davidodd

Lives in United Kingdom United Kingdom
Joined on Dec 14, 2010

Comments

Total: 28, showing: 1 – 20
« First‹ Previous12Next ›Last »
In reply to:

davidodd: Confused. If this mode moves the sensor 1 pixel then how do we get an increase in resolution above that from the actual photo-sites? If we are; then how is this different from enlarging a JPEG file (for which we have RGB at a pixel level)?

Apologies @deep7, I should have put an emoticon after the statement! I'm aware of the Foveon sensor, as well as the controversy and discussion related to claims by Sigma that it can derive a high spatial resolution than it has photosites (rather than just measure each site 'better' than a colour-filtered sensor).
So if we assume this is a 'Foveon' emulator; what's changed? I'm not doubting that this mode provides a super sharp image: but is it really more than the number of photosites and if so, how does it do it? This is a serious question and I am honestly interested in the math on how the additional colour information is rendered into additional spacial resolution.

Link | Posted on Nov 13, 2017 at 19:27 UTC
In reply to:

davidodd: Confused. If this mode moves the sensor 1 pixel then how do we get an increase in resolution above that from the actual photo-sites? If we are; then how is this different from enlarging a JPEG file (for which we have RGB at a pixel level)?

.... Now I can 'get' that guessing the missing R, G or B data for each photosite introduces uncertainty and therefore error, and therefore this method (or perhaps some magical sensor which records all RGB simultaneously!) would avoid this error and at the same resolution would look sharper (and more importantly) render without imputation of likely data from neighbouring pixels. This may in turn mean that any up-scaling is less likely to introduce/enhance these errors; but I don't see how FULL pixel shifting tells us ANYTHING about how much light hits the left/top half of the pixel so we can KNOW what's in-between them and increase the rendered resolution.

Link | Posted on Nov 13, 2017 at 16:33 UTC
In reply to:

davidodd: Confused. If this mode moves the sensor 1 pixel then how do we get an increase in resolution above that from the actual photo-sites? If we are; then how is this different from enlarging a JPEG file (for which we have RGB at a pixel level)?

Interesting, and don't get me wrong, and not saying the images aren't sharper than the sensor would suggest.... but I'm not sure these arguments explain it (or I don't get it).
e.g. If we take a monochrome sensor and say we can resolve more pixels than photosites, I'm not sure we'd be believed. However if you took just the green field of these pictures separately, that's exactly what we're told we have. How???

Link | Posted on Nov 13, 2017 at 16:33 UTC

Confused. If this mode moves the sensor 1 pixel then how do we get an increase in resolution above that from the actual photo-sites? If we are; then how is this different from enlarging a JPEG file (for which we have RGB at a pixel level)?

Link | Posted on Nov 12, 2017 at 19:17 UTC as 182nd comment | 9 replies
In reply to:

PaulSnowcat: I owned Tamron 16-300 and it is an unexpectedly good lens. Stopped down it produced a brilliantly sharp picture from corner to corner at every focus length! I've parted with this lens only because now I use FF. Tamron, please, PLEASE make a Full Frame 24-300 or even 24-450 lens with the same IQ that 16-300 PZD has!

I appreciate it doesn't go wide enough but the 28-300mm PZD is pretty good too.

Link | Posted on Jun 23, 2017 at 19:01 UTC

"This lets us compare field of view, depth of field and, to a reasonable approximation: low light image quality." - Yes. Yes I agree. BUT. Many of these pancakes vignette a massive amount. I have the Canon 22mm and while the middle of the shot might be around T3.2 (or so) the edges are a whole stop slower. I took a whole load of low-light shots at a party once only to find interesting (purple) noise patterns on the edges where the camera had cranked up the correction.

Link | Posted on May 15, 2017 at 15:45 UTC as 129th comment
On article Modern Mirrorless: Canon EOS M5 Review (1648 comments in total)

Fully agree with the review: it's a £1000 piece of kit with a hobbled (auto ISO) firmware. Unacceptable to me, and difficult to understand.

Link | Posted on Dec 21, 2016 at 08:27 UTC as 127th comment
On article Nikon reportedly eliminating 1000 jobs in Japan (518 comments in total)

This thread appears to have migrated onto "camera are doomed because everyone has a smartphone". To bring a (little) bit of science into this perhaps this page (http://petapixel.com/2015/04/09/this-is-what-the-history-of-camera-sales-looks-like-with-smartphones-included/) is worth a look? From a quick gander it looks like there was a massive spike in camera sales in the 2000's, but while that's dropping off the PRODUCTION of cameras still appears to be higher than the 1990's....

Link | Posted on Nov 8, 2016 at 16:41 UTC as 47th comment | 2 replies
On article Processor designer ARM acquires Apical (29 comments in total)
In reply to:

maljo@inreach.com: Sounds like a good move.
(I've never heard of either of these companies)

ARM based-processors are the most widely used in the world for mobile devices. You've probably half a dozen in your house (and one in your camera).

Link | Posted on May 20, 2016 at 14:37 UTC
In reply to:

Juck: That slow zoom had better be a hell of a performer for $1000.

I agree with Juck, and we need to wait for the reviews, but surely the closest comparison would be the (also) weather sealed, OIS, Tamron 28-300mm? Much cheaper and much lighter and pretty good reviews. In fact a 6D+28-300mm is pretty much the same weight as a A7+24-240mm.

Link | Posted on Mar 4, 2015 at 09:35 UTC

Maybe... yes?
There's a new body, an ultra wide, a standard and a tele zoom. All good lens too. And an excellent prime. Sure it's could do with a few more but that's surely a good start. AND you've got an excellent third party fish-eye and a travel zoom. AND you can put on canon's other primes if there's one you really need. Not saying it doesnt need developing, but it's a good start?

Link | Posted on Feb 11, 2015 at 17:56 UTC as 201st comment | 1 reply
In reply to:

G L: Not bad, but not very covenient too.
And also not cheap.
Not decided yet if it is worth it

It's free

Link | Posted on Jan 16, 2015 at 09:32 UTC

Have they released weights? For me that is far more important than size... and you can cram a lot of glass into a lens!

Link | Posted on Jan 11, 2015 at 20:44 UTC as 22nd comment
On article Lytro announces Illum light field camera (349 comments in total)

If Nikon sold this lens for their 1" cameras I'd think of buying into the system.

Link | Posted on Apr 22, 2014 at 18:55 UTC as 58th comment | 5 replies
On article Lensbaby releases 5.8mm F3.5 circular fisheye lens (20 comments in total)

Does an angle of 185 degrees mean it can take a picture of what's behind you?

Link | Posted on Apr 14, 2014 at 19:09 UTC as 5th comment | 2 replies

Maybe it's not the camera for me.... Too much for a 'one-trick' pony. HOWEVER, it shows that much of the kit needed for a FF camera can be squashed into a body about the size of an old 35mm camera (who'd have thought!). OK there's little things like a mirror box etc missing, but even bolting them on (or having an integrated EVF) it looks promising to me.
Next generation with SLT? Might make the plunge into Sony-world.....

Link | Posted on Feb 25, 2013 at 12:15 UTC as 11th comment
On article Just Posted: Canon EOS 6D In-depth Review (533 comments in total)

Honest question. One reason for the lower score (over the D600) is the autofocus system is 'less robust'. Can you expand on that? Granted there are less points, and not as many cross type (by a long way!), but the Nikon review hints at poor low light accuracy and failure, whereas the same section of the Canon suggests (even on the non-cross types) that auto-focus is accurate even in low light. Reading what is written in the two comparable sections I would probably conclude that the Canon has the 'more robust' system. Obviously this is not the case? Which has the FUNCTIONALLY better AF system?

Link | Posted on Feb 17, 2013 at 20:36 UTC as 27th comment | 1 reply
On article Just Posted: Canon EOS 6D In-depth Review (533 comments in total)

So what's better? A Nikon D600 or a Canon 6D? There's only one way to find out.... Fight!*

* Or perhaps a Challenge?

Link | Posted on Feb 14, 2013 at 16:33 UTC as 58th comment | 1 reply
On article Just Posted: Canon EOS 6D In-depth Review (533 comments in total)
In reply to:

davidodd: 337 comments and counting..... What's that about 'no such thing as bad publicity"?
Love mine. Love the way it feels, and functions. Great pictures. Good range of lens. Just about light enough to carry up a big hill without regretting it. Love the GPS and Wifi. Sounds mad but why not try one first???? Oh yes, because all the camera shops have closed....

Two very different answers. Isn't diversity wonderful!

Link | Posted on Feb 14, 2013 at 16:28 UTC
On article Just Posted: Canon EOS 6D In-depth Review (533 comments in total)
In reply to:

davidodd: 337 comments and counting..... What's that about 'no such thing as bad publicity"?
Love mine. Love the way it feels, and functions. Great pictures. Good range of lens. Just about light enough to carry up a big hill without regretting it. Love the GPS and Wifi. Sounds mad but why not try one first???? Oh yes, because all the camera shops have closed....

@agentul: Are you asking me? Yes indeed I've seen them. Got two in fact. Very useful tools, and for some purposes light, cheap and ideal. My GF3 with the Samyang 7.5mm has produced some of my nicest pictures recently, and all for less weight than a FF fish-eye.
On the other hand if you want shallow DOF then m43 isn't ideal, and can work out expensive (e.g. Panasonic 12-35mm for example). You can buy a second hand Canon 5D with an F5.6 kit lens and get SIMILAR (please let no-one start an 'equivalent' argument!) FOV/DOF pictures for £500.
If I was only allowed one kit? I'd keep the m43 because of the weight. What am I taking to on a photo-trip to Rome next week? Both!

Link | Posted on Feb 14, 2013 at 13:45 UTC
Total: 28, showing: 1 – 20
« First‹ Previous12Next ›Last »