Panasonic S1R first stack

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Panasonic S1R first stack

Post by Lou Jost »

There is a lot to say about this 47Mp full-frame camera. The manual is 500 pages, but nevertheless lots of the things we care about are not discussed. I'll write more about it later.

Getting the S1R to trigger from a stacking controller is a bit tricky. When I get a new camera I usually buy its remote and wire that to a socket that goes into the female WeMacro or Stackshot cord. But the S1R has a more complex mechanism in its remote. I didn't use its remote; the old Panasonic Lumix remotes for the FZ200/300/330 fit the S1R and are easier to modify. I clipped a slice of a US copper penny and inserted it between the half-press and full-press contacts, and then connected the third contact to one control wire, and the half-press contact to the other control wire. When the controller shorts the two control wires, the camera simultaneously executes a half-press and full press, and this works. I'll take some pictures of the finished product later.

For my first stack I used the exciting hi res mode, which was the main reason I bought the camera. It makes 187Mp images!!! I imagine that the actual effective number of informative pixels is 5/8 *187 = 116Mp, using the same factor as in the Olympus cameras which use the same kind of algorithm.

The uncompressed RAW images are 300Mb each, and are 16736 x 11168 pixels. These are crazy numbers, and stacking such images stresses my computer. But it is worth it!

A sharp, fairly fast lens is needed for this mode to work well. For this test I used a Repro-Nikkor 85mm f/1.0, used at f/2.0 and m=1, so with an effective aperture of f/4.0. I'll write more about this amazing lens elsewhere on the forum.

Image

Above is the full stacked image. I accidentally did not catch the whole image in the stack; some of the corners are not in focus. I stacked the RAW files, converted to Tiffs with default ACR sharpening turned to zero. After stacking I sharpened the final output strongly, and used some de-hazing, a small amount of clarity, and increased saturation. I did no retouchig, so some DMap artifacts are visible on the wing edges and some scales. The reduced image above doesn't do justice to the real image, which is bitingly sharp on my display.

I then resized the giant image by 50% so it is just an ordinary 47Mp image. Here are two 100% crops of that normal-sized image derived from the giant image. It differs from a normal FF image because every pixel has full RGB information.

Image
Image

There is an unprecedented level of detail here.

Below are a few 100% crops of the full 16736 x 11168 pixel hi-res image (Zerene DMap). Remember that to see this same area of the wing on an ordinary FF image at m=1, you would have to zoom in to 200% of the ordinary image.

Image

Image

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Lou,

Amazing detail at 1X! Speaks highly of the camera hi res mode and the lens!!

Would be interesting to see the same image area zooming in to near normal pixel level with hi res mode off and on.

That's interesting about the remote camera trigger using 3 active contacts, and has me wondering how this works? I'm familiar with the Nikon 2 contact approach where the first half-press (focus) contact shorts to the second full-press contact (take image) which initiates a camera trigger. Was able to get this to work with 2 open collector type opto-isolators, one to each contact, which pull each contact (the camera has internal pull up) to camera common ground and triggered in sequence with a slight delay between half and full press contacts.

How long is the effective exposure in hi res mode (time between 1st shift and last shift image), and how long does the camera require to process the shift images into a final RAW image before you can take another image?

Very impressive results!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Mike, the Panasonic S1R remote actually has four contact segments on its plug. Besides full press and half press, it also has an option to trigger video from a second push-button on the remote. Inside the remote is a tiny push-button on a circuit board, and another tiny contact switch. The FZ200/300/330 remote is much simpler and easy to deal with.

Good question about the timings. I woke up this morning realizing I had forgotten to post that important data. I set the WeMacro controller times to 1s plus 17s, and the MicroMate took at least a second to move from one position to the next. So probably about 19s per slice. This includes all processing and writing time on a fast XQD card. I took 138 slices and there were no missed frames for buffer issues. I do not think the buffer was being used at this frame rate. But this was right on the edge of the fastest possible sustainable rate.

Edit: I did not set the camera to take non-shifted RAW images along with the shifted images. The camera can be set to do this, and it can be useful if there are movement artifacts that need to be retouched from an upsampled normal RAW, but this would have taken longer to write to the card.
Last edited by Lou Jost on Sat Jun 29, 2019 9:21 am, edited 1 time in total.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Lou,

The Nikon connector has 10 pins, but only 3 (2 active, 1 ground common) are used for triggering an image, I have no idea what the other pins are for.

Suspect the push button you mention is probably a reset or load new firmware button as the more complex remote may be based upon a micro-controller.

Yikes, that's a long time between being able to capture images!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

No Mike, the push-button thing on the circuit board in the remote is triggered by the plastic push-button on the outside of the remote.

Yes, it does take a while to make the stack. But this is very close to what my MFT PEN F takes in 80Mp hi-res mode. The annoying part is how long it takes to batch-process the RAW files and run the stacks in Zerene. With each image weighing in at 300Mb+. every step takes a very long time. But worthwhile for special pictures.

By the way, the 138 shots took less than half the battery (with LCD turned off) And this camera has multiple external power options (pure AC, USB charging from a computer while the camera is on, an extra batter grip, and more) so power is not a problem. Also there was no noticeable heat build-up. This may change if I run a faster stack in normal res mode. Will try that soon.

Adalbert
Posts: 2491
Joined: Mon Nov 30, 2015 1:09 pm

Post by Adalbert »

Hi Lou,
Congratulations on buying a new camera!
Can you combine the internal focus bracketing with the hi res mode ?
BR, ADi

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

That is the big disappointment about this camera (see my earlier posts about the S1R).

First, focus bracketing doesn't work with hi-res mode.

Second, focus bracketing doesn't work with non-native L mount lenses. Even with the Sigma mount converter and Sigma Art lenses certified as compatible with the S1R, and even though Sigma and Panasonic are partners in this, it still doesn't work.

Third, it is unclear what it actually does, even with native L mount lenses. Focus bracketing may mean literally just bracketing a few frames on either side of a chosen focus point. That remains to be seen. The manula is ambiguous on this point, and the internet is silent about it. I eagerly wait for someone with deeper pockets than me to report some focus stacking tests with a native L mount lens.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

Lou,

Can you export the shifted images, or RAW images by WiFi or USB while stacking? This would allow you to do the RAW conversion in parallel with the stacking session, reducing the overall session time.

Good news about the power options and lack of heat buildup.

Looks like a very attractive camera for serious macro work at high resolutions.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Can you export the shifted images, or RAW images by WiFi or USB while stacking? This would allow you to do the RAW conversion in parallel with the stacking session, reducing the overall session time.
Yes Mike, both options are available, but I haven't tried them yet.

dmillard
Posts: 639
Joined: Tue Oct 24, 2006 7:37 pm
Location: Austin, Texas

Re: Panasonic S1R first stack

Post by dmillard »

Lou Jost wrote: A sharp, fairly fast lens is needed for this mode to work well. For this test I used a Repro-Nikkor 85mm f/1.0, used at f/2.0 and m=1, so with an effective aperture of f/4.0. I'll write more about this amazing lens elsewhere on the forum.
Thanks for sharing this Lou - the combination of the lens and the camera produce stellar results. Is this the same copy of the Repro-Nikkor that you tested against a 105mm Printing Nikkor last summer?
https://www.photomacrography.net/forum/ ... pro+nikkor

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Is this the same copy of the Repro-Nikkor that you tested against a 105mm Printing Nikkor last summer?
Yes, David, it is. You may recall that I concluded the lens was better at f/2.8 than at f/2.0, and that I thought the PN105 images at f/2.8 were slightly better.

But then I started thinking about the lensrentals.com tests on the spherical aberration caused by thick sensor filter packs on digital cameras. Those aberrations became important when lenses are faster than 2.8, and/or when the exit pupil is very close to the sensor. In those cases, lenses designed for film (no filter pack at all) will show aberrations when used on a digital sensor with a thick filter pack at fast apertures. The very fast Repro-Nikkor has an exit pupil very close to the sensor. I had tested it on an MFT camera with 4mm thick filter pack, so some of the softness or lack of contrast I was seeing at fast apertures might have been caused not by the lens but by the sensor pack!

To see if that might be the case, I examined the aerial image produced by the lens, by imaging it with a 20x objective. It turned out that the aerial image is sharpest (by a lot) at f/2.0! The amount of detail it resolved was quite striking, far outresolving a sensor at 1x.

FF cameras tend to have thinner sensor packs than MFT cameras, so I gave this lens a try at f/2.0 on the Panasonic. It did very well. Also, to solve this filter pack problem for fast film-era lenses on digital sensors, I sent my old FF Pentax K-1 to Kolari Vision, which removed the whole sensor pack and replaced it with a "coverslip" just 0.2mm thick. So I now have a unique Pentax K-1 which behaves more like a film camera with these special lenses.

This will be the subject of its own post when I find the time.

dmillard
Posts: 639
Joined: Tue Oct 24, 2006 7:37 pm
Location: Austin, Texas

Post by dmillard »

Thanks Lou - that makes a lot of sense. I'm sure you are delighted to have discovered the cause of the issue, and equally pleased that the S1R required no surgery. :D

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

David, I am not so sure about the S1R not needing surgery. I won't dissect it, because the modification ruins the infinity focus for digital lenses, but I do think I am seeing some defects in this image that are due to the sensor glass thickness.

I'm in the process of comparing the S1R and the K-1 with this lens, to see if the sensor glass matters.

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

I just finished the modified K-1 stack. Definitely cleaner, less astigmatism and color aberration. Unfortunately the K-1 just has 4-shot pixel-shifting, so nominal resolution is not increased, though every pixel is based on measured RGB values instead of a demosaicing extrapolation.

Lou Jost
Posts: 5991
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

There's one other possible explanation for the Panasonic aberrations. I used the same flange distance for both cameras, determined by requiring the K-1 mage to give exactly 1x. But because of the Panasonic's thicker sensor pack, that lens+camera has to be moved a bit (870 microns, to be exact) relative to the location of the Pentax+lens. It may not be at the lens' designed optimum. The more correct procedure would have been to use the K-1 to focus on a point on the subject, and then, without changing the lens position, replace the camera with the Panasonic S1R and add or subtract rings from the camera mount until the image was again sharp.

Post Reply Previous topicNext topic