leica sl2 Leica SL2 dynamic range

[Chaemono]

Advertisement (gone after registration)

vor 25 Minuten schrieb helged:

What about AF for the S1+SL 90-280 combo? Anybody out there....? 

It’s blazingly fast. 

[Chaemono]

What could be observed so far, is that in high contrast scenes that are sufficiently lighted one can easily underexpose by two stops at base ISO in order to protect highlights and push the files later.  What could also be observed is that in low-light scenes the same approach may not work because underexposing may lead to shadow clipping even at base ISO.  Fortunately, given how malleable the SL2 files are, one cannot only push shadows to the limit but also recover highlight details in what appear to be blown areas.  

DNGs are provided and, as a bonus, an S1R RAW file, too.  It's interesting that the S1R requires slightly slower shutter speed to capture an equal amount of light as the SL2.  It must be the extra glass on the sensor, or the different micro-lenses, or both.

Edit - less compressed JPEGs here: https://www.smugmug.com/gallery/n-tzdLng/

Let's take a look at one example and ask the question previously posed in a high contrast but sufficiently lighted scene.  Is it better to shoot like this?  Link to download the DNG here: https://cc2032.zenfolio.com/img/g473419243-o750076470.dat?dl=2&tk=MUJAZBLV2x9C0sEDsifgehZunC4lPdBIHweuNHI6tH8=

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

ISO 100 f/2 @1/320 sec.

Or like this?  Link to download the DNG here: https://cc2032.zenfolio.com/img/g345382036-o750076470.dat?dl=2&tk=7JOSpzlh07ItisN8yhrCvqLDp2TNqnZAFsGPl3bxOIU=

ISO 100 f/2 @1/50 sec.

And here the link to the S1R Raw file which was at 1/40 sec. to match more or less the SL2 exposure in the picture above: https://cc2032.zenfolio.com/img/g259271485-o750076470.dat?dl=2&tk=kkU59yw7fPrURDaHU1lZrcnB8RLyyL-VgWXYKrsqM9I=

Edited November 21, 2019 by Chaemono

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Chaemono]

And the answer is, in low-light high contrast scenes one should expose as far to the right as is allowed (just before the point of no return), even at base ISO, in order to capture as much shadow details as possible at the time the shot is taken.  It has been stated in another thread, that the SL2 struggles a bit with shadow details in low-light scenes that are underexposed too severely.  Recovering those details in post may lead to false colors, or a blue/green tint in the OOF area as shadows are pushed.  In addition, if one exposes in a low-light scene as far to the right as possible, meaning just before the point of no-recoverability of highlight detail, one won't need to apply any NR in post. 😁 And this is exactly the case here.

See the greenish tint on the left side of the OOF shadows in the first one?  See how clean the second picture looks and how much more balanced the tones are?  Keep in mind, WB settings are exactly the same.

Less compressed JPEGs here: https://www.smugmug.com/gallery/n-tzdLng/

Same as the first one in #62 with WB adjusted exactly the same as below to match, Exposure +2.6, Highlights -53, Shadows +100, Sharpening +40, zero NR

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

ISO 100 f/2 @1/320 sec.

Same as the second one in #62 with WB adjusted exactly the same as above to match, Exposure -0.12, Highlights -62, Shadows +100, Sharpening +40, zero NRISO 100 f/2 @1/50 sec.

Edited November 21, 2019 by Chaemono

[Mr.Q]

Since there was some discussion lately, could you try longer exposures (ie 10 seconds, 1 minute, 10 minutes) and try pushing the shadows?

[Chaemono]

vor 5 Stunden schrieb Mr.Q:

Since there was some discussion lately, could you try longer exposures (ie 10 seconds, 1 minute, 10 minutes) and try pushing the shadows?

Do you mean with LENR switched off in order to look for Chroma/Luma noise and hot pixels?

[Chaemono]

I posted this picture in the image thread a couple of days ago and didn't realize that I didn't make full use of the adjustment latitude it provides .  This is another version of it with Highlights -100 and Shadows +99!  I'm posting a link to the DNG, too.

Less compressed JPEGs here: https://www.smugmug.com/gallery/n-3STNss/

SL2 + 90 Summicron-SL DNG here: https://cc2032.zenfolio.com/img/g783264332-o750076470.dat?dl=2&tk=Fjn2L_XSfZI4c2DUCIuAcSOOjdmbz3O1mjg0TwWyl-o=

As shot

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

ISO 100 f/2.8 @1/320 sec.

Same as above with Exposure +0.9, Contrast +7, Highlights -100, Shadows +99, Whites +46, Blacks -20, Vibrance +15, Sharpening +60, NR -30

[sillbeers15]

Advertisement (gone after registration)

This was the DNG file captured by SL2 & exported into JPEG and what my eyes saw at the scene attached below.

 

After post LR (Temp:7112, Tint:-20, Exp:+1.80, Highlight:-100, Shadow: +91, Clarity: +39, Cropped frame):

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

untitled-1000280 by sillbeers15

[Donzo98]

10 minutes ago, sillbeers15 said:

This was the DNG file captured by SL2 & exported into JPEG and what my eyes saw at the scene attached below.

 

After post LR (Temp:7112, Tint:-20, Exp:+1.80, Highlight:-100, Shadow: +91, Clarity: +39, Cropped frame):

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

untitled-1000280 by sillbeers15

Very nice...

[Chaemono]

ISO 3200 pictures now, Z7 vs. SL2, for noise with +40 NR in LR for both.  No RAW files this time because if shadows are pushed to the max either one can be made to look bad. But one can check out the changes applied by running the JPEGs through an EXIF analyzer.  The key to keeping the SL2 shadows clean in high contrast scenes like these is to ETTR (expose to the right) as much as possible in order to minimize shot noise in the shadows.  Here, the SL2 picture is @1/1000 sec., the Z7 picture at 1/1250 sec.  It required pulling the highlights back by 90 in LR for the SL2 vs. -70 for the Z7.  I'll post crops later.

Less compressed JPEGs here: https://www.smugmug.com/gallery/n-tzdLng/

Z7 not processed except LR defaults and NR +40

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

ISO 3200 f/2 @1/1250 sec.

SL2 not processed except LR default, NR +40 and colors, WB adjusted a bit.

ISO 3200 f/2 @1/1000 sec.

[Chaemono]

What's the fuzz all about the ISO 3200 noise?

Less compressed JPEGs here: https://www.smugmug.com/gallery/n-tzdLng/

Z7 same as in #69 processed

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

SL2 same as in #69 processed

[Exodies]

Shot noise goes down as exposure time goes up. Is there a sensor independent graph of this?

[Arrow]

On 11/22/2019 at 1:58 AM, Mr.Q said:

Since there was some discussion lately, could you try longer exposures (ie 10 seconds, 1 minute, 10 minutes) and try pushing the shadows?

Here is a pretty much cropped in random detail. 60 seconds, Iso 50, in camera noise reduction turned off, pushed 5 stops in Lightroom. Hot pixels all over the place.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited January 8, 2020 by Arrow

[Arrow]

Here is the non pushed original ...

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[thighslapper]

44 minutes ago, Arrow said:

Here is a pretty much cropped in random detail. 60 seconds, Iso 50, in camera noise reduction turned off, pushed 5 stops in Lightroom. Hot pixels all over the place.

 

That's appalling.

I have put my SL2 for sale on eBay tonight. 

As we live in a coal mine, wear black clothes and have a black cat and dog this camera is going to be of no use to me whatsoever. 

[scott kirkpatrick]

8 hours ago, Exodies said:

Shot noise goes down as exposure time goes up. Is there a sensor independent graph of this?

Shot noise is just Poisson statistics.  If your pixel intensity in a region is the result of N photons reaching each sensor cell, then the signal in the pixels varies over a range of roughly sqrt(N).  So if a highlight is the result of capturing 64,000 photons, the variation is of order 256 photons.  That is noise that is 8 bits down below the average, invisible.  On the other hand, if your signal in the shadow is the response to only 4 photons, the cell to cell variation could be 2 photons, easy to see.  But that's photons.  What we read out is voltage, essentially the number of electrons generated, and that is affected by how strongly the signals are amplified, which is sometimes changed as we change ISO.

[Arrow]

8 hours ago, thighslapper said:

That's appalling.

I have put my SL2 for sale on eBay tonight. 

As we live in a coal mine, wear black clothes and have a black cat and dog this camera is going to be of no use to me whatsoever. 

Good Lord! Will not some happy little colour spots brighten up the tristesse? 😇

Edited January 9, 2020 by Arrow

[Arrow]

7 hours ago, scott kirkpatrick said:

Shot noise is just Poisson statistics.  If your pixel intensity in a region is the result of N photons reaching each sensor cell, then the signal in the pixels varies over a range of roughly sqrt(N).  So if a highlight is the result of capturing 64,000 photons, the variation is of order 256 photons.  That is noise that is 8 bits down below the average, invisible.  On the other hand, if your signal in the shadow is the response to only 4 photons, the cell to cell variation could be 2 photons, easy to see.  But that's photons.  What we read out is voltage, essentially the number of electrons generated, and that is affected by how strongly the signals are amplified, which is sometimes changed as we change ISO.

When constructing a light trap, it must be like trying to catch Srödinger´s cat with it. Just kidding 🐈 

[Arrow]

By the way, the hot pixels entirely disappear when shifting the single pixel slider in C1 at a value of 39. Hope Phase One will soon come up with the long awaited SL2 profile.