occasional magenta/green blotches

Started Feb 4, 2013 | Discussions
Uwe Steinmueller
Uwe Steinmueller Veteran Member • Posts: 3,414
occasional magenta/green blotches

In this thread we started a discussion about  the " occasional magenta/green blotches I see in random files.". I post the link here to reach a larger base for finding solutions (or not?).

http://thecameraforum.com/discussions/still-imagery-group7/cameras-lenses-lighting-other-photo-gear-forum5/sigma-dp2-merrill-field-report-thread10.0/#postid-347

Is there a sure way (besides B&W :-)) to reduce the effect?

-- hide signature --
SandyF Forum Pro • Posts: 14,941
Re: occasional magenta/green blotches

Uwe, I think the consensus of SD1 and  DPxMerrill users is that "magenta/green blotches" are basically due to underexposure. This was/is true of the older cameras too. I've seldom seen it, but I do recall some photos of mine where there is some blotching in underexposed areas. Example I'm thinking of is under roof edges/gutters where the wood is very, very dark + shadowy in photographs. I guess the solution is to hit the exposure right.... or perhaps use noise reduction... or meter for the shadows (?) if appropriate. I haven't tried that frankly.

Best regards, Sandy
http://www.pbase.com/sandyfleischman (archival)
http://www.flickr.com/photos/sandyfleischmann (current)

 SandyF's gear list:SandyF's gear list
Sigma DP2 Sigma DP1 Sigma DP1 Merrill Sigma DP2 Merrill Sigma DP3 Merrill +7 more
jrdigitalart
jrdigitalart Veteran Member • Posts: 3,899
As Sandy says, Uwe . . .
1

. . . under-exposure is the common demon and so, with users here declaring that the current Foveon sensor has the ability to recover 2 stops of blown highlights, exposure "to the right" seems to be one answer, at least.

-- hide signature --

Sincere regards, Jim Roelofs
Cherish your privacy? Avoid (sp)iPhones.
You are welcome to visit my portfolio here:
http://www.pbase.com/jrdigitalart

mike earussi Veteran Member • Posts: 8,418
other ways

Yeah, unfortunately there seems to be no real solution to this problem. All scenes with a high DR will experience underexposure and therefore will have those color blotches. In my experiments with the SD1 I tried to minimize them by pushing my exposure as much as possible just before blowing out the highlights and wound up shooting at an iso of between 64-80 (this will not work with the SD15 as it has much less recoverable highlights than the SD1). But this is real tricky as it's very easy to overexpose at this iso setting, so bracketing is suggested.

The other way to minimize this is to stack multiple shots (taken on a tripod) and average them out in PS (and perhaps other programs). This reduces noise and so helps the green blotches to cancel out the magenta ones--but it's not perfect, the blotches are merely smaller and therefore less noticeable.

A third way is to just darken the shadows until you no longer see them, as long as you don't mind blocked up shadows.

But there perhaps is some progress towards a solution as the darker/shadow in the DP3 shots seem to have much less blotches than the earlier cameras.

Uwe Steinmueller
OP Uwe Steinmueller Veteran Member • Posts: 3,414
Re: other ways
1

Thank you all.

I am not a big fan of OTR because it is easy to reach a point where you cannot recover.

-- hide signature --
MOD Kendall Helmstetter Gelner Forum Pro • Posts: 20,586
It can be safe

Uwe Steinmueller wrote:

Thank you all.

I am not a big fan of OTR because it is easy to reach a point where you cannot recover.

If you use ISO 200 you can easily shoot +0.7 exp and not lose highlights, since you can recover almost two stops.  So it's pretty safe to do so, using normal judgement as to when to increase or reduce that as a baseline.

If you don't need the DR you can back off to ISO 100 to improve noise.  Then you have to care more about overexposure.

With ISO 200 + overexposure you get more random noise (skies are slightly noisier for example), but fewer blotchy kinds of artifacts (especially because you are filling shadows a bit at time of exposure and not having to bring them up from base exposure as much).

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
Raist3d Forum Pro • Posts: 40,356
A couple of ideas...

- Use ISO 100

Use ISO 100 and expose to the right. You lose 1 stop of highlight recovery vs ISO 200 but you gain ~1 stop back in shadows (this is common for call cameras doing this underexpose, re-develop back up trick).

By using ISO 100 you have better control for fitting the dynamic range exposing to the right/letting the areas you want to "burn", burn. The result should be what will look like an under exposed shot but you will bring it "back up" in SPP.  This should provide more control than the automatic "bring backup" that ISO 200 does.

- Watch for temperature and length of use

While I don't know how much it affects at low ISO, saw a review where high ISO (6400) was severely affected by continuous use of the camera to the point that the shot looked like complete graphics trash.  But this was noted after continuous use, so it's temperature related.

Some of that may translate to the blotches at lower ISO, I would imagine if you were shooting say in Phoenix in the summer this could exacerbate the problem.

- Look for it the most in areas of neutral (grey)/shadow colors. You may have to desaturate manually a bit. SPP does this already to some extent as I have seen shots in shadow areas go monochromatic (actually a couple of the ones I did showed this).

The previous X3 sensor could do this but usually at higher ISO. Seems like the new X3 is able to keep more detail than the older as ISO goes up but the color constancy issue starts earlier (at even the lowest of ISO's).

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Raist3d Forum Pro • Posts: 40,356
ISO 200 won't help you here...
1

Kendall Helmstetter Gelner wrote:

Uwe Steinmueller wrote:

Thank you all.

I am not a big fan of OTR because it is easy to reach a point where you cannot recover.

If you use ISO 200 you can easily shoot +0.7 exp and not lose highlights, since you can recover almost two stops. So it's pretty safe to do so, using normal judgement as to when to increase or reduce that as a baseline.

ISO 200 won't help. It's ISO 100 under exposed and developed back up. You lose more control of the tonal range that way.  What If find it's best (applies to all cameras), is to find by spot metering at what point the data will become unrecoverable, then measure in the shot accordingly, and let the areas that you want to burn, burn.

I got pretty good at this with the Olympus e-420, 620, E-3 and Pentax Q.  You can control very well so that you avoid burns where you do not want them. Of course, if you have a really high contrast scene, you will still be losing shadows to noise when you bring them back up if you want them visible and not burn anything.

If you don't need the DR you can back off to ISO 100 to improve noise. Then you have to care more about overexposure.

The DR of ISO 100 and ISO 200 are the same. What changes is how it shifts, so you have more highlight range and less shadow range at ISO 200.  ISO 100 and ISO 200 are one of the same coin, it's just how the camera under exposes and develops back up to shift the total tonal range.

You will not gain more total DR by going to ISO 200, and you do lose more control over how to use the total DR (this again, applies to virtually all cameras that have this shift).  ISO 200 advantage is if you are in a hurry and you want to make sure you don't miss highlights and can't be methodical about watching for the highlights, then you can dial ISO 200 and not be as careful.

But then the shadows won't be as nice but that will depend on the shot.

With ISO 200 + overexposure you get more random noise (skies are slightly noisier for example), but fewer blotchy kinds of artifacts (especially because you are filling shadows a bit at time of exposure and not having to bring them up from base exposure as much).

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Raist3d Forum Pro • Posts: 40,356
Oh I find the suggestion of an IR filter...

and maybe even a UV filter interesting. The Foveon does not capture RGB, but its own tri-set of color wavelengths. It may be possible some of those filters may help.

Also I still have some hope Sigma may be able to do something in SPP and depending how the RAW pipeline is in-camera and programability, maybe in a firmware upgrade.

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

SigmaChrome Forum Pro • Posts: 11,859
Re: other ways

mike earussi wrote:

A third way is to just darken the shadows until you no longer see them, as long as you don't mind blocked up shadows.

There is a forth way also: Use the selective desaturation tool (in PSP at least). Or you can mask the shadow area and desaturate that way. The trick is to not overdo it.

-- hide signature --
 SigmaChrome's gear list:SigmaChrome's gear list
Sigma DP1 Sigma DP2 Merrill Sony RX100 Sigma DP3 Merrill Sigma dp0 Quattro +35 more
MOD Kendall Helmstetter Gelner Forum Pro • Posts: 20,586
Who to listen to, person with experience or other guy?

Raist3d wrote:

<...>

ISO 200 won't help. It's ISO 100 under exposed and developed back up.

I didn't even read the rest of what you said, since you are ignoring real facts backed up by my real world experience.

This very site reported that the SD-1 has the most DR at ISO 200.  My own experience has shown me that I can recover almost a full stop more highlight in ISO 200.  Common sense then says you can get nearly a stop better exposure in shadows simply by exposing to the right.  That is in turn backed up again, by real world experience and quite a lot of images.

Uwe, just try it instead of listening to any of the theory-only "professors" we have on staff.

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
Raist3d Forum Pro • Posts: 40,356
A person with experience?
1

Kendall Helmstetter Gelner wrote:

Raist3d wrote:

<...>

ISO 200 won't help. It's ISO 100 under exposed and developed back up.

I didn't even read the rest of what you said, since you are ignoring real facts backed up by my real world experience.

I have experience too.

This very site reported that the SD-1 has the most DR at ISO 200. My own experience has shown me that I can recover almost a full stop more highlight in ISO 200.

Indeed, you can recover more highlights at ISO 200 if you just straight-shoot and not expose to the right. I don't disagree with that. What you are missing is that total DR is not just highlight DR. Trying to expose to the right ISO 200 for shadows is really trying to bring it to an ISO 100.

What do you think over exposing ISO 200 means?

As for dpreview, you are talking about a website that scratched their head when they saw the Olympus E-30 "lost one stop of DR by doing ISO 100" and they didn't know why- and thought the DR was somehow better. It's not. It's the same DR.

Here's something you could try- whenever dcraw supports the DP2M files, shoot in manual exposure (complete manual) a shot at ISO 100 and at ISO 200. Convert in dcraw. You will see the exposures look the same.

Common sense then says you can get nearly a stop better exposure in shadows simply by exposing to the right.

Yes, you can get better exposure in the shadows ONLY if you control what is going on with the DR. If you dial ISO 200 and over expose by 0.7/1.0 EV, guess what happened to the highlight DR? You can't cheat total DR. Now, maybe you do want more highlight DR in an automatic way (as I pointed out), you can't meter as carefully and ISO 200 comes handy here.

That is in turn backed up again, by real world experience and quite a lot of images.

No, what you are doing is sacrificing highlight DR to get shadow DR. You are basically indirectly and with less control bringing ISO 200 back to an ISO 100. You should really consider thinking through a bit more what is going on and shoot accordingly to see what I mean.

Uwe, just try it instead of listening to any of the theory-only "professors" we have on staff.

I find your attempt at a personal attack childish and ridiculous here. Maybe you should try what I suggested and see it for yourself. You can get a nice primer here:

http://forums.dpreview.com/forums/thread/2590846

I encourage everyone reading this to NOT believe me and trying it out for themselves. But you can't use SPP because SPP will honor the ISO 100-> 200 relationship, just like you can't use the Olympus RAW converter in the example I mentioned of the e-620.

How is that for getting the experience and data, and being objective?

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

MOD Kendall Helmstetter Gelner Forum Pro • Posts: 20,586
Not correct

Raist3d wrote:

- Use ISO 100

Use ISO 100 and expose to the right. You lose 1 stop of highlight recovery vs ISO 200 but you gain ~1 stop back in shadows (this is common for call cameras doing this underexpose, re-develop back up trick).

This is not correct.  Why are you giving advice for a camera you don't even own?

I wouldn't even mind you giving completely wrong advice but at least admit up-front you don't own the camera first and the whole your your post is speculation.

It ignores the very real fact that at ISO 200 you have more DR.  It ignores the quality of the noise you get by overexposing and recovering.  It ignores how any of the Merrill cameras behave.

The rest of the advice is similarly more theory without any practical backing.  For instance the temperature thing is just not so (or at least the effect is so marginal as to be non-existant IN PRACTICE).

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
Raist3d Forum Pro • Posts: 40,356
Quite correct.
1

Kendall Helmstetter Gelner wrote:

Raist3d wrote:

- Use ISO 100

Use ISO 100 and expose to the right. You lose 1 stop of highlight recovery vs ISO 200 but you gain ~1 stop back in shadows (this is common for call cameras doing this underexpose, re-develop back up trick).

This is not correct. Why are you giving advice for a camera you don't even own?

Maybe because I believe I know what I am talking about? Why not? Do I need a license from you before I post here? I think not.

I wouldn't even mind you giving completely wrong advice but at least admit up-front you don't own the camera first and the whole your your post is speculation.

I don't need to own the camera, I can use one. Why do I need to own one?

It ignores the very real fact that at ISO 200 you have more DR. It ignores the quality of the noise you get by overexposing and recovering. It ignores how any of the Merrill cameras behave.

What do you think happens when you over expose ISO 200? This just tells me it doesn't seem you know what exposure on these cameras mean.

The rest of the advice is similarly more theory without any practical backing.

See my answer to the other reply.

For instance the temperature thing is just not so (or at least the effect is so marginal as to be non-existant IN PRACTICE).

You should argue that with one of the most professional of reviewers, that reviewed the camera, not me. That was his claim. Since apparently reviewers know their stuff no? I think he wrote an article comparing the DP2M, EM-5 and another camera recently. I think you know who I am talking about, look it up.

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

MOD Kendall Helmstetter Gelner Forum Pro • Posts: 20,586
Limited experience is not the same as that of an owner.

Raist3d wrote:

<...> What you are missing is that total DR is not just highlight DR. Trying to expose to the right ISO 200 for shadows is really trying to bring it to an ISO 100.

What you are missing is that TOTAL DR as measured by DPReview is higher at ISO 200, period.  So you can expose higher and get more exposure in the shadows (with less noise) than at ISO 100 while also having the same highlights.

Sorry but that's what actual shooting has ALSO shown to be true.  Your speculation that it acts some other way is not borne out at all in real shooting.  I wouldn't have switched to shooting ISO 200 almost all the time unless I saw a real advanatage to it.

I'm all done, I refuse to have an argument where you get to make up facts while I keep giving fact and experience based evidence.  Respond as you will, my point is made and people with cameras can simply try both ways and see what actually works.

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
Raist3d Forum Pro • Posts: 40,356
Re: Limited experience is not the same as that of an owner.

Kendall Helmstetter Gelner wrote:

Raist3d wrote:

<...> What you are missing is that total DR is not just highlight DR. Trying to expose to the right ISO 200 for shadows is really trying to bring it to an ISO 100.

What you are missing is that TOTAL DR as measured by DPReview is higher at ISO 200, period.

Total DR as measured by dpreview how? What? Stop quoting me dpreview, I already pointed and gave a good example they are not the end all be all.

So you can expose higher and get more exposure in the shadows (with less noise) than at ISO 100 while also having the same highlights.

Again, what do you think over exposing ISO 200 does? (hint: over exposed ISO 200 and bringing back down is like shooting ISO 100).

Sorry but that's what actual shooting has ALSO shown to be true. Your speculation that it acts some other way is not borne out at all in real shooting.

Sorry but I don't believe I am speculating. Imho of course.

I wouldn't have switched to shooting ISO 200 almost all the time unless I saw a real advanatage to it.

I'm all done, I refuse to have an argument where you get to make up facts while I keep giving fact and experience based evidence.

Of course, cover your ears, and don't run the experiment I suggested. It's the best way to learn.

Respond as you will, my point is made and people with cameras can simply try both ways and see what actually works.

Absolutely. I am a believer in that.

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Raist3d Forum Pro • Posts: 40,356
Funny, dpreview is saying what I am saying
1

Just in a different way:

" In contrast, at ISO 200 and 2 stops overexposure, the results are essentially identical to overexposing ISO 100 by one stop. This means that you'll get maximum latitude for overexposure at ISO 200; we'd estimate you can expect to recover at least 1.3 stops above the JPEG clipping point with full colour accuracy. This in turn means that in bright, contrasty conditions we'd recommend using ISO 200 by default."

They aren't saying you get more DR. In fact, they say pretty much what I am saying- you get more highlight range at ISO 200 (note there is no mention of shadows or total DR).

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Lin Evans
Lin Evans Forum Pro • Posts: 17,679
No....
1

Raist3d wrote:

Raist3d wrote:

- Use ISO 100

Use ISO 100 and expose to the right. You lose 1 stop of highlight recovery vs ISO 200 but you gain ~1 stop back in shadows (this is common for call cameras doing this underexpose, re-develop back up trick).

snip

The above is simply not correct. You do not gain anything in shadows by using ISO 100. The native ISO for the DPM's and SD1 is ISO200 where you get the most dynamic range, period.

Lin

-- hide signature --

learntomakeslidshows.net

Raist3d Forum Pro • Posts: 40,356
Re: No....
1

Lin Evans wrote:

Raist3d wrote:

Raist3d wrote:

- Use ISO 100

Use ISO 100 and expose to the right. You lose 1 stop of highlight recovery vs ISO 200 but you gain ~1 stop back in shadows (this is common for call cameras doing this underexpose, re-develop back up trick).

snip

The above is simply not correct. You do not gain anything in shadows by using ISO 100. The native ISO for the DPM's and SD1 is ISO200 where you get the most dynamic range, period.

ISO 100 does have less noise than ISO 200 no? What do you think that means?

The so called "native iso" of the e-620 is also 200. Turns out by dcraw examination ISO 100 = ISO 200.  All the cameras doing a DR shift are doing this trick.

I say run the Sigma RAWS at ISO 100/200 with a manual exposure that is identical in dcraw and see the results. Also consider: what over exposing ISO 200 to get better shadows really means (hint: it shifts it to an ISO 100).

Lin

-- hide signature --

learntomakeslidshows.net

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Lin Evans
Lin Evans Forum Pro • Posts: 17,679
Re: No....

Raist3d wrote:

Lin Evans wrote:

Raist3d wrote:

Raist3d wrote:

- Use ISO 100

Use ISO 100 and expose to the right. You lose 1 stop of highlight recovery vs ISO 200 but you gain ~1 stop back in shadows (this is common for call cameras doing this underexpose, re-develop back up trick).

snip

The above is simply not correct. You do not gain anything in shadows by using ISO 100. The native ISO for the DPM's and SD1 is ISO200 where you get the most dynamic range, period.

ISO 100 does have less noise than ISO 200 no? What do you think that means?

Lin

ISO 50 has less noise than ISO 100, what do "you" think that means? The dynamic range of the sensor has been measured and is best at ISO 200, period. All your arguments to the contrary go against the real world experience of numerous photographers who use the camera daily and professional review sites who "measure" the dynamic range. So who do we believe? Those who have real experience or the arm chair expert who takes photography lessons?

Lin

-- hide signature --

learntomakeslidshows.net

Keyboard shortcuts:
FForum MMy threads