Shooting an Event in a few Weeks - SL3 or SL2-S as second body?

[Dr. G]

Advertisement (gone after registration)

On 12/27/2024 at 2:42 AM, LocalHero1953 said:

Having shot the SL2-S and the Q3 43 side by side in low light, I don't think it can. Midlights and highlights are OK, but the shadow details just aren't there to the same degree of recoverability - noise levels rise dramatically. So the Q3 43 is OK at those ISOs (25,000) but at the expense of quite a lot of blocked shadows.

You have me rethinking things now.  Maybe I should ditch the q3 43, grab a Sigma 28-45 1.8 and just take two SL3 bodies with the zoom and 75 mounted and throw the 21 APO in the bag for the times that I need to take that wider shot.  

[jaapv]

Well, the Q3 is in some ways, as a daily camera, quite versatile, but for an assignment like yours two identical or at least similar bodies with appropriate lenses mounted seems far more practical.  You could even use an 28-90  and the 21 APO and forget about changing lenses altogether. Maybe throw a Summilux  APO 35 in the bag for those Bokeh shots.

[hdmesa]

The noise pattern in high ISO files from the SL2-S are still preferable to me over the SL3. The SL2-S noise is more uniform and filmic, while the SL3, while much improved over the SL2, is not quite as pleasing to me.

While scaled down results from the SL3 are very close to the SL2-S at high ISO, things are not so simple when photon count is extremely low and you need to lift the shadows.

Edited December 28, 2024 by hdmesa

[SrMi]

5 hours ago, hdmesa said:

The noise pattern in high ISO files from the SL2-S are still preferable to me over the SL3. The SL2-S noise is more uniform and filmic, while the SL3, while much improved over the SL2, is not quite as pleasing to me.

While scaled down results from the SL3 are very close to the SL2-S at high ISO, things are not so simple when photon count is extremely low and you need to lift the shadows.

A while ago, I posted a comparison between M11 (same sensor as SL3?) and SL2-S at ISO 12800, underexposed by 3 stops, and lifted in Lightroom +3. The difference is very small. You need to lift much more than that or have an even higher ISO to see a difference. I posit that one would be hard-pressed to see a difference in typical exposure settings.

 

[hdmesa]

43 minutes ago, SrMi said:

A while ago, I posted a comparison between M11 (same sensor as SL3?) and SL2-S at ISO 12800, underexposed by 3 stops, and lifted in Lightroom +3. The difference is very small. You need to lift much more than that or have an even higher ISO to see a difference. I posit that one would be hard-pressed to see a difference in typical exposure settings.

 

Isn't underexposing in good light different than shooting when photon count is very low? Perhaps I'm thinking of situations where read noise characteristics start driving the conversation. For example, in dim interior lamplight with large areas of dark shadows, I expose for the light source so it doesn't blow detail on the lamp, meaning the shadows are much more than three stops underexposed. 

For high ISO, one of my favorite things about both the SL2-S and SL3 is the noise character at the native resolution at the pixel level. Downsizing the 60mp file to 24mp loses that native "grain", IMO. So I prefer to judge both in low light in what a tester would call an unfair comparison. But the way I look at it is similar to film – I want to retain the native detail in the noise at all costs.

 

[SrMi]

53 minutes ago, hdmesa said:

Isn't underexposing in good light different than shooting when photon count is very low? Perhaps I'm thinking of situations where read noise characteristics start driving the conversation. For example, in dim interior lamplight with large areas of dark shadows, I expose for the light source so it doesn't blow detail on the lamp, meaning the shadows are much more than three stops underexposed. 

For high ISO, one of my favorite things about both the SL2-S and SL3 is the noise character at the native resolution at the pixel level. Downsizing the 60mp file to 24mp loses that native "grain", IMO. So I prefer to judge both in low light in what a tester would call an unfair comparison. But the way I look at it is similar to film – I want to retain the native detail in the noise at all costs.

The exposure is defined by shutter speed, aperture, and scene light. Low exposure means low photon count and read noise comes to fore. I do not see why one of the three components would have more influence than the others. However, a test would be useful.

[hdmesa]

Advertisement (gone after registration)

55 minutes ago, SrMi said:

The exposure is defined by shutter speed, aperture, and scene light. Low exposure means low photon count and read noise comes to fore. I do not see why one of the three components would have more influence than the others. However, a test would be useful.

Could be tested with the same camera against itself to set expectations. 25K ISO in well lit home interior scene in daylight versus same scene at night by single lamp light. Meter both to preserve highlights, then apply -3 stops exposure. Recover exposure from both in post, then recover shadows from both, which will mean much higher shadow recovery is required for the dark areas of the night shot. Read noise should overtake shot noise in the darker areas. 

I’ve seen this in practice by shooting at very high ISO in bright daylight in order to introduce grain character. But use that same high ISO at the end of blue hour, and you just can’t recover the dark areas of the image without seeing a lot of read noise. The exposure doesn’t determine how much light is hitting the scene and available to be recovered.

[FlashGordonPhotography]

2 hours ago, hdmesa said:

Could be tested with the same camera against itself to set expectations. 25K ISO in well lit home interior scene in daylight versus same scene at night by single lamp light. Meter both to preserve highlights, then apply -3 stops exposure. Recover exposure from both in post, then recover shadows from both, which will mean much higher shadow recovery is required for the dark areas of the night shot. Read noise should overtake shot noise in the darker areas. 

I’ve seen this in practice by shooting at very high ISO in bright daylight in order to introduce grain character. But use that same high ISO at the end of blue hour, and you just can’t recover the dark areas of the image without seeing a lot of read noise. The exposure doesn’t determine how much light is hitting the scene and available to be recovered.

This can happen. It’s not the sensor though. It’s that the spectrum of light seen is modified. Green is often affected and that’s HALF of the light a sensor is sensitive toe to. A tungsten bulb, sunset or blue hour light doesn’t contain all the green information.

Gordon

[Photoworks]

As i shoot events in low light often, I can tell you the noise is the smallest issue.

Using SL3, SL2, and SL2-s they all suffer from focusing speed and accuracy in low light.

I always have been using AF-S, for practical reasons I have been using the 24-90mm, or 24-70 in one camera and 50mm in the other camera.
I will use Flash with color filtration and external power pack. shooting people it is better to have a bit of light in the faces to make them look better and color equal.
Plus shooting people in small groups requires 5.6 to get all the faces in focus.

I find the Summicron-APO lenses AF speed ok in low light but not great, I got the SL3 to improve AF performance in low light, but it turns out it is not much improved at this time.
The apo lenses are often to contrasty for older people in bad light.

The 24-90 is a lens I will use for the reach, but the AF speed is not great in low light,  When using this lens I take 4-6 images of the subjects, A; for funny faces and blinking B: because the focus is not always accurate in this lens in low light, refocusing on every shot, and you may fire when it still finding focus. I would select only 1-2 images from that group.

The 50 is for atmosphere shots, and this days I am using the sigma 50 1.2, a great lens wide open and good AF speed.
The other camera this day uses more than 24-70 , the 2.8 helps with lower light on some occasions, I use the sigma because I like the quick MF switch.

SL3 and Flash.
I posted this before and reported it to Leica, when the Leica flash is mounted on the SL3, occasionally the camera is unresponsive and does not flash. I associate the issue with the hotshoe contact, the flash bumps and moves around with the larger diffusion on it, and crashes the camera, solution is to pull the battery and restart. It is a flow that happens 1-2 times a day on an event. I assume it is something similar when you mount a Nikon flash on a Leica, and the camera is totally confused and needs to restart.
I don't think this has been fixed on the SL3 firmware.

PS
Panasonic lenses performer better in low light, AF speed is fantastic, fast, and reliable. They are all not as sharp as the APO lenses, but great for people photos.
I use 18, 24, 50, 85. 
 

 

[hdmesa]

5 hours ago, FlashGordonPhotography said:

This can happen. It’s not the sensor though. It’s that the spectrum of light seen is modified. Green is often affected and that’s HALF of the light a sensor is sensitive toe to. A tungsten bulb, sunset or blue hour light doesn’t contain all the green information.

Gordon

True, I forgot about that.

My point to @SrMi was that high ISO performance should be evaluated in the worst possible scenarios and not just their evenly and well lit studio scene. 

[Dr. G]

7 hours ago, Photoworks said:

As i shoot events in low light often, I can tell you the noise is the smallest issue.

Using SL3, SL2, and SL2-s they all suffer from focusing speed and accuracy in low light.

I always have been using AF-S, for practical reasons I have been using the 24-90mm, or 24-70 in one camera and 50mm in the other camera.
I will use Flash with color filtration and external power pack. shooting people it is better to have a bit of light in the faces to make them look better and color equal.
Plus shooting people in small groups requires 5.6 to get all the faces in focus.

I find the Summicron-APO lenses AF speed ok in low light but not great, I got the SL3 to improve AF performance in low light, but it turns out it is not much improved at this time.
The apo lenses are often to contrasty for older people in bad light.

The 24-90 is a lens I will use for the reach, but the AF speed is not great in low light,  When using this lens I take 4-6 images of the subjects, A; for funny faces and blinking B: because the focus is not always accurate in this lens in low light, refocusing on every shot, and you may fire when it still finding focus. I would select only 1-2 images from that group.

The 50 is for atmosphere shots, and this days I am using the sigma 50 1.2, a great lens wide open and good AF speed.
The other camera this day uses more than 24-70 , the 2.8 helps with lower light on some occasions, I use the sigma because I like the quick MF switch.

SL3 and Flash.
I posted this before and reported it to Leica, when the Leica flash is mounted on the SL3, occasionally the camera is unresponsive and does not flash. I associate the issue with the hotshoe contact, the flash bumps and moves around with the larger diffusion on it, and crashes the camera, solution is to pull the battery and restart. It is a flow that happens 1-2 times a day on an event. I assume it is something similar when you mount a Nikon flash on a Leica, and the camera is totally confused and needs to restart.
I don't think this has been fixed on the SL3 firmware.

PS
Panasonic lenses performer better in low light, AF speed is fantastic, fast, and reliable. They are all not as sharp as the APO lenses, but great for people photos.
I use 18, 24, 50, 85. 
 

 

Great post, thank you.  I had the Sigma 50 1.2 but returned it as I liked the rendering of my 50 Summilux SL better.  But 50 1.2 would give me better low light performance and, as you point out, better focusing ability.  Separation would be similar to the 75 APO, although rendering would be different (I guess the Sigma 85 could be an option as well if it has the same AF performance).  I'm not opposed to cropping, and I'm familiar with the event layout and I could get away with a 50 instead of a 75.  24-70 could still be a decent option.

As far as flash goes, I'm not sure the contact is the problem.  I haven't tried it on the SL3, but on the Q3 43 the Godox Nano X3 is sometimes unresponsive, with the same issue - the shutter won't release at all at times.  In legacy mode on the flash transmitter this isn't an issue, it will always fire, but then you're in full manual.

 

[LocalHero1953]

I have had similar non-responsive behaviour of the Godox XPro trigger on the SL2-S. I've found that removing the trigger and screwing it in again firmly gets it going again. As it is irregular in occurrence, I have been unsure whether the problem is Leica's or Godox's. I was thinking of getting a X Nano as a replacement, but it sounds like that might not be the answer (and buying the upcoming SL3-S just for better behaviour in this regard would be an expensive solution).

Edited December 29, 2024 by LocalHero1953

[redrocksjoe]

I have an SL3 and a Q2, so not exactly your setup. I love the Q2, but if you're shooting indoors, the AF is just miserable on the Q2 compared to the SL3. I'd go with a second SL3 if it is easy to get. Otherwise I think your two camera combo is pretty neat 👍.

-Joe

 

Edited December 29, 2024 by redrocksjoe

[hdmesa]

17 hours ago, SrMi said:

A while ago, I posted a comparison between M11 (same sensor as SL3?) and SL2-S at ISO 12800, underexposed by 3 stops, and lifted in Lightroom +3. The difference is very small. You need to lift much more than that or have an even higher ISO to see a difference. I posit that one would be hard-pressed to see a difference in typical exposure settings.

 

I just noticed the P2P PDR chart shows the SL3 starts applying baked-in noise reduction at ISO 8,000 and up, while the SL2-S doesn't. We'd at least have to apply some NR in post to the SL2-S files at ISO 8,000 and up to make it a fair comparison. But how much? The SL3 in-camera menu has the option to affect NR: 1) for long exposures (on/off) and for 2) JPEGs (low/med/high) but not for the DNGs. Not sure we can make an apples-to-apples comparison at high ISOs.

https://www.photonstophotos.net/Charts/PDR.htm#Leica SL2-S,Leica SL3

The M11 doesn't bake in NR at these high ISO like the SL3. Perhaps the SL3 has higher internal temperatures leading to more read noise at high ISOs.

Edited December 29, 2024 by hdmesa

[Photoworks]

1 hour ago, Dr. G said:

As far as flash goes, I'm not sure the contact is the problem.  I haven't tried it on the SL3, but on the Q3 43 the Godox Nano X3 is sometimes unresponsive, with the same issue - the shutter won't release at all at times.  In legacy mode on the flash transmitter this isn't an issue, it will always fire, but then you're in full manual.

 

I was using the SF 60,Ihave so many of them, and they were the issue on the SL3. but for years I have been using them on the SL2 without any issues, the are a better fit on the SL2

Gogox triggers have not failed me, the only issue is the shooting delay of the preflash in TTL, but in Manual mode or legacy mode, it all goes away.

In reality, most of the time setting the flash to manual is an easy solution, the distance for the subjects is within range in most events, and the little over and under can be edited in post.

I would start with 3200 ISO and 1/125  f4-5.6 and the flash at 1/8 power and probably full CTO.

in the last few event, I have used the Godox AD200 II Pro and the head extension cord, and the head mounted on the L bracket plus with Gary Fong  the x3 Nano

 

[SrMi]

6 hours ago, hdmesa said:

I just noticed the P2P PDR chart shows the SL3 starts applying baked-in noise reduction at ISO 8,000 and up, while the SL2-S doesn't. We'd at least have to apply some NR in post to the SL2-S files at ISO 8,000 and up to make it a fair comparison. But how much? The SL3 in-camera menu has the option to affect NR: 1) for long exposures (on/off) and for 2) JPEGs (low/med/high) but not for the DNGs. Not sure we can make an apples-to-apples comparison at high ISOs.

https://www.photonstophotos.net/Charts/PDR.htm#Leica SL2-S,Leica SL3

The M11 doesn't bake in NR at these high ISO like the SL3. Perhaps the SL3 has higher internal temperatures leading to more read noise at high ISOs.

Comparing SL3's and M11's PDR graph and looking at the SL3's PDR Shadow Improvement Vs. ISO graph, it seems that the embedded NR becomes relevant only above ISO 12800.

[hdmesa]

4 minutes ago, SrMi said:

Comparing SL3's and M11's PDR graph and looking at the SL3's PDR Shadow Improvement Vs. ISO graph, it seems that the embedded NR becomes relevant only above ISO 12800.

Both the PDR and Shadow Improvement show the down arrow begins at ISO 8045 (8000).

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[SrMi]

7 minutes ago, hdmesa said:

Both the PDR and Shadow Improvement show the down arrow begins at ISO 8045 (8000).

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Yes, but we do not see any relevant effect of NR until after 12800. A relevant influence of NR is the deviation of the linearity of the PDR graph and increased shadow improvement:

https://www.photonstophotos.net/Charts/PDR_Shadow.htm#Leica SL3

 

[hdmesa]

33 minutes ago, SrMi said:

Yes, but we do not see any relevant effect of NR until after 12800. A relevant influence of NR is the deviation of the linearity of the PDR graph and increased shadow improvement:

https://www.photonstophotos.net/Charts/PDR_Shadow.htm#Leica SL3

 

It's possible that the NR is sustaining the linearity beginning at ISO 8000. It's not surprising to see more aggressive NR being applied above 12,800 because at those ISOs, we expect some tradeoffs like detail smearing and color integrity loss.

That said, we do have the M11 as a baseline without any baked in NR, and that seems to support your theory. Actually looks like to me that the sensor isn't designed to be analog above 12,800, but Leica has had Sony customize the sensor at these ISOs for both M11 and SL3. I say that because the A7R5 and GFX 100S both are outside the analog range above 12,800.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[hdmesa]

@SrMi Directly comparing the SL3 to the M11 at ISO 12,800 might give us some insight into how much NR is being applied to the SL3 files at that ISO.