My understanding is that the focus screen needs to be optimized for the AF system and the amount of light being passed to my eye is not accurate at wider apertures in regards to the actual depth of field (DOF). I think this also explains why you cannot see any changes in viewfinder brightness in the camera from f/1.4 until you reach f/2.8 when holding the DOF preview button. Can you (caution, bad pun ahead) shed some light on this?

Actually, focusing screens have nothing to do with conventional SLR autofocus systems. AF detection in most modern SLR cameras is carried out by a mechanism that's positioned in the base of the mirror chamber. Incoming light reaches this mechanism by passing from the lens through the semi-transparent main reflex mirror to a sub-mirror that redirects it downwards. The focusing screen, on the other hand, is positioned above the main mirror, which reflects incoming light towards it.

That said, it is true that most standard focusing screens for modern SLRs such as the EOS 5D are designed to provide a reasonable balance between viewfinder brightness and manual focusing capability. The Ee-A standard focusing screen for the EOS 5D is bright enough (and accurate enough) for manual focusing under most lighting conditions with virtually any EF lens regardless of maximum aperture. However, because of the design of the microlenses on the surface of the Ee-A, the depth of field shown through the viewfinder never appears be shallower than approximately f/2.8. Therefore, when using a lens faster than f/2.8, the depth of field in the resulting photograph may be shallower than what's shown in the viewfinder if a working aperture larger than f/2.8 is selected. This effect can be readily seen when comparing the viewfinder image to the LCD screen during replay, if you take time to look for it.

If this is an issue for you, Canon offers an optional focusing screen called the Ee-S Super Precision Matte Screen. This focusing screen uses more powerful microlenses than the standard Ee-A screen, with the result that out-of-focus areas in the viewfinder are more accurate to the actual depth of field in the resulting image. This has the effect of making it easier to determine the exact point of focus during manual focusing, especially with high-speed lenses like the EF50mm f/1.4 USM or EF50mm f/1.2L USM. However, it also has the effect of making the Ee-S focusing screen noticeably darker than the Ee-A screen when using lenses with maximum apertures smaller than f/2.8. (No free lunch!)

Thanks for your remarks in last month's Tech Tips column describing phase detection AF on the dedicated AF sensor vs. contrast detection AF on the image sensor. This inspires a question that has been on my mind for a while. Is there a physical difference between the electrical sensors that detect light on the AF sensor vs. the image sensor? Is there any reason why the image sensor can't record phase data for AF?

Dedicated sensors of the type used for phase detection AF in digital SLRs are somewhat similar to image sensors used for Live View in the respect that both contain photosites that are light-sensitive. If that's all there was to consider, then it would seem possible to use image sensors for phase detection autofocus measurements. However, the phase detection system also requires a set of secondary lenses to separate the light to be analyzed by an individual focusing point into pairs of beams that are then projected towards the AF sensor. Normally, there are no secondary lenses between the lens mounted on the camera and the image sensor, so phase detection AF with the image sensor is not available.

Having just bought a new PC with the Vista 64-bit operating system, I am disappointed that Canon has not produced a 64 bit CODEC for viewing their RAW files in Windows. This was possible in Windows XP and, it seems in 32 bit Vista. I realize other applications can be used but native Windows viewing is convenient. The Microsoft 'pro photography' Web site lists the support from various camera manufacturers and at least one - Olympus - has produced a Vista 64 CODEC. I think Canon is missing a trick here. It can't be that much work, can it? Are there any plans to release such a CODEC?

To date, there is no information to indicate that there are any plans to add support for 64-bit versions of Microsoft's Windows Vista OS to future versions of the Canon WIC codec, but that doesn't mean it will never happen. However, Canon's current proprietary RAW conversion software is compatible with both 32-bit and 64-bit versions of Windows Vista. There are two such applications: RAW Image Task, which is a sub-program that runs within ZoomBrowser EX, and Digital Photo Professional, also known as DPP. These software apps are bundled with current EOS Digital SLRs at no extra charge.

Although this is somewhat outside the scope of your questions, Canon also supports tethered operation for a select group of EOS Digital SLRs via Canon's EOS Utility software application. EOS Utility software is compatible with 64-bit versions of Windows Vista, but camera model compatibility in that OS is currently limited to EOS-1Ds Mark III, EOS-1D Mark III, EOS 40D, EOS Rebel XSi and EOS Digital Rebel XTi. All of these cameras are currently available except the Rebel XSi, which is due to be released this month.

I just took delivery of an EOS-1Ds Mark III and wanted ask a question about ISO to ensure I'm optimizing the quality of images from this fantastic camera. I've been doing studio work with my EOS-1D Mark II and II N on the extended ISO 50. Is there any quality compromise when using this lower ISO? I was using it thinking I would be getting better quality with the lower ISO, but I read something to the effect that it actually increases noise and that extending the range is only a method of light- not quality-management. Am I better off shooting at ISO 100?

The standard ISO range for the EOS-1Ds Mark III camera is 100 to 1600. Optional settings for ISO 50 and ISO 3200 are enabled via menu command, and are indicated on the camera's LCD data panel as "L" for Low and "H" for High, respectively. Part of the reason behind this is that these settings are not truly ISO 50 or ISO 3200. Instead, in the case of "L," you're looking at ISO 100 processed by the camera to achieve an effective rating of ISO 50. Similarly, "H" means that the camera is processing an ISO 1600 image to achieve an effective rating of ISO 3200. "L" images have about the same level of noise as ISO 100 images, but they have less "headroom" in their tone curves to recover overexposed highlights than do ISO 100 images captured with the 1Ds Mark III. This is typically not a problem with evenly lit scenes, but it may become an issue with scenes that have a lot of dynamic range to them, such as sunsets, backlit flash photos, etc.

ISO 50 can be helpful when you are trying to achieve a creative effect, such as a wide aperture to blur the background behind the main subject, for instance in portraiture. It can also be helpful when you are using studio strobes and you don't want to stop down the lens excessively, or in outdoor situations when you want a longer shutter speed to blur subject movement like the water in a waterfall. But if none of these situations apply, you are better off to use ISO settings in the 100 to 200 range in terms of overall image quality and maximum dynamic range.

I was wondering if you can answer a question I have regarding the High ISO Noise Reduction feature (C.Fn II-2) available on the Mark III cameras and the 40D. The following is a snippet from the last page of the Canon document called "Tailoring the Camera for Different Situations":

"High ISO images that are predominantly high-key subject matter (example: available-light ice hockey pictures in an arena) will sometimes show more of a 'salt and pepper' appearance from remaining luminance noise if the Mark III's High ISO Noise Reduction is applied. Users should run tests to see whether there's any negative impact when it's applied. This is one advantage of shooting RAW files; in either of the two supplied Canon RAW file processing software applications, Digital Photo Professional or RAW Image Task, you can apply or remove High ISO Noise Reduction, making it easy to see the same file 'before and after.'"

The snippet seems to imply that the high ISO noise reduction feature is a metadata tag which can be turned on and off via DPP or Raw Image Task and does not affect the RAW image that is actually recorded. However, my testing seems to indicate otherwise. When I shoot two RAW images, one with C.Fn II-2 turned off and the other with it turned on, and process it with Adobe's Camera Raw program, I get one image which has a lot of high ISO (chrominance) noise (C.Fn II-2 was set to 0 for this shot) and another with no high ISO noise (C.F II-2 was set to 1 for this shot). This seems to indicate that the Mark III actually does reduce the high ISO noise in the image before the RAW image is written out base on my assumption that ACR ignores the metadata tags in the RAW file and hence, should not be doing any type of processing to reduce the high ISO noise based on the setting of one of these metadata tags. Is my assumption wrong? Taking this a step further, if the high ISO noise reduction feature is implemented via a metadata tag, then how is this feature turned on or off in DPP? The only tool that I can find relating to noise reduction is the NR tab in the DPP tool palette.

Thanks for your question. Perhaps the article could have been clearer on this point, but according to my reading of it, the author was trying to compare in-camera high ISO noise reduction (by itself) to DPP noise reduction (by itself). In that comparison, DPP offers more control since it has separate sliders for luminance and chrominance noise reduction that are freely adjustable from 0 to 10 (in DPP 3.3) versus a simple Off and On setting in camera that makes no distinction between chrominance and luminance NR.

The article does not explicitly discuss whether in-camera high ISO noise reduction affects RAW image data with the Mark III cameras and the 40D. However, examination of comparison files in DPP clearly shows that it does not. I can't comment on another manufacturer's software application, but you may want to try the same experiment for yourself with DPP to verify this point.

Thanks for reading Tech Tips! That's it for now. See you in May!

You are invited to submit questions about photo equipment, imaging technology, or photo industry trends that may have a bearing on your work or interests. I cannot promise to answer all inquiries, but I pledge to do my best to address the issues that concern you. (Please use the e-mail link provided at the end of this article.)