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: 441, showing: 81 – 100
« First‹ Previous34567Next ›Last »
On HTC One M8 Camera Review post (53 comments in total)

Having a separate depth sensor means there are alignment issues, which are not helped by lower resolution. It would be interesting to see if the depth sensor could be used to stitch actual 3D models more accurately, but the simulated DoF effect looks too crude to be of much use -- at least for now. Obviously, shallow DoF "like professional cameras give" is what every tiny camera wants to fake this year....

Direct link | Posted on May 28, 2014 at 13:25 UTC as 12th comment | 2 replies
In reply to:

Elandreth: This is a bit confusing. As others have noted, there's subject motion between the individual shots of each scene. The description says 30,000 photographs were taken for this video. Even if each scene is composed of 50 photographs, that would still be 600 scenes, which is far more than appear in this video. So were the cameras shooting bursts at each scene?

As part of the post work done here, it says the photographs had to be slowed down and stabilized. Umm... what? How do you slow down a photograph, or stabilize one that's already been taken? Were these phones shooting images or video? There's an audio component to each scene, so perhaps they were actually shooting short video clips?

Is the twitching a timing error or a stitch error? (Normally, one interpolates between camera positions for this type of sequence.) Whatever the problem is, the video is an interesting effect, just not particularly well implemented. Much cheaper and easier to get high-precision sync using Canon PowerShots under CHDK, but then Microsoft probably wouldn't have helped. ;-)

Direct link | Posted on May 24, 2014 at 11:12 UTC
In reply to:

Kevin Sutton: I'm a chemist by training and this doesn't make much sense to me. Fluorine is a gas so do they mean that fluorine gas has been passed over the glass to make a fluoride-glass in situ? I'm assuming that the fluoride-glass is the hydrophobic material they are talking about??? Cheers Kevin

Yeah, "fluorine" obviously can't be more than a few atoms in the molecular formula of the coating, but the photographic industry has been touting fluoride glass for decades. I'm guessing this is really a heavy-metal fluoride glass, although those generally are known for their LACK of durability and resistance to water... so maybe Nikon has some secret sauce?

Impressive video anyway....

Direct link | Posted on May 23, 2014 at 11:01 UTC
In reply to:

Artistico: The Met states that the images are public domain - in fact they even point it out that that is what they believe.

As far as I can tell they therefore cannot legally prevent anyone else from making money from them. They have no copyright protection (Bridgeman Art Library v. Corel Corp), no database protection or sweat-of-brow protection (Feist Publications v. Rural Telephone Service). As long as the images are faithful reproductions, no new copyright exists.

What they could claim, though, is that by using their image database you are signatory to a contract stating that you cannot use them for commercial purposes. If this is in small print and not something you are forced to look at or acknowledge when entering the database, however, that would not really work either - and in any case, they cannot claim for more than breach of contract, not for breach of copyright. Since there is no money involved in the contract, I can't see how that would be legally feasible.

Many of the works I checked out were NOT PD... and they only have 394,253 total, not 400,000. ;-)

Seriously, this is a huge and wonderful resource for scholarly use and greatly broadens the impact of their collections. My concern is that people will now be able to "take the Met out of the loop" in dealing with key images, for example, all those WWW sites that sell big prints suddenly got some vastly superior things to print potentially without ever giving reference to the Met. The problem with that is that there is no "suggested donation" box attached to such uses as there is for access to view the PD things live, and the Met needs money to maintain and expand its collections. Perhaps the Met figures reaching a larger audience and the collection of email addresses will make up for that? I hope they're right.

BTW, most paintings are somewhat 3D and capture of a depth map along with a (sometimes multispectral) 2D image is becoming common in scholarly archiving... still not a new work.

Direct link | Posted on May 22, 2014 at 11:53 UTC
On Flickering fireflies in time-lapse article (34 comments in total)

Excellent work!

Really a great way to demo the various techniques, most of which are normally used for astrophotography. As impressive as the close-ups are, I don't think they fit-in as well... the transition is too abrupt. The write-up on the WWW site ties things together better.

Direct link | Posted on May 17, 2014 at 13:06 UTC as 17th comment
In reply to:

photoholiko: Just curious, does anybody really use f2.8 on a 300? to me that's a big price to pay for one f stop.

This is a 400mm f/2.8 (or 560mm f/4 with 1.4X converter).

If you want to get really far reach, imagine this on a little crop sensor... such as in a Nikon 1. ;-)

Just a pitty this lens has 2 more digits in its price tag than I'm likely to spend for a 400mm. Ok, it's actually 3 more digits than I spent for any of my current 400mm lenses....

Direct link | Posted on May 14, 2014 at 11:39 UTC
On Canon USA brings white EOS Rebel SL1 to America article (120 comments in total)

DPReview posts this as a separate item when even Canon's own press release just made it two sentences stuck at the end of the announcement of two lenses. Getting desperate for Canon to do something interesting in a body, are we?

PS: This also makes the claim that this is the "smallest and lightest digital SLR camera" -- to which footnote 1 adds the rather important qualifier "which use APS-C size equivalent sensors." Of course, Canon doesn't use standard-sized APS-C sensors, but slightly smaller ones. Arguably, smaller by about the same ratio as this body is. At least "Canon's research" specifying DSLR eliminates all the much smaller mirrorless cameras with actual APS-C sensors.... ;-)

Direct link | Posted on May 13, 2014 at 09:47 UTC as 54th comment
In reply to:

chiane: Why are all the 3rd party fast lenses manual focus?

A few more words: the electronics/mechanicals for AF and body control of aperture differ widely for different mounts, and most are not open specifications (i.e., must be reverse engineered with potentially undiscovered incompatabilities, as Sigma has found too late several times). Thus, fully manual is the best way to ensure usability on a wide range of mounts. Construction is also much simpler and potentially more precise, offering better optical performance as well as cost savings.

Direct link | Posted on May 9, 2014 at 14:19 UTC
On A travel-sized large-format 4x5 camera? article (219 comments in total)
In reply to:

steveh0607: At last, a true full frame camera! Now we don't have to suffer the poor depth of field the 35mm offers.

Yeah, they come much bigger than that. Biggest I've used was a 30x40" that had a vacuum back to hold the film really flat. Not very portable -- it weighed a couple of tons and was intended for making halftones, etc., for printing or making nameplates. Sadly, it ended-up being cut into pieces because the room it was in had been built around it, and there was no other way to get it out.

Direct link | Posted on May 9, 2014 at 11:33 UTC
On Seven photographic hacks, one short video article (34 comments in total)

Nothing new here, but a nice, quick, all-in-one-place presentation.

A few details left out.... The bokeh-shaping needs the cutout to fit within the entry pupil of the lens (50mm f/2 would be 25mm diameter), which usually means it must be surprisingly smaller than the front element -- if the cutout is too big, it will vignette instead. The flash diffuser design is similarly touchy if you want even lighting.

Also, the welding glass ND filter requires welding glass, which not everybody's got. The old standard alternative version used processed, but unexposed, color slide film. It was easier to get and usually had less of a color cast.

Direct link | Posted on May 9, 2014 at 03:46 UTC as 6th comment | 1 reply
On Samsung announces NX3000 mirrorless camera article (193 comments in total)
In reply to:

Retzius: Its the Leica T without its hand polished body ;)

No, the Leica has a 16MP sensor. However, I'm with you. Note the position of the contacts in the lens mounts....

Direct link | Posted on May 8, 2014 at 15:44 UTC
On Canon PowerShot G1 X Mark II Review preview (660 comments in total)

Kinda disappointing... but it's a Canon, so Silver Award anyway? ;-)

I actually used to be a big fan of the Canon G1 and then the G5, but this doesn't seem to have a great sensor and the touch screen is a poor excuse for not having either an optical finder or, better, an EVF. Actually, I think the touch screen is great for video (as an alternative to pulling focus), but this camera does mediocre video -- is Canon just trying to attract cell phone users who want some DoF control?

The G1X is supported by CHDK, so hopefully the Mark II will be, and that would be a plus. Then again, I can buy a very full-featured CHDK-supported "lesser" PowerShot for under $100. In sum, it's been a while since I've been excited by any of the G family members, and this is no exception.

Direct link | Posted on May 8, 2014 at 04:43 UTC as 128th comment
On National Park Service bans drones in Yosemite article (170 comments in total)

Many wild areas ban use of motor vehicles, this seems to logically be an extension of that to unmanned flying motor vehicles. Unfortunately, the way it is written would also prohibit things like flying kites with a camera attached. That's ok. One kite could still ruin the "wild" view for many visitors. This is really about the ongoing national park battle between preserving nature and allowing recreational use, and I think it is fine to favor preservation a little -- especially in a place as fragile as Yosemite.

Incidentally, I blame Disney Soarin' for the proliferation of drones in the parks. ;-)

Direct link | Posted on May 6, 2014 at 12:38 UTC as 54th comment | 1 reply
On Leica T (Typ 701) First Impressions Review preview (2300 comments in total)

I was wondering why the "lens corrections" page let Leica off so easily for falsely advertising that there were no software corrections applied to the images. However, I couldn't find any such claim in Leica's press release or at their WWW site. The claim seems to come from "During pre-launch briefings ... we [DPReview] heard that they relied on optical corrections."

The DPReview "lens corrections" page makes it sound like Leica was trying to mislead: "The only thing that we'd take issue with is the company claiming not to use this approach, when it so clearly is." Lacking any evidence Leica made any such claim in public, it sounds to me like somebody merely made a mistake in a preliminary verbal announcement or in interpreting it. Meh.

Direct link | Posted on May 2, 2014 at 21:06 UTC as 279th comment | 1 reply
On Create your own DSLR external battery pack article (66 comments in total)
In reply to:

backayonder: Plugging 9volts into a Canon? I wonder what the tolerance is?

The dummy battery connector he's using apparently has an active circuit inside and is designed for 7.4V-8.4V input, in which case 9V output from the battery is probably ok. Of course, that's based on sketchy Amazon specs which also say "4V-8.4V" without specifying for what. I don't put too much faith in those specs....

The real voltages are probably higher unless under heavy load, and there are probably regulators inside the camera that make that ok anyway. Those regulators might run hot, but having the battery outside the camera removes a major heat source.

In sum, probably ok, but I don't think I'd risk it with a $1000 camera unless I was truly desperate... even then I'd test it before trusting it.

Direct link | Posted on May 2, 2014 at 01:54 UTC
On Sony SLT-A77 II First Impressions Review preview (637 comments in total)
In reply to:

ProfHankD: All fine except the naming.

I don't get the "II" name? This and the RX100 II seem to be suggesting Sony's doing something new with the naming. The A7/A7R/A7S set is also a different naming scheme. And of course they dropped the NEX names. To add further to the confusion, long after announcing that all full-frame E-mount will be called FE, Sony's own web site uses phrases like "full-frame E-mount."

You'd think a company able to so cleanly engineer complex new camera systems would be able to come up with a better naming convention for them.... ;-)

Canon has bigger problems than dumb names. ;-)

The A77II would be a great upgrade from my A55... except I have my NEX-7 and A7 with LA-EA1/2/4 adapters for that. All I'm really missing is the sensor-movement anti-shake and the pivot on the LCD. In trade for that, I get the ability to use tons of lenses that can't be adapted to A-mount. Good trade for me.

Direct link | Posted on May 1, 2014 at 20:01 UTC
On Sony SLT-A77 II First Impressions Review preview (637 comments in total)

All fine except the naming.

I don't get the "II" name? This and the RX100 II seem to be suggesting Sony's doing something new with the naming. The A7/A7R/A7S set is also a different naming scheme. And of course they dropped the NEX names. To add further to the confusion, long after announcing that all full-frame E-mount will be called FE, Sony's own web site uses phrases like "full-frame E-mount."

You'd think a company able to so cleanly engineer complex new camera systems would be able to come up with a better naming convention for them.... ;-)

Direct link | Posted on May 1, 2014 at 15:58 UTC as 89th comment | 3 replies
On MIT algorithm predicts photo popularity article (97 comments in total)
In reply to:

ProfHankD: Reading the paper, correlation is no better than .4 based on the image, but as high as .77 based on social factors alone. In other words, image posters who are popular have more views of their images. So, it is who you know (and who knows you) more than what you do. Unfortunately, sounds about right.... ;-)

Well, I was being generous, because .4 is very unconvincing. I don't think I'd consider this PhD to be done yet if it was in ECE or CS at the University of Kentucky... whereas coming from MIT this causes a flurry of media coverage. ;-)

Direct link | Posted on Apr 30, 2014 at 18:02 UTC
On MIT algorithm predicts photo popularity article (97 comments in total)

Reading the paper, correlation is no better than .4 based on the image, but as high as .77 based on social factors alone. In other words, image posters who are popular have more views of their images. So, it is who you know (and who knows you) more than what you do. Unfortunately, sounds about right.... ;-)

Direct link | Posted on Apr 29, 2014 at 11:32 UTC as 40th comment | 3 replies
On Get more accurate color with camera calibration article (208 comments in total)

I often carry a 3" color checker card (or a neutral gray reference) and using it to calibrate the combination of scene lighting and camera can be productive. Unfortunately, smoothly-mixed lighting across a scene gives me the most trouble, and it isn't easily resolved even using a color checker. Fortunately, human viewers usually are not all that picky, and art doesn't require accurate color. ;-)

Just as a reminder, cameras capture weighted wavelength data, but color is not a straightforward function of wavelength. Land (the inventor behind Polaroid) did a lot of experiments exploring this. There is a wonderful little overview of Land's two-color experiments at http://www.wendycarlos.com/colorvis/color.html -- yes, for appropriate scenes, you can both see and photograph apparently full color images using just two color bands.

Direct link | Posted on Apr 28, 2014 at 16:38 UTC as 47th comment
Total: 441, showing: 81 – 100
« First‹ Previous34567Next ›Last »