DSLR vs P&S a replay of Film vs Digital?

"David J Taylor" wrote:
Where I had a problem was with the "film has a dynamic range not excceding
1000:1", and that does not agree with the publshed information.

Who, besides *you*, said that?

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

Floyd L. Davidson wrote:
"David J Taylor"
wrote:
Where I had a problem was with the "film has a dynamic range not
excceding 1000:1", and that does not agree with the publshed
information.

Who, besides *you*, said that?

You: "Film can't match a typical 12-bit digital camera, though it's a good
match for 10-bit cameras." I might agree were you to add "under typical
usage conditions".

Cheers,
David

"David J Taylor" wrote:
Floyd L. Davidson wrote:
"David J Taylor"
wrote:
Where I had a problem was with the "film has a dynamic range not
excceding 1000:1", and that does not agree with the publshed
information.

Who, besides *you*, said that?

You: "Film can't match a typical 12-bit digital camera, though it's a good
match for 10-bit cameras." I might agree were you to add "under typical
usage conditions".

So *I* said nothing that equates to "not exceding
1000:1", which is something *you* made up. A
manufactured quote...

"A good match for 10-bit cameras" means that the two are
close, and the film maybe even wins.

And what does "under typical conditions" have to do with
it that was *not* inferred already? Even under unusual
conditions film just doesn't have as much dynamic range
as the typical top of the line DSLRs today (Canon 1DsIII
and Nikon D3).

What you suggested is to compare to a film that was
discontinued a decade ago, and indeed one which
*doesn't* even get your fabricated 1000:1 ratio!

Here's another URL to review. Note the age of this
webpage, and the fact that DSLR's today are running more
than 2 fstops _better_ than the one shown.

http://www.normankoren.com/digital_t...#Dynamic_range

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

Floyd L. Davidson wrote:
"David J Taylor"
wrote:
Floyd L. Davidson wrote:
"David J Taylor"
wrote:
Where I had a problem was with the "film has a dynamic range not
excceding 1000:1", and that does not agree with the publshed
information.

Who, besides *you*, said that?

You: "Film can't match a typical 12-bit digital camera, though it's
a good match for 10-bit cameras." I might agree were you to add
"under typical usage conditions".

So *I* said nothing that equates to "not exceding
1000:1", which is something *you* made up. A
manufactured quote...

"A good match for 10-bit cameras" means that the two are
close, and the film maybe even wins.

10 bits - 1000:1 - first order

And what does "under typical conditions" have to do with
it that was *not* inferred already? Even under unusual
conditions film just doesn't have as much dynamic range
as the typical top of the line DSLRs today (Canon 1DsIII
and Nikon D3).

What you suggested is to compare to a film that was
discontinued a decade ago, and indeed one which
*doesn't* even get your fabricated 1000:1 ratio!

I chose the first transfer characteristic I was able to find.

Here's another URL to review. Note the age of this
webpage, and the fact that DSLR's today are running more
than 2 fstops _better_ than the one shown.

http://www.normankoren.com/digital_t...#Dynamic_range

which also shows Tri-X film as having somewhat more than a 10-bit (3 in
log base 10) range.

The difference is, I think, that the film's greater dynamic range occurs
in a less useful region than digital's. Note that in figure 4 the film
has not saturated when the graph stops.

Cheers,
David

"David J Taylor" wrote:
Floyd L. Davidson wrote:
"David J Taylor"
wrote:
Floyd L. Davidson wrote:
So *I* said nothing that equates to "not exceding
1000:1", which is something *you* made up. A
manufactured quote...

"A good match for 10-bit cameras" means that the two are
close, and the film maybe even wins.

10 bits - 1000:1 - first order

But "not exceding" is entirely a limitation of *your*
fabrication. I said no such thing.

And what does "under typical conditions" have to do with
it that was *not* inferred already? Even under unusual
conditions film just doesn't have as much dynamic range
as the typical top of the line DSLRs today (Canon 1DsIII
and Nikon D3).

What you suggested is to compare to a film that was
discontinued a decade ago, and indeed one which
*doesn't* even get your fabricated 1000:1 ratio!

I chose the first transfer characteristic I was able to find.

The fact that it doesn't support your claims is
significant, eh? Even the Kodak curves for Tri-X don't
support you.

Here's another URL to review. Note the age of this
webpage, and the fact that DSLR's today are running more
than 2 fstops _better_ than the one shown.

http://www.normankoren.com/digital_t...#Dynamic_range

which also shows Tri-X film as having somewhat more than a 10-bit (3 in
log base 10) range.

It shows Tri-X at *less* than 10 stops. About 2.8, not
3.0, or roughly about 631:1, for 56 dB of dynamic range.

The difference is, I think, that the film's greater dynamic range occurs

The film's *lesser* dynamic range.

That same page shows a Canon EOS-10D to have about 8.47
bits, or 355:1 for about 51 dB. But of course a lot of
water has gone under the bridge since then, and typical
digital SLRs today are all above 9 bits, many are above
10, and the best are over 11 (2048:1 or 66 dB).

Roger Clark lists 11 different cameras with more than 10
fstops of dynamic range, and that excludes of course all
of the newer models (Nikon D300 and D3, Canon Mark III):

http://www.clarkvision.com/imagedeta...ise/index.html

in a less useful region than digital's. Note that in figure 4 the film
has not saturated when the graph stops.

Here is Kodak's webpage with info on TriX. You are
welcome search for circumstances that look better:

http://www.kodak.com/global/en/profe...bs/f9/f9.jhtml

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

Scott W wrote:
Floyd L. Davidson wrote:
Here is Kodak's webpage with info on TriX. You are
welcome search for circumstances that look better:
http://www.kodak.com/global/en/profe...bs/f9/f9.jhtml


This may well me a case of what is useful range, where I
believe a DSLR wins, vs. range under any conditions. In
the transfer functions of density vs exposure they
always seem to stop the curve even though the density is
still clearly going up. One reason for this might well
be that they are already getting into such a noisy area
at high density that there is little point in showing
the curves that high.

I believe film has a very large DR, but I also believe
that to get it you have to way over expose and you will
not be happy with the photo.

I have seen a test someone did where they just kept
increasing the exposure to see how far they could still
get an image, pretty amazing how far they could go, but
you probably would not want any of the greatly over
exposed photos.

The point of course is dynamic range for useful
photography.

It isn't so much over exposure that is generally used to
increase the Dmax values, but "push" processing. Kodak
does discuss that, but doesn't provide graphs for the
details.

Basically if you increase Dmax to something above 3, the
grain and other noise become "excessive". It also
becomes more contrasty than was considered useful for
the range of papers available. (I say this all with
somewhat of a leer... my idea of fun came in a bottle
of Agfa Rodinal, and the idea that it was grainy wasn't
exactly a reason for concern! :-)

Regardless of that, dynamic range is something that is
not always optimized when using a digital camera, but
certainly setting the camera for the highest dynamic
range is a useful and typically used configuration. On
the other hand, the same is not true of film. Useless
"configuration" just to measure a dynamic range higher
than a digital camera is worthless if it cannot be used
for normal photography.

The whole point of course is that in normal photography
a digital camera today will generally have a higher
dynamic range than can be provided taking the same
pictures with 35mm film cameras. Not that you cannot
get a film to record a higher dynamic range than some
digital camera, or even that all images with digital
cameras have a higher dynamic range. Just that if
higher dynamic range is a criteria for selection, the a
digital camera will be the right choice.

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

In rec.photo.digital.zlr Scott W wrote:
Chris Malcolm wrote:
In rec.photo.digital.zlr Mr. Strat wrote:
In article , Ray Fischer
wrote:

And how many times will you inform us of your incredible expertise and
your ability to judge light levels just by looking at a scene?

I don't believe I used the word "incredible." But you amateurs never
seem to grasp the degree of competency required to produce quality
images consistently.

One of the reasons for that is that so many experts like yourself
spend so much time boasting about your skills and experience and
telling us what amateurs we are, and so little time telling us
anything useful which might help us improve.

Then there are all those people here who do exactly the same without
actually having the skills, just pretending, because if you never
communicate any of your skill it's so easy to pretend.

I sometimes wonder why a real professional would want to post anything
here, if all they ever wanted to post was sneers at the incompetence
of amateurs.

I read this back and forth and am surprised by a couple of things.
Chris you seem to believe that it is hard to get a well exposed shot
without using a histogram, have I got this right?

Nope. I find a live histogram useful enough that I wouldn't like to
have a camera without it, but probably only use it for a difficult 1%
of shots. My camera by the way is a Sony R1, probably somewhere
between a good P&S and a good DSLR in dynamic range because although
it has a DSLR sized sensor it is noisier than a DSLR, I suspect
because with live view it runs the sensor warmer. I find the rolling
zebra stripes on blown highlights useful more often than the
histogram, maybe 2% of shots. That doesn't mean these are not
important facilities: I probably only use my (35mm equiv) 19mm wide
angle 1% of the time, but consider that an essential lens.

--
Chris Malcolm DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]

"Floyd L. Davidson" wrote:

The point of course is dynamic range for useful
photography.

It isn't so much over exposure that is generally used to
increase the Dmax values, but "push" processing. Kodak
does discuss that, but doesn't provide graphs for the
details.

Basically if you increase Dmax to something above 3, the
grain and other noise become "excessive".

Push processing and overexposure are completely different; almost exactly
opposite in effect.

Noise in overexposed negative materials gets better, not worse with
overexposure. This agrees with Roger Clark's graphs showing gross noise at
the underexposed end of things.

The problem with overexposure is that the response curve isn't linear, so
the tonality gets compressed. Also, color response may be uneven in the
nonlinear range. As ScottW points out, you get an image, but not a quality
image.

It also
becomes more contrasty than was considered useful for
the range of papers available. (I say this all with
somewhat of a leer... my idea of fun came in a bottle
of Agfa Rodinal, and the idea that it was grainy wasn't
exactly a reason for concern! :-)

The excessive contrast problem is for push development; overexposure results
in less contrast.

David J. Littleboy
Tokyo, Japan

Floyd L. Davidson wrote:

It shows Tri-X at *less* than 10 stops. About 2.8, not
3.0, or roughly about 631:1, for 56 dB of dynamic range.

How do you figure that?

The Jones point is at -2.9
The DIN speed point is at -2.7

The graph goes into the positive log values and is still linear
when the graph cuts off. It is a reasonable assumption that the
8 minute line representing normal development would continue as
more or less linear up to a density of around 3.0. Thus there is
a fair amount of linear space to the right.

Peter
--

"David J. Littleboy" wrote:
"Floyd L. Davidson" wrote:

The point of course is dynamic range for useful
photography.

It isn't so much over exposure that is generally used to
increase the Dmax values, but "push" processing. Kodak
does discuss that, but doesn't provide graphs for the
details.

Basically if you increase Dmax to something above 3, the
grain and other noise become "excessive".

Push processing and overexposure are completely different; almost exactly
opposite in effect.

And that is why I pointed out the effects.

I can't see that you've added anything to the discussion
other than rewording it as if it hadn't been said.

Noise in overexposed negative materials gets better, not worse with
overexposure. This agrees with Roger Clark's graphs showing gross noise at
the underexposed end of things.

The problem with overexposure is that the response curve isn't linear, so
the tonality gets compressed. Also, color response may be uneven in the
nonlinear range. As ScottW points out, you get an image, but not a quality
image.

It also
becomes more contrasty than was considered useful for
the range of papers available. (I say this all with
somewhat of a leer... my idea of fun came in a bottle
of Agfa Rodinal, and the idea that it was grainy wasn't
exactly a reason for concern! :-)

The excessive contrast problem is for push development; overexposure results
in less contrast.

David J. Littleboy
Tokyo, Japan

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)