Sign in with
Sign up | Sign in
Your question

Question about ISO value ?

Last response: in Digital Camera
Share
May 25, 2005 4:15:05 AM

Archived from groups: rec.photo.digital (More info?)

I noticed that in manual mode I can choose 100, 200, or 400,
but on auto mode I can glso get 179, i.e. values in between.
What's the reason for this?

tia.

More about : question iso

May 25, 2005 4:15:06 AM

Archived from groups: rec.photo.digital (More info?)

Most cameras have ISO settable in third stops which would be:
100, 125, 160, 200 or
200, 250, 320, 400 etc
every doubling in the numbers is an increase of 1 stop.
I don't know where your camera gets a value of 179. I suppose some camera
might use quarter stops or even fifths of stops but if they did the numbers
would probably be rounded off.


--
http://www.chapelhillnoir.com
home of The Camera-ist's Manifesto
The Improved Links Pages are at
http://www.chapelhillnoir.com/links/mlinks00.html
A sample chapter from "Haight-Ashbury" is at
http://www.chapelhillnoir.com/writ/hait/hatitl.html

<olivia> wrote in message news:4293b634$1_4@mk-nntp-2.news.uk.tiscali.com...
> I noticed that in manual mode I can choose 100, 200, or 400,
> but on auto mode I can glso get 179, i.e. values in between.
> What's the reason for this?
>
> tia.
>
>
Anonymous
May 25, 2005 4:15:06 AM

Archived from groups: rec.photo.digital (More info?)

<olivia> wrote in message news:4293b634$1_4@mk-nntp-2.news.uk.tiscali.com...
>I noticed that in manual mode I can choose 100, 200, or 400,
> but on auto mode I can glso get 179, i.e. values in between.
> What's the reason for this?
>
> tia.
>
If by Auto Mode you mean the ISO is floating, I guess it's possible to get
any number between the highest and lowest available ISO. If you are
actually setting the ISO speed, the number 179 would be a bit odd.
Related resources
Anonymous
May 25, 2005 1:20:37 PM

Archived from groups: rec.photo.digital (More info?)

"Sheldon" <sheldon@XXXXXXXXsopris.net> wrote in message
news:b9OdnZTnFuPCSg7fRVn-1Q@comcast.com...
>
> <olivia> wrote in message
news:4293b634$1_4@mk-nntp-2.news.uk.tiscali.com...
> >I noticed that in manual mode I can choose 100, 200, or 400,
> > but on auto mode I can glso get 179, i.e. values in between.
> > What's the reason for this?
> >
> > tia.
> >
> If by Auto Mode you mean the ISO is floating, I guess it's possible to get
> any number between the highest and lowest available ISO. If you are
> actually setting the ISO speed, the number 179 would be a bit odd.
>
-------

Not odd at all in digital - you could get EXIF data showing a shot at ISO
142 + f 5.9 + shutter at 1/173s

Accurate? Reality? Who knows?

Journalist
Anonymous
May 25, 2005 11:27:54 PM

Archived from groups: rec.photo.digital (More info?)

"Journalist-North" <journalist-north@blueyonder.co.uk> wrote in message
news:FrXke.128389$Cq2.4876@fe2.news.blueyonder.co.uk...
>
> "Sheldon" <sheldon@XXXXXXXXsopris.net> wrote in message
> news:b9OdnZTnFuPCSg7fRVn-1Q@comcast.com...
>>
>> <olivia> wrote in message
> news:4293b634$1_4@mk-nntp-2.news.uk.tiscali.com...
>> >I noticed that in manual mode I can choose 100, 200, or 400,
>> > but on auto mode I can glso get 179, i.e. values in between.
>> > What's the reason for this?
>> >
>> > tia.
>> >
>> If by Auto Mode you mean the ISO is floating, I guess it's possible to
>> get
>> any number between the highest and lowest available ISO. If you are
>> actually setting the ISO speed, the number 179 would be a bit odd.
>>
> -------
>
> Not odd at all in digital - you could get EXIF data showing a shot at ISO
> 142 + f 5.9 + shutter at 1/173s
>
> Accurate? Reality? Who knows?
>
> Journalist
>
True, but could you actually set the ISO to 142? That was my explanation.
While it might be able to float to 142 in an automatic mode, I've never seen
a camera that would let you set the ISO to anything but a "normal" number.
Anonymous
May 27, 2005 10:30:00 PM

Archived from groups: rec.photo.digital (More info?)

Sheldon wrote:

> "Journalist-North" <journalist-north@blueyonder.co.uk> wrote
>>
>>Not odd at all in digital - you could get EXIF data showing a shot at ISO
>>142 + f 5.9 + shutter at 1/173s
>>
>>Accurate? Reality? Who knows?
>>
>>Journalist
>>
>
> True, but could you actually set the ISO to 142? That was my explanation.
> While it might be able to float to 142 in an automatic mode, I've never seen
> a camera that would let you set the ISO to anything but a "normal" number.

Apparently the auto ISO is fairly fine tuned (in some mysterious way) if
it's giving numbers like that. I suppose it doesn't matter much but
theoretically if the auto descision process was understood it could be
useful.

I think it would be cool to be able to program all sorts of custom modes
like curves & specific auto ISO response criteria. I guess some day that
might be an option though maybe just for programmer/linux geeks <g>. Or
be able to get plug-ins for cameras that tweak the firmware. So far I've
heard zero about D70 firmware hacks, only the Canon 300D.


--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
May 28, 2005 2:17:36 AM

Archived from groups: rec.photo.digital (More info?)

Sheldon wrote:
> True, but could you actually set the ISO to 142? That was my explanation.
> While it might be able to float to 142 in an automatic mode, I've never seen
> a camera that would let you set the ISO to anything but a "normal" number.

Sure you can, just shoot RAW.
Anonymous
May 28, 2005 2:17:44 AM

Archived from groups: rec.photo.digital (More info?)

Sheldon wrote:
> True, but could you actually set the ISO to 142? That was my explanation.
> While it might be able to float to 142 in an automatic mode, I've never seen
> a camera that would let you set the ISO to anything but a "normal" number.

Sure you can, just shoot RAW.
Anonymous
May 28, 2005 2:32:48 AM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:

>
> Sheldon wrote:
>
>>True, but could you actually set the ISO to 142? That was my explanation.
>>While it might be able to float to 142 in an automatic mode, I've never seen
>>a camera that would let you set the ISO to anything but a "normal" number.
>
>
> Sure you can, just shoot RAW.


Is that what exposure adjustments do in a raw converter? Makes sense.


--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
May 28, 2005 2:59:12 AM

Archived from groups: rec.photo.digital (More info?)

Paul Furman wrote:
> george_preddy@yahoo.com wrote:
> > Sheldon wrote:
> >
> >>True, but could you actually set the ISO to 142? That was my explanation.
> >>While it might be able to float to 142 in an automatic mode, I've never seen
> >>a camera that would let you set the ISO to anything but a "normal" number.
> >
> >
> > Sure you can, just shoot RAW.
>
> Is that what exposure adjustments do in a raw converter? Makes sense.

Within usable bounds (+/- 2 or 3 stops) finely adjusting the exposure
using a RAW converter is the general equivalent of tweaking the ISO in
tiny increments.

The better RAW converters, like SPP, even let you alter the ISO for
various pixels within any single image, based on such things as
luminousity, for example. You can pumping up the exposure in only the
shadows, for example, using X3FL or the Shadows slider, etc. Of course
you still get a noise penalty in the images or image-regions that you
push, but it can sometimes be less than if you set the ISO explicitly
higher.

For example, you might set ISO 100 with a -2EV camera exposure
adjustment, then push the exposure +2EV using the RAW converter.
Depending on the camera and the scene, you could get better overall
results than shooting ISO 400.

--George (not to be confused with Steven Scharf, posting pro Simga
using my name)
May 28, 2005 5:35:28 AM

Archived from groups: rec.photo.digital (More info?)

Seems like a basic bread and butter issue for photographers???
Anonymous
May 28, 2005 8:45:37 PM

Archived from groups: rec.photo.digital (More info?)

<george_preddy@yahoo.com> wrote in message
news:1117259952.342255.263170@g14g2000cwa.googlegroups.com...
SNIP
> Within usable bounds (+/- 2 or 3 stops) finely adjusting the
> exposure using a RAW converter is the general equivalent
> of tweaking the ISO in tiny increments.

Wrong. Decreasing the on camera ISO setting results (if the lightmeter
setting is followed) in increased exposure. That means that there are
more photons being captured. Increasing the ISO setting will likewise
decrease exposure and may affect the analog signal being amplified
before quantification (in the Analog Digital Converter).

Postprocessing can only work on the already quantified signal numbers
(with their inherent quantization noise added to the data before being
processed in software). This can lead to more damage to the image.

There is nothing that beats the quality of correct exposure, as can be
seen in e.g. the difference between a 2 or 3 stops underexposed image
adjusted by software, and the real thing, a correct exposure.

Bart
Anonymous
May 29, 2005 6:13:20 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117257456.223031.274890@f14g2000cwb.googlegroups.com>,
george_preddy@yahoo.com wrote:

>Sheldon wrote:
>> True, but could you actually set the ISO to 142? That was my explanation.
>> While it might be able to float to 142 in an automatic mode, I've never seen
>> a camera that would let you set the ISO to anything but a "normal" number.
>
>Sure you can, just shoot RAW.

There are some serious limits that you have to be aware of when you
push/pull to other exposure indexes. When you push, you wind up using a
smaller range of RAW values, so your data is more posterized, and this
tends to emphasize noise (and distort the signal). When you pull, your
posterization actually decreases, but you lose headroom in the
highlights, so you may clip your highlights.

On any camera with a decent gain amplifier, it is generally better to
use a higher gain than to push, if possible. George Preddy's claim a
couple of years ago that shooting at ISO 100 on the SD9 at -2 EV gave
better results than shooting straight at ISO 400 is bizarre, to say the
least, and the exact opposite of what happens with most cameras.

--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
May 29, 2005 6:15:28 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>For example, you might set ISO 100 with a -2EV camera exposure
>adjustment, then push the exposure +2EV using the RAW converter.
>Depending on the camera and the scene, you could get better overall
>results than shooting ISO 400.

That would require a defective or low-quality amplifier, or imply that
the camera was already doing the same, but discarding highlights.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
May 30, 2005 11:40:27 AM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117257456.223031.274890@f14g2000cwb.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
> >Sheldon wrote:
> >> True, but could you actually set the ISO to 142? That was my explanation.
> >> While it might be able to float to 142 in an automatic mode, I've never seen
> >> a camera that would let you set the ISO to anything but a "normal" number.
> >
> >Sure you can, just shoot RAW.
>
> There are some serious limits that you have to be aware of when you
> push/pull to other exposure indexes.

There are serious limits using high ISOs on the camera body as well.
Camera processors are very weak compared to desktop computers. So it
depends on the camera, the scene, the quality of the RAW processor and
other tools you may use as to which method produces the better end
result for any given image.

In general, Sigma's DSLRs produce very low noise results when pushing
ISO during RAW since ISO 100 noise is so much lower than other DSLRs.
So with the SD9 and SD10, it is almost always better to shoot ISO 100
-xEV and push during RAW conversion than it is to shoot high ISO.
Anonymous
May 30, 2005 5:03:29 PM

Archived from groups: rec.photo.digital (More info?)

Bart van der Wolf wrote:
> <george_preddy@yahoo.com> wrote in message
> news:1117259952.342255.263170@g14g2000cwa.googlegroups.com...
> SNIP
> > Within usable bounds (+/- 2 or 3 stops) finely adjusting the
> > exposure using a RAW converter is the general equivalent
> > of tweaking the ISO in tiny increments.
>
> Wrong. Decreasing the on camera ISO setting results (if the lightmeter
> setting is followed) in increased exposure. That means that there are
> more photons being captured. Increasing the ISO setting will likewise
> decrease exposure and may affect the analog signal being amplified
> before quantification (in the Analog Digital Converter).
>
> Postprocessing can only work on the already quantified signal numbers
> (with their inherent quantization noise added to the data before being
> processed in software). This can lead to more damage to the image.
>
> There is nothing that beats the quality of correct exposure, as can be
> seen in e.g. the difference between a 2 or 3 stops underexposed image
> adjusted by software, and the real thing, a correct exposure.

The overall EV value is obviously identical using both methods, or
there wouldn't be exact parameter equivalence. I thought that part was
obvious.
Anonymous
May 30, 2005 5:05:08 PM

Archived from groups: rec.photo.digital (More info?)

olivia wrote:
> Seems like a basic bread and butter issue for photographers???

You'll find that 99% of photographers don't understand the most basic
issues of digital photography.
Anonymous
May 30, 2005 5:06:18 PM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
> >For example, you might set ISO 100 with a -2EV camera exposure
> >adjustment, then push the exposure +2EV using the RAW converter.
> >Depending on the camera and the scene, you could get better overall
> >results than shooting ISO 400.
>
> That would require a defective or low-quality amplifier, or imply that
> the camera was already doing the same, but discarding highlights.

Or, that a desktop computer has more computing power available than a
camera.
Anonymous
May 30, 2005 5:06:18 PM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
> >For example, you might set ISO 100 with a -2EV camera exposure
> >adjustment, then push the exposure +2EV using the RAW converter.
> >Depending on the camera and the scene, you could get better overall
> >results than shooting ISO 400.
>
> That would require a defective or low-quality amplifier, or imply that
> the camera was already doing the same, but discarding highlights.

Or, that a desktop computer has more computing power available than a
camera.
Anonymous
May 30, 2005 5:25:29 PM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
> >For example, you might set ISO 100 with a -2EV camera exposure
> >adjustment, then push the exposure +2EV using the RAW converter.
> >Depending on the camera and the scene, you could get better overall
> >results than shooting ISO 400.
>
> That would require a defective or low-quality amplifier, or imply that
> the camera was already doing the same, but discarding highlights.

The camera doesn't have to be defective to shoot advantageously at -xEV
under some conditions, people who know what they are doing use the
option all the time. Here is an Sigma SD9 photo at ISO 1600...

http://www.pbase.com/isoequiv/image/44078397/large

It is much better than the camera's ISO 1600, because that ISO option
doesn't exist much like the original poster's question about how to
shoot at ISOs that aren't explicitly provided on camera.

Setting -EV in certain conditions is a basic technique that all good
photographers use at times, because it provides better results at
times. If you did't know that setting -EV is a valuable option under
some conditions, that leads me to believe you don't take many pictures.


For noise comparision purposes, here is the Canon 10D at ISO 1600,
taken in more than double the light of the image linked above...

http://www.pbase.com/imageprocessing/image/31095832/ori...
Anonymous
May 31, 2005 1:05:18 AM

Archived from groups: rec.photo.digital (More info?)

Bart van der Wolf wrote:
> Do you understand Analog Digital Conversion? Apparently not, *that* is
> obvious.

The exposure value is identical at (ISO 400 -2EV) or ISO 1600, the
sensor absorbs the exact same amount of light either way.
Anonymous
May 31, 2005 1:29:38 AM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:

>
> Bart van der Wolf wrote:
>
>>Do you understand Analog Digital Conversion? Apparently not, *that* is
>>obvious.
>
>
> The exposure value is identical at (ISO 400 -2EV) or ISO 1600, the
> sensor absorbs the exact same amount of light either way.


But it's clipped in the Analog Digital Conversion I believe. What you
said before about raw adjustment EV is not entirely true. You can tweak
but there are limits and tradeoffs. In this case, you always want to
overexpose as much as possible unless highlights are being blown it's a
bad idea to underexpose then pull up the shadows because there is more
data in the lighter values & once digitized into darker values, it
cannot be recovered fully.

Noever underexpose unless you are protecting highlights or want a
posterized grainy look.
Anonymous
May 31, 2005 2:03:45 AM

Archived from groups: rec.photo.digital (More info?)

Paul Furman wrote:
> george_preddy@yahoo.com wrote:
>
> >
> > Bart van der Wolf wrote:
> >
> >>Do you understand Analog Digital Conversion? Apparently not, *that* is
> >>obvious.
> >
> >
> > The exposure value is identical at (ISO 400 -2EV) or ISO 1600, the
> > sensor absorbs the exact same amount of light either way.
>
>
> But it's clipped in the Analog Digital Conversion I believe.

The exact same number of photons are registered either way.

> What you
> said before about raw adjustment EV is not entirely true. You can tweak
> but there are limits and tradeoffs.

That's exactly what I siad.

> In this case, you always want to
> overexpose as much as possible unless highlights are being blown it's a
> bad idea to underexpose then pull up the shadows because there is more
> data in the lighter values & once digitized into darker values, it
> cannot be recovered fully.
>
> Noever underexpose unless you are protecting highlights or want a
> posterized grainy look.

SD9 @ ISO 1600...
http://www.pbase.com/isoequiv/image/44078397/original

10D @ ISO 1600 (light is twice as bright)...
http://www.pbase.com/imageprocessing/image/31095832/ori...

How do you suggest taking an ISO 1600 pic with an SD9?
Anonymous
May 31, 2005 2:22:44 AM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:
>>
>>Never underexpose unless you are protecting highlights or want a
>>posterized grainy look.
>
>
> SD9 @ ISO 1600...
> http://www.pbase.com/isoequiv/image/44078397/original


It says ISO 400

>
> How do you suggest taking an ISO 1600 pic with an SD9?


Oh, not an option?
Anonymous
May 31, 2005 4:08:06 AM

Archived from groups: rec.photo.digital (More info?)

<george_preddy@yahoo.com> wrote in message
news:1117483409.649591.309230@g44g2000cwa.googlegroups.com...
>
>
> Bart van der Wolf wrote:
>> <george_preddy@yahoo.com> wrote in message
>> news:1117259952.342255.263170@g14g2000cwa.googlegroups.com...
>> SNIP
>> > Within usable bounds (+/- 2 or 3 stops) finely adjusting the
>> > exposure using a RAW converter is the general equivalent
>> > of tweaking the ISO in tiny increments.
>>
>> Wrong. Decreasing the on camera ISO setting results (if the
>> lightmeter
>> setting is followed) in increased exposure. That means that there
>> are
>> more photons being captured. Increasing the ISO setting will
>> likewise
>> decrease exposure and may affect the analog signal being amplified
>> before quantification (in the Analog Digital Converter).
>>
>> Postprocessing can only work on the already quantified signal
>> numbers
>> (with their inherent quantization noise added to the data before
>> being
>> processed in software). This can lead to more damage to the image.
>>
>> There is nothing that beats the quality of correct exposure, as can
>> be
>> seen in e.g. the difference between a 2 or 3 stops underexposed
>> image
>> adjusted by software, and the real thing, a correct exposure.
>
> The overall EV value is obviously identical using both methods, or
> there wouldn't be exact parameter equivalence. I thought that part
> was
> obvious.

Do you understand Analog Digital Conversion? Apparently not, *that* is
obvious.

Bart
Anonymous
May 31, 2005 5:05:47 PM

Archived from groups: rec.photo.digital (More info?)

"Paul Furman" <paul-@-edgehill.net> wrote in message
news:D MWdnfXKgeE7aQbfRVn-2A@speakeasy.net...
> george_preddy@yahoo.com wrote:
>>>
>>>Never underexpose unless you are protecting highlights or want a
>>>posterized grainy look.
>>
>>
>> SD9 @ ISO 1600...
>> http://www.pbase.com/isoequiv/image/44078397/original
>
>
> It says ISO 400

Correct, and with an exposure correction of -2 stops (otherwise the
dark background would have caused overexposure, since the camera was
multisegment metering in Programmed auto mode).

Bart
Anonymous
May 31, 2005 11:30:46 PM

Archived from groups: rec.photo.digital (More info?)

Bart van der Wolf wrote:
> "Paul Furman" <paul-@-edgehill.net> wrote in message
> news:D MWdnfXKgeE7aQbfRVn-2A@speakeasy.net...
> > george_preddy@yahoo.com wrote:
> >>>
> >>>Never underexpose unless you are protecting highlights or want a
> >>>posterized grainy look.
> >>
> >>
> >> SD9 @ ISO 1600...
> >> http://www.pbase.com/isoequiv/image/44078397/original
> >
> >
> > It says ISO 400
>
> Correct, and with an exposure correction of -2 stops (otherwise the
> dark background would have caused overexposure, since the camera was
> multisegment metering in Programmed auto mode).

Absolutely identical to selecting ISO 1600.
Anonymous
May 31, 2005 11:59:37 PM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:
>
> Bart van der Wolf wrote:
>
>>"Paul Furman" <paul-@-edgehill.net> wrote in message
>>news:D MWdnfXKgeE7aQbfRVn-2A@speakeasy.net...
>>
>>>george_preddy@yahoo.com wrote:
>>>
>>>>>Never underexpose unless you are protecting highlights or want a
>>>>>posterized grainy look.
>>>>
>>>>
>>>>SD9 @ ISO 1600...
>>>>http://www.pbase.com/isoequiv/image/44078397/original
>>>
>>>
>>>It says ISO 400
>>
>>Correct, and with an exposure correction of -2 stops (otherwise the
>>dark background would have caused overexposure, since the camera was
>>multisegment metering in Programmed auto mode).
>
>
> Absolutely identical to selecting ISO 1600.


Did you brighten the image in post processing? No, it was simply made
darker to retain the natural dark setting.
Anonymous
June 1, 2005 1:53:02 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117464027.252643.265790@g47g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>JPS@no.komm wrote:
>> In message <1117257456.223031.274890@f14g2000cwb.googlegroups.com>,
>> george_preddy@yahoo.com wrote:

>>>> >Sheldon wrote:
>> >> True, but could you actually set the ISO to 142? That was my explanation.
>> >> While it might be able to float to 142 in an automatic mode, I've never seen
>> >> a camera that would let you set the ISO to anything but a "normal" number.

>> >Sure you can, just shoot RAW.

>> There are some serious limits that you have to be aware of when you
>> push/pull to other exposure indexes.

>There are serious limits using high ISOs on the camera body as well.

Not really. Unless the support circuitry is defective, or poorly
designed, sensor noise should be a function of exposure index, not ISO
per se.

You generally get worse results at lower ISOs, if you maintain a fixed
exposure (f-stop and shutter speed), because the noise is about the
same, but the lower-ISO shot uses less RAW levels to digitize the analog
data, so it is more highly posterized, as is the signal also.
Posterization often appears as noise, even though it is technically a
distortion.

>Camera processors are very weak compared to desktop computers.

Changing the ISO on the camera is generally not computation; it is
analog amplification of the analog sensor charges. There are exceptions
of course; many camera don't use gain alone for their highest ISO or
two; they simply double (or quadruple) the digitized RAW values, and
that is not a task that requires a powerful CPU.

>So it
>depends on the camera, the scene, the quality of the RAW processor and
>other tools you may use as to which method produces the better end
>result for any given image.

As I've already said, if the camera gives less noise at lower ISO
settings for the same absolute exposure, something is wrong with the
gain circuitry.

>In general, Sigma's DSLRs produce very low noise results when pushing
>ISO during RAW since ISO 100 noise is so much lower than other DSLRs.

I'll agree that the x3f RAW data is low-noise, but it is only low noise
when taken as greyscale data. As soon as you interpolate color from it,
the hue-to-noise level is very low.

Anyway, you don't understand anything at all about how digital cameras
work. I'm not surprised.

>So with the SD9 and SD10, it is almost always better to shoot ISO 100
>-xEV and push during RAW conversion than it is to shoot high ISO.

Not so with my Canons; they give the cleanest results with the highest
ISO that can handle the f-stop and shutter speed without clipping, as it
should be, with a fairly clean gain circuit. Of course, the highest ISO
setting on my 20D is not all gain, and 2x worth of math, as is the two
highest on the 10D.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 1:55:26 AM

Archived from groups: rec.photo.digital (More info?)

In message <429b8ec4$0$24091$e4fe514c@news.xs4all.nl>,
"Bart van der Wolf" <bvdwolf@no.spam> wrote:

>Do you understand Analog Digital Conversion? Apparently not, *that* is
>obvious.

He seems to think that the RAW data is analog, and that the ISO setting
is a parameter that adds noise to post-processed signal, proportional to
the ISO value.

In other words, he understands nothing about digital capture.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 1:57:55 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117512318.889844.43540@g44g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>Bart van der Wolf wrote:
>> Do you understand Analog Digital Conversion? Apparently not, *that* is
>> obvious.

>The exposure value is identical at (ISO 400 -2EV) or ISO 1600, the
>sensor absorbs the exact same amount of light either way.

This is correct, but you don't get to use the sensor's analog data; you
only get it after it's been digitized, and the ISO 1600 RAW digitized
data is going to have 4x as many levels for each channel, and up to 64x
as many possible RGB colors.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 2:04:47 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117515825.367150.164960@g44g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>The exact same number of photons are registered either way.

Yep, but you never get to count all the electrons with your lower ISO.

Each RAW level at ISO 100 might represent 16 electrons; each RAW level
at ISO 400 will then represent 4 electrons; 1 electron at ISO 1600.
Which is more accurate?

It is remotely possible that your camera is really that bad, that
under-exposing for an equivalent EI gives better results, but I tend to
think that you're just mouthing off, and have never really tried this.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 2:14:21 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117515825.367150.164960@g44g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>SD9 @ ISO 1600...
>http://www.pbase.com/isoequiv/image/44078397/original

>10D @ ISO 1600 (light is twice as bright)...
>http://www.pbase.com/imageprocessing/image/31095832/ori...

The 10D image is grossly under-exposed. BTW, the 10D's "1600" is 8x
gain (relative to ISO 100), and 2x "math".

Your SD9 image has no midtones, as rendered; just black and highlights.
This really isn't an exposure index of 1600, because you left what
should be the midtones as nearly black; I just altered the gamma in PS,
and all kinds of big, blotchy color noise, and fine regular noise popped
up.

>How do you suggest taking an ISO 1600 pic with an SD9?

With the highest ISO setting.

--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 2:14:22 AM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:

> ISO 200 -1EV (pushed back in RAW) yields a superb ISO 400,
> unlike the noiser explicit ISO 400.


No, you've got it backwards. Try it and you will see. If it's not
obvious (100-400 all looks good on a canon) apply a contrast increase to
the true 400 & faked 400 and you will see. Try a 800 & drop it down to
match and that will look even better.
Anonymous
June 1, 2005 2:16:40 AM

Archived from groups: rec.photo.digital (More info?)

In message <429c4567$0$17211$e4fe514c@news.xs4all.nl>,
"Bart van der Wolf" <bvdwolf@no.spam> wrote:

>Correct, and with an exposure correction of -2 stops (otherwise the
>dark background would have caused overexposure, since the camera was
>multisegment metering in Programmed auto mode).

And from the looks of it, he didn't push it back in PP. This is, for
all intents and purposes, an EI of no more than 400 or so.

I just pushed it so that the bridge is a midtone, and it has big green
splotches all over it, and there is noise everywhere except in the
highlights.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 2:18:45 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117483578.722834.41610@g43g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>JPS@no.komm wrote:
>> In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
>> george_preddy@yahoo.com wrote:
>>
>> >For example, you might set ISO 100 with a -2EV camera exposure
>> >adjustment, then push the exposure +2EV using the RAW converter.
>> >Depending on the camera and the scene, you could get better overall
>> >results than shooting ISO 400.
>>
>> That would require a defective or low-quality amplifier, or imply that
>> the camera was already doing the same, but discarding highlights.

>Or, that a desktop computer has more computing power available than a
>camera.

There is no computing in an analog amplifier, Mr. Clueless, and when
gain is simulated with math, you can not improve on the quality of a bit
shift with a desktop.

You apparently understand none of this at all.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 2:29:05 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117484729.737178.322670@z14g2000cwz.googlegroups.com>,
george_preddy@yahoo.com wrote:

>
>
>JPS@no.komm wrote:
>> In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
>> george_preddy@yahoo.com wrote:
>>
>> >For example, you might set ISO 100 with a -2EV camera exposure
>> >adjustment, then push the exposure +2EV using the RAW converter.
>> >Depending on the camera and the scene, you could get better overall
>> >results than shooting ISO 400.
>>
>> That would require a defective or low-quality amplifier, or imply that
>> the camera was already doing the same, but discarding highlights.
>
>The camera doesn't have to be defective to shoot advantageously at -xEV
>under some conditions, people who know what they are doing use the
>option all the time.

Of course they do; they want to preserve highlights. However, you
sacrifice the quality of midtones and shadows, by doing so.

>Here is an Sigma SD9 photo at ISO 1600...
>
>http://www.pbase.com/isoequiv/image/44078397/large

No, you did not bring the exposure back up 2 stops in RAW conversion or
PP, so it is still effectively an ISO 400, as far as visible noise is
concerned.

>It is much better than the camera's ISO 1600, because that ISO option
>doesn't exist much like the original poster's question about how to
>shoot at ISOs that aren't explicitly provided on camera.

To call it ISO 1600, you need to quadruple the linear luminance of the
midtones, otherwise, it is simply exposure compensation used to keep the
image in a dark key.

>Setting -EV in certain conditions is a basic technique that all good
>photographers use at times, because it provides better results at
>times. If you did't know that setting -EV is a valuable option under
>some conditions, that leads me to believe you don't take many pictures.

I never said that it wasn't useful. What i said is that you do not get
less noise by using a lower ISO for the same absolute sensor exposure.

All subtlety is lost on mister schizophrenic know-it-all.

>For noise comparision purposes, here is the Canon 10D at ISO 1600,
>taken in more than double the light of the image linked above...

>http://www.pbase.com/imageprocessing/image/31095832/ori...

You think that we're really stupid enough to fall for this?

Try the same scene, idiot. The Canon shot has the subject as a shadow,
and the Sigma has the subject as highlights that would have washed out
had you not used -2 EC. You are truly pathetic.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 2:29:59 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117483508.462799.197750@o13g2000cwo.googlegroups.com>,
george_preddy@yahoo.com wrote:

>You'll find that 99% of photographers don't understand the most basic
>issues of digital photography.

That's probably not far off; I'll say 90%, and you are clearly one of
them.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 1, 2005 4:05:48 PM

Archived from groups: rec.photo.digital (More info?)

<george_preddy@yahoo.com> wrote in message
news:1117593046.582530.62410@g47g2000cwa.googlegroups.com...
>
> Bart van der Wolf wrote:
>> "Paul Furman" <paul-@-edgehill.net> wrote in message
>> news:D MWdnfXKgeE7aQbfRVn-2A@speakeasy.net...
SNIP
>> > It says ISO 400
>>
>> Correct, and with an exposure correction of -2 stops (otherwise
>> the dark background would have caused overexposure, since the
>> camera was multisegment metering in Programmed auto mode).
>
> Absolutely identical to selecting ISO 1600.

Of course not. If you could have selected ISO 1600, the analog (!)
signal would have been amplified in the ADC stage. Since you couldn't
select ISO 1600, but 400 instead, the amplification was less. That
produced lower luminance to be recorded in the Raw data (which
compensated for the metering of a lot of darkness, and avoided
overexposure).

Bart
Anonymous
June 2, 2005 1:48:56 AM

Archived from groups: rec.photo.digital (More info?)

Paul Furman wrote:

> >>Correct, and with an exposure correction of -2 stops (otherwise the
> >>dark background would have caused overexposure, since the camera was
> >>multisegment metering in Programmed auto mode).
> >
> >
> > Absolutely identical to selecting ISO 1600.
>
> Did you brighten the image in post processing? No, it was simply made
> darker to retain the natural dark setting.

Yes, it was pushed +2 stops in RAW.
Anonymous
June 2, 2005 1:49:53 AM

Archived from groups: rec.photo.digital (More info?)

Bart van der Wolf wrote:
> <george_preddy@yahoo.com> wrote in message
> news:1117593046.582530.62410@g47g2000cwa.googlegroups.com...
> >
> > Bart van der Wolf wrote:
> >> "Paul Furman" <paul-@-edgehill.net> wrote in message
> >> news:D MWdnfXKgeE7aQbfRVn-2A@speakeasy.net...
> SNIP
> >> > It says ISO 400
> >>
> >> Correct, and with an exposure correction of -2 stops (otherwise
> >> the dark background would have caused overexposure, since the
> >> camera was multisegment metering in Programmed auto mode).
> >
> > Absolutely identical to selecting ISO 1600.
>
> Of course not.

Yes it is. The exposure is identical.
Anonymous
June 2, 2005 1:57:17 AM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117515825.367150.164960@g44g2000cwa.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
> >The exact same number of photons are registered either way.
>
> Yep, but you never get to count all the electrons with your lower ISO.
>
> Each RAW level at ISO 100 might represent 16 electrons;

Wrong, it's at least 23 electrons.
Anonymous
June 2, 2005 2:08:55 AM

Archived from groups: rec.photo.digital (More info?)

Paul Furman wrote:
> george_preddy@yahoo.com wrote:
>
> > ISO 200 -1EV (pushed back in RAW) yields a superb ISO 400,
> > unlike the noiser explicit ISO 400.
>
>
> No, you've got it backwards. Try it and you will see. If it's not
> obvious (100-400 all looks good on a canon) apply a contrast increase to
> the true 400 & faked 400 and you will see. Try a 800 & drop it down to
> match and that will look even better.

Who said anything about Canon?
Anonymous
June 2, 2005 2:10:51 AM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117483578.722834.41610@g43g2000cwa.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
> >JPS@no.komm wrote:
> >> In message <1117259952.342255.263170@g14g2000cwa.googlegroups.com>,
> >> george_preddy@yahoo.com wrote:
> >>
> >> >For example, you might set ISO 100 with a -2EV camera exposure
> >> >adjustment, then push the exposure +2EV using the RAW converter.
> >> >Depending on the camera and the scene, you could get better overall
> >> >results than shooting ISO 400.
> >>
> >> That would require a defective or low-quality amplifier, or imply that
> >> the camera was already doing the same, but discarding highlights.
>
> >Or, that a desktop computer has more computing power available than a
> >camera.
>
> There is no computing

I think even you know how ridiculous that is.
Anonymous
June 2, 2005 2:13:03 AM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> >Setting -EV in certain conditions is a basic technique that all good
> >photographers use at times, because it provides better results at
> >times. If you did't know that setting -EV is a valuable option under
> >some conditions, that leads me to believe you don't take many pictures.
>
> I never said that it wasn't useful.

Great, you agree with me. Next time, agree up front so you don't waste
everyone's time.
Anonymous
June 2, 2005 4:59:49 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117593046.582530.62410@g47g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>Absolutely identical to selecting ISO 1600.

Nope. There are 4x as many RAW values to express the levels in each
channel with ISO 1600 due to gain than there is at ISO 400 and -2 EC.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 2, 2005 1:13:26 PM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:

>
> Paul Furman wrote:
>
>
>>>>Correct, and with an exposure correction of -2 stops (otherwise the
>>>>dark background would have caused overexposure, since the camera was
>>>>multisegment metering in Programmed auto mode).
>>>
>>>
>>>Absolutely identical to selecting ISO 1600.
>>
>>Did you brighten the image in post processing? No, it was simply made
>>darker to retain the natural dark setting.
>
>
> Yes, it was pushed +2 stops in RAW.

Oh so it is your photo. Hard to tell what's going on in a shot that's
mostly black but I see a heck of a lot of noise in the big lady's red
suit, for example and the green splotches on the bridge are bizzare.

Try a test like this:
<http://www.edgehill.net/1/?SC=go.php&DIR=Misc/photograp...;
Anonymous
June 2, 2005 1:16:07 PM

Archived from groups: rec.photo.digital (More info?)

george_preddy@yahoo.com wrote:

>
> JPS@no.komm wrote:
>
>>>Setting -EV in certain conditions is a basic technique that all good
>>>photographers use at times, because it provides better results at
>>>times. If you did't know that setting -EV is a valuable option under
>>>some conditions, that leads me to believe you don't take many pictures.
>>
>>I never said that it wasn't useful.
>
>
> Great, you agree with me. Next time, agree up front so you don't waste
> everyone's time.


It never makes sense to underexpose then push in post processing unless
you can't get the picture due to low light or are protecting highlights
from blowing out. Even then, boosting ISO would be the fist thing to try.
Anonymous
June 3, 2005 12:42:14 AM

Archived from groups: rec.photo.digital (More info?)

In message <1117687736.018440.157350@g44g2000cwa.googlegroups.com>,
george_preddy@yahoo.com wrote:

>Paul Furman wrote:

>> >>Correct, and with an exposure correction of -2 stops (otherwise the
>> >>dark background would have caused overexposure, since the camera was
>> >>multisegment metering in Programmed auto mode).

>> > Absolutely identical to selecting ISO 1600.

>> Did you brighten the image in post processing? No, it was simply made
>> darker to retain the natural dark setting.

>Yes, it was pushed +2 stops in RAW.

In the unlikely event that this is true, I would say that your camera's
metering is very strange then, indeed.

The median luminance of the image is 6; the mean is 24.78, both out of
255.

You are using the ambiguity of exposure to slip in an imaginary benefit
for your SD9.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 3, 2005 12:42:15 AM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1117687736.018440.157350@g44g2000cwa.googlegroups.com>,
> george_preddy@yahoo.com wrote:
>
>
>>Paul Furman wrote:
>
>
>>>>>Correct, and with an exposure correction of -2 stops (otherwise the
>>>>>dark background would have caused overexposure, since the camera was
>>>>>multisegment metering in Programmed auto mode).
>
>
>>>>Absolutely identical to selecting ISO 1600.
>
>
>>>Did you brighten the image in post processing? No, it was simply made
>>>darker to retain the natural dark setting.
>
>
>>Yes, it was pushed +2 stops in RAW.
>
>
> In the unlikely event that this is true, I would say that your camera's
> metering is very strange then, indeed.
>
> The median luminance of the image is 6; the mean is 24.78, both out of
> 255.
>
> You are using the ambiguity of exposure to slip in an imaginary benefit
> for your SD9.


It would have to have been shot way below -2 to produce a mostly black
photo with +2 compensation in raw conversion and the exif data says it
was shot at simply -2. That should remove any doubt.


--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
!