On 8/30/2018 4:49 PM, Mike Gordon via olympus wrote:
Looks pretty good for a huge crop though would require a crop from a full
sized tiff not already small jpeg of the original to compare. Would await
analysis if possible from peripatetic Maestro Manipulator Moose.
You left out the alliterative "Magic", ;-)
As I wrote before, Topaz A.I. Gigapixel wouldn't work on either of my machines. I installed a Zotac GeForce GTX 1050Ti
Mini in my desktop.* It seemed to me the most cost effective solution to a high end graphics processor with lots of memory.
Now Gigapixel not only works, but is faster than I'd expected. I suppose speed
depends a lot on the GPU and amount of VRAM.
In the meantime, in between time, Topaz released an update that can use the CPU, instead of a GPU. I imagine this will
run slowly and slow one's computer down. Mike reports that the slow isn't that bad. I'm not sorry I bought the card, as
I've been curious about the benefits of a separate, high performance video card for PS. Blending of a couple of focus
stacks in PS has seemed to go much faster than I'm used to.
The other thing the update brings is alternatives to the default processing to allow no NR or higher NR. This seems
important to me. This example, uprezzing by 400%, then down to 200%, shows the subtle difference between no NR and the
default setting. There's really no more detail, but the little bit of noise gives the impression of greater
sharpness/detail. <http://www.moosemystic.net/Gallery/tech/TopazAI_Gigapixel/AsianFalcon.htm>
I'm impressed, so far, both by my falcon and your eagle. It may be less wonderful for some other subjects. The Qimage
guy has posted a detailed comparison video that's very informative.
<https://www.youtube.com/watch?v=LGWEyG4DUOM&feature=youtu.be>
I'm not sure his approach is fully valid. Might be interesting to take an Oly HR shot, then try uprezzing the regular
rez to HR size. That way, one avoids the potential effect of the downsampling step. Other approaches using different
lenses or focal distances could work, but it would be impossible to fully register the two images.
In any case, he seems to have clearly shown the good and bad sides of both
tools.
Would rarely require a huge up res to print but the software may be very
interesting for crops like this when getting short of megapickles.
Yup. Going through images from several months, I've found very few where uprezzing is likely to add much. If I'm close
enough, I "upsize" with the zoom. I also think experience leads me not to take some shots where I know the subject will
be small in the frame.
Then, there's the effect of atmosphere. I suspect the ~40 m focal distance of the falcon is a larger contributor to
moderate detail than lens. I did find a few other examples of distant birds, but there was no detail to enlarge. The air
in Bhutan is a lot clearer than here.
So, I'd say it's a great tool, for limited uses.
Rez Up Moose
* I linked before to B&H, with a price of $170. When I went to buy, it was over $200. Amazon was $185, including tax. I
ordered at noon on Labor Day. Confirmation said delivery the next day. Then a notice to expect it same day. Delivered
before 8pm.
--
What if the Hokey Pokey *IS* what it's all about?
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|