Panasonic S1 vs S1R low light performance

[McCoy86]

While shooting the Milky Way last month with the Panasonic S1R and Sigma 14-24mm f/2.8, I was noticing more noise while processing my raw images in LR than with the A7Riii I was using last year. I broke down and purchased a S1 to replace my X-T2 camera earlier this month, as from all the reading, the S1 is much better at higher ISO due to the sensor (equivalent or same as the A7iii). I had debated on buying a used A7S, but in the end I really didn't want to support two camera systems and some of the features in the S1/S1R are really enjoyable... time lapse mode, display articulation, night assist features, and most of all... being able to power off a power pack.

I decided last night to perform a few tests with the S1 and S1R while shooting Comet NEOWISE behind the house, just to see the difference at ISO 6400 for myself. I was using ISO 6400, 14mm, f/2.8, 6 second exposure for both cameras and when doing a compare in LR at maximum zoom, here is the difference in noise. A little more than I was expecting to see at ISO 6400.





FWIW, here is my single exposure Milky Way photo from the S1R at ISO8000, 15mm, f/2.8, and 25s exposure. In LR I brought up the shadows to try and show some of the wheat field, but all I get is green static. I can get the Milky Way to look acceptable if I don't pull the shadows and apply some noise reduction. I could probably have done a composite photo if I had known what I know now, but that's in hind sight and I try to keep to single exposures as much as possible for time reasons.



Here is my Milky Way photo (single exposure) from the S1 last week at ISO 12,800, 14mm, f/2.8, 20s exposure. I was able to pull the shadows and can see the wheat fields with no just a little bit of noise. There is a little more light pollution in this photo, but should not affect the comparison by to much. I did take a few photos at ISO 25,600 and they looked similar to the S1R at ISO 8000.





FWIW, I am keeping both camera. For the majority of my landscape work, I am at the base ISO and the S1R has done great for me. For my time lapses and night photography I will be using the S1 going forward.

Disclaimer: This is the first time I have tried a side-by-side comparison, so might have done a few things wrong in my approach. With that said, there should be enough information here to help answer a few questions that viewers might have had about higher ISO performance of the S1 and S1R.

Regards,--
-Monty

 

[Off The Mark]

Thanks for posting your results.

That third photo of the Milky Way and the field is gorgeous.

 

[DBE]

Just out of curiosity, try saving out the S1R image as the equivalent 6000 x 4000 image as the S1 -- in other words, reduce the image directly when saving from Camera RAW. I tested that idea once with the ISO 6400 RAW files from the DPreview set of test images for both cameras, and found that the resulting noise levels were pretty close ...

 

[georgehudetz]

Thanks for posting. Great images! Other than the noise in that first one, LOL.

I've had similar results with the S1R & astro, and have been wondering about the S1. I had some amazing conditions the other night, with a fantastic foreground under the milky way, and I thought that 60 seconds at ISO 6400 would give me what I needed, but I was wrong. Live and learn.

The S1R really isn't very good at high ISO & deep shadows. I think there may be a long-exposure aspect to it as well. Like you, I bought the R for landscape, so I'm going to accept the tradeoff, but it's a little disappointing.

I think the key for using the R for astro is to either use a tracker, or do multiple exposures, including dark frames (see my thread on "Noise Grid" a few posts down). For the foreground, you really need to do long exposures and keep the ISO down below 3200. This will mean going beyond the limit of 60 seconds in the current firmware, so you have to use bulb mode or remote control, like the smartphone app. Or maybe do multiple 60 second exposures and do a median stack in PS.

--
http://georgehudetzphotography.smugmug.com/
My Flikr stream: http://flic.kr/ps/Ay8ka

 

[georgehudetz]

Just out of curiosity, try saving out the S1R image as the equivalent 6000 x 4000 image as the S1 -- in other words, reduce the image directly when saving from Camera RAW. I tested that idea once with the ISO 6400 RAW files from the DPreview set of test images for both cameras, and found that the resulting noise levels were pretty close ...

Here is my image, 60 seconds, f/2.8, ISO 6400, exposure pushed 0.6 stops, and then downsampled to 24 MP in Capture One. Ignore the sky as the exposure is too long & therefore shows star trails. I have 16 10 second frames I will be stacking soon for that. But, the clouds in the valley below me (this is at Rock Cut in Rocky Mountain National Park, at about 12,000 feet altitude) are quite noisy, even after downsampling.







--

George Hudetz Photography

Powered by SmugMug

georgehudetzphotography.smugmug.com

My Flikr stream: http://flic.kr/ps/Ay8ka

 

[McCoy86]

Just out of curiosity, try saving out the S1R image as the equivalent 6000 x 4000 image as the S1 -- in other words, reduce the image directly when saving from Camera RAW. I tested that idea once with the ISO 6400 RAW files from the DPreview set of test images for both cameras, and found that the resulting noise levels were pretty close ...

I've only used Camera RAW once from Photoshop. I did have a moment to take a copy of the image from the S1R into photoshop, resize to a maximum width of 6000 and from LR it does look cleaner for noise. With that said, this doesn't help when you are trying to recover detail from the shadows at higher ISO with the S1R.





--
-Monty

 

[McCoy86]

Thanks for posting. Great images! Other than the noise in that first one, LOL.

I've had similar results with the S1R & astro, and have been wondering about the S1. I had some amazing conditions the other night, with a fantastic foreground under the milky way, and I thought that 60 seconds at ISO 6400 would give me what I needed, but I was wrong. Live and learn.

I can say that the S1 was well worth the purchase. I was lucky enough to find a refurbished S1 on Adorama for under $1400 and scooped it up without a second thought.

The S1R really isn't very good at high ISO & deep shadows. I think there may be a long-exposure aspect to it as well. Like you, I bought the R for landscape, so I'm going to accept the tradeoff, but it's a little disappointing.

Same here, and as you I was a little disappointed in the low light performance. It's funny that the A7Riii and A7Riv both have low light performance that rivals the S1, it's just to bad the ergonomics don't work for my hands, the display doesn't articulate the way I like it, and it's lacking a lot of external buttons and knobs that the S1/S1R have and I enjoy.

-Monty

 

[McCoy86]

Just out of curiosity, try saving out the S1R image as the equivalent 6000 x 4000 image as the S1 -- in other words, reduce the image directly when saving from Camera RAW. I tested that idea once with the ISO 6400 RAW files from the DPreview set of test images for both cameras, and found that the resulting noise levels were pretty close ...

Here is my image, 60 seconds, f/2.8, ISO 6400, exposure pushed 0.6 stops, and then downsampled to 24 MP in Capture One. Ignore the sky as the exposure is too long & therefore shows star trails. I have 16 10 second frames I will be stacking soon for that. But, the clouds in the valley below me (this is at Rock Cut in Rocky Mountain National Park, at about 12,000 feet altitude) are quite noisy, even after downsampling.

The static noise in the darker areas looks very similar to my results. Still a beautiful photo non the less! Thanks for sharing!!!

--
-Monty

 

[georgehudetz]

The static noise in the darker areas looks very similar to my results. Still a beautiful photo non the less! Thanks for sharing!!!

Thanks Monty.

Agreed with your comments about the noise signature of the S1R sensor vs the Sony R cameras. It is unfortunate. I'm sure this is part of the reason for the low adoption rate of the S1R. Like you I really like the ergonomics of the S1 series - not to mention the hi-res mode - and I'm going to stick with it. I'd love to have an S1 as well, but right now I'm focused on getting lenses. I'm hoping that as far as astro is concerned, following best practices, (stacking, dark frames, etc) should overcome the noise issues with the S1R. But yeah, the S1 is likely the better camera for astro.

I was hoping to get out in the last few days, but too many clouds right now here in Northern Colorado.

 

[DBE]

"Agreed with your comments about the noise signature of the S1R sensor vs the Sony R cameras. It is unfortunate. I'm sure this is part of the reason for the low adoption rate of the S1R."

Interesting. I checked the DPR high ISO tests for the S1, S1R, Sony A7R IV and the Nikon D850 - screen shot below.



Unless you can see something I can't, it appears that the noise levels are about the same for the high-pixel sensors.

If you save out the S1 and S1R ISO 6400 test shots (no noise reduction) from Camera RAW at the same resolution, the S1 is slightly better, but possibly not enough to notice in a well-crafted print at normal viewing distances.




S1 - ISO 6400




S1R - ISO 6400 downsampled

I'm debating the purchase of the new Laowa 15mm L-mount to start learning the skills for astro landscape photography with the S1R. (There are some great locations up here in New Hampshire and Maine - check out Adam Woodworth's site at https://adamwoodworth.com/ I've met him personally and he does incredible work). If I come up with anything halfway decent using his techniques, I'll post some images.

 

[Tom Caldwell]

Years ago I bought a Sony A7R for the pixels and found it terminally useless in low light. I should have bought the A7S (the “S” was actually more expensive at the time ...)

Older and wiser I bought the S1 this time round - it seems that I was indeed better educated by experience (and it was cheaper than the pixeled-up S1R .... )

Lower light theatre capture (not happening much atm) is my speciality - not astro - looking forward to back to business whenever this might happen.

I would hate to have to drag my old Canon 5D out of retirement yet again (as I did with the A7R which had proved useless for theatre).

 

[georgehudetz]

Well, I can't honestly say that I've personally done a comparison between the Panasonic S1R and Sonys at equivalent ISOs. So you got me there. But I can say that I've been surprised by the noise levels in certain circumstances in the S1R.

One thing that is true is that the S1R sensor is not "ISOless" (see the DPR review). So, when you try to lift shadows or the entire exposure in high-ISO shots - as is common in astro - the results can be underwhelming. Yes, you can just push the ISO higher in the original shots, but then you risk getting results like I am seeing in my other thread (assuming my sensor is healthy).

I do feel that the high-ISO behavior of the S1R sensor is not up to par with the competition, and DPR reinforces that in their review. But as Tom points out above, one shouldn't be buying a pixel-monster for high-ISO performance anyway.

--
http://georgehudetzphotography.smugmug.com/
My Flikr stream: http://flic.kr/ps/Ay8ka

 

[McCoy86]

I do feel that the high-ISO behavior of the S1R sensor is not up to par with the competition, and DPR reinforces that in their review. But as Tom points out above, one shouldn't be buying a pixel-monster for high-ISO performance anyway.

To reinforce this one, David Brendan (don't know him) has tested both the S1, S1R, and several Sony camera's. I found his tests quite useful in determining that the S1 was the right direction to go. Just look at the amount of noise at 30 seconds... gulp!

https://www.brendandaveyphotography.com/

 

[georgehudetz]

One more point - if you activate the "Low Light" setting in the DPR comparison tool, you'll find that the S1R does not do as well as it's high-resolution competitors - particularly on the left edge of the image in the lighter areas. That, coupled with the fact that pushing the S1R in high-ISO conditions is problematic, is a 1-2 punch that's hard to recover from.

Also, again with Low Light active, you will see that the S1 does much better than the S1R, even when you use comparable resolutions.

 

[McCoy86]

Years ago I bought a Sony A7R for the pixels and found it terminally useless in low light. I should have bought the A7S (the “S” was actually more expensive at the time ...)

From reviews, it looks like Sony fixed the low light issue in the A7Rx series on the A7Riii. Yes, the A7S was high on my list a month ago, but was just missing to many features I have grown to enjoy. The A7Siii that was just released today should be an interesting one to watch, but that price!

Lower light theatre capture (not happening much atm) is my speciality - not astro - looking forward to back to business whenever this might happen.

Yep, I shoot photos at my kids concerts and indoor sports at times (use to should I say) and would have loved to have the S1 last year.

-Monty