Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
next PZ lens test report: Pentax SMC DA 35mm f/2.4 AL
#41
[quote name='Christos' timestamp='1303320073' post='7790']

Klaus I quite don't understand what the AA filter does?

[/quote]



Well, its function is the suppression of moiree effects (Anti-Aliasing).

This is an example for a moiree effect:

[url="http://en.wikipedia.org/wiki/File:Moire_on_parrot_feathers.jpg"]http://en.wikipedia.org/wiki/File:Moire_on_parrot_feathers.jpg[/url]

The AA filter does so by softening - it is a low-pass-filter in front of the sensor.





Your 5D II has a very weak AA filter for instance which is why you perceive it as very sharp. Some of this is "false" detail and moirees at times but my personal opinion on this is ... who cares as long as things look sharp. Others disagree here.
#42
As someone who has used both cameras extensively I found the k-5 to offer more usable high-iso shots due to lack of banding. The amount of vertical and horizontal banding in the 5d mark II at high sensitivities (ISO 6400+) in both stills and video is frankly ridiculous considering the price of the camera and how easily it may have been avoided by the manufacturer. The banding is also painfully present at low-iso images limiting the practical DR of the camera.

The lack of seemingly standard basic features like customizable auto-ISO, quick image magnification and zoomed multi-shot comparison (amongst others) was also disappointing on the 5d. Frankly the main advantage on the 5d is the size of its sensor and the different rendering it offers compared to APS-C. Aside from that, the lack of features and poor attention to quality in certain areas makes a camera not worthy of its premium price.



[quote name='Christos' timestamp='1303308552' post='7784']

Well received the K5 and DA* 16-50 2.8 and sent it back for a refund sorry but I didn't like it, and all that BS you here in pentax forums on how it is on par with FF is nonsense, my 5D II does circles around it in IQ,, even my 7D had better detailed images than the K5.

[/quote]
#43
LMAO, my 5D has no banding at all, also the K5 cannot compete with the 5D in IQ I have side by side test to confirm this as you probably don't, the 5D II has a 1 stop advantage at High ISO but the detail is huge difference.
#44
[quote name='Christos' timestamp='1303452600' post='7838']

LMAO, my 5D has no banding at all, also the K5 cannot compete with the 5D in IQ I have side by side test to confirm this as you probably don't, the 5D II has a 1 stop advantage at High ISO but the detail is huge difference.

[/quote]



Yes. On the APS-C side of life the Sony SLTs have an edge. On full format Canon is as good as Nikon. Sony is worse here.

This is about the level of details.
#45
I own both cameras. Just because you have not used your 5d in circumstance demanding enough to make the banding obvious and detrimental to image quality does not mean it is not there; I have examined files from multiple 5d's and they all have it, the numerous on-line complaints from other owners add further testimony. Simple test; point your camera in live view mode at a featureless wall in a darkened room with the gain turned up and see the tartan hatched banding for yourself. Whether you can live it is your own decision, but for me it is very poor performance for such an expensive camera, especially since the most basic of cameras no longer have problems with banding. If the banding was not there the 5d would indeed have the 1 stop advantage you claim at the image level(mainly due to double the sensor area since high-iso read noise is slightly worse than on the sony sensor used in the k-5), but unfortunately canon chose to leave it uncorrected. Also, the DxO sensor mark figures unfortunately do not take banding noise into account in their DR assessment (only measure averaged read noise over a small section of the frame) and as such give the 5d a unrealistically optimistic DR rating. Despite that, the low iso DR is still over 2 stops less than the K-5. The 5d does provide more detail on account of the larger pixel count, weaker AA filter and lower demand on lens resolution but that again is mostly due to a larger sensor - the 5d's only real advantage, which is spoiled by sloppy signal processing.



[quote name='Christos' timestamp='1303452600' post='7838']

LMAO, my 5D has no banding at all, also the K5 cannot compete with the 5D in IQ I have side by side test to confirm this as you probably don't, the 5D II has a 1 stop advantage at High ISO but the detail is huge difference.

[/quote]
#46
I will post a 5D II image when I get home mine has no banding, some were affected with it and others not is my opinion.
#47
[quote name='Klaus' timestamp='1303321059' post='7791']

Well, its function is the suppression of moiree effects (Anti-Aliasing).

This is an example for a moiree effect:

[url="http://en.wikipedia.org/wiki/File:Moire_on_parrot_feathers.jpg"]http://en.wikipedia.org/wiki/File:Moire_on_parrot_feathers.jpg[/url]

The AA filter does so by softening - it is a low-pass-filter in front of the sensor.





Your 5D II has a very weak AA filter for instance which is why you perceive it as very sharp. Some of this is "false" detail and moirees at times but my personal opinion on this is ... who cares as long as things look sharp. Others disagree here.

[/quote]

Its function is about suppression of aliasing, is more correct. Moire effects are patterns that can appear due to aliasing.



I can't remember seeing moire or moire effects in 5D mkII photos, so its AA-filter must not be all that weak (too weak). AA filter design/production is an art, and the costs of AA-filters are about as high as the sensors themselves.
#48
[quote name='Ayoh' timestamp='1303454190' post='7840']

I own both cameras. Just because you have not used your 5d in circumstance demanding enough to make the banding obvious and detrimental to image quality does not mean it is not there; I have examined files from multiple 5d's and they all have it, the numerous on-line complaints from other owners add further testimony. Simple test; point your camera in live view mode at a featureless wall in a darkened room with the gain turned up and see the tartan hatched banding for yourself. Whether you can live it is your own decision, but for me it is very poor performance for such an expensive camera, especially since the most basic of cameras no longer have problems with banding. If the banding was not there the 5d would indeed have the 1 stop advantage you claim at the image level(mainly due to double the sensor area since high-iso read noise is slightly worse than on the sony sensor used in the k-5), but unfortunately canon chose to leave it uncorrected. Also, the DxO sensor mark figures unfortunately do not take banding noise into account in their DR assessment (only measure averaged read noise over a small section of the frame) and as such give the 5d a unrealistically optimistic DR rating. Despite that, the low iso DR is still over 2 stops less than the K-5. The 5d does provide more detail on account of the larger pixel count, weaker AA filter and lower demand on lens resolution but that again is mostly due to a larger sensor - the 5d's only real advantage, which is spoiled by sloppy signal processing.

[/quote]

I do not get that people keep focussing on quite nonsensical DR measurements that much.... Photos with an actual high dynamic range are bland and contrastless. The ONLY time it matters that a camera is more noisy in shadows is when you have under-exposed severely and have to adjust a lot. That my 450D, by modern standards (compared to the 7D/60D/600D/D7000/D5100/K5/A580) is very noisy in shadows, that is evident. BUt the 5D mk II, that camera performs fine. Only when the photographer does something stupid, one might notice that "banding" you mention.



The K5 is not that great in high ISO, it gets as noisy as the current rest. Just as base ISO is is remarkably clean. One can still get superb results with 5D mk II's, 60D's and so on.
#49
[quote name='Brightcolours' timestamp='1303457015' post='7844']

I do not get that people keep focussing on quite nonsensical DR measurements that much.... Photos with an actual high dynamic range are bland and contrastless. The ONLY time it matters that a camera is more noisy in shadows is when you have under-exposed severely and have to adjust a lot. That my 450D, by modern standards (compared to the 7D/60D/600D/D7000/D5100/K5/A580) is very noisy in shadows, that is evident. BUt the 5D mk II, that camera performs fine. Only when the photographer does something stupid, one might notice that "banding" you mention.



The K5 is not that great in high ISO, it gets as noisy as the current rest. Just as base ISO is is remarkably clean. One can still get superb results with 5D mk II's, 60D's and so on.

[/quote]No ones going to underexpose an image by 4 stops anyways, anyway here is a 100% crop at ISO 6400 16mm 1/100th on both K5 and 5D II, even my 7D had better detail than the K5 but not like the 5D II,

5D II

[Image: 5DIIISO6400Lowepro.jpg]



K5

[Image: K5ISO6400Lowepro.jpg]
#50
While I have not used a 5D2 myself, I'm making an assumption the sensor technology is similar or same as the 50D which came out around the same time. Based on DxO mark results, the noise score differences between the 50D and 5D2 are about what you would expect if you scaled the same sensor design. Having said that, the 7D is... about 1.2x better per area, which if I'm calculating right, is just over half a stop.



The 50D did exhibit some visible banding under specific conditions. On that, shadow areas can show banding if you bought them up. At low ISO, it was a non-issue unless you're doing a ridiculous amount of shadow recovery, such as extremely under-exposed shots or attempting single file HDR. To me banding wasn't really significant until around ISO3200, where relatively minor shadow recovery (or similar exposure adjustment) started to have an increased chance of showing banding effects. The 18MP sensor replacing that doesn't show banding in similar conditions, although if you keep digging deep enough you can still find some if you really want.



On the user set auto-ISO limits, I'd expect that to be present in the 5D3 as it has started appearing on lower level Canon bodies now.
<a class="bbc_url" href="http://snowporing.deviantart.com/">dA</a> Canon 7D2, 7D, 5D2, 600D, 450D, 300D IR modified, 1D, EF-S 10-18, 15-85, EF 35/2, 85/1.8, 135/2, 70-300L, 100-400L, MP-E65, Zeiss 2/50, Sigma 150 macro, 120-300/2.8, Samyang 8mm fisheye, Olympus E-P1, Panasonic 20/1.7, Sony HX9V, Fuji X100.
  


Forum Jump:


Users browsing this thread:
1 Guest(s)