[Giklab]

I wish it was just that simple... although the result seems to indicate that it is underexposed. I tend to think it changes the tone curve and/or histogram - there must be some sort algorithm that Pentax uses to achieve the result.

I managed to almost match the results of in-camera HC JPEGs using an underexposed raw with PDCU - pushing by 1 stop and using the highlight compensation sliders. I'll try with Lightroom today.

[aleonx3]

I managed to almost match the results of in-camera HC JPEGs using an underexposed raw with PDCU - pushing by 1 stop and using the highlight compensation sliders. I'll try with Lightroom today.

If it were that simple, that would be a useless feature (as most Canikony users will ridicule) since one can always use EV compensation to adjust for that. What I am saying is that it is more than just underexpose in metering the scene... and something else. Hope someone who knows will chime in...

[Giklab]

If it were that simple, that would be a useless feature (as most Canikony users will ridicule) since one can always use EV compensation to adjust for that. What I am saying is that it is more than just underexpose in metering the scene... and something else. Hope someone who knows will chime in...

Whatever the background processing is or isn't, I wouldn't dream od calling the feature useless. For example in JPEGs this combined with Shadow Compensation extends the DR by a big fat 1-2 stops depending on the scene.

But what do you think is happening in the background? Maybe this is the cause of the occasional differences between JPEGs and raws! (Or more accurately, between JPEGs and raw preview JPEGs.)

[mcgregni]

The underexposure theory is interesting and worth testing for ... But its seems too limited to explain everything. I mean, why would 1 stop under deal with all blown highlights? I know plenty of my shots where I've used minus 1 ec and still had white clipping to a large extent, on bright clouds in particular.

But in similar scenes when shooting jpegs with HC on the camera processor has done a great job on the blown bits without any apparent under -exposure on the main parts of the image.

[aleonx3]

Whatever the background processing is or isn't, I wouldn't dream od calling the feature useless. For example in JPEGs this combined with Shadow Compensation extends the DR by a big fat 1-2 stops depending on the scene.

But what do you think is happening in the background? Maybe this is the cause of the occasional differences between JPEGs and raws! (Or more accurately, between JPEGs and raw preview JPEGs.)

Ok, I don't have a good explanation as I do not know the details. What I am saying is by turning both HC and SC ON is 'not' equivalent to keeping it OFF and simply use -1EV. The results from both settings would be different in JPEG and RAW. Sorry, someone who knows better can chime in, it is beyond me.

[mcgregni]

I believe that what you get after activating HC &/or SC in a camera jpeg is the result of image processing carried out on those brightest & darkest parts of the image, recovering what's possible by DR compression techniques. Obviously 1stop underexposure is not going to help recover shadows, but whether its part of the highlight recovery process....well, its a good theory that needs testing. So far what I've seen described here sounds more like the normal brightness difference between un-adjusted raw and the matching processed jpeg ... ?

[ScooterMaxi Jim]

Having done the tests extensively, aleon is closest to nailing the answer. First, the answer is absolutely not the same for highlight compensation, as opposed to shadow compensation. Shadow compensation has no effect on a RAW image, but will lift shadows on a JPEG (and you see added noise). Highlight compensation will have the same impact on RAW when using a proper converter (LR, Capture One) as what the camera processes into JPEG. The problem with not setting HC is that you really never know when it might be best to guard against running out of highlight dynamic range. If you really don't care about blown highlights, don't bother to test out your camera's ability to extend DR and be a "purist."

Unlike the CCD sensors (6, and 10 mp sensors), the CMOS sensors have virtually no roll off. They have a native ISO of 200 and will yield the greatest DR at that setting. Pentax implements the roll-off for HC at least as well as I can compensate for it in RAW post process. In HC-auto, the camera is very good at sensing when to implement HC to prevent blow-outs. You still will get the marginally better ISO 100 performance if DR isn't needed. I strongly recommend the HC auto setting. As a RAW+ shooter, I don't use shadow compensation, but might feel differently if I was relying only on JPEG.

It is one thing to accept what others have repeatedly stated on the forums, and quite another to learn how the camera and software actually operates by systematically testing out the comparison. Until someone actually shows that HC auto is somehow inferior to manually reducing exposure and raising ISO accordingly, I will continue to repeat what I have learned by actually performing the tests. The K-30 almost never will blow highlights with HC auto, yet holds to ISO 100 in the vast majority of high contrast scenes. The metering system is extremely trustworthy in that regard. Not so good, though, if you prevent the camera from performing at its best by shutting out HC entirely.

[TheGoldbloom]

A convenient way to see how camera settings affect the RAW data is to use RawDigger. Shooting the same scene, varying the camera settings, and bringing the resulting RAW files into RawDigger, it is possible to compare real RAW histograms and see if the particular camera setting affects the RAW data. That's what I do with every new camera.

[Mattco26]

Having done the tests extensively, aleon is closest to nailing the answer. First, the answer is absolutely not the same for highlight compensation, as opposed to shadow compensation. Shadow compensation has no effect on a RAW image, but will lift shadows on a JPEG (and you see added noise). Highlight compensation will have the same impact on RAW when using a proper converter (LR, Capture One) as what the camera processes into JPEG. The problem with not setting HC is that you really never know when it might be best to guard against running out of highlight dynamic range. If you really don't care about blown highlights, don't bother to test out your camera's ability to extend DR and be a "purist."

Unlike the CCD sensors (6, and 10 mp sensors), the CMOS sensors have virtually no roll off. They have a native ISO of 200 and will yield the greatest DR at that setting. Pentax implements the roll-off for HC at least as well as I can compensate for it in RAW post process. In HC-auto, the camera is very good at sensing when to implement HC to prevent blow-outs. You still will get the marginally better ISO 100 performance if DR isn't needed. I strongly recommend the HC auto setting. As a RAW+ shooter, I don't use shadow compensation, but might feel differently if I was relying only on JPEG.

It is one thing to accept what others have repeatedly stated on the forums, and quite another to learn how the camera and software actually operates by systematically testing out the comparison. Until someone actually shows that HC auto is somehow inferior to manually reducing exposure and raising ISO accordingly, I will continue to repeat what I have learned by actually performing the tests. The K-30 almost never will blow highlights with HC auto, yet holds to ISO 100 in the vast majority of high contrast scenes. The metering system is extremely trustworthy in that regard. Not so good, though, if you prevent the camera from performing at its best by shutting out HC entirely.

This has been my experience as well with the K-01.

[johngs]

I've always wondered if there's someone at Ricoh/Pentax who can be asked for a definitive answer to questions like this?

We can guess and theorize and test as much as we like, but surely it would be better to get an answer from whoever at Ricoh/Pentax designed and engineered the camera?

[ScooterMaxi Jim]

I've always wondered if there's someone at Ricoh/Pentax who can be asked for a definitive answer to questions like this?

We can guess and theorize and test as much as we like, but surely it would be better to get an answer from whoever at Ricoh/Pentax designed and engineered the camera?

The reason they don't tend to specify in the manual, or comment further is that the camera is always subject to some tweaking through firmware. HC is simply a marker - a profile to be applied by the RAW converter that happens to provide an excellent roll-off, broadening dynamic range at the cost of some minor shadow detail (at least at low ISO). I have no idea why a shadow compensation marker is not applied in RAW - possibly a design decision recognizing that the challenges of getting shadow right varies greatly with the situation. In any event, these applications are pretty obvious with even rudimentary testing and viewing the histograms on a good converter or image viewer.

[rp_dxn]

well that's surprising. I have two K30 bodies and i set them both with Shadow Correction turned ON and Highlight Correction turned OFF. I mostly shoot raw, and i still Shadow Correction is turned ON. There is a noticable difference even on raw file. That's what i observed.

[ScooterMaxi Jim]

well that's surprising. I have two K30 bodies and i set them both with Shadow Correction turned ON and Highlight Correction turned OFF. I mostly shoot raw, and i still Shadow Correction is turned ON. There is a noticable difference even on raw file. That's what i observed.

That's entirely possible with the K30. If the converter gets direction (a marker) to lift the shadows, then you will see a simulation to match the JPEG. Keep in mind that the camera is creating a JPEG thumbnail even if you are shooting only RAW. The question is whether a particular converter implements the marker from RAW.

I know that my own testing was done at the time of the K20D. When Shadow Correction was added to the K-7, those RAW files did not get altered for the shadows. Frankly, I'm not that interested in the shadow side of it because on a RAW file the shadow information is always going to be there, and conditions result in a lot of variation as to how you would want to do contoured shadow lifting. However, I find that the profile to preserve highlights (most often in sky or skin in sunlight) is going yield a pretty similar roll-off pretty much every time - and the Pentax HC marker is a very good one. Of course, with overexposure you can't get the information back.

Just curious why you would want Shadow Correction rather than Highlight Correction - especially in RAW shooting. If you take a look at any comparative review - Pentax with their 16 mp sensors are already better than competitors in handling the shadows without compensation, but the problem is in uncorrected highlights - headroom is limited. The K30 especially is prone to exposing just a bit hot in high dynamic range scenes. I find auto-HC to be very effective and gets utilized only when absolutely necessary.

[Lowell Goudge]

Hi guys I got a K50 a few days ago.
I'm familiarizing with all it's features, I'm quite impressed with it's video quality.
Putting aside Image Stabilization. Talkin about highlight and shadow correction usage. Does it worth to have them on?
I think this wont help if we're looking for natural shots. I must say I couldn't test it out carefully yet. I think I'm going to do some portraits to compare.

What do you gusy think?

Thank you in advance

The best way to test is with a scene shot without protection using the entire histogram

Shots must be in JPEG.

Meter using matrix metering on the scene with neutral settings for saturation, contrast etc in manual mode with the green button. ISO must be a minimum of 2x the cameras minimum ISO . Take one shot with corrections off, another with corrections on, without changing the shutter , aperture or ISO . Best to be tripod mounted to ensure exactly same image etc...

In a photo editor, that can measure greyscale and histogram on a selection (as opposed to the whole frame) select specific spots with different exposure in the original frame and then measure the same points exactly in the one using protection. You will likely see up to 45 greyscale value between the darkest and latest points of the pair of images representing at least 1 stop of additional dynamic range for both the light (half above 120 ) and dark (half below 120) parts of the total histogram.

If memory serves me correctly, neutral points ~120 greyscale remain neutral

What actually happens is the middle portion of the histogram is compressed from 45 to about 30 greyscale value per stop, reducing contrast, increasing the apparent dynamic range of the shot, but, note this can also lead to color banding if you make further adjustments to the image with a JPEG editor.

Sometimes it works sometimes not, depending upon the range of colours especially clear blue skies with small changes but finite across large areas

[ScooterMaxi Jim]

It really depends a lot on subject, RAW vs. JPEG shooting, and the kind of look you want. The problem with shadow correction is it can give you HDR look - compressed and a bit unnatural. If you use both SC and HC, you'll limit the effectiveness of both. Without trying to steer you, the K30/50 tends to run hot as has been indicated in reviews - especially in JPEG mode. That means you tend to run out of recordable information pretty easily. That's why HC or the conservatively implemented auto-HC can be very helpful - implemented only in high DR situations. Some folks here will tell you that no compensation is the most natural - and that's true until you have blown highlights and no way to recover them.

I'm yet to hear from anyone who has given HC auto a fair trial, and then goes back to no compensation. Still, plenty of folks say don't try it, though.