Bad Noise in Low Light ... is this normal

The dark vertical cables etc all have purple or blue fringing in the pod picture. The tokina is showing CA in the backlit tree branches.

Like you say it is possible to get very good pictures with manual lenses - I have a few myself, but common to most of them is either softness or CA. Nothing too bad, but noticeable.

I do find myself wondering why you bought 24mm and 135mm lenses when you already have the kit lenses. I've done something similar and now find myself thinking that it was a waste - fair enough a waste of only £30 or £40 in total.

Anyhow, nice shots of london life in tourist land.

Regards

Ian
 
That is what I just said, did'nt I?...I did not say they were a
perfect fit, but they DO fit and for Shoei thats all he need to get
up and running with M42 lenses.
Sooner or later one wants an adapter that does not wobble and goes for another one. Therefore the best one could do is pay the few extra bucks and get the real one right from the start.
And this is why you shoot it in manual mode with greatly varying
exposure bias?
No, I dont go by exposure bias at all....I check my histogram on
every shot and alter the aperture or shutter speed to compensate if
its a little underexposed.
Which is what Adam described as "guess via LCD". You said stopped down metering would work, hint it does not work all that great. Anyway if you check the histogram and shoot by that why do you still need to push some images nearly one f-stop in SPP?
So what?...I leave SPP2 set to auto and trust its image processing
decisions...
You said stop down metering would work, seeing how much exposure adjustment is done in SPP sort of shows that it does not work all that good.
To me its just right for MF lenses, but maybe my eyesite is just
very good?
No, not at all....I ALWAYS use maximum sharpening in SSP2 for EVERY
picture I take, and that goes for pics taken with my AF kit lenses
too, because I like pin sharp images with maximum detail.
If you would do that with a properly focussed image taken with a sharp lens it would show ugly halos and stairstepping. Your images don't show that, therefore they are either done with soft lenses or out of focus. And they are still not pin sharp. I got much sharper images with the Kit lenses and even with a really bad 24-70 f/3.5-5.6.
...Which is why I bought the SD10, in the first place!
Some Sigma owners moan about "jaggies" in their images when they
No not owners, I only saw very few owners complaining about jaggies. Most of the complaints came from Canon and Fuji users.
A magnifier can be cheap,
But not good in that case.
Faster in what way?...In focusing?...I can MF focus very quickly,
but then maybe I am just some kind of superhero?...Manual Focus
Man..perhaps?
You are just too full of yourself. Thats all.
Er...I think you will find that the dustspots you are referring to
are on the sensor, not the lens, so stopping down the lens has
nothing to do with it.
Ouch no you show how much you really know. Hint: nothing.

The more you stop down the more defined are the dustspots.

My point is the Vivitar lens was stopped down quite a lot (judging by the dustpots) and still shows very ugly CA. More CA than I ever saw from any Sigma Lens, even the Sony 828 shows less.
I think you are exagerating Dominic.
And you are very euphemistic, to put it in a nice way...
and detail is what I am trying to capture, and I IMO am getting a
massive head start by using the SD10 rather than a Bayer sensored
DSLR.
I don't think you need to argue about that point with me.

--
http://www.pbase.com/sigmasd9/dominic_gross_sd10

 
Yes, thats right, but if you read it properly you will see that I
inferred that, by not stating that they were all good, only that
many were.
I think many people will agree when I say that there are more junk lenses than good ones with m42 mount. Is this Vivitar 400mm one of the good Lenses? This imho belongs in the cr@p category as well...
Where exactly?....Going blind trying to find any!



Again, where exactly?....I am struggling to find any in 02656
This image is full of it. You said your eyesight was good??

that is worser than anything I saw till today from most lenses:



And here is another one:


Yes, on the highly reflective steel drum but certainly much less
than I would have got using a Sigma zoom
Sorry this is a ridicoulus claim, too bad I can't prove it by shooting the same situation with a Sigma zoom.
and there is no CA visible
anywhere else in the picture.
Err yes tell me where the green shiny edge comes from:



What is really noticable is that for example the Vivitar is stopped down quite a bit and we are looking on pictures taken in a situation with rather low contrast. Still there is a lot of CA, not just noticable CA, just too much of it. For example with the 18-50 Kit Lens I could take a picture of branches against very bright backlight (sunlight) and it showed about the same CA your Tokina shows on an overcast day between two equally bright objects (for example jeans and pavement on leicster square in the sample above).

--
http://www.pbase.com/sigmasd9/dominic_gross_sd10

 
I'm sticking with Bayer sensors. Not because I want to.
Your comments are articulate, interesting, fascinating and highly informative. I thank you for them. But the most telling statement is the one above. You've stayed with the Bayer technology because you HAVE to. As has been posted here and elsewhere, the digital Sigmas are not for everyone and not for every application...certainly not for yours.

But my SD-9 is exactly right for my needs and outperforms every other digital SLR I've tried, borrowed and read about, including several at twice the price. The pics it presents are almost always astonishing and most (99%) are without noise or artifacts of any kind.

Each to his own....
SOMEBODY has to be an early-adapter of emerging technologies,
otherwise those technologies stagnate and die. Remember the first
mobile phones? The bricks? Would you have one now? Not in your
right mind. At the time, however, they were great, and people took
them up anyway. Today's Foveon is the brick. It has great promise,
and I look very much forward to my first Foveon-style SLR.
Interesting thing about those "bricks:" they had 3-watt audio amplifiers instead of the .25 and .5 watts that are in most cell phones today and had much wider frequency response. I had no trouble whatsoever hearing the person at the other end of the phone, even when traveling at 70mph.

Just because something is a "brick" doesn't necessarily mean its evolved versions are superior in every respect. Early tv's were all vacuum tubes. If you were to compare the picture of an old (but high qaulity)vacuum tube color tv of yesteryear with any of the current crop of solid-state tv's out today, you would be amazed at the superior depth and realism of the vacuum tube tv picture. Other things about the current tv's are better, but not the sheer accuracy of the reproduction. Same thing with audio components: vacuum tube audio components, even very old ones, are often vastly superior to their newer solid-state counterparts when it comes to making real music sound like real music ... but they're far more expensive, too. So 99% of buyers opt for solid-state.

"Bricks" develop into smaller, cheaper more widely available products because the demand for the new technology increases substantially and because the demand is from the mass-market, who can't and won't pay the "brick" price. (The new Polaroid with Foveon is a good example.) But this wider distribution isn't an automatic guarantee of an improved or superior product.

And so, my old Canon G-III Q-17 (otherwise known as "the poor man's Leica" takes far superior film pictures than just about any point and shoot available today. The point-and-shoots, though, are easeir to operate than my Q-17, are lighter and hardly require as much care. But they do NOT take better pictures.

Picture wise, I believe that my Sigma digital SLR — in its price range and for its applications — will not be bettered for quite some time, and maybe not at all in my lifetime. Other cameras using Foveon technology — including any that Sigma has in mind for the near or far future — may be easier to use, smaller, lighter and more feature-laden, including built-in JPEG capture and even higher ISO. But it's not likely that the end result —the pics — will be superior. I will invest in better lenses before ever considering another digital camera.

Thanks again for your comments.....

— SteveG
 
Hi Shoei,

I have had quite a few images that look a lot like http://www.pbase.com/image/28520980

but never more than 2 consecutive. Am I understanding that you are able to repro this on demand?

If so might I suggest that you try it with the camera running from ac current instead of batteries and see if that helps. I'm beginning to suspect that low batteries may be contributing to this noise and to the occasional corrupt images I get.

thanks,

Mike
 
I took the liberty of applying the "Pattern" Noise Reduction Action for PS, that I developed, to this image.
Here is the Original as posted by Shoei



Here is the image after the PS action has been applied.



It is at least partially rescued by the action.

The action is posted on Gary Bainbridge's site
http://bainb.dyndns.org/Foveon/Removing_Foveon_ 'Pattern' Noise.html

Kent
Hi Shoei,

I have had quite a few images that look a lot like
http://www.pbase.com/image/28520980

but never more than 2 consecutive. Am I understanding that you are
able to repro this on demand?

If so might I suggest that you try it with the camera running from
ac current instead of batteries and see if that helps. I'm
beginning to suspect that low batteries may be contributing to this
noise and to the occasional corrupt images I get.

thanks,

Mike
--
Kent Dooley
 
I have written a PS Action which will clean up the blue sky noise problem.

Here is the original Image



Here it is acter using the PS Action



The action is described in this posting.
http://forums.dpreview.com/forums/read.asp?forum=1027&message=8544267

I can e-mail the action to anyone who is interested.

Kent
As another comment, it's very very dissapointing when my old 2MP
Bayer pattern Nikon Coolpix with no Apeture/Shutter priority on 8
second exposure completely wipes the floor with my new $2000 DSLR
when it comes to low light cityscapes.

e.g. this is from the old Nikon
http://www.pbase.com/image/28332449

This can't be right !!!

After reading the forum I thought this was not going to be a big
issue, appears I was wrong. In addition there was an observation of
a photo I posted a couple of days ago that the sky in it was
"blotchy". This was in full sunlight,
http://www.pbase.com/image/28456925

So .... I'm going to go for a walk around the block now and count
to ten
:-(

Cheers....

Shoei
So does this mean even though the Sigma comes with exposures up to
30 seconds and allegedly improved low light performance in the
SD10, you can't use them longer than about 2.5 seconds without
expecting a lot of really bad noise and above 5-10 secs it's
unusable !!!

Now I'm particularly unhappy. It also means Phils review of the low
light performance was spot on.

please someone tell me I am doing something wrong and it's not the
SD10 camera in general! Should I send it back to Sigma's
distributer in Australia to try to fix it or is it unfixable and I
should have gone for a different Camera. I really hope this can be
fixed (or at least alleviated).

in a word

AAaarrrgh !!!!

Cheers....

Shoei
a friend of me send his SD10 to sigma-germany - for killing the
noise-problem and set the "bulb"-funktion to the right funktion (so
long as you press)

now the cam ist back.......

the "bulb" is working with 15 secound (no change)
and the "green" noise is now a "blue" noise!!
a work of 3 weeks by sigma-service!

Ulli

(sorry about my english)
Hi All,

First of all thanks for the support, this group has been great.

Continuing my test of the SD10 in low light I went for a drive this
evening at dusk and shot a few images. After looking at the images
I've noticed a lot of Green noise. I thought this had been
corrected in the SD-10 so I thought I'd post and find out if it is
so.

Here are the first two "cityscape" shots. The first is as it came
out of the camera (no auto or custom SPP tweaks), the second (which
shows up the noise) was with a little "fill light" tweaking in SPP.

http://www.pbase.com/image/28520800
http://www.pbase.com/image/28520889

I was a little confused about this, and so I shot an F20 15sec
exposure of the front of my house. The results were absolutely
terrible (unusable). This is the result as it came from the Camera
(no auto or custom SPP tweaks).

http://www.pbase.com/image/28520980

Is this normal ?
Is anyone else seeing this
Is it something I'm doing wrong ?
have I picked up a lemon ?

Any assistance would be greatly appreciated

Regards,

Mark

--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--
Sorry, i will learn more english:-)
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--
Kent Dooley
 
Yes, thats right, but if you read it properly you will see that I
inferred that, by not stating that they were all good, only that
many were.
I think many people will agree when I say that there are more junk
lenses than good ones with m42 mount. Is this Vivitar 400mm one of
the good Lenses? This imho belongs in the cr@p category as well...
Ok, we will have to agree to differ on this because you obviously cannot see that we are talking about lenses costing on average less than £20 and you are trying to compare them to lenses costing on average £500 or more!
Where exactly?....Going blind trying to find any!
Edge of picture...Does not count!
Edge of picture...Does not count!
Again, where exactly?....I am struggling to find any in 02656
This image is full of it. You said your eyesight was good??

that is worser than anything I saw till today from most lenses:

Edge of Picture!...Does not count!
And here is another one:

Edge of picture...Does not count!...Why?...Because all of your examples can be cropped out and you would have no CA to moan about.

I dont know of a single M42 lens that is as sharp at the edges of the image as it is in the central part.

At the end of the day, M42 lenses were not designed for use on an APS sized sensor as used by Sigma.

The new DC lenses are an exception as they were designed to be sharp edge to edge over the whole image but you still get CA and that can be anywhere, across the whole image.

Show me some CA from anywhere in the central part of any of my M42 images, then I may concede your point.
Yes, on the highly reflective steel drum but certainly much less
than I would have got using a Sigma zoom
Sorry this is a ridicoulus claim, too bad I can't prove it by
shooting the same situation with a Sigma zoom.
Not rediculous at all...Are you suggesting that a zoom lens can have less CA than a Prime lens?....Now that is rediculous!

I have taken pictures of the same subject using my 18-50mm Sigma zoom lens which showed lots of CA and my 24mm Tokina which showed none... nuff said!
and there is no CA visible
anywhere else in the picture.
Err yes tell me where the green shiny edge comes from:

God!, you are a nit picker!...Its out of focus, so it does not matter or count but I will humour you anyway....The writing on his T-shirt is green and his jacket has a green lining...OK?
What is really noticable is that for example the Vivitar is stopped
down quite a bit and we are looking on pictures taken in a
situation with rather low contrast. Still there is a lot of CA, not
just noticable CA, just too much of it.
Now you really are exagerating!....What is your point?...That there is CA at the edges of my images?...Well whoppie do!...so what!....It can be cropped out, so its not important or noticable if you look at the central part which is where you are supposed to be looking at.
For example with the 18-50
Kit Lens I could take a picture of branches against very bright
backlight (sunlight) and it showed about the same CA your Tokina
shows on an overcast day between two equally bright objects (for
example jeans and pavement on leicster square in the sample above).
But my Tokina 24mm has no CA in the central part, where the 18-50mm has CA across the whole image!
 
Sooner or later one wants an adapter that does not wobble and goes
for another one. Therefore the best one could do is pay the few
extra bucks and get the real one right from the start.
Yes, but he only needs cheap one to try it out and see if he likes using M42 lenses, which he may not so why should he shell out for an expensive one for that?
Which is what Adam described as "guess via LCD". You said stopped
down metering would work, hint it does not work all that great.
Anyway if you check the histogram and shoot by that why do you
still need to push some images nearly one f-stop in SPP?
I dont push anything!...as I quite clearly stated...

"So what?...I leave SPP2 set to auto and trust its image processing
decisions..."

Not that it matters but you can therefore blame SPP2 for pushing it the one F-stop, not me!
You said stop down metering would work, seeing how much exposure
adjustment is done in SPP sort of shows that it does not work all
that good.
I have no problem with it...Nuff said!
No, not at all....I ALWAYS use maximum sharpening in SSP2 for EVERY
picture I take, and that goes for pics taken with my AF kit lenses
too, because I like pin sharp images with maximum detail.
If you would do that with a properly focussed image taken with a
sharp lens it would show ugly halos and stairstepping.
Your images
don't show that, therefore they are either done with soft lenses or
out of focus.
No, maybe its because I am using my camera and M42 lenses correctly?

? And they are still not pin sharp.

With the eception of the 400mm, they are very sharp and thats good enough.

The Tokina 24mm is sharper than the 18-50mm kit lens, has less CA and almost as much contrast, yet it was 3 times cheaper!

I got much sharper
images with the Kit lenses and even with a really bad 24-70
f/3.5-5.6.
Yeah, right!....Dream on!...Show me your evidence.
A magnifier can be cheap,
But not good in that case.
Good enough for me!
Er...I think you will find that the dustspots you are referring to
are on the sensor, not the lens, so stopping down the lens has
nothing to do with it.
Ouch no you show how much you really know. Hint: nothing.

The more you stop down the more defined are the dustspots.
Er...Now you are showing how little you know!...stopping down the lens makes absolutely no difference to the apperence of the dustspots from dust on the sensor, they simply show up more due to SPP2 boosting the contrast...i.e. Stopping down=less light coming in=under exposure=SPP2 boosting contrast.
They are still there, no matter what lens is used.
My point is the Vivitar lens was stopped down quite a lot (judging
by the dustpots) and still shows very ugly CA.
Yes, I always stop down my lenses to get greater DOF and sharpness.
More CA than I ever
saw from any Sigma Lens, even the Sony 828 shows less.
But where?,...At the edge of the images, which could be cropped out anyway so it does not really matter.
I think you are exagerating Dominic.
And you are very euphemistic, to put it in a nice way...
I will take that as a compliment!
 
Hi docmass,

Repeated the infamous "Shoei Test" of 5 x 15 second shots using the AC Transformer instead of batteries.

again, the sequrence is
1 x shot 15 sec - wait until the activity LED stops after the shot.

3 x shots 15 sec - as quickly one after the other as possible - after third shot wait until the activity LED stops
1 x shot 15 sec - wait until the activity LED stops after the shot.

and .....
...... the problem is 100% repeatable.

you get first good, second bad, third worse 4th disgusting 5th fine.

Darkframe processing interruption theory is holding

Batteries appear to not the problem. Good thought though, and worth investigating.

Cheers....

Mark
Hi Shoei,

I have had quite a few images that look a lot like
http://www.pbase.com/image/28520980

but never more than 2 consecutive. Am I understanding that you are
able to repro this on demand?

If so might I suggest that you try it with the camera running from
ac current instead of batteries and see if that helps. I'm
beginning to suspect that low batteries may be contributing to this
noise and to the occasional corrupt images I get.

thanks,

Mike
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
 
Thanks for that,

I'll have a look at it when I get some PS time :-)

Cheers...

Shoei


Here is the image after the PS action has been applied.



It is at least partially rescued by the action.

The action is posted on Gary Bainbridge's site
http://bainb.dyndns.org/Foveon/Removing_Foveon_ 'Pattern' Noise.html

Kent
Hi Shoei,

I have had quite a few images that look a lot like
http://www.pbase.com/image/28520980

but never more than 2 consecutive. Am I understanding that you are
able to repro this on demand?

If so might I suggest that you try it with the camera running from
ac current instead of batteries and see if that helps. I'm
beginning to suspect that low batteries may be contributing to this
noise and to the occasional corrupt images I get.

thanks,

Mike
--
Kent Dooley
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
 
Hi Kent,

Extremely interested

If you could send me the action that'd be great.

Cheers....

Shoei
Here is the original Image



Here it is acter using the PS Action



The action is described in this posting.
http://forums.dpreview.com/forums/read.asp?forum=1027&message=8544267

I can e-mail the action to anyone who is interested.

Kent
As another comment, it's very very dissapointing when my old 2MP
Bayer pattern Nikon Coolpix with no Apeture/Shutter priority on 8
second exposure completely wipes the floor with my new $2000 DSLR
when it comes to low light cityscapes.

e.g. this is from the old Nikon
http://www.pbase.com/image/28332449

This can't be right !!!

After reading the forum I thought this was not going to be a big
issue, appears I was wrong. In addition there was an observation of
a photo I posted a couple of days ago that the sky in it was
"blotchy". This was in full sunlight,
http://www.pbase.com/image/28456925

So .... I'm going to go for a walk around the block now and count
to ten
:-(

Cheers....

Shoei
So does this mean even though the Sigma comes with exposures up to
30 seconds and allegedly improved low light performance in the
SD10, you can't use them longer than about 2.5 seconds without
expecting a lot of really bad noise and above 5-10 secs it's
unusable !!!

Now I'm particularly unhappy. It also means Phils review of the low
light performance was spot on.

please someone tell me I am doing something wrong and it's not the
SD10 camera in general! Should I send it back to Sigma's
distributer in Australia to try to fix it or is it unfixable and I
should have gone for a different Camera. I really hope this can be
fixed (or at least alleviated).

in a word

AAaarrrgh !!!!

Cheers....

Shoei
a friend of me send his SD10 to sigma-germany - for killing the
noise-problem and set the "bulb"-funktion to the right funktion (so
long as you press)

now the cam ist back.......

the "bulb" is working with 15 secound (no change)
and the "green" noise is now a "blue" noise!!
a work of 3 weeks by sigma-service!

Ulli

(sorry about my english)
Hi All,

First of all thanks for the support, this group has been great.

Continuing my test of the SD10 in low light I went for a drive this
evening at dusk and shot a few images. After looking at the images
I've noticed a lot of Green noise. I thought this had been
corrected in the SD-10 so I thought I'd post and find out if it is
so.

Here are the first two "cityscape" shots. The first is as it came
out of the camera (no auto or custom SPP tweaks), the second (which
shows up the noise) was with a little "fill light" tweaking in SPP.

http://www.pbase.com/image/28520800
http://www.pbase.com/image/28520889

I was a little confused about this, and so I shot an F20 15sec
exposure of the front of my house. The results were absolutely
terrible (unusable). This is the result as it came from the Camera
(no auto or custom SPP tweaks).

http://www.pbase.com/image/28520980

Is this normal ?
Is anyone else seeing this
Is it something I'm doing wrong ?
have I picked up a lemon ?

Any assistance would be greatly appreciated

Regards,

Mark

--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--
Sorry, i will learn more english:-)
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--
Kent Dooley
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
 
What you're looking at is the physics of the image sensor itself.
This is not to say it's a flawed method; I look forward to the day
when all Bayer sensor cameras are in museums, but for the time
being, it's not good enough for what I do, which is a LOT of night
photography, with minimum exposure times of 20 sec, and now with my
new SLR, several minutes.
What are you using it for?....Astronomy?....Why on earth do you need such long exposure times for?
The Bayer sensor is very much like the CRT in this regard. Make a
up sandwich using dye-based inks, shove it in front of a sensor
that records monochrome only, and then go and fiddle the numbers
and see what we can come up with (hmmm sounds like a bunch of
corporate accountants). BUT.... it's a reasonably mature
technology. Sure, there are still gains to be made in terms of
resolution, sensitivity, and nose, for example, but the salient
point is that it produces damn good photos. And more importantly,
produces them NOW.
Just not as good as the Foveon X3 produces NOW,...unless you are prepared to pay comparatively huge sums of money and get a 1Ds .
PS: This is NOT an anti-Foveon post in any way whatsoever, merely
an observation from a semiconductor physics point of view.
Of course, as long as you realise that in good light only a 1Ds is the SD10's equal, so if your Bayer sensored camera is not a 1Ds then you would have been far better off going for an SD10.have an inferior camera.

PS: This is NOT an anti-Bayer post in any way whatsoever, merely
an observation from a semiconductor physics point of view.
 
1. when taking a night long exposure, get your initial exposure right. if the meter said 5 sec @ f5.6 don't push it to 8 sec. stay with 5s and f5.6 take a shot. then go f6.3 and take another shot. I wouldn't bother going to f5 and don't use AB with long exposure.

2. after you take the shot with MLU, tripod, and hopefully remote, you wait until it finishes processing. you already waited 10 sec, why not the extra 5 sec. (your tests proves this works)

3. try iso 400 or 800. don't believe the hype. they work. look in my gallery for examples if you want. plenty of night and high iso shots.

4. read this thread carefully: http://forums.dpreview.com/forums/read.asp?forum=1027&message=8091696

5. SD10 has its quirks. when you learn what they are, it's a wonderful camera to use and to take photos with. Remember, use the Camera, don't let it use you.

with that said Shoei, try your test with medium resolution. I remember when I took the fireworks shots I just snapped away and I was using MED because I forgot to bring the CF cards.

--
Chunsum
http://www.pbase.com/chunsum
http://www.pbase.com/sigmasd9
 
Hi Mark,

Thanks for trying.

Best,

Mike
Repeated the infamous "Shoei Test" of 5 x 15 second shots using the
AC Transformer instead of batteries.

again, the sequrence is
1 x shot 15 sec - wait until the activity LED stops after the shot.
3 x shots 15 sec - as quickly one after the other as possible -
after third shot wait until the activity LED stops
1 x shot 15 sec - wait until the activity LED stops after the shot.

and .....
...... the problem is 100% repeatable.

you get first good, second bad, third worse 4th disgusting 5th fine.

Darkframe processing interruption theory is holding

Batteries appear to not the problem. Good thought though, and worth
investigating.

Cheers....

Mark
Hi Shoei,

I have had quite a few images that look a lot like
http://www.pbase.com/image/28520980

but never more than 2 consecutive. Am I understanding that you are
able to repro this on demand?

If so might I suggest that you try it with the camera running from
ac current instead of batteries and see if that helps. I'm
beginning to suspect that low batteries may be contributing to this
noise and to the occasional corrupt images I get.

thanks,

Mike
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
 
What are you using it for?....Astronomy?....Why on earth do you
need such long exposure times for?
Astronomy? Not at all. Night photography doesn't have to be astronomy.
Just not as good as the Foveon X3 produces NOW,...unless you are
prepared to pay comparatively huge sums of money and get a 1Ds .
The Foveon, at 3MP, produces images with enough resolution to take on a 6MP Bayer sensor, no doubt about that. But it can't compete with the top of the crop such as 11 from Canon or 14 from Kodak. It's like having a 1 litre turbo take on a 5 litre V8 from the lights. It's simply out-muscled. Nothing wrong with that. As I said, the Foveon concept is a great one, but at this stage, it's got some way to go.
Of course, as long as you realise that in good light only a 1Ds is
the SD10's equal, so if your Bayer sensored camera is not a 1Ds
then you would have been far better off going for an SD10.have an
inferior camera.
Well, for what I do, the SD10 is inferior. Note: for what I do. Everybody's applications are different.
PS: This is NOT an anti-Bayer post in any way whatsoever, merely
an observation from a semiconductor physics point of view.
As they say, imitation is the sincerest form of flattery :)
 
Chunsum,

Thanks for your lessons!
I checked one of your works:

http://www.pbase.com/image/25434462

My question is, how did you meter the light to take this shot? From the far away buildings? The sky? The lakeshore? or the lamp nears you? And did you change to MF after getting the meter by using AF?

From the thread,
http://forums.dpreview.com/forums/read.asp?forum=1027&message=8091696

many experts took great nightshots without noise, but I took some with ISO100/200 last night and still had visible noise there.

Thank you for your answer!

-Izzy
1. when taking a night long exposure, get your initial exposure
right. if the meter said 5 sec @ f5.6 don't push it to 8 sec. stay
with 5s and f5.6 take a shot. then go f6.3 and take another shot. I
wouldn't bother going to f5 and don't use AB with long exposure.

2. after you take the shot with MLU, tripod, and hopefully remote,
you wait until it finishes processing. you already waited 10 sec,
why not the extra 5 sec. (your tests proves this works)

3. try iso 400 or 800. don't believe the hype. they work. look in
my gallery for examples if you want. plenty of night and high iso
shots.

4. read this thread carefully:
http://forums.dpreview.com/forums/read.asp?forum=1027&message=8091696

5. SD10 has its quirks. when you learn what they are, it's a
wonderful camera to use and to take photos with. Remember, use the
Camera, don't let it use you.

with that said Shoei, try your test with medium resolution. I
remember when I took the fireworks shots I just snapped away and I
was using MED because I forgot to bring the CF cards.

--
Chunsum
http://www.pbase.com/chunsum
http://www.pbase.com/sigmasd9
 
Chunsum,

Thanks for your lessons!
I checked one of your works:

http://www.pbase.com/image/25434462

My question is, how did you meter the light to take this shot? From
the far away buildings? The sky? The lakeshore? or the lamp nears
you? And did you change to MF after getting the meter by using AF?
because of all the light source, I used spot meter and meter for the lights on the building for the following reasons.

1.) My main concern was the building

2.) if I meter for the light right in front or the white lights along the shore, the buildings would have been under exposed.

3.) if I meter for the Sky, every thing else would have been over exposed. especially for the lights close by.

by metering for the light on the building I can get a good exposure on the skyline while only over expose the foreground slightly. this also give the sense of dept and layers in the composition.

I use AF and switch to MF to keep the focus. My eyes are only good enough to drive at night. you can continue to get meter readings even in MF mode.

you can then adjust in SPP.
From the thread,
http://forums.dpreview.com/forums/read.asp?forum=1027&message=8091696

many experts took great nightshots without noise, but I took some
with ISO100/200 last night and still had visible noise there.
When I first got the SD10, both Laurence Matson and Rick Decker told me this: alway try and get your exposure right. This is one of the biggest tip I received.

When you overexpose a night shot, which people tends to do( I know I did and still do it at times), you are going to get noise. many people think that you have to use Manual Mode to shoot a long exposure, so they tend to set the shutter a but too long thinking the longer it is, the better the shot gets exposed. you don't want to do that.

Use Manual only if you are sure you have a good reading. otherwise try AP and take the same shot twice. once with no exposure compensation, once +0.3. But I can say that msot of the time you'll end up using the 0 EC.

you want to get the exposure dead on or +- 0.3. granted with SPP you can get away with +- 0.7 or even +-1. but do try to get it as close as you can.

there are still some noise in the shot but they are minimal. you can't escape noise, you can only try to get it to a minimum.

Noise can also be produce in SPP. so becareful when using the sliders, especially exposure, sharpening, filllight, and saturation.

but the important is to not over expose it too much. the shot below is 13sec. the origial was a bit overexposed but salvaged in SPP. but you can still see the noice in the sky:



http://www.pbase.com/image/28097585

if you have to compensate exsposure in SPP for more than +- 0.6 then you're probably too far off. I use -highlight, -saturation, and -FL and keep exposure with in +-0.3.

SPP setting for the above shot:
Exposure: -0.1
Contrast: +0.4
Shadow: -0.1
Highlight: -1.6
Saturation: -0.7
Sharpness: +0.4
X3 Fill Light: -1.0
Color Adjustment: 0
Color Space: sRGB

here's a 8sec exposure and it's SPP Settings:



http://www.pbase.com/image/28097586

Exposure: +0.1
Contrast: +0.4
Shadow: -0.1
Highlight: -1.6
Saturation: -0.2
Sharpness: +0.4
X3 Fill Light: -1.0
Color Adjustment: 0
Color Space: sRGB

here's a pano made with 3 5sec shots:



http://www.pbase.com/image/28105323

now with this pano, you'll notice to the left there are noticable noise, that's becasue this shot was bit too over exposed. it's hard to meter a pano. here's the SPP setting I use for the 3 shots(they are the same):

Exposure: +0.2
Contrast: +0.4
Shadow: +0.0
Highlight: -1.5
Saturation: -0.4
Sharpness: +0.5
X3 Fill Light: -1.0
Color Adjustment: 0
Thank you for your answer!
I'm not very good at explaining what I do with these shots but I hope you get the idea. if not, I'll try again. remember, it's like shooting film and it's all about exposure. light doesn't change its properties for digital.

good luck.

--
Chunsum
http://www.pbase.com/chunsum
http://www.pbase.com/sigmasd9
 
Chunsum,

Thank you very very much for this useful instruction!

Izzy
Chunsum,

Thanks for your lessons!
I checked one of your works:

http://www.pbase.com/image/25434462

My question is, how did you meter the light to take this shot? From
the far away buildings? The sky? The lakeshore? or the lamp nears
you? And did you change to MF after getting the meter by using AF?
because of all the light source, I used spot meter and meter for
the lights on the building for the following reasons.

1.) My main concern was the building
2.) if I meter for the light right in front or the white lights
along the shore, the buildings would have been under exposed.
3.) if I meter for the Sky, every thing else would have been over
exposed. especially for the lights close by.

by metering for the light on the building I can get a good exposure
on the skyline while only over expose the foreground slightly. this
also give the sense of dept and layers in the composition.

I use AF and switch to MF to keep the focus. My eyes are only good
enough to drive at night. you can continue to get meter readings
even in MF mode.

you can then adjust in SPP.
From the thread,
http://forums.dpreview.com/forums/read.asp?forum=1027&message=8091696

many experts took great nightshots without noise, but I took some
with ISO100/200 last night and still had visible noise there.
When I first got the SD10, both Laurence Matson and Rick Decker
told me this: alway try and get your exposure right. This is one of
the biggest tip I received.

When you overexpose a night shot, which people tends to do( I know
I did and still do it at times), you are going to get noise. many
people think that you have to use Manual Mode to shoot a long
exposure, so they tend to set the shutter a but too long thinking
the longer it is, the better the shot gets exposed. you don't want
to do that.

Use Manual only if you are sure you have a good reading. otherwise
try AP and take the same shot twice. once with no exposure
compensation, once +0.3. But I can say that msot of the time
you'll end up using the 0 EC.

you want to get the exposure dead on or +- 0.3. granted with SPP
you can get away with +- 0.7 or even +-1. but do try to get it as
close as you can.

there are still some noise in the shot but they are minimal. you
can't escape noise, you can only try to get it to a minimum.

Noise can also be produce in SPP. so becareful when using the
sliders, especially exposure, sharpening, filllight, and saturation.

but the important is to not over expose it too much. the shot below
is 13sec. the origial was a bit overexposed but salvaged in SPP.
but you can still see the noice in the sky:



http://www.pbase.com/image/28097585

if you have to compensate exsposure in SPP for more than +- 0.6
then you're probably too far off. I use -highlight, -saturation,
and -FL and keep exposure with in +-0.3.

SPP setting for the above shot:
Exposure: -0.1
Contrast: +0.4
Shadow: -0.1
Highlight: -1.6
Saturation: -0.7
Sharpness: +0.4
X3 Fill Light: -1.0
Color Adjustment: 0
Color Space: sRGB

here's a 8sec exposure and it's SPP Settings:



http://www.pbase.com/image/28097586

Exposure: +0.1
Contrast: +0.4
Shadow: -0.1
Highlight: -1.6
Saturation: -0.2
Sharpness: +0.4
X3 Fill Light: -1.0
Color Adjustment: 0
Color Space: sRGB

here's a pano made with 3 5sec shots:



http://www.pbase.com/image/28105323

now with this pano, you'll notice to the left there are noticable
noise, that's becasue this shot was bit too over exposed. it's hard
to meter a pano. here's the SPP setting I use for the 3 shots(they
are the same):

Exposure: +0.2
Contrast: +0.4
Shadow: +0.0
Highlight: -1.5
Saturation: -0.4
Sharpness: +0.5
X3 Fill Light: -1.0
Color Adjustment: 0
Thank you for your answer!
I'm not very good at explaining what I do with these shots but I
hope you get the idea. if not, I'll try again. remember, it's like
shooting film and it's all about exposure. light doesn't change its
properties for digital.

good luck.

--
Chunsum
http://www.pbase.com/chunsum
http://www.pbase.com/sigmasd9
 
Much like Chromelight I, too, get nothing but a black image. However, we have SD9s. I don't think that should make a difference, though...
http://www.pbase.com/image/28540183

Regards,
Erik
I decided to do an "acid test" and do shutter times shots of the
back of the lens cap, reasoning being that this would be zero light
and hence all noise only.

These are the results

15 Seconds
http://www.pbase.com/image/28539215

10 Seconds
http://www.pbase.com/image/28539399

(Ran out of space on Pbase)

Anything above 5 seconds has a lot of Blue / Green noise in it,
almost like a pattern of hot pixels. at 3.2 seconds it's acceptable.

Could some people do the same so I can compare it with their SD10's
  • Set to S mode
  • Leave lens cap on
  • set focus to manual
  • Shoot at 15 sec down to 2 seconds (or wherever the noise ceases
to be a factor)
  • processs to jpg without SPP adjustments.
I really need to know if the Camera should go back for adjustment,
or just back.

Thanks in Advance

Cheers.....

Shoei
Here are the first two "cityscape" shots. The first is as it came
out of the camera (no auto or custom SPP tweaks), the second (which
shows up the noise) was with a little "fill light" tweaking in SPP.

http://www.pbase.com/image/28520800
http://www.pbase.com/image/28520889
When people talk about the Sigmas being noisy in poor light, I
think what they mean is that it gets noisy when the image is
underexposed, or when the shutter speed gets very slow, like more
than a few seconds. The SD10 does better than the SD9, but is
still not known for great low-light shooting.

Your sample above was shot at ISO 100, but "pushed" +1.4 stops in
SPP, and shadows brightened by +0.4 as well. So it's really more
like an ISO 400 shot, or two-stop underexposed for ISO 100. I
think that if you just set the shadow level back to zero, and let
the black areas be black, it will look a lot better.

Many settings can make noise show up more: Exposure push, shadow,
contrast, saturation, and especially Fill Light. If there's not
enough light in those dark areas, all you will see by pushing it is
noise.

For night shots, avoid Fill Light, or use it slightly negative to
clean up the black areas.
I was a little confused about this, and so I shot an F20 15sec
exposure of the front of my house. The results were absolutely
terrible (unusable). This is the result as it came from the Camera
(no auto or custom SPP tweaks).

http://www.pbase.com/image/28520980

Is this normal ?
Is anyone else seeing this
Is it something I'm doing wrong ?
have I picked up a lemon ?
That does NOT look normal. Sometimes the AWB messes up on long
exposures; try changing it to a fixed balance such as incandescent
and see if that helps. But probably something worse has gone bad
with that shot. Does it consistently do that at very long
exposures?

j
--
Enjoy Life ......
.... this is NOT a rehearsal !!!
--

All are lunatics, but he who can analyze his delusion is called a philosopher - Ambrose Bierce
 

Keyboard shortcuts

Back
Top