[luftfluss]

The IQ compromise of having video on a DSLR was laid bare by the K20D and K-7 models; the loss of 1 stop of DR and 1/2 stop of high ISO performance.

That may have had more to do with Samsung's 1st generation CMOS sensor design, rather than any inherent disadvantage with APS-C sized CMOS vs CCD sensors.

K-x (CMOS) proved to have a sizable leg up on K-m (CCD) in DR and high-ISO performance.

DxOMark - Compare cameras side by side

[audiobomber]

That may have had more to do with Samsung's 1st generation CMOS sensor design, rather than any inherent disadvantage with APS-C sized CMOS vs CCD sensors.

K-x (CMOS) proved to have a sizable leg up on K-m (CCD) in DR and high-ISO performance.

I wasn't discussing CCD vs CMOS, just the addition of video to DLSR's. K-x and K-m used different sensors. The point of the K20D vs. K-7 comparison is that it was the same basic sensor, but tweaked for video capability on the K-7. If anything the K-7should have been better because it was a second iteration, but video circuitry adds noise, hence the compromised SNR and DR. The same thing happens with CD vs DVD. Video noise compromises DVD sound quality.

[Rondec]

I wasn't discussing CCD vs CMOS, just the addition of video to DLSR's. K-x and K-m used different sensors. The point of the K20D vs. K-7 comparison is that it was the same basic sensor, but tweaked for video capability on the K-7. If anything the K-7should have been better because it was a second iteration, but video circuitry adds noise, hence the compromised SNR and DR. The same thing happens with CD vs DVD. Video noise compromises DVD sound quality.

There is always a trade off, but most people would trade the small high iso loss for the jump from 3 fps to 5 fps (video aside). The difference between my K20 and K7 was imperceptible in real life shooting situations.

Dxo mark gives difference as sports iso K20 of 640 versus K7 of 540 and dynamic range of K20 11.1 EVs versus K7 at 10.6. A measurable difference, but not a real world difference in my opinion.

[Laurentiu Cristofor]

Dxo mark gives difference as sports iso K20 of 640 versus K7 of 540 and dynamic range of K20 11.1 EVs versus K7 at 10.6. A measurable difference, but not a real world difference in my opinion.

Especially when comparing the performance with K-5.

And K10D had better DR (and overall score) than both K20D and K-7 - 11.6!

[ElJamoquio]

I guess at less than a cm, there isn't going to be an aperture ring.

I'd guess. As is there's barely a focus ring.

[luftfluss]

I wasn't discussing CCD vs CMOS, just the addition of video to DLSR's. K-x and K-m used different sensors.

As I stated in my post.

The point of the K20D vs. K-7 comparison is that it was the same basic sensor, but tweaked for video capability on the K-7. If anything the K-7should have been better because it was a second iteration, but video circuitry adds noise, hence the compromised SNR and DR. The same thing happens with CD vs DVD. Video noise compromises DVD sound quality.

OK, what threw me, then, was your use of the word "and", which is inclusive.

It's still hard to know Samsung's design process, and if they just quickly "tacked on" video capability rather than include it more organically. It's pretty darn likely that the core design of the sensor was not touched at all.

[audiobomber]

There is always a trade off, but most people would trade the small high iso loss for the jump from 3 fps to 5 fps (video aside). The difference between my K20 and K7 was imperceptible in real life shooting situations.

Dxo mark gives difference as sports iso K20 of 640 versus K7 of 540 and dynamic range of K20 11.1 EVs versus K7 at 10.6. A measurable difference, but not a real world difference in my opinion.

The added fps had nothing to do with video, that's the shutter. Lots of DSLR's had five or more fps before video.

DR and ISO differences depend on what ISO you use to compare. Looking at the graphs, the K20D is pretty close to midway between the K-x/K-r and the K-7 and look how much fuss people make over the slightly superior performance of the K-x vs the K20D (I've fought that one many times, but it's a hopeless cause). Even the K-5 is not that far ahead, but people go apeshit over the difference.

[Anvh]

As to the CCD versus CMOS debate, I think it is a non-starter. At this point CMOS is as good in every respect (and better in many) than CMOS ever was. The sensor in the K5 will match any output from the K10, even at iso 100.

What is the age difference between the two?
4 years... that's a lot of years in electronic years so not comparable.

You need to find two sensors with the same age if you want to compare.

[Fogel70]

The point of the K20D vs. K-7 comparison is that it was the same basic sensor, but tweaked for video capability on the K-7. If anything the K-7should have been better because it was a second iteration, but video circuitry adds noise, hence the compromised SNR and DR.

Do you know if the sensor in K20D has video capacility or not?

I think to remember that the biggest difference between sensors used in K20D and K7 was that K7 used a sensor with 4 channel output for higher frame rate (sensor in K20D had 2 channels).

The same difference was also between Nikon D90 (2 channel) and D300 (4 channel), but D90 had video and better image quality.

[Anvh]

I wasn't discussing CCD vs CMOS, just the addition of video to DLSR's. K-x and K-m used different sensors. The point of the K20D vs. K-7 comparison is that it was the same basic sensor, but tweaked for video capability on the K-7. If anything the K-7should have been better because it was a second iteration, but video circuitry adds noise, hence the compromised SNR and DR. The same thing happens with CD vs DVD. Video noise compromises DVD sound quality.

From what i read the K7 has a lower "blur" filter (sorry forgot the name but that's the effect) on the sensor.
The K7 therefore has more noise but also has more details.

[Winder]

I am less decided here. It is true, a new mount makes lenses for other systems available.

But a K mount forces Pentax to commit to make K mount lenses work flawlessly. I consider it a major failure of the industry that they release mirrorless cameras before they coded good enough CDAF algorithms which don't need special lenses with less moveable glass to function in a satisfactory manner.

Pentax lenses (esp. screw drive lenses) have a strong enough motor to work well with CDAF, assuming a good algorithm is used. We should not forget that Pentax PDAF uses a multistep approach to focus too.

Moreover, rumors have it that FA lenses would work, i.e., the K-01 would support screw drive AF. Of course, if it wants to profit from the Ltd. lenses.
.

By producing an adapter with the screw drive motor built it you gain all the advantages of K-mount support and gives you the advantages of using other lenses. People who don't own a single Pentax lens could pick-up the body and use it with other lens brands.

Olympus expanded its user base significantly as people who had no Olympus glass bought into M4/3 to use their existing lenses. I have a friend who bought the E-P2 to use with his collection of older lenses. He has now bought an E-P3 and is looking at the new OM-D that is coming out. He has bought the 12mm f/2, 45mm f/1.8, and a Panasonic 25mm. Had he not had the option to use his older glass he never would have picked up the Olympus. There is a larger group of people that will travel this route. I have been waiting on a larger sensor camera like the E-P3 series that gives me this option. Leica, Zeiss, & Voigtlander glass options become available. There are several older Olympus, Contax, Canon, Nikon, & Minolta lenses that are really good that could be come an option.

[Anvh]

Do you know if the sensor in K20D has video capacility or not?

Well yes actually.
You simply record the liveview feed....
The problem is that the sensor doesn't have enough lanes to transfer the data when you shoot at higher frames or pixels.

[Rondec]

The added fps had nothing to do with video, that's the shutter. Lots of DSLR's had five or more fps before video.

.

It has to do with the shutter, but it also has to do with read out speed. Even with the K7's shutter, the K20 still wouldn't do 5 fps, i am sure.

[ihasa]

The IQ compromise of having video on a DSLR was laid bare by the K20D and K-7 models; the loss of 1 stop of DR and 1/2 stop of high ISO performance.

I think that's a bit of an odd example. All the Sony sensors perform way better than the Samsung sensors, and also manage to include video with higher definition and higher frame rates. The inclusion of video also helps drive sales, which drives R&D, which raises the bar for IQ...

[audiobomber]

Do you know if the sensor in K20D has video capacility or not?

The K20 did not have video capability.

I think to remember that the biggest difference between sensors used in K20D and K7 was that K7 used a sensor with 4 channel output for higher frame rate (sensor in K20D had 2 channels).

Right, the extra channels were to accommodate video.

The same difference was also between Nikon D90 (2 channel) and D300 (4 channel), but D90 had video and better image quality.

I don't think so. AFAIK the D90 used a newer sensor than the D300. The D90 was from the same generation as the D300S.