5d vs k20d

Gooshin · 02-12-2008, 11:42 AM

Originally posted by attack11

that's impressive, my k10d doesn't come close to that in low light. 1600 is completely unusable, and we're talking about a k20d; not a k100d.

on paper, the CMOS unit surpasses both of the current CCD's.

Personaly i have no doubt that the new unit will outperform the K100D's 6mp unit, let alone the K10D's in terms of high ISO usage.

and if one was to be extreamly pessimistic, it cant possibly be worse than either, Pentax (or any other manufacturer) would not be so foolish at this stage in the game to permit such a thing.

Gooshin · 02-12-2008, 11:46 AM

Originally posted by attack11

that's impressive, my k10d doesn't come close to that in low light. 1600 is completely unusable, and we're talking about a k20d; not a k100d.

on the other hand, here is another 3200 ISO shot from my K100D, however this time around there simply was not any light to work with at all.

i'm only trying to point out that the current K20D iso pictures circling around could mean anything or nothing, i'm waiting for someone on this forum to do a comprehensive "try my very best" high ISO test with the K20D to settle the myths

however until that time i will trust in the spec sheets, and so far it looks promising.

jeffkrol · 02-12-2008, 12:07 PM

Originally posted by CarlCanary

Just curious how people think the k20d image quailty will compare with the canon 5d

Not my words and just a preliminary report of alleged performance of the K20 at various iso's.
I trust John's analysis and reasoning...... May not beat 5d at high iso's but may beat it in low iso's(my expectation) but it will really boil down to other factors ie resolution ect.. Let's all just wait and see... Anyone have an iso 100 K20d RAW file?
I have a K20D ISO 1600 RAW or two here; the pixel-level read noise is about 11.8 12-bit ADUs, which, while not as good as Canon DSLRs or the Nikon D3, is better than pretty much all others DSLRs out there (and about the same as the D300). with slightly more pixels than the D300, it should have slightly less image read noise.

The real unique beauty of the K20D, as I expect it, is its performance at low ISO, which should surpass any APS DSLR at the image level in read noise, and roughly equal them in shot noise, and surpass all DSLRs in image read noise at its lowest ISO. This assumes that the K20D has a similar readout system as the K10D (I didn't get any ISO 100 RAWs yet), which seems to be one single but very clean readout at base ISO at 22 bits, and then doing the higher ISOs with digital arithmetic.

Re: So if more megapixels is not the answer...: News Discussion Forum: Digital Photography Review
EDIT:Guess this is just an edited repeat of an earlier post......my apologies... but I'll leave it for now

benjikan · 02-12-2008, 01:59 PM

Originally posted by *isteve

It would be very difficult to tell them apart on poster size prints if they were processed to produce the same colour and saturation levels.

Ditto.... have had images blown up to 3.5 x 5 meters shot on a Canon 20D and Pentax K20D and the results were spectacular. I would beg to differ if anyone could tell the difference between a Canon 1DS Mk ll and the K20D.

Ben

SteveB · 02-12-2008, 01:59 PM

I heard that the K20D base ISO is actually 200, even though 100 is the lowest ISO.

benjikan · 02-12-2008, 02:02 PM

Originally posted by attack11

5d at iso3200. k20d doesn't come close.

exif data

the 5d is the better camera in every way. tighter/proper dof/bokeh, better iso performance. aps-c is a compromise, that anyone can learn to work with. i love my k10d, but if i could get a full frame at a reasonable price i'd switch in an instance because it'll let me get the images i want.

f1.4 will be f1.4

Have you tested it directly with the K20D with the firmware Version 1.00 yet? Until you do, I suggest that you postpone judgment until doing so. As an Ex Canon sponsor, I used all of their gear and the Pentax K20D competes admirably.

Ben

benjikan · 02-12-2008, 02:05 PM

Originally posted by jeffkrol

Not my words and just a preliminary report of alleged performance of the K20 at various iso's.
I trust John's analysis and reasoning...... May not beat 5d at high iso's but may beat it in low iso's(my expectation) but it will really boil down to other factors ie resolution ect.. Let's all just wait and see... Anyone have an iso 100 K20d RAW file?
I have a K20D ISO 1600 RAW or two here; the pixel-level read noise is about 11.8 12-bit ADUs, which, while not as good as Canon DSLRs or the Nikon D3, is better than pretty much all others DSLRs out there (and about the same as the D300). with slightly more pixels than the D300, it should have slightly less image read noise.

The real unique beauty of the K20D, as I expect it, is its performance at low ISO, which should surpass any APS DSLR at the image level in read noise, and roughly equal them in shot noise, and surpass all DSLRs in image read noise at its lowest ISO. This assumes that the K20D has a similar readout system as the K10D (I didn't get any ISO 100 RAWs yet), which seems to be one single but very clean readout at base ISO at 22 bits, and then doing the higher ISOs with digital arithmetic.

Re: So if more megapixels is not the answer...: News Discussion Forum: Digital Photography Review
EDIT:Guess this is just an edited repeat of an earlier post......my apologies... but I'll leave it for now

I can provide a RAW 100 iso K20D image. I can send via YouSendIt. Where shoud I send it?

Ben

attack11 · 02-12-2008, 02:38 PM

Originally posted by benjikan

Have you tested it directly with the K20D with the firmware Version 1.00 yet? Until you do, I suggest that you postpone judgment until doing so. As an Ex Canon sponsor, I used all of their gear and the Pentax K20D competes admirably.

Ben

even if the iso performance is close to being as good, you won't get the ff goodness that the 5d delivers. every iso comparison i've seen online shows it's pretty useless at/above 3200. if that can be corrected with firmware updates, great. otherwise it's another k10d imo.

nosnoop · 02-12-2008, 03:00 PM

Originally posted by attack11

even if the iso performance is close to being as good, you won't get the ff goodness that the 5d delivers.

Well, this just doesn't make sense to me.
Are you saying that you can tell FF vs APS-C just from the image without looking at the EXIF?

twinda1 · 02-12-2008, 03:13 PM

Originally posted by nosnoop

Well, this just doesn't make sense to me.
Are you saying that you can tell FF vs APS-C just from the image without looking at the EXIF?

Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus.

Gooshin · 02-12-2008, 03:30 PM

Originally posted by attack11

even if the iso performance is close to being as good, you won't get the ff goodness that the 5d delivers. every iso comparison i've seen online shows it's pretty useless at/above 3200. if that can be corrected with firmware updates, great. otherwise it's another k10d imo.

dont forget the 5D is 3 year old technology, the K20D is utalizing a newly developed chip.

are bigger pixels better than smaller pixels at capturing noise free light, in general, yes

but if someone goes out and develops a smaller pixel that can absorb light just as well, they just solved the problem didnt they?

you know regular CD roms? by your arguments Blue Ray disks are a pointless development.

attack11 · 02-12-2008, 03:37 PM

Originally posted by twinda1

Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus.

exactly. anyone saying there is no difference has never done a real comparison.

Gooshin · 02-12-2008, 03:42 PM

Originally posted by twinda1

Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus.

ofcourse they will be different if you start walking backwards with the cropped camera, thats understood.

and ofcourse the FF camera might produce a better image because you are physicaly closer to the subject if you try to force the same FOV on both cameras

and ofcourse your DOF will be different

but take the image at the same distance from both cameras, and then CROP the 5D to the same FOV as the cropped camera.

still different images?

nosnoop · 02-12-2008, 03:54 PM

Originally posted by twinda1

Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus.

But only if I give you a side-by-side comparison shots taken with two different cameras under the same situations. Otherwise, how can you tell just from any image with no reference or knowledge of lens and aperture used?

nosnoop · 02-12-2008, 03:58 PM

Originally posted by attack11

exactly. anyone saying there is no difference has never done a real comparison.

And that's the point. We are not doing a comparison. We are not giving you a side-by-side FF vs APS-C comparison.

You get difference even if you use two different lenses or two different APS-C cameras from different brand!

All I am saying is if you are given an image without an EXIF, can you tell me for sure whether it is taken by a FF or APS-C camera?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
For Sale - Sold: K20D, Battery Grip, NEW K20D battery, cable remote (Worldwide)	Albert Siegel	Sold Items	6	09-23-2010 08:02 AM
In Canada: Summer Rebates for K20d or K20d plus lens	Pentaxtic	Pentax DSLR Discussion	1	07-17-2009 11:34 AM
Magic Lantern Guides: Pentax K20D and MasterWorks: Jumpstart Guide for the K20D.	Reportage	Troubleshooting and Beginner Help	10	02-12-2009 10:24 PM

02-12-2008, 11:42 AM	#16
Gooshin Veteran Member Join Date: Sep 2007 Location: Toronto, the one in Canada. Posts: 5,610	Originally posted by attack11 that's impressive, my k10d doesn't come close to that in low light. 1600 is completely unusable, and we're talking about a k20d; not a k100d. on paper, the CMOS unit surpasses both of the current CCD's. Personaly i have no doubt that the new unit will outperform the K100D's 6mp unit, let alone the K10D's in terms of high ISO usage. and if one was to be extreamly pessimistic, it cant possibly be worse than either, Pentax (or any other manufacturer) would not be so foolish at this stage in the game to permit such a thing.

02-12-2008, 11:46 AM	#17
Gooshin Veteran Member Join Date: Sep 2007 Location: Toronto, the one in Canada. Posts: 5,610	Originally posted by attack11 that's impressive, my k10d doesn't come close to that in low light. 1600 is completely unusable, and we're talking about a k20d; not a k100d. on the other hand, here is another 3200 ISO shot from my K100D, however this time around there simply was not any light to work with at all. i'm only trying to point out that the current K20D iso pictures circling around could mean anything or nothing, i'm waiting for someone on this forum to do a comprehensive "try my very best" high ISO test with the K20D to settle the myths however until that time i will trust in the spec sheets, and so far it looks promising.

02-12-2008, 12:07 PM	#18
jeffkrol Veteran Member Join Date: Sep 2006 Location: Wisconsin USA Photos: Gallery Posts: 8,434	Originally posted by CarlCanary Just curious how people think the k20d image quailty will compare with the canon 5d Not my words and just a preliminary report of alleged performance of the K20 at various iso's. I trust John's analysis and reasoning...... May not beat 5d at high iso's but may beat it in low iso's(my expectation) but it will really boil down to other factors ie resolution ect.. Let's all just wait and see... Anyone have an iso 100 K20d RAW file? I have a K20D ISO 1600 RAW or two here; the pixel-level read noise is about 11.8 12-bit ADUs, which, while not as good as Canon DSLRs or the Nikon D3, is better than pretty much all others DSLRs out there (and about the same as the D300). with slightly more pixels than the D300, it should have slightly less image read noise. The real unique beauty of the K20D, as I expect it, is its performance at low ISO, which should surpass any APS DSLR at the image level in read noise, and roughly equal them in shot noise, and surpass all DSLRs in image read noise at its lowest ISO. This assumes that the K20D has a similar readout system as the K10D (I didn't get any ISO 100 RAWs yet), which seems to be one single but very clean readout at base ISO at 22 bits, and then doing the higher ISOs with digital arithmetic. Re: So if more megapixels is not the answer...: News Discussion Forum: Digital Photography Review EDIT:Guess this is just an edited repeat of an earlier post......my apologies... but I'll leave it for now

02-12-2008, 01:59 PM	#19
benjikan Veteran Member Join Date: Jan 2007 Location: Paris, France Photos: Gallery Posts: 4,308	Originally posted by *isteve It would be very difficult to tell them apart on poster size prints if they were processed to produce the same colour and saturation levels. Ditto.... have had images blown up to 3.5 x 5 meters shot on a Canon 20D and Pentax K20D and the results were spectacular. I would beg to differ if anyone could tell the difference between a Canon 1DS Mk ll and the K20D. Ben

02-12-2008, 01:59 PM	#20
SteveB Veteran Member Join Date: Mar 2007 Location: U.K. Posts: 685	I heard that the K20D base ISO is actually 200, even though 100 is the lowest ISO.

02-12-2008, 02:02 PM	#21
benjikan Veteran Member Join Date: Jan 2007 Location: Paris, France Photos: Gallery Posts: 4,308	Originally posted by attack11 5d at iso3200. k20d doesn't come close. exif data the 5d is the better camera in every way. tighter/proper dof/bokeh, better iso performance. aps-c is a compromise, that anyone can learn to work with. i love my k10d, but if i could get a full frame at a reasonable price i'd switch in an instance because it'll let me get the images i want. f1.4 will be f1.4 Have you tested it directly with the K20D with the firmware Version 1.00 yet? Until you do, I suggest that you postpone judgment until doing so. As an Ex Canon sponsor, I used all of their gear and the Pentax K20D competes admirably. Ben

02-12-2008, 02:05 PM	#22
benjikan Veteran Member Join Date: Jan 2007 Location: Paris, France Photos: Gallery Posts: 4,308	Originally posted by jeffkrol Not my words and just a preliminary report of alleged performance of the K20 at various iso's. I trust John's analysis and reasoning...... May not beat 5d at high iso's but may beat it in low iso's(my expectation) but it will really boil down to other factors ie resolution ect.. Let's all just wait and see... Anyone have an iso 100 K20d RAW file? I have a K20D ISO 1600 RAW or two here; the pixel-level read noise is about 11.8 12-bit ADUs, which, while not as good as Canon DSLRs or the Nikon D3, is better than pretty much all others DSLRs out there (and about the same as the D300). with slightly more pixels than the D300, it should have slightly less image read noise. The real unique beauty of the K20D, as I expect it, is its performance at low ISO, which should surpass any APS DSLR at the image level in read noise, and roughly equal them in shot noise, and surpass all DSLRs in image read noise at its lowest ISO. This assumes that the K20D has a similar readout system as the K10D (I didn't get any ISO 100 RAWs yet), which seems to be one single but very clean readout at base ISO at 22 bits, and then doing the higher ISOs with digital arithmetic. Re: So if more megapixels is not the answer...: News Discussion Forum: Digital Photography Review EDIT:Guess this is just an edited repeat of an earlier post......my apologies... but I'll leave it for now I can provide a RAW 100 iso K20D image. I can send via YouSendIt. Where shoud I send it? Ben

02-12-2008, 02:38 PM	#23
attack11 Veteran Member Join Date: Apr 2007 Location: Ottawa, ON - Canada Photos: Gallery Posts: 658	Originally posted by benjikan Have you tested it directly with the K20D with the firmware Version 1.00 yet? Until you do, I suggest that you postpone judgment until doing so. As an Ex Canon sponsor, I used all of their gear and the Pentax K20D competes admirably. Ben even if the iso performance is close to being as good, you won't get the ff goodness that the 5d delivers. every iso comparison i've seen online shows it's pretty useless at/above 3200. if that can be corrected with firmware updates, great. otherwise it's another k10d imo.

02-12-2008, 03:00 PM	#24
nosnoop Veteran Member Join Date: Feb 2007 Posts: 1,299	Originally posted by attack11 even if the iso performance is close to being as good, you won't get the ff goodness that the 5d delivers. Well, this just doesn't make sense to me. Are you saying that you can tell FF vs APS-C just from the image without looking at the EXIF?

02-12-2008, 03:13 PM	#25
twinda1 Senior Member Join Date: Jan 2008 Location: Boston Posts: 142	Originally posted by nosnoop Well, this just doesn't make sense to me. Are you saying that you can tell FF vs APS-C just from the image without looking at the EXIF? Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus.

02-12-2008, 03:30 PM	#26
Gooshin Veteran Member Join Date: Sep 2007 Location: Toronto, the one in Canada. Posts: 5,610	Originally posted by attack11 even if the iso performance is close to being as good, you won't get the ff goodness that the 5d delivers. every iso comparison i've seen online shows it's pretty useless at/above 3200. if that can be corrected with firmware updates, great. otherwise it's another k10d imo. dont forget the 5D is 3 year old technology, the K20D is utalizing a newly developed chip. are bigger pixels better than smaller pixels at capturing noise free light, in general, yes but if someone goes out and develops a smaller pixel that can absorb light just as well, they just solved the problem didnt they? you know regular CD roms? by your arguments Blue Ray disks are a pointless development.

02-12-2008, 03:37 PM	#27
attack11 Veteran Member Join Date: Apr 2007 Location: Ottawa, ON - Canada Photos: Gallery Posts: 658	Originally posted by twinda1 Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus. exactly. anyone saying there is no difference has never done a real comparison.

02-12-2008, 03:42 PM	#28
Gooshin Veteran Member Join Date: Sep 2007 Location: Toronto, the one in Canada. Posts: 5,610	Originally posted by twinda1 Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus. ofcourse they will be different if you start walking backwards with the cropped camera, thats understood. and ofcourse the FF camera might produce a better image because you are physicaly closer to the subject if you try to force the same FOV on both cameras and ofcourse your DOF will be different but take the image at the same distance from both cameras, and then CROP the 5D to the same FOV as the cropped camera. still different images?

02-12-2008, 03:54 PM	#29
nosnoop Veteran Member Join Date: Feb 2007 Posts: 1,299	Originally posted by twinda1 Absolutely; for the same scene (that is, field-of-view is the same), and the same aperture, the FF image will have shallower defpt-of-field at the plane of focus. But only if I give you a side-by-side comparison shots taken with two different cameras under the same situations. Otherwise, how can you tell just from any image with no reference or knowledge of lens and aperture used?

02-12-2008, 03:58 PM	#30
nosnoop Veteran Member Join Date: Feb 2007 Posts: 1,299	Originally posted by attack11 exactly. anyone saying there is no difference has never done a real comparison. And that's the point. We are not doing a comparison. We are not giving you a side-by-side FF vs APS-C comparison. You get difference even if you use two different lenses or two different APS-C cameras from different brand! All I am saying is if you are given an image without an EXIF, can you tell me for sure whether it is taken by a FF or APS-C camera?