Understanding Perceived Sharpness

[Playdo]

Hi,

First of all I'd like to say thanks to all those who contributed to my previous thread: http://luminous-landscape.com/forum/index....showtopic=38546. There was a lot of good information there and I'm hoping that I'll get as good an outcome with this post as there are obviously a lot of knowledgeable people here.

I'd like to know more about perceived sharpness between Canon and Nikon. The majority of Nikon photos that I see have more clarity/contrast to them, especially when viewed on the web. They seem to 'pop' more. This is more evident when looking at photos with a very shallow DOF. I'm not referring to resolving fine details (resolution) but perceived sharpness.

My thoughts on this are:

1. Most Nikon photos that I see have much more vibrant/saturated colours. This itself gives more contrast and can add to perceived sharpness.

2. Nikon lenses may have higher micro-contrast (in general).

3. Nikon may use softer AA filters.

I'm sure there are photographers that have noticed the same thing so I'd be interested to hear your reasonings.

One thing I'm not clear on. If an image is captured with a high micro-contrast lens without an AA filter, providing a sharp, contrasty looking image, can correct use of sharpening techniques in post-processing match it (for a low micro-contrast lens and strong AA filter)?

[feppe]

There are too many non-camera variables on web-sized JPEGs to be able to distinguish between two 35mm cameras. The resampling and sharpening method become much more important than the sensor. Unless you're doing a comparison without knowing which shot is from each camera before making a judgment, what you're seeing is pure bias.

I ran a test here a while back with web-resolution MFDB and 35mm photos. I had 50 MFDB and 50 35mm shots in the blind test. LL readers were unable to distinguish between them. So comparing two 35mm cameras together is even more difficult. Whatever you're seeing is placebo.

If you're actually running a blind test you can congratulate yourself, as you are an exceptional pixel peeper.

An alternative explanation is that you are seeing "Canon color" which some claim to be able to spot, and applying any preconceived notions to the shots accordingly. Another alternative is that Nikon users like to jack up the Saturation/Vibrance/Sharpening sliders more than Canon shooters smile.gif

Finally, there's a good recent thread about "micro-contrast" and another one started by me on "tonality" here as well - recapping those threads: both words are jedi mind tricks, and are not quantifiable. Don't be fooled by them.

[ErikKaffehr]

Hi,

1) Color rendition does not affect sharpness but may affect the perception of it.

2) I don't think so. The term microcontrast seems a bit underdefined to me.

3) Yes AA-filtering may matter. On the other hand I'd suggest that it's effect is overestimated. The choice of AA-filter thickness is to my understanding quite complex and closely related to the fill factor of the sensor. Increasing fill factor reduces aliasing but also resolution.

I guess the effect of the aliasing filter can be reduced by increased sharpening using high amount and small radius.

An ancient comparison of two cameras without/with AA filter can be seen here:
http://www.dpreview.com/reviews/kodakdcs14n/page18.asp
Today's raw converters are probably better on avoiding color moiré.

A more modern example can be found in this article: http://www.imx.nl/photo/leica/camera/page157/page157.html

Color moiré can be reduced by desaturating part of the image. Monochrome moiré cannot be eliminated to my best knowledge.

The impression I have is that scientist regard aliasing as an artifact while photographers may under circumstances see it as increased sharpness.

Finally, I suspect that AA-filtering on Canon and Nikon is similar, it seems that they are pretty close in resolution in most test I have seen.

Please keep in mind that there are many factors that are detrimental to sharpness, inexact focusingm camera vibration, lens aberrations and diffraction. In test all these parameters should be controlled in practical shooting this is seldom the case.

Image quality can only be judged at actual pixels or large prints.

Best regards
Erik

Best regards
Erik

Hi,

First of all I'd like to say thanks to all those who contributed to my previous thread: http://luminous-landscape.com/forum/index....showtopic=38546. There was a lot of good information there and I'm hoping that I'll get as good an outcome with this post as there are obviously a lot of knowledgeable people here.

I'd like to know more about perceived sharpness between Canon and Nikon. The majority of Nikon photos that I see have more clarity/contrast to them, especially when viewed on the web. They seem to 'pop' more. This is more evident when looking at photos with a very shallow DOF. I'm not referring to resolving fine details (resolution) but perceived sharpness.

My thoughts on this are:

1. Most Nikon photos that I see have much more vibrant/saturated colours. This itself gives more contrast and can add to perceived sharpness.

2. Nikon lenses may have higher micro-contrast (in general).

3. Nikon may use softer AA filters.

I'm sure there are photographers that have noticed the same thing so I'd be interested to hear your reasonings.

One thing I'm not clear on. If an image is captured with a high micro-contrast lens without an AA filter, providing a sharp, contrasty looking image, can correct use of sharpening techniques in post-processing match it (for a low micro-contrast lens and strong AA filter)?

[ErikKaffehr]

Hi,

You happen to have links to those two discussions?

Agree with your view. I don't know if microcontrast is a Jedi trick but also don't know any definition of it.

Best regards
Erik

There are too many non-camera variables on web-sized JPEGs to be able to distinguish between two 35mm cameras. The resampling and sharpening method become much more important than the sensor. Unless you're doing a comparison without knowing which shot is from each camera before making a judgment, what you're seeing is pure bias.

I ran a test here a while back with web-resolution MFDB and 35mm photos. I had 50 MFDB and 50 35mm shots in the blind test. LL readers were unable to distinguish between them. So comparing two 35mm cameras together is even more difficult. Whatever you're seeing is placebo.

If you're actually running a blind test you can congratulate yourself, as you are an exceptional pixel peeper.

An alternative explanation is that you are seeing "Canon color" which some claim to be able to spot, and applying any preconceived notions to the shots accordingly. Another alternative is that Nikon users like to jack up the Saturation/Vibrance/Sharpening sliders more than Canon shooters smile.gif

Finally, there's a good recent thread about "micro-contrast" and another one started by me on "tonality" here as well - recapping those threads: both words are jedi mind tricks, and are not quantifiable. Don't be fooled by them.

[Playdo]

Hi Feppe. Yeah I wondered how much of it was placebo too so recently I did blind tests and the majority of the time I could confidently notice the difference. Not pixel peeping but I think it's quite evident. I presume that a lot of it is to do with the colours. I think that the vibrancy/saturation seen more in nikon photos adds to the perceived sharpness. Different people with different perception eh. I'm intrigued to know if there's more to it.

[feppe]

Didn't save the links, searching my topic history will bring up the tonality post, micro-contrast topic might be harder to find as that's a term thrown around a lot here.

[dwdallam]

I believe Canon uses a stronger AA filter than does Nikon. In any event, if yuor shooting RAW, it doesn't matter. You can process them anyway you like.

[Playdo]

Please keep in mind that there are many factors that are detrimental to sharpness, inexact focusing

Hi Erik,

This is something else that I have been wondering about. On a lot of forums I hear of people needing to microadjust their Canon lenses. I rarely hear this regarding Nikon. I usually hear that they don't need adjustment. At shallow DOF focus is thin so there isn't much room for error. I don't know much about it myself but I have heard that microadjustment doesn't totally resolve the problem as focus can vary at different distances. If this is the case then a microadjusted lens could easily fall short of critical focus at different distances to the one it was microadjusted to?

[dwdallam]

Hi Erik,

This is something else that I have been wondering about. On a lot of forums I hear of people needing to microadjust their Canon lenses. I rarely hear this regarding Nikon. I usually hear that they don't need adjustment. At shallow DOF focus is thin so there isn't much room for error. I don't know much about it myself but I have heard that microadjustment doesn't totally resolve the problem as focus can vary at different distances. If this is the case then a microadjusted lens could easily fall short of critical focus at different distances to the one it was microadjusted to?

What else can you do? You could send the lens to the factory and have it calibrated. After that, if it's not good enough, you use manual focus. I tested my lens' at about 20 feet using the LCD screen test. 70-200IS L, 16-36L, 24-70L all perfect.

[Playdo]

I'd previously got my definition of micro-contrast from the luminous landscape website http://www.luminous-landscape.com/tutorial...-contrast.shtml

[gdanmitchell]

I'd like to know more about perceived sharpness between Canon and Nikon...

I think you are barking up the wrong tree if you are looking for generalized differences in "sharpness" between two brands like Canon and Nikon. There are great photographers using both brands and producing excellent photography using both, and there is no way to tell which brand was used by looking at the photos. Sure, one lens from brand X might be "sharper" than a similar lens from brand Y... and vice versa, but the the differences are often a matter of degree, cut both ways, and are between lenses that are often both quite good.

[ErikKaffehr]

Hi,

There are several aspect to this. One issue is sensor alignment. To achive absolute sharpens Sensor, DSL-mirror, autofocus mirror and the autofocus device msut be aligned within perhaps 10 microns (1/100 of a milimeter) no easy task.

Lately most models have micro adjustment, which probably works. I'd suggest that at least alignment related adjustment would work for all focusing distances.

It's entirely possible that one maker has greater tolerances than an another.

Best regards
Erik

Hi Erik,

This is something else that I have been wondering about. On a lot of forums I hear of people needing to microadjust their Canon lenses. I rarely hear this regarding Nikon. I usually hear that they don't need adjustment. At shallow DOF focus is thin so there isn't much room for error. I don't know much about it myself but I have heard that microadjustment doesn't totally resolve the problem as focus can vary at different distances. If this is the case then a microadjusted lens could easily fall short of critical focus at different distances to the one it was microadjusted to?

[ErikKaffehr]

OK,

MTF at low frequencies, right?

Erik

I'd previously got my definition of micro-contrast from the luminous landscape website http://www.luminous-landscape.com/tutorial...-contrast.shtml

[ErikKaffehr]

Hi,

There is a very good article on the Carl Zeiss site describing MTF in visual terms, it's hard to find so I set upo apge with links to it:

http://83.177.178.241/ekr/index.php/photoa...-and-perception

I'd check out this image: http://www.zeiss.de/C12567A8003B8B6F/Graph...le/Image_02.jpg

And the accompanying article: http://www.smt.zeiss.com/C12567A8003B8B6F/...Kurven_2_en.pdf

Best regards
Erik

I'd previously got my definition of micro-contrast from the luminous landscape website http://www.luminous-landscape.com/tutorial...-contrast.shtml

[ErikKaffehr]

Hi,

I'd say that sharpness can only be seen at actual pixels or in large prints. If we see differences in web-size pictures it would be something else. Color characteristics may differ between cameras especially in JPEG. With RAW endless processing options do exist. The same applies to tone curve.

It's well possible that one or another maker would be more successful in reducing flare, which is a major contribution to perceived haziness. Also different vendors take different approaches to noise reduction, but this is mostly a high ISO and JPEG issue.

Best regards
Erik

I think you are barking up the wrong tree if you are looking for generalized differences in "sharpness" between two brands like Canon and Nikon. There are great photographers using both brands and producing excellent photography using both, and there is no way to tell which brand was used by looking at the photos. Sure, one lens from brand X might be "sharper" than a similar lens from brand Y... and vice versa, but the the differences are often a matter of degree, cut both ways, and are between lenses that are often both quite good.

[Playdo]

I'd suggest that at least alignment related adjustment would work for all focusing distances.

Hi Erik,

Are you saying that microadjustment (an in camera function) would work for all focusing distances or do you mean actual calibration (sent in)?

Thanks for the links - very informative. I'll have an in-depth read when I get time.

[bjanes]

Hi Feppe. Yeah I wondered how much of it was placebo too so recently I did blind tests and the majority of the time I could confidently notice the difference. Not pixel peeping but I think it's quite evident. I presume that a lot of it is to do with the colours. I think that the vibrancy/saturation seen more in nikon photos adds to the perceived sharpness. Different people with different perception eh. I'm intrigued to know if there's more to it.

The vibrancy and saturation have more to do with the camera or raw converter settings than the qualities of the raw file produced by the camera. However, there are some differences in color rendering by different sensors as borne out by the DXO tests. To ensure a level playing field, you should use one raw converter and then calibrate the cameras to a neutral setting. For ACR, I would suggest a linear tone curve and the Adobe Standard profile tweaked by using the DNG Profile editor with a Color Checker. The results would look flat, but the playing field would be more equalized.

[ErikKaffehr]

My guess is that in camera function is superior to calibration by manufacturer. My guess is also that it may be adequate for all distances.

Please not that both statements are guesses, not without reason but still a guess.

Erik

Hi Erik,

Are you saying that microadjustment (an in camera function) would work for all focusing distances or do you mean actual calibration (sent in)?

Thanks for the links - very informative. I'll have an in-depth read when I get time.

[Wayne Fox]

Finally, I suspect that AA-filtering on Canon and Nikon is similar, it seems that they are pretty close in resolution in most test I have seen.

Here's a quote Graeme Nattress from this thread ... http://luminous-landscape.com/forum/index....showtopic=20367 which supports this ...

"Optical low pass filters (OLPF's) or Anti Alias filter don't come in different "strengths". The filter works in two passes, each layer splits the light in two either horizontally or vertically, so by combing them together, you get vertical and horizontal filtering. The distance of seperation of the two rays of light is governed by the thickness of that layer, so if you want, (or need to as you don't have square photosites) you can adjust the filter accordingly. You choose the thickness of the filter in relation to the spacing of the photosites on the sensor."

[Playdo]

I just came across this thread: http://www.fredmiranda.com/forum/topic/833110/0. The OP says it seems "photographers with Nikon cameras and lenses more easily create the unique special look (beautiful sharp, crisp, fresh images with dreamy look, and colors which pop!)". I've heard comments like this by others. This is similar to what I've noticed.

I know that different cameras and lenses give different 'looks'. I also believe that these different looks are better suited to shooting certain subjects.  

I'm not talking about resolution here ie. how much detail can be seen. As most of the images that I am viewing are on line so, as mentioned, I'm pretty sure that it is more to do with colour and contrast. Here's examples of what I mean: http://www.flickr.com/photos/tvchicklet/se...57600313161578/. All the images taken with the D3 have more colour saturation/vibrancy and contrast and tend to show that sharpness/crispness.