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: 338, showing: 81 – 100
« First‹ Previous34567Next ›Last »
On Enthusiast interchangeable lens camera 2013 roundup article (267 comments in total)

A little funny that the NEX-7 is Sony's contender here after the A7 and A7R have been released and A7 pricing is similar to the other contenders -- it looks like you stuck to APS-C sensors ignoring the A7/A7R FF, but then you included one and only one micro4/3...? Otherwise, this seems a reasonable list of contenders.

Of course, the about-to-be-replaced NEX-7 is the cheapest of this set and is really tied for best IQ with the other cameras using essentially the same 24MP APS-C sensor. Saving hundreds of dollars is arguably pretty good compensation for lack of wifi.... ;-)

Direct link | Posted on Nov 23, 2013 at 01:03 UTC as 84th comment | 1 reply
On Sony Alpha A7 / A7R preview (2381 comments in total)

Wow! It's beating the D800 fairly easily -- didn't expect that. It also blows away the Canon 5D III (there's more detail in the A7R at 6400 than in the 5D III at 50), but that was not entirely unexpected.

It looks like Sony is retaining detail really well in JPEGs while nicely removing somewhat more sensor color noise than the competition that shows up in the raw. This might be the best JPEG engine I've seen in a camera so far.... I will admit to also wondering if the color noise isn't noise at all, but artifacting due to the lack of an AA filter? If so, congrats to the lens used, because that would mean it's well past Nyquist for a 36MP FF sensor....

PS: Look at the Jack in the comparison scene. Interesting that only the A7R doesn't make the Jack's hair go the wrong direction with artifacting....

Direct link | Posted on Nov 15, 2013 at 06:35 UTC as 222nd comment | 7 replies
On The Q Camera wants you to 'shut up and shoot' post (27 comments in total)
In reply to:

rfsIII: LOVE IT! Who does not want an inexpensive, fun camera that will make people smile when you hold it up to take their pictures.
(One thing... I'm pretty sure that there are no more hipsters. That was a mid-century Brooklyn thing when Brooklyn was still a cheap and scary place. Brooklyn is now full of ex-husbands and middle management DBs. The cool people have moved to Queens.)

I agree with serenity now. The ONLY market for this would be kids too young to be trusted with a cell phone yet cheaply added onto the family usage plan. My 12-year-old wouldn't go for this, but at age 5 or so....

Direct link | Posted on Nov 10, 2013 at 14:18 UTC
In reply to:

Archiver: Interesting how this is available only for the C100 at this time. If this can be implemented with the C300, it will show that dual pixel AF is a function of the CMOS sensor shared by the C100, C300 and C500, rather than the C100's hardware.

Really? That would give terrible artifacting of out-of-focus regions in normal use. Also, the usual pattern is RG/GB. I know they have an unusual readout structure in the C100 sensor so they could reuse chips from earlier camcorders, so maybe...? I guess they could approximate things having the RG/GB pattern and still using the two G (the vertical offset shouldn't be too significant for PDAF). I really look forward to either Canon telling us what's up or the Magic Lantern people figuring-out the details....

Direct link | Posted on Nov 9, 2013 at 00:39 UTC
In reply to:

Leandros S: The idea that lenses need firmware still bothers me. Why would anybody engineer it that way? The lens should be a mindless slave to the camera imo.

No practical choice. Modern lenses commonly contain focus and aperture motors, the control logic to move them as commanded by the camera, and encoder feedback on things like where the aperture blades and focus position are. It is far more reasonable to put the lens-specific control logic in the lens than to require higher bandwidth communication and hope the body can do it all. Basically, this is the same reason there's a processor in your computer's mouse and another in the keyboard.

Direct link | Posted on Nov 7, 2013 at 12:06 UTC
In reply to:

Archiver: Interesting how this is available only for the C100 at this time. If this can be implemented with the C300, it will show that dual pixel AF is a function of the CMOS sensor shared by the C100, C300 and C500, rather than the C100's hardware.

Yes. Quote: "utilises the existing Super35mm sensor in the EOS C100." Really? Do they swap the microlenses? This is very strange to say the least....

Direct link | Posted on Nov 7, 2013 at 02:38 UTC
On Nikon Df preview (2817 comments in total)

Well, this may be overdoing the retro look a little, but so what? It seems well built and generally interesting, especially if you've got a pile of old Nikon lenses that wouldn't work on any Nikon FF digital camera.

The thing I don't get is the no-focus-aids fixed viewfinder screen. I'd take focus peaking over a good screen with a split and/or microprisim (I've made this comparison many times), but not even having a focus aid as an option in a camera bragging about support of manual lenses is idiotic. I can only imagine that it was a deliberate choice to give a brighter view and uncluttered view, which is why autofocus cameras switched to screens that are terrible for focus and don't even allow judging DoF. I predict a very large market for 3rd-party DF screens, assuming the screen isn't too permanently set in place....

Anyway, it's good to see Nikon doing something, and this does give Nikon manual lens users a distinctively different alternative to the Sony fleet. Canon?

Direct link | Posted on Nov 6, 2013 at 03:20 UTC as 407th comment
In reply to:

tjwaggoner: Why wasnt he asked if he had any info on when BSI sensors would migrate to the larger formats? This is a question that intrigues me, as I think it may be the next big leap in sensor performance over the very marginal increases you get from most modern sensors today. Just my opinion, maybe some of you are more informed on the issue?

Rear illumination involves bonding a new substrate on top and grinding away the original substrate -- neither a cheap nor high-yield process. It is easier to justify for tiny sensels on tiny dies,

Direct link | Posted on Nov 5, 2013 at 02:22 UTC
On Canon EOS 70D Review preview (563 comments in total)
In reply to:

justmeMN: I'm outraged! How DARE DPR say anything nice about a Canon camera! :-)

There is plenty nice to say about the 70D, but it doesn't really earn the scoring DPReview gave it. For example, according to DPReview, this beats the Sony NEX-7 83:81 primarily due to higher scores on "value" and "ergonomics" -- despite the fact that the 2-years-older NEX-7 is cheaper and still matches or outperforms it on just about every directly measurable property.

DPReview has a well-established history of giving Canon a few extra points just for being Canon... but many people give extra points to the sales leader in any market just for being the market leader. I pay more attention to the actual data/measurements DPReview & DxOMark give rather than the glowing prose and highly subjective final scoring summary....

Direct link | Posted on Oct 31, 2013 at 06:11 UTC
On Canon updates firmware for EOS-1D X and EOS-1D C news story (35 comments in total)
In reply to:

Howard S: "the ability to set exposure compensation when using Auto ISO in manual exposure mode." Can we have this on other models please?

You can have such things and much more on PowerShots using CHDK and Canon DSLRs using Magic Lantern. Free. Available now.

Direct link | Posted on Oct 22, 2013 at 22:43 UTC
On Canon updates firmware for EOS-1D X and EOS-1D C news story (35 comments in total)
In reply to:

HetFotoAtelier: available to download in January 2014?????
Now I know nothing to ask for christmast :-(

When companies like Sony announce awe-inspiring FF cameras, Canon has to announce something, right? ;-)

Direct link | Posted on Oct 22, 2013 at 22:41 UTC
On Sony Alpha A7 / A7R preview (2381 comments in total)
In reply to:

ProfHankD: One more comment: Sony named these as "7" models. The upcoming "9" must be one heck of a game changer!

Actually, I'm guessing 54+ MP among other things. It takes 54-56MP to give an APS-C crop with 24MP, thus matching the NEX-7 resolution. For lenses that only cover APS-C, or FF lenses that have lousy corners (which many do), the NEX-7 will still deliver higher IQ than the A7R or any other Sony product announced thus far.

Direct link | Posted on Oct 16, 2013 at 13:08 UTC
On Sony Alpha A7 / A7R preview (2381 comments in total)

One more comment: Sony named these as "7" models. The upcoming "9" must be one heck of a game changer!

Direct link | Posted on Oct 16, 2013 at 11:30 UTC as 631st comment | 5 replies
On Sony Alpha A7 / A7R preview (2381 comments in total)

Definitely an RX1 + some NEX features grafted on... which is fine, even if you can still see where parts are grafted on. ;-)

In the NEX forum (I guess we need A-mount/E-mount/FE-mount fora?), I took a poll as to what type of lenses NEX folks primarily use. Only half primarily use native E-mount lenses. 35% said they primarily use adapted SLR lenses, 10% said rangefinder lenses, and 3% adapted AF lenses (EF/EF-S or A mount). My guess is that means 45% of the NEX community here want the A7R and will not care one bit that the FE-mount lenses appear overpriced. Hopefully, those lenses aren't really overpriced, but are built to be compellingly good as compared to every SLR lens ever built -- which is truly their direct competition on these cameras. The weatherproofing is one step in that direction.

Direct link | Posted on Oct 16, 2013 at 10:59 UTC as 640th comment
In reply to:

ProfHankD: I can see why Nikon's annoyed because it sure looks like the J2/J3/S1 in body shape and placement of some features. The guts are functionally different, but that is not relevant for the "infringement of the design patent and trade dress right owned by Nikon." However, it is very clearly marked as Polaroid iM1836.

I assume Nikon is worried that it will lower their brand image. However, the only real harm I can see is potential confusion about compatibility. That's still enough reason to sue, I guess....

Anybody can sue... but the litigation easily could cost more than it's worth if they can't collect damages. I just don't see this being worth Nikon's effort. I guess Nikon sees Sakar differently? ;-)

Direct link | Posted on Oct 15, 2013 at 20:47 UTC

I can see why Nikon's annoyed because it sure looks like the J2/J3/S1 in body shape and placement of some features. The guts are functionally different, but that is not relevant for the "infringement of the design patent and trade dress right owned by Nikon." However, it is very clearly marked as Polaroid iM1836.

I assume Nikon is worried that it will lower their brand image. However, the only real harm I can see is potential confusion about compatibility. That's still enough reason to sue, I guess....

Direct link | Posted on Oct 15, 2013 at 18:31 UTC as 16th comment | 4 replies
In reply to:

DotCom Editor: I hope none of those screen were running Windows Vista

On the other hand, what's that Dell laptop doing there?

Product placement, I guess. ;-) Actually, I'd guess it's doing something like keeping the displays lit with something credible. There is probably more compute power in the laptop than in all the shuttle control computers combined....

Direct link | Posted on Oct 14, 2013 at 13:26 UTC
On Preview:canon-eos-70d (1333 comments in total)
In reply to:

ProfHankD: I really don't much care about PDAF. In fact, I usually prefer manual focus with peaking. However, I'm impressed to see that Canon's dual pixel structure doesn't seem to have sacrificed IQ for PDAF. According to both DxOMark and the DPReview sample images, this sensor is a significant improvement over the sensor in, for example, the 100D or 700D. It still has a half stop less dynamic range, but it's the first 1.6X crop sensor Canon has made that doesn't get trashed by the 1.5X crop one in the Sony NEX-5.

DPReview says the 70D competes with Canon's FF -- I cannot agree. The Sony NEX-7 outperforms the 6D at low ISO and looses to it at high ISO, but 6D IQ is always better than the 70D in color depth, dynamic range, and low light. Still, the 70D proves that Canon's dual pixels are a viable approach, competitive in APS-C IQ while incorporating a useful new capability. Now if only Canon would give us direct access to the 40M half-pixels.... ;-)

"Could a 20MP stereo pair be used to create a single 20MP 3D photo?" -- technically, we call such things 2.5D, but yes.

"Dual pixel could widen DR." -- not really. The interleaved ISOs trick in ML which works so well with some other Canons, will work less well if the 2 readout channels are the left/right sides of pixels because they are two different views. Theoretically, this problem might be correctable by using depth estimates in interpolation....

Incidentally, the ";-)" on my original comment was because the first thing I did was to contact Canon about getting access to the 40MP, and they said "no." The second thing I did was to contact one of the ML developers. :-)

Posted on Oct 12, 2013 at 00:47 UTC
On Preview:canon-eos-70d (1333 comments in total)
In reply to:

ProfHankD: I really don't much care about PDAF. In fact, I usually prefer manual focus with peaking. However, I'm impressed to see that Canon's dual pixel structure doesn't seem to have sacrificed IQ for PDAF. According to both DxOMark and the DPReview sample images, this sensor is a significant improvement over the sensor in, for example, the 100D or 700D. It still has a half stop less dynamic range, but it's the first 1.6X crop sensor Canon has made that doesn't get trashed by the 1.5X crop one in the Sony NEX-5.

DPReview says the 70D competes with Canon's FF -- I cannot agree. The Sony NEX-7 outperforms the 6D at low ISO and looses to it at high ISO, but 6D IQ is always better than the 70D in color depth, dynamic range, and low light. Still, the 70D proves that Canon's dual pixels are a viable approach, competitive in APS-C IQ while incorporating a useful new capability. Now if only Canon would give us direct access to the 40M half-pixels.... ;-)

See http://www.usa.canon.com/cusa/consumer/products/cameras/standard_display/daf_technology

Canon basically has 40MP with sensels paired under 20M microlenses so half-pixels separate left/right views through the lens. This is currently used only for PDAF, with values combined for the 20MP raws. However, with a little tweaking (but probably no change to the sensor), the 40MP could be used to capture 20MP stereo pairs. I've been looking at this in my research.

Posted on Oct 11, 2013 at 12:35 UTC
On Preview:canon-eos-70d (1333 comments in total)

I really don't much care about PDAF. In fact, I usually prefer manual focus with peaking. However, I'm impressed to see that Canon's dual pixel structure doesn't seem to have sacrificed IQ for PDAF. According to both DxOMark and the DPReview sample images, this sensor is a significant improvement over the sensor in, for example, the 100D or 700D. It still has a half stop less dynamic range, but it's the first 1.6X crop sensor Canon has made that doesn't get trashed by the 1.5X crop one in the Sony NEX-5.

DPReview says the 70D competes with Canon's FF -- I cannot agree. The Sony NEX-7 outperforms the 6D at low ISO and looses to it at high ISO, but 6D IQ is always better than the 70D in color depth, dynamic range, and low light. Still, the 70D proves that Canon's dual pixels are a viable approach, competitive in APS-C IQ while incorporating a useful new capability. Now if only Canon would give us direct access to the 40M half-pixels.... ;-)

Posted on Oct 11, 2013 at 10:54 UTC as 107th comment | 7 replies
Total: 338, showing: 81 – 100
« First‹ Previous34567Next ›Last »