Sony tells DSLR shooters they're idiots

On Thu, 6 Dec 2012 23:13:29 +1100, "Trevor" wrote:


wrote in message
.. .
On Wed, 5 Dec 2012 01:48:42 -0500, "Gary Eickmeier"

wrote:
wrote in message
...

The purpose of the 24/96 system is to aid in mixing multi-channel sound.
If you
want to mix down 2 tracks, for example, in 16/44 you would have to drop
the
levels by 3db. If you want more tracks, you have to cut even more, and
since
each bit represents 3db,

6dB actually. But don't forget each track you add also gives you 6dB more
level, so you lose no S/N by adding tracks *unless* they contain no signal.
The idea is to mute or "duck" the tracks where there is no signal.

6db in volts, but db (electricity) is always related to power, which is 3db.
Not really important here...


you eventually are left with 8 bit noisy crap!

I don't get this Bob - what does the number of channels have to do with
bitrate for each channel?


I'm not referring to the bit RATE here, but the bit DEPTH... the 24
compared to
the 16...

You can afford to drop 8 bits and still get a S/N of 96db.

Nope, since no 24 bit recorder actually does 24 bits (since you'd need
everything close to absolete zero temp to physically do so!)
A good 24 bit recorder does a maximum 18-20 bits real dynamic range (and
they are very close to what is physically possible without supercooling
now), giving you about 3 more bits to play with, which is still better than
not having it of course.

Trevor.


What I meant is that you can drop levels or add tracks in 24 bit and end up with
much more than you would in 16 bit... if you do this track by track you really
don't start or end with 24 bits, as you say that is sort of theory only, but you
should end up with a useful 16 bits. If you start mixing in 16 bits there is no
place to go... dropping a channel level by 3 bits here is a permanent loss, but
in 24 it's no problem.

On Thu, 06 Dec 2012 13:42:24 -0600, David Dyer-Bennet wrote:

writes:

On Wed, 05 Dec 2012 00:48:41 -0600, David Dyer-Bennet wrote:

writes:

On Tue, 4 Dec 2012 23:45:42 +0100, Alfred Molon wrote:

In article , David Dyer-Bennet says...
Yeah, right. Ask anybody these days. Nikon is making the best DSLRs in
the market.

If they are the best, why is there is 3/4 stop exposure mistake?

Exposure mistakes are in the eye of the beholder... the camera thinks
it did fine, but it doesn't have an incident light meter, which is the
only way to get perfection.

Incident metering is by no means perfect. In fact, it's less accurate
than reflected metering done carefully; it's a quick-and-dirty kludge
that's useful in some situations, especially with low-contrast lighting,
plus it's useful in the studio when reading the effects of individual
lights as you set things up. I do still have my separate light meter
(including flash meter), but I see no reason to use it these days.

Of course, the metering method must be tempered by the knowledge and
experience of the photographer, but the incident meter tells you how
much light is available, the camera only tells you the average of what
is reflected, which means nothing in a noon hour snow scene!

It means a lot -- it gives you the vital data you need to know that
you've avoided blowing out the highlights! With incident, you can apply
experienced intelligence to account for that, but you're doing it by
inference, whereas a reflected reading of a highlight tells you what the
actual brightness is. The reflected light is what the film actually
sees!

Actually, the camera measures the average of the reflected light, which in a
snow scene, is probably not the item you want to photograph... and you will end
up with a gray photo, rather than white snow.

If I know how much light is there, I can photograph a barn regardless if it's
surrounded by snow or manure... the camera will just average everything.

No, it records each individual spot independently -- meaning that the
highlights on the snow are probably blown.

Your camera will almost never blow out highlights in a snow scene, it will turn
everything down. An incident meter tells you how much light there is, and the
resultant photo will show whites and blacks if they exist, hence the snow will
be white, not gray.

I know what you're gonna say next - spot meter! Well, what if the object in
question has a wide tonal range? Spot meter no good....

Spot metering isn't generally used by taking *one* reading; it's used by
reading various important tones, and deciding where to place one or more
of them (yes, placing more than one requires adjusting light or
adjusting film and/or processing).

Now you aren't using the cameras spot meter to take the photo, but you are
making decisions about a manual exposure... not the same thing.

I've taken lots of photos with the camera meter and then with the Seconic
incident meter, and the incident is almost always better. I plan to buy a
new-fangled digital meter that can do flash as well.

Sometimes I have to pay a lot of attention to adjusting for the scene, but I
don't always do it.

Or you could just shoot manual!

I thought we were discussing methods of picking what manual settings to
make? I.e. that shooting manual was an initial assumption of this
thread.

Perhaps we got sidetracked...

On Thu, 06 Dec 2012 13:05:23 -0600, Doug McDonald wrote:

On 12/4/2012 8:02 PM, wrote: On Tue, 4 Dec 2012 23:45:42 +0100, Alfred Molon
wrote:

In article , David Dyer-Bennet says...
Yeah, right. Ask anybody these days. Nikon is making the best DSLRs in
the market.

If they are the best, why is there is 3/4 stop exposure mistake?

Exposure mistakes are in the eye of the beholder... the camera thinks it did
fine, but it doesn't have an incident light meter, which is the only way to get
perfection.

And ALL cameras have that problem, as if you didn't know.

Huh? In my experience an incident light meter almost NEVER ever
gets exposure right.

It's broken... or you used it for the wrong situation.

In principle it simply can't, since it
has no way to tell how light or dark the subject is.

That's the whole point... you set the exposure according to the light you have,
and normally reflecting objects, both black and white, will be exposed properly.
In a snow scene, snow will be white, buildings dark. A reflected average will be
terribly gray, and with a spot meter... what do you spot? The building? Then the
snow will be wrong. The snow? Then everything will be wrong.

And even worse ...
what if the incident light is different in different parts of the image.

That is illogical... does not compute... error... error...!

You use incident meters for the location of the objective... E.G. a group
portrait is perfect. Some shots of course you can't use one, too many different
locations in view.

The only way to get exposure exactly right, with metering, is
a spot meter and lots and lots of knowledge about the exact
correspondence between its reading and the actuality of your camera.

A spot meter does a good job... of the spot it's on... but a small spot means
only a small part of your exposure will be good.

These are only tools... if in doubt, bracket!

On Thu, 06 Dec 2012 20:55:34 -0500, wrote:

On Thu, 06 Dec 2012 13:42:24 -0600, David Dyer-Bennet wrote:

writes:

On Wed, 05 Dec 2012 00:48:41 -0600, David Dyer-Bennet wrote:

writes:

On Tue, 4 Dec 2012 23:45:42 +0100, Alfred Molon wrote:

In article , David Dyer-Bennet says...
Yeah, right. Ask anybody these days. Nikon is making the best DSLRs in
the market.

If they are the best, why is there is 3/4 stop exposure mistake?

Exposure mistakes are in the eye of the beholder... the camera thinks
it did fine, but it doesn't have an incident light meter, which is the
only way to get perfection.

Incident metering is by no means perfect. In fact, it's less accurate
than reflected metering done carefully; it's a quick-and-dirty kludge
that's useful in some situations, especially with low-contrast lighting,
plus it's useful in the studio when reading the effects of individual
lights as you set things up. I do still have my separate light meter
(including flash meter), but I see no reason to use it these days.

Of course, the metering method must be tempered by the knowledge and
experience of the photographer, but the incident meter tells you how
much light is available, the camera only tells you the average of what
is reflected, which means nothing in a noon hour snow scene!

It means a lot -- it gives you the vital data you need to know that
you've avoided blowing out the highlights! With incident, you can apply
experienced intelligence to account for that, but you're doing it by
inference, whereas a reflected reading of a highlight tells you what the
actual brightness is. The reflected light is what the film actually
sees!

Actually, the camera measures the average of the reflected light,

Nikons have measured the pattern of brightness and deduced the nature
of the subject since the late 1980s. I remember establishing that the
F801s could determine the correct exposure for snow but we were
puzzled as to how it did it. We were told that it had a database of
45,000 pictures from which it worked.

... which in a
snow scene, is probably not the item you want to photograph... and you will end
up with a gray photo, rather than white snow.

If I know how much light is there, I can photograph a barn regardless if it's
surrounded by snow or manure... the camera will just average everything.

No, it records each individual spot independently -- meaning that the
highlights on the snow are probably blown.

Your camera will almost never blow out highlights in a snow scene, it will turn
everything down. An incident meter tells you how much light there is, and the
resultant photo will show whites and blacks if they exist, hence the snow will
be white, not gray.

I know what you're gonna say next - spot meter! Well, what if the object in
question has a wide tonal range? Spot meter no good....

Spot metering isn't generally used by taking *one* reading; it's used by
reading various important tones, and deciding where to place one or more
of them (yes, placing more than one requires adjusting light or
adjusting film and/or processing).

Now you aren't using the cameras spot meter to take the photo, but you are
making decisions about a manual exposure... not the same thing.

I've taken lots of photos with the camera meter and then with the Seconic
incident meter, and the incident is almost always better. I plan to buy a
new-fangled digital meter that can do flash as well.

Sometimes I have to pay a lot of attention to adjusting for the scene, but I
don't always do it.

Or you could just shoot manual!

I thought we were discussing methods of picking what manual settings to
make? I.e. that shooting manual was an initial assumption of this
thread.

Perhaps we got sidetracked...
--

Regards,

Eric Stevens

"Eric Stevens" wrote in message
...
On Wed, 5 Dec 2012 15:20:52 +1100, "Trevor" wrote:


"Gary Eickmeier" wrote in message
...
OK, so I am iggerant. But you guys haven't been able to show me an
example
of a RAW image vs a JPG shot at the same time that demonstrates this
superiority of image.

If you are unable to demonstrate it for yourself, then it probably doesn't
matter to *you* what the difference is. The rest of us already know and
choose our work flow accordingly.

I have several times attempted to draw the attention of the ignoramus
to http://www.slrlounge.com/raw-vs-jpeg...e-visual-guide
which most definitely provides the information he says that he
requires. However he steadfastly refuses to either look at it or
acknowledge that it provides the information that he says he requires.
I think he is a troll.
--

Regards,

Eric Stevens

**** you and the horse you rode in on. I have looked at it several times.
No, it does not show any big difference in the images.

G

On 06/12/2012 22:13, Elliott Roper wrote:
In article , Doug McDonald
wrote:

On 12/5/2012 10:05 AM, Elliott Roper wrote:

(about Aperture and its preliminary RAW processing)
So I learned something from this discussion, even if it only why others
were inexplicably preferring JPG and lamenting the tedium of processing
RAW in post. Typical of Apple - "it just works".


Really? In my experience with Apple their slogan is
"Do it OUR way and like the result OR ELSE ... pray that we
will have, somewhere in some obscure menu item, some sort of
corrective setting." Which they do in maybe 40% of cases.

I will say that the big problem with the iPod Nano that actually
is a Shuffle concerning losing position in a playlist elicited
so many complaints that they issued a fix update. The didn't fix
complaints about how it shuffles.

That's what I like /so/ much about this group. Clear incisive argument
with deep insight into the subject at hand.

Doug /is/ right about Apple, though.
--
Cheers,
David
Web: http://www.satsignal.eu

"Trevor" wrote in message
...

What makes you think the average will always be the "perfect exposure"?
(it won't be)
A person behind the camera with a brain and some idea what s/he is doing
beats any automatic averaging system, especially since the incident and
reflected readings usually need to be taken at different locations. Spot
metering, and knowing where to point works pretty well already.

Frankly for autoexposure, a camera that always placed the total RAW
exposure just below white clipping would probably suit me better these
days. Given the number of matrix points used by some camera's these days,
it's almost possible with mirror mode as it already is with "live view".

Well, reflected is not always correct because you don't always want the
subject to average out to neutral gray. Take a black cat on a coal pile. If
you use a reflected meter, it will want to make the exposure too bright,
when what you want is to keep it nearer to black.

Now let's use an incident meter. It is incapable of evaluating the tones in
the subject, and may make it TOO black. But a combination of these two
exposures should be pretty damn close.

Another idea - why can't digital cameras use a histogram to adjust exposure?
With Live View you have live histograms at all times.

Gary Eickmeier

"nick c" wrote in message
...

Why couldn't the old method of using a Styrofoam coffee cup over the front
of the lens be used as an incident light meter? I haven't found a need to
try it but some other old timer might have thought of using that old
trick.

I have tried the Omnidisc a few times and it is no panacea. Camera's normal
metering was always better. I tried both pointing at the subject and
pointing at the light source.

Gary Eickmeier

On Fri, 7 Dec 2012 02:32:44 -0500, "Gary Eickmeier"
wrote:


"Eric Stevens" wrote in message
.. .
On Wed, 5 Dec 2012 15:20:52 +1100, "Trevor" wrote:


"Gary Eickmeier" wrote in message
...
OK, so I am iggerant. But you guys haven't been able to show me an
example
of a RAW image vs a JPG shot at the same time that demonstrates this
superiority of image.

If you are unable to demonstrate it for yourself, then it probably doesn't
matter to *you* what the difference is. The rest of us already know and
choose our work flow accordingly.

I have several times attempted to draw the attention of the ignoramus
to http://www.slrlounge.com/raw-vs-jpeg...e-visual-guide
which most definitely provides the information he says that he
requires. However he steadfastly refuses to either look at it or
acknowledge that it provides the information that he says he requires.
I think he is a troll.
--

Regards,

Eric Stevens

**** you and the horse you rode in on. I have looked at it several times.

Naughty naughty. You musn't lose your temper. The fact is that you
have never previously acknowledged that I have given you a URL leading
to a site which gives you the information you have kept bleating for.
You have missed so many times that you have put me in mind of Robert
A. Heinlein's "Once is happenstance, twice is coincidence and three
times is enemy action".

No, it does not show any big difference in the images.

Then you are blind.
--

Regards,

Eric Stevens

In rec.photo.digital.slr-systems David Dyer-Bennet wrote:
writes:

On Tue, 4 Dec 2012 23:45:42 +0100, Alfred Molon wrote:

In article , David Dyer-Bennet says...
Yeah, right. Ask anybody these days. Nikon is making the best DSLRs in
the market.

If they are the best, why is there is 3/4 stop exposure mistake?

Exposure mistakes are in the eye of the beholder... the camera thinks
it did fine, but it doesn't have an incident light meter, which is the
only way to get perfection.

Incident metering is by no means perfect. In fact, it's less accurate
than reflected metering done carefully; it's a quick-and-dirty kludge
that's useful in some situations, especially with low-contrast lighting,
plus it's useful in the studio when reading the effects of individual
lights as you set things up. I do still have my separate light meter
(including flash meter), but I see no reason to use it these days.

I still use mine sometimes. I use when I'm using a completely manual
lens with no auto features, I use it when setting up flashguns, and I
use it to decide which lens to use in awkward lighting circumstances
when my camera is still bagged, because it saves carrying out
experiments with the wrong lens.

It has two attachments for incident metering. It has a dome for doing
total incident light metering, and a flat disc for for doing incident
metering with respect to a specific direction of light. That
difference is sometimes important. Incident metering isn't necessarily
perfect, because the meter can't know what your purpose is. The
simplest example is a low contrast scene which you might wish to
expose to the right (of the histogram) as high key, or to the left as
a night scene.

Nor can a histogram-derived auto exposure always be perfect for the
same reason. For example, sometimes you want to expose to the left or
the right of the second histogram peak, not the first.

--
Chris Malcolm