Dynamic Range of D60?

[Cirkitdude]

With all the complaint threads here about underexposure, overexposure, comparisons of D60 output to the 1D and P&S's... seems that a very limited dynamic range is most often the culprit. Some of the "underexposed", "flat" images I've seen here look very much like film scans (from beautiful slides) with a squashed DR.

Certainly, can't expect the DR of the D60 to compare with negative film (which has a huge DR)... but does it compare to slide film? I would assume (hope?) it's better than either slide or negative scanned with a consumer-grade film scanner?

Are there any documented measurements of dynamic range for the D60? I can't find them. Perhaps experienced slide photographers now with D60's would comment on DR comparisons?

Cirkitdude

 

[John Harrison82293]

After seriously playing with my D-60 now for several hundred shots its obvious to me mine is at minimum of 1/2 f stop underexposed and and could be more. I took shot shots this am and thought this look much better and I realized that I had locked the compensation in at 1/2+ f stop.

I also feel there could be some problems with the dynamic range but not going to get all that started in here the way I was attacked on the other.

When I DL the images I feel as if I just scanned some slide or film and have to start processing. Hitting auto levels just works on occasion and I usualy have to spend time in curves or manual levels to get the shot close to what I remember it was.

Well will just keep shooting and see how it goes. But I definitely know my camera is underexposing and not just to preserve hilites.
John R

 

[Adam2]

Great question and one I have posed to several experts. I have yet to receive a reply. I would love to see a rating system of DR in # of stops for the digital camera market. I have been waiting for my D60 for 4 months and after seeing the hundreds of samples am not sure that its DR is going to be sufficient (for me). Certainly, it appears adequate for controlled lighting situations (eg. studio), but with the variable lighting conditions in which I like to shoot, I am not sure. One of beautiful things about digital photography is that you can really experiment and gain immediate feedback in terms of light and exposure.

With all the complaint threads here about underexposure,
overexposure, comparisons of D60 output to the 1D and P&S's...
seems that a very limited dynamic range is most often the culprit.
Some of the "underexposed", "flat" images I've seen here look very
much like film scans (from beautiful slides) with a squashed DR.

Certainly, can't expect the DR of the D60 to compare with negative
film (which has a huge DR)... but does it compare to slide film? I
would assume (hope?) it's better than either slide or negative
scanned with a consumer-grade film scanner?

Are there any documented measurements of dynamic range for the D60?
I can't find them. Perhaps experienced slide photographers now
with D60's would comment on DR comparisons?

Cirkitdude

 

[jrod7350]

check out an article on Fred Miranda's site on using a darkened exposure to control highlights mixed with a brighter exposure. You can also do this with a single exposure using the RAW and Linear RAW outputs of the D60 and mixing those together. With that technique you can get around 8-10 stops depending on the scene.

Jason Rodriguez

 

[George Montemayor]

If you shoot JPEG, it may help with the DR if you shot in RAW. In RAW you can make use out of the 12-bits saved in RAW during post-processing unlike JPEGs which are stored in 8 bits per pixel.

I don't use a D60 but read in this forum you don't seem to have to save them as linear TIFFs to get the most out of them (unlike the D30).

--
George
EOS D30 ~ 50mm f/1.4 ~ 200mm f/2.8 L II

 

[Cirkitdude]

Adam2:

I agree... a rating system would be great.

Also, I have never seen any serious discussion (or measurements) of the inherent dynamic range of CMOS versus CCD. I know something of these technologies, having worked with microelectronics for many years, but I don't know how their optical dynamic range specs pan out. Perhaps this is an inherent limitation of CMOS... one Canon obviously doesn't talk about?

The longer I follow this forum, the more I wonder how well my phantom D60 will replace film and scans. Hurry up Canon... send my toy before I over-analyze and back out!

With all the complaint threads here about underexposure,
overexposure, comparisons of D60 output to the 1D and P&S's...
seems that a very limited dynamic range is most often the culprit.
Some of the "underexposed", "flat" images I've seen here look very
much like film scans (from beautiful slides) with a squashed DR.

Certainly, can't expect the DR of the D60 to compare with negative
film (which has a huge DR)... but does it compare to slide film? I
would assume (hope?) it's better than either slide or negative
scanned with a consumer-grade film scanner?

Are there any documented measurements of dynamic range for the D60?
I can't find them. Perhaps experienced slide photographers now
with D60's would comment on DR comparisons?

Cirkitdude

 

[Mike Fulton]

Presuming a correct exposure, a "flat" image actually suggests that the dynamic range of the subject was LESS than the available dynamic range of the camera (or film).

When the camera's dynamic range is wider than the subject, the darkest areas of the subject don't register as black in the camera, and/or the lightest areas don't register as white. This is because the camera has the ability to capture even darker or lighter areas, but there aren't any in the scene. The result is a low contrast (aka "flat") image.

When the dynamic range of the camera is not wide enough for the subject, the result is an excessively contrasty picture lacking in either highlight detail or shadow detail, or both.

In the first situation, you can easily fix the contrast in Photoshop or whatever. That's where you notice the difference between an 8-bit image and a 12-bit image.

In the second situation, you're hosed, for the most part, because there's nothing you can do in Photoshop to recover detail that was never recorded by the camera (or film).

Mike

 

[Frank E.]

I am sure this will go over big but it realy is an under exposed problem with mixed lighting.

In the studio, I can meter normaly with great results. Head out to the church and the images are flat. Do not increase flash exposure but increase camera exposure and I think you get much better images.

I have also found the 550EX is more consistant than the 420EX.

I do not belive it is the dynamic range as you can duplicate it with the D30 if you under expose by about a stop.

BTW turn off the focus point light in the view finder.

Frank

 

[Mike Fulton]

I am sure this will go over big but it realy is an under exposed
problem with mixed lighting.

If the image is not properly exposed, then you'll have problems no matter what. What I was saying applies to properly exposed images. Or, more precisely, images where the exposure is centered on the midpoint between light and dark. That may not be considered the "right" exposure in all cases.

In the studio, I can meter normaly with great results. Head out to
the church and the images are flat. Do not increase flash exposure
but increase camera exposure and I think you get much better images.

Now I am confused. If your subject is the same distance from the flash in both cases, and you're not relying on your environment for bounce flash, then your exposure shouldn't be changing. Unless there's some other factor that you're not mentioning, such as the subject being in front of a highly reflective background. That can make a TTL flash shut down before the main subject is properly exposed.

Perhaps you can post a few example images?

Mike

 

[Lin Evans]

With all the complaint threads here about underexposure,
overexposure, comparisons of D60 output to the 1D and P&S's...
seems that a very limited dynamic range is most often the culprit.
Some of the "underexposed", "flat" images I've seen here look very
much like film scans (from beautiful slides) with a squashed DR.

Certainly, can't expect the DR of the D60 to compare with negative
film (which has a huge DR)... but does it compare to slide film? I
would assume (hope?) it's better than either slide or negative
scanned with a consumer-grade film scanner?

Are there any documented measurements of dynamic range for the D60?
I can't find them. Perhaps experienced slide photographers now
with D60's would comment on DR comparisons?

Cirkitdude

I haven't received my D60 yet, but I suspect its dynamic range will be quite similar to my D30's which is actually quite good, especially so in shadow detail. You may want to take a careful look at Michael Reichmann's treatise on both the D30 and D60 for some insight:

http://www.luminous-landscape.com/

I've found it to have equal or better DR to Provia. The difficulty I think is the tendency for users to alway shoot at ISO 100 unless absolutely necessary to go up. On my 1D I discovered that ISO 100 tends to shorten the DR and decided to test my D30 to see if the frequency of blown out highlights might have some relation to ISO 100 as it does in my 1D. I believe it does - and since a comon CMOS design is shared with the D60 I suspect that using ISO 200 might result in a greater DR as well.

Best regards,

Lin

http://204.42.233.244

 

[Markep]

Warning: I am not an expert in this subject matter.

However, doesn't 12 bit vs. 8 bit just offer you more gradations between darkest and lightest ? I don't think that you actually extend your DR by using 12 bits, I think that it just gives you more colors.

Can any experts comment on this ?

If you shoot JPEG, it may help with the DR if you shot in RAW. In
RAW you can make use out of the 12-bits saved in RAW during
post-processing unlike JPEGs which are stored in 8 bits per pixel.

I don't use a D60 but read in this forum you don't seem to have to
save them as linear TIFFs to get the most out of them (unlike the
D30).

--
George
EOS D30 ~ 50mm f/1.4 ~ 200mm f/2.8 L II

 

[Al Pacheco]

The average scene, without artificial lighting, has about 6-8 stops from the brightest highlights to the deep shadows.

The D60 can easily capture that range. As a matter of fact, even the cheapest digital made can probably do well. My original Olympus D22oL had no problem,
The issue is how the range of brightness is treated by the camera.

An increase in dynamic range means a decrease in tonality. This is how the digital world works.

Canon has decided that tonality is most important in the midtones.
I'm glad they did.

With all the complaint threads here about underexposure,
overexposure, comparisons of D60 output to the 1D and P&S's...
seems that a very limited dynamic range is most often the culprit.
Some of the "underexposed", "flat" images I've seen here look very
much like film scans (from beautiful slides) with a squashed DR.

Certainly, can't expect the DR of the D60 to compare with negative
film (which has a huge DR)... but does it compare to slide film? I
would assume (hope?) it's better than either slide or negative
scanned with a consumer-grade film scanner?

Are there any documented measurements of dynamic range for the D60?
I can't find them. Perhaps experienced slide photographers now
with D60's would comment on DR comparisons?

Cirkitdude

 

[Frank E.]

Sorry about the late reply, I have been very busy.

In the studio I do not use on camera flash, rather, I use monolights. Everything is metered and set manualy..... thus the exposure is or should be right on.

I have found the D60 to be a bit under-exposed in all conditions using the in camera meter with or without flash. Turning off the super-imposed focus selector light helps in low light conditions. Generaly speaking, you need to use flash in low light conditions. I have used a 380EX and 420EX and they are very inconsistan... meaning they do not expose the sensor the same every shot. Even two identical shots. The 550EX is much better.

Now, if you set your exposure manualy and set the flash manualy (metered with a hand-held meter) the images are exposed perfectly every time.

I also still have a clunky old D30 and it has none of these problems. Flash exposure very consistant no matter what flash is used. Not knocking the D60, it is an incredible camera, the metering just stinks. Maybe it is fixable with a firm ware update.

Frank

I am sure this will go over big but it realy is an under exposed
problem with mixed lighting.

If the image is not properly exposed, then you'll have problems no
matter what. What I was saying applies to properly exposed images.
Or, more precisely, images where the exposure is centered on the
midpoint between light and dark. That may not be considered the
"right" exposure in all cases.

In the studio, I can meter normaly with great results. Head out to
the church and the images are flat. Do not increase flash exposure
but increase camera exposure and I think you get much better images.

Now I am confused. If your subject is the same distance from the
flash in both cases, and you're not relying on your environment for
bounce flash, then your exposure shouldn't be changing. Unless
there's some other factor that you're not mentioning, such as the
subject being in front of a highly reflective background. That can
make a TTL flash shut down before the main subject is properly
exposed.

Perhaps you can post a few example images?

Mike

 

[Nill Toulme]

Could you also do this by converting twice and applying exposure compensation to one of the conversions, and merging the two? Or is the linear/non-linear combination superior for some reason?

Nill

check out an article on Fred Miranda's site on using a darkened
exposure to control highlights mixed with a brighter exposure. You
can also do this with a single exposure using the RAW and Linear
RAW outputs of the D60 and mixing those together. With that
technique you can get around 8-10 stops depending on the scene.

 

[jrod7350]

Nill

check out an article on Fred Miranda's site on using a darkened
exposure to control highlights mixed with a brighter exposure. You
can also do this with a single exposure using the RAW and Linear
RAW outputs of the D60 and mixing those together. With that
technique you can get around 8-10 stops depending on the scene.

No, you could do that also, but I guess in extreme situations the Linear might be better?? I'm not sure about this since I haven't needed to do it yet-just read about it.

Jason Rodriguez

 

[Frank G.]

Phil used to measure the dynamic range of each camera and for a time gave this spec for all cameras. I dont know why he dropped this from his excellant in depth reviews?

One thing I noticed when I first tested the Fuji S1, Wow, what a wide dynamic range. Fuji specifically developed their CCD to have the best dynamic range. Does anyone know why Phil stopped rating camera's this way?

 

[Ron Parr]

Phil used to measure the dynamic range of each camera and for a
time gave this spec for all cameras. I dont know why he dropped
this from his excellant in depth reviews?

There were problems with the way Phil was measuring dynamic range. He was measuring the range of values the sensor could record over multiple shots, not within a single shot, which is what most people care about.

The numbers were misleading people and there were many fights about them. Ultimately, he stopped doing the tests this way and said that he was researching a better way to measure DR that was more consistent with what people expect in terms of performance within a single frame.

I guess this never happened or is still under development.

--
Ron Parr
FAQ: http://www.cs.duke.edu/~parr/photography/faq.html
Gallery: http://www.pbase.com/parr/

 

[Ken Phillips]

Correct ... more steps in between the same light/dark.

However, doesn't 12 bit vs. 8 bit just offer you more gradations
between darkest and lightest ? I don't think that you actually
extend your DR by using 12 bits, I think that it just gives you
more colors.

Can any experts comment on this ?

If you shoot JPEG, it may help with the DR if you shot in RAW. In
RAW you can make use out of the 12-bits saved in RAW during
post-processing unlike JPEGs which are stored in 8 bits per pixel.

I don't use a D60 but read in this forum you don't seem to have to
save them as linear TIFFs to get the most out of them (unlike the
D30).

--
George
EOS D30 ~ 50mm f/1.4 ~ 200mm f/2.8 L II

 

[alfred_schiele]

Hi!

... Hitting auto levels...

I would recommend "auto contrast" only. It does not affect colour balance. In combination with an appropriate fixed white balance on the camera and a personal basic curve correction I believe it is possible to get more stable results.

Unfortnately "auto contrast" isn't provided with the light versions of PS but it is in my eyes a far better function.

Regards, A. Schiele

 

[alfred_schiele]

Hi!

Your explanation is perfect! I couln't do really better (-:

My impression when playing around with D60 sample images was that they look closer to film shots than every other camera. Because of the low noise levels it is easy to adjust the histogram without much quality loss. I was impressed by this possibilities even when playing with JPG files. The potential with RAW should be far wider. Many other cameras files react far more on this type of corrections.

But I think there is also a misunderstanding between dynamic range and underexposure. Maybe to better understand this you should read the grandmaster of B&W, Anselm Adams... as well. It is more than actual even with digital.

Presuming a correct exposure, a "flat" image actually suggests that
the dynamic range of the subject was LESS than the available
dynamic range of the camera (or film).

When the camera's dynamic range is wider than the subject, the
darkest areas of the subject don't register as black in the camera,
and/or the lightest areas don't register as white. This is because
the camera has the ability to capture even darker or lighter areas,
but there aren't any in the scene. The result is a low contrast
(aka "flat") image.

When the dynamic range of the camera is not wide enough for the
subject, the result is an excessively contrasty picture lacking in
either highlight detail or shadow detail, or both.

In the first situation, you can easily fix the contrast in
Photoshop or whatever. That's where you notice the difference
between an 8-bit image and a 12-bit image.

In the second situation, you're hosed, for the most part, because
there's nothing you can do in Photoshop to recover detail that was
never recorded by the camera (or film).

Mike