[audiobomber]

Agreed, a camera is much more than just a sensor but it helps a lot to start with a great sensor.

Which sensor used in a Pentax camera wasn't great? Are you saying that the Sony sensors are great compared to the Samsung sensors? When the K20D was released, it had the highest resolution and the best high ISO performance of any APS-C camera, including any with Sony sensors. It still compares well with the latest cameras. My K-x has better high ISO performance, but I still prefer the look of my K20 files at anything below 1600.

[zxaar]

Which sensor used in a Pentax camera wasn't great? Are you saying that the Sony sensors are great compared to the Samsung sensors? When the K20D was released, it had the highest resolution and the best high ISO performance of any APS-C camera, including any with Sony sensors. It still compares well with the latest cameras. My K-x has better high ISO performance, but I still prefer the look of my K20 files at anything below 1600.

Pentax Digital Cameras: Digital Photography Review

k20d - Announced 23-Jan-08


Nikon Digital Cameras: Digital Photography Review

D300 - Announced 23-Aug-07

Sensor rankings

DxOMark - Sensor rankings

d300 - rank 26
20d - rank 36


PS: By the way sony A700 - Announced 06-Sep-07 ranks 27 here on same page.

[ducdao]

Which sensor used in a Pentax camera wasn't great? Are you saying that the Sony sensors are great compared to the Samsung sensors? When the K20D was released, it had the highest resolution and the best high ISO performance of any APS-C camera, including any with Sony sensors. It still compares well with the latest cameras. My K-x has better high ISO performance, but I still prefer the look of my K20 files at anything below 1600.

I only own a K100 and a K10D and to be honest I prefer the images from my K100D better than the ones from K10D from ISO400 and above. I can't talk about the K20D since I don't own one. I think it all come down to personal preferences and maybe I'm being influenced by other forum members who say that Sony sensors are better than Samsung's and it happens that I'm also so happy with my K100D. I also heard a lot of praises for the k-x which also has a Sony sensor.

[Marc Sabatella]

DxOMark - Sensor rankings

d300 - rank 26
20d - rank 36

Those numbers are but one possible way to measure performance. And to get a sense of how well they correlate with most people's perception of reality, I'd note the K10D scores slightly above the K20D by this measure, too.

In any case, the existing of one *much* more expensive model that in one metric beat the K20D by an almost statistically insignificant margin (67 versus 65 in a logarithmic scale in which is takes a 5 point different to make a 1/3 stop difference) in no way disproves the notion that the K20D was basically at the top of its class.

---

Last edited by Marc Sabatella; 06-29-2010 at 05:56 PM.

[Marc Sabatella]

I only own a K100 and a K10D and to be honest I prefer the images from my K100D better than the ones from K10D from ISO400 and above.

That's a common enough perception - and note both are Sony sensors. I would say, though, that most people who draw this conclusion are making the mistake of comparing the images at very different sizes, comparing at "100%" for each camera rather than comparing at the same size. The differences become much less pronounced when comparing images at the same size. They also are less pronounced if you shoot RAW rather than rely on the in-camera JPEG engine.

But in any case, the actual story here isn't that the K100D is above average at high ISO - it's that the K10D is *below*. Pretty much every single other model Pentax has made does as well or better as the K100D when comparing images at the same size. Again, not that the K100D isn't a great camera, but it's greatly reduced resolution often fools people into thinking it's better at high ISO than it is, simply because they never blow images from the K100D up as large as they do the later cameras.

maybe I'm being influenced by other forum members who say that Sony sensors are better than Samsung's

People who say this are almost invariably talkng specifically about the K-x versus the K-7, not Sony and Samsung in general. As mentioned, the other Samsung sensor - the one in the K20D - beats every other Sony sensor by almost all accounts. And even the K-7 sensor is better than some Sony sensors by many accounts, if not as good as others. So I'd say you're still overgeneralizing.

Of course, we're mostly splitting hairs here. The lenses you use and your own technique have a *far* greater impact on the pictures taken than the sensor does, except at the very highest ISO levels.

[zxaar]

Those numbers are but one possible way to measure performance. And to get a sense of how well they correlate with most people's perception of reality, I'd not the K10D scores slightly above the K20D by this measure, too.

In any case, the existing of one *much* more expensive model that in one metric beat the K20D by an almost statistically insignificant margin (67 versus 65 in a logarithmic scale in which is takes a 5 point different to make a 1/3 stop difference) in no way disproves the notion that the K20D was basically at the top of its class.

As you say the difference is minuscule and it does not disproves the notion that k20d was basically at the top of the class, but the same time it does not prove that k20d was at the top.


Numbers even if they are small do say that other sensors released by sony were at least as good as k20d's.

plus sony has improved things very much since then but with k7 things become worse as k7's sensor ranks significantly lower than k20d on dxomark.

All in all this supports OP.

[audiobomber]

Sensor rankings

DxOMark - Sensor rankings

d300 - rank 26
20d - rank 36


PS: By the way sony A700 - Announced 06-Sep-07 ranks 27 here on same page.

Unfortunately DPR didn't check for raw noise when they reviewed the K20D, but from the DPR review of the K-7:
"Raw Noise graphs
"The graphs below confirm what we can see in the sample crops. The four cameras compared here (D300, 50D, K-7, K20D) produce similar amounts of noise at all ISO settings. The EOS 50D measures the largest amount of chroma noise and the K20D is producing the lowest RAW noise levels pretty much across the range."

DXOMark's score for the K20D is suspect, especially since they rated the GX20 sensor higher, when it's exactly the same design. DxOMark - Compare sensors

[audiobomber]

All in all this supports OP.

No it doesn't. The OP's position is that Sony sensors are superior to the other sensors used in Pentax cameras. The older Sony sensors don't compare very well to the Samsung sensor. The K-x has great high ISO performance, but I can't make it deliver images that look as good as the K20 flies out of the camera at anything below 1600.

The K-7 sensor was comparable with the APS-C state of the art when it was released. The K-x generation of sensors is newer and better in some ways. I predict the K-7 replacement sensor will be better than the sensor in the K-x, whether it's a Sony or not.

[ducdao]

That's a common enough perception - and note both are Sony sensors. i would say, though, that most people who draw this conclusion are making the mistake of comparing the images at very different sizes, comparing at "100%" for each camera rather than comparing at the same size. The differences become much less pronounced when comparing images at the same size. They also are less pronounced if you shoot RAW rather than rely on the in-camera JPEG engine.

But in any case, the actual story here isn't that the K100D is above average at high ISO - it's that the K10D is *below*. Pretty much every single other model Pentax has made does as well or better as the K100D when comparing images at the same size. Again, not that the K100D isn't a great camera, but it's greatly reduced resolution often fools people into thinking it's better at high ISO than it is, simply because they never blow images from the K100D up as large as they do the later cameras.



People who say this are almost invariably talkng specifically about the K-x versus the K-7, not Sony and Samsung in general. As mentioned, the other Samsung sensor - the one in the K20D - beats every other Sony sensor by almost all accounts. And even the K-7 sensor is better than some Sony sensors by many accounts, if not as good as others. So I'd say you're still overgeneralizing.

Of course, we're mostly splitting hairs here. The lenses you use and your own technique have a *far* greater impact on the pictures taken than the sensor does, except at the very highest ISO levels.

Making a lot of sense, Marc!

[zxaar]

No it doesn't. The OP's position is that Sony sensors are superior to the other sensors used in Pentax cameras. The older Sony sensors don't compare very well to the Samsung sensor. The K-x has great high ISO performance, but I can't make it deliver images that look as good as the K20 flies out of the camera at anything below 1600.

The K-7 sensor was comparable with the APS-C state of the art when it was released. The K-x generation of sensors is newer and better in some ways. I predict the K-7 replacement sensor will be better than the sensor in the K-x, whether it's a Sony or not.

how do you know the next sensor from k7 will be better. Think about it you could have made the same statement about k20d that the next would be better. But the next from k20d was k7 and sensor wise it was not better, at least from low light point of view.

As far as sony is concerned you must remember that sony a700 time sony was criticized for noise at high isos, so in terms of low light they WERE not good enough compared to competion.

But interesting thing is A700 still ranks higher in sensor marking than k20d. So their worse was on par with what samsung has to offer.

Now with new generation of sensors in kx, and some people are saying that NEX5 sensor is slightly better than kx in low lights.

All in all they do have indicated progress and today when we talk sony do have superior low light sensors. Easily better than anything samsung has to offer.

The K-x generation of sensors is newer and better in some ways. I predict the K-7 replacement sensor will be better than the sensor in the K-x, whether it's a Sony or not.

What makes you so sure that sony will be sleeping while this time. They already have lead, it might happen that gape is widen further. May be sony's new sensor might just be too good. Who knows.

[er1kksen]

As far as sony is concerned you must remember that sony a700 time sony was criticized for noise at high isos, so in terms of low light they WERE not good enough compared to competion.

But interesting thing is A700 still ranks higher in sensor marking than k20d. So their worse was on par with what samsung has to offer.

Now with new generation of sensors in kx, and some people are saying that NEX5 sensor is slightly better than kx in low lights.

You do know they introduced a firmware update quite a while ago for the A700 that makes it virtually identical in low light to the D300 (not surprising, since they use the same sensor)?

Sensor implementation, as far as the data pipeline (both hardware and software) can have as much impact on the quality of the output as the quality of the base sensor itself. Look at the K-x vs. the A500. Both use the latest generation of 12mp Sony CMOS (not quite the same model as the D90 and D5000 used, but newer). Pentax wrings noticeably better quality out of that sensor in the K-x than the A500 does (though it's not bad). There was a good bit of talk back in the day on how much better the K20D's output could have been with better A/D convertors (I remember someone running the numbers and finding that they added a large portion of the noise in the RAW files).

[ducdao]

So to go back to the original purpose of this thread do we have the final list of all Pentax DLSR with their respective sensor as below?

K-x: Sony CMOS, - 12mp
K-7: Samsung CMOS, 14.5mp (4-channel)
K20D: Samsung CMOS - 14.5mp (2-channel)
K2000/K-m: Sony CCD - 10mp
K200D: Sony CCD - 10mp
K10D: Sony CCD - 10mp
K100D Super: Sony CCD, 6mp
K100D: Sony CCD - 6mp
K110D: Sony CCD - 6mp
*ist: Sony CCD - 6mp

The K10D is really fuzzy to me, can someone post a link from official source confirming that the K10D really uses a Sony sensor?

[Rondec]

The K10, K200, and Km/2000 all have exactly the same sensor. It is funny, because while I think the Sony sensors are fine, the only one that I really struggled with in high iso situations was the K10. Samsung sensors are about the same, although the K7 probably took a (small) step back in order to get faster read out speeds.

[audiobomber]

how do you know the next sensor from k7 will be better. Think about it you could have made the same statement about k20d that the next would be better. But the next from k20d was k7 and sensor wise it was not better, at least from low light point of view.

I know it will be better because the technology is still advancing. There have been reports of a new and better Samsung sensor in the works. Sony is always working on something new. Foveon sensors are said to do some things bettter than any other. What's unknown is whether Pentax will stay with Samsung, go with Sony, or do something totally different (unlikely).

But interesting thing is A700 still ranks higher in sensor marking than k20d. So their worse was on par with what samsung has to offer.

According to DPR, the K20D has lower noise than a Nikon D300, which uses the same sensor as the Sony. Some of the (early?) K20D sensors were not up to scratch. If you read threads on K20D noise, a few people have serious banding at high ISO, while most don't. I suspect the DXOMark K20D was sub-par and that's why it didn't achieve the same mark as the GX20. I can't think of any other explanation, unless the explanation is that DXO measurements are flawed, which is what some people believe.

All in all they do have indicated progress and today when we talk sony do have superior low light sensors. Easily better than anything samsung has to offer.

For someone who shoots video, the K-7 sensor is better than the K20D sensor. IME, the K20D renders colours better and has a more 3-D look than the K-x, which makes it a better landscape shooter. Some portrait specialists rave about the K20D skin tones. The K-x is a stop better for low light. Clearly the K-7 is a better camera than the K20D and the K20D is a better camera than a K-x. I carry the K20, K-x, or both, depending on what I need for the day. It's nice to have variety.

What makes you so sure that sony will be sleeping while this time. They already have lead, it might happen that gape is widen further. May be sony's new sensor might just be too good. Who knows.

I'm glad that Sony entered the DSLR market. Their aggressive marketing and technological resources are helping to push the envelope. But these big companies don't own the market on innovation. The next revolutionary advance could come from left field.

[audiobomber]

Samsung sensors are about the same, although the K7 probably took a (small) step back in order to get faster read out speeds.

The buzz is that the K-7 has slightly higher noise than K20 due to the 4-channel design, implemented to accomodate video.