SL2 vs. S1R difference in IQ (image thread), or is the S1R a good back-up for the SL2 IQ-wise?

[Chaemono]

Advertisement (gone after registration)

I just handled the SL2 in the store this morning.  It looks and feels nice in a way that makes one want to grab it and go out and shoot it.  But the question is, do its files make one want to grab it and go out and shoot it.  I took, therefore, the S1R along to compare the files.  As a reminder, the $64,000 question we are after, to paraphrase from the DPR review of the SL four years ago here: https://www.dpreview.com/reviews/leica-sl-typ601-in-depth-camera-review/3, is 'put another way, is the SL2 ISO invariant?: does it have so little downstream read noise that one can shoot at base ISO and push the results later in order to give oneself extra highlight headroom?'  

Some comments on sharpness in the center only first.  I didn't see much difference and actually thought that the S1R produced crispier looking subjects, in the center, than the SL2 with the 50 Summicron-SL at ISO 100.  And even noise might be less with the S1R at ISO 100 as one adjusts brightness.  This was based on some quick shots and more thorough testing is required.  Still, my thoughts were as I was looking at some pictures that Panasonic did a great job with noise reduction and sharpening of the ISO 100 RAW files, and who cares that they are cooked.  

But this is where the good news ends for the 47 MPx sensor implementation in the S1R.  Assuming the SL2 uses the same sensor, it must be the way it's implemented because the files of the SL2 and the S1R behave quite differently.  First, I like the AWB of the SL2 a lot, not so much of the S1R.  This is software related, though, and anyone on this forum who can't use a gray card for custom WB should just be shooting with a cellphone (I, for one, can't, and I prefer AWB by the camera to look nice enough as a starting point in order to allow me to adjust it easily in LR).  Jaap explained in another thread what 'different implementation of the sensor' implies: "['sensor' IQ is affected by]..the processing pipeline: ADC, demosaicing, interpolation, etc. If the processor is working at the limit of its capablities, any power fluctuation may introduce artefacts. The SL2 should do much better with the new CPU."  And this is what could be happening.  

Let's look at the first example from the store this morning.  This is not a high contrast scene, but it serves to illustrate the difference in files.

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

S1R + 50 Summicron-SL from RAW, no adjustments except LR defaults, link to RW2 file here: https://cc2032.zenfolio.com/img/g641383526-o750076470.dat?dl=2&tk=ZDRRTS1EjBKOwtL9zzVh5YfwC_v-3RcTw_s2QDwP0Yw=

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 @1/500 sec.

SL2 + 50 Summicron-SL from RAW, no adjustments except LR defaults, link to DNG file here: https://cc2032.zenfolio.com/img/g554069433-o750076470.dat?dl=2&tk=deMbbRfwaa_Jnw3FyG7B0ZDO6mUhn6lf7WqwJzAWmOA=

ISO 100 f/2 @1/500 sec.

Edited November 7, 2019 by Chaemono

[Chaemono]

And now the brightened files with Shadows in LR pushed to +100.  Lots more artefacts/shadow banding with the S1R, green on the right side and blue on the left side, and the blue shadow banding is not a WB thing.  The SL2 picture is cooler with better WB.  Look at the white in the two Porsches.  The SL2 shows very little greenish cast on the right side but handles the tones in the shadows on the left side much better. 

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

S1R + 50 Summicron-SL with equal adjustments as the SL2 picture below to try to match

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 + 50 Summicron-SL WB adjusted to try to match.

Edited November 7, 2019 by Chaemono

[Chaemono]

This is my favorite trivia question.  Can one get a usable picture from this SL2 file?  Keep in mind, we are exposing to protect the highlights at base ISO.  

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

SL2 + 50 Summicron-SL from RAW, no adjustments except LR defaults, link to DNG file here: https://cc2032.zenfolio.com/img/g1066806095-o750076470.dat?dl=2&tk=qsWjvduEEoj6f0dY9hnbXppAaoN9QX4kONHGvij5Ddo=

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 @1/400 sec.

With adjustments except the Shadows slider at zero in LR

And now in addition with the Shadows slider at +100

Edited November 7, 2019 by Chaemono

[Donzo98]

How's the HIGH ISO performance??

[Chaemono]

vor 5 Minuten schrieb Donzo98:

How's the HIGH ISO performance??

I didn't check, Don, because the place wasn't dark enough.  If one bumps up the ISO and increases the shutter speed to 1/8000 sec., for example, in a well lighted place, one won't get the real high ISO performance as in a true dark setting. 

[bags27]

Thanks so much for doing this. Very impressive!

[Donzo98]

Advertisement (gone after registration)

51 minutes ago, Chaemono said:

I didn't check, Don, because the place wasn't dark enough.  If one bumps up the ISO and increases the shutter speed to 1/8000 sec., for example, in a well lighted place, one won't get the real high ISO performance as in a true dark setting. 

Yup..fo sure.

BTW... that AWB diff is very obvious.

 

Edited November 7, 2019 by Donzo98

[Tailwagger]

Yeah, the difference in WB is striking. To the point where I feel compelled to ask if both were set to auto?

[Chaemono]

vor 2 Stunden schrieb bags27:

Thanks so much for doing this. Very impressive!

Thanks. It will be interesting to compare the SL2 to the Z7 and the α7R III.  Both of these have significant PDAF banding when Shadows are lifted to +100 and Exposure to +5.  Sony deals with it by cooking the RAWs. But Nikon doesn’t. 😁 See how bad the pushed shadows banding of the Z7 is here: https://www.dpreview.com/forums/post/61811698.  This is probably why Leica is sticking to CDAF. 

Edit - I stand corrected, the α7R III has PDAF striping which, according to Jim Kasson, is an entirely different phenomenon. 🤔

BTW, this Jim Kasson guy does really thorough testing.  See his work on the α9 II’s ISO behavior, read noise, and ‘star-eating’ 😂 here if you’re interested: http://blog.kasson.com/
 

vor 21 Minuten schrieb Tailwagger:

Yeah, the difference in WB is striking. To the point where I feel compelled to ask if both were set to auto?

Yeees, of course.  Check the EXIF data.  It’s the best AWB I’ve seen, ever! 😁

Edited November 7, 2019 by Chaemono
I stand corrected, the α7R III has PDAF striping, not banding.

[Tailwagger]

12 minutes ago, Chaemono said:

Yeees, of course.  Check the EXIF data.  It’s the best AWB I’ve seen, ever! 😁

I figured so, just wanted it on record! 

[Donzo98]

1 hour ago, Chaemono said:

Thanks. It will be interesting to compare the SL2 to the Z7 and the α7R III.  Both of these have significant PDAF banding when Shadows are lifted to +100 and Exposure to +5.  Sony deals with it by cooking the RAWs. But Nikon doesn’t. 😁 See how bad the pushed shadows banding of the Z7 is here: https://www.dpreview.com/forums/post/61811698.  This is probably why Leica is sticking to CDAF. 

Edit - I stand corrected, the α7R III has PDAF striping which, according to Jim Kasson, is an entirely different phenomenon. 🤔

BTW, this Jim Kasson guy does really thorough testing.  See his work on the α9 II’s ISO behavior, read noise, and ‘star-eating’ 😂 here if you’re interested: http://blog.kasson.com/
 

Yeees, of course.  Check the EXIF data.  It’s the best AWB I’ve seen, ever! 😁

You have an S1 to compare it to?

[photonc]

Thanks for sharing these! I downloaded the RAW's and used the white edge inside the book to adjust for WB. On very close inspection, I'm not seeing much difference between the two. Noise from the SL2 looks better in some areas and worse in others. Not sure if I can draw any definitive conclusions but thanks for sharing the pictures!

[Donzo98]

11 minutes ago, photonc said:

Thanks for sharing these! I downloaded the RAW's and used the white edge inside the book to adjust for WB. On very close inspection, I'm not seeing much difference between the two. Noise from the SL2 looks better in some areas and worse in others. Not sure if I can draw any definitive conclusions but thanks for sharing the pictures!

Josh... seems like the S1R holds up just fine.

[photonc]

1 minute ago, Donzo98 said:

Josh... seems like the S1R holds up just fine.

Agreed... I think we will see the bigger differences between the two when the M-glass comparisons become available.  

[Mr.Q]

Yeah I don't see much difference, as expected. 

[sebben]

Those DNG files are huge...

[RoySmith]

13 minutes ago, sebben said:

Those DNG files are huge...

I just tried using Adobe Lightroom's "Lossy Compression" on a Q2 DNG file - it went from 90.7 MB to 24.2 MB !

Reports are that there is no loss of image quality with this. Sean Reid (Reid Reviews) uses Adobe Digital Negative Converter before importing DNG files into Lightroom. One can also accomplish this after importing by using the Library>Convert Photo To DNG  command.

[sebben]

If you do this you can’t use the enhanced details debayer feature in lightroom

Edited November 8, 2019 by sebben

[Chaemono]

Am 7.11.2019 um 22:13 schrieb Donzo98:

You have an S1 to compare it to?

Yes, I do but no SL2, yet. Some may trickle in before the 21st.  Let’s see. 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[thighslapper]

On 11/8/2019 at 2:12 PM, sebben said:

If you do this you can’t use the enhanced details debayer feature in lightroom

It does work with S1R RW2 files converted to DNG using Adobe's DNG converter. 

Having said that the feature is mainly to reduce artefacts when printing big ...... and with 47mpx files that means VERY BIG.