Joseph S Wisniewski

Lives in United States Detroit, MI, United States
Works as a Speech and sensor scientist
Has a website at http://www.swissarmyfork.com
Joined on Jul 12, 2002
About me:

I've been in love with photography for over 40 years.

My favorite form of photography is macro, and much of my gear is of my own design.

I've done professional photography on and off for over 20 years.
Taught 8 years at Midwest Photography Workshops.
Designed 3 scientific digital cameras and 6 lenses.

Author of mImage and (coming soon to the App Store and Google Play) ColorForEveryone.

Comments

Total: 615, showing: 361 – 380
« First‹ Previous1718192021Next ›Last »
On article Lytro Light Field Camera now works with Windows (39 comments in total)
In reply to:

guif: Sad sarcastic comments, tomorrow all cameras will be plenoptic ... Just as all cameras are now digital, back in the times the same stupid pepple were saying "that is not a camera yada ... Yada ... " and the same crowd cries after the performances of the new canikon toy ..

Now, I use my DSLR with lenses from f1.2 to f32. Remember, you said "tomorrow all cameras will be plenoptic". That's 9.5 stops or 710:1 decimation. Double that, like Ng's thesis says we have to, and you've got about 1400:1. To get a 6mp final output, which was what I had back in 2001, you need 6mp * 1400 squared...

That's a 12 terapixel sensor!

It also is an 8 nanometer pixel, 1/500 the wavelength of visible light. The practical limit for a pixel is 700nm, if it's going to be able to capture visible red light. That's 1.7 gigapixels. Which is only 17:1 decimation for a 6mp final image, or 8:1 usable range, according to Ng.

f1.4-f11 if the microlenses are built for that pitch, which means I lose at both the deep end and the shallow end, and give up 2 stops of low light ability (the math is real noisy), and deal with 3.4 gigabyte raw files, all for a dancing bear.

Microcameras like cell phones have the opposite problem. No optical resolution to decimate.

Link | Posted on Jul 25, 2012 at 14:05 UTC
On article Lytro Light Field Camera now works with Windows (39 comments in total)
In reply to:

guif: Sad sarcastic comments, tomorrow all cameras will be plenoptic ... Just as all cameras are now digital, back in the times the same stupid pepple were saying "that is not a camera yada ... Yada ... " and the same crowd cries after the performances of the new canikon toy ..

Now, ignoring that business reality, there's also a little technology issue keeping "all" or even "most" cameras from being plenoptic. It only actually works on a small, select subset of cameras. Plenoptic cameras work by decimation, they decimate a lens's wide-open DOF aperture by a certain factor, allowing various combinations of focus ranges and DOFs within the decimated DOF to be selected.

If you have an f1.4 lens and an f1.4 to f11 (8:1) decimating microlens array, you can, theoretically, select the entire f11 DOF at the focus distance (it's not "focus free") or you can select any subset: like f1.4 DOF near the front of the f11 DOF, or f8 DOF in the middle of the f11 DOF.

In practice, you miss the theoretic limit by over a stop. Ren Ng's first camera had an f4 lens and f4 to f52 decimating (13:1) microlenses, but it couldn't deliver f52 DOF, only f22, a 5.5:1, or 5 stop range.

Link | Posted on Jul 25, 2012 at 13:46 UTC
On article Lytro Light Field Camera now works with Windows (39 comments in total)
In reply to:

guif: Sad sarcastic comments, tomorrow all cameras will be plenoptic ... Just as all cameras are now digital, back in the times the same stupid pepple were saying "that is not a camera yada ... Yada ... " and the same crowd cries after the performances of the new canikon toy ..

And you made a sad, sarcastic (and poorly punctuated) response. Digital cameras, from the crude, low resolution, $40,000 monstrosities of 1990 to the DSLRs, P&Ss, and phones of today, filled a need for immediately available, transmittable, and editable results.

Plenoptic cameras don't fill any current or anticipated need. They are what technology forecasters call a "dancing bear". For a few minutes, you look and say "wow, a bear is dancing". Then, the novelty wears off and you say "why do we need a dancing bear?"

Link | Posted on Jul 25, 2012 at 13:35 UTC

Well, at least we know that the dpReview comment counter goes up to four digits.

Link | Posted on Jul 24, 2012 at 12:50 UTC as 49th comment | 2 replies
In reply to:

herebefore: Just another new camera with no VF of any kind.

Designed for those that can see an LCD screen close to the face in the bright sun.

In other words, a camera designed for the clueless, and near clueless.

So, you ordered one in each color?

Link | Posted on Jul 23, 2012 at 13:13 UTC
In reply to:

Pentax_Prime: Anyone notice that this is the only site that 'updates' you on 'Lytro'? (A company that has never actually released a product - nor made any sort of actual business sense.) Is this company Amazon owned or what?

Oh, if only it were so.

Link | Posted on Jul 17, 2012 at 13:12 UTC
In reply to:

Oahu Kamaaina: Here's an interesting link about the Dycam Model 1 digital camera released in 1990. It had 320x 240 resolution, held 32 pictures, fixed focus lens, and cost $995. [That's 1990 dollars]

http://www.cameracuriosities.com/2012_02_01_archive.html

I'm sure there were lots of comments by film photographers putting down the new technology then, similar to the comments in this blog. The Lytro camera may well fail, it's largely a proof of concept, but I'm surprised that the photographers who don't buy one aren't at least cheering their innovation rather than dissing their attempt to improve the technology.

17 megs is a small amount of data. Most P&S with raw capability produce more.

And again, the Lytro 3D demonstrations are highly contrived, as you'd see, if you cared to search. Single lens 3D techniques do not solve the occlusion problem, as you'd also see, if you cared to search.

No one has demonstrated "all in focus images", and yes, I've searched more than you have, or ever will, because facts would spoil your enjoyment of your cult.

"I joined this discussion thinking I could learn and share information with other Lytro owners"

Now, you know that's not true. All your posts have been defensive, and, as I correctly assessed, deceptive.

Link | Posted on Jul 17, 2012 at 13:11 UTC

Pretty cool, considering that it's 30g lighter and 1mm shorter than the f3.5 that it's replacing.

But this was already announced back in February, with an expected shipping date in June.

http://www.cameraquest.com/Voigt_SL2.htm

So, is this another announcement of the lens, or an admission that they blew June?

Then again, the 75mm f1.8 SL II was announced way back in February of 2011 (seriously) and is still not shipping...

Don't get me wrong: I love my Voigtlanders, the 40mm on FF, the 58mm on APS are two of my favorites, but they don't have the best track record for shipping what they announce.

Link | Posted on Jul 16, 2012 at 13:15 UTC as 26th comment | 3 replies
In reply to:

chris_j_l: And with this, Sigma signals that they have abandoned K mount for macro lenses.

Or that they have abandoned K mount for all new lenses. The big question is, are they still producing the existing K mount lenses, or when the supplies in the warehouse are gone, is Sigma out of the K mount business?

Tamron appears to have pulled out. Cosina (Zeiss and Voigtlander) has pulled out. That would leave Pentax the only camera brand with no third-party lens support.

Link | Posted on Jul 14, 2012 at 20:23 UTC
In reply to:

NancyP: Joseph Wisniewski: "Breathing" - that is a new term to me. Could you explain what you mean? Optical diagrams I get, but you can't readily show these in comments.

This can be a small amount, an overcompensated lens may breathe 1/2 as much as a unit focusing lens, and you only really notice the breathing because it's "wrong way" breathing, out as you focus in, in as you focus out. Or it can even be more than a unit focusing lens. That's why the Nikon 70-200mm VR2 gets a lot of comments from people shooting video, because it has a lot of that "wrong way" breathing that's so easy to notice. More than a "conventional" lens, and wrong way, to boot.

Link | Posted on Jul 14, 2012 at 20:16 UTC
In reply to:

NancyP: Joseph Wisniewski: "Breathing" - that is a new term to me. Could you explain what you mean? Optical diagrams I get, but you can't readily show these in comments.

To sum it all up. A unit focus lens is one "extreme". It breathes in, the angle of focus decreases, looking like it's "zooming in" as you "focus in" on closer objects. It's "not compensated" for breathing.

A "perfectly compensated" lens changes focal length exactly as much as would be required to focus, there's no movement of the lens's "rear nodal point" and you get no breathing, at all. Zeiss "master primes", at $15,000 each, are examples of perfectly compensated lenses.

An under-compensated lens changes focal length, but not enough to achieve focus, so it has to move a little, and the end result is it still breathes in as you focus in, but not as much as a unit focusing lens. The Nikon 85mm f1.4 is such a lens, Nikon eliminated about 70% of its breathing.

An overcompensated lens changes focal length more than it should to achieve focus, so there has to be "wrong way" motion to take up the slack. The breathing is more "noticeable" because they breathe out when you focus in.

Link | Posted on Jul 14, 2012 at 20:13 UTC
In reply to:

NancyP: Joseph Wisniewski: "Breathing" - that is a new term to me. Could you explain what you mean? Optical diagrams I get, but you can't readily show these in comments.

Note that I said "if it were perfect". In reality, it both zooms and moves as it focuses. So, it might use zooming to account for 5 of the 7.7mm that it needs to focus from infinity to 1m, and use movements of some elements near tie aperture to pick up the other 2.7mm. The result, the angle changes, but only 1/3 as much as it would if you used an old-fashioned unit-focus design.

So, the 85mm f1.4 Nikkor is "better" for video than the 85mm f1.4 Cosina Zeiss Planar. The Nikkor is not free of breathing, but it breathes about 1/3 as much as the Zeiss.

Some lenses overzoom. The Nikon 70-200mm VR II is an example. It actually changes focal length from 85mm to about 59mm when it goes from infinity to 1.4m. That's 25.5mm outward, when it only needs 5.5mm. So it "overzooms" about 20mm, and ends up breathing worse than a unit focusing lens that would only have moved 5.5mm.

The older version 1 zoomed much less, and has much less breathing.

Link | Posted on Jul 14, 2012 at 20:06 UTC
In reply to:

NancyP: Joseph Wisniewski: "Breathing" - that is a new term to me. Could you explain what you mean? Optical diagrams I get, but you can't readily show these in comments.

A lens that changes focal lengths as you focus can partially or totally compensate for breathing. The older Nikon 85mm f1.8 D, for example, has an "internal focusing" mechanism that moves the lens's rear inner elements to alter the focal length of the lens when the camera tries to focus it.

Instead of racking out from 85mm to 92.9mm when you focus from infinity to 1m, it changes focal length. At infinity, it's an 85mm lens, 85mm from the sensor. At 1m, if it were perfect, it would zoom, becoming a 77.3mm lens at 85mm from the sensor. Since the "hole in space" is always 85mm from the sensor, the lens is always covering 28.6 degrees.

Link | Posted on Jul 14, 2012 at 19:58 UTC
In reply to:

NancyP: Joseph Wisniewski: "Breathing" - that is a new term to me. Could you explain what you mean? Optical diagrams I get, but you can't readily show these in comments.

Brendon's on the right track.

Breathing is an old movie term. Still photographers don't pay as much attention to it, but it's a real problem when you focus in to blur a scene for a transition, or cross focus between two people at different distances, to add emphasis to a dialog.

It means the angle of coverage of a lens changes as you focus. Even a "unit focusing" lens that doesn't change focal lengths when you focus breathes.

Let's pick on my 58mm f1.4 Voigtlander. It doesn't change focal length when you focus: it's always 58mm. So, at infinity, the lens is a "hole" 58mm from the sensor. On a FF camera, that's 2*arctan(43.3mm/2/58mm) = 40.9 degrees.

When you focus to 1m, the lens extends 3.6mm. That hole is now 61.6mm from the sensor. So, the angle of coverage is now 38.7 degrees, a very visible difference from the 40.9 you had at infinity. Stuff gets larger looking.

Link | Posted on Jul 14, 2012 at 19:46 UTC
In reply to:

Frenske: Depth of field: 0.0004m at 0.47m and f/2.8. LOL ... even at f/16 it is just 3mm. I am not sure the world really needs lenses like this.

You're both welcome.

Link | Posted on Jul 13, 2012 at 16:10 UTC
In reply to:

The Scurvy Dog of PR: I noticed they did not state in the specs, how much it weighs. At F2.8 and 19 elements, it must be substantial.

1.64kg, 3 lb 9.8 oz

http://sigma-photo.co.jp/english/lens/macro/180_28_os/#/specification

Link | Posted on Jul 13, 2012 at 16:08 UTC
In reply to:

Frenske: Depth of field: 0.0004m at 0.47m and f/2.8. LOL ... even at f/16 it is just 3mm. I am not sure the world really needs lenses like this.

Actually, teleconverters are common with long macro lenses, to get more working distance around "skittish" subjects. Shaw's "Closeups in Nature" has a lot of good information on that technique. That's one reason for an f2.8 macro.

A lot of us do extreme shallow DOF macro. I often use a Nikon 85mm f1.4 (76mm f1.4 at its closest focus) with a Canon 500D diopter. That's effectively a 66mm f1.2. That's another reason why people need lenses like this.

Zeiss makes a pair of f2.0 macros, a 100mm and a 50mm.

Now, if you know so little about the subject, you really shouldn't be getting so rude with people who do know about it, or ridiculing equipment manufactures. Instead, ask intelligent questions, like why would they make this? What would you do with it?

Link | Posted on Jul 13, 2012 at 13:41 UTC
In reply to:

GreginVA: I am 2 for 2 with dud Sigma lenses, including a 150mm OS Macro which will periodically decide to have problems with autofocus and with image stabilization. I won't buy any more Ford cars and I won't buy any more Sigma lenses, particularly when there are reasonable alternatives from Canon and Nikon. The specs for this lens look good, but Sigma does not have adequate quality control.

I'm 6 for 5 (one lens went in twice, for two unrelated problems).

The HSM died in my 30mm f1.4.

The 14mm had a problem with its own lubricants coating the glass elements, and, as it was out of warranty, Sigma promptly pronounced it irreparable, but a local shop fixed it.

My 400mm 5.6 went in twice: once for the AF mechanism, once for an internal element that came unmounted.

My 8mm has a focus problem that Sigma says (in tiny print, in the manual, no less) is normal for that lens.

My SD-14 died the second or third time I turned it on, and I returned it for "fitness for purpose".

Link | Posted on Jul 13, 2012 at 13:30 UTC
In reply to:

semorg: I'll wait an see if they are truly f/2.8 at the 1:1 ratio position or if the the 2.8 is at longer range. Also I would be curious about the diaphragm design and the kind of bokeh this lens can produce. For me the best 180 lens and probably the best lens created remains to be leica 180/2.8 apo.

This one, like my Nikon 200mm, is an internal focusing type: the focal length drops as you get closer to 1:1. WillemB says 117mm, which makes sense. 180mm f2.8 is a 64.3mm aperture, so 117mm at 1:1 is f3.64, which is a lot better than the 150mm does, a "conventional" macro that goes from f2.8 to 5.6 at 1:1.

@sirkhann, aperture is reduced (see "bellows factor") but light transmission (see "t stop") is unaffected.

Link | Posted on Jul 13, 2012 at 13:24 UTC
In reply to:

IcyVeins: Just wondering but why would you need f/2.8 for doing macro? Don't you usually want as much DOf as possible since you are focusing so close?

Willem, it's close to the length you need to avoid breathing. If your lens doesn't change focal length at all, you get "breathing in" when you focus in. Say you've got a FF camera. At infinity, the angle of view is 2*arctan(43.3mm/2/180mm) is 13.7 degrees. Now, simply extend to 1:1, and you've got a pupil 360mm from the sensor, and 6.9 degrees.

If your lens zoomed to 90mm when it hit 1:1, you'd have the original 13.7 degrees, "breathing free", but you'd have a short working distance.

I guess 117mm is a "happy medium", you breath from 13.7 to 10.6 degrees, but still retain about 200mm of "working distance".

Oh, and of course, if it held its 180mm focal length, it would need to extend an additional 180mm to reach 1:1. It's already 204mm long: a 384mm (15 inch) long macro would be very hard to handle.

Link | Posted on Jul 13, 2012 at 13:10 UTC
Total: 615, showing: 361 – 380
« First‹ Previous1718192021Next ›Last »