Panasonic cameras are the worst cameras for lighting photography.

I don't understand your comment about Sony FF cameras.

I use Sony A7 24mpx sensor and A99ii 42mpx sensor with long exposure of sometimes more than 2 minutes.

If I don't activate the long exposure noise reduction option quickly the images are showing red/blue/green spots in the shadow parts of the image. Not hundreds but I have had ended up with some unusable images due to the amounts of spots.

Now if you are saying that you see these artifacts with exposures up to 15 seconds for eg (and long exposure noise reduction off) , then yes I would consider this also totally abnormal. For me it would prevent me to shoot fireworks for instance which any of my Sony cams are currently up to easily (being apsc or ff)
A7ii vs S1. Same settings, raw and noise reduction off. Not a single hot/stuck pixel on the Sony vs hundreds on the Panasonic. You can see why I'm peed off.

5fa42a4b53714279b426fdcd792bef9a.jpg


b6ea0cca03a942c9af3fa515977fe411.jpg
Have you played with the noise reduction in Lightroom?
I never do that myself for instance as I totally rely on the LENR of the Sony camera.

Do you know if this setting is available on my version of LR ? I use an older version 6, and eventually I could be interested to compare LENR with that in-LR feature.
LENR is totally not acceptable in that scenario
You mean in the scenario of lightning and/or fireworks shots I suppose.

If that is yes I concur as I did also some fireworks shots, and obviously I deactivated it.

It'd be a shame to miss 70% of the show because of this feature

--
 
I don't understand your comment about Sony FF cameras.

I use Sony A7 24mpx sensor and A99ii 42mpx sensor with long exposure of sometimes more than 2 minutes.

If I don't activate the long exposure noise reduction option quickly the images are showing red/blue/green spots in the shadow parts of the image. Not hundreds but I have had ended up with some unusable images due to the amounts of spots.

Now if you are saying that you see these artifacts with exposures up to 15 seconds for eg (and long exposure noise reduction off) , then yes I would consider this also totally abnormal. For me it would prevent me to shoot fireworks for instance which any of my Sony cams are currently up to easily (being apsc or ff)
A7ii vs S1. Same settings, raw and noise reduction off. Not a single hot/stuck pixel on the Sony vs hundreds on the Panasonic. You can see why I'm peed off.

5fa42a4b53714279b426fdcd792bef9a.jpg


b6ea0cca03a942c9af3fa515977fe411.jpg
Have you played with the noise reduction in Lightroom?
I never do that myself for instance as I totally rely on the LENR of the Sony camera.

Do you know if this setting is available on my version of LR ? I use an older version 6, and eventually I could be interested to compare LENR with that in-LR feature.
LENR is totally not acceptable in that scenario
I'm not sure. The more I zoom the more I find. I'd guess this photo has over 20k hot stuck pixels. I never would have touched Lumix if I'd know this.

fc6d0ab2796849cfbe721af12982257d.jpg
I don't consider this acceptable neither... from far. But the camera is just a tool and we have to use the right tool for the job to do.

--
 
Well, I prefer to use better NR in post than whatever a camera can irreversibly bake in. That's at least the logic behind Panasonic not doing it (at least not on that scale). Of course, it would be ideal if we were given a choice though.
I doubt Lightroom will ever be able to fix this.

f78df8a5ca0844d68cadd32a50cc4d23.jpg
 
Last edited:
Indeed... not cool at all...

What you're reporting is even concerning me a bit now, because I am staying on the side walk (talking about mirrorless gear) waiting for an eventual S1R mark II and finally enter the L-mount alliance platform.... But if the results are these, even if it would be a 100mpx FF sensor, that wouldn't be fitting at all with my needs :-|
This should frighten the crap out of every photographer.

d854abef21a34d2892e5639223ef4702.jpg


2365ceed6342401d9ad84d61695b0383.jpg
This looks very excessive.

Are you sure there is no problem with the camera sensor ?

Well if your 2 cams give same results ... damn ... I prefer o laugh of it seriously :-)

That's ridiculous.

--
 
Indeed... not cool at all...

What you're reporting is even concerning me a bit now, because I am staying on the side walk (talking about mirrorless gear) waiting for an eventual S1R mark II and finally enter the L-mount alliance platform.... But if the results are these, even if it would be a 100mpx FF sensor, that wouldn't be fitting at all with my needs :-|
This should frighten the crap out of every photographer.

d854abef21a34d2892e5639223ef4702.jpg


2365ceed6342401d9ad84d61695b0383.jpg
This looks very excessive.

Are you sure there is no problem with the camera sensor ?

Well if your 2 cams give same results ... damn ... I prefer o laugh of it seriously :-)

That's ridiculous.

--
https://www.actionphotopassion.com
That was after around 30 minutes of photos so the sensor was warm to hot. But even a cool sensor gives me hundreds of pixels. That pic was a freshly turned on camera that had been off for hours in a cool room at 100 iso. I'd love if anyone reading this could do the same test and post their results.



714e75e72ff14657a4ed655f3aabd923.jpg
 
Indeed... not cool at all...

What you're reporting is even concerning me a bit now, because I am staying on the side walk (talking about mirrorless gear) waiting for an eventual S1R mark II and finally enter the L-mount alliance platform.... But if the results are these, even if it would be a 100mpx FF sensor, that wouldn't be fitting at all with my needs :-|
This should frighten the crap out of every photographer.

d854abef21a34d2892e5639223ef4702.jpg


2365ceed6342401d9ad84d61695b0383.jpg
This looks very excessive.

Are you sure there is no problem with the camera sensor ?

Well if your 2 cams give same results ... damn ... I prefer o laugh of it seriously :-)

That's ridiculous.

--
https://www.actionphotopassion.com
Unless someone can convince me both my s1's are faulty I consider them simply unusable without LENR turned on. I wouldn't even risk shots at a few seconds.
 
I don't understand your comment about Sony FF cameras.

I use Sony A7 24mpx sensor and A99ii 42mpx sensor with long exposure of sometimes more than 2 minutes.

If I don't activate the long exposure noise reduction option quickly the images are showing red/blue/green spots in the shadow parts of the image. Not hundreds but I have had ended up with some unusable images due to the amounts of spots.

Now if you are saying that you see these artifacts with exposures up to 15 seconds for eg (and long exposure noise reduction off) , then yes I would consider this also totally abnormal. For me it would prevent me to shoot fireworks for instance which any of my Sony cams are currently up to easily (being apsc
 
Change out your shooting strategy, Timelapse , maybe 5 seconds on, 1 second off. Let it rip. Should be able to score plenty, adjust if needed.
I'm not the OP, I can't and don't talk for him. I'm just reading your answer and even if it was a nice workaround, the main question here would be: Why would long exposure work for other platform correctly and incorrectly only on the Lmount platform ...

If the work around is woking well that means that's the way to do it on this specific Lmount platform, but still the root cause is somewhere else. I know that we need to know our gear, but when in the field, and in the dark, and/or stressed by time and one has been used to use all his different gear 1 way and has to adjust for 1 single platform ... difficult
 
Change out your shooting strategy, Timelapse , maybe 5 seconds on, 1 second off. Let it rip. Should be able to score plenty, adjust if needed.
I'm not the OP, I can't and don't talk for him. I'm just reading your answer and even if it was a nice workaround, the main question here would be: Why would long exposure work for other platform correctly and incorrectly only on the Lmount platform ...

If the work around is woking well that means that's the way to do it on this specific Lmount platform, but still the root cause is somewhere else. I know that we need to know our gear, but when in the field, and in the dark, and/or stressed by time and one has been used to use all his different gear 1 way and has to adjust for 1 single platform ... difficult
 
All in all I think you've made a good point about one of the apparent weaknesses of the Lumix cameras for the particular type of photography you're doing, and in the shooting conditions you're mainly in (warm/hot/humid temps). However, without minimizing your issue, I think it is fair to point out that one only has to search "hot pixels long exposures Sony" to see that across various forums, the issue is reported and demonstrated within Sony's system as well.

That said, it is important to have an objective viewpoint on the strengths and weaknesses of a given brand or particular models within a brand. A while ago I made some posts about apparent compression artifact problems Panasonic seems to have with in-camera 8-bit (jpeg) stills and possibly 8-bit video, which also seems to be across the brand (at least in the FF cameras). Certainly I never saw the same issue in my years shooting Fuji, whose brand strength is high quality jpegs (though of course Fuji has other weaknesses too). For long exposures with Lumix, it seems like others here have some success with similar imagery, though using different techniques from yours. I wonder if using a lightning trigger (such as the ones made by MIOPS) with Lumix cameras would be helpful, since the sensor wouldn't need to be on/active for long periods of time. Shorter exposures would help keep the sensor cooler.

Anyway, that's my dad post of the day :) Good luck in your journey, at least you live in an area where a photography genre you enjoy can be done regularly enough to require the right equipment for it.

-- The grass isn't always greener, unless you shoot Velvia.
 
After shooting several night storms with the Lumix s1 I can honestly say this is the worst camera system I've ever used for shooting lightning. The first time I used the s1 every picture was unusable because of the shocking amounts of hot/stuck pixels in every raw photo. Even shooting at iso 100 for a few seconds would result in hundreds of red/green/blue pixels, and longer exposures would result in thousands. Ever bright well exposed photos were affected. I thought the camera must have a bad sensor until I started doing some research online and came to learn that this is surprisingly normal for Lumix cameras and the only way to prevent it is to turn on Long Exposure Noise Reduction in the settings. This fixed the pixel problem but created another massive problem. Waiting for the darn noise reduction buffer to clear! It has literally cut the amount of shooting time in half and I've already missed dozens of awesome lightning shoots while waiting for that stupid noise reduction. I've shot lightning with Nikon, Fujifilm & Sony and never had to worry about this and never had to turn on noise reduction. Yes now & then I would get a few hot pixels but nothing that would ever ruin my photos like the Lumix s1. I personally find it laughable that such a modern camera can be this bad at something Sony's first full frame mirrorless camera the a7 achieved in 2013 and the Nikon d3 from 2007, two cameras that I've owned since their release. The s1 is a phenomenal camera in so many ways but is truly backwards in some of the most basic. Even pixel peeping a raw photo in preview to check focus can't be done on the s1 which is about as basic at it comes and has been on every other brand camera for 20+ years.
UPDETE... After contacting Panasonic via email their camera department suggested manually doing DARK FRAME SUBTRACTION in post to avoid the long exposure noise reduction time problem. After learning how much BS it takes regarding conversation, importing and exporting into Photoshop (which I've never used) I'll be taking a BIG pass on that technique. I'll be going back to Sony for lightning.
Again, hot pixel reduction is automatic in LR, and nearly automatic in Capture One. I'm sure other packages offer it as well. In-camera LENR is completely unnecessary. No missed shots.

What PP software are you using that does not offer this?

<edit> I see that not everyone is happy with LR's hot-pixel removal, so perhaps it's not as ubiquitous as I thought. Perhaps folks who use LR (or other packages) can chime in here.

Anyway, good luck with your Sony setup, they make fine cameras & lenses. I guess I'm lucky that I picked CaptureOne years ago due to the way they handled X-trans, back when I was a Fuji X shooter.
What Panasonic should have done was give the user the option to have the long exposure noise reduction baked in of not. They told me in the email that they don't bake their noise reduction into the raw photos to benefit astro photography but how in the hell could I possibly use an unbaked s1 raw file full of hundreds to thousands of hot pixels anyway? And that's at low to base iso. Most astro photography is done at higher iso's which would be even worse.
Well, all I can tell you is that plenty of us easily make it work. I've taken thousands of long-exposure images using the S1 & S5 at elevated ISO and don't have any problems with getting excellent images.

Too bad about your resistance to using an S5. You could have what is probably the best system on the market for capturing lightning/fireworks/star trails. But we all have our own unique priorities, I suppose.
Are all of your long exposure shots done without LENR turned on?
Correct. I never use LENR.
 
These are lens cap shots. Lens cap installed during exposure.

100% views of three raw files. The first is a 60 second exposure at ISO 100 with the camera at ambient temp, inside my house, no sun exposure.

Second is another 60 second shot, but this one is at then end of a 60-shot sequence of 60-second exposures. Same conditions as above. So, the end of 1 hour of continuous exposures. The body was warm to the touch, but far from hot.

The third was a 4 second shot taken after the 60-shot sequence.

This first snapshot is from Capture One, 100% view, with the "Single Pixel NR" disabled. I'm zoomed into the lower left corner, where the hot pixels seem to be the worst.

5ce7e3b319904cfea9d67a0eae84f62d.jpg.png




Second is with the exposure pushed 4 stops, again, no single-pixel NR:

0395f927adf742b998a08f9356905ef8.jpg.png




Third is with the single-pixel NR enabled (you can see the slider in the left panel, set to 20 out of 100, which is stronger than I normally need to use it) along with the 4-stop push:

702033aab18741de8f18d89eec92ac64.jpg.png


So, as expected, pushing shadows on long exposures results in significantly more hot pixels.

However, my non-pushed images show much less noise than your examples, but I'm not sure if you were pushing those files. And of course ambient temps also play a role.

Also, it's clear that my body has no problem with a short exposure (4 seconds) even after extensive use. So while you may not be willing to trust your S1s with even a few second exposures, I am highly confident in mine (and my S5 as well) under those circumstances.

Some notes on Sony:

One of my regular astro shooting companions - who is highly accomplished in that genre - uses Sony & LR. I asked him if he ever gets hot pixels for astro images (keep in mind that we use trackers and often expose for several minutes at a time), and if so what he does about it, and he said yes, that some images have "a lot" of hot pixels. He also confirmed that LR does not have the equivalent of Capture One's single-pixel tool, so he sometimes has to use the clone/repair tools if he really wants to remove the hot pixels. Additionally, he felt that the current AI hot-pixel removal tool was not yet well suited to astro images, but perhaps it would work for your images.

It would be interesting to see how the various brands stack up with regard to hot pixels in long exposures in a hot environment. In general, I am quite content with Panasonic's approach, but if they ever do offer some kind of automated hot-pixel NR, I strongly would like it as an option that can be disabled. I understand the current approach could be frustrating for LR users, at least until the AI model can be trained better.

--
Capture One LUMIX FF feature request thread: https://www.dpreview.com/forums/thread/4658107#forum-post-66298057
 
Change out your shooting strategy, Timelapse , maybe 5 seconds on, 1 second off. Let it rip. Should be able to score plenty, adjust if needed.
I'm not the OP, I can't and don't talk for him. I'm just reading your answer and even if it was a nice workaround, the main question here would be: Why would long exposure work for other platform correctly and incorrectly only on the Lmount platform ...

If the work around is woking well that means that's the way to do it on this specific Lmount platform, but still the root cause is somewhere else. I know that we need to know our gear, but when in the field, and in the dark, and/or stressed by time and one has been used to use all his different gear 1 way and has to adjust for 1 single platform ... difficult
Nikon and Lumix are both using the 24mp sensor made by Sony. They all have this much hot pix and noise levels to start with, the only difference is the baked-in noise reduction for each brand. Using noise reduction results in loss of details and sharpness, but results in cleaner picture. I rather have more noise then a picture which lost too much details by noise reduction.

Lumix use the least or none, Nikon use mediocre noise reduction levels (if I recall correctly) and Sony the most of these 3 brands.
I see.

Well as a lazy person then Sony's algo suits me perfectly.
 
This image uses Panasonic's excellent "Live View Composite" and represents 60 seconds of net exposure time. I used a 2-second shutter speed, which works great for lightning, but let the LVC mode active for about 1 minute. So, a single 2-second exposure for the frame as a whole, and then about 30 2-second exposures to collect transient highlights for a total of 1 minute.

Shot inside a closet with me moving my smartphone's LED to simulate a brief-but-bright bright event (i.e., a lightning strike).

Again zoomed in 100%, with the exposure pushed four stops (exposed at ISO 400). No single-pixel NR used.

You can see there are effectively zero hot pixels here (the area I have zoomed in on was nearly black before the exposure was pushed. I know the OP is unlikely to go this way, but I post it as evidence to others that LVC really works brilliantly for lightning photography, as well as fireworks, star trails, etc., with virtually no hot pixels.

All raw files available upon request.

e2b9f958e0e546118da7e190cb39b721.jpg.png


--
http://georgehudetzphotography.smugmug.com/
Capture One LUMIX FF feature request thread: https://www.dpreview.com/forums/thread/4658107#forum-post-66298057
 
Last edited:
Let the S1 sit in a sunny window sill for 2 hours, running another set of 60-second exposures during the second hour. The raw files might be a little worse, but the change was very small. Very much like the first set.
 
These are lens cap shots. Lens cap installed during exposure.

100% views of three raw files. The first is a 60 second exposure at ISO 100 with the camera at ambient temp, inside my house, no sun exposure.

Second is another 60 second shot, but this one is at then end of a 60-shot sequence of 60-second exposures. Same conditions as above. So, the end of 1 hour of continuous exposures. The body was warm to the touch, but far from hot.

The third was a 4 second shot taken after the 60-shot sequence.

This first snapshot is from Capture One, 100% view, with the "Single Pixel NR" disabled. I'm zoomed into the lower left corner, where the hot pixels seem to be the worst.

5ce7e3b319904cfea9d67a0eae84f62d.jpg.png


Second is with the exposure pushed 4 stops, again, no single-pixel NR:

0395f927adf742b998a08f9356905ef8.jpg.png


Third is with the single-pixel NR enabled (you can see the slider in the left panel, set to 20 out of 100, which is stronger than I normally need to use it) along with the 4-stop push:

702033aab18741de8f18d89eec92ac64.jpg.png


So, as expected, pushing shadows on long exposures results in significantly more hot pixels.

However, my non-pushed images show much less noise than your examples, but I'm not sure if you were pushing those files. And of course ambient temps also play a role.

Also, it's clear that my body has no problem with a short exposure (4 seconds) even after extensive use. So while you may not be willing to trust your S1s with even a few second exposures, I am highly confident in mine (and my S5 as well) under those circumstances.

Some notes on Sony:

One of my regular astro shooting companions - who is highly accomplished in that genre - uses Sony & LR. I asked him if he ever gets hot pixels for astro images (keep in mind that we use trackers and often expose for several minutes at a time), and if so what he does about it, and he said yes, that some images have "a lot" of hot pixels. He also confirmed that LR does not have the equivalent of Capture One's single-pixel tool, so he sometimes has to use the clone/repair tools if he really wants to remove the hot pixels. Additionally, he felt that the current AI hot-pixel removal tool was not yet well suited to astro images, but perhaps it would work for your images.

It would be interesting to see how the various brands stack up with regard to hot pixels in long exposures in a hot environment. In general, I am quite content with Panasonic's approach, but if they ever do offer some kind of automated hot-pixel NR, I strongly would like it as an option that can be disabled. I understand the current approach could be frustrating for LR users, at least until the AI model can be trained better.

--
http://georgehudetzphotography.smugmug.com/
Capture One LUMIX FF feature request thread: https://www.dpreview.com/forums/thread/4658107#forum-post-66298057
Very good info thank you for taking the time to test. I tried Lightrooms new AI noise reduction in some of the pics I've posted and it was able to dull the hot pixels by around 30%, but most of that is simply softening/smoothing. Also that process took 12 minutes per photo. My friend has an old Nero lightning trigger that he said I can lone so that might solve my problem to a degree. It surprises me Panasonic allowed their raws to be uncooked when they are so shockingly noisey.
 
Wow, I am glad I stumbled on this thread as I have noticed this on my S5, I did some lens cap on tests and I have quite a big cluster of red pixels on mine, they are hot rather than stuck I assume as they are not there on normal shutter speeds but seem to kick in at 2 seconds, I did 2, 5 and 10 seconds and it was there and the same, have not tried anything longer yet but will. I'll post pictures later for anybody interested.

My questions are...

  1. Is this considered a fault or just a quirk/limitation and therefore expected behaviour? I am wondering whether to go down the warranty rabbit hole.
  2. Is it something that is seen mostly on the Sony made sensors?
  3. Linked to #2 are other manufacturers better?
I like long exposures so I'll admit that this worries me a little. I have a 30 minute exposure from the OG Canon 5D from circa 2006 that looks great for example.
 
Wow, I am glad I stumbled on this thread as I have noticed this on my S5, I did some lens cap on tests and I have quite a big cluster of red pixels on mine, they are hot rather than stuck I assume as they are not there on normal shutter speeds but seem to kick in at 2 seconds, I did 2, 5 and 10 seconds and it was there and the same, have not tried anything longer yet but will. I'll post pictures later for anybody interested.

My questions are...
  1. Is this considered a fault or just a quirk/limitation and therefore expected behaviour? I am wondering whether to go down the warranty rabbit hole.
  2. Is it something that is seen mostly on the Sony made sensors?
  3. Linked to #2 are other manufacturers better?
I like long exposures so I'll admit that this worries me a little. I have a 30 minute exposure from the OG Canon 5D from circa 2006 that looks great for example.
At this point in time I believe it's just a normal Panasonic thing. Very sh#tty design in my opinion. Worst noise I've experienced since I started photography in 2007, on FF of apsc.
 
...

My questions are...
  1. Is this considered a fault or just a quirk/limitation and therefore expected behaviour? I am wondering whether to go down the warranty rabbit hole.
  2. Is it something that is seen mostly on the Sony made sensors?
  3. Linked to #2 are other manufacturers better?
Well, it's probably expected behavior.

It's something you will see on all CMOS sensors. A few years back I have seen a test, where Canon seems much worse than Nikon or Sony. But Sony on the other hand does some noise reduction backed in what has other downsides (it also deletes stars for example). It's debatable what's the better solution. It's probably depending on the workflow and what you want to get.

As for the TO, the Sony solution seems quite good and better than the Panasonic way (what kind of hands the problem more to the user). Maybe the live composite of newer Panasonic bodies would be also a nice solution, but it's not available on the S1.
 
Last edited:
IMO Panasonic's approach is truer to what the raw format represents: raw, uncooked data that you can process as you wish.

As an astrophotographer I disdain Sony's baked in spatial noise reduction, the infamous "star eater".

One of the reasons I specifically chose Panasonic is because their raw output is as unprocessed as it could be.

I use the S5 with long exposures, typically at ISO 640, and I have zero issues with hot pixels.

Just use a tool that can deal with them effectively. I use Rawtherapee to convert my raws which has a specific setting for hot pixels. It's free app by the way.
 

Keyboard shortcuts

Back
Top