PIX 2015
ProfHankD

ProfHankD

Lives in United States Lexington, United States
Works as a Professor
Has a website at http://aggregate.org/hankd/
Joined on Mar 27, 2008
About me:

Plan: to change the way people think about and use cameras by taking advantage of cameras as computing systems; engineering camera systems to provide new abilities and improved quality.

Comments

Total: 725, showing: 41 – 60
« First‹ Previous12345Next ›Last »
On Feature film shot on iPhone 5 coming to cinemas post (64 comments in total)

Meh. There was serious stuff shot on 16mm years ago and there are now $30 dash cameras that can better that. Sounds like a PR gimick... that worked.

Direct link | Posted on Jul 9, 2015 at 04:54 UTC as 18th comment
In reply to:

Lan: I still can't understand why they've launched this lens. There are a lot of 100mm 1:1 macro lenses out there already, and many of those are quite affordable too... Those competitors often include auto apertures, autofocus, and image stabilizers.

A 2:1 macro lens would have been interesting, but this is just perplexing.

I'll concede that the geared version might make some sense for macro videographers, but how many of those are there?!

photominion: Peaking on a rear LCD isn't the same. Hopefully, you mean using Magic Lantern with the 60D, which helps a lot, but manual focus holding the camera away from you to view the LCD is pretty awkward. A magnifier hood for the LCD can approximate an EVF, but with that a DSLR like the 60D becomes truly huge.
fmian: Neither are the best focus screens -- and a 3rd-party split+microprisms screen was what I used with my last DSLR.
photominion: I used PDAF-based focus confirm before I changed the screen in a DSLR -- the screen change was a huge improvement.

Basically, only peaking + magnified EVF works over the whole frame with the camera braced against your face, clearly shows DoF, and even works with the lens stopped down.

In any case, my point is that Samyang has a real opportunity to distinguish themselves again if they tune this design for mirrorless, and manual focus isn't "2nd class" on mirrorless bodies with EVFs, so that might sell more lenses....

Direct link | Posted on Jul 7, 2015 at 09:14 UTC
In reply to:

Lan: I still can't understand why they've launched this lens. There are a lot of 100mm 1:1 macro lenses out there already, and many of those are quite affordable too... Those competitors often include auto apertures, autofocus, and image stabilizers.

A 2:1 macro lens would have been interesting, but this is just perplexing.

I'll concede that the geared version might make some sense for macro videographers, but how many of those are there?!

I own three 100mm macros and one 90mm macro... none of which go to 1:1. Samyang has been doing a great job of making modern fixed-focal-length manual lenses and I expect this will be a great lens. The catch is that I'm a Sony user, so I have an EVF with focus peaking and magnified view... DSLRs are much less pleasant to use with manual focus, so I'm not sure they'll sell many to DSLR users. I think it might have been wiser for Samyang to make a design geared specifically to mirrorless (e.g., more compact by utilizing the shorter flange distance).... maybe that's still to come, as they did for their fisheyes?

Direct link | Posted on Jul 7, 2015 at 05:20 UTC
In reply to:

Pandimonium: it smears detail a little less aggressively than my 'old' hx9v. Pics look like they've been taken in superior auto mode. It uses hdr mode a lot then resulting in less contrasty pics.

There are probably a lot of evil lens properties being corrected in the JPEG pipe, so HDR tone mapping or SLog JPEGs might give you nearly as much good data as the system can provide while saving users the pain of working-out a good raw-processing pipe. It still would be nice to have raws as an option....

Direct link | Posted on Jul 6, 2015 at 14:18 UTC

If presenting a view from a public place of a "copyrighted" building for monetary gain is not ok without permission, it wouldn't take much to say that taxi drivers need to get permission from all the buildings that a passenger might see through the window during a paid ride. That's nuts.

I can imagine issues with simulating a trademark, or implicit endorsement of a product, but existing law would handle those cases reasonably. Explicit restrictions on use of photography typically are about an expectation of privacy -- extending that concept to inanimate objects (buildings) is a huge stretch.

The national park and Disney examples are fundamentally different in that those involve views from specially-designated areas. For example, there are military bases where photography is restricted in marked "sensitive" areas. I've never seen a copyright notice posted on a building so that it is clearly visible from all public areas that might view the building....

Direct link | Posted on Jul 5, 2015 at 21:44 UTC as 49th comment | 9 replies
On Canon warns about dangers of counterfeit camera gear article (153 comments in total)

A counterfeit product is one that is falsely simulating another -- including being incorrectly labeled as the other product. Stuff clearly marked as coming from a third-party manufacturer, including batteries, is NOT counterfeit. It is always smart to know what you're purchasing before you buy it, but in the USA any product that doesn't match the written specs provided by the seller is generally returnable for a refund; for example, that's true of stuff bought from random folks on eBay. The trick is to get specs in writing -- don't ever buy something based on your guess as to what it is nor someone's verbal description.

Direct link | Posted on Jul 3, 2015 at 02:23 UTC as 47th comment
On Sony: An eye on focus article (763 comments in total)
In reply to:

Random Asian Guy: Wow, that's seriously impressive.

I'm already expecting all the critics saying 'Mehhh, I've been shooting Manual for 35 years and can focus on the eye faster'

or

'That's nothing my XXXXX camera is much better, Sony is not a real Camera Company'

The problem isn't that "Sony is not a real Camera Company" but that CaNikon are not real consumer electronics companies. ;-)

Actually, Sony has a long history as an industry leader in electronic imaging from back when that meant video. They also inherited and heavily leveraged the Minolta legacy in still cameras. Not really all that surprising that Sony+Minolta is more innovative than CaNikon; they always were from at least the 1970s (which is when I started doing photography professionally, primarily using a Minolta XK).

Direct link | Posted on Jul 2, 2015 at 13:49 UTC
In reply to:

DStudio: Charming, yet painful (the video).

802.11 takes a lot more power. Then again, there are about a dozen different low-power choices; this is a relatively fast one.

Direct link | Posted on Jul 2, 2015 at 12:05 UTC
In reply to:

DStudio: Charming, yet painful (the video).

The video is trying to define the "last centimeter" problem, for which there have been many solutions. There are two types of functions: ID exchange and high-speed data transfer. NFC and QR codes have won for IDs and 802.11 has won for high-speed transfer, whereas Transferjet could do both functions well.

The problem is that the USB2 dongle in the video is a terrible way to do this for a phone, etc. Sort of reminds me of the idea of having a car with cool upward-pivoting doors like a Delorean, but having them swing outward so much that you couldn't open them enough to squeeze through if a car was parked next to you. ;-) Sony is the consortium leader and Toshiba is making an SD card implementation... I'm sure the dorky dongle is not the preferred implementation.

Direct link | Posted on Jul 2, 2015 at 04:40 UTC

Meyer-Optik-Goerlitz went out of business in 1991. IMHO this is an attempt to use "soap bubble bokeh" to bring back the brand. You can still get an original Trioplan for less, the overcorrected SA that causes this (ugly) bokeh is found in other lenses, and saying it's available in "mounts for all modern DSRL- [sic] & mirrorless cameras" needlessly insults users of Pentax, Samsung NX, .... ;-)

Their WWW site lists a bunch of "made in Germany" lenses that sound more appealing and look oddly familiar. For example, their 9-bladed NOCTURNUS 50mm f/0.95 sure looks like the Mitakon Zhongyi product with identical specs. They carefully define "made in Germany" to mean "assembled in Germany using the best, globally available components and are precisely adjusted with great care." Is the Trioplan special for them because it's their first lens where the design came from Germany? I'd be happier if they just said that -- I welcome new lensmakers, but I hate marketing BS.

Direct link | Posted on Jun 30, 2015 at 00:03 UTC as 30th comment | 9 replies
In reply to:

Stephen Scharf: You've got to feel for Fuji....they do the right thing for customers, and they get bashed for it.

They put out a wonderful and significant enhancement to a superb camera FOR FREE, and they get bashed. They don't put out a free enhancement, and they get bashed.

They must feel they are damned if they do, and damned if they don't.

Fortunately they have a good heart and good spirit, so even though they get bashed either way, they still do the right thing.

Kudos, Fuji....you continue to lead by example!

AbrasiveReducer: Historically speaking, Sony is a consumer electronics company that also led in electronic imaging (video) while Fuji is a film company that also made cameras. I think you can still tell where both companies came from. Although both do more innovative things than CaNikon, I think Fuji is a lot less self-confident (maybe rightly so?) about bringing-out new technologies than Sony, and the retro styling strengthens the connection to the good old days of film for Fuji. Both companies make interesting cameras.

Direct link | Posted on Jun 29, 2015 at 20:09 UTC

Much more interesting with wifi. Still a bit pricey.

Direct link | Posted on Jun 26, 2015 at 18:56 UTC as 7th comment | 1 reply
On Canon EOS Rebel T6i Review preview (302 comments in total)

Ok, image quality is quite good in practical terms (as it is for most cameras).

However, if you ignore dynamic range, Canon's latest sensors are actually pretty competitive... so why do they do such a mediocre job holding detail in JPEGs? The gap between this and a (cheaper) Sony A6000 is much larger in JPEG than raw. I know it was a pretty good JPEG processing pipeline a decade ago, but I think it's getting to be as much of a handicap to Canon as their sensors are....

Direct link | Posted on Jun 26, 2015 at 16:34 UTC as 21st comment

Like Canon, but even moreso, Hasselblad is a company somewhat trapped by their historical success; they know they need to evolve, but they know any change will require lots of effort (development cost) and might cost them loyal customers. I'm happy to hear Hasselblad realizing that RED is a real competitor in a variety of ways. Partnering with Sony still makes perfect sense as a way forward, but they have firmly proved that minor tweaks on a Sony product will be seen as "not really Hasselblad" (except for the RX100, which is interestingly where Canon also was able to leverage some Sony guts).

In sum, this all sounds like Hasselblad is finally on the right track, but it also sounds a lot like the same track that led Minolta's camera business to be handed to Sony 2007. Either way, I expect to see some high-framerate sensors in V-like bodies soon.... ;-)

Direct link | Posted on Jun 26, 2015 at 14:04 UTC as 25th comment | 4 replies
On Opinion: Did Sony just do the impossible? article (1085 comments in total)
In reply to:

bernardf12: Hope Sony gets it right next time by using an 8 core processor, or even a quad core like Samsung does. It will give them the horsepower needed to shoot (hopefully) 10 fps and 14 bit RAW. It must be way cheaper to buy mobile processors than to build your own. Going onto mobile architecture looks to me like a free ride for camera platforms.

The peak framerate is a complex function of many components, least of which is the ARM cores (other hardware does most of the critical work). However, I do firmly agree that Samsung is the obvious challenger to Sony's dominance in camera electronics, and they have much more aggressive fab technology than Sony... or so it seemed until Sony's latest round of sensors. ;-)

Right now, I suspect the game is largely about power/heat management (a problem FF IBIS makes tougher) and getting rid of the mechanical shutter.

As for frame rate, many cameras do at least 240FPS video at low res, and 36MP video @ 60FPS is coming. The ultimate framerate issue is that if you're shooting at N FPS, you can't have a shutter speed slower than 1/N s, and photon shot noise can make that a problem. In my TDCI research, I've been working on frameless CAPTURE and processing in which individual pixels have integration periods independent of the virtual shutter speed of the RENDERED frames....

Direct link | Posted on Jun 24, 2015 at 06:57 UTC

I think this concept of dual special-purposes for APS-C and FF (in this case, shift wide and macro ultrawide), is really smart. My primary ultrawide is currently a Sigma 8-16mm, which works great on my NEX-7 (APS-C) and works with fixably dark corners around 16mm on my A7II (FF) -- another dual-purpose lens. If one carries a spare body anyway, it's nice to have a FF + APS-C spare and get something special out of both....

Direct link | Posted on Jun 24, 2015 at 05:55 UTC as 29th comment
In reply to:

Mike Gerstner: From a layman's point of view:

I'm beginning to think that Canon hasn't figured out the technology yet to get lower shadow noise from their sensors, or that the only technology that is "usable(?)" is owned by sony & canon won't cough up the dough to get licensed to use it. Something's goin' on that makes no sense, to me anyway. I'm still hopin' that Canon does have a clue & is gonna show some progress in the very near future.............but the hopes are fading.

Otherwise, very nice photos. Resolution is very nice, but I'd rather have 24M images & better shadows.

Canon's using an older fab and sensor design, but the Magic Lantern folks have figured out how to use alternating lines of different ISOs to get near-Sony DR from at least some Canons. I'm sure they'll do it ASAP IF it is technically possible....

Direct link | Posted on Jun 24, 2015 at 05:37 UTC
In reply to:

RuneMC: Stacked sensor sounds very interesting if it means something like the Sigma sensor/image quality.

4K video also interesting, but 8K would have been even more interesting to extract panorama images from the video with Microsoft ICE 2.0 for example.

And the EVF - please drop it to drop the price significantly.

Will probably keep my RX100 Mk1 and wait for RX200 (whenever that may be) - although if the stacked sensor is exceptional, then I might be tempted...

Nukunukoo: very difficult to say. In general, mixing analog/digital on a chip tends to produce noise problems that reduce any such benefit, but Sony has lots of experience with this, so ...? Certainly, this type of fab allows for very different chip architectures which can have many benefits -- e.g., the TDCI (time domain continuous imaging) research I'm doing really depends on having circuitry under the photosites.

Direct link | Posted on Jun 11, 2015 at 14:39 UTC
In reply to:

RuneMC: Stacked sensor sounds very interesting if it means something like the Sigma sensor/image quality.

4K video also interesting, but 8K would have been even more interesting to extract panorama images from the video with Microsoft ICE 2.0 for example.

And the EVF - please drop it to drop the price significantly.

Will probably keep my RX100 Mk1 and wait for RX200 (whenever that may be) - although if the stacked sensor is exceptional, then I might be tempted...

Nukunukoo: yes, "stacked" here basically means logic under photosites, not Foveon-like multi-layer photosites. The active area per pixel on this 20MP sensor is probably still only about half that of the 16MP E-M5II, but the E-M5II has a lower pixel count, non-BSI conventional sensors have significant pixel vignetting (the photosite is surrounded by tall wiring that blocks off-axis rays), pixel QE (quantum efficiency) is usually higher for smaller chips because more aggressive design rules still give good yields, and Sony gets to tune the lens+sensor together. In other words, sensor performance is likely to be very close to high-end micro4/3 overall even if per-pixel area still favors things like the E-M5II. ;-)

Direct link | Posted on Jun 11, 2015 at 14:24 UTC
In reply to:

cgarrard: DELICIOUS- RX10 II.

Oh boy.

At this point I'm wondering if Canons Semi-Conductor division can even catch up. Doesn't look likely. Sony is pulling really far ahead of competition, all of it.

lacikuss: Are you suggesting that CaNikon have superior optics or are leading optically in some way? I'd strongly dispute that. I think Sigma, Zeiss, Samyang, etc. are the optics leaders.

Anyway, I think IQ is mostly about photographic technique (both at capture and in postprocessing), then sensor, with optics in a relatively minor role. Actually, it was largely like that with film photography too. ;-)

I also don't think that film & camera coming from different companies implies sensors and bodies should too. Cameras are really consumer electronics (embedded computers), and the sensors are just part of Sony's expertise in that field. As for the optics, Sony's been in video optics for a very long time, and they acquired some really innovative stuff from Minolta, but I think they've rightly decided it's not really their thing and have heavily leveraged their relationship with Zeiss, producing some fantastic optics (at a higher cost).

Direct link | Posted on Jun 10, 2015 at 23:12 UTC
Total: 725, showing: 41 – 60
« First‹ Previous12345Next ›Last »