SL3 , am I the only one ?

[Dr. G]

Advertisement (gone after registration)

30 minutes ago, Photoworks said:

Considering that the base ISO on this camera is 100 for photos, ISO 50 is -1 stop pull and will reduce your highlight DR.

The SL3 sensor has dual gain, it is not specific in the manual but my impression is that it is at 100 and 400 ISO

For those of us who aren't as technical, what does that mean?  That dynamic range at 100 and 400 are the same?

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[BernardC]

3 minutes ago, Dr. G said:

For those of us who aren't as technical, what does that mean?  That dynamic range at 100 and 400 are the same?

Essentially yes, however it is distributed differently. EI 100 will have 2 stops more "shadow DR" than 400, and EI 400 has two stops more "highlight DR" than 100.

There are reasons to prefer one EI over another. Many people choose the lowest possible EI, and then protect for highlights ("expose to the right") because highlights usually have a hard maximum, whereas shadows can get deeper and deeper. That's what people are really saying when they tell you to shoot at "base ISO."

That's what people did for slide film, they spot-metered the brightest part of the image (or incident-metered the key light) to make sure that they weren't blowing the highlights. Shadows fell into place from there.

The downside to this technique is that your exposures will vary from shot to shot (because the brightest highlight varies). That's fine if you want to produce one image, but it's a bother if you want to deliver multiple images (eg: event/wedding, or video), because each shot needs to be processed individually. In that case it's better to pick a middle EI, like 400, knowing that you probably won't blow any highlights but your shadows will be noisier. You can then apply your basic corrections/look to a whole group of shots instead of tweaking each one. That's a huge time saving if you have thousands of images from one event.

[scott kirkpatrick]

33 minutes ago, BernardC said:

Essentially yes, however it is distributed differently. EI 100 will have 2 stops more "shadow DR" than 400, and EI 400 has two stops more "highlight DR" than 100.

 

That's not quite right.  It neglects the fact that increasing ISO is always at the expense of shadow depth.  It's true that <100 ISO gains you nothing.  I agree that when shooting at 100 ISO, it is a good idea to leave room in the highlights by underexposing a bit (since you have many stops of useful dynamic range).  The small bump in dynamic range starts just after ISO 200, but comes at a  cost in the shadows.  Here's one analysis:

 

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!

 

 

[BernardC]

29 minutes ago, scott kirkpatrick said:

Here's one analysis:

PtP measures dynamic range in a different way from anyone else (Arriflex, Leica, CineD, etc). I'm not sure why that is, you'll need to ask them.

It's easy enough to test. Set your camera to manual exposure and shoot a frame at EI 100. Shoot a second frame at EI 200 with the same manual exposure setting (aperture and shutter speed). Open both frames in your favourite raw converter, match the exposure. PtP's chart suggest that the second frame will have a stop less DR. That's not what I see when I do it, or what others have reported. Both frames are the same once normalized to the same exposure.

 

If you think about it, PtP's numbers don't make much sense. If you have 11 stops of DR at EI 100, and only 10 at EI 200, where does that stop go? Why would you ever shoot at EI 200 if the only difference is that you get a stop less DR than at 100? By that logic, the DR at 100 should contain everything that is in the EI 200 shot, plus more.

The bump at EI 400 in the chart is probably related to the dual-gain architecture.

[Chaemono]

PtP also shows that SL3 and SL2-S have the same DR at ISO 100 which is not the case. SL3 has more DR at ISO 100.

[SrMi]

5 hours ago, Chaemono said:

PtP also shows that SL3 and SL2-S have the same DR at ISO 100 which is not the case. SL3 has more DR at ISO 100.

P2P shows data for PDR, one of the variants of possible DR measurements (EDR being another popular one).

How do you measure SL3 as having more DR than SL2-S at ISO 100?

[SrMi]

Advertisement (gone after registration)

6 hours ago, BernardC said:

PtP measures dynamic range in a different way from anyone else (Arriflex, Leica, CineD, etc). I'm not sure why that is, you'll need to ask them.

Arriflex, Leica, and CineD are video cameras, and they probably measure DR for VLog, which does not relate to still photography.

6 hours ago, BernardC said:

It's easy enough to test. Set your camera to manual exposure and shoot a frame at EI 100. Shoot a second frame at EI 200 with the same manual exposure setting (aperture and shutter speed). Open both frames in your favourite raw converter, match the exposure. PtP's chart suggest that the second frame will have a stop less DR. That's not what I see when I do it, or what others have reported. Both frames are the same once normalized to the same exposure.

You misunderstood what P2P measures. It shows that the maximum exposure (without clipping) at ISO 100 has more DR than the maximum exposure (without clipping) at ISO 200. The reason is that the maximum possible exposure at ISO 200 is lower than the maximum exposure at ISO 100.

6 hours ago, BernardC said:

If you think about it, PtP's numbers don't make much sense. If you have 11 stops of DR at EI 100, and only 10 at EI 200, where does that stop go?

If the maximum exposure at ISO 100 is 1/100 @  f/5.6, then the maximum exposure at ISO 200 is 1/200 @ f/5.6. Exposure determines achieved DR (noise).

[Chaemono]

vor 2 Stunden schrieb SrMi:

P2P shows data for PDR, one of the variants of possible DR measurements (EDR being another popular one).

How do you measure SL3 as having more DR than SL2-S at ISO 100?

I took a couple of shots with both cameras and the same lens in high contrast scenes. The SL3 ISO 100 pictures were not only brighter but highlights were better controlled, too, from what I could see. It’s one of the main reasons I changed my mind about the SL3. Once I have the camera, I will post some DNGs.

[SrMi]

2 minutes ago, Chaemono said:

I took a couple of shots with both cameras and the same lens in high contrast scenes. The SL3 ISO 100 pictures were not only brighter but highlights were better controlled, too, from what I could see. It’s one of the main reasons I changed my mind about the SL3. Once I have the camera, I will post some DNGs.

Your observations are not related to DR.

[jaapv]

15 hours ago, scott kirkpatrick said:

That's not quite right.  It neglects the fact that increasing ISO is always at the expense of shadow depth.  It's true that <100 ISO gains you nothing.  I agree that when shooting at 100 ISO, it is a good idea to leave room in the highlights by underexposing a bit (since you have many stops of useful dynamic range).  The small bump in dynamic range starts just after ISO 200, but comes at a  cost in the shadows.  Here's one analysis:

 

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!

 

 

One thing that makes me wonder is that the camera appears to use traditional noise reduction. Surely dual gain would lead to better results? 

[Chaemono]

vor 5 Stunden schrieb SrMi:

Your observations are not related to DR.

Red Dot Camera Forum Talk at around 49 min refers to better highlight retention as being a function of DR. They show this at ISO 200 SL3 vs. SL2. They also show SL2-S files.

 

 

[BernardC]

9 hours ago, SrMi said:

If the maximum exposure at ISO 100 is 1/100 @  f/5.6, then the maximum exposure at ISO 200 is 1/200 @ f/5.6. Exposure determines achieved DR (noise).

My point is: take two shots at 1/100 F:5.6. The only difference between the two shots is that the camera is set to EI 100 for one, and EI 200 for the other.

According to PtP's graph, the EI 200 shot will have a full stop less DR! In other words, all the ISO setting does is give you less DR. PtP doesn't mention if this disappearing DR is from the highlights, shadows, or evenly distributed between the two.

[SrMi]

2 hours ago, Chaemono said:

Red Dot Camera Forum Talk at around 49 min refers to better highlight retention as being a function of DR. They show this at ISO 200 SL3 vs. SL2. They also show SL2-S files.

 

 

And yet, highlight retention is not related to DR. 
Highlight retention is often related to post processing, while DR refers to raw data only: from clipping point to a limit.

[SrMi]

58 minutes ago, BernardC said:

My point is: take two shots at 1/100 F:5.6. The only difference between the two shots is that the camera is set to EI 100 for one, and EI 200 for the other.

According to PtP's graph, the EI 200 shot will have a full stop less DR! In other words, all the ISO setting does is give you less DR. PtP doesn't mention if this disappearing DR is from the highlights, shadows, or evenly distributed between the two.

No, that is not what the graph says, and I explained it in post #47.
The graph shows the maximum possible PDR when using a given ISO. It does not show change in PDR when changing ISO alone.

DR is the difference between a maximum and a minimum. PDR’s minimum and maximum are defined on P2P’s site.

 

[SrMi]

2 hours ago, jaapv said:

One thing that makes me wonder is that the camera appears to use traditional noise reduction. Surely dual gain would lead to better results? 

The SL3 uses dual conversion gain above ISO 250. What do you mean by traditional NR. Many manufacturers use some (small?) amount of NR at very high ISOs.

[jaapv]

As opposed to dual native ISO at higher ISO values like 3200 or  6400. Yes, most manufacturers do but Leica used to apply none to very light, a policy they appear to have brought more into line with other manufacturers  

[Photoworks]

8 hours ago, Chaemono said:

I took a couple of shots with both cameras and the same lens in high contrast scenes. The SL3 ISO 100 pictures were not only brighter but highlights were better controlled, too, from what I could see. It’s one of the main reasons I changed my mind about the SL3. Once I have the camera, I will post some DNGs.

that is strange, all my tests show that the SL3 is 1/3 stop darker than SL2, and Sl2s. Leica said that this is caused by the new cover glass of the sensor.

not that it makes a difference for DR

[SrMi]

12 minutes ago, jaapv said:

As opposed to dual native ISO at higher ISO values like 3200 or  6400. Yes, most manufacturers do but Leica used to apply none to very light, a policy they appear to have brought more into line with other manufacturers  

No camera applies dual native ISO at that high ISO, which is mostly around 400 or 640, IIRC. If they did, the range from ISO 400 (typical DCG point) to 6400 (your suggested DCG point) would have too much noise, and it would not show any improvement at ISO 6400 and above.

Note that Leica M11 applies dual conversion gain (DCG) quite early at ISO 200.

It is unclear to me how the engineers picked the point to apply the DCG.

 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[SrMi]

20 minutes ago, Photoworks said:

that is strange, all my tests show that the SL3 is 1/3 stop darker than SL2, and Sl2s. Leica said that this is caused by the new cover glass of the sensor.

not that it makes a difference for DR

I ran an indoor test with SL2-S vs SL3, with the same exposure (1/30 @ f/4) and ISO 800 (M mode). The brightness seems identical.

However, I notice a difference in metering (A mode), with SL3's metering creating a darker exposure (1/50 vs. 1/40 sec). 

I do not think that the metering difference can be related to the new cover glass of the sensor.

[jaapv]

According to https://www.cined.com/panasonic-lumix-s5-ii-lab-test-rolling-shutter-dynamic-range-and-latitude/

640 and 4000 for  V-Log. 

 

23 minutes ago, SrMi said:

No camera applies dual native ISO at that high ISO, which is mostly around 400 or 640, IIRC. If they did, the range from ISO 400 (typical DCG point) to 6400 (your suggested DCG point) would have too much noise, and it would not show any improvement at ISO 6400 and above.

Note that Leica M11 applies dual conversion gain (DCG) quite early at ISO 200.

It is unclear to me how the engineers picked the point to apply the DCG.