[DSLRnovice]

Has anyone experimented with real world subjects to test the exposure limits of the K-1? Today was the first sunny day we have had this week, so I thought it would be a good time to test a contrasty scene.

In the film days the rule of thumb was f16 and the fraction of the film speed on a sunny day. It turned out that with my FA35 lens the camera called for 1/100 sec with f16 at ISO 200 instead of 1/200 based on the ASA. Anyway, based on that as the normal exposure, I found that I could get fairly close to the base exposure with RAW files in DCU from 3 stops under to 2 stops overexposed images. Mostly changed the exposure settings and some curves in the overexposed files.

This was not taking in consideration critical IQ but exposure. There was noticeable grain noise in the deeper solid areas at more than 2 stops underexposure, but only at 100%. I am not sure how much the camera settings affected the color in DCU. The camera was set for "natural" jpgs, which DCU uses when first opening the image.

I left everything but the exposure and curves alone, and was pleased to see the possibilities. It would be interesting to do a sort of "zone system" test to see what the true ISO setting would be, or at least find the best possible exposure for certain scenes. I don't think the in camera histogram is always sufficient for every subject and lighting.

[Tas]

Has anyone experimented with real world subjects to test the exposure limits of the K-1? Today was the first sunny day we have had this week, so I thought it would be a good time to test a contrasty scene.

In the film days the rule of thumb was f16 and the fraction of the film speed on a sunny day. It turned out that with my FA35 lens the camera called for 1/100 sec with f16 at ISO 200 instead of 1/200 based on the ASA. Anyway, based on that as the normal exposure, I found that I could get fairly close to the base exposure with RAW files in DCU from 3 stops under to 2 stops overexposed images. Mostly changed the exposure settings and some curves in the overexposed files.

This was not taking in consideration critical IQ but exposure. There was noticeable grain noise in the deeper solid areas at more than 2 stops underexposure, but only at 100%. I am not sure how much the camera settings affected the color in DCU. The camera was set for "natural" jpgs, which DCU uses when first opening the image.

I left everything but the exposure and curves alone, and was pleased to see the possibilities. It would be interesting to do a sort of "zone system" test to see what the true ISO setting would be, or at least find the best possible exposure for certain scenes. I don't think the in camera histogram is always sufficient for every subject and lighting.

I haven't tested for the limits per se, but I do have experience in managing exposures with the K-1. And I'm sure the more knowledgeable will be along to discuss the sunny 16 theory further who may have done some testing, in the meantime you're stuck with me.

If it's a nice sunny day, I'm aiming to use ISO 100. I normally shoot in DNG RAW as well as I know to get the best out of the K-1, RAW is the best option. This is the best way to reduce noise in your images and RAW gives greater lattitude where you need to dial in negative exposure due to highlights blowing out.

Personally I like to expose to the right (ETTR) as much as possible however sometimes there's sun on part of a cloud and the only way to control the exposure in camera is to drop the exposure then lift in PP. The Sony sensors are better used in this manner as they tend not to control highlights as well as the Canon sensor (according to most of the web) but will produce less noise if raising exposure in PP (when compared ot the Canon). Of course this when shooting RAW, so I can't say what jpeg would be like.

Tas

[DSLRnovice]

I haven't tested for the limits per se, but I do have experience in managing exposures with the K-1. And I'm sure the more knowledgeable will be along to discuss the sunny 16 theory further who may have done some testing, in the meantime you're stuck with me.

If it's a nice sunny day, I'm aiming to use ISO 100. I normally shoot in DNG RAW as well as I know to get the best out of the K-1, RAW is the best option. This is the best way to reduce noise in your images and RAW gives greater lattitude where you need to dial in negative exposure due to highlights blowing out.

Personally I like to expose to the right (ETTR) as much as possible however sometimes there's sun on part of a cloud and the only way to control the exposure in camera is to drop the exposure then lift in PP. The Sony sensors are better used in this manner as they tend not to control highlights as well as the Canon sensor (according to most of the web) but will produce less noise if raising exposure in PP (when compared ot the Canon). Of course this when shooting RAW, so I can't say what jpeg would be like.

Tas

Lately I have been trying out Raw Therapee from recommendations on this forum. I am finding out from RT that the unprocessed RAW files that come out of the camera are darker than shown in the camera's preview/histogram and on Digital Camera Utility. It looks like Ricoh is trying to prevent blown out highlights behind the scenes. So Exposing to the right might have more latitude than one would think.

OTTH, that also requires slower shutter speeds, higher ISO's or faster apertures that centering the histogram.

[RGlasel]

test to see what the true ISO setting would be

The subject of native ISO is a pointless rabbit hole filled with rubbish. Camera sensors have a bias voltage that is changed by an embedded processor, according to the luminace reported by the camera's meter and the sensor's signal is what's left when the bias voltage is removed. The ISO setting is digital gain applied to that signal, basically multiplying the signal by a number. The sensor manufacturer programs the embedded processor to match the camera manufacturer's requirements, so the ISO "sweet spot" is whatever the camera manufacturer wants it to be. The camera owner may not like where that "sweet spot" is located, but it doesn't change the fidelity of the signal from the sensor. In other words, let the camera select ISO instead of arguing with it. Histograms are good for telling you if the shutter speed and aperture you selected is going to cause exposure problems, but I don't think they are useful for outsmarting the camera's programming.

[DSLRnovice]

The subject of native ISO is a pointless rabbit hole filled with rubbish. Camera sensors have a bias voltage that is changed by an embedded processor, according to the luminace reported by the camera's meter and the sensor's signal is what's left when the bias voltage is removed. The ISO setting is digital gain applied to that signal, basically multiplying the signal by a number. The sensor manufacturer programs the embedded processor to match the camera manufacturer's requirements, so the ISO "sweet spot" is whatever the camera manufacturer wants it to be. The camera owner may not like where that "sweet spot" is located, but it doesn't change the fidelity of the signal from the sensor. In other words, let the camera select ISO instead of arguing with it. Histograms are good for telling you if the shutter speed and aperture you selected is going to cause exposure problems, but I don't think they are useful for outsmarting the camera's programming.

Thank you for the technical explanation. I am not sure about your point, but we are still talking about light and the limitations of either film or sensors to record it. I have already found that some highlight detail could not be recovered from some exposures in the camera. So I think it would be useful to know how much over/underexposure the sensor could take in any given lighting and ISO sensitivity the camera has set.

With jpg it is easy. What you see is what you get. But with RAW there are more possibilities. Knowing before hand how to interpret the histogram for how to get the best RAW exposure would help me a lot. Most of my outdoor photos are while on vacation and with friends, where there is less time for bracketing and experimentation.

[funktionsfrei]

Underexpose. Don't overexpose.

With overexposure, highlights are blown out, regardless of jpeg or raw. With underexposure, shadows go down the drain, but that is with jpeg mostly. With raw you can adjust exposure in PP for several f-stops without noticeable loss of detail. The dynamic range of the sensor is much higher than what a jpeg file can transport, so creating jpeg in-camera invariably destroys the chance of repairing things in PP.

So, avoid overexposure by all means, and use either raw or raw+ (if you really need those lossy jpeg files for quick review or whatever).

In LV, the camera can be set to highlight potentially blown-out highlights, so use that and adjust the exposure to get no blown-out highlights. If that's to much of a hassle - you can quite generously underexpose as a general setting. One or two f-stops are no problem at all, and it is really amazing what these camera sensors can deliver.

For Pentax, this started with the K-5, and the K-1 is quite an improvement on that.

[Tas]

Knowing before hand how to interpret the histogram for how to get the best RAW exposure would help me a lot.

I regularly find I change the exposure form the cameras zero exposure setting by dialing in exposure compensation. Most of the time it is perfectly okay to leave it, however as I mentioned earlier if I can capture a brighter scene I prefer to do this in camera and reduce exposure in PP if needed. Other times it does seem the camera is conservative with the exposure, however I often forget to change the exposure metering and leave the camera set to centre weighted and some of my adjustments may be uncessary if I changed it to allow the camera to meter the entire scene.

Typically once I"m on location I prefer to do some testing with the lens I'm using at the time to know whether I'm likely to be clipping highlights. I recheck exposure with lens changes too as I like to shoot primes and the field of view change might introduce a bright spot or even permit some lattitude to increase exposure compensation. I set the highlight warnings as a default and avoid relying on the main histogram, other than as a quick indicator of how much headroom I have before I lose highlight info.

I also prefer to use the RGB histogram as a more reliable indicator of whether I'm clipping any highlights. I've found from time to time the main histogram is close to correct but when I cycle over to the RGB histogram I might be clipping in one colour and the averaging in the main histogram has masked this. I then rely on managing the colour that is producing the peak highlights and from this avoid clipping.

If you go through the menus you should find some of the correction features are on by default including highlight correction. I've not noticed a lot of difference between them being on or off but it could be a case of not shooting in conditions where they would be useful to have on.

One other thing to be aware of is a characteristic of using the screen illumination option to increase or decrease the screen brightness for sunny or night conditions. This is a really handy tool so I use it regularly but you need to rely on the histograms more when using it to manage the exposure. If for example you're shooting in sunny conditions you might add +1 or + 2 brightness to the screen just to be able to see what you're looking at. When doing so the image review will also be brightened so you can see it and it's appearance would indicate the exposure is fine until you check the histogram and potentially find it under exposed. Alternatively it might look like the scene is too bright and clipping highlights. The only reliable way of knowing how well exposed the image is comes from relying on the histogram, again the RGB histogram is a better indicator of what's really happening in the scene. You can also return the screen brightness to normal and anything underexposed will become more obvious though this is more helpful when shooting at night than on really bright days.

Apologies for the novel, hopefully it's providing you some useful feedback.

Tas

[DSLRnovice]

Thanks for the tip on using RGB histogram. I have noticed in DCU later that one of the bands in the histogram was maxed out.

I regularly find I change the exposure form the cameras zero exposure setting by dialing in exposure compensation. Most of the time it is perfectly okay to leave it, however as I mentioned earlier if I can capture a brighter scene I prefer to do this in camera and reduce exposure in PP if needed. Other times it does seem the camera is conservative with the exposure, however I often forget to change the exposure metering and leave the camera set to centre weighted and some of my adjustments may be uncessary if I changed it to allow the camera to meter the entire scene.

Typically once I"m on location I prefer to do some testing with the lens I'm using at the time to know whether I'm likely to be clipping highlights. I recheck exposure with lens changes too as I like to shoot primes and the field of view change might introduce a bright spot or even permit some lattitude to increase exposure compensation. I set the highlight warnings as a default and avoid relying on the main histogram, other than as a quick indicator of how much headroom I have before I lose highlight info.

I also prefer to use the RGB histogram as a more reliable indicator of whether I'm clipping any highlights. I've found from time to time the main histogram is close to correct but when I cycle over to the RGB histogram I might be clipping in one colour and the averaging in the main histogram has masked this. I then rely on managing the colour that is producing the peak highlights and from this avoid clipping.

If you go through the menus you should find some of the correction features are on by default including highlight correction. I've not noticed a lot of difference between them being on or off but it could be a case of not shooting in conditions where they would be useful to have on.

One other thing to be aware of is a characteristic of using the screen illumination option to increase or decrease the screen brightness for sunny or night conditions. This is a really handy tool so I use it regularly but you need to rely on the histograms more when using it to manage the exposure. If for example you're shooting in sunny conditions you might add +1 or + 2 brightness to the screen just to be able to see what you're looking at. When doing so the image review will also be brightened so you can see it and it's appearance would indicate the exposure is fine until you check the histogram and potentially find it under exposed. Alternatively it might look like the scene is too bright and clipping highlights. The only reliable way of knowing how well exposed the image is comes from relying on the histogram, again the RGB histogram is a better indicator of what's really happening in the scene. You can also return the screen brightness to normal and anything underexposed will become more obvious though this is more helpful when shooting at night than on really bright days.

Apologies for the novel, hopefully it's providing you some useful feedback.

Tas

---------- Post added 04-11-2017 at 06:05 AM ----------

Underexpose. Don't overexpose.

With overexposure, highlights are blown out, regardless of jpeg or raw. With underexposure, shadows go down the drain, but that is with jpeg mostly. With raw you can adjust exposure in PP for several f-stops without noticeable loss of detail. The dynamic range of the sensor is much higher than what a jpeg file can transport, so creating jpeg in-camera invariably destroys the chance of repairing things in PP.

So, avoid overexposure by all means, and use either raw or raw+ (if you really need those lossy jpeg files for quick review or whatever).

In LV, the camera can be set to highlight potentially blown-out highlights, so use that and adjust the exposure to get no blown-out highlights. If that's to much of a hassle - you can quite generously underexpose as a general setting. One or two f-stops are no problem at all, and it is really amazing what these camera sensors can deliver.

For Pentax, this started with the K-5, and the K-1 is quite an improvement on that.

Agreed.

One thing I have noticed is that the camera is biased towards underexposure based on the scene. But that bias does not always appear because the software in the camera is correcting the preview. In Raw Therapee there is a neutral profile that takes the exposure back to the uncorrected RAW image.

Since I often used manual lenses, I cannot trust the camera's meter completely. It helps to know to expose for the highlights as in slide film. Overexposed white clouds or highlights in waterfalls definitely do not correct well!