Sony 256GB XQD Card Tests Good Twice with diglloydTools DiskTester, Puts Suspicion of Data Loss Bug Squarely on Panasonic S1R

About two weeks ago in Panasonic S1R: Data Loss with Sony XQD Card, I reported card errors but I wasn’t sure of the cause—camera or card. Since then I had card write errors 2 or 3 more times. I give credit to the Panasonic S1R for reporting the errors instead of silently failing. The card has been formatted many times in the camera, but the problem recurs.

I should have done this sooner, but I today tested the card in an XQD card reader using the fill-volume command of my software diglloydTools DiskTester (see also IntegrityChecker) The fill-volume command (which I use frequently for testing hard drives and SSDs) writes to 99% of the card capacity and then reads it back and verifies it byte-for-byte.

The tests ran to completion, and showed no errors. While not an absolute proof (could be a compatibility issue?), the rock-solid reliability of this exact-same card over the past year in the Nikon D850 and Nikon Z7 implicates the Panasonic S1R as the root cause.

Reader James K (NYC pro photog) writes:

This kind of problem is to be expected on a new top to the line system. Probably caused by not enough company testing under real world conditions. Not only could this affect you reputation but also your pocket book. If models are involved on a commercial advertising shoot you might be responsible for model fees and production costs. A pro would have to have at least two and possibly three bodies to feel halfway secure on a big money shoot. It also means that there are probably other hidden problems that will take 6 months of continuous use to uncover. Your readers should thank you for finding these significant errors. “Abandon all hope ye who enter here."

Lloyd you are really something. Why don’t the high end camera companies hire you to test their products BEFORE they unleash them on consumers? No one tests equipment like you, no one.

DIGLLOYD: one camera and one card do not make a pattern, but it is worrisome.

Test results from fill-volume command of diglloydTools DiskTester

Speed using a USB port on the 2018 Mac mini is rocking fast! Zero errors.

llcMini:DIGLLOYD lloyd$ disktester fill-volume SonyXQD
DiskTester 2.2.14 64-bit, diglloydTools 2.2.15, 2017-12-24 18:12
Copyright 2006-2017 DIGLLOYD INC. All Rights Reserved
Use of this software requires a license. See http://macperformanceguide.com/Software-License.html

OS X 10.14.4, 6 CPU cores @ 3000 MHz, 8192 MiB memory
Thursday, April 18, 2019 at 09:28:12 Pacific Daylight Time

Volume:               SonyXQD
Num files:            1000
Space to fill:        232.4 GiB
File size:            237 MiB
Transfer size:        131072 KiB
Fill with:            "0x0000000000000000"
Free space to remain: 1.40 GiB = 0.60%

Creating up to 1000 files of size 237 MiB on volume "SonyXQD"
Speed shown includes file system create/open/allocate/write-- real world time.
TARGET FOLDER: "/Volumes/SonyXQD/disktester-blobs"

Elapsed    File#        Qty   2sec   5sec  15sec   1min   2min   5min  15min  30min  60min 2s:15s  1m:5m ClockAvg
5s        4   1.05 GiB    226    219    219    219    219    219    219    219    219   1.03   1.00    212 
10s        9   2.21 GiB    230    230    225    225    225    225    225    225    225   1.02   1.00    221 
15s       14   3.37 GiB    232    231    227    227    227    227    227    227    227   1.02   1.00    224 
20s       19   4.52 GiB    229    230    230    228    228    228    228    228    228   0.99   1.00    226 
.........................
754s      994  230.2 GiB    317    315    316    314    313    315    312    312    312   1.00   1.01    312 

1000 files of size 237 MiB created in /Volumes/SonyXQD/disktester-blobs
Total data written: 231.4 GiB
Slowest write speed: 175 MiB/sec
Average write speed: 312 MiB/sec
Fastest write speed: 325 MiB/sec
Average write time per file: 758.50 milliseconds
.........................

Read 1000 files totaling 231.4 GiB in 681.08 seconds @ 348 MiB/sec
MiB/sec write/read for fill-volume on SonyXQD
  Mean: 312 MiB/sec	348 MiB/sec
StdDev: 	17.2 MiB/sec	2.90 MiB/sec

;

[low-res image for bot]

On the 2017 iMac 5K

Speed on a USB 3.1 port through a hub on the 2017 iMac 5K is severely impaired, but no errors.

# using SonyXQD card reader on 2017 iMac 5K, USB 3.1 gen 1
diglloyd-iMac:DIGLLOYD lloyd$ disktester fill-volume SonyXQD
DiskTester 2.2.15b 64-bit, diglloydTools 2.2.16 beta, 2018-01-24 07:42
Copyright 2006-2019 DIGLLOYD INC. All Rights Reserved
Use of this software requires a license. See http://macperformanceguide.com/Software-License.html

OS X 10.13.6, 8 CPU cores @ 4200 MHz, 65536 MiB memory
2019-04-17 at 22:24:31

Volume:               SonyXQD
Num files:            1000
Space to fill:        232.4 GiB
File size:            237 MiB
Transfer size:        131072 KiB
Fill with:            "0x0000000000000000"
Free space to remain: 1.40 GiB = 0.60%

Creating up to 1000 files of size 237 MiB on volume "SonyXQD"
Speed shown includes file system create/open/allocate/write-- real world time.
TARGET FOLDER: "/Volumes/SonyXQD/disktester-blobs"


Elapsed    File#        Qty   2sec   5sec  15sec   1min   2min   5min  15min  30min  60min 2s:15s  1m:5m ClockAvg
5s        0    128 MiB    nan    nan    nan    nan    nan    nan    nan    nan    nan    nan    nan   25.1 
13s        1    365 MiB    nan   25.7   25.5   25.5   25.5   25.5   25.5   25.5   25.5    nan   1.00   27.9 
20s        2    602 MiB    nan   35.9   31.4   29.5   29.5   29.5   29.5   29.5   29.5    nan   1.00   29.6 
........................
6734s      995  230.4 GiB    nan   35.6   35.6   35.6   35.6   35.6   32.6   34.1   34.8    nan   1.01   35.0 
6740s      996  230.6 GiB    nan   35.7   35.6   35.6   35.6   35.6   32.6   34.1   34.8    nan   1.01   35.0 
6747s      997  230.9 GiB    nan   35.7   35.6   35.6   35.6   35.6   32.6   34.1   34.8    nan   1.01   35.0 
6754s      998  231.1 GiB    nan   35.7   35.6   35.6   35.6   35.6   32.6   34.1   34.8    nan   1.01   35.0 
6760s      999  231.3 GiB    nan   35.7   35.6   35.6   35.6   35.6   32.6   34.1   34.8    nan   1.01   35.0 

1000 files of size 237 MiB created in /Volumes/SonyXQD/disktester-blobs
Total data written: 231.4 GiB
Slowest write speed: 21.3 MiB/sec
Average write speed: 35.0 MiB/sec
Fastest write speed: 37.2 MiB/sec
Average write time per file: 6.76 seconds

........................

Read 1000 files totaling 231.4 GiB in 6411.88 seconds of clock time @ 37.0 MiB/sec
MiB/sec write/read for fill-volume on SonyXQD

  Mean: 35.0 MiB/sec	37.0 MiB/sec
StdDev: 	1.89 MiB/sec	0.67 MiB/sec

File#	Write	Read
1	25.5	36.6
2	35.0	36.9
............ elided ............
997	35.5	37.0
998	35.6	37.0
999	35.6	37.0
1000	35.6	37.0

Wrote results summary to:
/Volumes/Boot/Users/lloyd/Documents/diglloydTools/2019-04-17_22.24.31.80671633-fill-volume-SonyXQD.txt
/Volumes/SonyXQD/2019-04-17_22.24.31.80671633-fill-volume-SonyXQD.txt

Command "fill-volume" executed in 13175.78 seconds on 2019-04-18 at 02:04:07

Panasonic S1R: Moving Water Image Defects with HighRes Mode2

See my L-mount mirrorless wishlist

The Single-Shot vs Multi-Shot High-Res Mode2 Motion Correction page looked at image quality using Motion Blur Processing = MODE2, showing excellent results, as did Multi-Shot High-Res MODE1 vs MODE2 with Moving Subject Matter.

This page shows that even using Mode2 for Panasonic S1R Multi-Shot High-Res mode, moving water can generate intolerable artifacts of at least two kinds: (1) patchy areas that show abrupt discontinuation from other areas, and (2) pixellation.

In diglloyd L-Mount Mirrorless:

Panasonic S1R: Moving Water Failures with HighRes Mode2

Includes images up to full camera resolution for HighRes vs single-shot, plus crops.

f4 @ 1.0 sec Multi-Shot HighRes, ISO 50; 2019-04-16 19:48:21 [Enhance Details, push 3.5 stops]
Panasonic S1R + Leica 50mm f/1.4 Summilux-SL ASPH

[low-res image for bot]

Panasonic S1R: a Way to get 2X more Magnification in Manual Focus

See my L-mount mirrorless wishlist

Just quick note, I’ve added a tedious but usable workaround for magnified Live View focusing that achieves 2X greater magnification.

Magnified Live View Magnification Precludes Reliable Manual Focus + Workaround

The workaround has serious drawbacks that make it worthwhile only in limited circumstances., but you do get 2X larger (4X in area) magnification for manual focusing.

The workaround is proof that Panasonic could easily fix the issue in a firmware update. And that someone in Panasonic engineering has poor judgment shipping the S1R with such an arbitrary magnification cutoff.

Panasonic S1R: Best Practices When Shooting in Multi-Shot High-Res Mode

See my L-mount mirrorless wishlist

This page discusses best shooting practices for obtaining the very best possible results from Panasonic S1R Multi-Shot High-Res mode. In diglloyd L-Mount Mirrorless:

Panasonic S1R: Best Shooting Practices for Multi-Shot High-Res Mode

See also:

Panasonic S1R: Workflow for Multi-Shot High-Res Mode Images in Adobe Camera Raw

Panasonic S1R Multi-Shot High-Res mode

Panasonic S1R: Multi-Shot High-Res MODE1 vs MODE2 with Moving Leaves and Water

See my L-mount mirrorless wishlist

Hey, if you’re reading this, subscribe now—you can’t get the stuff I do anywhere else. Help me keep doing it—not tomorrow or next month, but right now—thanks!

In Single-Shot vs Multi-Shot High-Res Mode2 Motion Correction, I showed that MODE2 of the Panasonic S1R Multi-Shot High-Res mode was the way to go for movement.

This page look at fast moving water and foliage moving around willy-nilly on a moderately windy day, the kind of thing I encounter all the time in field shooting and for which just 1/2 of one percent of the motion used here is the kiss of death for conventional pixel shift.

In diglloyd L-Mount Mirrorless:

Panasonic S1R: Multi-Shot High-Res MODE1 vs MODE2 with Moving Subject Matter

Includes HighRes image crops from Mode1 and Mode2 along with the corresponding auto-recorded standard-res crops.

Bad juju with Panasonic S1R Multi-Shot High-Res MODE1

f5.6 @ 1/15 sec Multi-Shot HighRes, ISO 50; 2019-04-13 17:44:29 [Enhance Details]
Panasonic S1R + Zeiss ZM 35mm f/1.4 Distagon

[low-res image for bot]

Testing on Panasonic S1R: Panasonic LUMIX S PRO 50mm f/1.4, Leica 50/1.4 Summilux-SL ASPH, Leica 16-35/3.5-4.5 Super-Vario-Elmar-SL ASPH

See my L-mount mirrorless wishlist and Leica M wishlist.

It is my goal to cover all the the Leica SL lenses on the Panasonic S1R as well as all of the Panasonic LUMIX S lenses too, and later the Sigma DG HSM Art lenses for L-Mount.

The following lenses should be in my hands by April 17:

• Leica 16-35mm f/3.5-4.5 Super-Vario-Elmar-SL ASPH
• Leica 50mm f/1.4 Summilux-SL ASPH
• Panasonic LUMIX S PRO 50mm f/1.4

Coverage will go into diglloyd L-Mount Mirrorless, currently included as part of diglloyd Mirrorless. I hope to also have the Leica 24-90mm f/2.8-4 Vario-Elmarit-SL ASPH OIS in time to compare it in the 24-35mm range against the 16-35. Subscribe to diglloyd Mirrorless....

     

Panasonic LUMIX S PRO 50mm f/1.4 and
Leica 50mm f/1.4 Summilux-SL ASPH and
Leica 16-35mm f/3.5-4.5 Super-Vario-Elmar-SL ASPH

Reader Comment: Panasonic S1R Multi-Shot High-Res Mode vs PhaseONE IQ4 150MP

See my L-mount mirrorless wishlist and Leica M wishlist.

Background: see Panasonic S1R Multi-Shot High-Res Mode: the Largest Advance in Image Quality in a Decade and It Works with Motion = OMG and Panasonic S1R Multi-Shot High-Res Mode: the Largest Advance in Image Quality in a Decade?.

For the best possible results with Panasonic S1R Multi-Shot High-Res Mode, see Panasonic S1R: Best Shooting Practices for Multi-Shot High-Res Mode.

Panasonic S1R HighRes mode

Reader Roy P writes about Panasonic S1R Multi-Shot High-Res mode:

Hi Lloyd,

I downloaded an RMA from B&H to return the Panasonic S1R and almost shipped it back on Friday, but my gut said I should play with it some more over the weekend, and especially check out the multi shot high res mode.

I’m glad I did. I took a bunch of test shots of the same scenes with both my PhaseONE IQ4 150MP + Schneider 80mm f/2.8 (~55mm equivalent) and the S1R with my Leica 50mm f/2 APO-Summicron-M ASPH, the only HQ lens I have that could work well with the S1R without introducing significant complications of its own. I tested both Mode 1 and 2 on the S1R. All my P1 shots were at f/11 (~f/6.8) and the S1R were at f/6.3.

While I’m not ready to declare medium format is dead, I think the high res mode in the S1R is for real and must be taken very seriously. For static scenes, it looks nothing less than spectacular, and at least on my iMac 5K 27” monitor, the images look every bit as good as my PhaseOne IQ140 images, but with even more pixels. The HR images from the S1R look slightly darker and more saturated than my P1 images, but it’s not a problem.

I have tried pushing the exposure up by 4 stops (the max that Capture One will allow), and maxing out the shadows, and nothing is breaking down: the S1R files appear to be holding up just as well as the P1 files. I don’t know if there might be a noticeable difference in print quality, but at least on my monitor, I see no issues at all.

Extremely precise focus is a problem with both cameras, so if you have an uneven surface with some features in the scene a little closer to the camera and some a little farther away, you can see the differences between the two cameras – different parts of the image are at the sharpest focus on the two images. From a distance of about 10 feet, a difference of ±2 inches proximity to camera seems enough to make a difference.

You almost need another merge that could combine the S1R and IQ4 images that could produce the perfect image! Of course, that would be impossible with two different lenses.

It looks like the S1R HR mode works best when you can contain all of the subject to within a fairly narrow DOF. It looks ideal for flat subjects, the flatter the better. In scenes with a lot of depth, the fall off from best focus is much more rapid in the HR composite image than either individual S1R images or P1 images. So if you have a lot of depth to a scene, maybe focus stacking HR images would be the ultimate! I am going to try this next.

So far, in Mode 2, I haven’t had the success you did, with features like leaves swaying with the wind even sharper in the composite image – the best result I’ve got is that there are no motion blurs, but without the sharpest focus either. The worst results I’ve seen are HR images with a bizarre patch that is badly OOF. [diglloyd: moving things will be blurred more or less depending on circumstances].

This maybe a premature guess, but I suspect the two best focal lengths to use with the HR mode maybe very wide (so movements are less perceptible, and the focal length buys you a great deal more DoF), and long (so the field is small enough that you can dodge movements, and the DoF is compressed to a relatively narrow range). A more normal focal length range (say 35mm to 70mm) is probably the most challenging to work with the HR mode.

[diglloyd: focal length is mainly about magnification, focal length applying mostly at close distance: a 70mm lens makes CoC blurs twice as large at the same aperture and distance to subject, but at twice the distance (which it must be for the same FoV), the CoC will be the same size as a 35mm lens at half the distance]

Net-net, I agree with your assessment, the HR mode has got to be considered as a breakthrough technology. Considering this is the very first generation of this technology, the results are already very impressive and highly usable, albeit with the proper selection of subject matter and careful technique (many opportunities to create new best practices using the HR mode!). It can only get better over time, with more compute power in cameras and other camera makers trying to one up Panasonic.

It makes it a bit messy for me personally, since I am already neck deep-committed to Sony and P1, and I expect I will have a hard time resisting a Sigma Foveon FF camera, and possibly a Leica M11 camera if it comes out with a high quality EVF. So the last thing I need is one more camera, and other than the HR mode, I have no use case for the S1R, and I don’t want to invest in L mount glass, if I can help it. So I need to figure out what I’m going to do with the S1R.

I can’t see how every other camera maker can afford to not provide it as a standard feature. All prior innovations from one camera maker have been copied by others in time (e.g., Live View, PDAF, AF points embedded in a sensor, focus peaking, IBIS, etc.) So unless Panasonic has some kind of a death grip IP ownership on this, I would expect the other camera makers to be working overtime to introduce their own versions in their next cameras. (Besides, the Japanese have a way of sorting out their IP issues). A Sony A7R IV with similar HR mode would make all my headaches go away! No speculations from any of the rumor sites yet. I suspect most of the photography world hasn’t quite woken up to the potential of the HR mode in the S1R.

DIGLLOYD: depth of field is critically important in HighRes mode. And yet Panasonic has made critical focus with manual focus impossible, or I should say it this way: I am not able to nail focus even a third of the time using magnified Live View. Never has any digital camera delivered a magnified Live View feature that is so destructive to need. So the #1 issue with the Panasonic S1R is its magnified Live View implementation—see Panasonic S1R: Magnified Live View is Fundamentally Flawed. Yet the flaw is fixable with a firmware update, so there is some hope that Panasonic will wake up.

See Panasonic S1R: Best Shooting Practices for Multi-Shot High-Res Mode.

Zeiss ZM 35mm f/1.4 Distagon Aperture Series: Moots MTB by Creek in Forest (Panasonic S1R)

See my L-mount mirrorless wishlist and Leica M wishlist.

Performance at wider apertures of Leica M lenses and Zeiss ZM lenses on all mirrorless cameras to date (~April 2019) is poor in outer zones. This series from f/1.4 through f/11 confirms that finding with one of the best M-mount lenses, with the key point being at which aperture the lens starts being nicely sharp.

In Walkway Through Grass at Night as well as the nighttime examples, the Zeiss ZM 35mm f/1.4 Distagon fails to impress in the outer zones, with f/4 the minimum aperture for performance to the edges. In other words, it is impaired substantially by the sensor cover glass at f/1.4, f/2, and f/2.8 and somewhat impaired at f/4. See MTF on Mirrorless Cameras for the huge loss of MTF caused by the sensor cover glass.

In diglloyd Leica M:

Zeiss ZM 35mm f/1.4 Distagon Aperture Series: Moots MTB by Creek in Forest (Panasonic S1R)

Includes images from f/1.4 through f/11 at up to full camera resolution including an frame at f/5.6 in Multi-Shot High-Res mode. Also includes intermediate 24 megapixel size for perspective vs Leica M10.

The Zeiss ZM 35/1.4 was mounted on the Panasonic S1R using the Novoflex Leica M Lens to Leica SL/T Camera Body Lens Adapter.

f5.6 @ 0.8 sec, ISO 50; 2019-04-13 18:24:06 [Enhance Details]
Panasonic S1R + Zeiss ZM 35mm f/1.4 Distagon

[low-res image for bot]

Panasonic S1R Multi-Shot High-Res Mode: the Largest Advance in Image Quality in a Decade and It Works with Motion = OMG

See my L-mount mirrorless wishlist

Update: while this example with moving water is impressive, moving water can be a problem.
...

Cameras with pixel shift are prone to severe and prominent artifacts like checkerboarding that are just about impossible to deal with, making pixel shift pretty much useless for field shooting, a finding from long experience and true of Pentax and Sony pixel shift (I have probably 50GB of ruined Pentax and Sony pixel shift files).

But the Panasonic S1R uses an 8-frame multi-shot high-res mode approach that performs in-camera merging of those 8 frames into single raw file with size equivalent to four frames. A smart merge, not a dumb-recording science fair project like Pentax and Sony.

Can the Panasonic S1R Multi-Shot High-Res mode be used without unwanted digital artifacts with subject motion? Here HighRes mode 2* is used and compared to a single-shot frame.

*Mode 1 = “Motion blur appears as afterimage in the picture.”
 Mode 2 = “An afterimage of motion blur is minimized.”

In diglloyd L-Mount Mirrorless:

Panasonic S1R: Single-Shot vs Multi-Shot High-Res Mode2 Motion Correction (Alpine Creek)

Includes HighRes and StdRes images along with crops.

Holy crap! For the first time besides an iPhone, I’m seeing a pro-grade field-usable computational photography feature. Huge kudos to Panasonic for unf**king what the other guys could not get right.

What a shame that Panasonic has screwed up magnified Live View manual focus. Someday a camera vendor will get it all right, maybe.

f5.6 @ 1.0 sec Multi-Shot HighRes, ISO 50; 2019-04-13 18:17:38 [Enhance Details]
Panasonic S1R + LENS_NA

[low-res image for bot] Zeiss ZM 35mm f/1.4 Distagon

Panasonic S1R: True ISO 50, not a Faux ISO — Kudos to Panasonic (UPDATED: "true ISO" is not a thing for RAW)

See my L-mount mirrorless wishlist

Update April 17: the question of ISO is quite confusing and even forbidden in the latest ISO specification. See update towards end with comments by Alex Tutubalin of RawDigger.

...

The Panasonic S1R has a real ISO 50—not a faux ISO "Lo" mode. That is established by the RawDigger analysis as well as the identical match to ISO 100 (except for lower noise).

In diglloyd L-Mount Mirrorless:

Panasonic S1R: ISO Series 50 to 51200 (Dolls)

The bonus is that the Panasonic S1R Multi-Shot High-Res mode has the noise of ISO 6.25 and that it is actually usable in the field even with motion and that image quality is improved.

Seriously? A REAL ISO 50? Dang!!!

I shot and provided raw files in 1/3 stop ISO from ISO 50 to 51200 to the RawDigger folks. The signal/noise analysis is shown below. ISO 50 has an SN of ~64 versus ~57 for ISO 100, showing that ISO 50 is notably superior to ISO 100 which is in turn much superior to ISO 200—which I showed directly in Poppy by Window.

Note the nice smooth curve even with these 1/3 stop ISO changes—true ISO at all settings.

Signal/noise ratio on target that had some dust—values would be somewhat higher if dust were removed and target defocused to eliminate image detail as source of “noise”.

Panasonic S1R signal/noise ratio from ISO 50 to ISO 51200, 1/3 stops

Slight brightness variation at +1/3 ISO values at higher ISO

While shooting, I noticed a slight fluctuation at ISO values 1/3 stop higher than whole ISO, starting at ISO 800 (e.g., the exposure bumps up at ISO 1000). This was consistent as shown, is visible right in the viewfinder, all this graph below does is verify what can be seen directly. The variation is also seen the the S/N graph above.

ISOs 1000, 2000, 4000, 8000 are really ~1100, 2200.... (assuming 800 is real 800), while in-camera metering meters exposure for formal '1000, 2000,...' So, this is results in slightly better exposure (minimal overexposure) at these ISOs and so better S/N ratio.

Panasonic S1R Green average vs ISO from ISO 50 to ISO 51200, 1/3 stops

What is “true ISO”

I asked Alex Tutubalin of Lib Raw (RawDigger) “what is true ISO?”:

To make things more complicated:

- ISO Sensitivity is defined in ISO 12232 standard

- Latest Edition of ISO 12232:2019, released this year, explicitly *forbids* ISO determination for RAW shots. It is defined only for JPEG (so shot + 'development').

So, to be completely correct, 'true ISO' means nothing for RAW.... according to standard.

Meanwhile, it is possible to use at least two measurement methods (out of three) described in standard:

- Saturation based method (because RAW saturates at some point /unless full-well limited at lowest ISO/
- Noise based method.

The third method (gray point) is not directly applicable to RAWs because RAW gray point is not fixed (and vendor may vary it, to get more/less room in highlights).

However (and not standard conformant), we may assume that RAW gray point should be at 3 stops from saturation point (or 3.5 stops, it does not matter, the main thing is always the same value) and use graypoint-based ISO estimation method for RAWs.

Of course, this is not ISO-standard conformant, because standard explicitly forbids to measure 'ISO sensitivity' for RAW files.

Also, it is possible to use some 'standard RAW developer' (like D76 or D23 in film days) and measure sensitivity for sensor + developer combo. The main thing here: avoid any hidden exposure compensation if these adjustment are different for different cameras (so, Adobe tools should not be used as standard developer, because Adobe adjusts graypoint to match in-camera JPEG)

I also asked why ISO 50 values go only to about ~8900:

My guess: camera sensor is 'full well' limited at ISO setting less than 100.

That means:

- sensor pixels are not get saturated (really: values are cut) at same value for all pixels, but results in bell-like histogram in highlights.

- because all pixels are different, this results in some highlight (output) non-linearity

- so Panasonic opted to cut all output values below lowest possible pixel saturation point, to avoid such non-linearity.

This will result in less 'highlights headroom', so ETTR exposure for ISO 50...80 should be corrected for this phenomenon. Apparently Panasonic opted to cut out non-linear part of pixel(s) response, to avoid bell-like histogram at very highlights (instead of usual 'hit the wall') This nonlinearity results in pink-colored highlights, so it should be cut either in camera or in RAW developer.

So what does all this mean for practical photography?

In terms of seeking peak image quality for RAW, the ISO value is irrelevant—all that matters in terms of exposure is not blowing highlights (except speculars) while also minimizing noise. In other words, preserving the range of detail by not obliterating it by blowing highlights or by obscuring detail by noise in dark areas.

So the only useful question to ask is whether ISO 50 is superior to ISO 100 for high dynamic range scenes. That ISO 50 is best is shown decisively in Poppy by Window: equally good highlights but visibly superior noise control at ISO 50, that is, superior dynamic range.

In essence, ISO 50 on auto exposure allows exposing (almost) a stop more, which inherently increases dynamic range by creating cleaner blacks. The trick is that there is no so-called “headroom” for error as with ISO 100—but all ISO 100 is doing is cutting the exposure in half and then scaling up to larger numbers. Assuming ISO 100 and ISO 50 as shot do not blow highlights, ISO 50 always wins because more exposure means lower noise.

f5.6 @ 1.0 sec Multi-Shot HighRes, ISO 50; 2019-04-11 19:19:52
Panasonic S1R + Zeiss Milvus 35mm f/1.4

[low-res image for bot]

Panasonic S1R Multi-Shot High-Res Mode: the Largest Advance in Image Quality in a Decade?

See my L-mount mirrorless wishlist

Hey, if you’re reading this, subscribe now—you can’t get the stuff I do anywhere else. Help me keep doing it—not tomorrow or next month, but right now—thanks!

I went out to do some motion assessment studies with the Panasonic S1R Multi-Shot High-Res mode, both Mode 1 and Mode 2. Finishing up, I took the shot below, which turned out to be so compelling that I’m showing it first.

Is Multi-Shot High-Res mode 2 a partial or total 'win' or something else? Is moderate subject motion an issue?

What are the real differences in the field at f/8 (where diffraction is a serious factor for HighRes mode), given that f/5.6 really is not enough depth of field for many real images, and f/11 mostly obliterates gains in HighRes mode*?

* Depth of field at 187 megapixels is two stops less on a per-pixel basis (e.g., in terms of resolving to the pixel level, f/11 is needed for 187MP pixels versus f/5.6 for 47 megapixels). And yet, f/11 obliterates most of the extra detail in Multi-Shot High-Res mode due to diffraction.

And the $50K judgment call: is this as good as a 100 megapixel medium format camera, in terms of detail, pixel quality, noise? For 1/10 the cost it certainly bears looking into.

This page looks at a detailed outdoor shot at f/8 which includes motion that would render the pixel shift mode of Pentax an Sony unusable by awful checkerboarding. But here, the Panasonic S1R Multi-Shot High-Res mode 2 is used. In diglloyd L-Mount Mirrorless:

Panasonic S1R: Standard vs Multi-Shot High-Res Mode (Moots by Creek)

Includes HighRes and StdRes images up to 125 megapixels along with numerous crops.

I’d say that Nikon and Canon and Sony have some catching up to do. Dang! Ever see noise effectively ZERO, undetectable to the eye? Multi-Shot High-Res mode is essentially ISO 6.25.

How does an Easton XC90 carbon fiber rim just decide to bulge out after sitting around for a few years at 5 PSI? Shame on Easton for a shitty defective product—I’m pissed off that the wheel is now unrideable—warranty long gone.

f8 @ 1.0 sec Multi-Shot HighRes, ISO 50; 2019-04-13 18:34:44
Panasonic S1R + Zeiss Milvus 35mm f/1.4

[low-res image for bot]

Zeiss ZM 35mm f/1.4 Distagon Examples: Night Shooting (Panasonic S1R)

See my L-mount mirrorless wishlist and Leica M wishlist.

I’ve long said that the Zeiss ZM 35mm f/1.4 Distagon is the best lens available for Leica M cameras, and I stand by that, so I thought I had better show its pedigree in spite of the degrading effects of ill-suited sensor cover glass.

Can a lens impaired by sensor cover glass still produce satisfying images? This page shows examples at various apertures in night shooting. The Zeiss ZM 35mm f/1.4 Distagon is such a gorgeous performer that even though it is technically impaired at wider apertures, its core rendering style is still rewarding.

In diglloyd Leica M:

Zeiss ZM 35mm f/1.4 Distagon Examples: Night Shooting (Panasonic S1R)

Includes images up to full camera resolution at various apertures, including 6000 pixel width for perspective versus the Leica M10/Leica M240.

The Zeiss ZM 35mm f/1.4 was mounted on the Panasonic S1R using the Novoflex Leica M Lens to Leica SL/T Camera Body Lens Adapter.

f2 @ 1.0 sec Multi-Shot HighRes, ISO 100; 2019-04-09 20:25:56 [Enhance Details]
Panasonic S1R + Zeiss ZM 35mm f/1.4 Distagon

[low-res image for bot]

Zeiss ZM 50mm f/2 Planar Aperture Series: Mosaic (Panasonic S1R)

See my L-mount mirrorless wishlist and Leica M wishlist.

This page evaluates performance of the Zeiss ZM 50mm f/2 Planar on the 47-megaixel Panasonic S1R from f/2 through f/8.

In diglloyd Leica M:

Zeiss ZM 50mm f/2 Planar Aperture Series: Mosaic (Panasonic S1R)

Includes images up to full camera resolution from f/2 through f/8, including 6000 pixel width for perspective versus the Leica M10/Leica M240.

f5.6 @ 2.5 sec, ISO 100; 2019-04-09 19:50:25 [Enhance Details]
Panasonic S1R + Zeiss ZM 50mm f/2 Planar

[low-res image for bot]

Leica 28mm f/2 Summicron-M ASPH Aperture Series: Mosaic (Panasonic S1R)

See my L-mount mirrorless wishlist and Leica M wishlist.

This page evaluates performance of the Leica 28mm f/2 Summicron-M ASPH on the 47-megaixel Panasonic S1R from f/2.8 through f/8.

In diglloyd Leica M:

Leica 28mm f/2 Summicron-M ASPH Aperture Series: Mosaic (Panasonic S1R)

Includes images up to full camera resolution from f/2.8 through f/8, including 6000 pixel width for perspective versus the Leica M10/Leica M240.

f5.6 @ 0.4 sec, ISO 100; 2019-04-09 19:35:03 [Enhance Details, LACA corrected]
Panasonic S1R + Leica 28mm f/2 Summicron-M ASPH

[low-res image for bot]

Leica 18mm f/3.8 Super-Elmar-M ASPH Aperture Series: Mosaic (Panasonic S1R)

See my L-mount mirrorless wishlist and Leica M wishlist.

With each new mirrorless camera system, hope springs eternal even as basic physical facts are ignored (sensor cover glass). No, micro lenss do not fix sensor cover glass issues. Leica M lens performance is always impaired by ray angle when the sensor cover glass of a mirrorless camera has differing (greater) thickness than the optical path designed for the lens. The negative effects increase with distance from optical center and can include sharpness losses and amplification of field curvature and astigmatism.

This page evaluates performance of the Leica 18mm f/3.8 Super-Elmar-M ASPH on the 47-megaixel Panasonic S1R from f/3.8 through f/11. We should expect poor performance at f/3.8, but what about f/8 and f/11?

In diglloyd Leica M:

Leica 18mm f/3.8 Super-Elmar-M ASPH Aperture Series: Mosaic (Panasonic S1R)

Includes images up to full camera resolution from f/3.8 through f/11, including 6000 pixel width for perspective versus the Leica M10/Leica M240. Also includes the f/11 frame with distortion correction.

f5.6 @ 1.6 sec IS=off, ISO 100; 2019-04-09 19:45:57
[LACA corrected, Enhance Details]
Panasonic S1R + Leica 18mm f/3.8 Super-Elmar-M ASPH

[low-res image for bot]

Panasonic S1R: Image Quality Gains at ISO 50, 100, 200 Plus ISO 50 in Multi-Shot High-Res Mode (Poppy by Window)

See my L-mount mirrorless wishlist.

When I shoot indoors or outdoors, I strive to achieve the best possible image quality. That means base ISO and an optimal ETTR exposure. But what of ISO 50 versus ISO 100—does it matter much? For that matter, is Multi-Shot High-Res mode worth the bother versus ISO 50? It would be easy to come to an erroneous conclusion for some average scene, but when it matters, it really matters. Namely in high dynamic range images, as here.

This page evaluates how image quality at ISO 50 compares to ISO 100 and ISO 200 as well as ISO 50 in Multi-Shot High-Res mode, showing compelling results that any Panasonic S1R user should be aware of.

In diglloyd L-Mount Mirrorless:

Panasonic S1R: Quality Gains with ISO 50 (Poppy by Window)

Includes images up to full camera resolution plus crops at ISO 50, 100, 200, plus ISO 50 HighRes mode.

Panasonic S1R readers should read this page— worth the price of admission methinks.

f5.6 @ 1.0 sec Multi-Shot HighRes, ISO 50; 2019-04-11 19:19:52
Panasonic S1R + Zeiss Milvus 35mm f/1.4

[low-res image for bot]

Panasonic S1R Noise / ISO Series from ISO 50 to ISO 51200

See my L-mount mirrorless wishlist

This page evaluates digital noise from ISO 50 through ISO 51200. Exposure was an optimal ETTR exposures which thus represents a best possible result.

In diglloyd L-Mount Mirrorless:

Panasonic S1R: ISO Series 50 to 51200 (Dolls)

Includes several series including two crops showing the performance in color as well as each individual color channel plus a Gray Gamma 2.2 conversion. The individual channels and gray gamma may be of particular interest to black and white shooters.

Noise vs ISO from ISO 50 to ISO 51200 for color and individual color channels

Novoflex Nikon F Lens to Leica SL/T Camera Body Lens Adapter Cannot Reliably Stop Down Zeiss Otus 55/1.4

I don’t yet know if this is true of my other Zeiss F-Mount lenses (Zeiss Milvus, Zeiss Otus) as I have as yet only used the Zeiss Otus 55mm f/1.4 APO-Distagon on the Panasonic S1R.

Issue: the Novoflex Nikon F Lens to Leica SL/T Camera Body Lens Adapter cannot reliably stop down the lens past f/8, and even f/8 might be iffy once in a while.

The coupling for the adapter is purely mechanical and since it is hard to believe that my Otus is at fault (zero issues for years on Nikon cameras), I blame the Novoflex adapter. But whether it is this particular sample of the Novoflex adapter or a general issue, I don’t know.

Lens Adapters for Panasonic SlR

Sigma DG HSM Art wish list • Deals on Sigma

$395 Leica M-Adapter L in Lens Accessories

$206 Novoflex Leica M Lens to Leica SL/T Camera Body Lens Adapter in Lens Accessories

$219 Novoflex Nikon F Lens to Leica SL/T Camera Body Lens Adapter in Lens Accessories

$249 Sigma MC-21 Mount Converter/Lens Adapter (Sigma EF-Mount Lenses… in Lens Accessories

Jason W writes:

If the Novoflex adapter doesn't work I'm inclined to throw out all ideas of purchasing a Panasonic S1R.

DIGLLOYD: the Nikon Z7 is a better platform with far superior lens adapter; I would go with the Z7 if the goal is Zeiss DSLR lenses.

Notes on the Panasonic S1R RAW RW2 File Format

First, I am not entirely happy with the file sizes from the Panasonic S1R; they are larger than need be as compressing with zip shows.

Every photographer keen on exposing optimally should be using RawDigger by LibRaw. Get RawDigger.

RawDigger will be supporting the Panasonic S1R file formats. Notes from the RawDigger folks on the Panasonic S1R raw file formats:

Panasonic S1R Multi-Shot High-Res mode and the (optional via settings) single-shot images associated with the multi-shot image: lossless, uncompressed 14-bit.

Panasonic S1R single shot mode: delta encoding, 14-bit base, 10-bit delta, same idea as in other Panasonic cameras, though older cameras are 12-bit base/8bit delta. Delta is scaled separately for each 3 pixels. Formally, this is lossy compression, but this will give precise results on smooth details, in worst case compression-related errors will be less than photon noise (and no Sony-style artifacts).

All crops are really same resolution, 8392x5620 visible area. The crop is performed via EXIF tags (instructing software to crop).

Lack of good compression in some formats aside, it looks like Panasonic has done well what Sony did and does not so well (see Sony raw file format and lossy compression).

Reader Question: eGPU for Photoshop and Lightroom?

Reader Darren K writes:

Long story short is I had a RAID failure and got me thinking about all of my old-ish gear (chain below if you’re inclined to read). Considering selling 2013 Mac Pro for a 2018 Mac mini.

Read your article about the eGPU issue. My friend Bob is a video guy, and doesn’t know about whether eGPU is available in Photoshop or Lightroom. Would be good to know if you find that out at some point, but bailing on the Mac Pro might be wise.

DIGLLOYD: using RAID-5 is plenty fast these days and the best way to avoid RAID issues as it offers fault tolerance. For most uses, the modest performance gains with RAID-0 striping (no fault tolerance) is not going to be worth the risk of drive failure (at least with hard drives).

For RAID, use SoftRAID in an OWC Thunderbay since failure of a hardware RAID can be a very painful experience when the enclosure fails—it might not even be available any more and with the drives for hardware RAID using proprietary formatting they are useless in any other enclosure without reformatting. Not so with SoftRAID—any enclosure will do. And as a daily losing proposition, hardware RAID has long had inferior performance compared to SoftRAID.

eGPU for Photoshop and Lightroom.

I think that some readers are missing important information at the top of this bloge.g., headlines from MacPerformanceGuide are shown for a week or so in smaller type.

An eGPU is Useless for Photoshop and Lightroom Unless...

Put another way: the eGPU is used for Photoshop and Lightroom if and only if that eGPU is driving the main screen. At least that is what Adobe online support says, an what I found when testing the 2018 Mac mini with macOS Mojave 10.14.2.

There might be one exception for an eGPU when not driving the main screen—the Adobe Camera Raw Enhance Details feature. Adobe states:

Enhance Details heavily uses the available GPU and can use an external GPU (eGPU) if available

Adobe states:

1. We tested with a MacBook with an Intel integrated GPU, both with and without an eGPU plugged in. Enhance Details ran much faster with the eGPU plugged in. We've not tested all eGPU scenarios, however. 2

2. At present, Enhance Details only uses a single GPU. The choice of GPU is actually determined by the operating system.

3. Enhance Details uses only one GPU at a time. In the built-in GPU + eGPU scenario, only one of these would be actively used to run Enhance Details (and the OS should be choosing the eGPU to do so).

See also: Adobe Camera Raw Enhance Details Feature: Integrated into my Workflow... How Much Does it Use the GPU?

Stephen G writes:

I'm puzzled by your finding that eGPU's aren't recognized by Photoshop CC and /or Lightroom. I have a newish 2018 Mac mini with 64GB of Apple memory and the 3.2 GHz Intel Core I7.

Below are details of the EGPU Thunderbolt 3 I use every day with these programs and the Radeon Card etc. I only use the one monitor on the Thunderbolt 3 Bus with HDMI. Both of these Adobe programs use the EGPU simultaneously and without fail. They are fast and reliable. Mac OS X 10.14.4

DIGLLOYD: the An eGPU is Useless for Photoshop and Lightroom Unless... post links to the Adobe support pages which confirm what I’ve stated: an eGPU is used for Photoshop and Lightroom if and only if that eGPU is driving the main display. Stephen G is driving the main display using the eGPU, hence it does get used by Photoshop/Lightroom.

At the time I tested the 2018 Mac mini, there were still bugs in macOS Mojave (since fixed, apparently) that made using the NEC PA271Q with the eGPU not work. With the Mac mini and an eGPU it is defensible that using the eGPU to drive the main display is the best choice. With a machine with a built-in display (iMac 5K, MacBook Pro), it gets iffy since the built-in display can only be driven by the built-in GPU.

Zeiss ZM 35mm f/1.4 Distagon Aperture Series: Walkway through Grass at Night (Panasonic S1R)

See my L-mount mirrorless wishlist

I have been tip-toeing up to the bad news that Leica M and Zeiss ZM lenses suck on the Panasonic S1R, notwithstanding the hallucinations surely being repeated yet again as with all prior mirrorless cameras. The special micro lenses on the S1R sensor are surely fodder for that—much ado about nothing in truth. Leica M land Zeiss ZM lens owners will have to hold out for an M11 with a 47MP sensor it seems, one tuned with sensor cover glass less destructive for image sharpness from ray angle.

That said, sharpness is not everything and there may be uses for lenses that have appealing rendering qualities, or cases where stopping well down is fine.

A goal of this series is to show that while impaired, the results can nonetheless be appealing for some types of images, and with stopping down the results can become quite strong, at least with lenses like the Zeiss ZM 35mm f/1.4 Distagon.

In diglloyd Leica M:

Zeiss ZM 35mm f/1.4 Distagon Aperture Series: Walkway through Grass at Night

Includes images up to full camera resolution from f/1.4 through f/5.6, including 24 megapixel size as relating to the 24MP resolution of the Leica M10/Leica M240.

f5.6 @ 30.0 sec IS=off, ISO 100; 2019-04-09 20:41:22
[LACA corrected, USM{8,50,0}, Enhance Details]
Panasonic S1R + Zeiss ZM 35mm f/1.4 Distagon

[low-res image for bot]

Panasonic S1R: Magnified Live View with its Gorgeous High-Res EVF = Total FAIL

See my L-mount mirrorless wishlist

When I see a 100% failure rate that destroys a night’s work, I get pissed off.

Which is what happened late last night when I reviewed focusing accuracy for my images from four f/1.4 and f/2 Leica M/Zeiss ZM lenses shot that evening. “Pretty close” is 100% useless for my work—and demonstrably inferior to what I obtained with the pathetic low-res Leica M240 EVF shot side by side with the S1R.

Panasonic S1R: Magnified Live View is Fundamentally Flawed

A firmware fix would address the issue, but who knows if that will ever happen.

Peter K writes:

So I read you have the same [bad] experience as I had with LEICA M-glass mounted on the new PANASONIC DC-S1R...something is strange with the ‘Gorgeous” High-Res EVF of the PANASONIC DC-S1R.

When I checked my files for focus accuracy I saw that a lot of my picture were not spot-on. And I am sure that Focus-Peaking told me that my subjects were in focus. Working with a LEICA M10 with a Visoflex I have nearly 100% perfect results [and I know focus is more critical with a higher megapixel camera!] Focus-Peaking with the LEICA M10 with a Visoflex works wonderful. Even better with the same lens mounted onto a LEICA SL! The viewfinder of the LEICA SL has a higher resolution than the Visoflex, but maybe this is the problem?

.Of course I do not know how Panasonic makes Focus-Peaking technical working. But something goes strange if I see the [different] results I have got with the Panasonic DC-S1R compared to the LEICA M10 [and LEICA SL].

DIGLLOYD: as I don’t use focus peaking (not quite good enough for critical work in my experience) I cannot say, but I can say that Panasonic has screwed the pooch for magnified Live View in general.

Panasonic LUMIX S 24-105mm f/4 Macro OIS Aperture Series @ 35mm: Mosaic (Panasonic S1R)

See my L-mount mirrorless wishlist

Whenever feasible, I like to cover the zoom range with some reference set of evaluations, because the use of a zoom lens usually involves compromises somewhere. The 99mm focal length fills out that picture for the Panasonic LUMIX S 24-105mm f/4 Macro OIS, completing the series.

This page looks at performance of the Panasonic 24-105mm f/4 Macro OIS at 99mm on a demanding planar (flat) target at medium distance. This particular subject lays bare all lens weaknesses, but particularly micro contrast across the frame.

In diglloyd L-Mount Mirrorless (included in subscription to diglloyd Mirrorless):

Panasonic 24-105mm f/4 Aperture Series @ 99mm: Mosaic

Includes images up to full camera resolution from f/4 through f/11, along with large crops showing the behaviors.

f5.6 @ 0.6 sec, ISO 100; 2019-04-04 19:36:53 [Enhance Details]
Panasonic S1R + Panasonic Lumix S 24-105mm f/4 Macro OIS @ 99mm

[low-res image for bot]

Panasonic LUMIX S 24-105mm f/4 Macro OIS Aperture Series @ 35mm: Mosaic (Panasonic S1R)

See my L-mount mirrorless wishlist

This page looks at performance of the Panasonic 24-105mm f/4 Macro OIS at 35mm on an extremely demanding planar (flat) target at medium distance. This particular subject lays bare all lens weaknesses, but particularly micro contrast across the frame.

In diglloyd L-Mount Mirrorless (included in subscription to diglloyd Mirrorless):

Panasonic 24-105mm f/4 Aperture Series @ 35mm: Mosaic

Includes images up to full camera resolution from f/4 through f/8, along with large crops showing the behaviors.

f5.6 @ 4.0 sec, ISO 100; 2019-04-04 19:49:52 [Enhance Details]
Panasonic S1R + Panasonic Lumix S 24-105mm f/4 Macro OIS @ 35mm

[low-res image for bot]

Adobe Camera Raw Enhance Details Feature: Integrated into my Workflow... How Much Does it Use the GPU?

Update: my testing shows that a slow integrated chipset GPU versus fast GPU makes a huge difference, about 8X slower on the 2018 Mac mini with its built-in Intel UHD Graphics 630 GPU versus the 2017 iMac 5K. This argues strongly against the 2018 Mac mini for photographers and argues in favor of getting a fast GPU, but it does not really clarify how much a "good/better/best" GPU choice on high-end machines impact performance.

...

The Adobe Camera Raw Enhance Details feature is now a core part of my image processing pipeline—always a win, never a loss seen yet.

What I am wondering and without a good answer yet is just how much a (claimed) faster GPU can benefit things. Shown below, CPU usage is modest and GPU usage is substantial (roughly 70%), but far from maximal. That implies a bottleneck such that a faster GPU would not be able to contribute as much s one might hope for.

My understanding is that the Adobe Camera Raw Enhance Details feature is GPU intensive (not all GPUs are supported), which might mean that upgrading the GPU is worthwhile. Or it might not. Until I can get a faster GPU and a slower one in the same type of machine, I cannot quantify the advantage or lack thereof. I suspect a small benefit (no more than 30%), but it quite possibly could be less than 10%.

Note that “GPU intensive” might not include the surrounding context (CPU and disk I/O), so gains in GPU speed do not necessarily translate into similar gains in time required to complete a task. Put another way, a GPU that is 30% faster might correspond to gains of 30%, or it might correspond to gains of 10%, in terms of the actual time saved.

See my discussion of recommended 2019 iMac 5K features in Apple 2019 iMac 5K: Two Hits with One Big Miss and also Reader Question: Radeon Pro Vega 48 GPU Upgrade for 2019 iMac 5K.

Continues below...

Shown below are GPU and CPU history usage on the 2017 iMac 5K 4.2 GHz with Radeon Pro 580 8192 MB while converting 97 Panasonic S1R raw RW2 files to enhanced versions.
CPU utilization averages about 1.5 CPU cores and GPU utilization averages roughly 70%. Why are CPU cores not utilized better (complicated to code for CPU cores and GPU also?) and why is the GPU not utilized at 100% ? Good I/O algorithms can eliminate I/O as a factor, and ideally both GPU and all CPU cores could be put to use as computational resources.

Adobe Camera Raw Enhance Details: CPU core usage and GPU Usage history processing 97 raw files
2017 iMac 5K 4.2 GHz with Radeon Pro 580 8192 MB

Below, the GPU and CPU utilization on the 2018 Mac mini. GPU utilization is 100%, excepting blips which are presumably inter-file blips to read and write data. The runtime per raw file on the 2018 Mac Mini with Intel UHD Graphics 630 GPU is approximately 8 times slower than on the 2017 iMac 5K.

Adobe Camera Raw Enhance Details: CPU core usage and GPU Usage history processing raw files
2018 Mac mini 3 GHz Intel Core i5 with Intel UHD Graphics 630

Robin D writes:

Lloyd you comments about GPU load are topical for me. The enclosed is a copy of my GPU load monitor which shows an approximate 96-100% load on my NVIDIA GeForce GTX i1070 - a highly rated GPU - for around 5 seconds on a 43MPix ARW file. Processing 1000 of such files is a significant bottleneck, especially as LR seems to bog down on batch processing. CPU load was low at around 14%. Friends told me I bought a GPU that was too powerful. Obviously not.

DIGLLOYD: the load makes sense to me given what Adobe has said about GPU usage for this feature, and the observations above.

But it does not necessarily follow that Robin’s choice of GPU is a better choice than another—to do that one has to compare actual performance to get the task done. For example using the 2019 iMac 5K, the +$450 AMD Radeon Pro Vega 48 versus the AMD Radeon Pro 580X. Is there a 5%, 10%, 30%, some other difference? I hope to have an opportunity to test and see.

Panasonic LUMIX S 24-105mm f/4 Macro OIS Aperture Series @ 24mm: Mosaic (Panasonic S1R)

See my L-mount mirrorless wishlist

This page looks at performance of the Panasonic 24-105mm f/4 Macro OIS at 24mm on an extremely demanding planar (flat) target at medium distance. This particular subject lays bare all lens weaknesses, but particularly micro contrast across the frame.

In diglloyd L-Mount Mirrorless (included in subscription to diglloyd Mirrorless):

Panasonic 24-105mm f/4 Aperture Series @ 24mm: Mosaic

Includes images up to full camera resolution from f/4 through f/8, along with large crops showing the behaviors.

f5.6 @ 5.0 sec, ISO 100; 2019-04-04 19:52:43 [Enhance Details]
Panasonic S1R + Panasonic Lumix S 24-105mm f/4 Macro OIS @ 24mm

[low-res image for bot]

Zeiss ZM 35mm f/1.4 Distagon on Panasonic S1R: Evaluating Applicability for Environmental Portraiture

See my L-mount mirrorless wishlist

This page is an exploratory probe into whether the Zeiss ZM 35mm f/1.4 Distagon can deliver good quality for close-range environmental portraiture when focused somewhat off-center. Used that way, the issues of field curvature and corner to corner performance are largely of little import. The issues of bokeh and color shading and secondary color are of perhaps more interest than sharpness.

In diglloyd Leica M:

Zeiss ZM 35mm f/1.4 Distagon Examples: Portraits (Panasonic S1R)

Includes images up to full camera resolution, but also 24 megapixel images downsampled from 46.7 megapixels. The 24MP variants are the same resolution as what can be captured on a Leica M240 or Leica M10, helping to provide perspective versus the rangefinder option.

f2 @ 1/320 sec handheld, ISO 100; 2019-04-08 11:56:09 [Enhance Details]
Panasonic S1R + Zeiss ZM 35mm f/1.4 Distagon

[low-res image for bot]

f2 @ 1/500 sec handheld, ISO 100; 2019-04-08 11:53:47 [Enhance Details]
Panasonic S1R + Zeiss ZM 35mm f/1.4 Distagon

[low-res image for bot]

Leica 50mm f/1.4 Summilux-M ASPH on Panasonic S1R: Evaluating Applicability for Environmental Portraiture

See my L-mount mirrorless wishlist

This page is an exploratory probe into whether the Leica 50mm f/1.4 Summilux-M ASPH can deliver good quality for close-range environmental portraiture when focused somewhat off-center. Used that way, the issues of field curvature and corner to corner performance are largely of little import. The issues of bokeh and color shading and secondary color are of perhaps more interest than sharpness.

In diglloyd Leica M:

Leica 50mm f/1.4 Summilux-M ASPH: Portraits Examples (Panasonic S1R)

Includes images up to full camera resolution, but also 24 megapixel images downsampled from 46.7 megapixels. The 24MP variants are the same resolution as what can be captured on a Leica M240 or Leica M10, helping to provide perspective versus the rangefinder option.

f1.4 @ 1/800 sec handheld IBIS=on, ISO 100; 2019-04-08 11:57:51
[LACA corrected, Enhance Details]
Panasonic S1R + Leica 50mm f/1.4 Summilux-M ASPH

[low-res image for bot]

Panasonic S1R Multi-Shot High-Res Mode: Evaluating How Diffraction Affects Captured Detail and Micro Contrast

See my L-mount mirrorless wishlist

For years I’ve written about the damaging effects of diffraction versus high-resolution imaging but the Multi-Shot High-Res mode of the Panasonic S1R offers an interesting opportunity to verify the limits that I have observed for 42/45/50 megapixel full frame cameras shot at native resolution, which are as follows:

• f/6.3 is the smallest aperture at which diffraction can be ignored.
• f/8 shows subtle but consistent dulling versus f/5.6.
• f/11 and f/13 invariably cause obvious dulling that damages macro and micro contrast, but the negative effects can be partially offset with diffraction mitigating sharpening.
• f/16 is so marginal as to be a very poor choice with virtually all image, while f/22 is an unmitigated disaster showing major damage to macro and micro contrast and loss of resolving power.

How about the limits for Multi-Shot High-Res Mode which in effect uses pixels 1/4 the size (1/2 linearly)?

This page looks at the the limits imposed by diffraction for Multi-Shot High-Res mode as well as standard resolution mode. In diglloyd L-Mount Mirrorless,

Panasonic S1R: Panasonic S1R: Diffraction Limits for Multi-Shot High-Res Mode

Along with the Panasonic S1R HighRes mode workflow page, this page is essential reading for anyone using Panasonic S1R Multi-Shot High-Res Mode.

Same image, different apertures = major quality loss from diffraction

Leica 28mm f/2 Summicron-M ASPH on Panasonic S1R: Evaluating Applicability for Environmental Portraiture

See my L-mount mirrorless wishlist

This page is an exploratory probe into whether the Leica 28mm f/2 Summicron-M ASPH can deliver good quality for close-range environmental portraiture when focused somewhat off-center. Used that way, the issues of field curvature and corner to corner performance are largely of little import. The issues of bokeh and color shading and secondary color are of perhaps more interest than sharpness.

In diglloyd Leica M:

Leica 28mm f/2 Summicron-M ASPH: Portraits Examples (Panasonic S1R)

Includes images up to full camera resolution, but also 24 megapixel images downsampled from 46.7 megapixels. The 24MP variants are the same resolution as what can be captured on a Leica M240 or Leica M10, helping to provide perspective versus the rangefinder option.

f2 @ 1/320 sec handheld, ISO 100; 2019-04-08 12:00:40 [Enhance Details]
Panasonic S1R + LENS_NA @ 0mm equiv (28mm)

[low-res image for bot]

Panasonic S1R: More Examples of Multi-Shot High-Res Mode vs Standard Resolution

See my L-mount mirrorless wishlist

Panasonic S1R High Resolution Mode

This page looks at the results when using the Panasonic S1R Multi-Shot High-Res mode.

Comparisons include standard vs HighRes mode images at native resolution as well as resampled to 125 megapixels (native sensor resolution is 46.7 megapixels)

In diglloyd L-Mount Mirrorless:

Panasonic S1R: Standard vs Multi-Shot High-Res Mode (Various)

Image can be toggled for A/B comparison to standard resolution.

At the least, studio shooters have an amazing new tool to work with! Anyone shooting products, artifacts, artwork, etc should be taking a hard look at this capability, but be sure to use only the very best lenses.

Below, I left the dust on this carved-stone Japanese storage box cover. At 125 megapixels (0.82X linear downsample from 187 megapixels), the detail is incredible. See my workflow for getting this kind of result.

Actual pixels crop from 125 megapixel image

f8 @ 1.0 sec Multi-Shot HighRes, ISO 100; 2019-04-06 19:15:33 [Enhance Details]
Panasonic S1R + LENS_NA @ 0mm equiv (55mm)

[low-res image for bot]

Panasonic S1R: Comparing Multi-Shot High-Res Mode to Standard Mode

See my L-mount mirrorless wishlist

Panasonic S1R High Resolution Mode

This page looks at the best possible results when using the Panasonic S1R Multi-Shot High-Res mode.

Comparisons include standard vs HighRes mode images at native resolution as well as resampled to 125 megapixels. Image can be toggled for A/B comparison. Observations on sharpness, noise, etc are made along with an estimate of the maximum possible gains.

In diglloyd L-Mount Mirrorless:

Panasonic S1R: Examples of Optimal Gains from Multi-Shot High-Res Mode (Dolls)

It’s clear after doing this comparison and also looking at others that these examples represent a “best possible” scenario—all sorts of things can degrade HighRes mode. Indeed, these examples shown above are compelling, but others I shot were less so.

Split view with standard vs HighRes (crop from 125 megapixels)

Panasonic S1R: Workflow for Multi-Shot High-Res Mode Images in Adobe Camera Raw

See my L-mount mirrorless wishlist

Panasonic S1R High Resolution Mode

As a prelude to presenting the Panasonic S1R Multi-Shot HighRes mode, I needed to sort out a solid workflow. This took some time and experimentation to work when and how much to sharpen as well as how much to downsample.

This page looks at a workflow in Adobe Camera Raw for Panasonic S1R Multi-Shot High-Res mode raw files. Use of Adobe Camera Raw Enhance Details is assumed, as it has subtle benefits; it is the best place to start. In diglloyd L-Mount Mirrorless:

Panasonic S1R: Workflow for Multi-Shot High-Res Mode Images in Adobe Camera Raw

Step by step instructions with screen shots are shown.

Resampling an image (downsampling)

Panasonic S1R: How Much to Sharpen During Raw Conversion in Adobe Camera Raw + Adobe Camera Raw Enhance Details

See my L-mount mirrorless wishlist

As a prelude to presenting the Panasonic S1R Multi-Shot HighRes mode, I felt it was important to critically evaluate the appropriate level of sharpening for standard frames. Factors to consider when choosing a sharpening value include:

• Starting level of micro contrast— a mediocre lens or diffraction losses might demand more sharpening even if there are downsides.
• The point at which the image starts to look “crunchy” or “stressed”. This effect is separate from accentuation of noise, though both tend to come on together.
• The point at which noise becomes accentuated producing a grittiness or worse.
• How well subtle differences in high-key values are distinguished. Often this can be at odds with other areas in that what looks good for high-key may be too much elsewhere.
• Presentation medium (prints of course, but even different computer displays!), personal preferences.

This page looks at how much sharpening to apply for standard single-shot exposures during raw conversion when converting using Adobe Camera Raw (ACR) in Adobe Photoshop (also applies to Adobe Lightroom). In diglloyd L-Mount Mirrorless:

Panasonic S1R: How Much to Sharpen During Raw Conversion in Adobe Camera Raw

Panasonic S1R: Adobe Camera Raw Enhance Details

Three different crops are shown across a range of sharpening values.

Note: on Retina displays, see the 2nd crop.

Enlarged crop with low and high sharpening

;

[low-res image for bot]