[darylk]

Thanks Lou

I would hate if there was anything wrong with my DA*300 - that would be worse than a K3 malfunction! But - I am thinking - if the problem had been slow aperture blades, wouldn't my exif still look like for instance 1/2000s, f 5.6, ISO 640, but the actual exposure being longer due to the slow blades?

My raw images DO look overexposed when I load them into lightroom but, WOW, shooting in raw definitely saved me this time! Here is a shot post processed with -2,4 EV and a bit of Topaz clarity - this is quite keepable, actually! I never thought the highlights could be restored this well!

If your aperture blades are slow, your shutter speed and ISO shown in your EXIF would remain accurate. What would not be accurate would be the indicated aperture. I.e., if your EXIF says F5.6, the actual f-stop may have been F4 if the blades are stuck/slow.

You can put the aperture question to rest pretty quickly...take your lens in your hand with the back open and flick the aperture lever (the one sticking straight out the back of the lens) with your finger. If it seems sluggish returning to its original (closed) position, you have a problem with the lens. If not, you have other issues.

Also, if all your lenses are exhibiting the same characteristic, that would rule out the lens....you need to do some troubleshooting to narrow down the potential problem, then get back with us with your findings.

Daryl

If the aperture was the culprit you would have had one stop over exposure and that does not look to be the case. Are you sure you have selected Matrix metering? Have you got the focus point and metering connected or default, C-menu 1 item 5 set to 1 and do you have C-menu 1 item 6 set to 1? Are you sure your auto iso range is 100-3200?

Thinking about it I recognise this issue from my Tamron 17-50 with stuck aperture blades therefor the measuring was thrown off so I might have to take back my first statement. DO check to see if they open up as they should. Be sure to hold the lens in all directions, i.e up, down, level. My Tamron only did weird things at an upwards angle... And clean the contacs on the body and lens to make sure all elecrons have a safe passage between the two.

Also if you are using live view the metering may screw you over.

[TedW]

I do the birds and also use TAv which, as you have found, generally works very well. My only thought is that you can set the maximum range of the ISO and maybe the lower setting got somehow changed from 100 to some higher number and the camera was not able to bring the exposure down to a low enough level for proper exposure. In other words check the minimum ISO to make sure it is still at 100. I would at least give a quick check although you seem well versed enough I suspect you probably checked this already.

I do have to add that I have had rare occasions with high shutter speeds where I will have one of the series totally blown out but never more than one exposure and less than 5 times total in the 1+ year I have been using it (K50). I hope you find the issue soon. I know how frustrating it would be to lose a great photo in these situations.

[IchabodCrane]

I do the birds and also use TAv which, as you have found, generally works very well. My only thought is that you can set the maximum range of the ISO and maybe the lower setting got somehow changed from 100 to some higher number and the camera was not able to bring the exposure down to a low enough level for proper exposure. In other words check the minimum ISO to make sure it is still at 100. I would at least give a quick check although you seem well versed enough I suspect you probably checked this already.

I do have to add that I have had rare occasions with high shutter speeds where I will have one of the series totally blown out but never more than one exposure and less than 5 times total in the 1+ year I have been using it (K50). I hope you find the issue soon. I know how frustrating it would be to lose a great photo in these situations.

I've also seen situations where I must have done something to have taken ISO out of Auto and fixed it to some value. My guess is the OP may be experiencing this, too, as ISO 3200 is pretty high for bright lighting.

[stevebrot]

if the problem had been slow aperture blades, wouldn't my exif still look like for instance 1/2000s, f 5.6, ISO 640, but the actual exposure being longer due to the slow blades?

No, the exposure time would be the same, but the amount of light striking the sensor would be somewhat more than if the aperture had actually been at f/5.6 for the full time. Still, though, the posted photos are at least two stops overexposed. Is there any chance you could post the photos with full exif intact to someplace like Flickr and link the images to this thread? The images above have been munged by the forum software and the exif rewritten in an edited form.


Steve

[ScooterMaxi Jim]

For all the doubters, I can confirm that Mette is reporting a problem that most certainly does exist.

In TAv, the K-3 has a significant tendency to overexposure in a wide variety of situations - most of which are not terribly challenging for multi-meter mode. I don't know if the situation is getting worse, but it certainly isn't getting better with the firmware updates. Clearly, Pentax felt ETTR was the right call, but I disagree strongly in the age of modern sensors. The appropriate workaround is using Auto-highlight compensation, but this solution is not nearly as well-implemented as it was in the K-30 series. My preference is to go with -0.7 underexposure compensation most of the time.

For me, this is the most evident design flaw in the K-3. I'm amazed it hasn't been more fully discussed. If I was a JPEG user, I'd be outraged. As a RAW+ user who processes in Capture One v8 with its far-superior highlight compensation capabilities, it ends up being mostly a nuisance rather than an outright failure.

[stevebrot]

In TAv, the K-3 has a significant tendency to overexposure in a wide variety of situations - most of which are not terribly challenging for multi-meter mode. I don't know if the situation is getting worse, but it certainly isn't getting better with the firmware updates.

Is this something new? If not, this is first I have heard of it*. Forgive me for thinking that it does not make sense. TAv is supposed to be nothing more than juggling the exposure triangle. The metering should be the same regardless of exposure mode.

My thoughts after reading through the thread are:

• There definitely is something funny going on here
• "Forcing" ISO 3200 in TAv is not the cause. The camera sets the ISO based on the light measurement and the set shutter speed and aperture. The EV is the same regardless of the numbers used
• I have seen similar overexposure with lazy aperture, but the examples show more than 1 stop over
• The available exif are not adequate to rule out user error or to shed light on other reasonable cause

One thing that might be helpful would be to do an in-camera JPEG using the "natural" custom image setting. That would rule out the RAW converter import preset as the source of the lightness.


Steve

* FWIW, my experience with the K-3 metering is that it does not ETTR, rather the matrix algorithms tend to bias to underexposure by a stop or more for mixed lighting with a blank white wall metering squarely to the center of the histogram. My experience may be atypical, however, since I have disabled both highlight and shadow protection.

---

Last edited by stevebrot; 06-01-2015 at 07:46 PM.

[stevebrot]

FWIW, I did the calculation and 1/2000s, f/5.6, ISO 3200 = EV(100) 12

This is a perfectly reasonable meter reading for many daylight situations (bright overcast), though about 2-3 stops overexposure if the subject was in full noonday sun. Specular highlights in the grasses tend to indicate that the sun was probably high and bright (exif says 11:30 AM) which would indicate an exposure of about EV(100) 15. The fault is squarely on the exposure values.

Is there any chance the AE lock was on or linked to AF lock*?


Steve


* Menu --> C --> 1 --> 4 AE-L with AF Locked

[MetteHHH]

You guys are great, you really are. A big THANKS for all your suggestions - TAV might be one reason I am happy I chose Pentax, but this forum is definitely another one! I have gone through your suggestions below, and I can rule out most of them:

You can put the aperture question to rest pretty quickly...take your lens in your hand with the back open and flick the aperture lever (the one sticking straight out the back of the lens) with your finger. If it seems sluggish returning to its original (closed) position, you have a problem with the lens. If not, you have other issues.
Daryl

The aperture lever is snappy. Phew! I love my DA* 300mm better than any camera body I have had, and I would HATE to send that in for repair and be without it for 3+ weeks!

Are you sure you have selected Matrix metering? Have you got the focus point and metering connected or default, C-menu 1 item 5 set to 1 and do you have C-menu 1 item 6 set to 1? Are you sure your auto iso range is 100-3200?

Thinking about it I recognise this issue from my Tamron 17-50 with stuck aperture blades therefor the measuring was thrown off so I might have to take back my first statement. DO check to see if they open up as they should. Be sure to hold the lens in all directions, i.e up, down, level. My Tamron only did weird things at an upwards angle... And clean the contacs on the body and lens to make sure all elecrons have a safe passage between the two.

Also if you are using live view the metering may screw you over.

Check, check and check. Matrix metering, not connected to the focus point, iso range 100-3200. I very rarely use live view, and definitely not for a fast moving little mammal in bright sun!

I do the birds and also use TAv which, as you have found, generally works very well. My only thought is that you can set the maximum range of the ISO and maybe the lower setting got somehow changed from 100 to some higher number and the camera was not able to bring the exposure down to a low enough level for proper exposure. In other words check the minimum ISO to make sure it is still at 100. I would at least give a quick check although you seem well versed enough I suspect you probably checked this already.

I do have to add that I have had rare occasions with high shutter speeds where I will have one of the series totally blown out but never more than one exposure and less than 5 times total in the 1+ year I have been using it (K50). I hope you find the issue soon. I know how frustrating it would be to lose a great photo in these situations.

Thanks TedW. I might go with restricting the ISO - or just defaulting -0.7 EV as suggested below. I feel a bit annoyed to have to, though - it is a brand new camera, and I think the problem ought to be solvable. But if it is randomly occuring, sending it in for service might not help anyway.

I've also seen situations where I must have done something to have taken ISO out of Auto and fixed it to some value. My guess is the OP may be experiencing this, too, as ISO 3200 is pretty high for bright lighting.

Actually, I have done that as well. It wasn't the case this time though - images taken before and after the stoat-incident are all fine, same settings.

No, the exposure time would be the same, but the amount of light striking the sensor would be somewhat more than if the aperture had actually been at f/5.6 for the full time. Still, though, the posted photos are at least two stops overexposed. Is there any chance you could post the photos with full exif intact to someplace like Flickr and link the images to this thread? The images above have been munged by the forum software and the exif rewritten in an edited form.

Can you post raw images on flickr? Otherwise I suppose some sort of conversion is inevitable? I did a "convert to jpg" in FastStone with all editing options disabled to keep it close to the original as I could. The previews on the camera screen are as blown out as the posted photos. I will attach a screenshot of the FastStone preview below showing the full EXIF.

Steve

...The appropriate workaround is using Auto-highlight compensation, but this solution is not nearly as well-implemented as it was in the K-30 series. My preference is to go with -0.7 underexposure compensation most of the time.
For me, this is the most evident design flaw in the K-3. I'm amazed it hasn't been more fully discussed. If I was a JPEG user, I'd be outraged. As a RAW+ user who processes in Capture One v8 with its far-superior highlight compensation capabilities, it ends up being mostly a nuisance rather than an outright failure.

I have not tried auto-highlight compensation. the -0.7 EV fix is simple, and I might go with that, even if it means a bit more post processing. Is Capture One better at highlight recovery than Lightroom? I was quite impressed with what Lightroom managed to do (see edited shot at the end)

...My thoughts after reading through the thread are:

• There definitely is something funny going on here
• "Forcing" ISO 3200 in TAv is not the cause. The camera sets the ISO based on the light measurement and the set shutter speed and aperture. The EV is the same regardless of the numbers used
• I have seen similar overexposure with lazy aperture, but the examples show more than 1 stop over
• The available exif are not adequate to rule out user error or to shed light on other reasonable cause

One thing that might be helpful would be to do an in-camera JPEG using the "natural" custom image setting. That would rule out the RAW converter import preset as the source of the lightness.

Thanks Steve - I did not try doing an in-camera JPEG, but I expect it would correspond to the preview I see on the camera screen? This is just as overexposed. To get the edited image in my last thread, I did -2.4 EV in Lightroom.

FWIW, I did the calculation and 1/2000s, f/5.6, ISO 3200 = EV(100) 12

This is a perfectly reasonable meter reading for many daylight situations (bright overcast), though about 2-3 stops overexposure if the subject was in full noonday sun. Specular highlights in the grasses tend to indicate that the sun was probably high and bright (exif says 11:30 AM) which would indicate an exposure of about EV(100) 15. The fault is squarely on the exposure values.

Is there any chance the AE lock was on or linked to AF lock*?


Steve


* Menu --> C --> 1 --> 4 AE-L with AF Locked

AE-L with AF Locked is off - I just checked. You are right that the sun was pouring down, and it was near noon. I didn't choose f5.6 and 1/2000s randomly either - I metered when we arrived at the area to see what shutter speed I could "afford" given the clear sunlight. I really find 1/2000s VERY useful for avoiding blurred bits on animal action shots.

Again, thanks to all of you! I am happy to have been able to rule out some of the possible explanations anyway, even if I didn't exactly find a solution. I expect I will keep shooting and keep an eye on whether it gets worse. For now I am really happy that Lightroom managed to get something usable out of some of the shots. I really did not expect that! (Another "rescued" image at the end)

[ScooterMaxi Jim]

Steve, I think your calculations are correct. However, I do run into situations where the K-3 simply does not adequately account for specular highlights or even larger areas of brightness in high contrast scenes. I should point out that the problem is entirely on the camera side of things. I did a shoot with a 21, FA35, and brand new 50 in quick succession and the problem was similarly evident with all three. Although the conversions were fine, providing overexposed JPEG proofs to the client was a bit embarrassing.

Mette, I probably overstated Capture One's capability. Obviously, if a highlight is totally blown there isn't much you can do (what isn't there can't be recovered). However, how a converter handles the graduation from blow-out to near-blown coloration is what your eyes pick up as unnatural (often in clouds, for instance); C1 handles this (and several other IQ issues) better than LR. LR still is a very good conversion program, but it hasn't been improved much at all in the past five years. I've been back and forth between the two for more than a decade, but C1 is now clearly ahead in speed and IQ (but not as an organizer). Phase One is a fine Danish company!

[stevebrot]

However, I do run into situations where the K-3 simply does not adequately account for specular highlights or even larger areas of brightness in high contrast scenes.

This would not surprise me. While I have found the metering to generally be good, I don't believe that Ricoh/Pentax have perfected how to best leverage the data received from the high resolution RGB meter sensor.


Steve

[MetteHHH]

Steve, I think your calculations are correct. However, I do run into situations where the K-3 simply does not adequately account for specular highlights or even larger areas of brightness in high contrast scenes.

This could be what is going on here, really. The metal plates where the stoat was hunting were corrugated iron, and to my eye they were dark brown (rusty) - but then my optical lobes would have HDR-processed that information to some degree. I guess the sun did reflect off of them to some degree, and maybe this was simply too much for the camera.

I like that possibility, since it would indicate that my K3 might actually be alright and that the problem is not likely to get worse just because of this series...

[trailer]

I recently purchased a demonstrator K3 which was said to be overhauled to over 90% of 'new' condition, but right away, I was having extreme overexposure issues when using the AV setting. Images were almost totally white. I gave up pretty quickly when even the store couldn't figure it out, and it has been sent in to Pentax for evaluation and repair. The problem was usually limited to f8 and above.

[Mikesul]

Did you have it on spot metering?

[Kobie]

Here's a test... Take a shot in Aperture mode, then crank up the shutter speed in TaV and compare. Pay attention to the ISO when shooting in TaV, make sure it's not flashing to tell you it won't expose correctly
I've had this happen a few times, I set the shutter speed too low (1/2500 Sec) and over exposed because the ISO was at 100 and couldn't go any lower. I should have been at around 1/4000 sec when that happened.