[ajuett]

I find the AF on the K100D Super just fine.

My kids are the fastest thing I photograph. I capture the kids running around all the time, some are good, some are blurry. Oh well, not even a D700 could 'predict' what a child is about to do or which way he/she will turn . It's all about timing, not FPS or AF speed.

[betaPhoto]

Exposure/Underexposure - k100d, Alfisti, Marc Sabatella and others interested

I'm not sure if all of the following is exactly accurate or based on provable science, but that would take months of research and a large sampling of product and testing equipment (money that Pentax has chosen to even skimp on (improved new AF system is to costly, so it will be left for a later time...)). That all being said, it doesn't disqualify what is being experienced first hand and noted second hand through others.

Many people have shared their underexposure experience and almost every time it is thrown out as inaccurate or untrue. k100d, Alfisti, RiceHigh (you may not like his negativity, but much of what he writes has some truth to it), a few others I have exchanged thoughts with, and myself including a series of about 4000 exposures in frustration of trying to figure out what is going on. To many spending that much time being frustrated over unpredictable exposure seems odd, but I've always been that way with gadgets I like. Mess with them long enough until you figure something out.

I believe there are two forms of underexposure taking place with some peoples equipment. The first type is accepted by most everyone and I will try to explain it below. The second type of underexposure will follow.

The first type is that Pentax tries to save blown highlights, or it's measured at 18% grey, the ISO standard, or switch the metering mode from matrix to center weighted to spot, or it's simply a matter of preference or maybe another idea or two.... Many Pentax users deal with this as it is and compensate as necessary, some even say it is better because it is predictable and allows for them to get the correct exposure and thus better art. To these users it's not a problem, it's just the nature of Pentax's exposure system.

I can agree with some of these ideas, but I believe other camera makers are not satisfied with this and are actively trying to improve the exposure system to compensate, so the user doesn't have to. Go to a camera store and pick up a Canon XTI or XSI and take a few shots, make sure your test shots are not just easy exposure with low light range (a range of only 5-7 f-stops). Take a few shots with a background window that is too bright or some other high dynamic range shots with uneven lighting (maybe 10-11 f-stop ranges). You will notice that almost every shot has the histogram exposed to the right and ending just before the histogram ends (a color channel may sometimes be clipped and that's what people point out when they say Canon clips highlights. If a desired highlight is clipped, you should compensate for it). Sometimes these shots are brighter than you like, but Canon knows that fully half of the sensors info collecting resides in the first and brightest f-stop range of the sensor (don't waste it or risk rising noise levels on the dark side of the exposure and decreased dynamic range), so they are trying to use as much of the sensor as they can (the Nikon D40 is quite good at this also). You can adjust some shots lower for artistic reasons with either exposure compensation (just like all cameras) or after the fact in PP. I don't know how they achieve the histogram exposed to the right so consistently, but in the 5-7 Canons I have tried they were all like this (they were recent models and I haven't tried a high end model yet, I need to do that). A few guesses would be 1) A ton of money and research and they have achieved exposing to the right incredibly well! End of story. 2) Pre-histogram guesses with the metering system. 3) mild (maybe up to 2/3 of a stop and unrecorded) ISO compensation during the shot, like the old TTL flash systems cutting off the flash when it metered enough light (incredible that it could do that) . 4) Pre-processing the data to push the histogram right to the edge before writing in Raw or Jpeg, kind of like the old method of "pushing" the film and telling your developer how much you pushed it so they can over develop the film (ideas 3 and 4 may or may not be the same thing). 5) Pentax has an underachieving exposure system and is making the other guys look good. Ouch!!! 6) Any other ideas???

The second type of underexposure is much more elusive and only strikes some cameras. The ones k100d, Alfisti, RiceHigh, others I have exchanged thoughts with, and myself have been unlucky enough to own. In addition to the first type of underexposure and the adjustments necessary, these cameras need additional exposure compensation because of poor quality control. This compensation is more erratic and unpredictable and sometimes exposes only 60% of the histogram. These shots are clearly well underexposed, and when these users tell about its consistent underexposure they are told it simply isn't true and the Pentax system is better than that. Well, I'm here to tell others it is poor metering and users with this problem learn to dial in +.7 compensation for most shots, but then for when you don't expect it you end up with overexposure or still underexposed. It's not very predictable and the only mild consistency is dimly lit shots are more underexposed than others (usually). The reasons for this I do not believe are as simple as poor calibration of the meter. It may have to do with poor calibration, or the metering components, or maybe even a poor sensor, or something else (k100d, you live in Canada and maybe have a longer warranty. Send it in! RiceHigh, with all your posts and gripes why didn't you figure this out?). I finally compared one camera to another and this was my conclusion. My *ist DS2 exposes with only the first type of exposure idiosyncrasies or benefits, depending on the way you use your camera and view such things.

I personally believe Canon has it right, consistently expose to the right and let the user apply negative compensation when they feel it is artistically correct. But, if Canon is doing the 3rd or 4th method of exposure compensation (mentioned above), please Pentax, don't do it. Just keep improving the exposure accuracy of the system and please admit that it could be improved!!!

And finally, respect that I am still just betaPhoto and really not sure about any of this???

[Marc Sabatella]

[QUOTE=Alfisti;509208]if you like compared to other cameras I have used the mirrir slap is a freakin huge kkkkkkeeeeerrrrrr clunk!!!!!
[/quot]

I agree - much louder. I just question the actual difference it makes. I think it incumbent on those who want to claim a specific disadvantage (eg, "one stop") to produce some numbers in evidence of this.

and Pentrax cameras produce an image that is darker than most of it's competitors

True - other cameras are more than happy to blow highlights. That doesn't make them right and Pentax wrong.

[Marc Sabatella]

Expose to the Right is always beneficial, the usage depends on whether your want to bother w/ it and sacrifice the shutter speed or not.

Well, exactly - that's why I didn't say it was *always* beneficial. It's only beneficial in cases where you don't mind the sacrifice in shutter speed or DOF and also don't mind the PP time. It's also sometimes not beneficial if the area you are most interested in *is* the highlights, as pushing them all the way to the right can result in diminished detail, as some sensors don't respond as well at the extreme as in the middle of their dynamic range.

In any case, my point remains: sure, ETTR is a neat idea, but it's *not* the proper standard for camera meters. If you wish to use that technique, it's up to *you* to take the necessary steps to do so. now, if Pentax were to offer an additional *option* to ETTR - including, perhaps, a parameter to control how much clipping you're willing to tolerate - that would be a fine enhancement. But it's silly to criticize Pentax for not doing this by default - it is *not* the proper default.

[Marc Sabatella]

To me these shots almost look like studio shots - except for the obvious action - I thought the light was that good....

I agree - excellent work! Of course, also contributing to the "studio" look is the background. Was that what was there, or is that the result of PP?

[UnknownVT]

Of course, also contributing to the "studio" look is the background. Was that what was there, or is that the result of PP?

Thanks Marc.

No post processing just as-is - other than fairly standard brightness/contrast and a little sharpening.

The stage used a white backdrop curtain which lights were shone on - that gave the background lighting - in the rest of that Album -
one can see in these shot1; shot2; shot3 and others of the full stage, give an idea of what the set up was like. These from the side of the stage side1; side2 - show that the "white" lights were mainly over-head - unlike most concert venue lighting where they are mostly in front pointing at the artists.

Actually I found these two shots (my "discards" and not meant to be posted) that hopefully show the set up better -

Both from the side of the stage -

this shows the back-drop curtain lit up red in this shot - I guess the (red) lights were from the top just in front of the back screen behind the baffles - see other pics of the full stage in the links above.

Even more intersting were the overhead lights -

I now see there are quite a few fluorescent panels - that may explain why AWB (Auto White Balance) worked so well for me last night - whereas actual Tungsten balance (which works better with traditional concert incandescent lighting) came out too blue/cool.

Interesting.... notice the metadata which I re-attached with PhotoME shows both these shots with White Balance Auto (Daylight) - whereas on the shots posted above from this set - the green background gets Auto (White Fluorescent) and the red background shots get Auto (Tungsten)....

This was one of the few cases where the K100D AWB performed better than any other set white balance I tried.

Thanks for asking - as it brought out things I would probably not have looked for.

[Class A]

so to capture as much information as possible, you want to overexpose to account for this. then decreasing the exposure in RAW and you'll get a better picture.

Neither you nor the luminous landscape article explain why that should be the case. Yes, there are more levels available for the brighter parts of an image but scaling an overexposed image down in luminance doesn't create more levels in the shadow areas. The technique compares favourably to pulling up shadow details of an underexposed / correctly exposed image, but you claim another advantage to which I see no explanation basis.

if you do this the reverse way of underexposing (expose to the left) and then increasing exposure, your picture will not lok as good.

Sure.

this only works in RAW because RAW is has all of the captured information. JPEG compression throws out a lot of this information and therefore does not have the same allowance for exposure adjustments.

You have more latitude for such correction when using RAW but there is nothing that does rule out the same corrections/benefits for JPEG in principle.

Expose to the Right is always beneficial, the usage depends on whether your want to bother w/ it and sacrifice the shutter speed or not.

I'm not convinced it is always beneficial. I believe it is only better than pulling brightness up.

Sometimes these shots are brighter than you like, but Canon knows that fully half of the sensors info collecting resides in the first and brightest f-stop range of the sensor (don't waste it or risk rising noise levels on the dark side of the exposure and decreased dynamic range), so they are trying to use as much of the sensor as they can (the Nikon D40 is quite good at this also).

As Marc said, ETTR is not the correct approach to metering. I wouldn't want to be forced to pull all these overexposed images down in post-processing. Also, I like the camera to err on the side of caution. Rather than trying ETTR and then sometimes miss by a bit, resulting in blown highlights that cannot be rescued anymore, it should rather meter a bit more conservatively. BTW, I really don't like how the D40 sample images at DPreview are exposed. The Pentax images retain a lot more detail in comparison.

The second type of underexposure is much more elusive and only strikes some cameras. The ones k100d, Alfisti, RiceHigh, others I have exchanged thoughts with, and myself have been unlucky enough to own.

I highly doubt that there is merit to this idea. All the people you mentioned take issue with several other issues of the K100D and/or Pentax cameras in general. I think it is much more likely that this is a critical and outspoken bunch of people rather than a group that happened to receive lemons.

[k100d]

i read the article i was convinced, if you're not, then no worries, i wouldn't think the differences are always visually apparent, but it is the best way the capture as much of the scene with the camera.

edit: it's not some stuff that reichmann made up, google it and you'll see lots of people using this techinque. again i must stress, read the article they address things like not blowing highlights, why you don't use JPEG, other people's comments, digital sensors.

[betaPhoto]

Originally Posted by betaPhoto -
Sometimes these shots are brighter than you like, but Canon knows that fully half of the sensors info collecting resides in the first and brightest f-stop range of the sensor (don't waste it or risk rising noise levels on the dark side of the exposure and decreased dynamic range), so they are trying to use as much of the sensor as they can (the Nikon D40 is quite good at this also).

As Marc said, ETTR is not the correct approach to metering. I wouldn't want to be forced to pull all these overexposed images down in post-processing. Also, I like the camera to err on the side of caution. Rather than trying ETTR and then sometimes miss by a bit, resulting in blown highlights that cannot be rescued anymore, it should rather meter a bit more conservatively. BTW, I really don't like how the D40 sample images at DPreview are exposed. The Pentax images retain a lot more detail in comparison.

If you look at how I wrote about the Canon results you'll see the word "sometimes", which I don't believe means "forced to pull all these overexposed images down in post-processing". If you go and try out one in the store you'll see that most of the images are ready to print without adjustment, some, but not all of them could benefit with negative adjustment. Also, many Pentax images if not compensated for beforehand will need positive adjustment. As you agreed in response to k100d "if you do this the reverse way of underexposing (expose to the left) and then increasing exposure, your picture will not look as good"

I highly doubt that there is merit to this idea. All the people you mentioned take issue with several other issues of the K100D and/or Pentax cameras in general. I think it is much more likely that this is a critical and outspoken bunch of people rather than a group that happened to receive lemons.

I am making assumptions about other peoples cameras because what they write is exactly what happened to me with poor exposure. When I read someone else's experience and it is nearly identical to mine I get the feeling this problem is more widespread than people believe. When I got a different camera the results were much better, just as I wrote "My *ist DS2 exposes with only the first type of exposure idiosyncrasies or benefits, depending on the way you use your camera and view such things".

[k100d]

I am making assumptions about other peoples cameras because what they write is exactly what happened to me with poor exposure. When I read someone else's experience and it is nearly identical to mine I get the feeling this problem is more widespread than people believe. When I got a different camera the results were much better, just as I wrote "My *ist DS2 exposes with only the first type of exposure idiosyncrasies or benefits, depending on the way you use your camera and view such things".

to be fair, i'm not ALWAYS at +0.7 EV if i'm taking a non-ETTR picture. bright sunny day the camera works fine for me.

[Alfisti]

[QUOTE=Marc Sabatella;509909]

if you like compared to other cameras I have used the mirrir slap is a freakin huge kkkkkkeeeeerrrrrr clunk!!!!!
[/quot]
True - other cameras are more than happy to blow highlights. That doesn't make them right and Pentax wrong.

I knew you'd respond with that and to a degree it is true, however I think it is the right approach as it results in better prints and better screen images. Unless you're prepared to select areas in pp and adjust exposure then you have TO blow some highlights to get an overall balanced image.

Also, i don't think it is just a case of 'they are more willing to blow highlights'. Tomorrow I will post an image of a black sock on a cream coloured carpet floor in dim lighting. I expose off the sock and yet the image is still severely under exposed, none of the data is in the last third of the histogram.

[k100d]

realized that we had totally diverted from Richard's post of contentment, i initially only wanted to offer some counterpoints.
believe it or not, i maintain the K100D user group, and you are all welcome to join. PentaxForums.com - K100D Series Users

one photo and i'm done

[betaPhoto]

Originally Posted by betaPhoto
I am making assumptions about other peoples cameras because what they write is exactly what happened to me with poor exposure. When I read someone else's experience and it is nearly identical to mine I get the feeling this problem is more widespread than people believe. When I got a different camera the results were much better, just as I wrote "My *ist DS2 exposes with only the first type of exposure idiosyncrasies or benefits, depending on the way you use your camera and view such things".

to be fair, I'm not ALWAYS at +0.7 EV if i'm taking a non-ETTR picture. bright sunny day the camera works fine for me.

I agree, with my old camera, bright sunny outdoor shots were more likely to be correct or close enough to accept. [yours may benefit from warranty work?]

[Marc Sabatella]

Tomorrow I will post an image of a black sock on a cream coloured carpet floor in dim lighting. I expose off the sock and yet the image is still severely under exposed, none of the data is in the last third of the histogram.

Well, why would you expect it to be? A picture of a cream colored carpet should come out with everything a little left of center. Depending on how dominant the black sock is, that *might*push the histogram a bit to the right. But unless the picture is mostly sock with just a little carpet, I would be shocked if that would push the histogram so far to the right that anything made it to the right hand third.

[Deni]

The main gripes I have with the k100d is that some times it is very slow as I tend to shoot only raw, and exposure in broad day light is a nightmare.

Otherwise it is an excellent camera.

This photo was shot at iso 3200 then adjusted EV 1+ in lightroom. Adjusted the whitebalance. Noise reduction performed in Lighroom, Luminance 40 and Color 100.