Pages: 1 ... 5 6 [7] 8 9 10   Go Down

Author Topic: Nikon D7000 Dynamic Range  (Read 117208 times)

JR

  • Newbie
  • *
  • Offline Offline
  • Posts: 24
Re: Nikon D7000 Dynamic Range
« Reply #120 on: December 07, 2010, 05:51:15 am »

http://www.sony.net/SonyInfo/technology/technology/theme/cmos_02.html)
The K5 images are _really_ impressive imho.

Thank you for the link. I agree. Sony has been in business for decades and when they now have entered the DSLR market the increased competition will benefit us all. As for the Pentax K5 it seems it is very close or identical to the D7000. I donīt have the camera so I can only speculate. I have however downloaded a bunch of raw files just to have a look at them and the only thing I am not so sure about is the apparent softness. The D7000 appears a little sharper than the K5. I have no idea why. Could be a stronger AA filter or deliberately under-sharpening. Whatever it is both cameras have excellent shadow detail. Thatīs what I would like to see in my Canon camera  :)
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: Nikon D7000 Dynamic Range
« Reply #121 on: December 07, 2010, 06:48:01 am »

Ray, just a comment on HTP as it concerns DR :)

I am familiar with HTP. I know a lot of people donīt like the feature, they say they get to much noise in the shadows. I donīt agree with that. I like to push all my equipment to their limits, and my personal findings are completely in line with the findings of Bob Atkins.

I find that HTP gives m a little more than one stop of DR in the highlights with no extra noise in the shadows. When I say no extra noise I am looking at it from a practical point of view. I know it can be measured. I know I can see it if I enlarge the image 400% and study the shadows closely but I canīt see it in an A3+ print. I have tried and it is not possible. From a practical point of view it is a non issue for me. I therefore leave HTP on all the time. It gives med extra highlight detail at no cost.

Hereīs an image from Bobīs site. As you can see you have to push pretty hard and look pretty hard to see the extra shadow noise produced by HTP. It is as Bob said, "there is an increase in the shadow noise if you look closely enough."


JR,
That is interesting, isn't it! I would have thought that the noise in the shadows would have been greater than that, with HTP on.

I have a lot of respect for Bob Atkins. He seems to know his stuff. If he'd had more time, or the inclination, I'm sure he would not only have compared images at ISO 200 with HTP on, and off, but also compared the same exposure at ISO 100 to see how the shadows compared with the shot at ISO 200 with HTP on.  

He might also have tried comparisons at higher ISOs with HTP on, and off, to see if the difference in shadow noise was similarly negligible.

Out of curiosity, I checked the measurement data from DXO for the Canon 40D, which featured in Bob Atkins HTP test.

Surprise! Surprise! All is now clear. The reason why Bob couldn't detect any significant increase in shadow noise with HTP on, at ISO 200, is because in this model of camera, the 40D, the SNR curves for ISO 100 and ISO 200 merge into one, in the shadows.

Have a look at the attached image of the full SNR information for the 40D at various ISOs, from the DXOMark website. You should notice that the SNR curves for ISO 100 and ISO 200 merge at a point about 2/3rds of the way towards the x-axis.

However, this is not the case at any ISO above 200. If you were to compare shadow noise at ISO 3200, with HTP on, and off, I predict you would observe a substantial increase in shadow noise in the image with HTP on. At ISO 400, there would be a less substantial increase in shadow noise, with HTP on, but still noticeable.

By the way, this is yet another example of how accurate DXO tests really are.
« Last Edit: December 07, 2010, 07:02:33 am by Ray »
Logged

telyt

  • Guest
Re: Nikon D7000 Dynamic Range
« Reply #122 on: December 07, 2010, 10:55:42 am »

I have a lot of respect for Bob Atkins. He seems to know his stuff.

He lost any shred of respect I had for him when he "graded" a camera he has never used, and defended his "grade" of it in the face of contrary evidence from people who had used the camera.
« Last Edit: December 07, 2010, 04:12:55 pm by telyt »
Logged

Dennis Carbo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 163
Re: Nikon D7000 Dynamic Range
« Reply #123 on: December 07, 2010, 01:15:58 pm »

He lost any shred of respect I had for him when he "graded" a camera he has never used, and defended his "grade" it in the face of contrary evidence from people who had used the camera.

Exactly How I feel about all the people that speak in absolutes about specs and charts, this system vs. that system etc.  without ever actually putting their hands on that system. Its like expressing your opinion about a restaurant you never ate at !
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: Nikon D7000 Dynamic Range
« Reply #124 on: December 07, 2010, 06:07:01 pm »

Exactly How I feel about all the people that speak in absolutes about specs and charts, this system vs. that system etc.  without ever actually putting their hands on that system. Its like expressing your opinion about a restaurant you never ate at !


But in this example, Bob did put his hands on the camera, the 40D, and his results in respect of increased shadow noise with HTP on, were in fact very accurate at ISO 200 (according to the DXO graphs). What Bob should have done was extend his testing to higher ISOs, and then he would have seen the anomaly that shadow noise is the same in the 40D at ISO 100 and 200 only, irrespective of HTP considerations.

It almost seems as though Canon have deliberately designed the 40D this way in order to give the impression that HTP has no downside, knowing that most shots by most peopleare taken at either ISO 100 or 200.
Logged

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: Nikon D7000 Dynamic Range
« Reply #125 on: December 08, 2010, 05:00:27 am »

It almost seems as though Canon have deliberately designed the 40D this way in order to give the impression that HTP has no downside, knowing that most shots by most peopleare taken at either ISO 100 or 200.
Do you mean that they willingly limited the performance of ISO100 so as to hide the drawbacks of a mode that few (?) people use. I find that hard to belive. Is there no physical mechanism that explains the anomaly of iso100/200 at low input levels?

-h
Logged

ejmartin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: Nikon D7000 Dynamic Range
« Reply #126 on: December 08, 2010, 08:03:02 am »

Do you mean that they willingly limited the performance of ISO100 so as to hide the drawbacks of a mode that few (?) people use. I find that hard to believe. Is there no physical mechanism that explains the anomaly of iso100/200 at low input levels?

-h

Yes there is.  Canon cameras have a read noise profile (the noise the camera adds when reading the sensor) that looks like this as a function of ISO (this for my 1D3):



Since ISO 2oo (ie ISO 200 w/HTP) internally in the camera uses the amplification of ISO 100, the noise relative to signal is approximately double what one would get by using straight ISO 200.  Read noise is most noticeable in shadows (low signal/noise) so that is where the difference will show most noticeably.  By the time one gets to ISO 1600, there is essentially no difference between that and ISO 16oo (using internally the amplification of ISO 800, the difference is very slight).

Logged
emil

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: Nikon D7000 Dynamic Range
« Reply #127 on: December 08, 2010, 10:38:47 am »


Since ISO 2oo (ie ISO 200 w/HTP) internally in the camera uses the amplification of ISO 100, the noise relative to signal is approximately double what one would get by using straight ISO 200.  Read noise is most noticeable in shadows (low signal/noise) so that is where the difference will show most noticeably.  By the time one gets to ISO 1600, there is essentially no difference between that and ISO 16oo (using internally the amplification of ISO 800, the difference is very slight).

Emil,
There's something odd going on here. The DXO Full SNR graphs for each ISO show that the 40D at ISO 100 has no better SNR than at ISO 200, in the shadows.

I interpret that as meaning that a given exposure at ISO 200 with HTP activated, will produce the same results as at ISO 100, where HTP is not an option, using the same exposure.

The principle has been with Canon cameras, that it's always better to fully expose at a higher ISO than use the same exposure to underexpose at a lower ISO, regarding shadow noise.

This doesn't appear to be the case with the 40D at ISO 200.

If one examines the DXO full graphs for SNR at each ISO, one sees that the earliest Canon APS-C camera that DXOMark have tested, the 6mp 10D, has a clear distinction in SNR between ISO 100 & 200, in the shadows.

The 20D has a much less significant distinction, the 30D even less, and the 40D none at all, except very close to the camera's noise floor where there's a slight blip in favour of ISO 100.
I'm speculating that this situation has arisen due to increased pixel count and smaller pixels.

The 50D seems to reflect the reality of this situation by making ISO 100 and ISO 200 the same sensitivity. I didn't realise this until I checked the DXO test results.

Here are the DXO graphs for the 40D and 50D. Could you explain what's going on here, please.  :)

Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3366
Re: Nikon D7000 Dynamic Range
« Reply #128 on: December 08, 2010, 02:37:56 pm »

Emil,
There's something odd going on here. The DXO Full SNR graphs for each ISO show that the 40D at ISO 100 has no better SNR than at ISO 200, in the shadows.

Here are the DXO graphs for the 40D and 50D. Could you explain what's going on here, please.  :)

While waiting for Emil's reply, I will add this observation which may be pertinent. Looking at the ISO plots for the 40D and 50D, I see that the ISO of the 50D, as measured by DXO, is the same for ISO 200 and ISO 100, which leads me to believe that the real base ISO of the 50D is 200 and that the ISO of 100 is merely overexposure. This is what I have observed for my Nikon D3, where the results ab the base ISO of 200 are exactly the same as for the LoISO setting of 100.

For the 40D, the real base ISO appears to be close to a nominal value of 100 (the measured ISO is actually 87 and the other measured ISOs are also proportionally smaller than the nominal values). What a wealth if information in the DXO database!

Regards,

Bill

Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: Nikon D7000 Dynamic Range
« Reply #129 on: December 08, 2010, 09:40:11 pm »

While waiting for Emil's reply, I will add this observation which may be pertinent. Looking at the ISO plots for the 40D and 50D, I see that the ISO of the 50D, as measured by DXO, is the same for ISO 200 and ISO 100, which leads me to believe that the real base ISO of the 50D is 200 and that the ISO of 100 is merely overexposure. This is what I have observed for my Nikon D3, where the results ab the base ISO of 200 are exactly the same as for the LoISO setting of 100.

For the 40D, the real base ISO appears to be close to a nominal value of 100 (the measured ISO is actually 87 and the other measured ISOs are also proportionally smaller than the nominal values). What a wealth if information in the DXO database!

Bill, I should thank you for bringing to my attention that secod tier of headings which is not available in the 'Compare Sensor' mode which I've always previously referred to.

I've taken several thousand shots with my 50D; most of them on a recent trip to Europe and Russia, and I'm embarrassed to admit I had no idea that ISO 100 on this camera has identical performance to ISO 200, regarding highlight detail and shadow detail, at appropriately different exposures.

In other words, 1/50th at ISO 100 produces the same result as 1/100th at ISO 200; identical for all practical purposes.

The reason why DXO do not show an SNR curve for the 50D at ISO 100 is because it would be impossible to see it. It would lie exactly underneath the SNR curve for ISO 200.

I've just stepped outside onto the verandah to take a few bracketed shots of the cloudy sky, at ISO 100 & 200. An ETTR exposure at 1/640th and ISO 200 produces the same result as 1/320th at ISO 100. The same degree of negative EC in ACR is required in each case. The same amount of detail is visible in the clouds. The histogram of each shot is approximately the same, indicating both exposures are a full ETTR.

I then stepped back into the house and took some severely underexposed shots of Jonathan's DR Test Target, at ISO 100 and 200, expecting the ISO 100 shot at 1/50th might display noticeably less noise than the ISO 200 shot at 1/100th.

Nope! The images are indistinguishable. They are identical for all practical purposes.

Well I never!  ;D  It's seems that this camera has two base ISOs.  ;D
Logged

JR

  • Newbie
  • *
  • Offline Offline
  • Posts: 24
Re: Nikon D7000 Dynamic Range
« Reply #130 on: December 09, 2010, 04:57:23 am »

Thatīs right, the 50D seems to have two "base" isoīs. Good to know if you are an iso 100 photographer but need the speed of iso 200  :)

If you look closely at the SNR number provided by Dxomark you will see that even if the max SNR between selected cameras does not differ that much the SNR in the deep shadows does. Below are screen shots of a few models, looking at the SNR in the deep shadows, just above 1% gray. I think 1% gray qualifies as "deep" shadows.  :)

The 60D has marginally improved in the shadows.

The measured points are not exactly the same but they are close enough.

Below is max SNR for the same cameras. Note the difference between the 50D and the D700/K5.

50D -          SNR:    40,5dB
5D Mark II - SNR:    43,1dB
D700 -        SNR:    44,8dB
D7000 -      SNR:    44,3dB
K5 -           SNR:    44,7dB
60D -         SNR: 41,1dB
« Last Edit: December 09, 2010, 05:24:59 am by JR »
Logged

JR

  • Newbie
  • *
  • Offline Offline
  • Posts: 24
Re: Nikon D7000 Dynamic Range
« Reply #131 on: December 09, 2010, 05:00:48 am »

D700 and 60D SNR @ 1% gray
« Last Edit: December 09, 2010, 05:22:32 am by JR »
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: Nikon D7000 Dynamic Range
« Reply #132 on: December 09, 2010, 10:45:45 am »

Thatīs right, the 50D seems to have two "base" isoīs. Good to know if you are an iso 100 photographer but need the speed of iso 200  :)

It's good to know whatever type of photographer you are. The only reason I can think of why anyone would want to use ISO 100 on the 50D is if they wanted to deliberately create a blur for a particular effect, such as depiction or emphasis of movement. A shutter speed which is unnecessarily fast is generally preferrable to one which is too slow.

Quote
If you look closely at the SNR number provided by Dxomark you will see that even if the max SNR between selected cameras does not differ that much the SNR in the deep shadows does. Below are screen shots of a few models, looking at the SNR in the deep shadows, just above 1% gray. I think 1% gray qualifies as "deep" shadows.


No. This is a logarithmic scale. The deep shadows would be around 0.1% and below. Don't ask me to explain. I'm not a mathematician. However, DXOMark state specifically that a change of one stop (or 1 EV) of exposure results in a change of about 3dB in SNR in the midtones, but a change of about 6dB in the shadows.

Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
1% gray is 4-stops below normal midtones, in the deep shadows
« Reply #133 on: December 09, 2010, 11:20:01 am »

With a typical mid-tone placement, traditionally 18% gray (18% reflectance), in practice often a bit lower, like 13%, the 1% gray level is about four stops below the midtones. To be pedantic, it is 4 stops below 16% gray.

That level of "fours stops below midtones" is what is the zone system describes as providing only vague textural information, not real subject detail, which typically ends 3 stops below mid-tones. To give you an idea of how dark 1% gray looks, it is below the reflectance of the purest black of most or all photographic printing papers. Also, four stops below mid-tones is the level at which the ISO definition of film speed requires signal to be just barely above the "noise level" of "film-base plus fog", so more or less the floor for useful "signal" with film used at its rated ISO speed.

So yes, I would say that 1% gray is deep shadow.


P. S. Here is some information on the related topic of albedo, the percentage reflectance of sunlight, which seems a natural measure for daylight photography
http://en.wikipedia.org/wiki/Albedo
One highlight:
"Albedos of typical materials in visible light range from up to 0.9 for fresh snow, to about 0.04 for charcoal, one of the darkest substances."
So 1% gray is two stops blacker than charcoal, and this real-world reflectivity range from fresh snow to charcoal (critical when photographing snowmen!) is about 23:1, or 4.5 stops.
« Last Edit: December 09, 2010, 11:29:48 am by BJL »
Logged

JR

  • Newbie
  • *
  • Offline Offline
  • Posts: 24
Re: 1% gray is 4-stops below normal midtones, in the deep shadows
« Reply #134 on: December 09, 2010, 01:53:39 pm »

BJL,

Thank you. Excellent explanation.

Even if we look at SNR at lets say 3-4% gray or 0.1% ( as Ray suggests ) the latest generation sensors from Sony has better SNR in the shadows compared to some Canon cameras.

-John

With a typical mid-tone placement, traditionally 18% gray (18% reflectance), in practice often a bit lower, like 13%, the 1% gray level is about four stops below the midtones. To be pedantic, it is 4 stops below 16% gray.

That level of "fours stops below midtones" is what is the zone system describes as providing only vague textural information, not real subject detail, which typically ends 3 stops below mid-tones. To give you an idea of how dark 1% gray looks, it is below the reflectance of the purest black of most or all photographic printing papers. Also, four stops below mid-tones is the level at which the ISO definition of film speed requires signal to be just barely above the "noise level" of "film-base plus fog", so more or less the floor for useful "signal" with film used at its rated ISO speed.

So yes, I would say that 1% gray is deep shadow.


P. S. Here is some information on the related topic of albedo, the percentage reflectance of sunlight, which seems a natural measure for daylight photography
http://en.wikipedia.org/wiki/Albedo
One highlight:
"Albedos of typical materials in visible light range from up to 0.9 for fresh snow, to about 0.04 for charcoal, one of the darkest substances."
So 1% gray is two stops blacker than charcoal, and this real-world reflectivity range from fresh snow to charcoal (critical when photographing snowmen!) is about 23:1, or 4.5 stops.
« Last Edit: December 09, 2010, 01:55:21 pm by JR »
Logged

ejmartin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: Nikon D7000 Dynamic Range
« Reply #135 on: December 09, 2010, 05:22:26 pm »

Emil,
There's something odd going on here.

[snip]

Here are the DXO graphs for the 40D and 50D. Could you explain what's going on here, please.  :)

Low ISO read noise drops by nearly a factor of two (in photon equivalents, ie electrons) between ISO 100 and 200 on most Canons, and since signal in these SNR plots is relative to clipping, in terms of absolute exposure one has half the read noise and half the signal, for about the same SNR.  When the read noise drops to being more level as a function of ISO, then the curves separate, since the read noise is the same but signal drops with ISO for a fixed percentage of clipping.  But note that even at ISO 100, there is still a SNR advantage above deep shadows since the read noise is less important and it comes down to more photons being gathered at a given percentage of clipping, when the ISO is lower.
Logged
emil

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: 1% gray is 4-stops below normal midtones, in the deep shadows
« Reply #136 on: December 09, 2010, 09:34:02 pm »

With a typical mid-tone placement, traditionally 18% gray (18% reflectance), in practice often a bit lower, like 13%, the 1% gray level is about four stops below the midtones. To be pedantic, it is 4 stops below 16% gray.

That level of "fours stops below midtones" is what is the zone system describes as providing only vague textural information, not real subject detail, which typically ends 3 stops below mid-tones. To give you an idea of how dark 1% gray looks, it is below the reflectance of the purest black of most or all photographic printing papers. Also, four stops below mid-tones is the level at which the ISO definition of film speed requires signal to be just barely above the "noise level" of "film-base plus fog", so more or less the floor for useful "signal" with film used at its rated ISO speed.

So yes, I would say that 1% gray is deep shadow.


P. S. Here is some information on the related topic of albedo, the percentage reflectance of sunlight, which seems a natural measure for daylight photography
http://en.wikipedia.org/wiki/Albedo
One highlight:
"Albedos of typical materials in visible light range from up to 0.9 for fresh snow, to about 0.04 for charcoal, one of the darkest substances."
So 1% gray is two stops blacker than charcoal, and this real-world reflectivity range from fresh snow to charcoal (critical when photographing snowmen!) is about 23:1, or 4.5 stops.

BJL,
What an excellent justification for not needing a better camera than a Point & Shoot, as regards SNR and dynamic range.
However, some of us, notwithstanding the reflectivity of charcoal, wish for lower noise and better detail in the shadows than a P&S or even an Olympus 4/3rds can provide.  Some of us even feel the need to merge different exposures to HDR in order to get a better SNR in the shadows.

Comparing the SNR on the log scale at 1% input for various cameras at base ISO, I find that the Canon G12 manages an SNR of about 19 dB, The Olypus E-620 only 18dB, the Canon 50D 21dB, the Canon 60D 22dB and the D7000 26dB.

We can argue all day about subjective impressions of what may be considered a moderately deep shadow, a deep shadow or a very deep shadow, but a degree of shadow in which a P&S has slightly lower noise than the Olympus E-620 does not seem particularly deep to me.

What you've written above also appears to be at variance with DXOMark's own explanations of the meaning of their results.

I invite you to look at the following extract from their 'insights'  articles at
http://www.dxomark.com/index.php/en/Our-publications/DxOMark-Insights/Noise-characterization/Summary

Quote
As a consequence, the SNR also has three regimes:

Shadows:  the SNR increases 6dB for every EV and loses 6dB for each doubling of the ISO setting.

Midtones:  the SNR increases 3dB for every EV and decreases by 3dB for each doubling of the ISO setting.

Highlights:  the SNR is constant and does not depend on the ISO.

Now it so happens the SNR graphs for the D7000 lend themselves admirably to demonstrate these DXO statements, because, as you know, the D7000 can be used for all purposes at base ISO. The loss in SNR due to doubling ISO is the same as reducing exposure by one stop at base ISO

In light of the above, we can make some reasonable deductions as to what point along the x-axis of the DXO graphs the midtones and the shadows lie.

If we examine the graph for SNR at 18% grey, for the D7000, we find that there is indeed a loss of approximately 3dB for each doubling of the ISO setting.

If we search for the point along the x-axis on the Full SNR graph where a doubling of ISO produces a fall of approximately 3dB in SNR, ie. the point where 18% grey would lie, we find it's about midway between 100% and 10% on the x-axis. That point represents the midtones, approximately the tonal range of skin tones in a portrait.

If we examine the fall in SNR with each doubling of ISO at the 1% point on the x-axis, which you describe as the deep shadows, we see that between ISO 100 and ISO 6400, the SNR has fallen from 26dB to 4dB with a total change of 6 EV in exposure.

22/6=3.7dB. I hardly think this constitutes the deep shadows, an increased drop in SNR of just 0.7dB compared to the rate of fall in the midtones of 3dB.

If we examine the fall in SNR at the 0.1% point on the x-axis, which I would describe as the point where the deep shadows begin, we find there's a fall of approximately 14dB between ISO 100 and ISO 800, or 4.7dB with each doubling of ISO setting, which is still short of the 6dB mentioned by DXO.

The graphs below show the Full SNR details for the D7000, and specifically the SNR at 18% grey.

Draw your own conclusions, but I know what my eyes tell me and I've taken the trouble to shoot tests with the D7000 to demonstrate the improvement in SNR in the shadows, and the consequent imrovement in visible detail compared with other cameras in identical circumstances.
« Last Edit: December 09, 2010, 09:39:09 pm by Ray »
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10235
Re: Nikon D7000 Dynamic Range
« Reply #137 on: December 09, 2010, 10:16:10 pm »

Low ISO read noise drops by nearly a factor of two (in photon equivalents, ie electrons) between ISO 100 and 200 on most Canons, and since signal in these SNR plots is relative to clipping, in terms of absolute exposure one has half the read noise and half the signal, for about the same SNR.  When the read noise drops to being more level as a function of ISO, then the curves separate, since the read noise is the same but signal drops with ISO for a fixed percentage of clipping. 

Emil,
I understand that's the case with most Canons, which is why I've been using ISO 100 with the 50D as the default, base ISO. I'm now having second thoughts and think I should set the camera at ISO 200 as base.

Quote
But note that even at ISO 100, there is still a SNR advantage above deep shadows since the read noise is less important and it comes down to more photons being gathered at a given percentage of clipping, when the ISO is lower.

I understand from DXO's 'insight' articles and other explanations, that photonic shot noise is generally more of a problem in the midtones.

I've already compared images of bright clouds, and images of deep shadows, at ISO 100 & 200 with the 50D, and have failed to observe any significant difference in noise, detail or tonal range.

I haven't, however, compared portraits of lovely models with creamy smooth skin, taken at those ISOs.

Could we reasonably predict that the ISO 100 shots of such models, using the 50D, would show visibly smoother skin tones than the ISO 200 shots?
 
(Unfortunately, no such models are currently at hand for me to test this  ;D )
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11288
    • Echophoto
Re: 1% gray is 4-stops below normal midtones, in the deep shadows
« Reply #138 on: December 10, 2010, 12:52:05 am »

Hi,

Except that the white snowman may be in the sun, melting away, while that piece of charcoal may be in a deep shadow.

The image below illustrates real world outdoor contrast range. It was taken using a Konica Minolta 7D in 2005. The image was shot in raw and processed in Lightroom. Better processing would certainly be possible.

Best regards
Erik

With a typical mid-tone placement, traditionally 18% gray (18% reflectance), in practice often a bit lower, like 13%, the 1% gray level is about four stops below the midtones. To be pedantic, it is 4 stops below 16% gray.

That level of "fours stops below midtones" is what is the zone system describes as providing only vague textural information, not real subject detail, which typically ends 3 stops below mid-tones. To give you an idea of how dark 1% gray looks, it is below the reflectance of the purest black of most or all photographic printing papers. Also, four stops below mid-tones is the level at which the ISO definition of film speed requires signal to be just barely above the "noise level" of "film-base plus fog", so more or less the floor for useful "signal" with film used at its rated ISO speed.

So yes, I would say that 1% gray is deep shadow.


P. S. Here is some information on the related topic of albedo, the percentage reflectance of sunlight, which seems a natural measure for daylight photography
http://en.wikipedia.org/wiki/Albedo
One highlight:
"Albedos of typical materials in visible light range from up to 0.9 for fresh snow, to about 0.04 for charcoal, one of the darkest substances."
So 1% gray is two stops blacker than charcoal, and this real-world reflectivity range from fresh snow to charcoal (critical when photographing snowmen!) is about 23:1, or 4.5 stops.
« Last Edit: December 10, 2010, 01:04:35 am by ErikKaffehr »
Logged
Erik Kaffehr
 

JR

  • Newbie
  • *
  • Offline Offline
  • Posts: 24
Re: 1% gray is 4-stops below normal midtones, in the deep shadows
« Reply #139 on: December 10, 2010, 06:42:03 am »


Where is this, Erik? Var är detta?


The image below illustrates real world outdoor contrast range. It was taken using a Konica Minolta 7D in 2005.

Best regards
Erik

Logged
Pages: 1 ... 5 6 [7] 8 9 10   Go Up