View Single Post
  #36  
Old December 9th 12, 04:16 AM posted to rec.photo.digital.slr-systems
Wolfgang Weisselberg
external usenet poster
 
Posts: 5,285
Default D600 review isn't that great (some aspects)

Me wrote:
On 29/11/2012 10:54 a.m., Wolfgang Weisselberg wrote:
Me wrote:
On 23/11/2012 4:33 a.m., PeterN wrote:


[D800 file size and processing]


I process them easily on a three year old quad core i7, with 8 gig of
RAM. It takes a lot longer for Bridge to assemble the viewing
information, than the time for my D300 files. I anticipate even longer
times when I start shooting 14 bit color depth.


FWIW, this is resource monitor screenshot using the (cheap - probably
about US$6-700 in the US?) laptop with 3612QM cpu, opening a 14 bit
D800E lossless compressed raw file, then rotating (levelling tool), then
applying some colour control points, applying some d-lighting (high
quality), then some USM, then some noise reduction (high quality), then
saving again as an *.nef file and closing Capture NX.
http://i50.tinypic.com/qoxaw2.png


Says little, except that the app makes use of additional CPUs.


True,


Case closed.

but the CPU isn't "maxing out" for any extended period, and
there's the RAM use chart as well. Dynamic CPU clock-speed monitor
widgets don't seem to report ivybridge CPU properly, a basic monitor app
from Intel, shows clock speed jump to about 2.8 GHz for brief instants
when working on the D800 raw files, then drop back to idle at 1.2GHz.
It seems happy to sit at 3GHz for 5 minutes or so while transcoding some
video, but once the cooling system is full of dust over time...


And what do you suppose that tells us, other that your computer
is suboptimal (i.e. slower than it needs be) for the program
in question?


Now, do the same with a D300 or similar file and then we
can compare.


I have a Core 2 duo (T7500) here - it's "adequate" with D300 files.


Let me rephrase that! "Now, do the same with a D300 ON THE
SAME COMPUTER, DAMNIT, and then we can compare."


CPU speed is the "bottleneck", and if that isn't
adequate, more RAM, faster (ie SSD) drive, or physically separate HD for
the NX cache isn't likely to make much difference.


I don't know if CaptureNX can handle a GPU, but that would be
the obvious next step.


I don't know either.


Find it out.

I suppose there are some geek tools used by gamers
etc to monitor GPU, VRAM etc. I think some of the new Macbook pros (and
others?) even have a separate graphics card, but can use the intel
on-chip graphics to save power, and the external GPU kicks in only when
needed.


Google: Nvidia Optimus.

But does the GPU actually assist with raw conversion etc, or assist with
displaying (zooming, viewing multiple layers etc) graphics for which
that "number crunching" has already been done?


That depends on the program.

PeterN (above)reports that "It takes a lot longer for (Adobe) Bridge to
assemble the viewing information, than the time for my D300 files".
I've read similar comments elsewhere, that even with very fast machines,
before D800 files can be (pre)viewed at 100%, there's a couple of
seconds lag. With ViewNX, rendering D800 files is effectively instant -
ie when clicking the image preview to bring up 1:1 pixel view on screen.
I suspect that's an "issue" with Adobe bridge.


You're comparing fresh water fish and sunrays, so that means
nothing.

Meanwhile, I played with a D600 and D800 yesterday (and a Canon 6d).
RichA's bull**** making a big deal about build quality compromise and
ergonomics is of course bull****.


Tell news!

-Wolfgang