7D dynamic range?

[Julio]

The dynamic range of a digital camera depends on one's acceptance of the noise. In turn, the acceptance often depends on the setting: one is accepting more noise in a very low-light setting, like a night stage shot than in a daylight landscape.

Thus it is not reasonable to talk about a specific number as if the DR would be something universal measurement.

Add to this, that the appearance of the noise too is a factor; think of the 5D2's pattern noise.

You're talking about an individualized subjective evaluation of dynamic range. I can buy that there is a "usefull DR" that is subjective according to individual taste and acceptance of noise.

But there is also a more objective measure of DR where the noisy end is not so subjective and is dependant on a measurable noise floor with (0 signal )compared to the full-well capacity of a pixel...says so right here in the glossary so it must be true

 

[Julio]

Edit: found it... seems ISO200 has either the same or slightly better DR in those measurements... what do you think?

There is nothing like that there. ISO 100 is better than 200 with all present Canons, which is pretty natural.

Your explanations usually make sense to me and I'm ready to listen to what you have to say about bronxbombers' data. But this is what I'm seeing:

5D
ISO100: 11.24
ISO200: 11.22

50D
ISO100: 10.96
ISO200: 11.07

40D
ISO100: 11.22
ISO200: 11.32

20D
ISO100: 10.56
ISO200: 10.51

7D
ISO100: 11.14
ISO200: 11.26

So..what am I missing?

 

[GaborSch]

If your argument is correct, then it would be best for me to use a digicam to shoot wide EV subjects because most sensors have clean ISO 800

but the DR with ISO 800 is quite much lower than with ISO 100.

Now does that mean film has less DR?

This is not so simple, for film and digital behave very differently. Their DRs can not be compared straightforward.

DR, for me, is the range of values where a media is able to render the dark and light areas in an image. A medium with a good DR renders these light and dark areas in such as way as to preserve the detail...

This is the crux. You are expecting that it "preserves the detail" - but what does that mean? When is the detail preserved? The darker you reach into the recorded data, the more noise and less detail you find - but it is not a sudden change, there is no boundary between the detail and the noisiness.

In other words, the proportion between details and noise decides if you accept that part of the image. This proportion depends on your preferences, on your budget, on the shooting situation, on the subhect, ...

The following captures show the raw channels of a 5D2 shot with pattern noise.

The average of the red pixels in the selected area is at the 10.6th EV of the dynamic range (i.e. close to the end of the 11th stop); the green is at 9.84 EV and the blue is at 10.3 EV.

The first one is composite (all four channels), the others show separated channels. It is obvious, that the noisiness is increasing with the decreasing pixel intensity, and the details vanish. As the green is the most intensive, the green channels carry enough details to make it reasonable to push the intensity to the end of the 10th stop of the DR, but the pattern noise of the red channel ruins the result (even though the green too exhibits pattern noise at that level, but much less, than the darker red).

The red channel does not carry any detail on this area, but the blue does some. Would you accept that level of noise, i.e. so few details? You may or you may not - but this is the consequence of noise.

and render the transition from one EV to the next smoothly or "gracefully," so that the eyes do not see a harsh boundary where the transition from one EV to the next happened

There is no such thing; the "boundary" between EVs is a purely human construction. However, this plays no role in the present context.

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

 

[GaborSch]

You're talking about an individualized subjective evaluation of dynamic range. I can buy that there is a "usefull DR" that is subjective according to individual taste and acceptance of noise

This does not mean, that it can not be measured. It means, that if you put the bar higher (greater SNR), the DR becomes lower (pattern noise, streaks, etc. pose different problems).

For example I measured the noise level of the 5D2 at different ISOs. Instead of giving specific numbers, I give the curves; if you pick a specific level of noise, you can see the dynamic range:

But there is also a more objective measure of DR where the noisy end is not so subjective and is dependant on a measurable noise floor with (0 signal )compared to the full-well capacity of a pixel...says so right here in the glossary so it must be true

I measure that too; see http://www.panopeeper.com/Noise/ReadNoise.txt

However, its usefulnes is limited. Some believe, that that is a constant value accross the intensity range, but it is not so. Anyway, the number you receive for DR but that measurement is pretty meaningless, except for engineers.

Instead, I created samples of different noise levels demonstrating, how a certain level looks (they look strange, because the samples are much smaller and I replicated them, so that the result be larger). 25% corresponds to SNR=4; 107% is somewhat worse than SNR=1.









--
Gabor

http://www.panopeeper.com/panorama/pano.htm

 

[GaborSch]

So..what am I missing?

Most of those numbers are based on the ReadNoise data I compiled (linked to in the other post). These numbers are so close, that one must not make any conclusion based on their differences. For example I measured much greater differences with my 40D (same ISO) between shots made only minutes apart.

Anyway, all those numbers are based on the noise measured on the masked pixels, and their meaning is, well, see my other post. I prefer to measure tzhe noise on targeted shots. Example:



The graphs of the 5D2 I linked to in the other post shows, that ISO 100 is a tiny bit better than ISO 200 (in my eyes it is not enough to use ISO 100).

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

 

[bronxbombers]

Edit: found it... seems ISO200 has either the same or slightly better DR in those measurements... what do you think?

There is nothing like that there. ISO 100 is better than 200 with all present Canons, which is pretty natural.

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

hmm i found it slightly better at ISO200 for the 40D/50D/7D
although that chart didn't take into account how it applied gain....

 

[bronxbombers]

So..what am I missing?

Most of those numbers are based on the ReadNoise data I compiled (linked to in the other post). These numbers are so close, that one must not make any conclusion based on their differences. For example I measured much greater differences with my 40D (same ISO) between shots made only minutes apart.

they were based on tests I made on blackframes using IRIS for the read noise and saturation point at lower ISO quite a few of the saturation point values at intermediate and higher did come from your table though (however differences there from camera to camera basically don't make any change at all to DR)

Anyway, all those numbers are based on the noise measured on the masked pixels, and their meaning is, well, see my other post. I prefer to measure tzhe noise on targeted shots. Example:

only the 7D #'s were based on masked pixels, the rest were averaged over the entire visible area.

anyway, that said the differences are so small as to be realistically meaningless yes, but since the results showed a trace better for ISO200 for so many cameras (not the 5D2) there may be something to it (even if it doesn't realistically matter)

 

[bronxbombers]

Anyway, all those numbers are based on the noise measured on the masked pixels, and their meaning is, well, see my other post. I prefer to measure tzhe noise on targeted shots. Example:

This is measuring something other than DR though (not that it is not important however and in some sense it is closer to the giving a sense of the avg look a photo, this is basically SNR at various luminances, which is an important measure and tends to put the 5D2 more above the APS-C at lower and intermediate ISOs than the DR measurements do)

 

[nikrphoto]

so what does that mean for a photographer. can he be a bit more relaxed while setting up the exposure?

 

[Julio]

You're talking about an individualized subjective evaluation of dynamic range. I can buy that there is a "usefull DR" that is subjective according to individual taste and acceptance of noise

This does not mean, that it can not be measured. It means, that if you put the bar higher (greater SNR), the DR becomes lower (pattern noise, streaks, etc. pose different problems).

For example I measured the noise level of the 5D2 at different ISOs. Instead of giving specific numbers, I give the curves; if you pick a specific level of noise, you can see the dynamic range:

That chart helps. I can visualize what you're saying.

 

[GaborSch]

they were based on tests I made on blackframes using IRIS for the read noise

Does not IRIS perform black level correction before calculating the standard deviation?

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

 

[photog7320]

I also was pretty surprised when I heard that digital cameras have 9 stops of DR, as I learned negativ film has 7-9 stops DR.

Unfortunatly I can't find the website right now, but I once saw a comparison on a website that showed that film is more like 14 stops DR.

Here's a pretty good test:
http://forums.dpreview.com/forums/read.asp?forum=1031&message=32891438

That's a horrible test. It's not even a test of DR, only highlight range, and it doesn't even test that correctly. Most of a DSLR's range is on the shadow side.

The proper, measurable, scientific way to test DR is the same way dpreview.com does it: a single properly exposed frame of a Stouffer transmission step wedge.

Films vary dramatically in their characteristics. DSLRs have better DR than slide film, but print films are all over the place. DSLRs are better than some, worse than others, roughly equal to the rest. DSLRs are also, in a sense, all over the place depending on settings and processing. But you can pull a lot out of a 14-bit RAW file, especially if you're willing to develop and blend two versions. If you're shooting JPEG with the contrast setting shoved to the wall, don't expect nearly as much.

In my experience when a scene exceeds the DR of a DSLR RAW file, it really exceeds it, by so much that no print film would help. That's what neutral grad filters, exposure blending, and HDR are for.

 

[John Sheehy]

Again, that is only true of specific bodies. The 20D did not have enough highlight headroom for the RAW values to go as high at ISO 100 as they did at ISO 200, and read noise was roughly the same.

It was back in the 20D days when I got in the habit of shooting @ 200... I could have sworn there was a graph posted here by someone in the last couple of weeks that showed a slight bump up in DR for several Canon bodies @ 200, including the 7D but I can't seem to find the thread atm... I'll look for it.

Edit: found it... seems ISO200 has either the same or slightly better DR in those measurements... what do you think?

http://forums.dpreview.com/forums/read.asp?forum=1019&message=33018892

I don't have the time to look through all those figures, what they mean, and their ramifications. I can say this, though, SNR at 18% has no direct connection to DR.

It is possible that ISO 100 has less headroom on the 7D. I only replied to you because you seemed to suggest that there is some kind of "natural ISO", or electronic optimization in canons, in general, at ISO 200. The only thing which has ever given ISO 200 slightly higher DR is limited ISO 100 headroom.
--
John

 

[Julio]

I don't have the time to look through all those figures, what they mean, and their ramifications. I can say this, though, SNR at 18% has no direct connection to DR.

It is possible that ISO 100 has less headroom on the 7D. I only replied to you because you seemed to suggest that there is some kind of "natural ISO", or electronic optimization in canons, in general, at ISO 200. The only thing which has ever given ISO 200 slightly higher DR is limited ISO 100 headroom.

And ironically, I developed that misconception, if it is a misconception, largely in part from reading your posts (here and at the Luminous Landscape) and those of a few others back when the 20D was the new APS-C DSLR on the block.

Conventional wisdom here by the technically-minded for the 20D was most definitely that ISO 200 had the better DR. If I had the time I would research and find those 5-yr-old posts.

And yes, being an old fart I'm always ready to admit to misremembering things, but I don't think I am in this case.

 

[rwbaron]

I don't have the time to look through all those figures, what they mean, and their ramifications. I can say this, though, SNR at 18% has no direct connection to DR.

It is possible that ISO 100 has less headroom on the 7D. I only replied to you because you seemed to suggest that there is some kind of "natural ISO", or electronic optimization in canons, in general, at ISO 200. The only thing which has ever given ISO 200 slightly higher DR is limited ISO 100 headroom.

And ironically, I developed that misconception, if it is a misconception, largely in part from reading your posts (here and at the Luminous Landscape) and those of a few others back when the 20D was the new APS-C DSLR on the block.

Conventional wisdom here by the technically-minded for the 20D was most definitely that ISO 200 had the better DR. If I had the time I would research and find those 5-yr-old posts.

And yes, being an old fart I'm always ready to admit to misremembering things, but I don't think I am in this case.

I have to agree with you somewhat. What I remember were posts showing the "true" base ISO for cams like the 40D and 50D to be ISO 160 and that was where the DR peaked on the chart. DXOMark Sensor shows this for the 40D although it's slight and for the 50D they show the first ISO reference as 160 on the DR chart.

Bob
--
http://www.pbase.com/rwbaron

 

[GaborSch]

The question was if ISO 200 is better (creates lower noise) than ISO 100. No, it does not.

Another issue is, that the base ISO is sometimes a fractional ISO; in such case that may be the best setting, but that too is not self-evident.

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

 

[JimH]

Most of us probably all remember that the 20D and even the 40D use numeric processing to achieve the "in between" ISOs.

And thus, if you're shooting RAW, there is little point in using anything but the "full stop" ISO settings (100, 200, 400, etc.).

But does anyone know if the 7D continues this design philosophy, or if it might actually use analog amplification ahead of the A/D to arrive at the intermediate ISOs?

--
Jim H.

 

[Julio]

Slow day at work and researching this issue is more fun, so...

Here's a link to an exchange between you and Victor Engel 4.5 yrs ago. You, in fact, are skeptical about ISO 200 being better and Victor is referring to this being posted in the forums "several times."

I haven't traced the original source (or culprit if you wish) yet but I'll keep plugging away.

At any rate, it's obvious to me now that I missremembered you as the source of the "200 ISO has greater DR" since your statements 5 years ago are consistent with everything you're saying now...my apologies.

Funny thing is that I have been using ISO 200 as my default for 5 years--quite happily I might add since I can use that extra stop for my shaky hand-holding--based on wrong information...not that it's that much of a big deal since ISO 200 does seem to be very, very close to 100 with respect to DR.

Quite fittingly, while I was searching I ran into an unrelated post by you where you said something to the effect that "It's a fact that missinformation is often repeated as fact..." true, that

The link: http://forums.dpreview.com/forums/read.asp?forum=1019&message=12379036

 

[JimH]

Those are handy graphs and examples.

I think most people want to see hard, cold, objective measurements so that they can make judgments comparing cameras or ISOs, etc., so that they can come up with easy rules for what ISOs to use, etc.

But the problem is that all of this is subjective in the end.

It's the same with audio, too.

The point at which you suddenly decide that the SNR is too bad to tolerate depends on a lot of factors and personal taste, etc.

Back in the days of analog tape, there was a lot of personal taste involved in how "hot" to record things.

Record hotter, and the tape noise ends up lower with respect to the average signal level. But you have higher levels of distortion in the loud passages.

Record lower, and the tape noise ends up being higher with respect to the average level, but you have lower distortion in the loud passages.

So different people, using the exact same recording system, ended up with different signal to noise ratios in their final product. But who's to say which "sounded better"? I tended to find tape hiss less objectionable than distortions. So I tended to record fairly low. Others preferred to lower the tape hiss by recording hotter, and could tolerate the distortion in the loud areas in trade for that.

And then we get into the question of "pattern noise" versus random noise. I suspect this ends up being very similar between audio and digital image capture, too.

If the non-signal artifacts in the "recording" have no pattern to them (that is, they're truly random in nature), then our eyes or ears can put up with them at a much higher level than if that non-signal artifact has a discernible pattern to it.

Vertical or horizontal banding in an image is a lot like having a low-level "whine" in the background of a recording versus random noise. The relative power level of that "whine" may be the same as the power level of the random noise, but your ear/brain can pick it out easily, and it'll be annoying. And the relative power level of a bunch of lines in an image might be no higher than another image's random noise, but the image with the lines in it will be very ugly to most people.

Our brains tend to look for or listen for patterns. So reducing pattern noise is paramount IMO.

It appears that the 7D is pretty good in this regard.

To me, therefore, I'd probably tend to find its DR to be greater than a camera with the same "measured" DR but which had a more defined pattern to it's "noise".

If we got a batch of film that was made when the coating machine was having a problem, and it had an uneven coating thickness such that there were horizontal "bands" in it, people hated it. And yet, the intensity of those bands didn't need to be very great for it to be unacceptable. You might even argue that this "banding" was below the "noise level" of the grain. But it still looked awful.

So I think the nature of the "noise" in question can have a huge impact on whether or not it's acceptable.

Two cameras or ISO settings might have the same DR or SNR by some measurement, yet the images may not look the same.

--
Jim H.

 

[Julio]

Good analogy, I liked that.

In digital photos I tolerate fine luminance noise and leave some there if I have to in order to preserve fine detail in the dark areas. I also have a dislike of blown-out areas so I often find myself trying to recover (or maybe remove them altogether in PS) even small hot spots since I find them to be attention grabbers.

If you put that all together, I guess I'm already maximizing the DR in all my captures.

Back in the days of analog tape, there was a lot of personal taste involved in how "hot" to record things.

Record hotter, and the tape noise ends up lower with respect to the average signal level. But you have higher levels of distortion in the loud passages.

Record lower, and the tape noise ends up being higher with respect to the average level, but you have lower distortion in the loud passages.

So different people, using the exact same recording system, ended up with different signal to noise ratios in their final product. But who's to say which "sounded better"? I tended to find tape hiss less objectionable than distortions. So I tended to record fairly low. Others preferred to lower the tape hiss by recording hotter, and could tolerate the distortion in the loud areas in trade for that.