Raw vs jpeg
Lyndon
be a modified camera that I intend to primarily only shoot
in monochrome mode and see little reason to ever shoot in color.
However, the camera I most wanted does not do the RAW
format. So, with intent to only shoot in monochrome, is there
any disadvantage in not having the raw format available?
Lyndon
--
********************
lyn...@Some-Isp.com
Scott W
> I am interested in buying a new dig camera. It will
> be a modified camera that I intend to primarily only shoot
> in monochrome mode and see little reason to ever shoot in color.
> However, the camera I most wanted does not do the RAW
> format. So, with intent to only shoot in monochrome, is there
> any disadvantage in not having the raw format available?
>
What is the modification you are going to me to the camera?
Charles
"Lyndon" <lyn...@Some-Isp.com> wrote in message
news:lyndonr-202B92...@news.newsguy.com...
>I am interested in buying a new dig camera. It will
> be a modified camera that I intend to primarily only shoot
> in monochrome mode and see little reason to ever shoot in color.
> However, the camera I most wanted does not do the RAW
> format. So, with intent to only shoot in monochrome, is there
> any disadvantage in not having the raw format available?
Yes and no. It all depends on your goals and your final output ...
electronic, prints ...
More info. about this "modified camera" would be helpful.
I shoot RAW about 20% of the time. Just me.
tony cooper
wrote:
>I am interested in buying a new dig camera. It will
>be a modified camera that I intend to primarily only shoot
>in monochrome mode and see little reason to ever shoot in color.
>However, the camera I most wanted does not do the RAW
>format. So, with intent to only shoot in monochrome, is there
>any disadvantage in not having the raw format available?
>
You really need to supply more information to get a good answer. Like
"What will the photographs be used for?" and "What software do you
have?".
If you are taking photographs for illustrations in a manual or
documentation of something, RAW may not be useful. If you are taking
photographs as "art", then RAW offers a sophisticated level of
post-processing.
If you have a fairly sophisticated editing program, you may be able to
do sufficient post-processing to a .jpg.
If you expect to be able to take the images from the camera and use
them without any significant adjustments other than size, then you
don't really need RAW.
--
Tony Cooper - Orlando, Florida
al-Farrob
Well, if you intend to use monochrome, one of the reasons to use raw
is out of the way, the white balance.
Though there is more reasons why you'd want to use raw, maybe the most
important is to take advantage of a wider dynamic range.
José Varela
www.al-farrob.com
nospam
<bab1836d-e39e-42f6...@t10g2000vbg.googlegroups.com>,
al-Farrob <alFa...@gmail.com> wrote:
> Well, if you intend to use monochrome, one of the reasons to use raw
> is out of the way, the white balance.
actually no it doesn't, since the sensor shoots in colour and
converting it to monochrome can be done in a variety of ways. if you
set the camera to jpeg, you get only one monochrome conversion.
John Navas
>I am interested in buying a new dig camera. It will
>be a modified camera that I intend to primarily only shoot
>in monochrome mode and see little reason to ever shoot in color.
>However, the camera I most wanted does not do the RAW
>format. So, with intent to only shoot in monochrome, is there
>any disadvantage in not having the raw format available?
You'll have more flexibility and can get better results if you shoot in
color and convert to B&W in post-processing.
--
Best regards,
John (Panasonic DMC-FZ28, and several others)
Martin Brown
> I am interested in buying a new dig camera. It will
> be a modified camera that I intend to primarily only shoot
> in monochrome mode and see little reason to ever shoot in color.
If it is a specialist scientific camera with no Bayer filter mask then
shooting in colour requires 3 exposures with different filters.
Dedicated astronomical cameras tend to be of this designs (although a
few do single shot colour they are not as good for photometry).
> However, the camera I most wanted does not do the RAW
> format. So, with intent to only shoot in monochrome, is there
> any disadvantage in not having the raw format available?
Possibly. You should at least in theory have a few more bits of dynamic
range in the raw data whereas after in camera JPEG conversion you are
stuck with just 8 bits. If photometry is your aim then RAW is best.
If you described your application and the camera(s) you were considering
it would be a lot easier to give sensible advice.
Regards,
Martin Brown
Robert Coe
: I am interested in buying a new dig camera. It will
: be a modified camera that I intend to primarily only shoot
: in monochrome mode and see little reason to ever shoot in color.
: However, the camera I most wanted does not do the RAW
: format. So, with intent to only shoot in monochrome, is there
: any disadvantage in not having the raw format available?
No.
al-Farrob
> In article
> <bab1836d-e39e-42f6-8c6a-ed544e563...@t10g2000vbg.googlegroups.com>,
>
> al-Farrob <alFar...@gmail.com> wrote:
> > Well, if you intend to use monochrome, one of the reasons to use raw
> > is out of the way, the white balance.
>
> actually no it doesn't, since the sensor shoots in colour and
> converting it to monochrome can be done in a variety of ways. if you
> set the camera to jpeg, you get only one monochrome conversion.
To be exact, yes, you are right.
José Varela
www.al-farrob.com
RAW Can Be Nonsense
>In article <lyndonr-202B92...@news.newsguy.com>,
> Lyndon <lyn...@Some-Isp.com> wrote:
>
>> I am interested in buying a new dig camera. It will
>> be a modified camera that I intend to primarily only shoot
>> in monochrome mode and see little reason to ever shoot in color.
>> However, the camera I most wanted does not do the RAW
>> format. So, with intent to only shoot in monochrome, is there
>> any disadvantage in not having the raw format available?
>>
>> Lyndon
>
>This is very simple. If you intend to do a lot of photo manipulation in
>Photoshop or some other photo editor, shooting in raw is a good idea. If
>you do not intend to do much manipulation in a photo editor, raw is
>overkill.
Let us be reminded too where the popularity of RAW formats originated. In
the earlier years of DSLRs their in-camera RAW to JPG conversion process
was abominable. People spending that much on cameras couldn't tolerate the
color-balance, compression artifacts, and sharpening decisions that were
being made for them by the pencil-pushing tech-head firmware programmers,
those who knew absolutely nothing about real photography. Purchasers of
these cameras demanded access to the original sensor data so they could do
the conversion themselves, painstakingly trying to repair their camera's
glaring firmware programming problems and other imaging defects.
Many DSLRs still suffer from this poor conversion problem because owners of
them are more than happy to be jumping through needless editing hoops
trying to repair what their camera fails to provide in the first place. As
long as they are willing to do all this then the camera companies see no
need to make the conversion process better in the camera itself. Instead
they are quite content to also sell them overpriced and required RAW
editing software so the DSLR owner can repair all these defects on their
own. Just another way to make more money from fools. They primarily succeed
on the principle of, "There's one born every minute."
Today's cameras designed to sell based on their JPG output alone do a
remarkable job of the in-camera conversion where little to no extra editing
is ever needed. You would be hard-pressed to champion what these cameras
provide in one step by processing the RAW data externally on your own. If
you buy a good camera that handles this conversion process correctly in the
first place then access to a RAW file format is far far less important then
the lens affixed to that camera and other capabilities of that camera.
Robert Coe
That is bullshit, as anyone who has used RAW mode on a good camera well knows.
Bob
Lyndon
tony cooper <tony_co...@earthlink.net> wrote:
> You really need to supply more information to get a good answer. Like
> "What will the photographs be used for?" and "What software do you
> have?".
I was looking to buy a Fuji camera (possibly the s-10000) from
LDP (maxmax.com). After being a long time film infrared shooter
and the diminishing availability of infrared film, I would like to
convert over to digital (plus the extended digital range of infrared
sensitivity). But the s-1000 does not have the raw format. Since the
infrared blockers are removed from the camera, it would be possible
but busy work to be attaching the proper filters to be able to take
"normal" pictures with the camera. I would most likely get the camera
in the UV-IR capable mode as I already own the filters to block IR,pass
UV(very dark violet filter), and pass film level IR (very dark red).
Software I would use could be anything from GraphicConverter and
PhotoshopCS1.
Thanks for all the previous answers
Lyndon
--
********************
lyn...@Some-Isp.com
RAW Can Be Nonsense
Yes, they would know. That's how I know that the above is perfectly true.
This is also why you just outted yourself as never having used any of these
cameras that force the user to have to depend on the use of RAW to get
worthwhile images from them. This is precisely why they shout the RAW
mantra incessantly--they can't get decent JPG output from their crippled
and poorly programmed cameras. Then again, most of them are happier
fiddling at their desk on their computer than actually taking pictures that
are ready to print. Always hoping that if they fiddle with the just the
right filter, just the right auto-fix tool, that they too one day will
finally have worthwhile images. Get just the right RAW editor and they'll
have instant talent. Buy the most expensive camera and they too will look
like and be a "pro" photographer. In their delusions only.
Some would rather be taking exceptional photos, rather than pushing buttons
and selecting options from menus in the hopes their photos will look
passable one day. You don't need more than 8-bit color-depth for any print
job if the resulting image looks good to begin with. No printer can produce
that many shades and hues. Didn't you know this? I guess not or you would
have known better than to post your ignorant reply. Maybe you should go buy
any of the decent cameras with very high quality JPG output so you can
educate yourself a bit. Get a little real-world experience under your belt.
John Navas
<270q35daaumavr5rh...@4ax.com>:
I've long used both RAW and out of camera JPEG, and I personally think
(when stripped of pejoratives) that it's pretty much on the money.
Martin Brown
> On Sat, 20 Jun 2009 08:50:21 -0400, Shawn Hirn <sr...@comcast.net> wrote:
>
>> In article <lyndonr-202B92...@news.newsguy.com>,
>> Lyndon <lyn...@Some-Isp.com> wrote:
>>
>>> I am interested in buying a new dig camera. It will
>>> be a modified camera that I intend to primarily only shoot
>>> in monochrome mode and see little reason to ever shoot in color.
>>> However, the camera I most wanted does not do the RAW
>>> format. So, with intent to only shoot in monochrome, is there
>>> any disadvantage in not having the raw format available?
>>>
>>> Lyndon
>> This is very simple. If you intend to do a lot of photo manipulation in
>> Photoshop or some other photo editor, shooting in raw is a good idea. If
>> you do not intend to do much manipulation in a photo editor, raw is
>> overkill.
>
> Let us be reminded too where the popularity of RAW formats originated. In
> the earlier years of DSLRs their in-camera RAW to JPG conversion process
> was abominable. People spending that much on cameras couldn't tolerate the
> color-balance, compression artifacts, and sharpening decisions that were
> being made for them by the pencil-pushing tech-head firmware programmers,
> those who knew absolutely nothing about real photography. Purchasers of
Your insults here should be hurled at the marketing men who insisted
(correctly as it turns out) that sharpness sells to end users. Most
cameras are by default slightly over sharpened in the firmware.
> these cameras demanded access to the original sensor data so they could do
> the conversion themselves, painstakingly trying to repair their camera's
> glaring firmware programming problems and other imaging defects.
The maximum quality JPEG even out of early cameras like the Kodak DC-120
wasn't that far off the mark from a standards point of view. It did have
a RAW mode though so that you could do a more complex demosaic and white
balance afterwards. Despite only being a 1Mpixel camera it found a place
for a while in scientific photography because of this feature.
There are still faults in the JPEG *decoders* that do lead to problems
when shooting images that have fine black detail against saturated red
or blues in them. Flowers and jazz or rock concerts for instance. In
these circumstances RAW wins hands down because you do not get the JPEG
subsampled chroma problems damaging the image resolution.
>
> Many DSLRs still suffer from this poor conversion problem because owners of
> them are more than happy to be jumping through needless editing hoops
> trying to repair what their camera fails to provide in the first place. As
> long as they are willing to do all this then the camera companies see no
> need to make the conversion process better in the camera itself. Instead
> they are quite content to also sell them overpriced and required RAW
> editing software so the DSLR owner can repair all these defects on their
> own. Just another way to make more money from fools. They primarily succeed
> on the principle of, "There's one born every minute."
The current crop of JPEG converters in modern cameras have no difficulty
in making near perfect high quality JPEGs that are the optimum encoding
of their input data. There are no compromises on accuracy for speed now.
RAW gives you the option to sort out white balance and retain both deep
shadow detail and highlights afterwards whereas the decision of the JPEG
converter is pretty much final. The initial 12bit image has the dynamic
range to support detail in both the brides dress and the grooms black
velvet suit but not in a JPEG conversion unless you are very lucky.
>
> Today's cameras designed to sell based on their JPG output alone do a
> remarkable job of the in-camera conversion where little to no extra editing
> is ever needed. You would be hard-pressed to champion what these cameras
> provide in one step by processing the RAW data externally on your own. If
Pretty much agree unless you are shooting a scene with high dynamic
range and details in the extremes that really matter to the content. Or
in a challenging light environment that is outside what the autowhite
balance can handle reliably - sports under HPS lights for instance.
> you buy a good camera that handles this conversion process correctly in the
> first place then access to a RAW file format is far far less important then
> the lens affixed to that camera and other capabilities of that camera.
The continual increase in the number of pixels in the small cameras is
getting pretty silly these days. Again marketting knows that a simple
number sells on a mine is bigger than yours basis. But we have reached
the point where more pixels doesn't mean more useful data.
Regards,
Martin Brown
John Navas
<|||newspam|||@nezumi.demon.co.uk> wrote in
<SGH%l.9757$cz1....@newsfe17.iad>:
>RAW Can Be Nonsense wrote:
>> Let us be reminded too where the popularity of RAW formats originated. In
>> the earlier years of DSLRs their in-camera RAW to JPG conversion process
>> was abominable. People spending that much on cameras couldn't tolerate the
>> color-balance, compression artifacts, and sharpening decisions that were
>> being made for them by the pencil-pushing tech-head firmware programmers,
>> those who knew absolutely nothing about real photography. Purchasers of
>
>Your insults here should be hurled at the marketing men who insisted
>(correctly as it turns out) that sharpness sells to end users. Most
>cameras are by default slightly over sharpened in the firmware.
True, but to be fair, that's largely because earlier photo printers gave
their best results with slightly over-sharpened images.
>> these cameras demanded access to the original sensor data so they could do
>> the conversion themselves, painstakingly trying to repair their camera's
>> glaring firmware programming problems and other imaging defects.
>
>There are still faults in the JPEG *decoders* that do lead to problems
>when shooting images that have fine black detail against saturated red
>or blues in them. Flowers and jazz or rock concerts for instance. In
>these circumstances RAW wins hands down because you do not get the JPEG
>subsampled chroma problems damaging the image resolution.
With all due respect, that's really a fault of the encoding.
>> Many DSLRs still suffer from this poor conversion problem because owners of
>> them are more than happy to be jumping through needless editing hoops
>> trying to repair what their camera fails to provide in the first place. As
>> long as they are willing to do all this then the camera companies see no
>> need to make the conversion process better in the camera itself. Instead
>> they are quite content to also sell them overpriced and required RAW
>> editing software so the DSLR owner can repair all these defects on their
>> own. Just another way to make more money from fools. They primarily succeed
>> on the principle of, "There's one born every minute."
>
>The current crop of JPEG converters in modern cameras have no difficulty
>in making near perfect high quality JPEGs that are the optimum encoding
>of their input data. There are no compromises on accuracy for speed now.
That's not what I see when comparing the JPEG encoding of good computer
software with camera JPEG output, and not terribly surprising given the
short amount of time and limited processing power available to the
camera.
>RAW gives you the option to sort out white balance and retain both deep
>shadow detail and highlights afterwards whereas the decision of the JPEG
>converter is pretty much final. The initial 12bit image has the dynamic
>range to support detail in both the brides dress and the grooms black
>velvet suit but not in a JPEG conversion unless you are very lucky.
Dynamic range management is actually one area in which current camera
JPEG conversion performs very well.
>> you buy a good camera that handles this conversion process correctly in the
>> first place then access to a RAW file format is far far less important then
>> the lens affixed to that camera and other capabilities of that camera.
>
>The continual increase in the number of pixels in the small cameras is
>getting pretty silly these days. Again marketting knows that a simple
>number sells on a mine is bigger than yours basis. But we have reached
>the point where more pixels doesn't mean more useful data.
Tests actually show these cameras to produce significantly better
results at greater resolution than earlier cameras with lower
resolution. The reasons are that sensors are better and that image
quality has come to be dominated by in camera processing. For that
matter, you could make the same complaint about dSLR cameras, and it's a
bit hypocritical not to do so.
Martin Brown
No it isn't. Information is correctly stored in the chroma subsampled
JPEG encoded coefficients. But there is a significant and *unnecessary*
loss of information during standard JPEG decoding. The resulting image
is inconsistent with its original JPEG encoding and accelerates
generational losses.
It happens whenever a strong saturated red or blue is next to black or a
strong saturated yellow or cyan next to white. These produce the worst
case loss of fidelity. This is *not* a fault in the encoder. I can post
a sample line art demonstration if you wish.
It is an error or omission in the original JPEG specification for chroma
upsampling during the reconstruction of an image. A few of the newest
MP4 TV codecs are now doing what they call 4:4:4 chroma. I suspect they
are at last doing it right.
>
>>> Many DSLRs still suffer from this poor conversion problem because owners of
>>> them are more than happy to be jumping through needless editing hoops
>>> trying to repair what their camera fails to provide in the first place. As
>>> long as they are willing to do all this then the camera companies see no
>>> need to make the conversion process better in the camera itself. Instead
>>> they are quite content to also sell them overpriced and required RAW
>>> editing software so the DSLR owner can repair all these defects on their
>>> own. Just another way to make more money from fools. They primarily succeed
>>> on the principle of, "There's one born every minute."
>> The current crop of JPEG converters in modern cameras have no difficulty
>> in making near perfect high quality JPEGs that are the optimum encoding
>> of their input data. There are no compromises on accuracy for speed now.
>
> That's not what I see when comparing the JPEG encoding of good computer
> software with camera JPEG output, and not terribly surprising given the
> short amount of time and limited processing power available to the
> camera.
I would like to see and example where the cameras encoding is inferior
to what a PC manages. At least one common PC imaging application has a
broken encoder but few people have noticed. Again it is a fault in the
chroma sampling (but this time serious irreversible ones during encode).
>> RAW gives you the option to sort out white balance and retain both deep
>> shadow detail and highlights afterwards whereas the decision of the JPEG
>> converter is pretty much final. The initial 12bit image has the dynamic
>> range to support detail in both the brides dress and the grooms black
>> velvet suit but not in a JPEG conversion unless you are very lucky.
>
> Dynamic range management is actually one area in which current camera
> JPEG conversion performs very well.
You still only get one chance with a JPEG whereas from the RAW you keep
your options open.
>
>>> you buy a good camera that handles this conversion process correctly in the
>>> first place then access to a RAW file format is far far less important then
>>> the lens affixed to that camera and other capabilities of that camera.
>> The continual increase in the number of pixels in the small cameras is
>> getting pretty silly these days. Again marketting knows that a simple
>> number sells on a mine is bigger than yours basis. But we have reached
>> the point where more pixels doesn't mean more useful data.
>
> Tests actually show these cameras to produce significantly better
> results at greater resolution than earlier cameras with lower
> resolution. The reasons are that sensors are better and that image
> quality has come to be dominated by in camera processing. For that
> matter, you could make the same complaint about dSLR cameras, and it's a
> bit hypocritical not to do so.
dSLR camera sensors are not yet at the theoretical limit to be a
resolution match for the very best lenses at their optimum aperture.
Regards,
Martin Brown
John Navas
<|||newspam|||@nezumi.demon.co.uk> wrote in
<2X%%l.7341$dg5...@newsfe25.iad>:
>John Navas wrote:
>> On Mon, 22 Jun 2009 09:56:16 +0100, Martin Brown
>> <|||newspam|||@nezumi.demon.co.uk> wrote in
>> <SGH%l.9757$cz1....@newsfe17.iad>:
>>> There are still faults in the JPEG *decoders* that do lead to problems
>>> when shooting images that have fine black detail against saturated red
>>> or blues in them. Flowers and jazz or rock concerts for instance. In
>>> these circumstances RAW wins hands down because you do not get the JPEG
>>> subsampled chroma problems damaging the image resolution.
>>
>> With all due respect, that's really a fault of the encoding.
>
>No it isn't. Information is correctly stored in the chroma subsampled
>JPEG encoded coefficients. But there is a significant and *unnecessary*
>loss of information during standard JPEG decoding. The resulting image
>is inconsistent with its original JPEG encoding and accelerates
>generational losses.
Proof?
>It happens whenever a strong saturated red or blue is next to black or a
>strong saturated yellow or cyan next to white. These produce the worst
>case loss of fidelity. This is *not* a fault in the encoder. I can post
>a sample line art demonstration if you wish.
How about even one authoritative supporting citation?
>> That's not what I see when comparing the JPEG encoding of good computer
>> software with camera JPEG output, and not terribly surprising given the
>> short amount of time and limited processing power available to the
>> camera.
>
>I would like to see and example where the cameras encoding is inferior
>to what a PC manages. ...
Very easy to do with a camera that produces both RAW and JPEG output.
Open the RAW in Photoshop and save as JPEG. Depending on output
settings, you can get a comparable quality JPEG in a much smaller file,
or better quality JPEG in the same size file.
>> Dynamic range management is actually one area in which current camera
>> JPEG conversion performs very well.
>
>You still only get one chance with a JPEG whereas from the RAW you keep
>your options open.
If that one chance works 99% of the time, saves a great deal of
unnecessary post-processing, and allows transfer or printing wherever
I happen to be, that's often a worthwhile tradeoff to me. But as
always, YMMV.
>> Tests actually show these cameras to produce significantly better
>> results at greater resolution than earlier cameras with lower
>> resolution. The reasons are that sensors are better and that image
>> quality has come to be dominated by in camera processing. For that
>> matter, you could make the same complaint about dSLR cameras, and it's a
>> bit hypocritical not to do so.
>
>dSLR camera sensors are not yet at the theoretical limit to be a
>resolution match for the very best lenses at their optimum aperture.
Irrelevant to the issue at hand.
Martin Brown
> On Tue, 23 Jun 2009 08:59:07 +0100, Martin Brown
> <|||newspam|||@nezumi.demon.co.uk> wrote in
> <2X%%l.7341$dg5...@newsfe25.iad>:
>
>> John Navas wrote:
>>> On Mon, 22 Jun 2009 09:56:16 +0100, Martin Brown
>>> <|||newspam|||@nezumi.demon.co.uk> wrote in
>>> <SGH%l.9757$cz1....@newsfe17.iad>:
>
>>>> There are still faults in the JPEG *decoders* that do lead to problems
>>>> when shooting images that have fine black detail against saturated red
>>>> or blues in them. Flowers and jazz or rock concerts for instance. In
>>>> these circumstances RAW wins hands down because you do not get the JPEG
>>>> subsampled chroma problems damaging the image resolution.
>>> With all due respect, that's really a fault of the encoding.
>> No it isn't. Information is correctly stored in the chroma subsampled
>> JPEG encoded coefficients. But there is a significant and *unnecessary*
>> loss of information during standard JPEG decoding. The resulting image
>> is inconsistent with its original JPEG encoding and accelerates
>> generational losses.
>
> Proof?
In general for images containing areas of white or black with sharp
edges there are values in the reconstruction after inverting the
quantised coefficients through the DCT that are outside the allowed
range of 0..255. The default method is to top and tail the decoded image
which introduces a marked systematic error in the DC component as well
as any other components that contribute to the adjusted image point(s).
Xerox have patents on this method of reducing the ringing and smoke in
blocks that contain sharp lines on a plain ground by using an iterative
method to find an image that when JPEG encoded exactly matches the
supplied JPEG coefficients subject to some regularisation rule (eg.
smoothness or entropy).
See for example US patents 5,521,718 and 5,757,975 in the latter case
they have effectively been granted a patent on the mathematical identity
X + (-X) = 0. The prior art for these techniques existed in radio
astronomy long before JPEG was even imagined.
>
>> It happens whenever a strong saturated red or blue is next to black or a
>> strong saturated yellow or cyan next to white. These produce the worst
>> case loss of fidelity. This is *not* a fault in the encoder. I can post
>> a sample line art demonstration if you wish.
>
> How about even one authoritative supporting citation?
It is work in progress. Easy enough to demonstrate the problem though.
Flood a toy image with pure red and draw a few fine black lines on it
(or vice versa). Lines need to be single pixel and sharp to maximise the
visible effect. Then encode it as chroma subsampled JPEG and decode
(Photoshop Levels 0..5).
If you decode again to a colour image you will see the luminance error
smearing from the chroma and convert that to monochrome and you do not
get the right answer. By comparison if you decode the luminance channel
of the JPEG coefficients directly to a greyscale image the result is
more or less correct (subject to the other effect mentioned above).
Apples recent patent 7,483,037 describes some of the issues and one way
of dealing with it. I cannot say more in public.
>
>>> That's not what I see when comparing the JPEG encoding of good computer
>>> software with camera JPEG output, and not terribly surprising given the
>>> short amount of time and limited processing power available to the
>>> camera.
>> I would like to see and example where the cameras encoding is inferior
>> to what a PC manages. ...
>
> Very easy to do with a camera that produces both RAW and JPEG output.
> Open the RAW in Photoshop and save as JPEG. Depending on output
> settings, you can get a comparable quality JPEG in a much smaller file,
> or better quality JPEG in the same size file.
But that is due to a time consuming size optimisation in the lossless
phase of the image compression. Photoshop uses optimised custom Huffman
tables matched to the image being processed. Most digital cameras that I
have seen use the default Huffman tables from the JPEG standard App K.3.
There is a modest hit in the size of the image but it is much faster and
it doesn't affect the quantisation artefacts at constant quality. One
other advantage of using the default Huffman tables in a digicam is that
if something goes wrong on the media you still know how to decode the
JPEG stream without having to resort to forensic techniques.
It is also a lot faster to save with a predefined Huffman table since
optimising it for frequency requires two full passes across all the
image coefficient data. Higher frame rate is probably more important to
most people. Digital media is very cheap!
PS Sad to note the passing of Kodachrome film yesterday. The 25 ASA
version was one of my favourites for high colour resolution work - even
if some bright reds did seem to stand out of the image plane.
>
>>> Dynamic range management is actually one area in which current camera
>>> JPEG conversion performs very well.
>> You still only get one chance with a JPEG whereas from the RAW you keep
>> your options open.
>
> If that one chance works 99% of the time, saves a great deal of
> unnecessary post-processing, and allows transfer or printing wherever
> I happen to be, that's often a worthwhile tradeoff to me. But as
> always, YMMV.
Fair enough. I mostly use JPEG unless I think it will cause problems.
>
>>> Tests actually show these cameras to produce significantly better
>>> results at greater resolution than earlier cameras with lower
>>> resolution. The reasons are that sensors are better and that image
>>> quality has come to be dominated by in camera processing. For that
>>> matter, you could make the same complaint about dSLR cameras, and it's a
>>> bit hypocritical not to do so.
>> dSLR camera sensors are not yet at the theoretical limit to be a
>> resolution match for the very best lenses at their optimum aperture.
>
> Irrelevant to the issue at hand.
Not at all. P&S sensors have crossed the boundary where the diffraction
limited lens is too small to make full use of the sensor array. And
there is no sign that the race to have ever more pixels is abating.
Regards,
Martin Brown
And the idiots go on ...
<|||newspam|||@nezumi.demon.co.uk> wrote:
>
>Not at all. P&S sensors have crossed the boundary where the diffraction
>limited lens is too small to make full use of the sensor array. And
>there is no sign that the race to have ever more pixels is abating.
>
>Regards,
>Martin Brown
Go educate yourself on what "diffraction limited" truly means. This is not
a defect, this describes the best optics that are possible by the physics
of the universe itself. But since all ignorantly biased fools like to
continually perpetuate the egregious misuse of this term, why should you be
any different. I've yet to see even one photo from larger lenses that is
pixel-detail sharp, because those larger lenses are not figured to
diffraction-limited precision. It is cost prohibitive for them to attain
diffraction-limited precision on larger lenses. Those
non-diffraction-limited larger lenses can't focus details down to just one
sensor-site on those larger sensors. A much larger margin of error and they
still can't hit it.
Even if those larger lenses were diffraction-limited they'd still provide
more diffraction defects in the image because they have to use longer
focal-lengths for the same FOVs. Diffraction spreads out as a function of
distance, a physical constant. With smaller diffraction-limited precision
lenses of shorter focal-lengths on smaller sensors, pixel-detail sharp
imaging is commonplace. On average there is 1/6th the amount of diffraction
in the image from small sensor cameras than when using all 35mm lenses
(even if they could create those larger lenses with diffraction-limited
precision). Because the smaller sensor cameras use lenses that are 1/6th
the focal-length for the same FOVs. This means there is 1/6th the distance
from lens to imaging-plane for the diffraction to spread and be recorded on
the sensor. I'd never buy a camera that couldn't provide their
pixel-detail-sharp quality routinely.
Go peddle your blatantly biased DSLR ignorance elsewhere. Those who know
more than you ever will aren't buying it. And oh-so-tired of attempting to
educate you mindless DSLR idiots as often as is needed.
btw1: I just did your jpg-defects tests. Apparently you like typing things
that sound intelligent but when put to real-world tests your
"intellectualized fiction" is just that. Either that or Photoline, one of
the best editors ever created, has already solved the problem of which you
attempt to speak.
btw2: One of the more fun and at times very useful aspects of Photoline--it
has a "while editing" JPG compression setting for each individual layer if
needed. Its function was designed primarily for when creating PDF documents
or large complex collages with many images. Allowing you to choose whatever
compression level you want for any individual image in the document
independently of all others, during the creation of the document. This
makes it easy to preview all JPG compression artifacts on-the-fly on any
layer/image before even saving. Found under Layout > Image > JPG
Compression. (Turn off the handy antialiasing preview option too so you'll
see the true compression effects. The antialiasing preview is primarily
used to see how your printer is going to degrade the image before sending
it through the printer's own DPI interpolation. Most apparent and useful
when printing vector-graphics (fonts, line-art, etc.).) Some images
withstand large compression settings (content dependent), others don't.
You're not forced into using a one-size-fits-all compression setting on the
whole document with Photoline, as all other less-intelligently designed
programs force you to do.
John Navas
<|||newspam|||@nezumi.demon.co.uk> wrote in
<tsm0m.454$Ai6...@newsfe09.iad>:
>John Navas wrote:
>> How about even one authoritative supporting citation?
>
>It is work in progress. ...
I'm seriously interested, so let me know when you have something (by
email if you like).
>> Very easy to do with a camera that produces both RAW and JPEG output.
>> Open the RAW in Photoshop and save as JPEG. Depending on output
>> settings, you can get a comparable quality JPEG in a much smaller file,
>> or better quality JPEG in the same size file.
>
>But that is due to a time consuming size optimisation in the lossless
>phase of the image compression. Photoshop uses optimised custom Huffman
>tables matched to the image being processed. Most digital cameras that I
>have seen use the default Huffman tables from the JPEG standard App K.3.
>
>There is a modest hit in the size of the image but it is much faster and
>it doesn't affect the quantisation artefacts at constant quality.
Actually a pretty big size hit in all the cases I've looked at, and
there is a hit on quality as well because camera manufacturers are
keeping the file size down to conserve memory card space. TANSTAAFL.
>PS Sad to note the passing of Kodachrome film yesterday. The 25 ASA
>version was one of my favourites for high colour resolution work - even
>if some bright reds did seem to stand out of the image plane.
[SOB!] Loved both 25 and 64, the latter being my primary film for
sports photography, producing awesome images, including a couple of
national magazine covers.
>>> dSLR camera sensors are not yet at the theoretical limit to be a
>>> resolution match for the very best lenses at their optimum aperture.
>>
>> Irrelevant to the issue at hand.
>
>Not at all. P&S sensors have crossed the boundary where the diffraction
>limited lens is too small to make full use of the sensor array.
I respectfully disagree; e.g., my Panasonic DMC-FZ28 is objectively
better than my DMC-FZ8.
>And
>there is no sign that the race to have ever more pixels is abating.
I personally think there has been a slowing at the 10 MP level.
All this said, I personally see no reason not to increase the sensor
resolution as long as in camera processing keeps pace -- I've seen some
interesting papers on how smaller photosites can be aggregated to
produce better results than larger photosites, in part because of Bayer
issues. I think it's quite possible we might see (say) 16 MP sensors
aggregated down to (say) 8 MP output that produce better results than 8
MP sensors, especially for everyday shooting.
Wolfgang Weisselberg
> al-Farrob <alFa...@gmail.com> wrote:
>> Well, if you intend to use monochrome, one of the reasons to use raw
>> is out of the way, the white balance.
> actually no it doesn't, since the sensor shoots in colour and
> converting it to monochrome can be done in a variety of ways. if you
> set the camera to jpeg, you get only one monochrome conversion.
If you set the camera to JPEG you still get a colour JPEG.
You'd need to set the camera deliberately to monochrome and JPEG
for your claims to be true, for from a colour JPEG you will still
have a variety of ways to convert to monochrome.
-Wolfgang
Wolfgang Weisselberg
Parents can be cruel. Have you sued yours yet or are you
going postal soon and will run amok?
BTW, has RegistrarAds, Inc. given you leave to use their
domain for your email? Surely they have not, as the domain
doesn't accept emails (no MX server) ... so you are nothing
but a fake email addres troll, too fake to even stand with your real name for
what you say, shivering that someone might want to send *you*
mail, too cheap to buy your own domain.
Maybe we should mail <ab...@cpinternet.com>. Not only can they
decode their X-Trace stamp, they also know which customer used
their dynamic IP '169.220.61.64.modem.dynamic.cptelecom.net'
at Sat, 20 Jun 2009 08:28:58 -0500 ...
> Let us be reminded too where the popularity of RAW formats originated.
Yes, lets. I think it was with William Henry Fox Talbot
(1800-1877) and his invention of the negative-positive process
of photography. Before that you only had things like the
daguerrotype where every single image was a unique positive.
Now, with the calotype (1840) process, duplication, even mass
duplication was possible for the first time. The first negative,
not yet using the improved calotype process, was made in 1835.
Direct-positive images have become very rare, with Polaroids
going out of business, even though they have very interesting
and important possibilities. (Note that Polaroid offered some
films that not only had an immediately developed positive image,
but also a negative, all in one.)
Out of the negative process evolved all the magnification and
lab work, enhancing final positives by dodging and burning,
using unsharp masks to enhance acutancy --- the very same process
actually also happens to the negative during development with
more modern films and developers --- and so on and so on.
That's where and when the popularity of RAW formats originated.
> Today's cameras designed to sell based on their JPG output alone do a
> remarkable job of the in-camera conversion where little to no extra editing
> is ever needed.
For a troll you see quite clearly that there is a difference
between "needed" (photography isn't needed, for example) and
"wanted". You just close your eyes ...
-Wolfgang
RAW Can Be Nonsense
<ozcv...@sneakemail.com> wrote:
>
>Maybe we should mail <ab...@cpinternet.com>. Not only can they
>decode their X-Trace stamp, they also know which customer used
>their dynamic IP '169.220.61.64.modem.dynamic.cptelecom.net'
>at Sat, 20 Jun 2009 08:28:58 -0500 ...
You do that. See how far it gets you.
LOL!!!!!!!!!!