[Peter Fang]

Hi Steve,

I am also fairly sure that the SDM algorithm has been enhanced and is much more precise (less hunting) than on the K10D. This is just a matter of tweaking the settings in the AF computer but it could make a big difference and not to test this out is sheer laziness. To cut and paste number from a different camera is not just laziness however, its downright lying.

I agree with you that using cut and paste in this case is indeed lying. And the way they tried to brush away the skeptism and critism by their readers makes it even worse. If I were the publisher I'd fire those who were involved in this.

Peter

[dantuyhoa]

Sloppy work indeed . Popphoto is not my mag to buy any more . I shall pass the words !

[benplaut]

Hardly fair then, given that the Canon versions are USM lenses. The 50 1.4 is not the fastest focusing lens in the Pentax camp either - its contrast wide open is not that great and it affects AF performance. Also many of the FA lenses are slow to focus.

You cannot say anything about AF performance unless you compare with a cross section of popular lenses on a variety of targets under controlled conditions and publish the results. I have never really seen such a test performed by any magazine.

That's completely fair! There are generally two lenses that every brand has--a kit lens, and a 50/1.4. If you're going to test, the kit lens is essentially ruled out because there the results will almost always be sub-par to other lens. Pentax's 50 is old. Not much we can do about that.

[ricardobeat]

Hi Steve,



I agree with you that using cut and paste in this case is indeed lying. And the way they tried to brush away the skeptism and critism by their readers makes it even worse. If I were the publisher I'd fire those who were involved in this.

Peter

I'm not defending PopPhoto in any way, and I'll sure would like to see numbers from a fresh AF test with the K20D, but I think many people over here are overreacting on this. I think you didn't quite get the explanation:

Let's say that the measured AF speeds for the K10D were 0.41, 0.45 and 0.48 for given light levels. Then they go out and measure the speeds for the K20D, coming out at 0.40, 0.46 and 0.47. How can you verify that these are real differences and not sample variation? You can't, unless you redo all the tests including the K10D. That's what I think they meant with "avoiding confusion".

What I think is really unfair is that SDM performance is untested. They should give it a go just for comparison. There is no updated 50mm with SDM in the foreseeable future, as it was dropped in favour of 55mm.

[thibs]

PopPhoto may have chosen not to test the AF because supposedly it is the same as K10D's AF and place a link to the K10D review.

They chose to copy/paste which is at the very least (and being totally independant of the fact that the camera is a Pentax or even a low-end Kodak) a absence of any professionalism: doing this proves they are journalists wannabe and should be regarded as poor kids playing with toys.

[lol101]

Did it occur to anyone that they might have tested the K20 to be slightly slower than K10, thus the copy/paste to "avoid confusion"...?

In any case, they should post the data they got, no matter what they look like.

[thibs]

If the K20D isslower than the K10D it should be tested and written as such in the review, period !

[Rickster]

I wrote to the PP editors ... It will be interesting to see what their response is, if any.

[ricardobeat]

PopPhoto may have chosen not to test the AF because supposedly it is the same as K10D's AF and place a link to the K10D review.

They chose to copy/paste which is at the very least (and being totally independant of the fact that the camera is a Pentax or even a low-end Kodak) a absence of any professionalism: doing this proves they are journalists wannabe and should be regarded as poor kids playing with toys.

I think you may have missed this post. He says that they started testing of the K20D but didn't finish it as the results were "similar" to the old ones. A full test, even if the result is the same, would be preferable though.

--- quote begins ---

Executive Technology Editor Michael J. McNamara explains: Pentax openly admits that the K20D AF system is identical to the AF system on K10D with no claims for speed or sensitivity improvements. As noted in our several stories on the K20D, there are multiple design and performance elements that are identical between the two cameras since the K20D is actually a significant upgrade to the K10D but not a totally new camera design. Our preliminary lab findings supported the similarity in AF systems, and therefore we decided to incorporate the AF results from the K10D to minimize confusion. Had there been slight variations between the two camera models (as there would be), we would have received even more complaints after claiming both AF systems were the same. However, when the test data was incorporated into the K20D Test Results graphic in print and online, it is was accidentally transposed by one place. We have since updated the chart to correct this and believe the chart to be an accurate appraisal of how the K20D AF system performs compared to other DSLRs tested in the same manner.
--- quote ends ---

And a little insight on their lab (and that's the guy): DivePhotoGuide Visits Popular Photography Magazine's Testing Lab :: DivePhotoGuide.com - The Underwater Photo & Video Portal

[Rickster]

-- quote ---

" ... Our preliminary lab findings supported the similarity in AF systems, and therefore we decided to incorporate the AF results from the K10D to minimize confusion. ..."

This is such a self-serving statement. I think it is 'made out of whole cloth'.


Def: Made out of whole cloth - "utterly without foundation in fact, completely fictitious."

[ogl]

I wrote to the PP editors ... It will be interesting to see what their response is, if any.

Ask him to retest K20D with DA*16-50 and K10D with 1.30 firmware + DA*16-50 so we can compare with another systems.

[thibs]

[QUOTE=ricardobeat;182945]I think you may have missed this post. He says that they started testing of the K20D but didn't finish it as the results were "similar" to the old ones. A full test, even if the result is the same, would be preferable though.

Yes Ricardo, I saw it and I stand by my comments: you either provide numbers you got properly are ou do not post any numbers. Anything else is a fraud.

[*isteve]

That's completely fair! There are generally two lenses that every brand has--a kit lens, and a 50/1.4. If you're going to test, the kit lens is essentially ruled out because there the results will almost always be sub-par to other lens. Pentax's 50 is old. Not much we can do about that.

No it is not fair it is completely pointless. You are testing the lens more than the camera.

[ftpaddict]

In order to accurately measure AF speed, wouldn't it be most wise to use at least 3-4 different lenses in their tests? Say... the most popular focal lengths?

Even better, buy an identical third party lens that works with most mounts (I'm thinkin' something along the lines of Tamron's 18-250, in K, F, EF, Alpha mounts). That way the lenses are almost identical, excluding one very important variable.

[mattdm]

[QUOTE=thibs;182995]

Yes Ricardo, I saw it and I stand by my comments: you either provide numbers you got properly are ou do not post any numbers. Anything else is a fraud.

I think it'd have been "okay" (of questionable value but not dishonest) if they had labeled the table as a reprint of K10D data. As it is, I think it really is dishonest.

The best thing to do is going to be to write polite letters to the editor expressing disappointment and suggesting that a test with the SDM 50-135 would be interesting to see. (Not that it's their "moral obligation" to do so or anything, because that line of argumentation isn't likely to get anything done practically.)