Interesting DR Test and BG Noise Test of the K20D

[RiceHigh]

The Pentax K20D: a RAW review, page 3

DR = Dynamic Range
BG = BackGround

[Zewrak]

Lots of babbling, no pictures.

[RaduA]

Actually this is months old ...

Radu

[attack11]

so glad i didn't jump on the k20d bandwagon.

[Gooshin]

the author starts off his essay by comparing the human eye to a digital sensor and is expecting them to be idnetical,

lets see how "well lit" his room is after someone blinds him with a 500 watt flashlight, hahaha.


testing dynamic range in extreme cases like a dark room with a sunny window slit is like seeing if your lexan water bottle will break if you drive an 18 wheeler over it.

while techincaly correct (or correct as far as his own personaly analysis of his own work goes), in his own conclusion it seems that the only people that would care for this are astrophotographers and what not, and i guess Pentax is not the best tool for astrophotography, ohh well.

[attack11]

Originally Posted by Gooshin

the author starts off his essay by comparing the human eye to a digital sensor and is expecting them to be idnetical,

lets see how "well lit" his room is after someone blinds him with a 500 watt flashlight, hahaha.


testing dynamic range in extreme cases like a dark room with a sunny window slit is like seeing if your lexan water bottle will break if you drive an 18 wheeler over it.

while techincaly correct (or correct as far as his own personaly analysis of his own work goes), in his own conclusion it seems that the only people that would care for this are astrophotographers and what not, and i guess Pentax is not the best tool for astrophotography, ohh well.

actually, i'm the type of photographer that cares about this type of test. almost every k20d noise test i've seen on here hasn't been much of a test for the conditions and style i shoot.

i under expose at low iso with the intention of pushing to get high shutter and low noise; i know the limits of my k10d but have been curious about the k20d. i see it's worse. i'll wait for the k30d.

[Gooshin]

Originally Posted by attack11

actually, i'm the type of photographer that cares about this type of test. almost every k20d noise test i've seen on here hasn't been much of a test for the conditions and style i shoot.

i under expose at low iso with the intention of pushing to get high shutter and low noise; i know the limits of my k10d but have been curious about the k20d. i see it's worse. i'll wait for the k30d.

well, i just got a K20D and i'm waiting for my boss to lend me his D300, then i will test the crap out of both of them in identical conditions.

[attack11]

Originally Posted by Gooshin

well, i just got a K20D and i'm waiting for my boss to lend me his D300, then i will test the crap out of both of them in identical conditions.

shoot a bright room with a very dark area. that should show a difference.

[morfic]

Originally Posted by attack11

actually, i'm the type of photographer that cares about this type of test. almost every k20d noise test i've seen on here hasn't been much of a test for the conditions and style i shoot.

i under expose at low iso with the intention of pushing to get high shutter and low noise; i know the limits of my k10d but have been curious about the k20d. i see it's worse. i'll wait for the k30d.

why underexpose at low iso to get a higher shutter? makes no sense, you reduce the signal in the shot while the noise stays constant, then you push in PP and there you push signal and noise.....
I could see that thinking right between 1600 and 3200 if you say you can push it better than what pentax does to get 3200, but that is not what you use pushing for.....

[attack11]

Originally Posted by morfic

why underexpose at low iso to get a higher shutter? makes no sense, you reduce the signal in the shot while the noise stays constant, then you push in PP and there you push signal and noise.....
I could see that thinking right between 1600 and 3200 if you say you can push it better than what pentax does to get 3200, but that is not what you use pushing for.....

if you have to ask, you don't understand your camera or what the end result is.

for instance, when shooting friends at a bar i use this method to avoid horrible shadow noise at higher iso.

when shooting hockey i get my 1/250 at iso 400 in the worst light conditions and it looks far better than anyone shooting high iso.

i do this for anything that's moving fast.

it's a very old technique that a generation seems to have forgotten in the quest for silly iso's with horrible noise.

it can also be the difference between 1/5 and 10sec; or a 'lost shot' due to hand shake. it also gives you higher iq, especially if you disable in body noise reduction and do that as your final pp step with something like noise ninja.

i'm all about iq.

[attack11]

oh yeah, if this still makes no sense re-read what that article is talking about. the 'push' available with my k10d is greater than a k20d; i'll get a better final image.

[morfic]

Originally Posted by attack11

if you have to ask, you don't understand your camera or what the end result is.

for instance, when shooting friends at a bar i use this method to avoid horrible shadow noise at higher iso.

when shooting hockey i get my 1/250 at iso 400 in the worst light conditions and it looks far better than anyone shooting high iso.

i do this for anything that's moving fast.

it's a very old technique that a generation seems to have forgotten in the quest for silly iso's with horrible noise.

it can also be the difference between 1/5 and 10sec; or a 'lost shot' due to hand shake. it also gives you higher iq, especially if you disable in body noise reduction and do that as your final pp step with something like noise ninja.

i'm all about iq.

The reason i don't see the need to do it is simple, the K20D is by far better than anything i am used to from my previous camera, thus your effort seems unnecessary in that relationship, and i pushed Provia 400X 2 stops before since i saw no ISO1600 film giving me suitable results, there i saw a necessity, which i don't see on the K20D, that's all. It's good enough to leave noise reduction always off and treat it in PP *if* required.

If you are all about IQ, then i would think you would want to test the K20D to your K10D yourself under the conditions you use it, to make sure you don't miss out on an opportunity just because perhaps the reviewer did not test to your standards? ever thought about that?

[RiceHigh]

Originally Posted by morfic

The reason i don't see the need to do it is simple, the K20D is by far better than anything i am used to from my previous camera, thus your effort seems unnecessary in that relationship, and i pushed Provia 400X 2 stops before since i saw no ISO1600 film giving me suitable results, there i saw a necessity, which i don't see on the K20D, that's all. It's good enough to leave noise reduction always off and treat it in PP *if* required.

I think you can't compare slide to digital as slide is well known to be of narrower latitude than any DSLR.

If you are all about IQ, then i would think you would want to test the K20D to your K10D yourself under the conditions you use it, to make sure you don't miss out on an opportunity just because perhaps the reviewer did not test to your standards? ever thought about that?

I think IQ consists of mainly four things: resolution, colors, DR and then noise. If you are to compare against the K10D with the K20, I think the 20 will win for resolution and colors and maybe noise as well at higher ISO but DR is not better according to the test.

Do note that the noise is better (than the K10) also doesn't mean that it is good enough, when it is compared with other models like 40D and D300.

[attack11]

Originally Posted by morfic

If you are all about IQ, then i would think you would want to test the K20D to your K10D yourself under the conditions you use it, to make sure you don't miss out on an opportunity just because perhaps the reviewer did not test to your standards? ever thought about that?

i'd rather not gamble around $1k on "tests" in evenly lit rooms. that's just me. however, thanks to the person referenced in the link above, i don't need to do any tests. his program did it for me.

[RiceHigh]

Originally Posted by RaduA

Actually this is months old ...

Radu

There is no text.