoseconds,
>> 80 MHz repetition rate, are in the $10K - $60K range, each. If only one
>> microscope rig, implies point scanning (or TriMscope or similar). That
>> is, 1 or 3 lasers for potentially a lot less than a tunable Ti:Sapphire
>> laser (and could start with one fiber laser at most important wavelength
>> ... could be launched into a conventional 1p confocal scanner).
>>
>> You could have both point scanning (MPEF) and spinning disk confocal on
>> same microscope. I strongly urge investing in GPU deconvolution
>> software, optimized for each mode you will use.
>>
>> FYI - In May 2019 at JHU, Nathan Shaner (UCSD) started his talk on
>> AausFP1 -- 5x brighter than EGFP -- with a multi-day timelapse video
>> acquired at 6 min intervals (sped up on playback).
>>
>> I suggest starting to make AausFP1 constructs, amino acid sequence in
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.biorxiv.org_content_10.1101_677344v2&d=DwIBaQ&c=ogn2iPkgF7TkVSicOVBfKg&r=DtXPX1Vw9uh8rqlCEC9BTBr2oYBP4dEG1ecsgK6l-4k&m=5O_UqzQCsaTecA6gtZOvYZ0L387YferRuYhXvzpJlyA&s=_jzLWYEbanIKBXFqUswseorsYDNhxaZEGUfZvquDgyg&e=  (obligate dimer ... so
>> tandem dimer and/or monomerization mutation should be installed ... also
>> codon optimize). Preprint does not discuss Yellow version, should be
>> doable with one mutation (re: EGFP --> EYFP), so could have two colors.
>> Tandem dimer of G-->Y should be outstanding for FRET (3rd color).
>>
>> Also start thinking about optimized filter set(s) and/or confocal/MPEF
>> capabilities, to get the most of the narrow excitation and emission
>> peaks of AausFP1(green), AausFP1(Yellow), AausFP1(green-->Yellow).
>>
>> FYI - our Olympus FV3000RS confocal microscope worked well for GCaMP6,
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.physiology.org_doi_abs_10.1152_ajpcell.00351.2018&d=DwIBaQ&c=ogn2iPkgF7TkVSicOVBfKg&r=DtXPX1Vw9uh8rqlCEC9BTBr2oYBP4dEG1ecsgK6l-4k&m=5O_UqzQCsaTecA6gtZOvYZ0L387YferRuYhXvzpJlyA&s=ET9N1EbSlzPvMprIjdbbjKfvVSDG0Q9_gZmoGixrSlo&e=  (our Leica
>> SP8 confocal would have worked fine, but purchased without 37 C and
>> environmental control). Here in the U.S., the microscope vendors often
>> provide a nice trade-in credit, so I encourage asking your new place if
>> they have an old confocal to trade in toward purchase.
>>
>> best wishes,
>>
>> George
>>
>> p.s. see also current optical clearing methods for fixed specimens, and
>> expansion microscopy.
>>
>>
>> On 12/13/2019 12:18 PM, Francesco Pasqualini wrote:
>>> *****
>>> To join, leave or search the confocal microscopy listserv, go to:
>>> https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.umn.edu_cgi-2Dbin_wa-3FA0-3Dconfocalmicroscopy&d=DwIBaQ&c=ogn2iPkgF7TkVSicOVBfKg&r=DtXPX1Vw9uh8rqlCEC9BTBr2oYBP4dEG1ecsgK6l-4k&m=5O_UqzQCsaTecA6gtZOvYZ0L387YferRuYhXvzpJlyA&s=AJjv8jedDfsaIj6mFFgKL-FINkuWS7WNz9z1rbZvgjU&e=
>>> Post images on https://urldefense.proofpoint.com/v2/url?u=http-3A__www.imgur.com&d=DwIBaQ&c=ogn2iPkgF7TkVSicOVBfKg&r=DtXPX1Vw9uh8rqlCEC9BTBr2oYBP4dEG1ecsgK6l-4k&m=5O_UqzQCsaTecA6gtZOvYZ0L387YferRuYhXvzpJlyA&s=mggAMhnKbnPe2ZdbN-8CjPUuE8RUMwNtFynTTmgt63Q&e=  and include the link in your
>> posting.
>>> *****
>>>
>>> Hi,
>>> I will start my ERC-funded lab in Italy in 2020, which means I am
>> fortunate
>>> to be shopping for a confocal instrument right about now.
>>>
>>> Since I do mostly live experiments (more details below) I was going to
>> get
>>> a spinning disk system. But, I realized that the new scanning confocals
>> are
>>> also relatively fast and gentle. The question is how much (if at all)
>>> slower and harsher are they with respect to spinning disks?
>>>
>>> More details:
>>> - I do a lot of live experiments (traction force microscopy and
>>> voltage/calcium-sensitive dyes) on engineered cells (300x300 um FOV) and
>>> tissues (600x600 um FOV) in 2D or 3D (<300 um thick) that I complement
>> with
>>> immunostainings after fixation. Needed acquisition rates go from <1 fps
>> to
>>> 100s depending on the application.
>>>
>>> - Originally, I was oriented towards a spinning disk confocal
>>> (Yokogawa/SORA, Crest) and was looking at ways to deal with the issues of
>>> confocality on the 3D tissue (600x600x300 um volumes) case. But, I
>> realized
>>> all vendors have new point scanning confocal (980+Airyscan2, FV3000, A1R)
>>> with resonant scanners that acquire full frames/ROIs at 10s/100s of fps
>>>
>>> - Of course, structured illumination (Elyra-7, N-SIM) and light-sheet
>>> microscopes (QuVi) are also appealing but relatively untested in my
>>> applications of interests....
>>>
>>> The application specialists from all vendors in my area have been great
>> to
>>> work with but since my lab is not up and running, yet, I can't demo these
>>> systems directly. Therefore, I could use help and feedback, especially
>> from
>>> people who have had related experiences in the recent past.
>>>
>>> Thanks,
>>> Francesco
>>>
>