Sigma's Evaluative Metering

[xpatUSA]

I'm wondering how Sigma does the "evaluation"?

For example, the DP2 Merrill splits the scene into 256 separate areas.

If the evaluation just takes the average value of those areas then it is just a crude version of the old-fashioned 'average' metering. Therefore, some sort of fancy weighting must be involved.

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

 

[xpatUSA]

The Manual says:

"Even under strong back-lighting or complex lighting situations, the camera will give you the correct exposure" ...

... a grand promise! But I found this on a non-Sigma site:


Result not surprising to this old man ...

Anyone here shot a similar scene successfully without using in-camera EC?

 

[D Cox]

In your example, the Spot version has burnt out highlights. The Evaluative has dark shadows which can easily be lightened. I would use a Curves tool.

Don

 

[xpatUSA]

In your example, the Spot version has burnt out highlights.

Not unexpected when exposing for a strongly back-lit subject

The Evaluative has dark shadows which can easily be lightened.

If lightened to be like the Spot version would probably burnout highlights. And bring out the noise due to the lower exposure.

I would use a Curves tool.

Sounds like you're defending evaluative metering?

They are not my images, I got it them an article on the net

 

[D Cox]

In your example, the Spot version has burnt out highlights.

Not unexpected when exposing for a strongly back-lit subject

The Evaluative has dark shadows which can easily be lightened.

If lightened to be like the Spot version would probably burnout highlights. And bring out the noise due to the lower exposure.

I would use a Curves tool.

Sounds like you're defending evaluative metering?

They are not my images, I got it them an article on the net

I would lighten the shadows and not the highlights, using a Curves tool.

Don

 

[xpatUSA]

In your example, the Spot version has burnt out highlights.

Not unexpected when exposing for a strongly back-lit subject

The Evaluative has dark shadows which can easily be lightened.

If lightened to be like the Spot version would probably burnout highlights. And bring out the noise due to the lower exposure.

I would use a Curves tool.

Sounds like you're defending evaluative metering?

They are not my images, I got it them an article on the net

I would lighten the shadows and not the highlights, using a Curves tool.

I would have metered the subject and not have had to do that ;-)

 

[xpatUSA]

In your example, the Spot version has burnt out highlights.

Not unexpected when exposing for a strongly back-lit subject

The Evaluative has dark shadows which can easily be lightened.

If lightened to be like the Spot version would probably burnout highlights. And bring out the noise due to the lower exposure.

I would use a Curves tool.

I would lighten the shadows and not the highlights, using a Curves tool.

I just now tried Curves but I'm not good at them and couldn't match the spot shot at all.

However, the GIMP Exposure tool worked quite well:

 

[jande9]

I'm wondering how Sigma does the "evaluation"?

For example, the DP2 Merrill splits the scene into 256 separate areas.

If the evaluation just takes the average value of those areas then it is just a crude version of the old-fashioned 'average' metering. Therefore, some sort of fancy weighting must be involved.

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

Good question. I've often wondered this myself.

Jan

 

[D Cox]

The result from spot metering here is very poor. The background is badly overexposed.

What is wanted is that the person should be lighter and the background unchanged.

Don

 

[xpatUSA]

The result from spot metering here is very poor. The background is badly overexposed.

What is wanted is that the person should be lighter and the background unchanged.

We seem to be missing the point of the posted image.

To me it shows clearly that Spot exposed the subject well enough and that that camera's Evaluative underexposed the subject significantly.

Comments as to how either shot can be fixed and as to how the background looks are side issues.

Can we restrict our comments to the rendition of the subject without red herrings?

 

[Stillton]

I'm wondering how Sigma does the "evaluation"?

For example, the DP2 Merrill splits the scene into 256 separate areas.

If the evaluation just takes the average value of those areas then it is just a crude version of the old-fashioned 'average' metering. Therefore, some sort of fancy weighting must be involved.

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

I bet we could ask ChatGPT to design a test where a monitor would display a set of patterns to a camera with a given manufacturer/metering type, take photos, and then, based on those images, it could approximate the logic behind the algorithm.

 

[Stillton]

The result from spot metering here is very poor. The background is badly overexposed.

What is wanted is that the person should be lighter and the background unchanged.

Don

Well, if the spot was on the model, the results exactly are like what I would expect. The model is exposed somewhat correctly (or rather the spot area is exposed correctly), and everything else falls by the wayside.

 

[Stillton]

I'm wondering how Sigma does the "evaluation"?

For example, the DP2 Merrill splits the scene into 256 separate areas.

If the evaluation just takes the average value of those areas then it is just a crude version of the old-fashioned 'average' metering. Therefore, some sort of fancy weighting must be involved.

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

 

[xpatUSA]

I'm wondering how Sigma does the "evaluation"?

For example, the DP2 Merrill splits the scene into 256 separate areas.

If the evaluation just takes the average value of those areas then it is just a crude version of the old-fashioned 'average' metering. Therefore, some sort of fancy weighting must be involved.

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

I didn't know what "ChatGPT" was, so I signed up and asked it:

what is the best camera exposure metering?

It said:

The best camera exposure metering mode depends on the specific shooting conditions and your creative intent. There are several common metering modes available on most cameras, including:

1. Matrix (Evaluative) Metering: This mode analyzes the entire frame and takes into account various factors like subject brightness, color, and composition. It's a good all-around choice for general shooting because it considers the entire scene.
2. Center-Weighted Metering: This mode places more emphasis on the center of the frame and is useful for situations where the subject is in the center of the frame or you want to prioritize the exposure in the center while letting the edges of the frame go darker.
3. Spot Metering: Spot metering measures the exposure based on a very small area, usually at the center of the frame. It's ideal for precise metering on a specific subject or area, especially when the subject is significantly brighter or darker than the surrounding scene.
4. Partial Metering: This mode is similar to spot metering but covers a slightly larger area than the spot. It's useful when you need to meter a larger portion of the scene while still giving priority to the center.
5. Average or Center-Spot Metering: Some cameras offer these hybrid metering modes that combine aspects of spot and center-weighted metering, providing more versatility.

The "best" metering mode depends on what you're shooting:

• For general or everyday shooting, matrix metering is often a safe choice as it evaluates the entire scene.
• For portraits or subjects with a significant difference in brightness from the background, spot metering can help ensure the subject is properly exposed.
• For landscapes, center-weighted or matrix metering might be more suitable, as you typically want an even exposure across the frame.
• In challenging lighting conditions, it may be necessary to use exposure compensation to adjust the metered exposure based on the creative effect you want to achieve.

Ultimately, the best metering mode is the one that helps you achieve the exposure you desire for your specific photography goals. Experimenting with different modes and learning how they interact with exposure compensation can help you become a more versatile and creative photographer.

Pretty good, I thought,

Don, see bold.

 

[xpatUSA]

I'm wondering how Sigma does the "evaluation"?

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

I didn't know what "ChatGPT" was, so I signed up and asked it:

what is the best camera exposure metering?

Then I asked: what is Matrix (Evaluative) Metering?

It said:

Here's how Matrix (Evaluative) metering typically works:

1. Scene Analysis: The camera's built-in light meter divides the frame into numerous zones or segments, often using a complex algorithm to assess the overall scene.
2. Data Evaluation: It evaluates the brightness, color, and contrast of the different zones within the frame. Some zones may contain the subject, while others may represent the background.
3. Weighted Calculation: The metering system assigns different levels of importance or weight to each zone based on its analysis. For example, it may give more weight to the center of the frame or areas with the most significant contrast.
4. Exposure Calculation: Based on the data from the scene analysis and the weighted calculation, the camera's metering system calculates the optimal exposure settings, which typically include shutter speed, aperture, and ISO sensitivity.

Although it told me not much more that I got yesterday from many, many search engine results it did in seconds what took me about an hour!

Item 3. was especially significant re. my comment about weighting in the OP.

 

[xpatUSA]

I'm wondering how Sigma does the "evaluation"?

For example, the DP2 Merrill splits the scene into 256 separate areas.

Initially I wrote 16x16 but that's not what the Manual says. Neither does it say that the areas are equal in size.

 

[Jeffry7]

I'm wondering how Sigma does the "evaluation"?

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

I didn't know what "ChatGPT" was, so I signed up and asked it:

what is the best camera exposure metering?

Then I asked: what is Matrix (Evaluative) Metering?

It said:

Here's how Matrix (Evaluative) metering typically works:

1. Scene Analysis: The camera's built-in light meter divides the frame into numerous zones or segments, often using a complex algorithm to assess the overall scene.
2. Data Evaluation: It evaluates the brightness, color, and contrast of the different zones within the frame. Some zones may contain the subject, while others may represent the background.
3. Weighted Calculation: The metering system assigns different levels of importance or weight to each zone based on its analysis. For example, it may give more weight to the center of the frame or areas with the most significant contrast.
4. Exposure Calculation: Based on the data from the scene analysis and the weighted calculation, the camera's metering system calculates the optimal exposure settings, which typically include shutter speed, aperture, and ISO sensitivity.

Although it told me not much more that I got yesterday from many, many search engine results it did in seconds what took me about an hour!

Item 3. was especially significant re. my comment about weighting in the OP.

Since you are new to these AI language model agents, I would like to warn you that they sometimes spit out reasonable sounding lies. Or occasionally gibberish.

They can be fun. Some people have gotten amazing results with the AIs that generate images. But they scrape the internet for their source material and we all know what kind of BS is easily found there.

 

[Stillton]

I'm wondering how Sigma does the "evaluation"?

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

I didn't know what "ChatGPT" was, so I signed up and asked it:

what is the best camera exposure metering?

Then I asked: what is Matrix (Evaluative) Metering?

It said:

Here's how Matrix (Evaluative) metering typically works:

1. Scene Analysis: The camera's built-in light meter divides the frame into numerous zones or segments, often using a complex algorithm to assess the overall scene.
2. Data Evaluation: It evaluates the brightness, color, and contrast of the different zones within the frame. Some zones may contain the subject, while others may represent the background.
3. Weighted Calculation: The metering system assigns different levels of importance or weight to each zone based on its analysis. For example, it may give more weight to the center of the frame or areas with the most significant contrast.
4. Exposure Calculation: Based on the data from the scene analysis and the weighted calculation, the camera's metering system calculates the optimal exposure settings, which typically include shutter speed, aperture, and ISO sensitivity.

Although it told me not much more that I got yesterday from many, many search engine results it did in seconds what took me about an hour!

Item 3. was especially significant re. my comment about weighting in the OP.

Caveat, occasionally it will completely lie to you when answering, so keep an eye on that.
Then you can correct it and it says "I am sorry, you are absolutely correct" and then it repeats back your correction. It is hilarious.

 

[xpatUSA]

I'm wondering how Sigma does the "evaluation"?

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

I didn't know what "ChatGPT" was, so I signed up and asked it:

what is the best camera exposure metering?

Then I asked: what is Matrix (Evaluative) Metering?

It said:

Here's how Matrix (Evaluative) metering typically works:

1. Scene Analysis: The camera's built-in light meter divides the frame into numerous zones or segments, often using a complex algorithm to assess the overall scene.
2. Data Evaluation: It evaluates the brightness, color, and contrast of the different zones within the frame. Some zones may contain the subject, while others may represent the background.
3. Weighted Calculation: The metering system assigns different levels of importance or weight to each zone based on its analysis. For example, it may give more weight to the center of the frame or areas with the most significant contrast.
4. Exposure Calculation: Based on the data from the scene analysis and the weighted calculation, the camera's metering system calculates the optimal exposure settings, which typically include shutter speed, aperture, and ISO sensitivity.

Although it told me not much more that I got yesterday from many, many search engine results it did in seconds what took me about an hour!

Item 3. was especially significant re. my comment about weighting in the OP.

Since you are new to these AI language model agents, I would like to warn you that they sometimes spit out reasonable sounding lies. Or occasionally gibberish.

They can be fun. Some people have gotten amazing results with the AIs that generate images. But they scrape the internet for their source material and we all know what kind of BS is easily found there.

Thanks for the caveat, Jeffry!

The simple questions I've asked have so far returned credible results. I asked one about Stock Price charts and it too got credible answers. I don't intend to ask it to create anything yet and especially not images!

 

[Stillton]

I'm wondering how Sigma does the "evaluation"?

Unfortunately, each manufacturer has a different method - rarely if ever published:

https://en.wikipedia.org/wiki/Metering_mode#Multi-zone_metering

Any ideas on the Sigma method?

BTW, this is why I prefer to use histogram metering with multizone metering for my slow shots. That way the metering provides me with a base "average" exposure by default, and then my brain does the rest based on what I want exposed/underexposed/preserved.
I found it to be very effective and I remove this "wt* is this thing doing now?" moment. Instead, I already know what to expect from evaluation, and then I just quickly adjust based on the picture/histogram. This is much quicker than fiddling with the spot as you move it around the shot.

I didn't know what "ChatGPT" was, so I signed up and asked it:

what is the best camera exposure metering?

Then I asked: what is Matrix (Evaluative) Metering?

It said:

Here's how Matrix (Evaluative) metering typically works:

1. Scene Analysis: The camera's built-in light meter divides the frame into numerous zones or segments, often using a complex algorithm to assess the overall scene.
2. Data Evaluation: It evaluates the brightness, color, and contrast of the different zones within the frame. Some zones may contain the subject, while others may represent the background.
3. Weighted Calculation: The metering system assigns different levels of importance or weight to each zone based on its analysis. For example, it may give more weight to the center of the frame or areas with the most significant contrast.
4. Exposure Calculation: Based on the data from the scene analysis and the weighted calculation, the camera's metering system calculates the optimal exposure settings, which typically include shutter speed, aperture, and ISO sensitivity.

Although it told me not much more that I got yesterday from many, many search engine results it did in seconds what took me about an hour!

Item 3. was especially significant re. my comment about weighting in the OP.

Since you are new to these AI language model agents, I would like to warn you that they sometimes spit out reasonable sounding lies. Or occasionally gibberish.

They can be fun. Some people have gotten amazing results with the AIs that generate images. But they scrape the internet for their source material and we all know what kind of BS is easily found there.

Thanks for the caveat, Jeffry!

The simple questions I've asked have so far returned credible results. I asked one about Stock Price charts and it too got credible answers. I don't intend to ask it to create anything yet and especially not images!

Shoot, I should have read this comment, before typing my own warning, lol. Oh well.