Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-19-2009, 02:17 AM   #91
Senior Member




Join Date: Jun 2008
Location: Atlanta, Georgia
Photos: Gallery
Posts: 182
This is where I have an issue with DxOMark composite numbers using subjectively set thresholds. Say e.g. that their Low-Light ISO score would have been given by the following definition:

"Low-Light ISO is then the highest ISO setting for the camera such that the SNR reaches this 30dB value while keeping a good dynamic range of 8EVs and a color depth of 18bits."

Here I put in my (subjective) assumption on what dynamic range is acceptable in low light situations, and suddenly their "composite DxOMark" for low light would be almost identical for these cameras.

As a reference one could use the dynamic range of the popular slide film Velvia 50 which is said to be 5-6 EV (in good situations). So if 5-6 EV is good enough to capture fantastic images on that film, should not 8 EV be more than enough, at least in low light situations?

Another issue I have with their dynamic range measure, is that it only measures the noise in the lowest black level and not the capability to resolve different tones at that level. It would be rather easy to improve the DxOMark score for dynamic range by having a non-linear gain in the "darkest" part of the curve such that the gain is lower for the low signals. This would reduce noise in the darkest areas, but at the expense of the possibility distinguish different tones at the darkest areas (which is not captured by DxOMark for dynamic range)

So for me, the best way is to look at the images (produced under same conditions), and then determine which look more noisy. And then I see no difference. Likely because there are only very small portions of the image that is represented by the extreme dark tones and the impression of noise is given by the noise in the other "99%" of the image that is not "pitch dark".

So to answer the original question "K7 high ISO - will it be competitive?"

For me the answer is Yes - since already the K20D is competitive.

Best regards,
Haakan

07-19-2009, 05:23 AM   #92
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,596
Original Poster
QuoteOriginally posted by Haakan Quote
So to answer the original question "K7 high ISO - will it be competitive?"

For me the answer is Yes - since already the K20D is competitive.
By now, after a thousand Pentax pixel-peepers have looked at the two camera's output, the consesnus seems to be that K20D = K7 in terms of sensor performance across all measures, give or take a few percentage points for one variable or another.

Assuming nothing weird happens for the K7 once it goes through the DXOLabs, I am also assuming that the K7 will get about the same low-light ISO score as the K20D.

So that certainly makes the K7 competitive with the K20D

The issues of dynamic range etc I understand a little bit but don't feel expert enough to debate. But while DXO may have set DR at a fairly high standard, I can only assume they had a reasonable basis for doing so, and they have also at least applied that measure consistently to all sensors tested.
07-19-2009, 06:19 AM   #93
Senior Member




Join Date: Jun 2008
Location: Atlanta, Georgia
Photos: Gallery
Posts: 182
I think it has been reported that chroma noise is reduced with K7, but I agree that also K20D is likely to be as good performer as D90 and EOS50D.

When it comes to DR, there are contradictory numbers given by different review sites. E.g. DP Review gives the following numbers for usable DR:

K20D: 9.1 EV
D90: 8.3 EV
EOS50D: 8.3 EV

Using RAW and some "extreme" ACR settings they could get slightly above 10 EV on K20D and 9.7 EV for D90.

So they rate the K20D to outperform D90 in terms of DR

This just highlights the concern I have of using DR as the cut off performance measure for Low Light performance.

I think most people perceive low light performance through how "noisy" the image looks like (i.e. more related to the SNR value), and to a much lesser extent the DR. And looking at the "RAW" images (which includes a grey scale for DR), IMHO I feel that actually the K20D gives the overall cleanest perception (based on a cleaner look in the lighter areas)

But since the differences are very subtle, I can accept that someone with a preference for another camera brand, are equally correct in claiming a small advantage for their brand.

In reality we are talking about small differences that are over shadowed by other factors (such as proper exposure) and far smaller than should to be driving a user to one or the other brand, or have a brand be rated as not competitive.

Best regards,
Haakan
12-11-2009, 09:15 AM   #94
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,596
Original Poster
FINALLY DXO Labs have updated their database to include the K-7.

The results for the K-7 aren't good. It's low-light ISO score relative to most of it's competitors did not impress DXO Labs... In fact across the range - on color-depth, dynamic range and low-light ISO, the K-7 scores are below those of the K20D...

It will be interesting to see how the K-x fares. Cameras with a similar Sony sensor (the Sony A500 and the Nikon D300s) have fared rather well. Not surprisingly, due to it's crowded sensor, the Canon 7D's numbers, while good, weren't that great either.

Pls don't shoot the messenger...



12-11-2009, 10:02 AM   #95
Veteran Member




Join Date: Jan 2009
Location: Frozen white North
Photos: Albums
Posts: 845
So if I want an upgrade to my K10D, the K20D is a better choice than the K-7 for low-light, high-ISO photography?

I haven't paid much attention to it, but are K20D photos generally less noisy at high ISO?
12-11-2009, 10:11 AM   #96
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,596
Original Poster
QuoteOriginally posted by sebberry Quote
So if I want an upgrade to my K10D, the K20D is a better choice than the K-7 for low-light, high-ISO photography?

I haven't paid much attention to it, but are K20D photos generally less noisy at high ISO?
Generally the answer would be, yes, I think.

According to DXOMark, the K20D is better than the K10D at low-light ISO by quite a margin, which usually can be translated into better IQ at higher ISO. If you want the full picture, you can look at a comparison of the results for the K10D and K20D at the DXOMark database Pentax page:

Pentax

If you really are looking to upgrade for low-light high ISO, get the Pentax K-x. It hasn't been tested by DXOMark yet, but it has the same sensor as the Sony A500 so it's performance at low-light ISO should be at least as good as the Sony on that metric. My guess though is that the K-x will probably do even better than the A500.
12-11-2009, 10:18 AM   #97
Veteran Member




Join Date: Jan 2009
Location: Frozen white North
Photos: Albums
Posts: 845
QuoteOriginally posted by rawr Quote
If you really are looking to upgrade for low-light high ISO, get the Pentax K-x. It hasn't been tested by DXOMark yet, but it has the same sensor as the Sony A500 so it's performance at low-light ISO should be at least as good as the Sony on that metric. My guess though is that the K-x will probably do even better than the A500.
If the K20D and the K-7 are on-par with eachother then I'd be inclined to pick the K20D.

It seems like the K-x is turning into a good little performer. Too bad it has one of the cheapest feeling bodies around.
12-11-2009, 10:38 AM   #98
Veteran Member
PentaxPoke's Avatar

Join Date: Jan 2008
Location: Oklahoma
Posts: 1,411
QuoteOriginally posted by rawr Quote
Not surprisingly, due to it's crowded sensor, the Canon 7D's numbers, while good, weren't that great either.
Wow you are right about the K-7. That is not good. I have to disagree about the 7D though. It is an 18MP sensor, and it is still 3% better than the 12MP D300s, and 51% better than our 15MP K-7! It is commonly thought that increasing megapixels on a given sensor size increases noise. So Canon increased their MP by 20%, yet decreased their noise by 16%. I think that is impressive. What is odd is the D90, which I believe uses the same sensor as the D300s.

To my understanding that ISO benchmark number is the iso where the signal to noise ratio reaches 30db. So in other words the signal to noise ratio on the K-7 at an ISO of 536 will be the same as that on the 7D at an iso of 809. That is approaching one full stop, which is about what people have qualitatively concluded based in comparometer and other picture comparisons.


Last edited by PentaxPoke; 12-11-2009 at 11:11 AM.
12-11-2009, 03:22 PM   #99
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,596
Original Poster
QuoteOriginally posted by PentaxPoke Quote
Wow you are right about the K-7. That is not good. I have to disagree about the 7D though. It is an 18MP sensor, and it is still 3% better than the 12MP D300s, and 51% better than our 15MP K-7! It is commonly thought that increasing megapixels on a given sensor size increases noise. So Canon increased their MP by 20%, yet decreased their noise by 16%. I think that is impressive. What is odd is the D90, which I believe uses the same sensor as the D300s.
I think the K7 results, while disappointing, were broadly in accord with people's experience. It is still a fine instrument and you can get great results with it, but for a flagship system it would have been nice for its numbers to have been better, naturally.

In terms of the Canon 7D, I guess I was comparing the 7D with it's predecessors like the 50D rather than the K7, and also referencing some of the rather exuberant commentary I had read about how it's low-light ISO matched some of the full-frame Canon's. Clearly while it's a very clean and efficient design, and quite a technical achievement, the 7D still isn't APS-C class leading in performance on these metrics.

I agree the D90/D5000 vs the D300s result is interesting. If I had paid a lot to upgrade from the D90 to the D300s in search of superior low-light IQ, I'd be disappointed.
12-11-2009, 03:38 PM   #100
Veteran Member
Pentaxor's Avatar

Join Date: May 2009
Location: Vancouver, B.C.
Photos: Gallery
Posts: 6,513
QuoteOriginally posted by rawr Quote
I think the K7 results, while disappointing, were broadly in accord with people's experience. It is still a fine instrument and you can get great results with it, but for a flagship system it would have been nice for its numbers to have been better, naturally.

In terms of the Canon 7D, I guess I was comparing the 7D with it's predecessors like the 50D rather than the K7, and also referencing some of the rather exuberant commentary I had read about how it's low-light ISO matched some of the full-frame Canon's. Clearly while it's a very clean and efficient design, and quite a technical achievement, the 7D still isn't APS-C class leading in performance on these metrics.

I agree the D90/D5000 vs the D300s result is interesting. If I had paid a lot to upgrade from the D90 to the D300s in search of superior low-light IQ, I'd be disappointed.
I believe this is also the reason why Pentax should release a FF body with a great sensor. there is nothing like doubling the effect or advantage. true that the K-7 can manage it's own and could produce great results even at ISO 3200 with proper post-processing techniques (elimination unwanted noise). if ever Pentax releases a FF body, I'd be happy having the K-7 as a secondary body, unless Pentax decides to release a better sensor with better specs than the K-7 that I might consider selling it. otherwise, the K-7 will do for 3 more years.
12-11-2009, 04:39 PM   #101
Veteran Member




Join Date: Jul 2009
Location: kobe/japan
Posts: 510
QuoteOriginally posted by rawr Quote
FINALLY DXO Labs have updated their database to include the K-7.

The results for the K-7 aren't good. It's low-light ISO score relative to most of it's competitors did not impress DXO Labs... In fact across the range - on color-depth, dynamic range and low-light ISO, the K-7 scores are below those of the K20D...

It will be interesting to see how the K-x fares. Cameras with a similar Sony sensor (the Sony A500 and the Nikon D300s) have fared rather well. Not surprisingly, due to it's crowded sensor, the Canon 7D's numbers, while good, weren't that great either.

Pls don't shoot the messenger...

so k7 sensor ends up in bottom of the pile. Nothing surprising for me, have been saying this for some time. Seeing the ranking of d5000 there is going to be HUGE gape between rankings of k7 and kx. Just a matter a little wait, it will be in black and white soon.
12-11-2009, 05:04 PM   #102
Pentaxian
LeDave's Avatar

Join Date: Jun 2009
Location: Minneapolis - St. Paul
Photos: Gallery
Posts: 2,896
QuoteOriginally posted by sebberry Quote
If the K20D and the K-7 are on-par with eachother then I'd be inclined to pick the K20D.

It seems like the K-x is turning into a good little performer. Too bad it has one of the cheapest feeling bodies around.
One of the cheapest bodies around? Are you smoking something? I used to have the K2000 and it is in no way cheap, sure it is plastic but it feels just as solid as my K-7 if you don't take the magnesium alloy in consideration. It feels very well built and the buttons are very solid just like all the previous WR Pentax DSLR's, the only difference in build if I were to think of is the K-x and K2000 aren't weather-sealed. Many people are put off by the K-x's build just because it's not a WR and is plastic compared to the K-7, but if they were to actually see and touch it in person, they would definitely know how well it is built.
12-11-2009, 05:18 PM   #103
Pentaxian
raider's Avatar

Join Date: Jul 2007
Location: Perth, Australia
Photos: Gallery
Posts: 1,931
It would seem that the k20D has better ISO performance than K-7. I am curious how could this be as both cams spot the same sensor. Hmm....

While K20D may be suitable for night shots and stuff, it also lack an Auto focus beam. And before the arrival of the K-7, there are plenty of "comments" that the k20D is unable to lock focus accurately and fast enough in low light situations. In this case, would the high iso performance be sorta redundant then? I mean, u get high iso performance but u might get misfocus pictures.

As for why would the K-x has high iso performance and pentax's flagship K-7 doesnt, the best I can think of is the "analogy" between Starship Defiant vs Starship Enterprise D. Any trekkies here?

Starship Enterprise is the flagship. It is bigger, better, more features, more powerful and faster, BUT
Starship Defiant is design to counter the Borg and it has the cloaking device, ablative armour and is the testing ground for the latest in weaponary and technology.

So the defiant is better in some ways but still loses out in others. U really need both vessels in the Federation space to maintain peace against the romulans.

So there you are, the K-x is the defiant and K-7 is the enterprise. The K-7 has dual dials, AF beam, batt grip, AF tuning, wr body, etc etc. The K-x is smaller, has better iso but otherwise lacks everything the flagship has. sorta

Sorry, out of topics here. Just want to get this one out
12-11-2009, 06:03 PM   #104
Veteran Member
er1kksen's Avatar

Join Date: Oct 2008
Location: Staten Island, NY
Photos: Gallery | Albums
Posts: 1,663
QuoteOriginally posted by sebberry Quote
If the K20D and the K-7 are on-par with eachother then I'd be inclined to pick the K20D.

It seems like the K-x is turning into a good little performer. Too bad it has one of the cheapest feeling bodies around.
Have you handled a Canon rebel lately? My K-x feels about as "solid" construction-wise as my K20D did, just less dense and weighty. The rubber used on the K20D grip was a lot nicer though.

QuoteOriginally posted by raider Quote
It would seem that the k20D has better ISO performance than K-7. I am curious how could this be as both cams spot the same sensor. Hmm....
Well, as you can see, DXOmarks has a habit of giving different scores to identical sensors. See also K-m vs. K200D, K20D vs. GX-20, Oly E-410 vs. E-510, Nikon D90 vs. D300s vs. D5000. DXOmark scores are simply not that relevant to real-world performance.

HOWEVER: you might be surprised to learn that the K20D and K-7 do not, in fact, use the same sensor. They share the same approximate resolution, but the K-7's sensor was further optimized for readout speed, enabling 5fps and video. The high-ISO noise from each sensor differs in both quantity and character.
12-11-2009, 06:18 PM   #105
Veteran Member
Pentaxor's Avatar

Join Date: May 2009
Location: Vancouver, B.C.
Photos: Gallery
Posts: 6,513
QuoteOriginally posted by raider Quote
It would seem that the k20D has better ISO performance than K-7. I am curious how could this be as both cams spot the same sensor. Hmm....

While K20D may be suitable for night shots and stuff, it also lack an Auto focus beam. And before the arrival of the K-7, there are plenty of "comments" that the k20D is unable to lock focus accurately and fast enough in low light situations. In this case, would the high iso performance be sorta redundant then? I mean, u get high iso performance but u might get misfocus pictures.

As for why would the K-x has high iso performance and pentax's flagship K-7 doesnt, the best I can think of is the "analogy" between Starship Defiant vs Starship Enterprise D. Any trekkies here?

Starship Enterprise is the flagship. It is bigger, better, more features, more powerful and faster, BUT
Starship Defiant is design to counter the Borg and it has the cloaking device, ablative armour and is the testing ground for the latest in weaponary and technology.

So the defiant is better in some ways but still loses out in others. U really need both vessels in the Federation space to maintain peace against the romulans.

So there you are, the K-x is the defiant and K-7 is the enterprise. The K-7 has dual dials, AF beam, batt grip, AF tuning, wr body, etc etc. The K-x is smaller, has better iso but otherwise lacks everything the flagship has. sorta

Sorry, out of topics here. Just want to get this one out
true. there is no debate regarding the High ISO performance of the k-x from the start. nobody would argue that one certainly.

however, regarding the the noise between the K20D and the K-7, the difference maybe due to the sensor filtering difference, not necessarily the sensor. Yes, it does appear that the K-7 is noisier than the K20D, but what does look is that the luminance noise is better controlled as compared to the K20D. from a qualitative stand point, the K-7 is better, but as per quantitative, the K20D is better. I dont know if DxO performs or considers a qualitative standpoint. having said that, both cameras still need a good RAW post-processing job when shooting above ISO 1600.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
aps-c, camera, d90, dslr, iso, k7, k7 high iso, pentax, performance, photography
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
There is more to the K-5 than high ISO, right? jake14mw Pentax K-5 5 10-16-2010 09:33 AM
People High ISO K7 milesy Post Your Photos! 6 10-11-2010 01:36 AM
K7 at high ISO zelovoc Pentax DSLR Discussion 30 09-14-2010 07:23 AM
K-7 high ISO vs K20D high ISO supa007 Pentax DSLR Discussion 72 05-10-2010 04:24 PM
Night photography with K10D - High ISO short exposure VS Low ISO long exposure pw-pix Pentax DSLR Discussion 10 02-03-2008 01:37 AM



All times are GMT -7. The time now is 12:23 AM. | See also: NikonForums.com, part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top