Pentax Camera Image Controls vs Computer Editing

I'm finding these sort of threads can go their own mysterious way Andrew! I try to keep things on track as best I can, but usually the extra points are useful and add things into the mix.

The purpose of the 'input' / 'output' classification of camera controls is to lay down some firm guidelines, and help in how we consider post processing actions, and hopefully achieve better agreement on the appropriatness of specific types of 'photoshopping'. To me there is a clear division between those camera controls that operate physical camera functions, or 'burn' in certain characteristics ... (this would include ISO I'd say), and those that are effectively 'post processing' options, the sort of things that are the normal part of image adjustment (and these are in the main the custom image settings).

There's been some debate about white balance here .... the consensus seems to be that, at least for RAW files, it is a variable that can be changed in processing without quality loss ... therefore by my definition it is an 'output' control.

Now, as an example, an 'input' control (or its effects) could not be adjusted in a routine manner without quality loss ... this includes aperture, time value and ISO. These settings are really 'burned' in at the point of capture. You cannot change the aperture, or time value, or the ISO after the shot has been taken. You might mimic the effects of doing so by image manipulation, but thats unlikely to create results that are as good as if it had been done on the camera at capture ... thats the 'input' controls.

So my theory says that if we are dealing with something that was created by an input control, then its likely any 'photoshopping' to change it could more fairly be seen as less acceptable by most ... a manipulation that some might term 'cheating'

But, if what we are changing is the result of one of my 'output' controls, (eg a change to sharpness, or saturation), then that is something that is a normal and expected part of digital image processing, and should not be critisised as wrong in any way.

Everyone will still have their own opinions of course, but we have a clearer way of identifying the source of the effect that is being adjusted ... was it from a camera 'input' control ( in which case it ideally should have been set right at capture), or was it from an 'output' control (or its computer software equivilent), and therefore its is a personal matter of taste and judgement for the photographer.

Obviously much more to read here later, but I think you'll find the HDR option in camera is more of an 'input' control. The camera (at least the K7) takes multiple shots to combine rather than processing a single RAW file - which I know is possible for producing HDR with the K5.

I would say the AA simulation of the K3 & SR settings are also 'input' settings, and I suspect there are probably a few others that in some way effect the way the RAW file is created and not just the (imbedded or otherwise) JPEG conversion.

Hi petro, good to have you joining in The division I'm defining here between 'input' and 'output' is really based on whether the particular control or function is a 'physical' or optical function (like aperture, time value, focus) or a quality 'burned' in to the sensor data (like ISO) ... these are 'input' controls .... or if the feature is really a form of software processing that is applied to the image data immediately after it is recorded ... applied by the Prime Processor - these features will have their equivalents in computer software programmes ... so these 'output' features are all the things like saturation, hue, sharpness etc that we could equally adjust successfully using normal editing tools.

I included the camera HDR controls in the 'output' category because, as I see it, this too is a software process applied by the camera processing ... sure, the camera takes the bracketed exposures, and it does this according to the input controls set up by the user ... but the significant point is that the HDR feature merges these using software to produce a single JPEG. This process equally could be performed on a computer if we took the brackets manually. So within the parameters of my definitions here, HDR is an 'output' control.

The AA filter is an interesting one also! You're right that its effect would be really embedded into the image data, in much the same way as the effect of the ISO setting would be. It poses an interesting academic challenge though, because you could argue that the effect of the AA filter (ie the amount of blurring added or not added) could in fact be adjusted in post processing by additional fine sharpening or blurring filters. But, my theory goes that anything that could cause quality loss in post processing should best be left to the camera at capture ... so, thanks ... from now on the official designation for the AA filter controls are as 'inputs' !

Your original statement on this "Input Controls directly affect the 'physical' optical image capture" is wrong if you include ISO in your list of input controls. ISO does NOT affect the amount of light captured by the sensor - only aperture and shutter speed do that.

You are correct, of course that in current cameras, you cannot change the ISO value after image capture. This is, through, a pretty academic point as cameras having the Sony 16mpx sensor (e.g. Nikon D7000, pENTAX k500, k50, k30 and all the k5 series) are almost isoless. You can deliberately grossly underexpose in camera and then correct the exposure in PP with little noise penalty - which is equivalent to changing ISO after capture.

Earlier today, I tried an experiment with my K30. I took one correctly exposed shot at f/2.8 1/125sec ISO 3200 - and then a shot of the same image that was 5 stops under exposed, at f/2.8 1/125sec ISO 100. I pushed the exposure of the ISO 100 shot in Lightroom by 5 stops to produce an image of the same brightness. There was a (just) noticeable difference in noise levels between the two images - but that disappeared after noise reduction was applied in Lightroom. Adjusting exposure during RAW conversion produces a similar result to that which would have been achieved by adjusting ISO pre-capture.

Andrew

I'm looking at ISO as being a setting that 'burns in' certain IQ characteristics at point of capture - noise levels being the most obvious, but there are tonal and colour changes at higher ISO's also. I see it as in the same category as aperture, time value, focus and flash lighting because it has no direct post-capture adjustment (ie the sensor sensitivity cannot be changed for that particular image after its taken).

Although I accept that post-processing brightness adjustments can be very successful in creating a similar result to actual ISO differences shot by the camera, this depends heavily on the individual camera, and this is still an image manipulation that leads to some loss of quality - as you said it was necessary to apply secondary corrective processes .... not really ideal. So for this reason, by my definitions, it falls firmly within the 'input' category.

Did you see that thanks to petrochemist we now have the AA filter controls as 'inputs' also (although HDR remains 'output' ).

Your original definition of Input Control was: "1) Input Controls directly affect the 'physical' optical image capture ...how the camera optics perform and the way the sensor records that optical image data.".

The ISO setting has no effect on the amount of light captured by the sensor - so you cannot use your original definition and regard ISO as an input control.

Andrew

One has to setup (as in pick one of the HDR profiles) pre taking the shot when you do HDR in camera. It produces a JPG only (not RAW), so the end result is well and truly baked in when the final HDR JPG is created. Plus the camera doesn't provide you with the 3 different exposures as individual files either when you produce the HDR in camera, so you can't manipulate the HDR level etc after the shot has been taken. It's not like sharpness or other JPEG creation tweak s etc.

Does anyone ever really use HDR in camera?

It is rare with the K30 (or any camera with the 16mpx Sony sensor) that you'd really need to worry much about high contrast scenes. Shooting in RAW, spot metering the brightest highlight and opening up 2.5 stops, followed by "normalising" the exposure in Lightroom usually does the trick without excessive noise.

In very rare extreme circumstances, the scene's dynamic range might be too high, even for that approach - which is why I usually carry some ND grads around with me.

For scenes where using ND grads is not possible, e.g. taking a photo from inside a cave but near the cave exit where a sunlit area is present in frame, taking two shots and merging them in PP works well and avoids having to use HDR software. I am well aware that there are HDR evangelists around who seem to want to generate an HDR "effect" as the end result - no doubt that can be achieved in camera.

Andrew

If we want to do HDR Andrew, then the best way is make multiple exposures and put them through a program such as Photomatix. The in-camera effect is a bit weak really, IMHO.

HDR is often used for excessive effects, but for severe dynamic range, such as church interiors, it can be used to deliver impressive, natural looking images with full detail throughout.

Interesting comment!

Do you remember this thread?

The concensus conclusion, which included the HDR evangelist who authored the current thread, was that HDR could not deliver as good a result as merging two images that included the whole dynamic range ,manually, in Photoshop (or Elements 8 in my case).

I have no doubt that Photomatix can produce a result that, if viewed in in isolation can appear realistic. The church windows thread is the only time I have seen a direct A-B comparison between images produced from the same RAWs with and without HDR software. In that thread, Photomatix came up short.

Have you got or are you aware of any other direct A-B comparisons between post processing methodologies where HDR software produces the better result? If you are aware of such a comparison, then maybe I should have another look at HDR, but for now it definitely stays on the back burner.

Andrew

Have you tried this Andrew https://www.pentaxuser.com/forum/topic/newcastle------again-----51425#comm542413

I think it should be used in moderation Andrew, but any HDR program can be used in a subtle way.

I use Photomatix to make HDR images, but not often, only when the subject suits it. I also use it to make black and white conversions where I'm looking for dramatic, high contrast results.

For me, an HDR image is made up of seven exposures all 1 stop apart, so -3, -2, -1, 0, +1, +2 and +3.

The images in that thread use HDR and look very nice - but, without a direct A-B comparison between the HDR version and one that was optimally processed, but without HDR, it is not possible to draw conclusions about the value of the HDR processing.

Andrew

As obviously you will know, mono images whether prints or digital images do generally need to be dramatic and high contrast to do well in club comps. I haven't tried that - but maybe should.

Andrew

Happy for you to leave it on the back burner then