Nikon D800/E $3000, cheaper than I thought

On Thu, 08 Mar 2012 19:53:23 -0900, (Floyd L.
Davidson) wrote:

Eric Stevens wrote:
On Thu, 08 Mar 2012 16:54:53 -0900, (Floyd L.
Davidson) wrote:
There is nothing on page 38 that suggests anything wrong
with what I said.

Re-read what I said just above. Get that square before you start
spouting nonsense again.

I wish you wouldn't give these indirect references. What on earth are
you getting at?

I have always understood you to refer to the action of the paper and
head feed drives as though they were stepper motors. In fact you now
go on to say just that.

Those are of course DC stepper motors. They are fed
precisely controlled *pulses*, at specific frequencies,
to cause movements.

Stepper motors are called that for the reason that their pole design
is such that they step around one pole with each pulse. Stepper motors
are not usually associated with external encoders. Their controller
knows where they are from the number of pulses they have been given.

Yep. But "usually" does not mean "always". Eh?

They are only required where slippage of the stepper motor can have
catastrophic consequences.

You haven't looked at page 40. There Table 2-2 clearly gives the Paper
Feed (PF) Motor as "Type DC Motor, Voltage 42 VDC +/- 5%". It also
states that the Carriage Return (CF) Motor is "Type DC Motor, Voltage
42 VDC +/- 5%". Just to make certain, there are stepper motors in the
printer e.g. the Pump Motor is a "4-Phase 48 Pole PM Stepping Motor,
Voltage 42 VDC +/- 5%". The Paper Feed and Carriage Return motors are
not stepping motors. Nor are _any_ of the motors fed with 10.08 kHz or
20.16 kHz of the various drive frequencies.

So... just exactly what do you suppose various "drive
frequencies" related to the dots per inch specifications
are used for?

That's side-tracking the issue. Why do you seem to be arguing that
some stepper motors are described as stepper motors and others are
deescribed as just DC Motors?

(You want confusion... I've give you
confusion: The minimum amount that the paper can actually
be moved is 1/8640 inch.)

Apparently you didn't catch the significance of that
particular comment.

That's another of your indirect references. What on eaarth are you
driving at?

Look at page 40. See both the diagram showing print
head physical configuration and dimensions, and at
"Table 2-5 Print Modes (Drive Waveforms)" where it gives
the "Drive Frequencies" etc.

Nor the significance of that.

Neither of us know the signific ance of the Drive Waveforms or their
frequencies.

...
There are three different drive electronic programs generated within
the printer. See page 40

Use you *head* Eric. What do you see there that tells
you what scale the CR Encoder works at?

If you look at "Table 2-3 List of Sensors/Encoders/CSIC"
on page 39 you'll find that it works at "180 lpi".
Hence it provides accurate location information for
pixels at 360 pixels per inch.

It will also be sensitive to the direction of movement. This is
likely to be achieved by a vernier reading arrangement. If its got
vernier reading capability its resolution will be better than 180/360
lines per inch. At this point I'm guessing as by some means or another
the CR encoder has the ability to read to 1/1440" and the PF will go
to 1/2880".

You might stick with a little bit of reality here. Stop
imagining and stick with facts.

Then you explain how it can discriminate to 1/360" inch.

VSD1 looks after 360 x 360, 720 x 360 and 720 x 720.
VSD2 looks after 1440 x 720 (used for Gerry Eskin's tests).
VSD3 looks after 2880 x 1440.

So did you notice that each of those is related to a
different drive frequency as well as a different dot
size?

Only VSD3 has a different drive frequency, being twice that of the
frequency used for the others.

Each is specified independently. Sorry to have confused
you with simple terms again.

You have confused me with your sentence construction. Each 'what' is
specified independently of 'which'.

The printer is able to generate a number of different dot sizes (see
page 11).

VSD1 can generate 22.8, 13.8 and 6.6 where the unit of measure is
picolitres.
VSD2 can generate 13.2, 5.9 and 3.5.
VSD3 can only generate 3.5.

And the drive frequency is 10.8 KHz for all except VSD3,
which is 20.16.

Is this starting to make any sense to you yet?

No.

You don't understand it, but still want to tell me I'm
wrong.

I'm very tempted to tell you that you are glossing over those bits you
don't understand, but I won't.


I don't have the information to fully understand what is going on but

Yes you do. You just don't have the ability to
understand the information. Primarily because you are
still confused by the very basics.

PIXELS ARE NOT DOTS! Pixel per inch is not the same as
DPI.

I never said they were. All have done is (1) doubt that the 3800 is
limited to 360 x 360 pixels and (2) wonder whether pixels are the
right description for what goes on on the paper.

Try learning, and stop telling anyone what it does.

I've decideed to give up trying to learn from you. You can't give a
straight forward explanation to savve your life. I'll go and find
someone who really knows what he is talking about.

it seems that between between the carriage feed and paper movement
controls, the printer has the ability to position dots at any of the
stated resolutions

That's nonsense. It has the ability to move the head to
provide the stated resolution, which is *not* to say
that it can position just any dot at any position.

I don't see why not, within the limits of the resolution of the
various measurements.

Because when you finally understand what this is all
about, it will be obvious that 2880 dots per inch when
there are 8 different colored dots means simply 2880/8
is how often each color can appear.

Get out your calculator and divide 2880 by 8.

I've done all that.

Dots are at various resolutions. But a red nozzle
cannot put down dots at greater than every 8th dot.

I'm not sure what you mean by that. What you have just said applies if
all dots are firing. But whether or not a (say) black fires has
nothing to do with what the red does.

Use your head. Or get a 12 year old to explain it.

And
therefore the number of *pixels* per inch cannot be
greater than 2880 dot divided by 8 nozzles...

That's assuming a single print pass.

No **** Sherlock. (Re-read what I've said previously.)

which is 360 *pixels* per inch.

and is not limited to an input resolution of 360 x
360.

Yes it is you silly twit.

"Input resolution".

Don't be silly.


What's silly about that?

It is quite easily able to switch to 1440 x 720 and no doubt
others with the appropriate firmware. See
http://people.csail.mit.edu/ericchan...tml#native_res

When it switches to 1440 or 720 *dots per inch* it also
makes the dots larger, to fill up more space.

Not according to table 1-2 on page 11.

You can't read?

Look at Table 1-2 on page 11, and also at Table 2-5 on page 40.

What's the significance of 2-5 in this context? You are just dancing
around trying to give the impression you have an argument.

We've already been there, and I have no need to repeat it.

I wish you you wouldn't keep on assuming that my response to not
understanding you is to assume you are wrong.

Then stop arguing with me about what is right!

Are you saying I have to accept what you say?

At the most it means
that I will not accept that you are right.

Twit.

So far the only thing I
have seen where I am confident you are wrong is over the use of
stepper motors for Pf and CR. I don't believe they are.

They are used as stepper motors, with pulsed DC.

You are guessing and almost certainly wrong.

Eric, this topic has been explained to you in detail.
There is no point in going over and over the same
basics again and again. You've found relevant web
resources, I've explained it in detail 3 or 4 times
over, and you've read Epson's service manual.

Your attitude has always sucked, and this time is no
exception. I just have no reason at all to force feed
these concepts into someone who is not nice and refuses
to be taught.

I'm abandoning this discussion. I'm going to talk to someone who knows
what they are talking about and has the ability to explain it.

Regards,

Eric Stevens

Floyd L. Davidson wrote:

Yes, exactly. Nozzle squirts are not pixels. 2880x1440
nozzle squirts in an inch for 360 pixels is... an 8x4
matrix for each pixel.

Look on page 11 of the Service Manual where it shows the
horizontal resolution of 2880 is for a print area of 17.39"
and that it can lay down 50,029 dots of ink. With 8 colors,
that is one nozzle from each color into 360 pixels per inch.

Look on page 40 of the Service Manual where it shows that
in the vertical direction the head spacing is 1/360", and
that there are 4 pairs of nozzle rows vertically, each offset
horizontally from each other by 280/720" while each pair is
separated by 65/720". And vertically the nozzles in each pair
of rows are 1/360" apart.

It prints at 360 Pixels Per Inch. No more, and no less.

This simply isn't true: it's possible to print a clearly discernible
720ppi grid (i.e. 360 cycles per inch) on an Epson 9900. I can
provide a scan if you don't believe me, or you could try it yourself.

Andrew.

nospam wrote:
In article , David Dyer-Bennet
wrote:

Big prints are all the rage, among those who make prints at all.

no, they most definitely are not. the most popular print size is 4x6,
and has been for a long time. most people don't own a printer that can
go beyond 8.5x11 (or perhaps longer with roll paper). plus, with
digital, most people don't print anymore, they upload to facebook or
flickr or just email it.

Go look in any gallery exhibition. Those prints are getting bigger.

gallery prints are a tiny, tiny fraction of photography.

Sure, lots of people still make 4x6 prints for their albums, that's not
changing that much.

actually, it is changing quite a bit. people no longer print or put
into albums. instead, they upload to facebook, flickr, etc. and/or
carry their photos on a smartphone or tablet. the albums are now
digital.

But art prints are getting much bigger.

so what. most photos don't ever end up as art prints.

I print few of my photos, and those I do print are mostly 6"x4". Less
than 0.5% of my photos end up on walls as A3 or larger prints. But it
takes me a while to decide which ones. And sometimes a neglected old
photograph suddenly becomes topical and interesting. So I take most of my
photographs aiming at an image quality good enough for A2 with a bit
of cropping.

--
Chris Malcolm

Floyd L. Davidson wrote:
Eric Stevens wrote:
On Thu, 08 Mar 2012 16:54:53 -0900, (Floyd L.
Davidson) wrote:

Because when you finally understand what this is all
about, it will be obvious that 2880 dots per inch when
there are 8 different colored dots means simply 2880/8
is how often each color can appear.

Get out your calculator and divide 2880 by 8.
[...]
And
therefore the number of *pixels* per inch cannot be
greater than 2880 dot divided by 8 nozzles...

That's assuming a single print pass.

No **** Sherlock. (Re-read what I've said previously.)

So you do agree that with an ink coverage of 300% (very
common!) any dot position can be occupied by 3 different
colours.[1] (ignoring any different ink amounts of the dots)

With 8 colours (k (matte or gloss) lk llk c lc m lm y) not all
colours will be needed at the same place. It can be assumed that
only one of the 3 blacks are used, and either the light or the
normal version of cyan and magenta.

Thus there are roughly only 4 colours used in any pixel, and the
mixture in same parts of cyan, magenta and yellow can be replaced
by black and vice versa.

You can now calculate for yourself how many different dot positions
you need per pixel.

Your "obvious that 2880 dots per inch when there are 8 different
colored dots means simply 2880/8 is how often each color can
appear" is clear bull and shows you have very little idea of
how the stuff works.

-Wolfgang

[1] Depending on the maximum repeat rate of the nozzle it
may need several passes of the head to print all dot
positions with one ink.