[Patricia1066]

Hi I have been taking a photography course, and am discovering manual controls of the camera K5ii

The "ISOless" sensor - PentaxForums.com

This discussion in 2011 which refers to research done in 2010 has been interesting, particularly to someone currently engaged in testing the device settings. My tests and others I have seen show real world no downside to using ISO up to 6400.

I would like to check if I understand what I read, and is it possible to create a user setting that ignores the exposure rating of the camera and provides a jpg that shows what pp can make of the RAW file taken at 100?

Original film photography used film emulsions which used a trade off between grain and details, so a faster shutter speed on ASA 400 has higher levels of grain, better exposure value than ASA100 film. A digital sensor counts photons. Changing ISO sensitivity on a DSLR is amplifying that light, after capture and conversion to an electrical signal.

As this can be done in post production, the exposure triangle theory that was inherited from the film exposure needs rework. The Pentax K5ii firmware is based on the exposure triangle theory, and with Auto ISO range set, will increase amplification even though it will lose detail (increase noise) by so doing. At least I can set a range from 80-6400 on auto-ISO.

Are there any benefits to high ISO like 51200? The jpg shown on the LCD will have a good exposure after the ISO has been increased, though this has poor dynamic range by comparison to the RAW file.

[Ian Stuart Forsyth]

This discussion in 2011 which refers to research done in 2010 has been interesting, particularly to someone currently engaged in testing the device settings. My tests and others I have seen show real world no downside to using ISO up to 6400.

There is a downside as to how much light you can store in the sensor, above iso 1000 the camera really becomes isoless and increasing above that really would not give you any benefit other than showing you an image you can review on your LCD

I would like to check if I understand what I read, and is it possible to create a user setting that ignores the exposure rating of the camera and provides a jpg that shows what pp can make of the RAW file taken at 100?

Nothing really as of yet, it would also be nice if there was a UniWB, this way you can see how each channel saturates differently depending on the light source.

There is also the problem of findings the settings with the raw file converter that would truly show you what is contained within the raw file. If you are using abode you need t find out the baseline conversion for that camera


https://photographylife.com/where-are-my-mid-tones-baseline-exposure-compensation

There is also the problem with how WB of a camera works, what it does is apply a multiply to the different color channels of the sensor, to see how this data would look in the raw file and its histogram you would need to set the WB to a UniWB

http://www.dslrbodies.com/cameras/camera-faq/what-is-uniwb.html

combined with overriding the presets of the raw converter then you can better understand what data is found in the raw file. At this time I would recommend using something like fastrawveiwer.
This will better help you to understand what is contained with in the raw file, where the camera clips data in the raw file and how the color channels saturate under different light sources .

will increase amplification even though it will lose detail (increase noise) by so doing.

Raising the iso decreases the noise that the camera produces, its the decease in the exposure to accommodate the increase in iso that causes noise

http://www.photonstophotos.net/Charts/RN_e.htm#Pentax%20K-5%20IIs_14

Are there any benefits to high ISO like 51200?

not really if you are shooting raw, and if you are shooting raw with the k5II really shooting above iso 1000 really doesn't help you much and can cause you the greater risk of clipping data if you don't set your exposure correctly up to the clipping point.

[/URL]

[victormeldrew]

Are there any benefits to high ISO like 51200?

Yes. Where there is insufficient light to otherwise get an acceptable exposure and you really must get a (any) shot.

[interested_observer]

I dabble in a bit of astrophotography - well the Milky Way over landscapes. High ISO just reduces dynamic range, clipping more stars and losing their color.

Are there any benefits to high ISO like 51200?

I use 51200 to check my framing / composition, focus, levelness of the camera, etc., a test shot/throwaway - then turn the ISO back down for the real shot.

[Patricia1066]

Thanks Stuart, a reply that gives me plenty to think on.

I'll certainly investigate fastrawviewer as it seems like a good tool for RAW, and has a RAW histogram. I didn't realize that our amazing cameras are interpreting jpg not RAW when giving preview and playback views.

"Raising the ISO decreases the noise that the camera produces, its the decrease in the exposure to accommodate the increase in iso that causes noise."

I have a question about this statement; i understand that within the native ISO range of the camera, raising the ISO improves the exposure, depending on the conditions.

By Decrease in exposure, do you mean -1EV? This creates noise?

[pres589]

You can always set the camera to TAv mode and have the auto-ISO limited to 6400 or 3200. I believe the camera will only output RAW files in TAv mode and you're probably better off learning how to edit RAW's anyway.

What I personally do in low light with my K-5 II is run in manual mode, know about how steady I can hand-hold without showing motion blur induced by my own unsteadiness and use that as my exposure time, know how the lens in question performs, and then set ISO high enough to decently capture the image. I shot a lot of photos of bands performing in a small bar over the years (which I miss doing, but that's another story) and figured out that with my DA 50 1.8, wide open at 1.8 was acceptably sharp for the job, I could usually hand hold down to 1/15th of a second (although keepers were more common at 1/30th), and ISO 1600 was bright enough. ISO 3200 isn't what I would call "noiseless" on my K-5 II but it can be useable. Some folks on this forum have blown me away with their ability to kill the noise while maintaining the deal in post at such high ISO's.

[dms]

The K-5 will produce the widest DR at lowest iso, so of course the lowest iso is best as regards DR, if you can manage the stop and exposure time.

But if you are limited in f-stop and exposure time--e.g., theater photographs and you cannot open more than f/2.8 and to avoid too much motion blur you want to stay above 1/90 sec, you have extremes of choices (if using raw):
-- use iso 1600 and risk blowing out high lights
-- use iso iso 80 and have 4+ stops headroom in highlights, and then increase exposure in ACR (or whatever raw pp software you use)
-- use iso 400 and have 2 stops of headroom in highlights, and then increase exposure in ACR

They all will give same results, except margin to blowing highlights, although iso 400 will let you better use the camera display, and 4+ stops exposure compensation (w/ iso 80) lkely will be beyond your softwares options. So lowering iso has a huge benefit, if the camera has a DR that drops I stop for every stop increase in iso, although there may be other practical limits.

BTW easy to verify this, take two shots at different iso and same f/stop and same exposure time, and compare them. I did with K-5, K-x, K-20d, at iso 1600 and iso 100; and see no difference for K-5, some difference for K-x, and large difference for K-20d--thus if I am limited in exposure--I use lowest practical iso for K-5, iso 400 for K-x, and iso 1600 for K-20d.

Sorry, I just realized you are talking about using capture in jpg (which makes little sense in this scenario), but anyway in this case isoless, is not AFAIK, a useful idea. But you have a dslr, and can check it yourself.

[GUB]

I have been utilising Iso invariance by staying at base Iso since about the time I started this thread. Iso invariance - PentaxForums.com
I have not found a way to see the corrected results in camera and I think the only way to utilise it is via raw.

[Patricia1066]

@dms I don't take jpg images, I was wondering how the jpg that is embedded in dng could be tweaked to more closely match the profile of the dng.

I will try the settings you mentioned.

@gub ISO invariance is another great thread, thanks for the link.

The most useful statement is that K5 is isoless above 1000, as this gives me more latitude on taking a camera approved exposure. I don't like having dark images on LCD, or having pp on every photo.

I found a tip on processing of underexposed dng files in Photoshop. I haven't tried it yet, any other beginner tutorials out there?

https://photographylife.com/fixing-partly-underexposed-images-in-photoshop

[GUB]

The most useful statement is that K5 is isoless above 1000,

I treat my K01 and K1 as iso-less full stop.

or having pp on every photo.

The only gain you can get from utilising Iso-invariance is from the personalised modification of the tone curves in pp. Otherwise you may as well just leave it to the incamera Iso.

I don't like having dark images on LCD

Yes the only downside. I always have a lcd viewfinder mounted and I can easily see -3 EV shots. It would be good if you could take bracketed shots based on Iso.

[pres589]

This seems less useful than just letting the K-5 or whatever pick the proper ISO to expose the shot per normal user commands. Post-production workload seems largely the same, the camera's indications of exposure accuracy are no longer useful, etc.

One feature that Pentax DSLR's have over other brands seems to be low-light/high-ISO IQ. I think I would let the engineers that built my K-5 II call the shots. Where this might be more useful is a camera like my sister's Canon SL1 which has comparatively poor high-ISO noise/IQ issues assuming this "ISO-less" concept applies to that camera.

[Ian Stuart Forsyth]

Thanks Stuart, a reply that gives me plenty to think on.
I'll certainly investigate fastrawviewer as it seems like a good tool for RAW, and has a RAW histogram. I didn't realize that our amazing cameras are interpreting jpg not RAW when giving preview and playback views.
"Raising the ISO decreases the noise that the camera produces, its the decrease in the exposure to accommodate the increase in iso that causes noise."
I have a question about this statement; i understand that within the native ISO range of the camera, raising the ISO improves the exposure, depending on the conditions.
By Decrease in exposure, do you mean -1EV? This creates noise?

The noise is created from using a higher iso because that selection in that iso causes the camera to use a smaller exposure. You have to understand that the exposure is what happen in front of the sensor scene luminosity, shutter speed and aperture. The iso is basically what happens within the sensor and the processing done for that exposure.


When selecting a higher iso you are telling the camera to use a smaller exposure and to do this it needs to develop the image differently and also shift the cameras meter for the less light that is captured because of this smaller exposure. It is this smaller exposure that causes the increase in noise ( this is do to the noise found in the very light its self shot noise) that you see when increasing the iso, the camera on the other hand generates less noise ( read noise being the most prominent) as you increase the iso and with some cameras this can be very pronounce. With the K5 you will see some decrease in read noise that it produces from iso 100-1000, going above iso 1000 does not decrease the read noise.

[Patricia1066]

In very low light outdoors I contrasted two ISO levels, 51200 and 6400; aperture, shutter speed and flash were constant.


The high ISO picture had details in the far background, that were missing in the lower ISO picture, even after post processing to lift the shadows.


Indoors and in better light the low ISO has better results up to 3200.


I used the wrong ISO above - I was comparing ISO 51200 and 640, not 6400.

[dms]

Above iso 1600 apparently the K-5 uses additional modifications to the raw file (including noise reduction) w/ raw capture, thus the comparison should not go above 1600, as you are losing resolution above iso 1600. (e.g., see Photographic Dynamic Range versus ISO Setting)

And presumably you used camera raw?

I can tell you I performed the same test going from iso 100 to iso 1600, and the images were identical after raising the exposure in pp w/ bridge/adobe camera raw in photoshop.

[dms]

Here are two raw results at 100%--the first at iso 80, the second at iso 1600--same exposure for both ( same fstop and shutter speed)