[omega leader]

Strange I use spot for most of my images, I find I like the subject to be properly exposed and screw the rest.

[RiceHigh]

I know you wont like it or believe it Michael, but I have seen under exposed shots from the Canon 5d, which as I said, is probably more of a reflection on the operator than the camera.

So, that what you saw but not what you used.

But at the same time I also note that the 5d is a $5000 camera here in Australia, some 3 1/2 times the price of the K20D so any comparison lacks credibility from the canon viewpoint.

How about the 300D then?

Also if you took the time to actually read a few posts here and other places you would find contributions from people that own Pentax and Canon or Nikon and report exposure & AF issues with the other brands.

Yes, but the problem is not as common as Pentax's and as serious as Pentax, frankly. It should be noted that how many Pentax cameras and lenses are sold and how many of those reports have already been there.

The point: no one brand is perfect, each has it foibles, so move on please.

Yes, there is nothing perfect on this planet. But something is better and something are worse. I do hope Pentax should make again better DSLR bodies as they had some of the best glass ever made in this planet.

The Canon 5d was released in Aug of 2005.....who was it that said the K10D was obselete after 12 months?????....please remind me.

You're right that the K10D is "obsolete" but the case does not hold true for the 5D, which *is* still selling widely in many camera shops right now. Just go figure out yourself if you still don't believe.

[arbutusq]

I'm wondering if the problem (if there is one, which I am not convinced as I tend not to use EV and still get good results) is in the lenses. Here is a quick and dirty test (non scientific, I'll admit) comparing my 16-45 with my FA 28-70 f4.

Both images were shot on Manual at f8, 1/60 sec, 28 mm flISO 200,RAW converted with Lightroom with no PP.

First the 16-45 image


And the 28-70...


The 16-45 needed 2/3 EV boost to equal the 28-70. This has been borne out in other tests I have done.

[RiceHigh]

Both images were shot on Manual at f8, 1/60 sec, 28 mm flISO 200,RAW converted with Lightroom with no PP.

The 16-45 needed 2/3 EV boost to equal the 28-70. This has been borne out in other tests I have done.

True, as I have told above, the digital lenses are darker and tend to underexpose than the film ones. At least even Ned Burnnell's samples prove and shown the same.

So, it's just a design and/or QC problem afterall.

[arbutusq]

Actually RiceHigh my experience is that this leads to overexposure on the FA lensesas opposed to underexposure on the DA lenses. I frequently need to dial a negative EV with my FA lenses to avoid blown highlights.

The issue that my photos show is not the Pentax Exposure/metering system but the construction of the lenses. At f8 and 1/60s the 28-70 lets in more light than the 16-45. On my K100d this leads to generally accurate exposures with my DA lenses and blown highlights on my FA lenses.

[falconeye]

Both images were shot on Manual at f8, 1/60 sec, 28 mm flISO 200,RAW converted with Lightroom with no PP.

Interesting. The histograms do differ indeed.

It would be interesting to have versions with full EXIF to see the 16 metering values of both images. Either the metering differs (one of the two lenses has an incorrect max. aperture) or the aperture blades are somewhat lame or inaccurate on one.

You could try to shoot a series at constant EV but varying time/aperture to check that histograms indeed turn out identical for either lens.

[arbutusq]

I've noticed a variation between the light let in at identical values between most of my lenses, and I noticed it when shooting Nikon as well. I remember a thread on the Nikon Cafe forums where this was discussed and the conclusion was that the shape and number of aperture blades, the number of glass elements and the presence of ED glass could all lead to this kind of variation.

Since f number is focal length/aperture, having more/different glass elements would reduce light without reducing f number and thus exposure.

---

Last edited by arbutusq; 06-14-2008 at 09:33 AM.

[stevebrot]

Hi,
This thread is wonderfully entertaining!

I have been involved in photography for about 40 years and remember reading an article some time ago regarding Edward Weston and his approach to exposure. He would take his hand-held incident* meter, walk around the subject, take several reading from different angles, scratch his head, set aperture and shutter speed, and take the picture. The final combination of settings was ultimately based on his experience and artistic intent with a few hints supplied by the meter.

Now, maybe it is reasonable to expect the camera's electronics to provide the equivalent of Weston's experience and artistic eye and maybe it is not. And maybe Pentax (compared to other brands) does a worse job at this and maybe they do not. A quick survey of Canon/Nikon forums indicate a high degree of grumbling about exposure issues with those products as well. The fact remains that many (most?) subjects will include light values that challenge the dynamic range of even the best sensors or films.

Now, the old rule of thumb with film was to expose for the highlights while being sensitive to providing adequate exposure to maintain detail in the shadows. This often would result in a negative that would be somewhat underexposed overall unless you tweaked the film's response curve by using some form of "compensating" development to generate additional density in the thinner parts of the negative.

From what I can tell, the situation with the digital sensor is strictly analogous. The metering system makes its best guess for the highlights and mid-range values and we attempt to salvage the shadows in post-processing. The good news is that the modern sensor's low light sensitivity is far superior to the films I am familiar with.

My experience with my K10D is that the metering system generally takes the above approach. I seldom see blown-out highlights and usually am able to lift the shadows, if needed, by curve manipulation in post-processing. I have not seen any overall tendency to underexposure. This may be due to my habit of shooting primarily RAW and using Lightroom for RAW development. Those shooting JPEG (in-camera processing) or using a different RAW converter may see something completely different.

Steve

* Measures the light hitting the subject rather than the light reflected from it.

BTW...Point and shoot digital cameras do the curve compression/expansion for the user in the camera since the assumption is that the user is incapable of doing it for themselves. The question is...what sort of user are you?

[jeffkrol]

True, as I have told above, the digital lenses are darker and tend to underexpose than the film ones. At least even Ned Burnnell's samples prove and shown the same.

So, it's just a design and/or QC problem afterall.

False conclusion as usual:
A design choice maybe not "problem". QC well there could be variation but the odd thing is considering the "old kit" lens had a pretty consistent track record as to exposures I'd conclude it's not a QC problem just a design consideration.
Or conversly, the "old" lenses are not behaving to the "new" standards...

[jeffkrol]

I'm wondering if the problem (if there is one, which I am not convinced as I tend not to use EV and still get good results) is in the lenses. Here is a quick and dirty test (non scientific, I'll admit) comparing my 16-45 with my FA 28-70 f4.

Both images were shot on Manual at f8, 1/60 sec, 28 mm flISO 200,RAW converted with Lightroom with no PP.

The 16-45 needed 2/3 EV boost to equal the 28-70. This has been borne out in other tests I have done.

Actually comparing the histograms is a bit revealing:
Adding a 1/2 stop EC to the one would equal the other, though it seems a straight EC would only move that tighter blob to the right. They would still look different until you spread the histogram out a bit........

OR adjusting the contrast/gamma;

[stevebrot]

False conclusion as usual:
A design choice maybe not "problem". QC well there could be variation but the odd thing is considering the "old kit" lens had a pretty consistent track record as to exposures I'd conclude it's not a QC problem just a design consideration.
Or conversely, the "old" lenses are not behaving to the "new" standards...

The problem with the rice guy is that he speaks without knowledge regarding the basic nature of light, optics, and photographic media. Yes, he knows the currently available products, but from a marketing perspective. All this talk about "obsolete" makes me chuckle. My 40-year old film SLR takes pictures that are on a par with my K10D...it just takes more effort!

I suspect that his posts would be different if he were to spend some time with the old Ansel Adams photo books followed by some time shooting in manual mode with a hand-held meter followed by hours in the darkroom creating the final product.

Steve

BTW...have you ever seen some examples of ricehighs work? He has his pictures posted somewhere...probably worth taking a look to get some perspective...

Correction: I just took a gander at RiceHigh's homepage on geocities. Despite the content of most of his e-mails, he does have some knowledge and and some familiarity with dosimetry (surprise!), though I have doubts about the validity of his methodology and his conclusions regarding his *ist D.

More... Here is his image directory on geocities:

Index of /ricehigh/Gallery

[stevebrot]

Actually comparing the histograms is a bit revealing:

Wow!

Not only is there a left shift of the curve, but there is also a significant difference in contrast!

[arbutusq]

Very interesting. Here are the LR full color histograms.
16-45


28-70


It would appear that the lens elements exert a fairly large influence on the overall lightness and contrast of an image (kinda obvious though if you think about it) If I was the kind of guy who enjoyed doing tests i would probably be spending hours devising intricate methodologies to discover what is going on.

[monochrome]

Actually RiceHigh my experience is that this leads to overexposure on the FA lenses as opposed to underexposure on the DA lenses. I frequently need to dial a negative EV with my FA lenses to avoid blown highlights.

The issue that my photos show is not the Pentax Exposure/metering system but the construction of the lenses. At f8 and 1/60s the 28-70 lets in more light than the 16-45. On my K100d this leads to generally accurate exposures with my DA lenses and blown highlights on my FA lenses.

Ahah!

I made a couple hundred shots outdoors in Boca Raton last week at all times of day from blazing sun to rising/setting to overcast. Almost all were in verdant greenery of some interesting feature or landscape/trail, and every one had a bright highlight somewhere, whether it be an opening in the canopy or a bright walk with light hitting it or a bright item in the background or the petals of a white flower.

Every single image had blown highlights - nice, centered histogram except for a spike through the roof at the right or left margin (and blinkies in every preview).

All were A, F or FA lenses, usually in AV or M (I even tried G for the first time ever on this camera), always set to "A" and using the e-dial for aperture control. I messed with ISO - everything I could think of!

Digital newbie that I am I tried filters (CPL helped for full beach shots), changing metering patterns and even hand metering a gray card with a Studio Deluxe (I still think film) - until I despaired of getting it right.

For the Canikon fanboys lurking here I NEVER thought it was my K10D - I just gut frustrated by user error.

Now I know what to try in similar conditions!!! (How could I forget to "pull" the sensor?)

P.S. - RiceHigh is on my ignore list, so I didn't see his comments.

[EDIT] Next day - I just experimented with a K200/4 and a K50/1.4 (settings are irrelevant) shooting blooming Hosta (white flower stalk against deep green leaves in shade). By intentionally reducing exposure 2 steps I was able to eliminate overexposure.

[Big Dave]

Steve,
Could the difference be the result of the so called digital coatings? The image appears slightly brighter to the metering cells which are optomize for these. This would account for the slight increase in exposure with film FF lenses. I did a test with my Nikon 55mm micro and my kit lens at 5.6 and 55mm with available light. The Nikon 55mm did have slightly greater exposure. This is why I like shooting with manual lenses at the working appeture. This allows the use of AV mode and shows actual depth of field during shooting. The exposure compensation is available in AV mode, where it isn't in manual. Auto diaphram is better for low light flash work of course.

Dave