Sign in with
Sign up | Sign in
Your question

A concept that I had a few years ago (virtual resolution e..

Last response: in Digital Camera
Share
Anonymous
December 28, 2004 11:11:31 AM

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

I have noticed that there is an interest in high resolution digital
images and thought that I would share my concept that I developed back
in the early 90s.

I was never able to get anything going with this project but I am
curious as to whether anyone came up with something like this.

The method that I proposed was to expose the CCD to the image multiple
times but with a slightly different location. This (I theorized) would
make it possible to enhance the resolution of the CCD by 4, 9, or even
16 times.

This could provide you with 160 Mega Pixels with a conventional CCD.
If interested, I could post a Word document that illustrates the
concept very clearly. I didn't want to post it here as this is not a
binary group.

Please remove the extra characters in my email address if you want to
use email.
Anonymous
December 28, 2004 11:11:32 AM

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

Lorne Tontegode <lorne___tontegode@@@hotmail...com> writes:
> I was never able to get anything going with this project but I am
> curious as to whether anyone came up with something like this.
>
> The method that I proposed was to expose the CCD to the image multiple
> times but with a slightly different location.

Yes, this is a well known technique.
Anonymous
December 28, 2004 11:11:32 AM

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

There were a few pro cameras that used this technique, but it was very
expensive, and as far as I know it is no longer being used. The
mechanism that moves the image plane needs to move on a few microns-
this is extremely difficult and expensive.

It is not possible to move the whole camera in angle- the gimballing
system would be horrendously expensive and there are cheaper
alternatives.
Anonymous
December 28, 2004 11:47:49 AM

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

You probably don't mean this, but the same basic technique can be done by
moving the whole camera (including the ccd) between exposures and stitching
the images together with software. There is even a market for the tripod
accessories that move the camera in measured steps. Search for the following
terms in Google:

Kaidan Quickpan
QTVR
Photo Mosaic
Stitched Panorama
Panoramic tripod head
Pano Tools
PTGui
PT Assembler

Eric Miller

"Lorne Tontegode" <lorne___tontegode@@@hotmail...com> wrote in message
news:rvl2t0h5afg7r3itssa4a40ln3f3bfgv5b@4ax.com...
> I have noticed that there is an interest in high resolution digital
> images and thought that I would share my concept that I developed back
> in the early 90s.
>
> I was never able to get anything going with this project but I am
> curious as to whether anyone came up with something like this.
>
> The method that I proposed was to expose the CCD to the image multiple
> times but with a slightly different location. This (I theorized) would
> make it possible to enhance the resolution of the CCD by 4, 9, or even
> 16 times.
>
> This could provide you with 160 Mega Pixels with a conventional CCD.
> If interested, I could post a Word document that illustrates the
> concept very clearly. I didn't want to post it here as this is not a
> binary group.
>
> Please remove the extra characters in my email address if you want to
> use email.
Anonymous
December 28, 2004 1:41:33 PM

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

Actually, this is a different method. What you are referring to I
think is a mosaic. The method that I had proposed was to move the CCD
half a pixel over and then half a pixel down and then coninue.

There would have to be a mask placed over the CCD to limit exposure to
the scene so that only a quarter of the sample is taken. This would
result in a higher actual resolution and not require stitching it back
together.

There were a couple of other response to this and I am curious if
anyone has a link to their sites.


On Tue, 28 Dec 2004 08:47:49 -0600, "Eric Miller"
<ericmiller@cox-internet.com> wrote:

>You probably don't mean this, but the same basic technique can be done by
>moving the whole camera (including the ccd) between exposures and stitching
>the images together with software. There is even a market for the tripod
>accessories that move the camera in measured steps. Search for the following
>terms in Google:
>
>Kaidan Quickpan
>QTVR
>Photo Mosaic
>Stitched Panorama
>Panoramic tripod head
>Pano Tools
>PTGui
>PT Assembler
>
>Eric Miller
>
>"Lorne Tontegode" <lorne___tontegode@@@hotmail...com> wrote in message
>news:rvl2t0h5afg7r3itssa4a40ln3f3bfgv5b@4ax.com...
>> I have noticed that there is an interest in high resolution digital
>> images and thought that I would share my concept that I developed back
>> in the early 90s.
>>
>> I was never able to get anything going with this project but I am
>> curious as to whether anyone came up with something like this.
>>
>> The method that I proposed was to expose the CCD to the image multiple
>> times but with a slightly different location. This (I theorized) would
>> make it possible to enhance the resolution of the CCD by 4, 9, or even
>> 16 times.
>>
>> This could provide you with 160 Mega Pixels with a conventional CCD.
>> If interested, I could post a Word document that illustrates the
>> concept very clearly. I didn't want to post it here as this is not a
>> binary group.
>>
>> Please remove the extra characters in my email address if you want to
>> use email.
>
Anonymous
December 28, 2004 1:41:34 PM

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

OIC

Eric Miller

"Lorne Tontegode" <lorne___tontegode@@@hotmail...com> wrote in message
news:0ev2t0lnq0g781532fj9g1g8e7gfjbk3as@4ax.com...
> Actually, this is a different method. What you are referring to I
> think is a mosaic. The method that I had proposed was to move the CCD
> half a pixel over and then half a pixel down and then coninue.
>
> There would have to be a mask placed over the CCD to limit exposure to
> the scene so that only a quarter of the sample is taken. This would
> result in a higher actual resolution and not require stitching it back
> together.
>
> There were a couple of other response to this and I am curious if
> anyone has a link to their sites.
>
>
> On Tue, 28 Dec 2004 08:47:49 -0600, "Eric Miller"
> <ericmiller@cox-internet.com> wrote:
>
> >You probably don't mean this, but the same basic technique can be done by
> >moving the whole camera (including the ccd) between exposures and
stitching
> >the images together with software. There is even a market for the tripod
> >accessories that move the camera in measured steps. Search for the
following
> >terms in Google:
> >
> >Kaidan Quickpan
> >QTVR
> >Photo Mosaic
> >Stitched Panorama
> >Panoramic tripod head
> >Pano Tools
> >PTGui
> >PT Assembler
> >
> >Eric Miller
> >
> >"Lorne Tontegode" <lorne___tontegode@@@hotmail...com> wrote in message
> >news:rvl2t0h5afg7r3itssa4a40ln3f3bfgv5b@4ax.com...
> >> I have noticed that there is an interest in high resolution digital
> >> images and thought that I would share my concept that I developed back
> >> in the early 90s.
> >>
> >> I was never able to get anything going with this project but I am
> >> curious as to whether anyone came up with something like this.
> >>
> >> The method that I proposed was to expose the CCD to the image multiple
> >> times but with a slightly different location. This (I theorized) would
> >> make it possible to enhance the resolution of the CCD by 4, 9, or even
> >> 16 times.
> >>
> >> This could provide you with 160 Mega Pixels with a conventional CCD.
> >> If interested, I could post a Word document that illustrates the
> >> concept very clearly. I didn't want to post it here as this is not a
> >> binary group.
> >>
> >> Please remove the extra characters in my email address if you want to
> >> use email.
> >
>
Anonymous
December 28, 2004 1:49:12 PM

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

Lorne Tontegode <lorne___tontegode@@@hotmail...com> writes:

> The method that I proposed was to expose the CCD to the image multiple
> times but with a slightly different location. This (I theorized) would
> make it possible to enhance the resolution of the CCD by 4, 9, or even
> 16 times.

Take a look at Keith's Image Stacker, which can do as you wish if one moves
the camera:
http://www.unm.edu/~keithw

and see the Sony Qualia 016 (US$4,000), which combines 4 consecutive frames
for each shot.
http://www.dpreview.com/news/0306/03061002sonyqualia016...
for brief overview
http://shopping.rednova.com/catalog/product_16825_Sony_...
to buy
--
Phil Stripling | email to the replyto address is presumed
The Civilized Explorer | spam and read later. email from this URL
http://www.cieux.com/ | http://www.civex.com/ is read daily.
Anonymous
December 28, 2004 3:36:27 PM

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

Lorne Tontegode <lorne___tontegode@@@hotmail...com> writes:

> I have noticed that there is an interest in high resolution digital
> images and thought that I would share my concept that I developed back
> in the early 90s.
>
> I was never able to get anything going with this project but I am
> curious as to whether anyone came up with something like this.
>
> The method that I proposed was to expose the CCD to the image multiple
> times but with a slightly different location. This (I theorized) would
> make it possible to enhance the resolution of the CCD by 4, 9, or even
> 16 times.
>
> This could provide you with 160 Mega Pixels with a conventional CCD.
> If interested, I could post a Word document that illustrates the
> concept very clearly. I didn't want to post it here as this is not a
> binary group.

There's an easier way --
<http://www.tawbaware.com/maxlyons/gigapixel.htm&gt;.
--
David Dyer-Bennet, <mailto:D d-b@dd-b.net>, <http://www.dd-b.net/dd-b/&gt;
RKBA: <http://noguns-nomoney.com/&gt; <http://www.dd-b.net/carry/&gt;
Pics: <http://dd-b.lighthunters.net/&gt; <http://www.dd-b.net/dd-b/SnapshotAlbum/&gt;
Dragaera/Steven Brust: <http://dragaera.info/&gt;
Anonymous
December 28, 2004 4:47:50 PM

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

Lorne Tontegode wrote:
> I have noticed that there is an interest in high resolution digital
> images and thought that I would share my concept that I developed back
> in the early 90s.
>
> I was never able to get anything going with this project but I am
> curious as to whether anyone came up with something like this.
>
> The method that I proposed was to expose the CCD to the image multiple
> times but with a slightly different location. This (I theorized) would
> make it possible to enhance the resolution of the CCD by 4, 9, or even
> 16 times.
>
> This could provide you with 160 Mega Pixels with a conventional CCD.
> If interested, I could post a Word document that illustrates the
> concept very clearly. I didn't want to post it here as this is not a
> binary group.
>
> Please remove the extra characters in my email address if you want to
> use email.

I believe the technique is well known but has few practical
applications.

--
Joseph Meehan

26 + 6 = 1 It's Irish Math
Anonymous
December 28, 2004 9:43:38 PM

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

"Lorne Tontegode" <lorne___tontegode@@@hotmail...com> wrote in message
news:rvl2t0h5afg7r3itssa4a40ln3f3bfgv5b@4ax.com...
SNIP

> The method that I proposed was to expose the CCD to the image
> multiple
> times but with a slightly different location.

As others have noted, this is a known technique. In Astronomical
applications it is referred to as "drizzling"
(http://astrosurf.com/buil/us/spe9/lrgb22.htm), and it is also used on
Hubble Space Telescope images.

The reason it works well in astronomy is because the images consist of
many point light sources that, when randomly multi-sampled (in
sometimes turbulent atmosphere), approximate a Gaussian type of point
spread function.

Bart
Anonymous
December 28, 2004 9:55:33 PM

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

lorne___tontegode@@@hotmail...com (remove redundant characters) writes:

>The method that I proposed was to expose the CCD to the image multiple
>times but with a slightly different location. This (I theorized) would
>make it possible to enhance the resolution of the CCD by 4, 9, or even
>16 times.

I have seen several cameras that used this technique at SIGGRAPH over
the years, but none recently. It made sense when the highest resolution
CCDs that you could get were ones intended for video cameras, for
example 720x576 pixel PAL video CCDs. These days, just using a
higher-resolution CCD solves the problem for most pictorial photography.

The one camera actually used some sort of mechanical micro-mover to
shift the CCD by a fraction of a pixel while the rest of the optics
remained stationary. The other camera had a thin glass plate between
the lens and CCD (e.g. a microscope slide cover glass) which was rotated
to a variety of angles to shift the image. The CCD itself was
stationary.

All of these cameras are suitable only for stationary subjects like a
still life or product photography. They take a long time to capture an
image, and a bunch of computer power to process it (since the sensing
areas overlap, high frequencies are attenuated). You'd also want to use
a B&W CCD with colour filter wheel for colour; applying the process with
a Bayer sensor would just give you a cluster of red samples, a cluster
of green ones, and a cluster of blue ones, rather than RGB samples
interleaved with each other.

Basically, it's a feasible way to get more resolution under some
conditions with some subjects when you just cannot get a larger CCD or
shoot a mosaic. But there's usually a better way available.

Dave
Anonymous
December 28, 2004 11:41:37 PM

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

Thanks to all that have responded. It was about 1992 that I thought
about this but wasn't able to manage anything more than a patent
search at the time. Too many other things to worry about at the time.

I especially found the reply from Phil Stripling interesting. I had
thought about the principle behind the image stacker at about the same
time and thought that it must be possible to get higher resolution
from video. After all, the video is sampling many small increments at
a time.

I have thought about it for a while whether there was anything lost
when I was not able to promote it further. Who knows, maybe there was
but that is not to worry about.

Regards, Lorne
Anonymous
December 28, 2004 11:41:38 PM

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

Lorne Tontegode <lorne___tontegode@@@hotmail...com> writes:

> thought about the principle behind the image stacker at about the same
> time and thought that it must be possible to get higher resolution
> from video. After all, the video is sampling many small increments at
> a time.

Now with digital video, there would be discrete frames of the image
available on the fixed disk (or other recording medium). The only issue
would be whether one wanted to have pixel-sized movements of the CCD or
other sensor.
--
Phil Stripling | email to the replyto address is presumed
The Civilized Explorer | spam and read later. email from this URL
http://www.cieux.com/ | http://www.civex.com/ is read daily.
Anonymous
December 29, 2004 12:23:28 AM

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

On Tue, 28 Dec 2004 08:11:31 -0500, Lorne Tontegode
<lorne___tontegode@@@hotmail...com> wrote:

>The method that I proposed was to expose the CCD to the image multiple
>times but with a slightly different location. This (I theorized) would
>make it possible to enhance the resolution of the CCD by 4, 9, or even
>16 times.

Lorne,

unless you take the same picture thousands of times, you will
never get close to those claimed resolution enhancements.

Otherwise the method is sound and does yield a somewhat higher
resolution.

Hans-Georg

--
No mail, please.
Anonymous
December 29, 2004 12:58:11 AM

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

In message <GieAd.7392$Ov3.3788@bignews1.bellsouth.net>,
"Eric Miller" <ericmiller@cox-internet.com> wrote:

>You probably don't mean this, but the same basic technique can be done by
>moving the whole camera (including the ccd) between exposures and stitching
>the images together with software. There is even a market for the tripod
>accessories that move the camera in measured steps. Search for the following
>terms in Google:

That's not what the OP was talking about. He was talking about moving
the sensor/subject registration by fractions of a pixel, like so:

1) on-pixel
2) 1/4 pixel to right
3) 1/2 pixel to right
4) 3/4 pixel to right
5) 1/4 pixel down
6) 1/4 pixel down, 1/4 pixel to right
....
16) 3/4 pixel down, 3/4 pixel to right

This would work best with a camera with no anti-aliasing filter, and the
final output would not be aliased, most likely.

A properly anti-aliased camera would have low pixel-to-pixel contrast,
and require a boost at the high frequencies.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
December 29, 2004 1:02:00 AM

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

In message <1104248369.047138.260100@z14g2000cwz.googlegroups.com>,
stauffer@usfamily.net wrote:

>There were a few pro cameras that used this technique, but it was very
>expensive, and as far as I know it is no longer being used. The
>mechanism that moves the image plane needs to move on a few microns-
>this is extremely difficult and expensive.
>
>It is not possible to move the whole camera in angle- the gimballing
>system would be horrendously expensive and there are cheaper
>alternatives.

If there are some distinct, hi-contrast registration points in the
image, software could upscale the images and register them at a
sub-original-pixel level, so you could feed it any amount of frames and
it would get more and more detail the more frames you delivered.

Again, this would work best with a weak AA filter, or none at all
(provided you really did capture a lot of sub-original pixel
information).
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
December 29, 2004 1:44:03 PM

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

In article <1104248369.047138.260100@z14g2000cwz.googlegroups.com>,
stauffer@usfamily.net says...
> There were a few pro cameras that used this technique, but it was very
> expensive, and as far as I know it is no longer being used. The
> mechanism that moves the image plane needs to move on a few microns-
> this is extremely difficult and expensive.

Not really - the CCD can be precisely moved with tiny piezo actuators,
which are quite cheap. The Minolta 8MP compact prosumer camera uses
these actuators to reduce camera shake (image stabilisation).
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5060 resource - http://myolympus.org/5060/
Olympus 8080 resource - http://myolympus.org/8080/
Anonymous
December 29, 2004 6:50:37 PM

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

On Tue, 28 Dec 2004 20:41:37 -0500, Lorne Tontegode
<lorne___tontegode@@@hotmail...com> wrote:

>it must be possible to get higher resolution
>from video. After all, the video is sampling many small increments at
>a time.

Lorne,

I'm sure that this is possible in theory. However, we are
currently short of processing power for that and, even more
importantly, we cannot currently program the very complex
abstraction algorithms that would be required.

We will see this in the future, however.

An example would be an algorithm that uses several pictures and
uses them, including light and shadow and 3D effects, to
reconstruct a 3D scene and all the objects in it. Using a
complete 3D model, the algorithm can then use information from
several different pictures to refine the 3D model far beyond the
information contained in just one picture.

Going from the 3D model, the algorithm could then refine each
picture.

One side effect of this realization is that, in expectation of
this capability, it may not be necessary to increase TV and
video camera resolution by much.

Hans-Georg

--
No mail, please.
Anonymous
December 29, 2004 6:50:41 PM

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

On 28 Dec 2004 17:52:03 -0800, Phil Stripling
<phil_stripling@cieux.zzn.com> wrote:

>Lorne Tontegode <lorne___tontegode@@@hotmail...com> writes:

>> thought about the principle behind the image stacker at about the same
>> time and thought that it must be possible to get higher resolution
>> from video. After all, the video is sampling many small increments at
>> a time.

>Now with digital video, there would be discrete frames of the image
>available on the fixed disk (or other recording medium). The only issue
>would be whether one wanted to have pixel-sized movements of the CCD or
>other sensor.

Phil,

it is not necessary, and actually not possible with moving
objects, to obtain several pictures that are offset exactly by a
certain subpixel distance.

Instead it is possible to juxtapose a series of pictures that
are slightly different, for example by relying on normal camera
shake.

The technology of taking a whole series of photos, rather than
one, has some other advantages as well. For example, it could
remove shutter lag or make it negative.

I predict that we will see this technology soon.

Hans-Georg

--
No mail, please.
Anonymous
December 29, 2004 6:50:42 PM

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

Hans-Georg Michna <hans-georgNoEmailPlease@michna.com> writes:

> it is not necessary, and actually not possible with moving
> objects, to obtain several pictures that are offset exactly by a
> certain subpixel distance.

Hi, Hans,

My apologies -- I was certainly not clear. My understanding from the
original poster was that he wanted to photograph stationary subjects
several times, moving the CCD or other sensor a pixel or so with each
discrete image capture. I was suggesting using video to capture a
stationary object over several frames, while moving the sensor a bit.

>SNIP<
> For example, it could
> remove shutter lag or make it negative.

I have had trouble with negative shutter lag for years, now.

--
Phil Stripling | email to the replyto address is presumed
The Civilized Explorer | spam and read later. email from this URL
http://www.cieux.com/ | http://www.civex.com/ is read daily.
Anonymous
December 29, 2004 11:49:37 PM

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

On 29 Dec 2004 10:31:42 -0800, Phil Stripling
<phil_stripling@cieux.zzn.com> wrote:

>Hans-Georg Michna <hans-georgNoEmailPlease@michna.com> writes:

>> For example, it could
>> remove shutter lag or make it negative.

>I have had trouble with negative shutter lag for years, now.

Phil,

hehe, but there was one camera (Olympus 1R or similar) that
actually had negative shutter lag. When you pressed the shutter,
you got the picture that had been taken shortly before.

The idea got lost, unfortunately, but I predict that it will
reappear as processor and memory chips become more powerful and
cheaper.

Hans-Georg

--
No mail, please.
Anonymous
December 29, 2004 11:49:38 PM

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

Hans-Georg Michna <hans-georgNoEmailPlease@michna.com> writes:
> hehe, but there was one camera (Olympus 1R or similar) that
> actually had negative shutter lag. When you pressed the shutter,
> you got the picture that had been taken shortly before.

E-100RS, and some Casio had it too. It wasn't as useful as it sounds.
If you want to buy me E-100RS, let me know.
Anonymous
December 29, 2004 11:49:39 PM

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

Paul Rubin <http://phr.cx@NOSPAM.invalid&gt; writes:
> If you want to buy me E-100RS, let me know.

Oops, typo. I was offering to sell my E-100RS, not asking you to buy
one for me. I should put the thing on ebay while it still has some value.
Anonymous
December 30, 2004 1:17:56 AM

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

In message <hh24t0l1fkgg7pud6rvrtunqv81on29rd8@4ax.com>,
Lorne Tontegode <lorne___tontegode@@@hotmail...com> wrote:

>I especially found the reply from Phil Stripling interesting. I had
>thought about the principle behind the image stacker at about the same
>time and thought that it must be possible to get higher resolution
>from video. After all, the video is sampling many small increments at
>a time.

Well, that's partly why video isn't too offensive. The brain creates a
higher-resolution image than the one that is present on the screen at
any given time. A video still is a horrible thing to look at
(especially one from VHS).
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
December 30, 2004 1:19:28 AM

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

In message <3qu0q5j6r5.fsf@shell4.tdl.com>,
Phil Stripling <phil_stripling@cieux.zzn.com> wrote:

>My understanding from the
>original poster was that he wanted to photograph stationary subjects
>several times, moving the CCD or other sensor a pixel or so with each
>discrete image capture.

It would have to be a 1/n (n>=2) fraction of a pixel.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
December 30, 2004 1:20:36 AM

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

In message <l326t0pu7gsantgbmgt1refnm3eevb5693@4ax.com>,
Hans-Georg Michna <hans-georgNoEmailPlease@michna.com> wrote:

>hehe, but there was one camera (Olympus 1R or similar) that
>actually had negative shutter lag. When you pressed the shutter,
>you got the picture that had been taken shortly before.
>
>The idea got lost, unfortunately, but I predict that it will
>reappear as processor and memory chips become more powerful and
>cheaper.

Yes, that was a 1.3MP camera. I almost bought one, just for the 15 fps
burst it offered, and buffered pre-shutter capture.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
December 30, 2004 1:20:37 AM

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

JPS@no.komm writes:
> Yes, that was a 1.3MP camera. I almost bought one, just for the 15 fps
> burst it offered, and buffered pre-shutter capture.

Olympus E-100RS. I have one but don't use it much. It's for sale.
Anonymous
December 31, 2004 12:05:37 AM

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

Kibo informs me that JPS@no.komm stated that:

>In message <3qu0q5j6r5.fsf@shell4.tdl.com>,
>Phil Stripling <phil_stripling@cieux.zzn.com> wrote:
>
>>My understanding from the
>>original poster was that he wanted to photograph stationary subjects
>>several times, moving the CCD or other sensor a pixel or so with each
>>discrete image capture.
>
>It would have to be a 1/n (n>=2) fraction of a pixel.

If your 'shutter speed' is high enough, jittering the camera randomly
has the same effect.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------
Anonymous
December 31, 2004 7:56:13 AM

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

In message <9lk7t0tgch36nhb8u282rlmjmok53flbdo@4ax.com>,
usenet@imagenoir.com wrote:

>Kibo informs me that JPS@no.komm stated that:
>
>>In message <3qu0q5j6r5.fsf@shell4.tdl.com>,
>>Phil Stripling <phil_stripling@cieux.zzn.com> wrote:
>>
>>>My understanding from the
>>>original poster was that he wanted to photograph stationary subjects
>>>several times, moving the CCD or other sensor a pixel or so with each
>>>discrete image capture.
>>
>>It would have to be a 1/n (n>=2) fraction of a pixel.
>
>If your 'shutter speed' is high enough, jittering the camera randomly
>has the same effect.

It's computationally simpler, though, if they all fall in a nice grid.

Did you see my new thread in another group about the 10D's ISO 1600? I
did some further checking of the RAW data, and it appears that ISO 1600
is, in fact, multiplied by 8 in the amplifier, and doubled digitally.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
December 31, 2004 5:11:11 PM

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

Hans-Georg Michna wrote:

> On Tue, 28 Dec 2004 08:11:31 -0500, Lorne Tontegode
> <lorne___tontegode@@@hotmail...com> wrote:
>
>
>>The method that I proposed was to expose the CCD to the image multiple
>>times but with a slightly different location. This (I theorized) would
>>make it possible to enhance the resolution of the CCD by 4, 9, or even
>>16 times.
>
>
> Lorne,
>
> unless you take the same picture thousands of times, you will
> never get close to those claimed resolution enhancements.
>
> Otherwise the method is sound and does yield a somewhat higher
> resolution.
>
> Hans-Georg
>
This has already been done. It was done with the
Mars Pathfinder lander and from the Mars orbiters,
and with tens of images, not thousands. Higher signal
to noise images would give better results.

Google: "Mars Pathfinder" super resolution

some sites:

SUPER-RESOLUTION RESULTS FROM PATHFINDER IMP. B. Kanefsky,
T.J. Parker, and P.C. Cheeseman.
NASA Ames, M/S 269–2, Moffett Field CA 94035, 2JPL, 3RIACS, NASA Ames.
http://mars.jpl.nasa.gov/MPF/science/lpsc98/1536.pdf

http://mpfwww.jpl.nasa.gov/MPF/ops/super-res-index.html

http://ic.arc.nasa.gov/projects/bayes-group/Atlas/Mars/...

Roger Clark
Anonymous
January 1, 2005 11:05:04 PM

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

On Fri, 31 Dec 2004 14:11:11 -0700, "Roger N. Clark (change
username to rnclark)" <username@qwest.net> wrote:

>Hans-Georg Michna wrote:

>> On Tue, 28 Dec 2004 08:11:31 -0500, Lorne Tontegode
>> <lorne___tontegode@@@hotmail...com> wrote:

>>>The method that I proposed was to expose the CCD to the image multiple
>>>times but with a slightly different location. This (I theorized) would
>>>make it possible to enhance the resolution of the CCD by 4, 9, or even
>>>16 times.

>> unless you take the same picture thousands of times, you will
>> never get close to those claimed resolution enhancements.
>>
>> Otherwise the method is sound and does yield a somewhat higher
>> resolution.

>This has already been done. It was done with the
>Mars Pathfinder lander and from the Mars orbiters,
>and with tens of images, not thousands. Higher signal
>to noise images would give better results.

Roger,

thanks! Interesting.

Note that I didn't say the method always requires thousands of
pictures. I only said that you may need thousands of pictures to
achieve a resolution of 4, 9, and 16 times higher than the
physical one. Actually a factor of 4 may not be so difficult to
achieve.

Hans-Georg

--
No mail, please.
Anonymous
January 1, 2005 11:05:05 PM

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

"Hans-Georg Michna" <hans-georgNoEmailPlease@michna.com> wrote in message
news:gdrdt0l8mee3vhkvbn8aj93tfkaq3nc8e1@4ax.com...
> On Fri, 31 Dec 2004 14:11:11 -0700, "Roger N. Clark (change
> username to rnclark)" <username@qwest.net> wrote:
>
>>Hans-Georg Michna wrote:
>
>>> On Tue, 28 Dec 2004 08:11:31 -0500, Lorne Tontegode
>>> <lorne___tontegode@@@hotmail...com> wrote:
>
>>>>The method that I proposed was to expose the CCD to the image multiple
>>>>times but with a slightly different location. This (I theorized) would
>>>>make it possible to enhance the resolution of the CCD by 4, 9, or even
>>>>16 times.
>
>>> unless you take the same picture thousands of times, you will
>>> never get close to those claimed resolution enhancements.
>>>
>>> Otherwise the method is sound and does yield a somewhat higher
>>> resolution.
>
>>This has already been done. It was done with the
>>Mars Pathfinder lander and from the Mars orbiters,
>>and with tens of images, not thousands. Higher signal
>>to noise images would give better results.
>
> Roger,
>
> thanks! Interesting.
>
> Note that I didn't say the method always requires thousands of
> pictures. I only said that you may need thousands of pictures to
> achieve a resolution of 4, 9, and 16 times higher than the
> physical one. Actually a factor of 4 may not be so difficult to
> achieve.
>
> Hans-Georg
>
> --
> No mail, please.\

Am I missing something? This is just what panorama software does.

Tom
Anonymous
January 1, 2005 11:05:06 PM

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

Tom Scales wrote:

> "Hans-Georg Michna" <hans-georgNoEmailPlease@michna.com> wrote in message
> news:gdrdt0l8mee3vhkvbn8aj93tfkaq3nc8e1@4ax.com...
>
>>On Fri, 31 Dec 2004 14:11:11 -0700, "Roger N. Clark (change
>>username to rnclark)" <username@qwest.net> wrote:

>>>Hans-Georg Michna wrote:
>>
>>>>On Tue, 28 Dec 2004 08:11:31 -0500, Lorne Tontegode
>>>><lorne___tontegode@@@hotmail...com> wrote:
>>
>>>>>The method that I proposed was to expose the CCD to the image multiple
>>>>>times but with a slightly different location. This (I theorized) would
>>>>>make it possible to enhance the resolution of the CCD by 4, 9, or even
>>>>>16 times.
>>
>>>>unless you take the same picture thousands of times, you will
>>>>never get close to those claimed resolution enhancements.
>>>>
>>>>Otherwise the method is sound and does yield a somewhat higher
>>>>resolution.
>>
>>>This has already been done. It was done with the
>>>Mars Pathfinder lander and from the Mars orbiters,
>>>and with tens of images, not thousands. Higher signal
>>>to noise images would give better results.
>>
>>Roger,
>>
>>thanks! Interesting.
>>
>>Note that I didn't say the method always requires thousands of
>>pictures. I only said that you may need thousands of pictures to
>>achieve a resolution of 4, 9, and 16 times higher than the
>>physical one. Actually a factor of 4 may not be so difficult to
>>achieve.
>>
>>Hans-Georg
>>
>>--
>>No mail, please.\
>
>
> Am I missing something? This is just what panorama software does.
>
> Tom

Panorama software is for mosaicing multiple images together,
but does not improve the spatial resolution of a single field of view.
Say you have a 50mm lens and take a picture. If you want
to double the resolution and have the same field of view,
you can:

1) quadruple the number of pixels and use the same lens
(assuming the lens has the resolution),

2) double your focal length, take 4 frames to cover the
field of view and mosaic them together with panorama
software, or

3) take multiple images with the one lens, moving the
sensor a fraction of a pixel and use software to sort out
the small differences in the multiple images to increase
spatial resolution.

This thread is talking about #3.

Roger
Anonymous
January 2, 2005 3:38:12 AM

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

"Tom Scales" <tom@scalesfamily.com> wrote in message
news:T%EBd.4240$Me4.3554@bignews4.bellsouth.net...
SNIP
> Am I missing something? This is just what panorama software does.

What the OP and a few other posters are talking about is small pixel
offsets, e.g. less than a pixel. Many samples at slightly different
positions will allow to calculate higher resolution images than would
be possible to record with a single exposure form the same sensor.

Bart
Anonymous
January 2, 2005 3:38:13 AM

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

"Bart van der Wolf" <bvdwolf@no.spam> wrote in message
news:41d73466$0$6219$e4fe514c@news.xs4all.nl...
>
> "Tom Scales" <tom@scalesfamily.com> wrote in message
> news:T%EBd.4240$Me4.3554@bignews4.bellsouth.net...
> SNIP
>> Am I missing something? This is just what panorama software does.
>
> What the OP and a few other posters are talking about is small pixel
> offsets, e.g. less than a pixel. Many samples at slightly different
> positions will allow to calculate higher resolution images than would be
> possible to record with a single exposure form the same sensor.
>
> Bart
>

Yes, I understand that now. It seems that the effort to do so would soon be
overwhelmed by the increases in the MP of sensors; at least at a reasonable
price point.
Anonymous
January 2, 2005 3:38:14 AM

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

Tom Scales wrote:
> "Bart van der Wolf" <bvdwolf@no.spam> wrote in message
> news:41d73466$0$6219$e4fe514c@news.xs4all.nl...
>
>>"Tom Scales" <tom@scalesfamily.com> wrote in message
>>news:T%EBd.4240$Me4.3554@bignews4.bellsouth.net...
>>SNIP
>>
>>>Am I missing something? This is just what panorama software does.
>>
>>What the OP and a few other posters are talking about is small pixel
>>offsets, e.g. less than a pixel. Many samples at slightly different
>>positions will allow to calculate higher resolution images than would be
>>possible to record with a single exposure form the same sensor.
>>
>>Bart
>>
>
>
> Yes, I understand that now. It seems that the effort to do so would soon be
> overwhelmed by the increases in the MP of sensors; at least at a reasonable
> price point.
>
>
Not necessarily. For example, the Mars landers: you can't
upgrade once launched. Here is another example, where the
limit is not the sensor itself, but atmospheric turbulence
and lens quality:

Saturn with a Telephoto Lens:
http://clarkvision.com/astro/saturn.03.02.2004

99 images were were registered (an automated process
with some software), averaged, and filtered to increase
spatial resolution by about 3 times. A sensor with
more pixels would not help.

Roger
!