Canon has quite a confusing mess of drivers and software for their digital cameras. I’m going to try to sort them out here. Recently in September of 2006 Canon updated some of this stuff, so it’s worth paying a visit to their web site to check it out. On the, you will have to make some selections on pull down menus:. Select a Category: EOS SLR Camera Systems. Select Product Type: Digital EOS Cameras.
Select Mode: Pick your specific camera model For this example, I picked the EOS Digital Rebel XT. Note that you need different software and drivers depending on your particular camera model, and your particular computer operating system.
The Best Free Camera Drivers app downloads for Mac: Nikon Camera Control Pro Canon PowerShot ImageBrowser Canon EOS Digital Rebel/EOS 300D Digital Can. To see if your product is compatible with Mac OS X, please click on the [+] below next to your model series. Visit our Self-Help Center landing page for other OS-compatibility information for your product. Canon will continue to update this page with the most up-to-date information regarding software and drivers for our products. As is typical for canon sd1000 digital ELPH series, this point-and-shoot camera lacks manual aperture and shutter speed controls, but offers a series of scene modes, exposure compensation, and canon sd1000 and auto ISO.
Most of these pieces of software are updates for previous versions. You generally must have a previous version installed for the updater to work. The original versions are usually supplied with the camera when you buy it. On the next page that comes up, select the Drivers and Software link. The following drivers and software are listed for the Rebel XT: Windows Drivers. WIA Driver 5.6.0 for Windows XP. TWAIN Driver 5.6.0 for Windows 98/Windows 2000.
WIA Driver 5.6.0 / PTP WIA Driver 1.0 for Windows Me Windows Software. EOS Utility 1.1 for Windows XP and Windows 2000? Ha, that and a bunch of other things they should do if they are really serious! Below is a revised list from an old posting 1) 30sec exposures via sdk 2) Ability via sdk to take darks (shutter stays down for exposure) 3) Control of EF lens focusing via sdk – attain critical focus via software autofocus techniques 4) Document exactly what processing their DigicII processor does to the RAW – although I would guess that they would treat this as a trade secret, but it might provide some insight into better ways to handle the noise post processing. For example (if it does this) open up the information as to how to mark hot pixels so we can remap the sensor. At the very least release this kind of info to the likes of Mike Unsold under NDA. 5) Release full details of the RAW format.
I know from a friend and I’s experimentation with dcraw sources that an there is an unexposed strip of pixels left and top. This could be used for very accurate dark frame matching. What else might lurk in there? 6) Build an intervalometer into the firmware, controllable via the menus. It would be trivial for them to do this, but might impact sales of their remote.
7) Not specific to astrophotography, but expanding the dynamic range would be a useful improvement: 16bit rather than 12bit sensor, perhaps coupled to finer control over the gain/ISO setting. On to perhaps even less likey ideas 7) Put a temperature sensor next to the sensor and record in RAWs/exif. We can then more easily match darks against a library. Through plain crazy 8) Some medium format digital camera’s (Mamiya IIRC) now have exchangeable filter screens (in the same way you can change viewfinder screens) – so you can remove the high pass (and IR cut) filter for sharper shots. Allowing us to do this would give the best of both worlds – a full sensitivity astro cam, and a camera capable of being used without compromise during the day too! Might be a tad difficult to make this work out with the latest 400D’s dust removal system:) To certifiable 9) The canon sensors and DIGIC image processors are pretty spectacular peices of technology. I would love see Canon sell monochrome and bayer versions of them on the open market.
Many camera makers use Sony CCDs for example so why not get a slice of that general market – and let sbig/starlight/et al build a dedicated mono cooled camera. Those are some great ideas Paul!
But they will never do that stuff because they don’t make a camera for astrophotography. Oh, wait, yes they do!
Or did at least. You can forget about 4 and 5 becuase those are proprietary secrets, and, yes, 9 is never going to happen. I like the idea of a changeable filter in 8 a LOT. I would trade that feature for the “shaking the dust off feature” in a minute.
I think the shaking the dust off thing is kind of a gimmic. I shoot with pro digital cameras every day, and have for 4 years, and I change lenses a lot, and the dust has never been a problem. Of course, I don’t shoot landscapes with clear skies, so maybe that’s why I don’t notice.
Just going from 12 bits to 16 bits will not expand the dynamic range. The bit depth is the number of steps the dynamic range in divided into. The dynamic range is determined by the full well depth and the noise floor. From something I read somewhere else, I think they can fully encode the info they have with just 12 bits.
They don’t need 16. Just don’t ask me to remember where I read this Nikon has an intervalometer built into the camera for the D2H.
Canon Drivers For Mac Download
Perhaps I wrote that. I once made a calculation. Under a moderately dark sky and for sky noise limited (cf. Read noise limited) exposures, 12 bit already samples the noise quite well, and it is not too practical to sample the noise with the even finer 16 bit AD conversion. 16 bit is only needed when the noise is beat down by cooling (for thermal noise) and narrowband filters (for sky noise) at the same time.
This is why 16 bit makes sense on cooled CCDs. I think for DSLR, 12 bit is quite enough for most situations.
Yes, I know that too. And I also think that I know what’s going on there. I think one key is in this link: What Nikon does is to skip a few levels when the brightness level is high. The number of levels gets skipped is proportional to the square root of the brightness. This square root is what the “parabolic-like relationship” means in the above link. The reason for this square root is Poisson noise.
We all should know that Poisson noise is the altimate limit to the image quality. When such noise is big, there is no point to retain a level of AD conversion or color depth that is much finer than the noise level. For example, at ISO 400, Nikon D200 has a gain of about 0.4 DN/e-, according to Terry Lovejoy’s test. Thie means, for example, a brightness level of 4000 in the raw file corresponds to 10000 electrons, and therefore a +/-100 electrons of Poisson noise. In other words, the signal is 10000+/-100 electrons, or 4000+/-40 in the raw file. 4000+/-40 means, there is no difference between (for example) 4040 and 3960.
They are both consistent with the number 4000 (at 1 sigma level). Pixels with birghtness 4040 and 3960 may indeed have the same true brightness and it is the Poisson noise that makes them look different.
Since there is not a statistical difference between 3960 and 4040, why don’t we just write all of them as 4000 and save some file size? In my above example, I skip all the levels between 3960 and 4040 and write all of them as 4000. Nikon doesn’t skip this many levels. It still fina-samples the Poisson noise to a fair degree. And since Poisson noise increases as the square root of the brightness, the number of levels gets skipped can also increase as the square root.
This is what Nikon does. The lost levels actually do not have any impact to image quality. The image quality is limited by Poisson noise, not the sampling of brightness levels. Wei-hao, from your description the Nikon approach seems pretty sensible, and should provide a wider dynamic range for a fixed bit depth RAW. I guess in theory it could be readily converted into a linear representation if required (for photometry for example?). Jerry, when talking of expanding to 16bits I was expecting that the full well depth would be correspondingly increased. Having 16bits available should improve the dynamic range at all ISOs.
Currently to get enough dynamic range if bright objects are in the frame (and if star colour is important) we have to reduce the ISO (gain) or exposure duration. This is a tradeoff against the better sensitivity (due to lower noise) available at the higher ISOs. Hence we could shoot longer using higher ISOs and still achieve a good dynamic range, avoiding overexposure of the bright objects. Higher ISO’s also provide better sampling of the (all important) dimmest portions of the image (20D ISO1600.8e/ADU ISO400 3.1e/ADU). In theory this should lead to smoother dim nebulous areas without posterization.
Wei Ho, could you expand on your sampling the noise point – given a relatively dark sky, moderately low temperature, and plenty of frames – do you think the noise levels in the current crop of DSLRs are still generally too high to benefit from this aspect of higher ISOs? I’d like to understand the math, but its not my stongest point! П™‚ Shooting widefield at ISO1600 with a Ha filter and a fast lens (my current setup) I could benefit from the extra dynamic range as this would allow me to expose for longer.
Suprised #3 doesn’t light peoples candles, I would love this capability!