Much of the discussion regarding Sigma's cameras, with their Foveon sensors, has focused on the low megapixel count and "high per pixel sharpness" of the native output files. Yet at the print sizes most of us tend to use, neither of these characteristics, one of which tends to balance the other, dominates image quality. On the other hand, the dynamic range of the output and the range of tones within that range have a significant impact on image quality at any print size.

The dynamic range (DR) capability of a camera depends on the sensor technology used (see more discussion here and here); but in general, for a given type of sensor technology, DR will correlate with sensel (sensor "pixel") size. The DP1 has a very large sensel area. Given this and anecdotal reports from Sigma DSLR users about the DR of those cameras, I was optimistic about the DP1 DR.

Dynamic range test results can often be confusing. Two reputable sites will use the same method, such as Imatest, and come up with significantly different results. There are rarely practical examples shown to demonstrate the detected differences. Furthermore, dynamic range testing is often done on in-camera JPEGs, so the information gained doesn't reflect the entire sensor capability and is of limited practical use to me as a primarily RAW shooter.

My method is pretty straightforward. I take a single photo of a high DR scene, one which exceeds the DR capability of the camera sensor. I then "push" the image to +2EV and examine the shadow detail. Next, I push the same image to -2EV and look at the highlight information. The ability of the camera to simultaneously capture both the shadow and highlight detail from a challenging scene can thus be determined in a practical manner. It is critical that a single capture be used to look at both shadow and highlight latitude. With any camera, it would be easy to demonstrate good highlight detail from one capture and good shadow detail from another, since ambient lighting conditions and camera exposure settings can be vary between two captures.

I chose to compare the DP1 DR to that of the Canon 5D and E-420 since those are the most compact digital representatives of the 35mm and Four Thirds formats, respectively. The 5D was used with the Canon EF 28mm f/2.8 lens, and the Olympus with the ZD 14-42mm f/3.5-5.6 zoom at 14mm. The Canon image was taken at f/11 and 1/100s. The Olympus and Sigma images were at f/5.6 (for a similar depth of field to the Canon) and 1/400s (for an equal exposure). The three captures (one from each camera) occurred within a minute of one another, and the lighting conditions were not appreciably changing. All were at ISO 100. EXIF information is not embedded in the crops, but the RAW files are available for download at the end of the post. For details regarding processing workflow, please refer to the shootout introduction.

First, let's look at the highlight information. The specific area of examination is depicted by the yellow box in the following resized DP1 photo:



All crops shown are 100% without resizing. Here is the Canon 5D example, with the baseline conversion on the left and -2EV on the right:



As you can see, the tonality in the sky and car windshield was completely lost in the baseline capture and partially restored at -2EV. Likewise, a number of the small branches were restored.

Next we'll take a look at the E-420 performance in this area:



The baseline crop from the E-420 is similar to that of the 5D. However, the E-420 does not gain as much highlight detail by comparison at -2EV.

Finally, here is the DP1 sample:



Unlike the 5D and E-420, the DP1 has managed to retain some of the variation in the sky and car windshield, as well as the small branch definition, even in the baseline capture. At -2EV, there some further gain.

Looking at these results for highlight latitude, it is tempting to speculate that the DP1 was underexposed compared with the other cameras. However, as the EXIF in the RAW files will testify, the exposure settings were equivalent for the three cameras. Is it possible that in the moments between captures the ambient light dropped just in time for the DP1 capture? That is possible, though no such change was apparent. One way to get at this possibility is to see how the DP1 did from a shadow standpoint in the very same capture. One would expect that if a lower exposure is the exaplanation for the strong retention of highlight detail, that shadow detail would be compromised.

For shadow detail, I examined the two regions depicted in the following image:



Here is the Canon 5D examples, with the baseline conversion on the left and +2EV on the right in each of the two regions examined:



The additional apparent shadow detail at the +2EV setting comes, predictably, at the expense of greater noise.

Next we'll take a look at the E-420 performance in this area:



The comments from the 5D crop apply here as well.

Finally, here is the DP1 sample:



The pushed DP1 shadows were, to me, surprisingly clean. Clean enough to make me wonder whether some noise reduction is being applied by SPP during RAW conversion.

Based on this comparison I would conclude that DP1 exposure latitude at the base ISO is similar to that of the 5D and exceeds that of the E-420. Other comparisons I have done have been consistent with this observation.

You can download the RAW files for the three images used in this comparison by clicking the links below.



In Part 3 of the shootout, we'll take a look at the overall detail captured using the DP1 compared with the Canon 5D and Olympus E-420. Part 4 will compare the low light, high ISO performance of the same three cameras in addition to the current compact camera high ISO champ, the Fuji F31. The results may surprise you!

Addendum:
Sean Reid of Reid Reviews reminded me to discuss the impact of overall lens contrast on effective dynamic range. A lower contrast lens will tend to lighten the tones in the shadows (move them away from pure black) and, in doing so, may provide separation between shadow detail and the noise floor. The noise floor is the level at which the detail in deep shadows is lost among the file's digital noise. The effective dynamic range that a digital camera can record, therefore, is noticeably affected by lens contrast.

To truly isolate the DR ability of the cameras being compared, one would have to use the same lens on each camera. Clearly this is not possible with the three cameras being compared. Thus it is best to think of the results shown above as reflecting the particular camera-lens combinations used.

Read More......

Part 1 - Introduction
Part 2 - High ISO Noise
Part 3 - Dynamic Range

In practical terms, the dynamic range (DR) of a camera refers to it's ability to capture image detail in both very dark and very well lit areas of a single exposure. Using conventional sensor technology, this capability tends to correlate with sensor pixel (sensel) size. Working with RAW files generally allow one to access a greater portion of the DR captured in an exposure, and combining multiple exposures is a commonly-used method for capturing the dynamic range of a scene. For more on DR and the explanation of why pixel size matters, this DPReview article is a good place to start.

With the S3 and S5 DSLRs, Fujifilm implemented a novel sensor technology which incorporates two types of pixels of different sizes. The larger pixels are of the conventional DSLR type, whereas the smaller pixels capture additional "bright area" information. The simultaneous capture using these two different pixel types essentially allows the camera to obtain two different apparent exposures in a single image, and the result is class-leading DR based on additional highlight detail.

To date, no other manufacturer has introduced a production camera with such an effective system for improving the DR capability of a sensor. Instead, the approach taken by others has largely been to underexpose and then selectively map tones to incorporate greater highlight detail without actually increasing the DR. Highlight Tone Priority (HTP), featured on the latest Canon DSLRs, is an example of such an approach. The way HTP works is that the camera underexposes the shot to preserve highlights and then applies a proprietary contrast adjustment to "push up" low and midrange tones. For example, a Canon DSLR set to ISO 200 with HTP engaged will use aperture and shutter speed settings corresponding to those necessary for a proper apparent exposure at ISO 200. Meanwhile, the analog gain applied to the sensor is that of ISO 100 rather than ISO 200. Thus, the actual apparent exposure will be one stop underexposed. However, the in-camera JPEG processing is such that everything but the highlights is brightened as needed to simulate a proper apparent exposure in those areas. The result is that the highlights are properly exposed, and the expense is that the low and midtones are noisier because they represent ISO 100 (analog gain) "pushed" to ISO 200 (equivalent) using digital gain. This application of digital gain results in increased noise as explained in this rather technical thread and demonstrated in these examples from Joe Mama.

It is worth noting that the actual DR captured in a RAW file taken at ISO 200 with HTP enabled is no different than that of a corresponding RAW file taken at ISO 100 set for 1 stop of negative exposure compensation with HTP disabled. Nevertheless, this feature provides great value to the JPEG shooter for scenes prone to highlight clipping and may also benefit the RAW shooter who does not want to invest the time and energy in adjusting levels/curves after the fact (Canon's RAW conversion software will do the manipulation for you on RAW files obtained with HTP enabled). Click here to see Ron Purdy's excellent example of the benefit of HTP. Nikon's approach, termed Active D-Lighting, is also based on underexposure, though the implementation is different.

Fujifilm's press release for the F100fd states the following:

NEW Wide Dynamic Range

Dynamic Range, or the gradations of light that exist between bright and dark in any scene, can be easily detected by the human eye, but not by most cameras. Fujifilm first expanded dynamic range capture functionality on the professional imaging side with its FinePix S3 Pro digital camera. The technology, now perfected in the EISA Award-winning FinePix S5 Pro, has been brought to the consumer in the FinePix F100fd.

Through a combination of advances in Fujifilm’s new Super CCD HR VIII sensor and RP (Real Photo) Processor III, this wide dynamic range dramatically broadens tonal capability and guarantees exceptional rendition of photos with both bright highlights and dark shadows. This expansion of sensitivity captures greater detail, and subtle nuances of brightness and tonality that bring the photo closer to what the human eye actually sees helping to avoid the ‘white sky’ effect caused by the loss of fine highlight detail.

What isn't clear from this press release is exactly how the F100fd achieves "wide dynamic range" despite having tiny pixels (sensels). Does it truly, as they imply, feature a similar technology to that seen in the S3 and S5 DSLRs? Seems doubtful given that Fuji hasn't mentioned anything about different pixel types in the sensor. On the other hand, if it employing the same underexpose/tonemap approach being used by Canon and others, it would seem somewhat disingenuous to bill it as "wide dynamic range." Perhaps they are implementing an entirely novel way to boost actual DR. All they really say is that the wide DR is made possible by advances in the new sensor and onboard image processor. My guess is that the new sensor has implemented slightly lower read noise and that the rest of what we're getting here is similar to Canon's HTP.

Consider the following excerpt from the F100fd manual:


The relevant portions are as follows:
"When dynamic range is set to 200 % or 400 %, some [ ISO] settings cannot be selected. Also, some [ DYNAMIC RANGE] settings may not be selectable depending on the [ ISO] setting."
"Although wider dynamic range settings will allow you to take pictures with good levels of contrast, they will also increase the amount of speckling in your images. Choose the dynamic range setting that best suits the conditions."

So, basically, the use of "wide dynamic range" settings requires setting the camera to higher ISO values (probably ISO 200 minimum for 200% DR and ISO 400 minimum for 400% DR) and increases noise. To me, this strongly suggests that the wide DR function uses a similar approach to Canon's HTP, namely underexpose the whole image and then selectively "push up" everything but the highlights. Note that this approach to tone mapping does not preclude the possibility of other true contributions to improved DR such as measures to decrease read noise.

Let's look at an actual example. The following crops were taken from images provided at http://f100fd.seesaa.net/. Each image in the comparison was taken at a nominal ISO 400, f3.3, 1/100s, and these are 100% crops from the 6MP output mode of the F100fd. Click the image below to view at the intended size.


Don't know about you, but I am impressed. If Fuji is accomplishing this the way I think they are, then the noise we're seeing in the shadows and midtones of the DR 400% crop actually represents ISO 100 noise pushed two full stops. Yet to my eye, it's only barely more prominent than the noise in the native ISO 400 (DR 100%) crop. In fact, the most disturbing noise to me is the subtle pink and green blotching in the flat gray regions, and this is no different between the two crops. The highlight detail has been effectively reclaimed, and the rest of the image is virtually undisturbed with "good contrast levels" just as the manual claimed.

The Fuji F100fd has several strikes against it from the standpoint of a "serious" user, some or all of which may or may not matter to any given photographer. It lacks manual controls for aperture and shutter speed, RAW capability, and a viewfinder. The lens, in exchange for covering a long range, is relatively slow. The high ISO noise reduction cannot be disabled. However, as I mentioned in Part 1 of this series, there are now a number of full-resolution samples online to demonstrate that the F100fd output is of high quality, up to the standards for "serious" use. Add to that the fact that the high ISO output is unusually low in noise without what I'd consider an obnoxious level of smearing (though this is highly subjective) and that the highlight preservation is rather effective (albeit limited to use at high ISO settings), and I'd imagine that the Fuji F100fd will be the compact of choice for many of the photographers who don't enjoy postprocessing. We'll soon find out. I look forward to learning more with you when the F100fd hits the streets.

Read More......

Part 1 - Introduction
Part 2 - High ISO Noise
Part 3 - Dynamic Range

When I shop for a pocket camera, there are three questions I ask.

1) How well does the camera allow me to do my part? This is largely dependent on features such as those that follow:
- Aperture-priority, shutter speed priority, and true manual mode (independent aperture and shutter speed control)
- Focal length range
- Maximum lens aperture
- RAW capability
- Manual focus capability
- Quality of LCD and/or viewfinder
- Ergonomics
- Physical controls
- Operational speed (including autofocus speed, shutter lag, RAW buffer, etc)
- Camera size

2) How well does the camera do its part when I do mine? This consists of the following:
- Lens sharpness, distortion, contrast, bokeh, color
- Sensor ability to capture detail in optimal light
- Sensor detail relative to noise in suboptimal light
- Sensor dynamic range (effect of lens omitted here for discussion)
- Sensor effects on color
- Presence and effectiveness of image stabilization
- Battery life
- Weather resistance

The two categories above are largely what define the "seriousness" of a compact to me. Yet there is another category.

3) How well does the camera do its part if I don't want to do my part? Here we consider such features as face detection, which are of greatest importance to the person who doesn't want complete control over the picture-taking process.

One area which broadly falls under categories 2 & 3 is that of in-camera JPEG handling. There is no doubt that some compact camera photographers prefer to do "serious" work while shooting in JPEG mode, whereas others prefer RAW. For the former group, the in-camera handling of JPEGs is critical and raises the following questions:
- How flexible are in-camera settings for sharpness and noise reduction?
- How flexible are in-camera settings for color and contrast?
- What other operations is the camera performing with regards to tone mapping, and how much control can I exert over these?

What does all this have to do with the Fuji F100fd, and does the F100fd deserve to be called a "serious compact"? The second question is worth asking because the F100fd lacks important features from category 1, namely control of aperture/shutter speed and RAW capability. In this sense, it isn't a very serious compact at all. It is the heir of the F20/F40 more so than that of the more capable F30. Yet, having viewed a number of image samples from the F100fd, it is clear to me that this camera has the potential to produce "serious" results. In fact, despite having left out aperture and shutter speed controls, Fuji is referring to the F100fd as "the ultimate ‘F Series’ digital camera."

Boiled down to it's essence, the F100fd offers the promise of two potentially exciting capabilities. First, it is said to have a 1/1.6" 8th Generation Super CCD sensor capable of both recording 12MP of detail as well as maintaining high detail relative to noise at high ISO values in low light. Second, Fuji claims that "[the wide dynamic range] technology, now perfected in the EISA Award-winning FinePix S5 Pro, has been brought to the consumer in the FinePix F100fd." Before dismissing these claims out of hand, one ought to give Fuji the respect they are due. In terms of detail/noise, Fuji truly brought high ISO performance to compact cameras with the F10/11 and refined this capability with the F30. With regards to high dynamic range imaging, Fuji again successfully pushed the field with their S3 and S5 DSLRs. However, with today's technology, the F100fd's 5x zoom and shirt pocket size means a small sensor size; and while 1/1.6" is larger than most other small sensors, it isn't that much larger. Putting 12MP in a small sensor means a tiny sensel size. Part 2 of this series will address the question, "Can the tiny-senselled F100fd have a detail/noise performance similar to that of the 6MP Fuji F30?" Part 3 will address the F100fd dynamic range claims. Disclaimer: I have no inside information from Fuji, nor have I had the priviledge to yet test the F100fd. What follows is simply a discussion of the known technology with a measured dose of good-natured speculation.

Read More......

There has been quite a bit of discussion about the DP1 performance compared with other compact cameras. Much of it has been negative. Some have pointed out that the low ISO DP1 images upscaled to 12MP fail to capture noticeably more detail than those from the excellent Canon G9 viewed at native size. Others have pointed to the disappointing high ISO results demonstrated by in-camera JPEGs from the pre-production DP1 used in the Imaging Resource Comparometer. Personally, I have been most impressed with the tonal range I'm seeing in the image samples we've seen from capable photographers. Download Sigma Photo Pro (SPP), get Carl Rytterfalk's DP1 RAW pack, and play. If you use a Mac, get SPP 3.1 here. Carl might just be able to help out the PC users ;-).

Addendum: Carl has kindly allowed me to use his RAW files as examples. Image #6 of 10 (SDIM0408.X3F) in the RAW pack, the photo of shoes on a table, shows quite a bit of exposure latitude despite being shot at ISO 400. The following animated GIF demonstrates the highlight headroom. Watch as the highlight slider in SPP is adjusted. The GIF is just over 1MB in size, so it may take a bit to load depending on your connection. Also, please note that the apparent noise is actually due to posterization, which results from the conversion to a 256-color GIF.



2nd Addendum: It seems that the ISO 400 and 800 DP1 images have a shocking amount of highlight headroom during RAW processing, whereas the shadows are nothing special (in fact, an early impression is that they seem rather blocked up). Click here for an example of the highlight latitude. The likely explanation for this is that the high ISO files actually represent "pushed" underexposed low ISO files. This would explain how the highlights are restorable with negative EV in post, which would act to simply reverse the push. I'd like to see an evaluation of the DP1 dynamic range at the base ISO.

Read More......

I think most of us feel that sensor dynamic range (DR) is an important thing to assess in a camera. Is anyone doing it well? Of all the sites which tell us about DR on one camera vs another, I just don't have confidence in any of their methods. I don't mean any offense to any of the reviewers on the big sites, not that they're reading this anyway =). I'll try to explain myself here as well as I can.

I'll start with the table found 2/3 way down on this page at Imaging-Resource.
Those results just don't remotely jibe with my real world experience having owned or extensively used a bunch of cameras on that list. Furthermore, their methods - using the lowest ISO possible for each camera and different ACR versions for different cameras - just don't strike me as valid.

Next, consider DPReview. First look here, where Phil Askey, Editor of DPReview, had the following to say regarding the DR of the 5D versus the 40D: "Naturally the 5D's full frame sensor is going to give you better DR, IQ will depend more on your lenses than the camera." Now after reading that, look at what the DPReview 30D and 5D reviews show for dynamic range here, here, and here. Their testing method shows essentially no difference in DR between the 30D and the 5D.

Most like to say that DR correlates best to pixel size/density. That agrees pretty well with my real life experience, but not all that well. My 4-year-old camera with a 5MP 2/3" sensor seems to hold it's own in DR with my new camera with a 10MP 4/3" sensor, which has half the pixel density. I may be wrong about that; it's a preliminary opinion based on regular shooting without any rigorous testing. My old Fuji F30 had a lower pixel density than many of its competitors, but DR always seemed to be in limited supply with that camera. I'll chalk that up to in-camera processing since there was no RAW file available. My personal experience, again without rigorous testing, is that the 5D has noticeably more sensor DR than the 30D. Still, none of the rigorous testing on DPR and other sites seems to demonstrate a tight correlation between pixel size and DR.

Here's how DPR measures dynamic range in their words. From the 5D review:
"Our new Dynamic Range measurement system involves shooting a calibrated Stouffer Step Wedge (13 stops total range) which is backlit using a daylight balanced lamp (98 CRI). A single shot of this produces a gray scale wedge from (the cameras) black to clipped white (example below). Each step of the scale is equivalent to 1/3 EV (a third of a stop), we select one step as 'middle gray' and measure outwards to define the dynamic range. Hence there are 'two sides' to our results, the amount of shadow range (below middle gray) and the amount of highlight range (above middle gray)."

Sounds great, right? I think it would be fair to say that based on the DSLR test results on this DPReview, the Fuji S5 has the most DR, the Oly E-410 the least, and all the other DSLRs about the same. JPEG results don't matter much to me as I shoot almost exclusively RAW. Also, JPEG results don't tell us as much about sensor DR. Here's the deal from the E-410 review conclusion: "Secondly we were kind of disappointed with the dynamic range performance - dropping three quarters of a stop of highlight range can be the difference between a beautiful blue sky and one which is white and washed-out". Now take a look at the "RAW headroom" graphs on this page from the E-410 review and this one from the 400D/XTi review. I don't see where this 3/4 stop highlight loss is coming from. Even in their JPEG performance curves, it doesn't look like 3/4 stop (except vs the Canon), but in RAW - where it counts IMO - it's almost a dead heat.

I'm no Olympus fanboy. In fact, as I alluded to above, my early impression with the E-410 is that DR in RAW is more limited than I had expected it to be. However, I believe the following are probably true:
1) None of the major review sites are assessing DR in a way that helps us learn what we need to know.
2) The commonly taught principle that sensor DR correlates very closely with pixel size, while as useful a way as any to guess about a given camera's DR capability, is probably far from perfect in practice.
3) Perhaps some of us should give greater consideration to how lenses can maximize delivery to the sensor in such a way that makes the most of the sensor DR we're given, whether that be using ND Grad filters with a bright sky or keeping around an old, low-contrast prime for a day of digital black and white.

Read More......

The "megapixel war" continues. This week, both Panasonic and Casio have announced 12 megapixel sensors in their newest compact models. In general, high megapixels on a small sensor means small pixels, which translates to higher noise. While many people realize this, fewer realize that another sacrifice being made is in dynamic range. Unfortunately, none of the review sites do a great job of analyzing dynamic range performance, and few buyers consider this characteristic in choosing a compact camera. However, poor dynamic range performance forces one to choose to sacrifice either highlights or shadows when encountering a contrasty scene. The resulting dynamic range is poor, and though the tests don't show it well, the photos certainly do.

Here is an example of a quick snap with my G7, a camera into which Canon has stuffed 10 megapixels into a relatively small 1/1.8" sensor.

The colors are true and accurate, and as you can see from the full size image, the resolution is excellent. However, the highlights have been hopelessly blown. This is undoubtedly my fault, not the G7's. I could have exposed for the highlights by consulting the histogram. I could have used more negative exposure compensation and set the in-camera contrast lower knowing that the G7 has a tendency to clip highlights in such challenging conditions. However, I can tell you with complete certainty, than none of my other cameras would have clipped the highlights so severely in that image.

By comparison, here are two photos taken under similar lighting conditions with my four-year-old Leica D2. The full size versions can be viewed here and here.



Like the G7, the D2 has a small sensor. The D2 sensor is slightly larger, at 2/3"; the key difference is that it has only 5 megapixels. As a result, the pixel pitch is several times larger than that of the G7. The fact that, at the pixel level, G7 noise is not significantly different from D2 noise is a testament to the dramatic improvement of sensor technology since 2003. However, in the examples shown above, the D2 has nicely captured the dynamic range of a scene the G7 could not hope to have handled. Furthermore, both D2 examples shown here were shot at default JPEG settings. One can gain even more flexibility in dealing with dynamic range by shooting in RAW mode on the D2. Sadly, Canon has removed this option from their compact cameras, further compromising the already weak dynamic range performance.

Provided that the lens can resolve the detail, a 10MP sensor has enough pixels to capture an image from which one can crop out and discard 75% of the picture and still be left with an area that can be printed to 8x10" with good results. Most of us don't need more pixels in out pocket cameras. Poor dynamic range performance has been a weakness of digital photography since its inception, will flaw a photo even at 4x6" print sizes, and is largely ignored by makers of compact cameras as they compete for our money. Just something to consider when shopping for our next compact digital camera.

Read More......