From: Bruce on
On Sun, 16 May 2010 08:51:23 +0800, "Wilba"
<usenet(a)CUTTHISimago.com.au> wrote:
>Bruce wrote:
>> Wilba wrote:
>>> Bruce wrote:
>>>>
>>>> Some AF systems are even worse; Canon's being a
>>>> case in point, with fundamental errors built in to the system.
>>>
>>> Where can I find out more about this?
>>
>> There have been many discussions on Usenet. Search in Google Groups
>> for contributions by David Kilpatrick, who gave the clearest possible
>> explanation of Canon's AF problems. You will find them either in
>> rec.photo.digital.slr-systems or rec.photo.equipment.35mm.
>>
>> The problem is caused by Canon's choice of site for the AF sensors.
>> Nikon chose a different site and, as a result, their AF system has far
>> fewer problems than Canon's. But David Kilpatrick's explanation is
>> far clearer and more concise than anything I could manage.
>
>I asked David and he said that all he can recall saying is that the AF
>points are too close to the edges of the frame on the 7D, which means that
>the wrong thing gets the focus when all points are active.
>
>I was interested in the idea I've heard here that the configuration of
>Canon's AF sensors is fundamentally different to how other manufacturers do
>it, and that that explains some problems. David only mentioned something
>like that with the 5D II, and how the 7D seems to be tuned for speed rather
>than accuracy. He's not aware of any systemic issue across the brand that
>means that single AF point focus is any more erratic or inaccurate than for
>any other brand.
>
>Maybe we're thinking of two different things...


Maybe we are. If so, I apologise. I wish I had archived David's
postings on the subject.

From: Bruce on
On Sat, 15 May 2010 15:15:49 -0700, Paul Furman <paul-@-edgehill.net>
wrote:

>Bruce wrote:
>> Paul Furman wrote:
>>> Bruce wrote:
>>>
>>>> The same guy also had a useful contribution to make recently when we
>>>> were discussing cleaning DSLR sensors. Putting aside his perfectly
>>>> valid point that compact P&S and bridge cameras don't need their
>>>> sensors cleaned, he pointed out that some methods of cleaning DSLR
>>>> sensors could induce a static charge on the sensor causing it to
>>>> attract dust.
>>>
>>> That was BS.
>>
>> I phoned a materials scientist friend who confirmed it.
>>
>> He's a university professor at a Russell Group research university - the
>> nearest equivalent to your Ivy League. He thought it would make a
>> good final year project for an undergraduate student.
>
>
>Confirmed that it's possible?


Confirmed that the basic principle is valid. He seemed quite keen on
getting one of his students to look into it further.
undergraduate student.


>>> It's probably possible some combination of materials could
>>> create troublesome static but I tried the masking tape idea the other
>>> day and it worked fine.
>>
>> On a sensor?
>
>Yes. Just one quick test. It was awkward, the tape had to be cut thinner
>and hard to lay it down all the way to the edge, so I did two strips and
>a test shot showed that I left a dirty stripe down the middle so three
>strips and I still missed some of the edges & corners. It didn't get all
>the dust, some still glued on, I think from cigarette smoke :-(


Brave man!


>>> The intent was destructive, as usual. That would be fine to say: "watch
>>> out, tape might cause static" but he said: "it won't work, you're an idiot".
>>
>> Quite a few people posting here have a similar mentality, including
>> some who accuse him of being a troll.
>
>OK but that doesn't change anything.


No, it doesn't. But it establishes a yardstick. If you are looking
for pleasant discussions with polite, charming people, on balance it
would be best to give Usenet a miss.

Sometimes you do find polite, charming people (as in your case) but
that's an occasional bonus rather than the norm.


>>>> It caused me to consider different methods of sensor cleaning for the
>>>> Kodak DCS Pro 14n I recently bought, whose sensor is a dust magnet.
>>>
>>> What methods?
>>
>> I tried the paint-on goo we discussed a couple of weeks ago. It
>> seemed to work reasonably well on the rear LCD, so I tried it on the
>> sensor. To be honest, using it on the sensor was pretty terrifying. I
>> am clumsy at the best of times but I was particularly nervous because
>> of the risk of getting the liquid where it shouldn't go. So I didn't
>> cover the whole sensor.
>>
>> It wasn't much fun and I won't be using it again. For someone more
>> confident, more skilled and less clumsy, it might be OK, but I
>> wouldn't recommend it to anyone I like. ;-)
>
>Thanks for the report!


You're welcome. While I wouldn't use it on sensors, it is very good
on lenses. It's very similar to a produce I used a few years ago that
was purpose designed for lens cleaning. I stopped using it because,
on one lens, the plastic tab pulled off leaving the solidified goo
behind - with no easy way of removing it!

From: Chris Malcolm on
In rec.photo.digital Bruce <docnews2011(a)gmail.com> wrote:
> On Thu, 13 May 2010 19:27:04 -0500, Rich <none(a)nowhere.com> wrote:
>>
>>Depends if you want to preserve 3-dimensionality of the image. Using a
>>450mm lens would basically flatten it.

> Exactly. There is a reason why most lenses used for portraiture are
> within a very short range of focal lengths. Anything significantly
> longer or shorter gives a rendition that most people consider is
> neither natural nor pleasant.

Short focal lengths exaggerate nearer features, but I can't see the
problem with long focal lengths. They're usually inconvenient, because
you'd have to get very far back, they're a lot heavier, and need
higher shutter speeds or a tripod. But some portrait photographers use
them sometimes with perfectly good results. In terms of perspective
"distortion" over a face or body there's hardly any difference between
100mm and 500mm, whereas there's a huge difference between 100mm and
20mm. For outside portraits the narrow angle of view makes it easier
to select a specific kind of backgrouind and throw it out of focus.

--
Chris Malcolm
From: Chris Malcolm on
In rec.photo.digital DanP <dan.petre(a)hotmail.com> wrote:
> On 15 May, 16:34, Bruce <docnews2...(a)gmail.com> wrote:
>> On Sat, 15 May 2010 06:59:27 -0700 (PDT), DanP <dan.pe...(a)hotmail.com>
>> wrote:
>> >No, I do not think of camera design at all.
>> >Instead I think of binoculars and telescopes where the bigger lens
>> >diameter gives a better IQ.
>>
>> Does it really? ?Do binoculars and telescopes really offer better IQ
>> than camera lenses?

> I am not saying that. I said the bigger the lens diameter the better
> IQ is.

I can't imagine why. In binoculars and terrestial telescopes the
aperture is in effect provided by the iris of the human eye using it,
and optically the larger lens diameters are used to accomodate larger
iris openings (exit pupils) to accommodate lower light levels.

--
Chris Malcolm
From: Chris Malcolm on
In rec.photo.digital Paul Furman <paul-@-edgehill.net> wrote:
> Bruce wrote:
>>
>>> This is why diffraction-limited quality lenses are made regularly for P&S
>>> cameras. That much resolution is required or they wouldn't be able to
>>> resolve details down to their ~2um size photosite levels.
>>
>> Fascinating! Thanks for taking the time to reply.

> A sensor with 2 micron pixels would need an f/1.4 lens to operate
> without diffraction losses. At f/5.6 the smallest detail resolvable
> would be more than 3x the pixel density.

> A lens designed for 15x zoom isn't going to perform all that well at
> most focal lengths and distances. Surely you know this.

Modern design is changing that. For example, Tamron's famous 18-250mm
zoom performed so much better than their earlier 18-200mm that some
reviewers suggested it seemed to perform better than was theoretically
possible. In terms of comparison with good prime lens quality it's
very obviously inferior in image quality at its wider apertures, but
when stopped down to f8 and backed off from the extreme edges of its
zoom can produce images hard to distiguish from a prime on an A4
print.

--
Chris Malcolm
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Prev: Nikon Raw 64 bit codec
Next: Good news for shills...