PDA

View Full Version : Why are digital images taken by Mars explorers so low quality and why don't they use a camcorder??


Perdit
January 5th 04, 11:10 AM
Also why were images from the moon (1st landing) so low quality?

Davoud
January 5th 04, 12:21 PM
Perdit:
> Also why were images from the moon (1st landing) so low quality?

Talk about your nattering nabob of negativism!

The television images of the first humans to walk on the moon were the
best live TV I had ever seen from the moon -- or from anywhere else,
for that matter.

As for the current photos from Mars, they have been excellent, too. If
you'll be patient, and if all goes well, they'll get better. You want
they should have sent an Eisenstadt?

Davoud

--
usenet *at* davidillig dawt com

Chris L Peterson
January 5th 04, 02:23 PM
On 5 Jan 2004 03:10:20 -0800, (Perdit) wrote:

>Also why were images from the moon (1st landing) so low quality?

Well, how about because they were the product of a low power, analog broadcast
from a quarter million miles away? Normal TV broadcast reception wasn't so
wonderful back then, either. The film images certainly weren't low quality!

And what is wrong with the Mars images? They look good to me. Certainly many of
the first batch were taken under poor lighting conditions, and the posted raw
images need a little processing. I took a few into Photoshop and stretched the
range and contrast a little. Maybe you are just used to seeing the images after
JPL has processed them.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

Gary Heath
January 5th 04, 09:11 PM
Good Questions.
First .. they didn't have Richard Crisp taking them!! ;-)
Actually, the technology of the time during the late 60's/early 70's was
just not all that well developed, to make the lunar videos sent from the
moon to earth, as good as we can do these days.
Just plain a result of the technology of the time.
I would add that the still images and film brought back to earth is as good
as any of today (IMO).
The situation with Sprit is a bit more complex.
The initial data being sent back is coming from a low gain antenna that is
capable of about 1kb/sec.
It has to share this slow data stream with a lot of other non image data
this is crowding the initial flow from the rover.
The high gain antenna is up and running (as of last night) and as the
mission develops further, and until more defined objectives (that remain
fluid until checkout, systems calibration, and ultimately spectroscopy) are
accomplished, the picture quality will be on a ramp from low res, black &
white, to what I believe will be the truly awe inspiring color images to
come.
A lot of data needs to be funneled through the antenna., and even the high
gain is only on par with a 12K modem.
That, and (of course) ... they need Mr. Crisp!!
Awesome stuff you do Richard.

Gary


"Perdit" > wrote in message
m...
> Also why were images from the moon (1st landing) so low quality?

Perdit
January 6th 04, 12:03 PM
Ok, thanks. And it looks like an well-informed answer to me.

However I do not quite understand the relation between 'modem' speed
and image quality. 1k/s would be enough to produce a good quality
image in little time, above all if you use jpg compression. The
non-martian landscape in front of my house which I snap with my 2mp
kodak looks wonderful at 150-200k (a 3 minute download at 1k/s).


Also why can we see the 'stitches' in the panoramic picture
composites??

I think there's still something we are missing here. I wish somebody
from Nasa was so kind to give me a minute of his time and honor us
with a response.

Martin Brown
January 6th 04, 01:27 PM
In message >, Perdit
> writes
>Ok, thanks. And it looks like an well-informed answer to me.
>
>However I do not quite understand the relation between 'modem' speed
>and image quality. 1k/s would be enough to produce a good quality
>image in little time, above all if you use jpg compression. The
>non-martian landscape in front of my house which I snap with my 2mp
>kodak looks wonderful at 150-200k (a 3 minute download at 1k/s).

I think you will find modem speeds are specified in bits per second so
you are being roughly an order of magnitude optimistic about transfer
times. It would take it around half an hour to send your image back.
>
>Also why can we see the 'stitches' in the panoramic picture
>composites??

Almost certainly it is policy to use the images without any cosmetic
molestation. Quicker to do and makes it obvious where the frame edges
are. Scientific images are not necessarily pretty provided they are fit
for purpose.

Regards,
--
Martin Brown

Chris L Peterson
January 6th 04, 03:41 PM
On 6 Jan 2004 04:03:53 -0800, (Perdit) wrote:

>Also why can we see the 'stitches' in the panoramic picture
>composites??

Because nobody has taken the time to construct a high quality mosaic off these
original images. I'm sure it isn't a high priority at this point.

I'm still curious in what way you find the images poor. They look excellent to
me- good resolution, low noise, what more do you want? Obviously, many were
made under low light conditions. I have been bringing some of the raw images
into Photoshop and adjusting the curves for better brightness and contrast, but
that's just basic processing. Other than the small size, which I attribute to
limited bandwidth, the images are great.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

Roger Hamlett
January 6th 04, 04:17 PM
"Perdit" > wrote in message
om...
> Ok, thanks. And it looks like an well-informed answer to me.
>
> However I do not quite understand the relation between 'modem' speed
> and image quality. 1k/s would be enough to produce a good quality
> image in little time, above all if you use jpg compression. The
> non-martian landscape in front of my house which I snap with my 2mp
> kodak looks wonderful at 150-200k (a 3 minute download at 1k/s).
Part of the 'key' to how modems achieve the rates they do, is that the link
is bidirectional, and if data is lost, the receiver can ask for a
retransmission. This is not possible with a space probe (think of the light
speed delays involved), so normally data is encoded to allow sections to be
lost, and yet the final results to be reliably retrieved. The 'downside' of
this, is that this has the opposite effect to compression, making the amount
of data that has to be sent increase. Speaking scientifically, any 'lossy'
compression like jpeg, would not be acceptable either. The data rate also
has to be shared with other sensors on the craft. The craft data rate, is
typically around 8000bits/second, which with encoding to protect against
data loss, a typical picture from a good quality scientific imager (say 5M
pixels, at 16bits/pixel, shot through three seperate colour filters, and
then using a hamming code), would take 5625 seconds to send!. Even a single
video frame (say 307000 pixels, each using 8bits/pixel), would take 307
seconds to send. This can be bettered (very significantly), using lossless
compression, but pictures will still take a _long_ time. The fastest
link 'back', is by sending to the orbiter, then from this to Earth, which
can receive about 7.5Mbytes of data in 8 minutes, but only once per Mars
day. Direct transmissions to Earth, are also limited to a maximum of about
three hours per day, by power (and heat) limitations.
The link back, is also 'shared' by the various craft, further restricting
the amount that each craft can send.
The panoramic camera, can generate an image, that is 4000*24000 pixels. To
shoot colour, would require three such images, giving one 'image', needing
576MB of data. Even if carefully compressed, then sent, this would occupy
the entire transmission time for a couple of days!.

> Also why can we see the 'stitches' in the panoramic picture
> composites??
Because it takes a human time, to tidy up such joints.

> I think there's still something we are missing here. I wish somebody
> from Nasa was so kind to give me a minute of his time and honor us
> with a response.
You are missing, two things. The first is just how restricted the 'link' is,
and the second is that pictures that look pretty, will often have lost a lot
of the scientific data they may contain. The pictures being sent, are
targetted towards making accurate colour measurements, not images that
necessarily look pretty. Hence the massive compressions typically used on
consumer cameras, are not suitable for spacecraft data. The pretty pictures
will appear, as processors on the ground re-assemble the data, and cover the
joints...

Best Wishes

Gary Heath
January 6th 04, 05:57 PM
Thanks Roger.
This is a much better answer than I managed to provide.

Gary

"Roger Hamlett" > wrote in message
...
>
> "Perdit" > wrote in message
> om...
> > Ok, thanks. And it looks like an well-informed answer to me.
> >
> > However I do not quite understand the relation between 'modem' speed
> > and image quality. 1k/s would be enough to produce a good quality
> > image in little time, above all if you use jpg compression. The
> > non-martian landscape in front of my house which I snap with my 2mp
> > kodak looks wonderful at 150-200k (a 3 minute download at 1k/s).
> Part of the 'key' to how modems achieve the rates they do, is that the
link
> is bidirectional, and if data is lost, the receiver can ask for a
> retransmission. This is not possible with a space probe (think of the
light
> speed delays involved), so normally data is encoded to allow sections to
be
> lost, and yet the final results to be reliably retrieved. The 'downside'
of
> this, is that this has the opposite effect to compression, making the
amount
> of data that has to be sent increase. Speaking scientifically, any 'lossy'
> compression like jpeg, would not be acceptable either. The data rate also
> has to be shared with other sensors on the craft. The craft data rate, is
> typically around 8000bits/second, which with encoding to protect against
> data loss, a typical picture from a good quality scientific imager (say 5M
> pixels, at 16bits/pixel, shot through three seperate colour filters, and
> then using a hamming code), would take 5625 seconds to send!. Even a
single
> video frame (say 307000 pixels, each using 8bits/pixel), would take 307
> seconds to send. This can be bettered (very significantly), using lossless
> compression, but pictures will still take a _long_ time. The fastest
> link 'back', is by sending to the orbiter, then from this to Earth, which
> can receive about 7.5Mbytes of data in 8 minutes, but only once per Mars
> day. Direct transmissions to Earth, are also limited to a maximum of about
> three hours per day, by power (and heat) limitations.
> The link back, is also 'shared' by the various craft, further restricting
> the amount that each craft can send.
> The panoramic camera, can generate an image, that is 4000*24000 pixels. To
> shoot colour, would require three such images, giving one 'image', needing
> 576MB of data. Even if carefully compressed, then sent, this would occupy
> the entire transmission time for a couple of days!.
>
> > Also why can we see the 'stitches' in the panoramic picture
> > composites??
> Because it takes a human time, to tidy up such joints.
>
> > I think there's still something we are missing here. I wish somebody
> > from Nasa was so kind to give me a minute of his time and honor us
> > with a response.
> You are missing, two things. The first is just how restricted the 'link'
is,
> and the second is that pictures that look pretty, will often have lost a
lot
> of the scientific data they may contain. The pictures being sent, are
> targetted towards making accurate colour measurements, not images that
> necessarily look pretty. Hence the massive compressions typically used on
> consumer cameras, are not suitable for spacecraft data. The pretty
pictures
> will appear, as processors on the ground re-assemble the data, and cover
the
> joints...
>
> Best Wishes
>
>
>

Perdit
January 7th 04, 08:21 PM
OK, Thanks I now better understand the limitations but you can agree
with me that it should be a priority to get some crystal-clear images,
now or after procecessing or whenever.

Even the latest 12Mp image that is being hyped is good, right, but it
does not look like 12MB clarity.

This is the CCD imagin forum. I am sure many of you have a decent 3-5
Mp digital camera. Take a picture of any surface landscapè. The
clarity from this 3-5 cameras is amazing. I am impressed by the 12Mp
Mars image, but not because of the clarity but because it is Mars and
it had never seen like that before.


Looking at other images in the same set what i can see is lightly out
of focus stones. That's pretty obvious and any photo enthusiast can
see that.

Gary W. Swearingen
January 20th 04, 01:36 AM
(Perdit) writes:

> Looking at other images in the same set what i can see is lightly out
> of focus stones. That's pretty obvious and any photo enthusiast can
> see that.

The cameras are fix-focused on infinity, so close stuff is fuzzy.
Except, IIRC, the microscopic camera, which has some focusing
ability (in addition to moving the arm that carries it).

The cameras are 1024 x 1024, more-or-less state-of-the-art when
the system was designed. The two pancams shoot through different
narrow-band filters (except one can shoot through no filter and
they both can shoot through one particular (infrared?) "color").

The Martian air is dusty, but I don't know if it's enough to be
noticible, other than in the sky.

Communication is limited to a few hours a day, IIRC, as Mars rotates
and the relay satellites orbit. Some of the flow is for a bit of
engineering data and other instruments like spectrometers.