Always shoot at highest megapixel setting?

Paul Furman <paul-.RemoveThis@-edgehill.net> wrote:
> David J Taylor wrote:
>>
>> To be honest, as soon as you said "web pages" I immediately thought "shoot
>> at the lowest resolution and lowest quality setting (i.e. smallest file
>> size). Depending on your exact needs, that may be more than adequate.
>> Generally, for Web use, it's better to produce the fastest loading (i.e.
>> smallest file size) you can, even if a few JPEG artefacts are visible. If
>> these is any chance you will want to do something else with the photos,
>> that's different.

> If you are going to make adjustments to brightness & contrast there may
> be some benefit to shooting higher quality but generally web images are
> *vastly* smaller than the smallest image you could get out of any modern
> camera. The final use on the web page could make a significant difference.

Depends on what is meant by "web image". That can range from a tiny
illustrative image the size of a large postage stamp, to a full screen
image. A Full screen image is typically approx 1200 pixels in width,
which is only a few times smaller than the native resolution of a good
digital camera. I would say that unless you were taking great care in
the photography, and using the very best of lenses, for an image with
highest quality at the highest resolution you will often want to cut
the out-of-camera size down by 50%. Then if you also want to do some
cropping, you can find yourself getting close to full screen
resolution.

What is more, sharpening and other local contrast manipulations are
best done at the target resolution, so they are best done at the last
stage of editing rather than in camera.

In other words, if what you mean by "web image" is a screen filling
high quality image, then you probably want to set your camera to its
highest resolution and lowest processing options, and do the resizing,
sharpening, etc., in your image editor.

But if what you mean by "web image" is something like a passport
photograph, you're probably wasting your time using a camera. Your
mobile phone is probably good enough

--
Chris Malcolm cam.RemoveThis@infirmatics.ed.ac.uk DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]
 >> Stay informed about: Always shoot at highest megapixel setting? 

John Navas <spamfilter1 RemoveThis @navasgroup.com> wrote:
> On Wed, 13 Feb 2008 11:41:06 -0800 (PST), Rob <robbelljr RemoveThis @aol.com> wrote
> in <e4af1acc-f41a-40ad-af9f-81b361ff9d39 RemoveThis @b2g2000hsg.googlegroups.com>:

>>I understand your statement that sometimes using higher resolutions
>>with greater compression could produce a clearer "native" result, but
>>assuming a production environment of photos for display in web pages
>>it is usally practical to perform at least some post-processing
>>reduction. Under those circumstances the level of compression applied
>>by the camera seems might be more important than the resolution at
>>which the photos were shot. In other words, I may have asked the
>>wrong question to begin with.

> The level of compression and the amount of downsampling are two entirely
> different things. The former is best down in post-processing, but the
> latter is usually done better in camera, because the camera is working
> from the RAW sensor output, not a converted image.

I think it depends on what kind of camera output you're working
with. If you're working with RAW output, then doing it in editor
should always be better. If you're doing it on the camera jpeg output,
it depends how good that is. Some cameras which can output RAW, for
example, assume that's what you'll be using for high quality work, and
produce rather inferior jpegs. Some of the undeservedly bad reputation
that jpegs have got in the "always RAW for quality" crowd are due to
the inferior jpegs those cameras produce. In that case letting the
camera do the compression and downsampling will give better
quality. But if the camera is capable of high enough quality jpegs,
which some certainly are, then you'll get better quality downsampling
and compression in a good editor than in the camera.

--
Chris Malcolm cam RemoveThis @infirmatics.ed.ac.uk DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]
 >> Stay informed about: Always shoot at highest megapixel setting? 

On 14 Feb 2008 10:31:19 GMT, Chris Malcolm <cam.TakeThisOut@holyrood.ed.ac.uk> wrote
in <61ijjnF1phbneU3.TakeThisOut@mid.individual.net>:

>John Navas <spamfilter1.TakeThisOut@navasgroup.com> wrote:
>> On Wed, 13 Feb 2008 11:41:06 -0800 (PST), Rob <robbelljr.TakeThisOut@aol.com> wrote
>> in <e4af1acc-f41a-40ad-af9f-81b361ff9d39.TakeThisOut@b2g2000hsg.googlegroups.com>:
>
>>>I understand your statement that sometimes using higher resolutions
>>>with greater compression could produce a clearer "native" result, but
>>>assuming a production environment of photos for display in web pages
>>>it is usally practical to perform at least some post-processing
>>>reduction. Under those circumstances the level of compression applied
>>>by the camera seems might be more important than the resolution at
>>>which the photos were shot. In other words, I may have asked the
>>>wrong question to begin with.
>
>> The level of compression and the amount of downsampling are two entirely
>> different things. The former is best down in post-processing, but the
>> latter is usually done better in camera, because the camera is working
>> from the RAW sensor output, not a converted image.
>
>I think it depends on what kind of camera output you're working
>with. If you're working with RAW output, then doing it in editor
>should always be better. If you're doing it on the camera jpeg output,
>it depends how good that is. Some cameras which can output RAW, for
>example, assume that's what you'll be using for high quality work, and
>produce rather inferior jpegs. Some of the undeservedly bad reputation
>that jpegs have got in the "always RAW for quality" crowd are due to
>the inferior jpegs those cameras produce. In that case letting the
>camera do the compression and downsampling will give better
>quality. But if the camera is capable of high enough quality jpegs,
>which some certainly are, then you'll get better quality downsampling
>and compression in a good editor than in the camera.

I wrote downsampling is best done in camera, not compression, assuming
JPEG output.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
 >> Stay informed about: Always shoot at highest megapixel setting? 

John Navas <spamfilter1.RemoveThis@navasgroup.com> wrote:
> On 14 Feb 2008 10:31:19 GMT, Chris Malcolm <cam.RemoveThis@holyrood.ed.ac.uk> wrote
> in <61ijjnF1phbneU3.RemoveThis@mid.individual.net>:

>>John Navas <spamfilter1.RemoveThis@navasgroup.com> wrote:
>>> On Wed, 13 Feb 2008 11:41:06 -0800 (PST), Rob <robbelljr.RemoveThis@aol.com> wrote
>>> in <e4af1acc-f41a-40ad-af9f-81b361ff9d39.RemoveThis@b2g2000hsg.googlegroups.com>:
>>
>>>>I understand your statement that sometimes using higher resolutions
>>>>with greater compression could produce a clearer "native" result, but
>>>>assuming a production environment of photos for display in web pages
>>>>it is usally practical to perform at least some post-processing
>>>>reduction. Under those circumstances the level of compression applied
>>>>by the camera seems might be more important than the resolution at
>>>>which the photos were shot. In other words, I may have asked the
>>>>wrong question to begin with.
>>
>>> The level of compression and the amount of downsampling are two entirely
>>> different things. The former is best down in post-processing, but the
>>> latter is usually done better in camera, because the camera is working
>>> from the RAW sensor output, not a converted image.
>>
>>I think it depends on what kind of camera output you're working
>>with. If you're working with RAW output, then doing it in editor
>>should always be better. If you're doing it on the camera jpeg output,
>>it depends how good that is. Some cameras which can output RAW, for
>>example, assume that's what you'll be using for high quality work, and
>>produce rather inferior jpegs. Some of the undeservedly bad reputation
>>that jpegs have got in the "always RAW for quality" crowd are due to
>>the inferior jpegs those cameras produce. In that case letting the
>>camera do the compression and downsampling will give better
>>quality. But if the camera is capable of high enough quality jpegs,
>>which some certainly are, then you'll get better quality downsampling
>>and compression in a good editor than in the camera.

> I wrote downsampling is best done in camera, not compression, assuming
> JPEG output.

Assuming poor jpeg output. Otherwise I can't think of any reason why
that should be so. More sophisticated algorithms, and the processor
power to run them, are available in editors, so I'd expect the best of
them to do a better job, provided the camera in question can output a
good enough quality jpeg to begin with, which at least wasn't always
the case with certain past models of DSLR.

--
Chris Malcolm
 >> Stay informed about: Always shoot at highest megapixel setting? 

On 15 Feb 2008 12:12:44 GMT, Chris Malcolm <cam.DeleteThis@holyrood.ed.ac.uk> wrote
in <61ldtsF1uv6mdU2.DeleteThis@mid.individual.net>:

>John Navas <spamfilter1.DeleteThis@navasgroup.com> wrote:

>> I wrote downsampling is best done in camera, not compression, assuming
>> JPEG output.
>
>Assuming poor jpeg output. Otherwise I can't think of any reason why
>that should be so. More sophisticated algorithms, and the processor
>power to run them, are available in editors, so I'd expect the best of
>them to do a better job, provided the camera in question can output a
>good enough quality jpeg to begin with, which at least wasn't always
>the case with certain past models of DSLR.

The differences are that:

(a) the camera is downsampling from RAW sensor output, which gives it a
big leg up on an editor working only from out of camera JPEG; and

(b) the JPEG compression algorithms in many digital cameras are
relatively crude, presumably due to limited time and processor power.
(I've tested them against PC software.)

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
 >> Stay informed about: Always shoot at highest megapixel setting? 

John Navas <spamfilter1 RemoveThis @navasgroup.com> wrote:
> On 15 Feb 2008 12:12:44 GMT, Chris Malcolm <cam RemoveThis @holyrood.ed.ac.uk> wrote
> in <61ldtsF1uv6mdU2 RemoveThis @mid.individual.net>:

>>John Navas <spamfilter1 RemoveThis @navasgroup.com> wrote:

>>> I wrote downsampling is best done in camera, not compression, assuming
>>> JPEG output.
>>
>>Assuming poor jpeg output. Otherwise I can't think of any reason why
>>that should be so. More sophisticated algorithms, and the processor
>>power to run them, are available in editors, so I'd expect the best of
>>them to do a better job, provided the camera in question can output a
>>good enough quality jpeg to begin with, which at least wasn't always
>>the case with certain past models of DSLR.

> The differences are that:

> (a) the camera is downsampling from RAW sensor output, which gives it a
> big leg up on an editor working only from out of camera JPEG; and

There should be little difference if there's little detail lost
between the RAW and the jpeg, which is the case in the the highest
quality lowest compression jpegs.

> (b) the JPEG compression algorithms in many digital cameras are
> relatively crude, presumably due to limited time and processor power.
> (I've tested them against PC software.)

The crudity of jpeg compression is an adjustable factor. Many digital
cameras, especially early ones, compressed a lot more than modern
ones. What matters is whether the same level of compression produced
different results in image quality depending on whether it was done in
camera or in a computer image editor. Did you make your comparisons at
the same level of compression?

--
Chris Malcolm cam RemoveThis @infirmatics.ed.ac.uk DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]
 >> Stay informed about: Always shoot at highest megapixel setting? 

On 16 Feb 2008 04:04:24 GMT, Chris Malcolm <cam.DeleteThis@holyrood.ed.ac.uk> wrote
in <61n5m8F1vo3ihU2.DeleteThis@mid.individual.net>:

>John Navas <spamfilter1.DeleteThis@navasgroup.com> wrote:
>> On 15 Feb 2008 12:12:44 GMT, Chris Malcolm <cam.DeleteThis@holyrood.ed.ac.uk> wrote
>> in <61ldtsF1uv6mdU2.DeleteThis@mid.individual.net>:
>
>>>John Navas <spamfilter1.DeleteThis@navasgroup.com> wrote:

>> The differences are that:
>
>> (a) the camera is downsampling from RAW sensor output, which gives it a
>> big leg up on an editor working only from out of camera JPEG; and
>
>There should be little difference if there's little detail lost
>between the RAW and the jpeg, which is the case in the the highest
>quality lowest compression jpegs.

Direct downsampling from RAW is a totally different algorithm from
downsampling from converted JPEG. It's not amount of compression that
makes the difference, but working from RAW sensor output. The only way
PC software could do as well or better would be to use RAW output, not
JPEG.

>> (b) the JPEG compression algorithms in many digital cameras are
>> relatively crude, presumably due to limited time and processor power.
>> (I've tested them against PC software.)
>
>The crudity of jpeg compression is an adjustable factor. Many digital
>cameras, especially early ones, compressed a lot more than modern
>ones. What matters is whether the same level of compression produced
>different results in image quality depending on whether it was done in
>camera or in a computer image editor. Did you make your comparisons at
>the same level of compression?

My comparisons were file size for comparable quality. Photoshop beat
all cameras I tested by 20-30%. In other words, camera file size at
highest quality setting is larger than it needs to be and thus quality
not as good as it could be, and at a given file size quality is not as
good as PC software compression. No disrespect to camera firmware,
because it has to work very fast with much less processing power. It's
one reason for using RAW when that's an available option, but the
downside of RAW is no in-camera correction of lens defects.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
 >> Stay informed about: Always shoot at highest megapixel setting? 

John Navas <spamfilter1 DeleteThis @navasgroup.com> wrote:
> On 16 Feb 2008 04:04:24 GMT, Chris Malcolm <cam DeleteThis @holyrood.ed.ac.uk> wrote
> in <61n5m8F1vo3ihU2 DeleteThis @mid.individual.net>:

>>John Navas <spamfilter1 DeleteThis @navasgroup.com> wrote:
>>> On 15 Feb 2008 12:12:44 GMT, Chris Malcolm <cam DeleteThis @holyrood.ed.ac.uk> wrote
>>> in <61ldtsF1uv6mdU2 DeleteThis @mid.individual.net>:
>>
>>>>John Navas <spamfilter1 DeleteThis @navasgroup.com> wrote:

>>> The differences are that:
>>
>>> (a) the camera is downsampling from RAW sensor output, which gives it a
>>> big leg up on an editor working only from out of camera JPEG; and
>>
>>There should be little difference if there's little detail lost
>>between the RAW and the jpeg, which is the case in the the highest
>>quality lowest compression jpegs.

> Direct downsampling from RAW is a totally different algorithm from
> downsampling from converted JPEG.

If the software has been properly written by a competent professional
that should not be the case, because the first thing to be done should
be the mapping of the presented image data, either RAW or jpeg, into a
simple field of pixels indexed directly by position in the image. Then
exactly the same downsampling algorithm can be applied to both
images. Of course remapping the downsampled image back to the original
format can be done losslessly for RAW, but not losslessly for jpeg,
although at high levels of jpeg quality the losses should be pretty
hard to detect visually.

If a totally different downsampling algorithm is used because the
downsampling is mixed up with the different manner of presentation of
the image between RAW and jpeg, then that is extremely incomptetent
software design which both invites errors and makes them hard to find
and fix.

That's not to say that such a thing has not happened. But I'd be very
dismayed indeed to find that it had happened, and if I found out that
any camera company was using software programmers of that degree of
incompetence I'd become extremely suspicious of that company in general.

> It's not amount of compression that
> makes the difference, but working from RAW sensor output. The only way
> PC software could do as well or better would be to use RAW output, not
> JPEG.

>>> (b) the JPEG compression algorithms in many digital cameras are
>>> relatively crude, presumably due to limited time and processor power.
>>> (I've tested them against PC software.)
>>
>>The crudity of jpeg compression is an adjustable factor. Many digital
>>cameras, especially early ones, compressed a lot more than modern
>>ones. What matters is whether the same level of compression produced
>>different results in image quality depending on whether it was done in
>>camera or in a computer image editor. Did you make your comparisons at
>>the same level of compression?

> My comparisons were file size for comparable quality. Photoshop beat
> all cameras I tested by 20-30%. In other words, camera file size at
> highest quality setting is larger than it needs to be and thus quality
> not as good as it could be,

I don't understand the "thus" step in your argument. From wht I know
of jpeg implementations that does not necessarily follow. It certainly
could be the case in a particular instance, but whether it was would
have to be decided by experiment.

> and at a given file size quality is not as
> good as PC software compression.

That's a well known difference between different jpeg
implementations. It's also the case that some kinds of operations on
jpeg images can be done with computational economy on the jpeg data,
leaving a poor compression which is larger than it needs to be at that
quality, or it can be recompressed to reduce the file size. When
working at high levels of jpeg quality there should be little if any
difference in image quality between those two options.

> No disrespect to camera firmware,
> because it has to work very fast with much less processing power. It's
> one reason for using RAW when that's an available option, but the
> downside of RAW is no in-camera correction of lens defects.

I know that the effects you describe exist. What I'm disputing is
whether they necessarily exist in visible form in good implementations
when using high quality large size jpeg compressions. I suspect that
at least some modern cameras can do it properly, by which I mean if
there are any visible jpeg losses of detail, they will only be
discoverable by determined pixel peeping at specially chosen awkward
cases.

Of course there will always be the well known RAW-2-jpeg losses in
dynamic range and colour representation, but that's a general feature
of the difference between RAW and jpeg and not something that would
specifically happen as a result of downsampling.

So I suupose. It's possible, howver, that I'm ignorant of some
important quirk of jpeggery or some specific folly of digital camera
makers

--
Chris Malcolm cam DeleteThis @infirmatics.ed.ac.uk DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]
 >> Stay informed about: Always shoot at highest megapixel setting? 

On 19 Feb 2008 12:04:45 GMT, Chris Malcolm <cam DeleteThis @holyrood.ed.ac.uk> wrote
in <61vuutF20ubkfU1 DeleteThis @mid.individual.net>:

>John Navas <spamfilter1 DeleteThis @navasgroup.com> wrote:

>> Direct downsampling from RAW is a totally different algorithm from
>> downsampling from converted JPEG.
>
>If the software has been properly written by a competent professional
>that should not be the case, because the first thing to be done should
>be the mapping of the presented image data, either RAW or jpeg, into a
>simple field of pixels indexed directly by position in the image. Then
>exactly the same downsampling algorithm can be applied to both
>images. Of course remapping the downsampled image back to the original
>format can be done losslessly for RAW, but not losslessly for jpeg,
>although at high levels of jpeg quality the losses should be pretty
>hard to detect visually.
>
>If a totally different downsampling algorithm is used because the
>downsampling is mixed up with the different manner of presentation of
>the image between RAW and jpeg, then that is extremely incomptetent
>software design which both invites errors and makes them hard to find
>and fix.
>
>That's not to say that such a thing has not happened. But I'd be very
>dismayed indeed to find that it had happened, and if I found out that
>any camera company was using software programmers of that degree of
>incompetence I'd become extremely suspicious of that company in general.

Technical documentation I've seen suggests that downsampling from RAW is
different from and superior to downsampling dematrixed and anti-alised
output, no matter what the compression quality, and not something that
would be considered bad programming practice -- just the opposite.

>> My comparisons were file size for comparable quality. Photoshop beat
>> all cameras I tested by 20-30%. In other words, camera file size at
>> highest quality setting is larger than it needs to be and thus quality
>> not as good as it could be,
>
>I don't understand the "thus" step in your argument. From wht I know
>of jpeg implementations that does not necessarily follow. It certainly
>could be the case in a particular instance, but whether it was would
>have to be decided by experiment.

My "thus" means that quality could be better at a given file size, or
that quality could be the same at a significantly smaller file size.

>> and at a given file size quality is not as
>> good as PC software compression.
>
>That's a well known difference between different jpeg
>implementations. It's also the case that some kinds of operations on
>jpeg images can be done with computational economy on the jpeg data,
>leaving a poor compression which is larger than it needs to be at that
>quality, or it can be recompressed to reduce the file size. When
>working at high levels of jpeg quality there should be little if any
>difference in image quality between those two options.

I can usually see the difference, comparing RAW output converted to JPEG
in Photoshop (with no other processing) to camera JPEG output at the
same file size.

>> No disrespect to camera firmware,
>> because it has to work very fast with much less processing power. It's
>> one reason for using RAW when that's an available option, but the
>> downside of RAW is no in-camera correction of lens defects.
>
>I know that the effects you describe exist. What I'm disputing is
>whether they necessarily exist in visible form in good implementations
>when using high quality large size jpeg compressions. I suspect that
>at least some modern cameras can do it properly, by which I mean if
>there are any visible jpeg losses of detail, they will only be
>discoverable by determined pixel peeping at specially chosen awkward
>cases.

Each of us must decide what level of quality we want, and different
cameras are of course different. I'm personally a bit disappointed in
the quality of "high quality" JPEG output from some cameras (e.g., my
Olympus D-40Z), but not all cameras (e.g., my Panasonic DMC-FZ8, which
is noticeably better than my previous DMC-FZ5).

>Of course there will always be the well known RAW-2-jpeg losses in
>dynamic range and colour representation, but that's a general feature
>of the difference between RAW and jpeg and not something that would
>specifically happen as a result of downsampling.

Of course. Not what I'm talking about. When testing, I try to be
careful to avoid such issues.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)
 >> Stay informed about: Always shoot at highest megapixel setting? 

John Navas <spamfilter1 DeleteThis @navasgroup.com> wrote:
> On 19 Feb 2008 12:04:45 GMT, Chris Malcolm <cam DeleteThis @holyrood.ed.ac.uk> wrote
> in <61vuutF20ubkfU1 DeleteThis @mid.individual.net>:

>>John Navas <spamfilter1 DeleteThis @navasgroup.com> wrote:

>>> Direct downsampling from RAW is a totally different algorithm from
>>> downsampling from converted JPEG.
>>
>>If the software has been properly written by a competent professional
>>that should not be the case, because the first thing to be done should
>>be the mapping of the presented image data, either RAW or jpeg, into a
>>simple field of pixels indexed directly by position in the image. Then
>>exactly the same downsampling algorithm can be applied to both
>>images. Of course remapping the downsampled image back to the original
>>format can be done losslessly for RAW, but not losslessly for jpeg,
>>although at high levels of jpeg quality the losses should be pretty
>>hard to detect visually.
>>
>>If a totally different downsampling algorithm is used because the
>>downsampling is mixed up with the different manner of presentation of
>>the image between RAW and jpeg, then that is extremely incomptetent
>>software design which both invites errors and makes them hard to find
>>and fix.
>>
>>That's not to say that such a thing has not happened. But I'd be very
>>dismayed indeed to find that it had happened, and if I found out that
>>any camera company was using software programmers of that degree of
>>incompetence I'd become extremely suspicious of that company in general.

> Technical documentation I've seen suggests that downsampling from RAW is
> different from and superior to downsampling dematrixed and anti-alised
> output, no matter what the compression quality, and not something that
> would be considered bad programming practice -- just the opposite.

We're not talking about the same thing. You're referring to whether or
not the downsampling is done from a RAW file. I'm talking about how
the program which does that is designed; there's a well known good way
and a well known bad way, and saying that a different algorithm is
employed does sound like a reference to the bad way of doing it,
although it's too vague to be completely sure.

>>> My comparisons were file size for comparable quality. Photoshop beat
>>> all cameras I tested by 20-30%. In other words, camera file size at
>>> highest quality setting is larger than it needs to be and thus quality
>>> not as good as it could be,
>>
>>I don't understand the "thus" step in your argument. From wht I know
>>of jpeg implementations that does not necessarily follow. It certainly
>>could be the case in a particular instance, but whether it was would
>>have to be decided by experiment.

> My "thus" means that quality could be better at a given file size, or
> that quality could be the same at a significantly smaller file size.

So you're using "quality" in a more general sense than I am. From my
point of view, I regard two different jpeg files which give rise to
visibly indistinguishable images to be of equivalent quality, even
though one file is twice the size of the other. I agree that the
smaller file was produced by a higher quality of compression, but that
difference hasn't affected image quality.

>>> and at a given file size quality is not as
>>> good as PC software compression.
>>
>>That's a well known difference between different jpeg
>>implementations. It's also the case that some kinds of operations on
>>jpeg images can be done with computational economy on the jpeg data,
>>leaving a poor compression which is larger than it needs to be at that
>>quality, or it can be recompressed to reduce the file size. When
>>working at high levels of jpeg quality there should be little if any
>>difference in image quality between those two options.

> I can usually see the difference, comparing RAW output converted to JPEG
> in Photoshop (with no other processing) to camera JPEG output at the
> same file size.

I'm not surprised you find such a difference, because it's not
surprising that PC software would be more efficient at the compression
than the camera. I do suspect that some cameras are capable of
producing jpeg image quality which can't be much bettered by PC
software, even though the jpeg file sizes might be bigger.

In deciding how much to use RAW with my Sony R1 I checked out a number
of test images very carefully. My conclusion was that if I turned the
camera jpeg processing options down to a minimum there was no easily
discernible loss of detail, unless the detail was present at the
extremes of dynamic range which could be brought into visibility by
selective procesing of the RAW file. So I use RAW in cases of
unusually high dynamic range or awkward colour balancing, or if I know
I'm going to spend a long trying to squeeze the ultimate quality from
the image in editors, but otherwise not.

--
Chris Malcolm cam DeleteThis @infirmatics.ed.ac.uk DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]
 >> Stay informed about: Always shoot at highest megapixel setting?