[GerryL]

There is this issue of Shutterbug magazine that reviews the K20D and one of the shots that was tried was for a skiing event.
The reviewer was testing the exposure and the fps if it would be able to capture this type of a sport. Glad to say that it did perform.
I think this is the November 2008 issue.
They also have the magazine online but is a more condensed version of the review (it has the pix of the skiing but not all of it and not the whole commemtary).
You can go to Shutterbug: Home Page for the K20D review.
Hope this helps.

[RioRico]

Just be glad you didn't have to make a camera from two matchboxes and a tin beer can: self_made_camera

[RiceHigh]

The recent low prices have me very interested in a K20D, and I've read about as much material (reviews, etc.) as exists on the camera. While I hear lots about the AF (not much of which is good) I'd like to hear more specifics...does it hunt?, is it just slow to focus, are some lenses better than others and if so which ones?

Its really sloooow indoor, the latest entry level Pentax K-m is much faster:

RiceHigh's Pentax Blog: AF Speed Difference of K-m Vs K20D

I also see some negative comments about underexposing...is that correctable via a permanent setting (essentially as a preference?) or only via +/- exposure adjustments?

Is the exposure metering up to modern snuff? How does it compare (qualitatively, not numerically) to Nikon's 3D Matrix system?

Nikon's matrix metering is far more intelligent and accurate. The K20D's metering does not only tend to underexpose, but it does overexpose. It is just not accurate enough and results can be rather inconsistent:

RiceHigh's Pentax Blog: K20D Exposure Inaccuracy and Inconsistency

I don't care about the limitations of live view or the comparative 'slow' FPS. I know the 2.7" LCD is both smaller and lower resolution than many, and can deal with that.

Me too. But a 3"+ mon with VGA resolution would be nice.

...just trying to get a better handle on the minuses to this otherwise impressive camera..perhaps I will just have to buy one from B&H or elsewhere (and a lens) and give it a go for a week or so and pay the re-stocking fee if it doesn't work out...

You should note about the "hot pixel issue" and "forced Dark Frame Subtraction" of the K20D too. Relevant read by me:

RiceHigh's Pentax Blog: Issue of Hot/Dead Pixels of the K20D

RiceHigh's Pentax Blog: Noise "Reduction"?

In short, I won't buy a K20D or GX-20 *now*. A K20D replacement might be announced just within a few days at the PMA. I am sure it will have some of the above problems improved and probably the new model will be available before or in Summer. So, just wait if you can, my humble advice.

[rfortson]

Its really sloooow indoor, the latest entry level Pentax K-m is much faster:

RiceHigh's Pentax Blog: AF Speed Difference of K-m Vs K20D



Nikon's matrix metering is far more intelligent and accurate. The K20D's metering does not only tend to underexpose, but it does overexpose. It is just not accurate enough and results can be rather inconsistent:

RiceHigh's Pentax Blog: K20D Exposure Inaccuracy and Inconsistency



Me too. But a 3"+ mon with VGA resolution would be nice.



You should note about the "hot pixel issue" and "forced Dark Frame Subtraction" of the K20D too. Relevant read by me:

RiceHigh's Pentax Blog: Issue of Hot/Dead Pixels of the K20D

RiceHigh's Pentax Blog: Noise "Reduction"?

In short, I won't buy a K20D or GX-20 *now*. A K20D replacement might be announced just within a few days at the PMA. I am sure it will have some of the above problems improved and probably the new model will be available before or in Summer. So, just wait if you can, my humble advice.

You just couldn't help yourself, could you?

BTW, to the OP, take RH's blog with a large grain of salt. Let's just say a lot of his findings aren't supported by real world users, especially the many people here happy with the K20D.

[kenyee]

You should note about the "hot pixel issue" and "forced Dark Frame Subtraction" of the K20D too. Relevant read by me:
RiceHigh's Pentax Blog: Issue of Hot/Dead Pixels of the K20D

AFAIK, this is no longer a problem, at least w/ the K20D I bought at the end of last year. I put a comment in your blog about my testing and results.

[Ben_Edict]

The recent low prices have me very interested in a K20D, and I've read about as much material (reviews, etc.) as exists on the camera. While I hear lots about the AF (not much of which is good) I'd like to hear more specifics...does it hunt?, is it just slow to focus, are some lenses better than others and if so which ones?

The AF is fast and reliable. It is not quite as fast as Canon's and Nikon's, especially at lower light levels. But depending on the lens you use, it locks more reliably than Canon's. The main shortcoming is with the continous AF, but even that is adequate for many things in my experience. Hunting is usually only a problem with slow lenses under dim lighting. This also should put the complaints some people have into the right perspective: usually they are using only the slow kit lens or sup(p)er-zooms, which have very slow apertures at the long end. Use a sensible lens, and AF performance improves.

I also see some negative comments about underexposing...is that correctable via a permanent setting (essentially as a preference?) or only via +/- exposure adjustments?

I have read loads of comments. Usually by people repeating, what they have heard or read elsewhere. From my own experience I would say: In high contrast conditions the K20 tends to underexpose slightly, in order to prevent blown out highlights.- It is easier (though it increases noise) to bring back shadow detail from the then underexposed parts, than bring back non-existent (aka washed-out) details from the highlight areas. On my K10 I had exposure compensation sat to +1/3 all the time, you can do so with the K20, if the results please you more.

Otherwise, in all usual lighting conditions, the K20 exposes nicely.

I cannot deny, that there is a certain variance in expsoure, if you take a series of images. In between there are single shots, with a visible under - or overexposure (though usually within the limits of the RAW contrast ratio and thus can be optimised in post-processing). This somewhat erratic behaviour is much more pronounced in the K10D and you should not overestimate its significance. Firstly it is fixable and secondly it is not the norm, but a rare, though undiniable, occurence.

Is the exposure metering up to modern snuff? How does it compare (qualitatively, not numerically) to Nikon's 3D Matrix system?

No and yes. Pentax matrix metering has been around for the last 15 years or so. It is not as sophisticated as the Nikon system (i.e. less measuring points), but it works nicely.

I don't care about the limitations of live view or the comparative 'slow' FPS. I know the 2.7" LCD is both smaller and lower resolution than many, and can deal with that.

The lcd resolution is a good issue, typical for the Pentax marketing, which always understates. All other vendors simply count each screen pixel. Pentax always only counts three pixels (the primary colours of the screen RGB) as a single image forming pixel. Thus the misconception, that the lcd would sport a much lower resolution as the competitor's. It is simply not true. The lcd resolution is en par with nearly any current DSLR on the market and significantly better, than that of many cheaper modells. I have so far seen this mentioned only once in a magazine test, so far...

Ben

[rfortson]

AFAIK, this is no longer a problem, at least w/ the K20D I bought at the end of last year. I put a comment in your blog about my testing and results.

Agreed, even before I upgraded my firmware, I tried the "recipe" posted online for seeing the problem, and my K20D wouldn't show it. So, while the issue was acknowledge and corrected by Pentax, it did not affect all K20Ds. In fact, it was several months after the K20D was released before this "issue" was even brought to light.

[Hobo]

I have no complaints at all of the K20D. Sure, other cameras may have faster AF, but I doubt if they have better ergonomics and user interface. That right there can cancel out the advantages of faster AF - at least under many conditions. I also have to wonder about this need for speed. I mean, many of the very best sports-oriented photos in the past 50 years were taken with manual focus cameras - hell, some even with flash bulb (i.e., Ali standing over his fallen foe with the smoky arena air behind him - one of the best captures of all times). Makes we wonder how critical fast AF really is. Perhaps to capture action yes - but certainly not to capture life - or more specifically - the drama of sport.

Then again I don't depend on my camera to make a living - I depend on it to make me smile - and in that - it performs exceedingly well.

[pingflood]

The AF is fast and reliable. It is not quite as fast as Canon's and Nikon's, especially at lower light levels. But depending on the lens you use, it locks more reliably than Canon's.

I have to respectfully disagree. There's this general concept floating around of Pentax AF being slow but "more accurate" -- I've found that to pretty much not be the case. I've never had any focusing inaccuracies or issues with any of my Canon bodies (300D, 2 10Ds, 1DsII), and the same goes for my brother's D40. Pretty much any modern DSLR will focus accurately in "one shot" mode unless there's something wrong with it.

[pentaxmz]

- AF tracking is behind the competition
- Flash strobes in low light to get a lock on a target

True, but with in the absence of a focus assist light, this doesn't bother me.

- Mirror slap is louder than competition

Definitely true but only when compared to Canon's brilliant mirror control algorithms on the 40D and 50D (perhaps other Canons as well?).

- It does under expose but IMHO that is so easy to fix, just dial in +0.3 to 0.7 as a default and you're all set. This is really a non issue as it is very, very consistent and predictable.

All this talk about the K20D purposefully underexposing to protect highlights sounds more like nonsense to protect a potential metering issue. True, it is easy to correct with a +EV. But for my full-frame Sigma's and Tamron's I have to add 1 1/2 to 2 stops of + compensation. My APS-C lenses need about 1/2 stop. Underexposure will introduce more noise, while over exposure will wash out highlights. I don't want either (if I can help it), so... why not meter correctly in the first place?

- Some clasic designs are eye watering in price (think an 85/1.4 for example, an 80-200 ... prices are just absurd)

I cannot agree with you here. Expensive glass is expensive for a reason. And good glass prices are pretty much universal across all manufacturers.

That's the bad news, I just bought a Nikon and good news for Pentax is that the SR is just the bees dick, honestly it's great. Pentax offers unique shooting experiences with the tiny, brilliant primes like the 21mm pocket lens, the 77/1.8 is tiny etc etc . No one else offers this, not even close.

Pentax isn't the only camera with in-body SR and access to good primes.

Some lenses are crazy cheap, the 55-300 is cheap for the quality of the optic (slow AF though) and the 12-24 is a chunk cheaper than the Canikon options.

Quality is subjective so I'm not sure what your criteria is.

But prices, for non-IS lenses, built for other manufacturers are similar to PK mount lenses.

To OP.... overall the design of the K20D is good. The camera does suffer from some flaws. But, as many have pointed out, you may not notice these flaws unless you have specific requirements. It is my opinion that many of these flaws could be fixed with better software (firmware). To list those that I know:

Hardware flaws (likely not fixable)

- sensor noise (allow truly visible in longer exposures)
The K20D sensor is hot compared to other cameras and that is why
a Dark Frame Subtraction is necessary on any BULB exposure (or > 15 sec.).

Specialized application: The K20D is not suitable for long exposures
(e.g. astrophotography, etc.)

Software flaws (likely fixable)

- Auto Focus. Another controversial issue but it is absolutely true that the
K20D is slower (by way of more seeking) than even less expensive DSLRs.
Even my 10 year old Pentax MZ-S is faster.

See this page for a video comparing AF on two Pentax cameras. It is clear
that the less expensive K-M is faster than the K20D:
??? | ?????? ??

Is it really THAT bad? No! It's good enough unless you specifically require
super-fast focusing.

I believe the software could be improved, if Pentax was so inclined.

- Live View. Pentax clearly 'slapped' this feature on as a half-thought out
marketing tool. This could be corrected in a firmware upgrade.

But there are good attributes about the K20D and it would only be fair to balance with the good. Again, unless you have 'do not past go' specific requirements... the K20D is a good value.

See here for the good attributes: https://www.pentaxforums.com/forums/pentax-dslr-discussion/50138-happy-k20d-users-post-here.html

And my specific post: https://www.pentaxforums.com/forums/504931-post118.html

[Marc Sabatella]

All this talk about the K20D purposefully underexposing to protect highlights sounds more like nonsense to protect a potential metering issue.

No, the effect is real and easily measurable. It doesn't take but a few minutes of experimentation to get a pretty clear idea of how the meter works in any of three modes (multisegment, center-weighted, and spot) and to verify that its behavior is consistent, predictable, and adheres admirably well to ISO standards for exposure.

why not meter correctly in the first place?

It does, by the classic and ISO-standardized definition of correct exposure. If you understand how the metering works and why it works that way, you'll find it perfectly consistent and accurate. Only if you insist on wanting it to read your mind and produce the exposure you personally happen to want instead of the exposure it was designed to create does there appear to be an "issue". Hint: if you don't want it to protect highlights, don't use multisegment mode.

[Ben_Edict]

I tend to agree with some of your statements, but not all of them. And ofcourse to some degree the experiences reflected in this thread are subjective, so I add my own practical experience.

All this talk about the K20D purposefully underexposing to protect highlights sounds more like nonsense to protect a potential metering issue. True, it is easy to correct with a +EV. But for my full-frame Sigma's and Tamron's I have to add 1 1/2 to 2 stops of + compensation. My APS-C lenses need about 1/2 stop. Underexposure will introduce more noise, while over exposure will wash out highlights. I don't want either (if I can help it), so... why not meter correctly in the first place?

Personally I have never seen that much underexposure with the K20. With the K10 I almost always used a +1/3 EV compensation and I do sometimes on the K20, when the scenes include a fair amount of brighter background. But all lenses, be they full-frame or APS-C format, be the Pentax or Sigma, behave the same, with the exception of some old manual glass. Ofcourse I have to apply more exposure compensation when shooting against bright light sources, but that's the same for all cameras. All in all I am very satisfied with the K20's exposure.

But it might just be, that I am dumb enough to be easily satisfied after three decades of photography with all the different formats...

Pentax isn't the only camera with in-body SR and access to good primes.

The post, you replied to, aimed especially at the small size of the DA Limiteds, and these are indeed something special. You will find single "pancake" lenses from almost all manufacturers (Oly just introduced one), but none of them has the comparatively broadand current(!) portfolio of compact and high-quality primes, that Pentax has on offer.

To OP.... overall the design of the K20D is good. The camera does suffer from some flaws. But, as many have pointed out, you may not notice these flaws unless you have specific requirements. It is my opinion that many of these flaws could be fixed with better software (firmware). To list those that I know:

Hardware flaws (likely not fixable)

- sensor noise (allow truly visible in longer exposures)
The K20D sensor is hot compared to other cameras and that is why
a Dark Frame Subtraction is necessary on any BULB exposure (or > 15 sec.).

Specialized application: The K20D is not suitable for long exposures
(e.g. astrophotography, etc.)

I find (and this is subjective) the K20's noise to be on a very good level. Under normal circumstances at low ISO settings it is not noticeable, unless you search particularily sensible areas. I don't do that, as I am photg and not a noise-searcher. I even find the noise up to ISO 1000 on a very good level, producing photographies, which I can easily print large. Above that, I take compromises, as with any other APS-C camera.

Oh, yes and I just used the K20 for short exposure astrophotography (30 s at ISO 3200). And while I find the automatic dark frame substraction annoying and would wish, that I could switch it off, if works nicely. Averaging a series of these short expsoures also gets rif of the remaining noise in the background, which is indeed not nice at that ISO setting.

Software flaws (likely fixable)

- Auto Focus. Another controversial issue but it is absolutely true that the
K20D is slower (by way of more seeking) than even less expensive DSLRs.
Even my 10 year old Pentax MZ-S is faster.

Yes, it is slower than some other cameras. Nevertheless it is s very noticeable advancement over any previous Pentax and at a level, which makes it very useable under most circumstances. My experience with the MZ-S is too limited to include that particualr modell, but the PZ-1p, which is considered to be the latest modell, that made some steps into a pro-modell direction, is certainly slower.

See this page for a video comparing AF on two Pentax cameras. It is clear
that the less expensive K-M is faster than the K20D:

Is it really THAT bad? No! It's good enough unless you specifically require
super-fast focusing.

I believe the software could be improved, if Pentax was so inclined.

- Live View. Pentax clearly 'slapped' this feature on as a half-thought out
marketing tool. This could be corrected in a firmware upgrade.

But there are good attributes about the K20D and it would only be fair to balance with the good. Again, unless you have 'do not past go' specific requirements... the K20D is a good value.

Comparing the K20 to the K-m must take into account, that the K-m is 18 months younger and even Pentax made some progress with their AF system in the mean time.

All in all, I think quite like you: the K20 is a highly respectable tool, if you need it for your applications. For some niches, I would not consider it in the first place (pro sports photography). And yes, if I would make astrophotos every other night, I would buy an addtional optimized Canon body (converted for extended red sensibility etc.). But for occasional use, I stay with the K20.

Ben

[pentaxmz]

You just couldn't help yourself, could you?

BTW, to the OP, take RH's blog with a large grain of salt. Let's just say a lot of his findings aren't supported by real world users, especially the many people here happy with the K20D.

I must disagree with you!

While I don't always agree with RiceHigh, he does an excellent job bringing to light both the good and the bad in Pentax's products. The posting you quoted was sound information for the OP to consider.

I am curious if you are allowing your obvious negative bias toward Ricehigh to get in the way of the facts. What exactly do you not agree with in Ricehigh's quote?

[pentaxmz]

No, the effect is real and easily measurable. It doesn't take but a few minutes of experimentation to get a pretty clear idea of how the meter works in any of three modes (multisegment, center-weighted, and spot) and to verify that its behavior is consistent, predictable, and adheres admirably well to ISO standards for exposure.

It does, by the classic and ISO-standardized definition of correct exposure. If you understand how the metering works and why it works that way, you'll find it perfectly consistent and accurate. Only if you insist on wanting it to read your mind and produce the exposure you personally happen to want instead of the exposure it was designed to create does there appear to be an "issue". Hint: if you don't want it to protect highlights, don't use multisegment mode.

Hello Marc,

You quoted me out of context. I was speaking of my own experience. If I have to keep my K20D set for + 2 stops of EV, that seems weird. Wouldn't you agree?

And I never said it was a 'show-stopper' for me. It's really not a big deal that I must do this... having instant feedback makes this problem moot. If this were a film camera, obviously I'd be screaming bloody murder!

I do have a very good understanding of how the K20D metering works...but EV compensation is the worst on my Tamron SP 24-135mm gold 50th anniversary lens. Does anyone have this 1999 lens?

[rfortson]

I must disagree with you!

While I don't always agree with RiceHigh, he does an excellent job bringing to light both the good and the bad in Pentax's products. The posting you quoted was sound information for the OP to consider.

I am curious if you are allowing your obvious negative bias toward Ricehigh to get in the way of the facts. What exactly do you not agree with in Ricehigh's quote?

Just been around enough forums where RH has been a constant complainer about Pentax. I think I'm one of the more balanced posters, as RH doesn't bug me out of hand, but 3-4 years of seeing him on a variety of forums has shown his bias to me (and most everyone else). BTW, I haven't seen much "good" about Pentax from RH, but maybe you and RH have the same "calibration".