[str8talk83]

Also, in regards to Thom Hogan preferring his D300 over his D3 and D700, I haven't read the article but perhaps, it's simply the additional crop factor and higher density pixels that he finds preferable when needing to shoot really long (like in Africa on a safari)?

I bet that's it. Having the crop factor is very handy when you are looking for reach. It makes composing and focusing through the viewfinder easier and allows the user to utilize the shorter(cheaper) telephoto lenses. This is especially important since Nikon's FX sensor is only 12MP. A lot of 5D Mark II users just heavily crop rather than using a APS-C sensor camera, but they have 21MP to mess around with.

[str8talk83]

Interesting read, but I'm not sure I understand how Mike Johnston's article supports the idea that he prefers his K20 more than a D700 at low ISOs. The article simply tests shooting a scene with the D700 (at ISO 6400 and 1/60 shutter speed) versus a K20 (at ISO 1600 and 1/15 shutter speed - both done handheld. The K20 had a DA 35 at @ F4, while the Nikon had a 35mm F2 @F4. When the pic was cropped, the K20 pic looked better.

Frankly, it seemed like a silly test, but what does it have to do with the K20 and D700 both at low ISOs?

I think that the point he was going for, was that IS can be more useful than upping the ISO. The Pentax was able to stay at a lower ISO due to it's SR mechanism while the Nikon had to use a higher ISO to get a sufficient shutter speed. The Pentax was able to stay at a low enough ISO with SR to give it the slight edge over the Nikon. I get his point on this.

I guess I'm safe though since I have a 17-55 IS with f/2.8. Nikon doesn't have this though so the point still stands.

[jonlee]

ISO 1600 is not low ISO. Misleading at best.

[str8talk83]

ISO 1600 is not low ISO. Misleading at best.

It's probably usable. I use 1600 ISO on my 40D and it is about the same as the K20D from what I've seen. The problem most people have is that they underexpose and bring up the exposure in pp, adding noise. I usually overexpose slightly and then bring down the exposure, giving me less noise.

[Cambo]

That's because by the time Pentax had focused on the spot where the skaters were, they'd already done a lap and come back to the exact same place where the camera was pointed...

nice one!

Cheers,
Cameron

[Rondec]

ISO 1600 is not low ISO. Misleading at best.

The point that Mike Johnston makes is not that you can shoot at low iso with SR, but that you can shoot at lower iso with SR. 2 stops makes a difference, even when comparing a full frame to a crop frame camera.

This is the classic question. SR versus higher iso. Clearly high iso can freeze motion, where SR cannot, but with a static image, SR is probably better because more detail will be preserved with the lower iso image. With the 35mm on the Nikon, it is not available with VR and so the only option would be to push the iso up to the point where camera shake ceases to be an issue (or use a tripod).

[jonlee]

The point that Mike Johnston makes is not that you can shoot at low iso with SR, but that you can shoot at lower iso with SR. 2 stops makes a difference, even when comparing a full frame to a crop frame camera.

This is the classic question. SR versus higher iso. Clearly high iso can freeze motion, where SR cannot, but with a static image, SR is probably better because more detail will be preserved with the lower iso image. With the 35mm on the Nikon, it is not available with VR and so the only option would be to push the iso up to the point where camera shake ceases to be an issue (or use a tripod).

Thanks, I understood Mike's article. My point was addressing another comment stating that the K20 was more preferred to the D700 at low ISO, and Mike's article being used to support that statement. Mike's article does nothing to show this, but rather, as you confirmed, tests out a very specific scenario where high iso isn't as effective as SR.

It's an interesting read, but what if the scene required one more stop, or one less stop? His test was using ISO 6400 vs 1600 with the relative shutter speeds of 1/60 vs 1/15, both camera's at F4. Say the scene could have been shot only requiring ISO 3200 vs 800 using the same shutter speeds, I wonder if the results would have been the same. Or, what if the scene required an additional 1-stop exposure. I don't think the shot could have been taken hand-held with either system. While the D700 can do ISO 12800, it's hardly usable. Not sure if the K20 can even do 3200. The other option of shooting the D700 at 1/30th hand-held with no VR even at 35mm would be challenging, as would the K20 at 1/8 even with SR. And, on top of all that, Nikon just release a 16-35mm F4 that has VR, so at least VR is an option at the wide angle now.

[Rondec]

The answer is that 3200 on the K20 (shooting RAW) will look better than 12,800 on the D700. There is a little more than a stop benefit in high iso shooting between full frame and cropped frame (that's assuming RAW shooting). As I said above, if you need to freeze motion, SR is not going to get you where you need to be.

Just because you can shoot 12,800 on the D700 doesn't mean that it will look great and if you get a good exposure, iso 3200 will look decent on the K20. If you don't believe it, see the high iso gallery by Jeweltrail.

I agree that the article has nothing to do with preferring the K20 to the D700 at low iso. My best guess would be that he likes the smaller size, or a particular lens, but who knows?

[cwood]

Walter’s conclusion was the Pentax Predictive Focus Tracking is better for his skating subjects."

In your face, AF whiners

odd since k20d does not have predictive AF.

K-7 does though.

[*isteve]

odd since k20d does not have predictive AF.

K-7 does though.

Actually, they both do, to the extent that it allows for time-lag before the shutter opens. However, with fewer reference points, its much less effective at close quarters, low contrast or random movement.

However I generally have no issues with motorsport, air-shows or other similar events provided the light is good and the background is not too busy. For bird or wildlife photography its not good at separating subject and background, and doesnt help you maintain focus on a small object when panning like a multipoint system does.

Is the Nikon D300 better in more cases, yes of course. However in some situations the Pentax system is entirely adequate such that it may even come down to the lens used. My K7 is certainly a major improvement over my K20 in that it takes a lot more smaller increments when focusing continuously (even if you dont press the shutter).

Incidentally, having used the 70-200 F2.8VR on a D2X I cant say its all that fast. Both Canon versions are faster, especially the 70-200 F4 which is really quick.

[pentaxdude]

So it was nothing to do with the fact he had a 50-135mm f2.8 on the Pentax and a Sigma 70-210mm f/4-5.6 on the Nikon?

The sigma Iwas using was a 70-200mm 2.8 hsm.

[cwood]

Actually, they both do, to the extent that it allows for time-lag before the shutter opens. However, with fewer reference points, its much less effective at close quarters, low contrast or random movement.

However I generally have no issues with motorsport, air-shows or other similar events provided the light is good and the background is not too busy. For bird or wildlife photography its not good at separating subject and background, and doesnt help you maintain focus on a small object when panning like a multipoint system does.

Is the Nikon D300 better in more cases, yes of course. However in some situations the Pentax system is entirely adequate such that it may even come down to the lens used. My K7 is certainly a major improvement over my K20 in that it takes a lot more smaller increments when focusing continuously (even if you dont press the shutter).

Incidentally, having used the 70-200 F2.8VR on a D2X I cant say its all that fast. Both Canon versions are faster, especially the 70-200 F4 which is really quick.

From both of what I have read and experienced there is no "predictive" capability in k20d but there is in k-7

these are all shot in AV mode (at f2.8 on 50-135) AF-C in default jpg settings with ZERO post processing (i.e. no extra sharpening). The rider is going a stead 20mph

K20D


100% crop


K-7


100%crop

[*isteve]

From both of what I have read and experienced there is no "predictive" capability in k20d but there is in k-7

these are all shot in AV mode (at f2.8 on 50-135) AF-C in default jpg settings with ZERO post processing (i.e. no extra sharpening). The rider is going a stead 20mph

Like I said the K7 takes more reference points and the K20 is not very good at acquiring focus to start with, but if it does it will compensate for motion. The car in this picture was doing over 100 and the shot was taken with a 400mm lens - you can see from the road that the focus point is spot on despite the speed of the car. And this with a K10D which was even worse than the K20D.

There were times when the K10D simply failed to acquire or react fast enough to movement, but it is predictive - just not very effective.

http://i.pbase.com/o6/89/257389/1/80885621.JN5VyWrs.IMGP2341postsmall.jpg

Note the K7 is way better even if it does slow down the frame rate a bit.

[cwood]

Like I said the K7 takes more reference points and the K20 is not very good at acquiring focus to start with, but if it does it will compensate for motion. The car in this picture was doing over 100 and the shot was taken with a 400mm lens - you can see from the road that the focus point is spot on despite the speed of the car. And this with a K10D which was even worse than the K20D.

There were times when the K10D simply failed to acquire or react fast enough to movement, but it is predictive - just not very effective.

http://i.pbase.com/o6/89/257389/1/80885621.JN5VyWrs.IMGP2341postsmall.jpg

Note the K7 is way better even if it does slow down the frame rate a bit.

the photo you posted is really nice but it has a VERY large DOF. I don't think that photo shows that k10 has predictive focus.

On your other point - you are sure correct about AF-C slowing down the frame rate on K-7. I think I've dropped below 2fps in some situations.

[gazonk]

The sigma Iwas using was a 70-200mm 2.8 hsm.

So you're the photographer from that interview? Great to have you here!
Do you use auto af selection with af.c on the k20d, or just center point? I've had very good results with the k10d myself, using af.c and center point. E.g. Shooting my kids snow-sledding at high speed. I used the da70 for that, and I think, as several have pointed out, that the lens may be the most important part of the equation.