I've been considering buying an inexpensive "carry body" to replace my aged Nikon D70 and my aged Canon 20D--those are the two cameras I actually carry and use when I want to take a camera along on risky trips, like saltwater salmon fishing, or on day trips to the zoo, where I want to have some "snaps", but not have to pack a big, heavy 3.8 pound pro Nikon body + lens, or a gripped Canon 5D and 24-105L, for days when I will be carrying a kit lens, or a 50mm prime, or 85mm prime, or a small,light wide-to-tele zoom like a 24-85 AF-S Nikon or my 18-125 Sigma for Canon APS-C.
So, I did a quick bit of research on the D60 and the XSi at Imaging Resource's web site. I found out the XSi's extra speed is largely only on paper.
The D60 raw buffer is stated as 6 images, or 100 Large Fine JPEG images at 3 fps, then it continues shooting at 50% reduced speed
The XSi raw buffer is stated as 6 images, or 45 JPEG images at 3.5 fps.
Please note that the XSi's ACTUAL,tested,real world RAW shooting allows it to shoot 5 frames in RAW, then it pauses to flush its buffer and to write to the memory card, whereas the D60's ACTUAL,tested, real-world RAW behavior is to shoot 7 raw files, and then write to the memory card.
Go to Imaging resource.com web site and look at the actual,tested specifications of the XSi and the D60. In Continuous shooting mode shooting Large Fine JPEGs, the XSi shoots one frame every 0.29 seconds. Shooting a 13 frame burst, it took 5 seconds to clear those shots to the memory card. The Nikon D60 on the other hand was slightly slower, at 0.33 seconds per frame for an over 20 frame burst, and it took 3 seconds to clear the shots to the memory card. Canon 13 frames, 5 seconds flush; Nikon 20, 3 seconds.
The cameras are very comparable,and the actual, real world difference in shooting speed per frame is four one-hundredths of a second slower for the Nikon, unless you need to shoot rather long,extended sequenes, where the Nikon's 100-frame JPEG buffer is just over twice as deep as the Canon's 45-shot JPEG buffer.
Note that the autofocus tests were done with a 70mm f/2.8 Sigma Macro lens on the XSi, a very fast-focusing, prime lens with the AF system optimized for a single focusing length, ie 70mm, while the D60 was tested with the "doggy"18-55 f/3.5~5.6 kit lens that it comes with. And yet, in continuous autofocusing mode, the Nikon's timing is 0.145 seconds per frame, while the XSi's timing is a slower 0.192 seconds per frame.
For indoor shooting, full AF with flash enabled, the Nikon again beat the Canon, taking 0.32 second to focus and make an expsoure, while the XSi took 0.328 seconds. In manual focus, with the AF shut off, the Nikon took -.095 seconds to make an exposure, the XSi took 0.115 seconds. The only time where the D60 took longer to focus and shoot was when the doggy 18-55 kit lens was used, taking 0.26 seconds from pressing the shutter button in Single-servo mode, where the Canon XSi took 0.160 seconds with the Sigma 70mm f/2.8 lens. From a "cold start" the $500 70mm prime 2.8 lens managed to beat the "free" f/3.5~5.6 kit lens. Surprising? Hardly.
And yet, in actual tests of things like Continuous AF, manual focus shooting speed, and full AF with flash indoors, the D60 still managed to beat the XSi in terms of pressing the button, achieving focus, and making an exposure, beating the XSi,either comfortably, or marginally. I think it's a little deceptive to test an XSi with a fast-focusing f/2.8 70mm telephoto macro prime lens versus a Nikon D60 with an 18-55 f/3.5~5.6 lens; the Nikon system is being handicapped with a rather cheap, slow zoom lens with lower magnification, which leads to much more difficult in-focus, out-of-focus discernment for the Phase Detection AF system. The Sigma 70mm 2.8 tested with the Canon XSi retails at $499, while the XSi is selling for $639 today at
B&H, making the 70mm macro a very unlikely actual "buy" for the typical XSi user. The D60 with 18-55 is selling for $599 today at
B&H.
I would love to see both the XSi and D60 tested on perfectly equal lens footing, like both driving their own maker's 200mm f/2 stabilized supertelephotos...at least, that would be like against like...