Dynamic Range: D5100/D7000

Started May 21, 2011 | Discussions
binary_eye
Veteran MemberPosts: 4,290
Like?
Re: D7000 Tonal Compression PP Methodology
In reply to Zane Paxton, May 24, 2011

Zane Paxton wrote:

Keep in mind that the 'Highlight blinkies' on the camera's LCD are a JPG version of the RAW file and don't have the same wide DR as the RAW files (which there is no way to view in-camera, the JPG is a substitution, thus the need to learn to use the Histogram).

The histogram is based on the JPEG as well, so looking at it doesn't really offer anything over the blinking highlight indicator.

Reply   Reply with quote   Complain
rhlpetrus
Forum ProPosts: 23,507Gear list
Like?
In two words: a lot! With examples.
In reply to drwho9437, May 24, 2011

drwho9437 wrote:

For those of you who use similar techniques I am curious how much of an improvment you feel you get with a D5100 or D7000 compared with your old DSLR.

Coming from a D80, it's just amazing the potential for recovery that the D7k's files have. I was checking MM behavior for another thread and shot a cityscape in two ways.

1- Focus point in darkest areas of image, sky blown by about 1.3 stop:

2- Focus point on brightest cloud, foreground underexposed by about 1.7 stop:

3- Now first with HLs recovery in CNX2:

4- Second with LLs recovery in CNX2:

Now, I also shot same scene with D80, that gave about 1/3 stop brighter exposure in both situations (D7k has a better MM system), and tried to recover the darker image, here are the details from both at 100% res of D80, I don't need to tell you which is which. Observe that the D80's image needed 1/3 stop less of recovery. All are at ISO100.

1-

2-

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Nikon D7000 Nikon 1 V1 Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF Nikkor 35mm f/2D Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +3 more
Reply   Reply with quote   Complain
rhlpetrus
Forum ProPosts: 23,507Gear list
Like?
Zane's shots, a question about the spotmeter calibration to you and ENO
In reply to Zane Paxton, May 24, 2011

The first one is very very impressive, the other are also good examples. I agree with you and Eduardo, at ISO100 the D7k is really outstanding regarding it's ability to recover shadow detail.

I want to get your views on this: at least my sample seems to have a different calibration re spotmetering than I'm used to. If you center the exposure scale with SM, I get the central peak to the right of the midvalue of 127 for the histograms. That's pretty consistent across the color range. A simple way to test it is to put a piece of white paper against a dark background, SM it and check histogram. Consistently I get the peak at 1/3 to 2/3 above central line, with the D80 I get exactly centered.

It's not a shutter calibration, the exposure is actually higher on the D7k. Have you both tested your spotmeter readings on the D7k?

Then I check how far I go with exposure up until blinkies start to show: same distance for both, about 2 1/3 stops. But when I add the extra 1/3-2/3 stop I started with for D7k, that means that from same reference, D80 gives 2 1/3 stops of distance from midtone to clipping and D7k gives 2 2/3 to 3 stops.

We know the blinkies actually appear before RAW clips, so the RAW file has more than that. In my tests I found that going 2/3 above clipping point with D7k still provides for total color recovery, so that's a practical 3 1/3-2/3 HLs range from midtone.

If you couple that with the LLs' recovery abilities of the D7k, which seems to go forever, as Zane and others have shown, including Eduardo's test showing it keeps colors and noise better than D700 in that respect, we have an outstanding tool for high DR shooting. I'm pretty confident of sending bright clouds in a landscape to +2 2/3 stops over central reading with spot on my D7k (which is at least 3 stops up from typical midtone rendition, as I mentioned) and still recover detail and color balance fully.

Now try to do that to a D80. A D300/D90, well, I can't do it, but I think the D7k beats them as well.

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Nikon D7000 Nikon 1 V1 Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF Nikkor 35mm f/2D Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +3 more
Reply   Reply with quote   Complain
binary_eye
Veteran MemberPosts: 4,290
Like?
Re: Zane's shots, a question about the spotmeter calibration to you and ENO
In reply to rhlpetrus, May 24, 2011

rhlpetrus wrote:

A simple way to test it is to put a piece of whoite paper against a dark background, SM it and check histogram. Consistently I get the peak at 1/3 to 2/3 above cental line, with the D80 I get exactly centered.

You're referring to the in-camera histogram? If so, it likely means the D7000's tone curve is lifting the mid-tones more as compared to the D80's tone curve.

Reply   Reply with quote   Complain
_sem_
Veteran MemberPosts: 4,313
Like?
Re: In two words: a lot! With examples.
In reply to rhlpetrus, May 24, 2011

rhlpetrus wrote:

drwho9437 wrote:

For those of you who use similar techniques I am curious how much of an improvment you feel you get with a D5100 or D7000 compared with your old DSLR.

Coming from a D80, it's just amazing the potential for recovery that the D7k's files have. I was checking MM behavior for another thread and shot a cityscape in two ways.

Renato, I think the recovery potential may be even more amazing if you use DxO or Adobe converters. Or if you grab two CNX conversions with different EC, one for the shadows and the other for the highlights, and blend them in Photomatix. Not that I'd have sth against CNX, I just think it is just not made for heavy adjustments and quasi-HDR.

For what I see, the sensor is considerably improved at base ISO, but the processing algorithms are still not ready for reasonable work with harsh light. I mean,

  • there is no automation of exposure for highlights protection,

  • there is no raw histogram for optimizing exposure manually,

  • when exposing for saving the highlights, the preview images are bloody dark.

So you have more DR but you're still on your own.
On the upside, the lower noise level does help ADL look better.

Reply   Reply with quote   Complain
Zane Paxton
Veteran MemberPosts: 6,937Gear list
Like?
Re: D7000 Tonal Compression PP Methodology
In reply to binary_eye, May 25, 2011

binary_eye wrote:

Zane Paxton wrote:

Keep in mind that the 'Highlight blinkies' on the camera's LCD are a JPG version of the RAW file and don't have the same wide DR as the RAW files (which there is no way to view in-camera, the JPG is a substitution, thus the need to learn to use the Histogram).

The histogram is based on the JPEG as well, so looking at it doesn't really offer anything over the blinking highlight indicator.

The histogram is way more precise; over time I can relate what the histogram is telling me and what the RAW files results will be in a highly predicable way.

-- hide signature --

Zane
http://www.pbase.com/devonshire
Nikon D7000 & D2x
NAPP Member

'Life is not measured by the number of breaths we take, but by the number of moments it takes our breath away.” ~ Anonymous

 Zane Paxton's gear list:Zane Paxton's gear list
Nikon D7000 Nikon AF-S DX Nikkor 16-85mm f/3.5-5.6G ED VR Nikon AF-S Nikkor 17-35mm f/2.8D ED-IF Nikon AF-S Nikkor 70-200mm f/4G ED VR +2 more
Reply   Reply with quote   Complain
Zane Paxton
Veteran MemberPosts: 6,937Gear list
Like?
Re: Zane's shots, a question about the spotmeter calibration to you and ENO
In reply to rhlpetrus, May 25, 2011

rhlpetrus wrote:

The first one is very very impressive, the other are also good examples. I agree with you and Eduardo, at ISO100 the D7k is really outstanding regarding it's ability to recover shadow detail.

I want to get your views on this: at least my sample seems to have a different calibration re spotmetering than I'm used to. If you center the exposure scale with SM, I get the central peak to the right of the midvalue of 127 for the histograms. That's pretty consistent across the color range. A simple way to test it is to put a piece of white paper against a dark background, SM it and check histogram. Consistently I get the peak at 1/3 to 2/3 above central line, with the D80 I get exactly centered.

It's not a shutter calibration, the exposure is actually higher on the D7k. Have you both tested your spotmeter readings on the D7k?

No, but I should!

I've learned to shoot with matrix mode then I dial in EV to compensate by eye and by using the histogram. I've shot enough that I can pretty reliably predict what the RAW file will be like by using the histogram and avoiding the hard-right edge spikes. Usually, -0.3 EV is sufficient to control highlights, sometimes -0.67 is needed. In a few rare cases I've shot up to -1.0 EV, but usually that is too much back at the computer in ACR.

Then I check how far I go with exposure up until blinkies start to show: same distance for both, about 2 1/3 stops. But when I add the extra 1/3-2/3 stop I started with for D7k, that means that from same reference, D80 gives 2 1/3 stops of distance from midtone to clipping and D7k gives 2 2/3 to 3 stops.

We know the blinkies actually appear before RAW clips, so the RAW file has more than that. In my tests I found that going 2/3 above clipping point with D7k still provides for total color recovery, so that's a practical 3 1/3-2/3 HLs range from midtone.

If you couple that with the LLs' recovery abilities of the D7k, which seems to go forever, as Zane and others have shown, including Eduardo's test showing it keeps colors and noise better than D700 in that respect, we have an outstanding tool for high DR shooting. I'm pretty confident of sending bright clouds in a landscape to +2 2/3 stops over central reading with spot on my D7k (which is at least 3 stops up from typical midtone rendition, as I mentioned) and still recover detail and color balance fully.

Now try to do that to a D80. A D300/D90, well, I can't do it, but I think the D7k beats them as well.

The DR is really quite remarkable.... The Fuji S5 has finally been bested!
http://snapsort.com/compare/Fujifilm_FinePix_S5_Pro-vs-Nikon_D7000

For me, since I shoot mostly at ISO 100, the D7000's DR is way more useful to me than high ISO-Low Noise shooting.

-- hide signature --

Zane
http://www.pbase.com/devonshire
Nikon D7000 & D2x
NAPP Member

'Life is not measured by the number of breaths we take, but by the number of moments it takes our breath away.” ~ Anonymous

 Zane Paxton's gear list:Zane Paxton's gear list
Nikon D7000 Nikon AF-S DX Nikkor 16-85mm f/3.5-5.6G ED VR Nikon AF-S Nikkor 17-35mm f/2.8D ED-IF Nikon AF-S Nikkor 70-200mm f/4G ED VR +2 more
Reply   Reply with quote   Complain
Cobras
Regular MemberPosts: 493
Like?
ACR insturction?
In reply to Zane Paxton, May 25, 2011

Zane Paxton wrote:

3. In ACR, I usually reduce the exposure around 0.8 or more depending on the shot. (-.3 EV + -0.8 in ACR = 1.1 stops)

4. I'm finding that a combo of "Fill Light" and "Brightness" works best to pull up the shadows even more.

I see ACR in Photoshop CS5. I could not find the instruction for using ACR. I looked up Photoshop help. I don't see any instruction on Fil Lightight or other parameters. Could you please suggest me where to look for an ACR insruction? TIA.

Reply   Reply with quote   Complain
Zane Paxton
Veteran MemberPosts: 6,937Gear list
Like?
Re: ACR insturction?
In reply to Cobras, May 26, 2011

Cobras wrote:

Zane Paxton wrote:

3. In ACR, I usually reduce the exposure around 0.8 or more depending on the shot. (-.3 EV + -0.8 in ACR = 1.1 stops)

4. I'm finding that a combo of "Fill Light" and "Brightness" works best to pull up the shadows even more.

I see ACR in Photoshop CS5. I could not find the instruction for using ACR. I looked up Photoshop help. I don't see any instruction on Fill Lightight or other parameters. Could you please suggest me where to look for an ACR insruction? TIA.

I'm more of a "jump in and figure it out" kind of guy, so I never used any instructions, so can't really recommend anything that I actually have had any real experience with....

I did a web search for "Adobe Camera Raw Books" and found these:

http://www.amazon.com/Adobe-Camera-Digital-Photographers-Only/dp/0764596837

http://my.safaribooksonline.com/book/photo-and-graphic-manipulation/9780131385061

-- hide signature --

Zane
http://www.pbase.com/devonshire
Nikon D7000 & D2x
NAPP Member

'Life is not measured by the number of breaths we take, but by the number of moments it takes our breath away.” ~ Anonymous

 Zane Paxton's gear list:Zane Paxton's gear list
Nikon D7000 Nikon AF-S DX Nikkor 16-85mm f/3.5-5.6G ED VR Nikon AF-S Nikkor 17-35mm f/2.8D ED-IF Nikon AF-S Nikkor 70-200mm f/4G ED VR +2 more
Reply   Reply with quote   Complain
Robert Hoy
Senior MemberPosts: 1,572
Like?
Re: In two words: a lot! With examples.
In reply to _sem_, May 26, 2011

sem wrote:

rhlpetrus wrote:

drwho9437 wrote:

For those of you who use similar techniques I am curious how much of an improvment you feel you get with a D5100 or D7000 compared with your old DSLR.

Coming from a D80, it's just amazing the potential for recovery that the D7k's files have. I was checking MM behavior for another thread and shot a cityscape in two ways.

Renato, I think the recovery potential may be even more amazing if you use DxO or Adobe converters. Or if you grab two CNX conversions with different EC, one for the shadows and the other for the highlights, and blend them in Photomatix. Not that I'd have sth against CNX, I just think it is just not made for heavy adjustments and quasi-HDR.

For what I see, the sensor is considerably improved at base ISO, but the processing algorithms are still not ready for reasonable work with harsh light. I mean,

  • there is no automation of exposure for highlights protection,

  • there is no raw histogram for optimizing exposure manually,

That's because RAW is a file format not an image format. A RAW file only provides a histogram once conversion settings are applied to the RAW data. If you want to see what's in your RAW file without lots of contrast settings added, choose as conservative a image/picture style as possible that has the least contrast, saturation, and least steep tone curve applied.

  • when exposing for saving the highlights, the preview images are bloody dark.

So you have more DR but you're still on your own.
On the upside, the lower noise level does help ADL look better.

Reply   Reply with quote   Complain
_sem_
Veteran MemberPosts: 4,313
Like?
Re: In two words: a lot! With examples.
In reply to Robert Hoy, May 26, 2011

Robert Hoy wrote:

  • there is no raw histogram for optimizing exposure manually,

That's because RAW is a file format not an image format. A RAW file only provides a histogram once conversion settings are applied to the RAW data. If you want to see what's in your RAW file without lots of contrast settings added, choose as conservative a image/picture style as possible that has the least contrast, saturation, and least steep tone curve applied.

I can't wholly agree conceptually. NEF contains image data, which clearly is "histogramizible", check Photobola Rawnalyze. But it is true that it makes sense to apply a "picture control" (tone curve) to make the histogram more "legible". UniWB also produces an approximation of the raw histogram for the highlights end, but it certainly is clumsy. It is a big question how this should be implemented in a camera to make it most useful and not disturbing for certain groups of users. One option I've seen somewhere was to extend the RGB histogram on both ends, so that any spikes indicating blown highlights (right) or shadows (left) are expanded to indicate the content of the raw headroom in white and the size of the headroom for each channel.

Notice mostly no PC is conservative enough - Nikon's processing is not designed for processing high DR. I guess considerable modifications are required to change this.

Reply   Reply with quote   Complain
rhlpetrus
Forum ProPosts: 23,507Gear list
Like?
Re: Zane's shots, a question about the spotmeter calibration to you and ENO
In reply to binary_eye, May 27, 2011

No, not histogram only, the exposure is higher, 1/3 to 2/3 more.

binary_eye wrote:

rhlpetrus wrote:

A simple way to test it is to put a piece of whoite paper against a dark background, SM it and check histogram. Consistently I get the peak at 1/3 to 2/3 above cental line, with the D80 I get exactly centered.

You're referring to the in-camera histogram? If so, it likely means the D7000's tone curve is lifting the mid-tones more as compared to the D80's tone curve.

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Nikon D7000 Nikon 1 V1 Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF Nikkor 35mm f/2D Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +3 more
Reply   Reply with quote   Complain
rhlpetrus
Forum ProPosts: 23,507Gear list
Like?
Re: Zane's shots, a question about the spotmeter calibration to you and ENO
In reply to Zane Paxton, May 27, 2011

Zane Paxton wrote:

The DR is really quite remarkable.... The Fuji S5 has finally been bested!
http://snapsort.com/compare/Fujifilm_FinePix_S5_Pro-vs-Nikon_D7000

For me, since I shoot mostly at ISO 100, the D7000's DR is way more useful to me than high ISO-Low Noise shooting.

For me too, then the LLs recovery abilities when protecting HLs is juts very nice.

'Life is not measured by the number of breaths we take, but by the number of moments it takes our breath away.” ~ Anonymous

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Nikon D7000 Nikon 1 V1 Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF Nikkor 35mm f/2D Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +3 more
Reply   Reply with quote   Complain
vin ramundo
New MemberPosts: 17Gear list
Like?
Re: D7000 Tonal Compression PP Methodology
In reply to Tom Zimmer, May 27, 2011

I have been getting good results using raw files. I typically save the raw file at 0, +2, --2, sometimes +\-4

I have found that if you shoot low iso and you don't have a tripod using the raw file works better than shooting exposure brackets.

 vin ramundo's gear list:vin ramundo's gear list
Sony RX100 II Nikon D7000 Sony Alpha NEX-5N
Reply   Reply with quote   Complain
rhlpetrus
Forum ProPosts: 23,507Gear list
Like?
Re: In two words: a lot! With examples.
In reply to _sem_, May 27, 2011

I don't think you need RAW histograms, jpegs is enough and a few tests indicate that when no channel blows by more than 1/2 stop you are in safe territory, that's my experience so far, but I am even more conservative, only 1/3 stop and in very difficult situations. Otherwise I prefer to see no channel clipped in the in-camera histograms (blinkies are not good enough in-camera, since it uses only the mixed channles values, won't indicate a single channel clipped).

Re CNX2, it's true its recovery tools and exposure control only go a certain distance, but more than enough if one controls exposure as I mentioned above. Using too much LLs recovery takes some of the naturality of final image for my style. Those that like a somewhat harder HDR effect will use the appropriate tools, as you mention. I never go beyond 50 in either of the recovery controls in CNX2.

sem wrote:

Robert Hoy wrote:

  • there is no raw histogram for optimizing exposure manually,

That's because RAW is a file format not an image format. A RAW file only provides a histogram once conversion settings are applied to the RAW data. If you want to see what's in your RAW file without lots of contrast settings added, choose as conservative a image/picture style as possible that has the least contrast, saturation, and least steep tone curve applied.

I can't wholly agree conceptually. NEF contains image data, which clearly is "histogramizible", check Photobola Rawnalyze. But it is true that it makes sense to apply a "picture control" (tone curve) to make the histogram more "legible". UniWB also produces an approximation of the raw histogram for the highlights end, but it certainly is clumsy. It is a big question how this should be implemented in a camera to make it most useful and not disturbing for certain groups of users. One option I've seen somewhere was to extend the RGB histogram on both ends, so that any spikes indicating blown highlights (right) or shadows (left) are expanded to indicate the content of the raw headroom in white and the size of the headroom for each channel.

Notice mostly no PC is conservative enough - Nikon's processing is not designed for processing high DR. I guess considerable modifications are required to change this.

-- hide signature --

Renato.
http://www.flickr.com/photos/rhlpedrosa/
OnExposure member
http://www.onexposure.net/

Good shooting and good luck
(after Ed Murrow)

 rhlpetrus's gear list:rhlpetrus's gear list
Nikon D7000 Nikon 1 V1 Nikon AF-S DX Nikkor 12-24mm f/4G ED-IF Nikon AF Nikkor 35mm f/2D Nikon AF-S Nikkor 70-300mm f/4.5-5.6G VR +3 more
Reply   Reply with quote   Complain
_sem_
Veteran MemberPosts: 4,313
Like?
Re: In two words: a lot! With examples.
In reply to rhlpetrus, May 27, 2011

rhlpetrus wrote:

I don't think you need RAW histograms, jpegs is enough and a few tests indicate that when no channel blows by more than 1/2 stop you are in safe territory, that's my experience so far, but I am even more conservative, only 1/3 stop and in very difficult situations. Otherwise I prefer to see no channel clipped in the in-camera histograms (blinkies are not good enough in-camera, since it uses only the mixed channles values, won't indicate a single channel clipped).

I'm not saying I can't live without them, I'm just saying there is room for improvement With such raw histogram I'd instantly know I've blown 1/2 stop but recoverably. The way it is I just know I've blown and I must take one more shot with a stop more underexposure to see how much I've blown. And the way NX is made, I agree that pushing exposure to the edge of the highlights headroom is the best idea, and mostly this is really not required. But this is important if you want to do harsh light without doing multiple exposures, which the new sensors are inviting us to play with

Reply   Reply with quote   Complain
Zane Paxton
Veteran MemberPosts: 6,937Gear list
Like?
Re: In two words: a lot! With examples.
In reply to _sem_, May 27, 2011

sem wrote:

rhlpetrus wrote:

I don't think you need RAW histograms, jpegs is enough and a few tests indicate that when no channel blows by more than 1/2 stop you are in safe territory, that's my experience so far, but I am even more conservative, only 1/3 stop and in very difficult situations. Otherwise I prefer to see no channel clipped in the in-camera histograms (blinkies are not good enough in-camera, since it uses only the mixed channles values, won't indicate a single channel clipped).

I'm not saying I can't live without them, I'm just saying there is room for improvement With such raw histogram I'd instantly know I've blown 1/2 stop but recoverably. The way it is I just know I've blown and I must take one more shot with a stop more underexposure to see how much I've blown. And the way NX is made, I agree that pushing exposure to the edge of the highlights headroom is the best idea, and mostly this is really not required. But this is important if you want to do harsh light without doing multiple exposures, which the new sensors are inviting us to play with

My experience is that -1/3 stop EV is sufficient for most situations, regardless of what the (JPG) derived histogram suggests.

Generally speaking it is good practice to "expose to the right" as much as you can without blowing out the highlights..

-- hide signature --

Zane
http://www.pbase.com/devonshire
Nikon D7000 & D2x
NAPP Member

'Life is not measured by the number of breaths we take, but by the number of moments it takes our breath away.” ~ Anonymous

 Zane Paxton's gear list:Zane Paxton's gear list
Nikon D7000 Nikon AF-S DX Nikkor 16-85mm f/3.5-5.6G ED VR Nikon AF-S Nikkor 17-35mm f/2.8D ED-IF Nikon AF-S Nikkor 70-200mm f/4G ED VR +2 more
Reply   Reply with quote   Complain
Cytokine
Contributing MemberPosts: 626
Like?
Re: Recent findings
In reply to Zane Paxton, May 27, 2011

Zane Paxton wrote:

My discipline around bright skies in landscapes is still to use the ND grads where possible. Getting the exposure right in the first place is always better than trying to compensate in PP.

Soooooo Right!

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads