Pointy Hat

[Peter Ward]

This one is from the vault.



I had a massive PC failure. Major “blue screen of death” event.



Lost 9Terabytes of data due a failing motherboard taking out 3 hard drives. I think I said “Golly! Gosh!” or words to that effect at the time.



Took me a week to recover the data…but in the process I un-earthed some long forgotten images that I had entered into the CWAS/Malins that did not make the cut…..like this one from 2011 !



My fancy new dual booting i7 iMac (Mac and Win10x64) simply rips through various Photoshop and PixInsight routines. Hence I gave the data a contemporary make-over.



The link is here (http://www.atscope.com.au/BRO/gallery495.html)

[peter_4059]

I think Thor's helmet is my favorite object and this is a great rendition.

Sorry to hear about the pc but regarding the new imac...to quote Stutodd...

Cinebench R20, or it didn't happen...

[Peter Ward]

Ta. The Mac came in with a Cinebench mark of 2400, my old PC clocked in at 607.

That said, the top Cinebench marks seem to be circa 20,000.

Might be time to build a new PC based on an i9 chip ;)

P.S. There is some irony to the linked image and my old PC both being 2011 vintage.
One wonders how powerful PC's will be in another decade or so :question:

[peter_4059]

Not sure about Photoshop however Pixinsight benefits a lot from multi threads. Bang for buck, it is hard to ignore AMD Ryzen 9 processors. I went for the 12 core/24 thread R9 3900X but there is also the R9 with 16 cores/32 threads for around $1200. That's a similar price point to the i9 10900 with 10 cores/20 threads.

With the R9 3900X my Cinebench mark was 7252:
http://www.iceinspace.com.au/forum/showpost.php?p=1479246&postcount=12

The other important factors (at least for PI) are swap file speed, RAM and operating system. M.2 NVMe SSD's use the PCIe interface are therefore much faster than SATA SSD's and I've got two just for swap file on my new build. I went for PCIe gen 3 for cost reasons but you can get gen 4 now that boast double the bandwidth of gen 3.

Pixinsight also has an internal benchmark script. Results here (takes a while for the page to load) however this shows the system build info:

https://pixinsight.com/benchmark/index.php?sort=total&os=windows

[Placidus]

Data loss: Pain and horror! Glad that you were able to bring it back from the dead.


Thor: You are the master of the sharp image! Lovely to see.



(And lovely to see a nice summer object on such a freezing morning. It was down to 1.5C here when I went up to the observatory to collect last night's data).

[Andy01]

Hmmm - Thanks for posting this blast from the past. :)

The Ha & O3 detail is incredible - especially the "brain" bubble of the WR star - I can't recall seeing that much detail before. :question:
That said, the stars are very average by today's standards - no definition, blown out and circled with coloured wide outer haloes.
If you have an RGB set it would be a great exercise to swap out these NB ones.

The pursuit of excellence is one of the core attractions of competitions like the Malins, IAPOTY etc. and you've deservedly been at the forefront every year. :thumbsup:
Interesting to see how far even a highly skilled, awarded and experienced imager like yourself has come over the past decade. ;)

[Peter Ward]

Thanks Andy.

While I have had some success at the Malin's ...I never take it for granted as so many guys in OZ are churning out excellent images these days.

I must admit I am a big fan of David Malin's take on Astrophotography.
While the occasional NB image has got up from time to time, he is terse judge on accurate colour...as it says so much about the physical processes going on within an object.
Fiddle with that and you are unlikely to place well.

My current Alluna 'scope is quite a step up from the RCOS I was using back then.
Extremely rigid 100 mm diameter instrument rings, 90mm diameter field corrector, even more rigid OTA and effectively perfect optics delivers some amazing data.

Not sure whether I'll pull the trigger on a large format sCMOS camera just yet. SBIG's AC4040 looks very seductive as does the QHY600.
I struggle with the low-level random fixed pattern noise in all CMOS sensors...something totally absent in CCD's which unfortunately are being phased out, not by engineers saying CMOS is better, but by bean counters.

Interesting times ahead ! :thumbsup: