[kenyee]

Since you are a measurbator, please tell me what percent of the canon shots are missed ?

I'd also like to see this with various Canon cameras (1D, 40D, 5D) and Nikons (D80, D300, D3). With enough light, someone w/ a K10D said the 40D gets more photos, but they're not necessarily perfectly in focus. With the 1D I'd expect more in-focus shots. With the 5D, I'd expect less shots in focus than the 40D since it's so old.

Suggestions for test: get a remote controlled car (not too fast), start it at some point 200' away, use the same lens (e.g. 70-200/2.8 or some other lens w/ big aperture like a 50/1.4), have someone press the car's drive button while you hold down the shutter on the camera. Record the #shots taken by each camera, then mark the ones you think are exactly in focus aimed at the front of the car. Repeat 3 times. Average the results in Excel and put up a graph of the #keepers.
Repeat the test for panning (car moves horizontally from left to right), and diagonally (car moves left to right and towards you).
Repeat for low light (you'd need some sort of warehouse you can control lighting in).

Up for it RiceHigh? I don't know of anyone who's tried to quantize AF performance like this, so you could make history


ken

[artxas]

Pentax using same SAFOX VIII AF module about 5 years!
hy not upgrade AF sensor?

I've made some AF tests today:
http://www.pocej.com/AFtest.jpg
I'll add K20D results next week.

[kenyee]

artxas: what do the numbers mean?

[wlachan]

Some may find this useful. I don't remember where this came from. My personal experience with the DS is that once the EV drops below 4EV (measured by Sekonic L308S), the AF starts to have trouble locking focus (AF-C/AF-S won't matter). There seems to have a sudden drop on its ability to AF below EV4. I pulled up a picture that my DS had problem AF with the DA16-45/4 and PhotoME indicated the AF sensor was pointing at 2.5EV at the time. A couple was walking down the stairs very slowly at the time (waiting for me to get the shot) and the entire time the camera was unable to lock focus at all. The best one I got was BF because it was attracted by the brighter ceiling light. Kind of embrassing.

[miriya]

I've made some AF tests today:
http://www.pocej.com/AFtest.jpg
I'll add K20D results next week.

where is under EV4?

[miriya]

Quote:
Is there any problem of Pentax AF module?
Yes, slugglish low light performance and serious AF huntings, Front Focusing when shot in tungsten lightings with faster and longer lenses, inability to trace fast moving object at a speed towards and away from the camera, inability to trace continuously for a randomly moving object (at least as compared to Canon's and Nikon's upper DSLR models)


Sorry, I mistyped original text.
[Is there any problem of Pentax AF module?] to..
[Is there any problem of developing/producing Pentax AF module?]

I know Pentax AF module's low light performance is seriously bad.



I taking a photo of CEO's breakfast meeting at hotel grandballroom,

so I always in low light situation.

It's too difficult to taking a photo of 'moving people' with my K10D.



I use 18-55 for wide, sigma 30.4 for nomal, 77.8 Limited for portrait.

shutter speed can cover with ISO and my 58AF-1.

but K10D's AF can't. I can't take a photo without AF.

there is no high quality photo without shot.

MF? moving people with f/1.4 or f/1.8?

If I can, I can take a photo of night scene without tripod.

Am I god? Pentax must improve their AF performance.

They use same AF module about 5 years.

Are they maniac??

use M42, use MF, Carl Zeiss ZS is cheap, oh yes SR can improve MF too!

or go Nikon, go Cannon, go Sony, go Olympus..



sorry for my poor English, I am bit angry.

doomed SAFOX VIII!! enough!

[mattdm]

I can't help but noticing that the lines for the 4/3rds systems just stop at points where at least Pentax works. Not saying the numbers are good, mind you, but at least they're there.

It's also worth noting that this chart really needs a third dimension — accuracy of results. Focusing in 1/10th of a second in near-darkness is nice and all, except when it gets the wrong thing.

shutter speed can cover with ISO and my 58AF-1.

Does the assist light from the Metz flash help at all? Or is that too obnoxious for the setting?

but K10D's AF can't. I can't take a photo without AF.

Wait, what? Why not?

[lol101]

I'd also like to see this with various Canon cameras (1D, 40D, 5D) and Nikons (D80, D300, D3). With enough light, someone w/ a K10D said the 40D gets more photos, but they're not necessarily perfectly in focus. With the 1D I'd expect more in-focus shots. With the 5D, I'd expect less shots in focus than the 40D since it's so old.

Suggestions for test: get a remote controlled car (not too fast), start it at some point 200' away, use the same lens (e.g. 70-200/2.8 or some other lens w/ big aperture like a 50/1.4), have someone press the car's drive button while you hold down the shutter on the camera. Record the #shots taken by each camera, then mark the ones you think are exactly in focus aimed at the front of the car. Repeat 3 times. Average the results in Excel and put up a graph of the #keepers.
Repeat the test for panning (car moves horizontally from left to right), and diagonally (car moves left to right and towards you).
Repeat for low light (you'd need some sort of warehouse you can control lighting in).

Up for it RiceHigh? I don't know of anyone who's tried to quantize AF performance like this, so you could make history


ken

I don't know if RH is up for it but I might.

Beside my K10, I also use a 30D from time to time (from a photo club) but as I didn't get to use it in an extended period of time, I didn't post much comparisons (plus their 17-55f2.8 died on me last time...).

But next week, they'll buy a new pair of 40D to loan their customers soooo.... it appears I managed to get their "old" 30D on "unlimited loan" and, if I like it, I'll buy it from them in a few weeks!

With a 70-200f4 IS and a 50f1.4, I'll be able to compare with my FA50 and DA 50-135.

[aegisphan]

Quote:

MF? moving people with f/1.4 or f/1.8?

Why would you want to take picture of moving people at such aperture? Since you have flash, even stopping down, you still have a decent shutter speed. From the second shot you provided, I don't see any problem with the center frame subject, which I assume the bowing person on the right hand side.
You might also want to learn to do manual focus. It's an essential skill regardless of any platform you're going to use.

[miriya]

Quote:
[shutter speed can cover with ISO and my 58AF-1.]
Does the assist light from the Metz flash help at all? Or is that too obnoxious for the setting?

- no, I say shutter speed, not AF speed.
I don't use 58AF-1's assist light for AF. It works not good.
And red assist light for my model(CEOs..) looks so bad, It is not manner.
I just use 58AF-1's P-TTL light.

Quote:
[but K10D's AF can't. I can't take a photo without AF.]
Wait, what? Why not?

- Pentax's indoor low light situation's AF performance is very bad.
If you can't feel it, take a photo of 3~4 year old children at indoor.
(ofcourse it is really difficult situation)
with K10D, I can take a photo of 'people doing exchanging name card' or 'bowing people'.
but just 'physically can', in real situation, K10D's AF starts hesitating.
bubububububububu----.
finally AF is done, people already moves away, just 0.5~1.5second.


sorry for my poor English

[miriya]

Why would you want to take picture of moving people at such aperture? Since you have flash, even stopping down, you still have a decent shutter speed. From the second shot you provided, I don't see any problem with the center frame subject, which I assume the bowing person on the right hand side.
You might also want to learn to do manual focus. It's an essential skill regardless of any platform you're going to use.

I have flash, but non flash image looks natural.
Second image is successed image.(just I showing you like this image is difficult to shoot.)
in many case I can't shot, because AF hesitating.

I don't use MF, because AF is more accurate and fast.
but, some situation(slowly moving person..) AF starts hesitate,
I have 2 choice.

1. wait until AF is done
2. change MF and shoot

but It just 0.5~1.5second. How can I choose?
Just full time MF or full time AF with some AF lack.

[wlachan]

I don't know if RH is up for it but I might.

Beside my K10, I also use a 30D from time to time (from a photo club) but as I didn't get to use it in an extended period of time, I didn't post much comparisons (plus their 17-55f2.8 died on me last time...).

But next week, they'll buy a new pair of 40D to loan their customers soooo.... it appears I managed to get their "old" 30D on "unlimited loan" and, if I like it, I'll buy it from them in a few weeks!

With a 70-200f4 IS and a 50f1.4, I'll be able to compare with my FA50 and DA 50-135.

Would love to hear your impression. I have been considering 40D with EF-S17-55/2.8IS & EF70-200/2L IS too. Dim light AF is my primary concern. But at the same time, the common err99 sudden death has me worried.

[VHDEL]

I don't know if RH is up for it but I might.

Beside my K10, I also use a 30D from time to time (from a photo club) but as I didn't get to use it in an extended period of time, I didn't post much comparisons (plus their 17-55f2.8 died on me last time...).

But next week, they'll buy a new pair of 40D to loan their customers soooo.... it appears I managed to get their "old" 30D on "unlimited loan" and, if I like it, I'll buy it from them in a few weeks!

With a 70-200f4 IS and a 50f1.4, I'll be able to compare with my FA50 and DA 50-135.

Would there be any chance of you testing the AF tracking speed/accuracy of the K20D compared to the 30D/40D

Over 80% of my photos are of my two young kids & although I have tried testing the AF speed using a DA*16-50, i have difficulty capturing a sequence of shots of my 4 yr old running towards me.

Right now I'm on the fence deciding if I should look at changing systems.

I'm sure that the next Pentax model will focus on further AF improvments but there will be long wait & my kids are growing up fast.

Members here who have the K20 say it does focus faster & in lower light which has me waiting for further conclusive tests pic's.

Cheers

Del

[lol101]

Would there be any chance of you testing the AF tracking speed/accuracy of the K20D compared to the 30D/40D

Over 80% of my photos are of my two young kids & although I have tried testing the AF speed using a DA*16-50, i have difficulty capturing a sequence of shots of my 4 yr old running towards me.

Right now I'm on the fence deciding if I should look at changing systems.

I'm sure that the next Pentax model will focus on further AF improvments but there will be long wait & my kids are growing up fast.

Members here who have the K20 say it does focus faster & in lower light which has me waiting for further conclusive tests pic's.

Cheers

Del

Over the last two weekends, I have tried getting some pics of my 5 year old and his mom learning rollerskating.

So far, my experience on these really slowish (but nor really constant speed) subject have been pretty bad with the K10 + 50-135 SDM. Maybe 10 to 15 % of in focus shots, no matter what I try.

I have taken possession of my new 30D + 70-200f4L IS USM (+ 50f1.4) yesterday and will test both systems this weekend (if weather permits) and report my feelings.

From the few shots I took yesterday at home with the 30D + 50f1.4 and K10 + 50f1.4 (no flash for both) all I can say is that when things are moving and lights are dim there are 0% chance of getting anything good with both systems.

When flash are used, the 30D focuses faster and can track focus better than K10 but that's still not sufficient to get a good hit rate.

I'll try to post some pics to show you the kind of light (or absence thereof...) I am talking about but with both cams, I was using 1/15s f1.4 ISO 1600 without the flash....

Another thing is: the 70-200f4.... WOW! Incredible lens on all counts!

[RiceHigh]

For the application you mentioned, I don't think any Pentax DSLRs can do it, K20D neither, I bet.

Would there be any chance of you testing the AF tracking speed/accuracy of the K20D compared to the 30D/40D

Over 80% of my photos are of my two young kids & although I have tried testing the AF speed using a DA*16-50, i have difficulty capturing a sequence of shots of my 4 yr old running towards me.

Right now I'm on the fence deciding if I should look at changing systems.

I'm sure that the next Pentax model will focus on further AF improvments but there will be long wait & my kids are growing up fast.

Members here who have the K20 say it does focus faster & in lower light which has me waiting for further conclusive tests pic's.

Cheers

Del