Determining if a sensor is "ISOless"

[jonathanlung]

First off, sorry if it would have been better to revive this old thread or even hijack a more recent thread. Secondly, please feel free to correct me anywhere. My assumptions/understanding may be wrong entirely!

I (think I) understand Guillermo's posts on the topic of ISO-less sensors, and I definitely understand the practical implications. On the other hand, it seems like a bit of work to figure out if a sensor is ISOless using the data provided by DxO (ultimately, via sensorgen) using the method he uses. I'm just wondering if the following approach is a valid and easier way to make the same determination using readily-available DxO calculations.

The intuition: the primary reason to increase sensor gain is (from a data-collection perspective rather than a viewing perspective, for which not having a dark image might be the main reason), for a fixed exposure, to have less noise in the shadows. As such, one is potentially sacrificing highlight data, or at least not making optimal use of the full-well capacity of the sensor (by maximizing SNR due to shot noise). In other words, the reason why we crank up the ISO is because we expect that when we sacrifice some headroom in the brighter part of the scene, we gain at least a little in the shadows.

So if, after applying gain to the sensor data and assuming sufficient light, the most luminous parts of the scene will be at (approximately) the same value as it would have been had the exposure been adjusted instead of the sensor gain, would it make sense to assume that a sensor's DR, for a given ISO, is limited by its performance in the darker parts of the scene? If so, instead of having to work with raw SNR data as Guillermo has done, if DR is measured by finding when SNR drops below a certain point, would it instead be easier to determine whether a sensor is ISOless using published DR measurements? Or, in other words, does sacrificing a stop from the high end buy us anything in the shadows, or have we simply lost a stop of DR?

The method would look like this: take two pairs of ISO-DR values, (ISO1, DR1) and (ISO2, DR2). If (log(ISO1/ISO2)/log(2))/(DR2-DR1) ≈ 1, then the sensor is basically ISOless for those pairs of ISOs and the lower ISO of the pair should be used for taking a shot. If, instead of being approximately 1, the left-hand side is, say, >1.3, then the sensor is not ISOless. Some values:
Canon 1D MkIV: ISO100 vs. ISO3200: 2.3
Nikon D3s: ISO200 vs. ISO6400: 2.1
Canon 5D MkII: ISO100 vs. ISO6400: 2.0
Pentax K5*: ISO200 vs. ISO1600: 1.2
Sony A77: ISO100 vs. ISO3200: 1.1
Nikon D7000*: ISO100 vs. ISO3200: 1.0
Nikon D300: ISO200 vs. ISO3200: 1.0
Leica M8: ISO160 vs. ISO1250: 1.0

The *-ed camera (sensors), I think, are ones that come up in ISOless discussions.

If this method is valid, then it looks like ISOless sensors have actually been with us for a while (the Nikon D2X, for example). For what it's worth, I shot a pair of RAW images on my D300 at the same aperture and shutter speed, varying only the ISO - ISO200 for one and ISO3200 for the other. I processed both in RPP and one of the full-resolution demosaicing algorithms, but using a +4EV compensation on the ISO200 image, and compared the output. They seemed to have roughly (visually estimated) the same amount of noise, as I would expect based on the DR values at DxOMark. There was only a tad more chroma noise in the ISO200 image.

[Guillermo Luijk]

would it instead be easier to determine whether a sensor is ISOless using published DR measurements? Or, in other words, does sacrificing a stop from the high end buy us anything in the shadows, or have we simply lost a stop of DR?

That is totally correct. An isoless sensor loses 1 entire stop (or nearly) of DR for every extra ISO stop, straight from base ISO. So any sensor with this behaviour can be considered isoless.

This plot clearly reflects that the K5 is an ISOless sensor, while Nikon and specially Canon require the user to cleverly make use of the ISO setting to get the best from their cameras:

[jonathanlung]

Thanks! That completely changes the way I'm using my camera from now on when shooting indoor sports! I normally to boost ISO to decrease noise in the shadows, but end up with highlight clipping, but since I'll end up with about the same noise, I'll avoid the clipping by staying at base ISO and pushing later.

[Guillermo Luijk]

That should be the way to work with isoless cameras.

Just to illustrate, this is a quite high dynamic range scene, with well defined local highlights (the window and the lamp):



After post processing:


In front of the scene we could have followed two approaches:
- Stay at base ISO100 obtaining an underexposed RAW file, with the idea to lift the shadows in post processing
- Push ISO to get better exposure in the important part of the scene (the inner subjects) and hence less noise

In an isoless camera, the 'less noise' argument becomes false, we won't have less noise for pushing ISO as long as shutter/speed (i.e. light impacting the sensor) remain the same. The result of both approaches would be:



- Staying at base ISO100 we preserved well the lamp's surroinding area. We nearly have the same noise in the deep shadows.
- Pushing ISO100 we ruin the lamp's surrounding area, but we hardly improve noise in the shadows.

An obvious but important consideration is that in order to post process the shadows, we should lift them with some highlight preserving strategy. Using the exposure slider in the RAW developer will clip the highlights. The 'Bright' tool in ACR for instance pushes exposure in the shadows softly preserving the highlights.

The bad thing of this philosophy is that we can get dark camera displays that can make difficult to check framing, sharpness,... image in general. If our camera in particular has ISO as pure metadata (like the MFDBs discussed), we can still expose for base ISO, but in the end set a higher ISO in order to 'see' better the camera's display.

[jonathanlung]

Thanks for the extra info!

That should be the way to work with isoless cameras.

I started this thread because I was surprised to find my D300 ought to be ISOless!

The bad thing of this philosophy is that we can get dark camera displays that can make difficult to check framing, sharpness,... image in general. If our camera in particular has ISO as pure metadata (like the MFDBs discussed), we can still expose for base ISO, but in the end set a higher ISO in order to 'see' better the camera's display.

If only we had the equivalent of CHDK for all cameras! That's one thing that _could_ get me to ditch my flash and glass and switch to a different camera system offering similar ergonomics, price, and image quality.

[BobFisher]

This is similar to the concept of whether it's worthwhile to ETTR above base ISO that Francisco discussed in this thread, isn't it?  The math Jonathan outlined really doesn't need to be done then, does it?  Based on the graph GL provided and the similar graphs DxO provides for the cameras it reviews, one could determine whether the sensor is ISO-less by looking at the shape of the DR curve, no?  As long as the slope of the curve is less than 1 (i.e., less than a 1 stop drop in DR for a 1 stop increase in ISO) then it can be beneficial to increase ISO as opposed to push the ISO in post.

[FranciscoDisilvestro]

Another way is to look at the graphs available at sensorgen.info (mentioned in the OP) which are derived from the DxOMark data.

Attached are the sensorgen graphs for the Canon 5D MKII and the Nikon D7000. In the 5D MKII graph, the read noise is higher at low ISO up to ISO 1600, and the DR stays about the same up to ISO 400. Instead, in the case of the D7000, read noise is almost constant at any ISO value and DR drops 1 stop for every doubling of ISO.

The D7000 can be considered ISOless in this case

[JonathanRimmel]

Hmm. it would seem I still don't quite understand ISO.

Bottom line. Should I or shouldn't I use higher ISO's? Or should I just underexpose like crazy and boost exposure in post? (seems to go against ETTR).

[Schewe]

Bottom line. Should I or shouldn't I use higher ISO's? Or should I just underexpose like crazy and boost exposure in post? (seems to go against ETTR).

It depends on the camera and how well it does or doesn't amplify the signal when you increase the ISO. If the ISO settings is merely a metadata tag then you won't get any real benefit by increasing the ISO and doing ETTR. If the camera does a good analog amplification then you prolly will up to a point. It all depends on the camera. So, what camera?

[jonathanlung]

Based on the graph GL provided and the similar graphs DxO provides for the cameras it reviews, one could determine whether the sensor is ISO-less by looking at the shape of the DR curve, no?  As long as the slope of the curve is less than 1 (i.e., less than a 1 stop drop in DR for a 1 stop increase in ISO) then it can be beneficial to increase ISO as opposed to push the ISO in post.

Same thing, yes.

[JonathanRimmel]

It all depends on the camera. So, what camera?

Currently I use (personally) a Nikon d300, d70 and (at work) Canon 7d.  Right now I am mostly concerned about the 7d, since I do quite a bit of low light shooting at the college I work for, whereas in my personal photography, not so much.

[Guillermo Luijk]

Currently I use (personally) a Nikon d300, d70 and (at work) Canon 7d.  Right now I am mostly concerned about the 7d, since I do quite a bit of low light shooting at the college I work for, whereas in my personal photography, not so much.

In all Canons it's mandatory to make a clever use of ISO, i.e. push ISO as soon as RAW underexposure happens at ISO100. The DR of Canon cameras at ISO200 is nearly the same as at ISO100. This also means that in terms of deep shadows noise, pushing from ISO100 to ISO200 on a Canon camera means almost the same improvement as pushing exposure by 1 stop by optical means (i.e. increasing photon count through aperture or shutter).

As a rule of thumb, on a Canon camera we can (almost) forget about ISO100 and consider ISO200 our base ISO. In fact I usually do that when I get too long exposure times when shooting HDR interiors at f/16 with my 5D. And I guess the decission of considering ISO200 as base ISO in many cameras (Nikon, Olympus) is closely related to this behaviour. Basically not having ISO100 is only a drawback when long exposures are the goal.

Some HDR composites taken at ISO200 in the Vitoria cathedral, an impressive restoration work in progress in the North of Spain:

A 100% crop of the last shot:

Regards