[solar1]

Is there an originating post or thread for his 6400 iso pic?

It's pretty darned good looking! (at least on my mac laptop)

Well, actually I think there is. But I also think any converters need some tweaking before they get most out of the new sensor, and as I do not have that I actually haven't tried it. In stead I've used jpegs only.
Everybody is paying a lot of attention to the Pop photo test. But several pictures with noise reduction on and off and with firmware 1.0 were available on the net before te pop photo test to help you judge for yourself.
Personally, I tend to believe the pop photo noise results for the time being, despite the fact that it's not clear if NR was on or off ( though the low iso 800 value may indicate NR kicking in). Noise levels without NR are higher than the d300 at least, but the noise is very very fine. Combined with the high resolution even at iso 6400 I have been able to PP the jpegs to produce very clean pictures with vbery good detail. Thus,despite showing the verdict' uneacceptable' in pop photo, I tend to say 'don't judge it until you've tried yourself'.

Actually, PP D300 pictures (with NR on btw) did not give me much different results. PP the A700 pictures (with NR on btw) never gave me the same quality sofar because of the patchy color reproduction at high iso.

So I suggest we all wait until we get our hands on a decent K20d with firmware 1.0, and see what happens if NR is on or off, throughout the entire iso range, and how these pictures respond to PP.

I personally would have rated pop photo's testing much higher if they would have actually found that 'high speed AF-function' and tested it properly. That would have been news to me. But I guess they really never looked for it. That actually tells me more about how they test cameras than anything thet produce from a standardized measurement (however small the samples may be).

This is iso 6400 on the k20d (thanks to Blende8 who gave us the original jpeg without NR), but after light PP with Neat Image:



I can live with this, because I'm used to the k10d high iso performance after NR.....


lock

[pareto]

Is there an originating post or thread for his 6400 iso pic?

It's pretty darned good looking! (at least on my mac laptop)

That image has been processed with noise reduction software and resized from 14,6MP to 0,7MP. I really don't understand the point in such examples. They don't tell anything about noise performance. Any camera would appear noise free under such circumstances. I have the full size original as it has been posted in the German forum digitalfotonetz.de and it's very bad. There is severe horizontal pattern noise (see the brown cabinet even in the edited and resized image) and a strong magenta shift in the shadows, so everyone who says K20D's 6400 is like 1600 on the K10D is really insulting the K10D.

Don't get me wrong the K20D will be a great camera, and the ISO 3200 version of that image looks okay. But some people suggest it to be a marvel, which will lead to disapointment. In the noise department it does improve over the K10D, but don't get to optimistic, ISO 6400 is only for emergencies where image quality is of subordinate importance.

Have a nice day

[regken]

Don't get me wrong the K20D will be a great camera, and the ISO 3200 version of that image looks okay.
Have a nice day

Is there any way I can get the ISO 3200 RAW file in DNG? Is there someplace I can download it?

Thanks,

Ken

[regken]

Is there an originating post or thread for his 6400 iso pic?

It's pretty darned good looking! (at least on my mac laptop)

I downloaded the original files and think you should look at those before you draw any conclusions. There is a K10D shot of the same at ISO 1600 along with 3 from the K20D at ISO 1600, 3200, and 6400. In Lightroom the ISO 1600 shots look almost identical. The K20D shot was at a shutter speed of 1/125 while the K10D was at 1/60 which will make a little difference in noise but not much. The higher ISO ones are also about one stop under exposed.

While one set of photos doesn't make a positive conclusion these seem to indicate there is little or no difference in noise between the two cameras at ISO 1600.

Regards,

Ken

[lock2nl]

I read that it does not matter much in terms of noise if one sensor gets twice as much light than the other sensor. I disagree......

Furthermore, I must say that Blende8 has a rather noise free version of the k10d if you would compare it to mine (yes, there's variation in the high iso performance between one k10d and the other).

But I uploaded both files at iso 6400 to see for yourself.

without NI :http://aycu07.webshots.com/image/43566/2001314002645639094_rs.jpg

With NI:
http://aycu17.webshots.com/image/42896/2001303150668072413_rs.jpg


Those who are interested in the k10d en k20d at iso 1600 (remember the exposure time is twice as long on the k10d) here are the links:
k10d:http://aycu14.webshots.com/image/45453/2001965730078055671_rs.jpg

k20d:http://aycu40.webshots.com/image/44479/2001979153995393889_rs.jpg

[regken]

I read that it does not matter much in terms of noise if one sensor gets twice as much light than the other sensor. I disagree......

Furthermore, I must say that Blende8 has a rather noise free version of the k10d if you would compare it to mine (yes, there's variation in the high iso performance between one k10d and the other).
without NI :http://aycu07.webshots.com/image/43566/2001314002645639094_rs.jpg

With NI:
http://aycu17.webshots.com/image/42896/2001303150668072413_rs.jpg

Those who are interested in the k10d en k20d at iso 1600 (remember the exposure time is twice as long on the k10d) here are the links:
k10d:http://aycu14.webshots.com/image/45453/2001965730078055671_rs.jpg

k20d:http://aycu40.webshots.com/image/44479/2001979153995393889_rs.jpg

I tend to agree with you that higher shutter speed MIGHT create more noise but to be fair we are talking about 1 stop in these examples. As you stated it is twice as long but still only one stop.

I'm not ready to say there is no difference between the two cameras at ISO 1600 based on one set of photos but I do think it's important for us K10D owners to take a very close look at what the real difference is between these two in determining the value of upgrading. And for the first time buyer a comparison of the K20D to the GX 10/K10D is important considering the $600-$700 difference. Or to put it in your terms, twice the price.

BTW, when I opened up the K10D sample my thought was it looks just like the noise I get out of mine.

Regards.

Ken

[*isteve]

I tend to agree with you that higher shutter speed MIGHT create more noise but to be fair we are talking about 1 stop in these examples. As you stated it is twice as long but still only one stop.

I'm not ready to say there is no difference between the two cameras at ISO 1600 based on one set of photos but I do think it's important for us K10D owners to take a very close look at what the real difference is between these two in determining the value of upgrading. And for the first time buyer a comparison of the K20D to the GX 10/K10D is important considering the $600-$700 difference.

BTW, when I opened up the K10D sample my thought was it looks just like the noise I get out of mine.

Regards.

Ken

No comparison is even slightly valid unless both comparisons are taken at the same exposure and ISO of the same scene at the same time with the same or equivalent lens and processed using the same settings. Everything else means precisely nothing.

[feronovak]

There is one thing I dont understand over these years, why the h*ll RiceHigh doesnt change system and keeps bashing something he hasn't tried yet. Did RiceHigh lose money in Pentax shares or what?

[regken]

No comparison is even slightly valid unless both comparisons are taken at the same exposure and ISO of the same scene at the same time with the same or equivalent lens and processed using the same settings. Everything else means precisely nothing.

That is precisely why I am waiting for exact samples. However, what we have seen so far has made me not pre order as I did with the K10D.

Regards,

Ken

[lol101]

For those who still don't understand, my suggestion is to compare the test results against other DSLR tests. It's just as simple as that.

Btw, I have summarised the PopPhoto test results of the K20D, 40D, D300 and A700 in single tables for direct and easy comparison, just look and judge yourself:-

RiceHigh's Pentax Blog: First K20D Production Camera Formal Test

How a man, priding himself as a "scientific", could even think of doing that?

An uneducated newby wanting to buy a DSLR might be tempted I know, but it's really not worthy of you!

In addition to eventual "on-chip" hardware noise reduction, there are 4 settings for NR on the K20: off, weakest, weak and strong... even if they turned it "on" (and I don't think they did), it doesn't tell us which setting they chose.

If they used jpegs for these tests, it is necessary to know these settings in order to draw any kind of conclusion.

If they used RAW (.dng) and converted it in ACR, it's even worse as ACR is not yet properly calibrated for the K20D. If they used PPL3.5, they'd have to set the NR parameters the same way you would in-camera so that doesn't change anything either...

Knowing the procedure is crucial in order to be able to assess the meaning of the results.

To me, these results are worthless until we know precisely how they were done: as a scientific (and if you wanted to be objective), you should at the very least put some reservations on them.

Putting tests results side by side is pretty sloppy if you know nothing about the "experimental conditions"...

For reference, look at Imaging Ressource review of the D300, there is an interesting part on the effect of the different NR settings (with pictures this time): that should tell you how important it is at ISO 1600-6400 (just compare ISO 3200 NR off whith NR normal and tell me if it has any impact on noise/resolution).

If Pentax follows more or less Nikon's grading in NR, the corresponding K20 NR setting to D300's "normal" should be "weak".


Come on, let's make some sense and at least wait for K20D's review from Imaging Ressource and other sites/magazines before jumping to conclusions.

[lock2nl]

I haven't seen many jumping to conclusions. We are all aware that the materials we have seen sofar are difficult to compare, at least to some extent. But they did give us the impression it's probably better to wait cause sofar things are far from mind blowing without NR. The only thing I found out with the prelease pictures and firmware 1.0 pictures of the k20d is the effectivity of NR software. That allone is a big plus for me.
So yes, let's wait. And hope they test that 'fast AF' option too....

lock

[*isteve]

There is one thing I dont understand over these years, why the h*ll RiceHigh doesnt change system and keeps bashing something he hasn't tried yet. Did RiceHigh lose money in Pentax shares or what?

Oh he did.

He allegedly bought a Canon 5D just so he could say how much better it was in every possibly way. Of course no proof is offered, so we are supposed to take his word for it.

Just put him on your blocked list. Life is so much nicer.

[RiceHigh]

Asking the testers or reviewers to change the camera settings to get the most "optimised" test results is in fact silly and unrealistic. How they bother to do that but not simply test the factory setting, which should be the default one recommended by the manufacturer? How many different combinations should they try before going on to do the test?

In fact, if the default setting won't do good, the camera or its maker deserves for the poor results. They should leave the optimisation works to the users, but not to the testers.

How a man, priding himself as a "scientific", could even think of doing that?

An uneducated newby wanting to buy a DSLR might be tempted I know, but it's really not worthy of you!

In addition to eventual "on-chip" hardware noise reduction, there are 4 settings for NR on the K20: off, weakest, weak and strong... even if they turned it "on" (and I don't think they did), it doesn't tell us which setting they chose.

If they used jpegs for these tests, it is necessary to know these settings in order to draw any kind of conclusion.

If they used RAW (.dng) and converted it in ACR, it's even worse as ACR is not yet properly calibrated for the K20D. If they used PPL3.5, they'd have to set the NR parameters the same way you would in-camera so that doesn't change anything either...

Knowing the procedure is crucial in order to be able to assess the meaning of the results.

To me, these results are worthless until we know precisely how they were done: as a scientific (and if you wanted to be objective), you should at the very least put some reservations on them.

Putting tests results side by side is pretty sloppy if you know nothing about the "experimental conditions"...

For reference, look at Imaging Ressource review of the D300, there is an interesting part on the effect of the different NR settings (with pictures this time): that should tell you how important it is at ISO 1600-6400 (just compare ISO 3200 NR off whith NR normal and tell me if it has any impact on noise/resolution).

If Pentax follows more or less Nikon's grading in NR, the corresponding K20 NR setting to D300's "normal" should be "weak".


Come on, let's make some sense and at least wait for K20D's review from Imaging Ressource and other sites/magazines before jumping to conclusions.

[codiac2600]

Asking the testers or reviewers to change the camera settings to get the most "optimised" test results is in fact silly and unrealistic. How they bother to do that but not simply test the factory setting, which should be the default one recommended by the manufacturer? How many different combinations should they try before going on to do the test?

In fact, if the default setting won't do good, the camera or its maker deserves for the poor results. They should leave the optimisation works to the users, but not to the testers.

Really... then you're not a true tester.

If you really want to test something you do it over and over again and test it in every possible way until you get the optimal results and note what needs to be done to achieve those optimal results. Car mags do it, Gun mags do it, camera mags... they test at factory defaults but mention they have options to improve the settings.

Phil even shoots his test shots at the lens optimal aperture so why can't he hit the menu buttons and play with other modes? Yes, he does state that he physically shoots his 50mm 1.4's at differing apertures and keeps the results from the best one even though they contrast heavily from Photozones tests of the lenses.

RH, the only time you say anything is if it makes you seem right and everyone else seem wrong. Thats fine and dandy, but it fills the site with more nonsense to dig through. You've already switched to Canon, but still tag along for some reason and again thats fine, but you can't base factual results without really comparing the camera yourself. Basically you're breaking the law... Pop photo and all of them have their results and you're using them for your site, isn't their some sort on infringement there. It's as if Motor Trend used another mags review and summarized it, but swung it in their opinion. IDK RH, you're really stretching your name as measurebator.. you're more of a copy&paster.

[RiceHigh]

Oh he did.

He allegedly bought a Canon 5D just so he could say how much better it was in every possibly way. Of course no proof is offered, so we are supposed to take his word for it.

Just put him on your blocked list. Life is so much nicer.

I can never understand why people need to believe that mankind can only own ONE camera brand at a time. Would this be called "brand loyalty"?

I have a 5D and a small Canon system comprised of two lenses (1 prime and 1 L zoom, 1 flash and 1 remote switch);

I have a K100D with a *huge* Pentax system (12 original Pentax AF lenses and other K-mount lenses and convertors/adaptors, all Pentax remote switches ever made! two IR controller, still 4 bodies, original Pentax flash gun and accessories and so on - that cannot be mentioned all here..;

I also have Pentax, Oly, Fuji, Sony digital photographic equipment as well as Ricoh, Canon, Pentax, Olympus or even Agfa film SLR/P&S/Rangefinder cameras. So what then? What conclusion can you make here? Afterall, the truth is I am *still* using my Pentax gear despite it has quite some persistent unresolved problems for *their DSLR bodies*.

Last Sunday, I was still using my MZ-S to shoot a roll of slide. Weeks ago, I was still using my K100D to shoot in a concert:-

https://www.pentaxforums.com/forums/pentax-dslr-discussion/19649-money-more-c...tml#post169126

All in all, we have not signed a life-time contract with any camera brand that we can only use their gear alone, have you? I think that (blind) "brand loyalty" is just a very stupid thing. There is no company on Earth loyal to their customers but in fact only some customers are unexplainably loyal to their company as a customer!