Re: 2nd bodies
- From: dnichols@xxxxxxxxxxx (DoN. Nichols)
- Date: 16 Oct 2005 03:36:22 GMT
According to Roger <Delete-Invallid.stuff.groups@xxxxxx>:
> On 14 Oct 2005 05:24:10 GMT, dnichols@xxxxxxxxxxx (DoN. Nichols)
> wrote:
>
> >According to David Dyer-Bennet <dd-b@xxxxxxxx>:
> >> dnichols@xxxxxxxxxxx (DoN. Nichols) writes:
> >
> > [ ... ]
> >
> >> > Yep. I believe that there was some mention here a while back
> >> > about a device which attaches to the camera's eyepiece and displays on a
> >> > LCD screen. That would help deal with such problems -- but I don't know
> >> > whether it would be sufficiently affordable to make it practical.
> >>
> >> "Video assist" for still cameras. I do remember reading about this
> >> somewhere, and it was remarkably cheap (though not cheap enough to get
> >> me to grab it right away).
> >
> > O.K. I may yet try it.
>
> Many digital still cameras have a video out. You simply plug it into
> a monitor.
Not much use when you want to use it for awkward angle
composition and focusing with a DSLR, which can't disply the image until
it *takes* it.
The device in question scans the view through the eyepiece, and
displays it on a LCD at an angle more convenient to you. (Rather like
the swivel LCDs on many of the P&S cameras.
> >
> > [ ... ]
> >
> >> > I suspect so. I really need to scan all of my old negatives,
> >> > but it takes forever with the scanner which I have. It is a Nikon model
> >> > from prior to the CoolScan models. Very high resolution, but it
> >> > requires a GPIB interface, and can only be run on a Windows or a Mac
> >> > (with the supplied software). It produces only Targa or BMP images, at
> >> > about 72MB for a full resolution scan. My GPIB card is ISA interface,
> >> > and it took absolutely forever to scan on Windows 3.11. Part of the
> >> > problem was that it insisted on scanning to *memory* and that box
> >> > couldn't handle that much memory. So -- it used virtual memory, instead
> >> > of scanning directly to disk.
>
> XP, and a good computer will take care of that.
XP might well -- but I refuse to use an OS which insists on
being re-licensed every time I change a board or two -- especially since
I refuse to let any Windows system touch the net from my IP address
block, so requesting the new license key is a serious pain.
> I'm scanning 60 to 120 meg files in less than a minute.
> Virtual memory is writting to disk, now if that could just be saved
> that way.
In other words -- if it would write directly to a file, instead
of insisting on using virtual memory at a temporary way-station?
> >> > So -- it took about one hour to scan the 72MB image, another
> >> > half-hour to save to disk, a quarter-hour to copy over the net to one of
> >> > my unix boxen, and perhaps 6-7 minutes to copy between two of the unix
> >> > boxen with a 10BaseT network interface. I've not checked the speeds
> >> > since most systems have moved to 100BaseT networking. But, I am
>
> 100 base T? I've been running a gigabit network for about 6 months.
Perhaps so -- but then, you are spending money on the latest
computers all the time. I collect Sun workstations which are perhaps
two or three years past their introduction point, and seldom spend more
than $200-$300 per computer -- sometimes with a full 2GB of RAM
installed.
> >> > convinced that the majority of the time burden was thanks to the Windows
> >> > virtual memory implementation and disk file system. :-)
>
> This computer has 2 Gigs of 400 MHz DDR RAM. If I drop to one meg I
> see no difference on most scans unless scanning 5 at 16 bit color
> depth while running other apps.
> OTOH if I drop to 512 Megs it's the difference between night and day.
> The system starts page file swapping and takes for ever.
And the Windows computer which I used for that was limited to
something like 64MB of RAM maximum, so it *had* to swap with a 72MB
image file, even making the invalid assumption that none of the memory
was being used by the OS. :-)
> >>
> >> Now *that's* a slow scanner.
> >
> > I really think that the majority of the slowness came from the
> >combination of the Windows filesystem and the virtual memory in that
>
> It came from an old version of Windows that can't handle enough
> memory.
Agreed -- and old hardware which could not supply sufficient
memory.
However -- the GPIB interface card is an ISA card, and too
expensive to replace at new prices. (National Instruments *loves* their
cards and the driver software. :-)
> >system. I am interested in seeing how much faster it would be directly
> >on a Sun, which deals gracefully with both large files and virtual
> >memory.
>
> Just put it on a newer Win machine with one gig of RAM and I'll bet
> you'll be surprised,
How -- the newer ones don't support the ISA bus for the GPIB
board.
> but OTOH if you could hear it scanning line by
> line and then slow down when it started page file swapping that was
> one slow scanner.
A tick as it moved to each new line, and then scanned through
three successive color filters. The ticks were along the 36mm
dimension, and a scanning mirror for the 24mm dimension.
And if I did not use the GPIB interface, it would have to got
through a serial port -- I forget whether that topped off at 9600 baud,
or at 19.2k -- in any case -- way too slow for transporting 72MB of
image data. I figure about 10.42 hours with no overhead in the data
transfer. :-)
[ ... ]
> >> I've been using a Nikon LS-2000 for a while now, can't remember
> >> exactly when I got it. But my new LS-5000 arrived today, *with* the
> >> slide feeder (and reports I've read indicate the slide feeder actually
> >> *works* on the 5000, unlike earlier versions). Haven't even hooked it
> >> up yet, but I anticipate hard drives becoming full very rapidly here
> >> soon!
>
> The auto feeder depends on how good your mounts are. It will not
> handle curved paper mounts, it will not handle paper mounts with edges
> that have spread and there are a number of plastic mounts wihich will
> only feed one way. So, you learn to iron your paper mounted slides
> <:-)) smooth the edges and make sure the slide mounts go in the
> proper way, reguardless of what direction the transparencies is
> situated in the mount.
Most of my slides are glass mount, with various formats of
carriers, many of which would probably not go through the auto-feeder.
Of course the negatives are 6-frame strips, and easier to mount in the
old LS-3500, which has a nice negative/full-length-film holder.
> They can run for hours with out a fault and they can jam on nearly
> every slide depending on the mount and its condition.
The odds of jamming increase with your need to have it run
unattended. :-)
> I have an LS5000ED and the SF210 feeder. I've put well past 20,000
> slides, negatives, and transparency film strips through it. It was
> worth the money, but it is not without fault. You do not put 50
> slides in the feeder, walk off, and expect to come back and find all
> 50 have gone through the scanner. You might, but that would be your
> lucky day.
:-)
> >
> > Hmm ... what was that one which I have? O.K. -- LS-3500.
> >
> >> > If I ever get the drivers for my Sbus GPIB cards for my Suns, I
> >> > hope to write my own programs to run it, as the full command set for the
> >> > scanner was included in the manual. :-)
> >>
> >> Running it directly off the Sun should simplify the workflow
> >> considerably.
> >
> > It should also speed up the acquisition of the image. (Once I
>
> You are probably goig to be I/O bound by the output speed of the
> scanner
Even more so with the serial port interface, which is the only
alternative to the GPIB.
Enjoy,
DoN.
--
Email: <dnichols@xxxxxxxxxxx> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---
.
- Follow-Ups:
- Re: 2nd bodies
- From: Eugene
- Re: 2nd bodies
- References:
- 2nd bodies
- From: Gordon
- Re: 2nd bodies
- From: Roger
- 2nd bodies
- Prev by Date: Re: To RAW or not to RAW?
- Next by Date: Re: Nikon autofocus lenses
- Previous by thread: Re: 2nd bodies
- Next by thread: Re: 2nd bodies
- Index(es):
Relevant Pages
|
