kevinconnery

June 5, 2006

<ul>Kevin Connery:

Can you elaborate on how an interchangeable sensor would differ from a "digital back"? (KC)

"Backs" are usually solid and moreover thicker. Sensors are thin, like memory cards.

The Hubble Telescope (I`ll call it that or HT) is, with all its complexity, a ``plug and play`` device. When something goes wrong, they fly the new part up, take out the old, insert the new and restore power, like a bus sized Windows PC/camera. (EG)</ul>

You're not familiar with chip design or D/A signal subsystems, are you?

Replacing a specific sensor with an identical one wouldn't require the imaging subsystem to be changed; replacing a sensor with a different one would--or would require a much more elaborate imaging engine, with added complexity and cost.

<ul>Your operative phrase (above) then is ``existing sensors`` when Wundercamera is a device of the future. (EG)</ul>

I await new imaging models to be developed. <a href="http://www.tanner.com/Labs/research/technologies/mcm/default.htm">MCM</a> could make this possible, but MCM tech has been around for nearly 15 years, and is still not in widespread use. Nevertheless, doing so would require more than replacing the "sensor"; you'd still be replacing the imaging system, no matter what you called it.

Similarly, while chip yields improve, larger chips will still cost more than smaller ones. Not just because of direct real estate (e.g. 10 chips per die vs 100), but failure rates in fab. (if a die has 1 defect and 10 sensors fit on that die, you lose 1 chip--netting 9 sensors. If a die has 1 defect and one sensor fits on that die, your net is zero. The larger the defect rate, the more impact this has.)

I'm curious as to when--if ever--a 645 sized sensor would becomes the norm, even in your so-called wundercam for professional use.

<ul>

with an intrinsicly higher implementation cost will replace a lower-complexity design (interchangeable lenses) is a long shot. (KC)

Not for you or me to say. Again, the ``Wundercamera`` will be modular, like the HT, plug and play in that those engineers will have TOTB. Most automobiles today are themselves modular, even the high end models, though there is a large degree of hand finishing for the Lexus types. (EG)</ul>

If you can make guesses without experience on the engineering side, I can make estimates with experience. In that light, it's not for you to tell me it's "Not for you or me to say" when you've already had your say.

<ul>Talking DOF is a shibboleth often used by DSLR folks to denigrate P&S cameras. (EG)</ul>

How so? A 35mm P&S camera had the same DoF as a 35mm SLR did, with the good and bad accruing.

Choice of DoF is a legitimate issue. Just because you claim you "don't have one "Artistic" bone in [your] body" doesn't mean that technical/craft issues of that nature aren't important to others. (You did, after all, claim superiority for the large depth of field--dismissing a rebuttal as being a shibboleth is disingenuous. (But I do see that this is your MO.)

<ul>I don't think very many people fooling themselves into thinking they need a DSLR, when a P&S digicam would do. (AR)

Your point not mine. But I agree. (EG)</ul>

Perhaps. You, yourself, though, insisted "Nothing less than a new or gently used "1" type body will do for weddings and social events", so it becomes even less clear whether your 'floozi wundercam' is for professionals or joe sixpack: you've dismissed both the casual shooter's needs, AND the professional shooters needs.

<ul>And I think you don't know anything about modern front mounted teleconverters (EG)</ul>

I believe that is a correct statement. Your understanding of optics may be greater than your knowledge of digital imaging engineering, but it's not apparent here.

<ul>At close up distances, DOF is severely limited by the shallow plane of focus. (AR)

Not with P&Ss (EG)</ul>

Make up your mind. Is your wundercam a small sensor or large? If you're claiming a 645 sensor, you lose this argument--but you do gain access to the shibboleth of DSLR users. Consistency?

<ul>Low-end DSLRs are what they are, but they are not 'severely crippled'. Any low end 6 MP DSLR can comfortably blow most any 8-10 MP digicam for image quality.(AR)

totally unsubstantiated. The 8MP FZ30 make perfect, perfect 11 x 14 photos. Which is all we need to prove. (EG)</ul>

My "small" prints are 8x10. My typical prints are 16x20. My large prints are 30x40. Andrew's statement is "totally substantiated" in that environment. In a journalistic environment, with 60-85 line screen printing, or for a casual user who will rarely make a print larger than 8x10, it may well be true. But that is not the entire universe of photography, as much as you seem to want it to be--"all we need to prove" is a false assertion.

<ul>

1.5 KG in weight? Why the heck would I choose it over a DSLR? (AR)

Because it would kick DSLR butt. (EG)</ul>

Circular reasoning. You're saying your wundercam will displace DSLRs because "it would kick DSLR butt", and it kicks DSLR butt because it ... kicks DLSR butt. If you're going to be that sloppy, at least add one or two steps to hide the circularity.

<ul>No one reading this could have predicted a camera with a 36-432mm image stabilized f/2.8 throughout the zoom range lens until Panasonic surprised every damn body with the FZ20. (EG)</ul>

Unsupported. That lens is an extension of known art, given the physical requirements--imaging circle being one of them. Nevertheless, it's not a 36 to 432mm lens; it's a 6mm-72mm lens.

<ul>Working ``pros` (I still carry my association card and PJ ID, though I'm retired) will absolutely love this camera, at least the PJs will. PJs will be able to dump 30 pounds out of their bags, with no serious penalty accruing to their shooting ability. (EG)</ul>

For that purpose, it sounds like a great camera. PJ, however, is not the limits of photography, no matter how often you prattle about your shots being "mostly unplanned", or that "anything that veers too far from "Real life" seems, as I noted before: "pretentious."

You may find your current panasonic wundercam "wunderful". That doesn't make it a universal truth. Nor are your "predictions" automatically correct just because you find the results to your taste. Understanding some of the technology inside the current boxes would help you avoid thinking so far out of the box that reality--current or future--is no longer within sight.

June 4, 2006

<ul>"When(ever) discourteous behavior/language/inferences surface, rancor ensues." (EG)</ul>

Yes, it does. Claiming something was discourteous when there's little evidence of it decreases communication: "I was having fun until your bile surfaced to spoil it."

<ul>"And ``Wundercamera`` would not have or utilize a ``digital back`` per se, but interchangeable sensors. Then again, I don't know every camera out there." (EG)</ul>

Can you elaborate on how an interchangeable sensor would differ from a "digital back"? No existing sensors carry the support electronics, and each sensor requires different logic, if only to handle different cell geometry or size--much less differences in filtration, if a Bayer mosaic (or similar, such as Sony's RGBE) is used.

<ul>

"Most digital capture devices have fixed lenses, like cellphones and the zillion P&S digitals." (KF)

"Better not let the DSLR people hear you spout such blasphemies!" (EG)</ul>

How so? Kelly identified a fact. You're mocking a position that hasn't been shown.

<ul>" but they are still in reality very short lenses with proportionately large depth of field." (AR)

"DOF which can be put to good use. Limited depth of field is also a severe handicap in macro imaging for instance, wherein I have to use either expensive ``macro`` lenses on my SLR or attach expensive diopter lenses. With my FZ20 for example, great macro with great depth of field comes with the lens, with no additional costs attending."</ul>

That's great if (and only if) a strong depth of field is desired. While that's beneficial in some circumstances, it's equally a disadvantage in others. With tiny sensors, selective DoF is not practical. Cherry picking`` photographic features from the world of photography to support one view is silly, and further,it makes for a disingenuous argument.

<ul>"What about tilt / shift," (AR)

"You win. But then, how many ``Joe Sixpack`` SLR owners also own tilt-shift lenses? How many have even seen a Tilt/shift lens?" (EG)</ul>

Again, you're cherrypicking. You called it the death of the DSLR, which is--today--largely in the hands of advanced amateurs and professionals--not "joe sixpack". (Film SLRs had a much wider penetration model than DLSRs do.)

Do I believe your Panasonic "Wundercamera" won't occur? It doesn't particularly matter, unless it comes at the expense of professional level tools. Note: that doesn't automatically mean DSLRs; I've stated numerous times that SLRs as built today use a kludgy hack for a viewing system--one that will most likely disappear in the years ahead.

However, predicting that an increased-complexity design (interchangeable sensors) with an intrinsicly higher implementation cost will replace a lower-complexity design (interchangeable lenses) is a long shot.

June 4, 2006

An epiphany is the comprehension or sudden understanding of reality; (Me)

You have no hope in hell of telling what ``reality`` is to me, or will be or is to others. (EG)

Apparently so.

Imaging circle issues intrude on the prediction; the imaging circle of lenses are limited by the physical materials available for lenses. A larger imaging area requires a larger image circle--how "fast" are the lenses designed for LF cameras? How large are they physically?

If you meant tiny sensor units (the 6 X 5 sized sensor reference isn't clear as to whether that's very large [inches], very small [mm], or somewhere in the middle), the trade-off would be focus selectivity: "functional" depth of field goes up as media size goes down.

It IS possible that manufacturers will go for a more modular approach--as the MF cameras did for decades (and still do...), and have interchangeable sensor/processors. Cost increase for that modularity hasn't been considered worthwhile by most vendors--eFilm/Silicon Film folded, and Leica's R9 digital back isn't price-competitive for the functionality--but that may change.

Sensors and sensor subsystems will almost certainly improve. No epiphany there, I'd hope; that's the same step-by-step process digital cameras have made since they first appeared.

And my ``forecast`` is cast; in stone: where is yours?(EG)

My predictions (stored on electrons):

Cameras may become more decoupled from their viewfinder systems, perhaps with the camera physically (or wirelessly) tethered to a high-resolution detachable viewfinder (heads-up display). Or with an imaging module (lens+sensor), processing/storage modules, and viewing modules. That's practical from an engineering standpoint today, and offers significant improvements for some photographers--probably not enough to justify the expense, though, today.

Higher-sensitivity sensors will appear, vastly reducing the need for added light. (Control will still be desirable, but a "good exposure" will be less limited by the quantity of light)

Mirrors will probably fade away; they've always been a workaround to the problem of viewing through the taking lens, and as sensor quality increases, devoting some of that real-estate to real-time viewing becomes practical. The box shape, as you guessed, will change as a result.

Improved signal processing will permit increasing or decreasing depth of field and varying focus after the capture. This is currently available in some COTS software, but it's expensive and compute-intensive. If it does become available, the DoF limitations of tiny sensors are removed.

The above combined with the faster sensors may permit shooting at the same aperture for all images, and adjust DoF to taste afterwards.

No epiphanies, alas. Merely guesses and extrapolation of current trends, along with some knowledge of chip manufacturing and design, and the rough rate of change in the field.

Kevin didn't seem the least bit angry. (MD)

Thanks, Mike. I certainly hadn't meant it to sound angry.

June 3, 2006

It's my damn epiphany and I'm sticking to it-Bwahaahaahaa!

No, Ed. It's your guess, your prediction, your desire or hope. An epiphany is the comprehension or sudden understanding of reality; it's not "the ability to forecast a future", however accurate or senseless that forecast may be.

June 3, 2006

There's a few thousand ways, and the ones that work for some images won't work as well for others. When I teach my Converting to B/W class in Photoshop, I cover a few basic methods:

<ul>

<li>Image > Mode > Grayscale

<li>Image > Desaturate

<li>Desaturate in Color mode

<li>L channel from LAB

<li>Gradient Map

<li>Channel Mixer

<li>Channel Blending via Apply Image or Calculations

<li>Greg Gormanﾒs Luminosity mask and blend

<li>Russell Brownﾒs Multilayer stack (Hue/Saturation*2)

<li>Glenn Mitchellﾒs Multilayer stack (Desaturate+Selective Color)

<li>Jeff Scheweﾒs Multilayer stack (Multiple layers of B/W channels)

</ul>

(A web search will give links to most of these.)

Grayscale and Desaturate tend to give the worst results for most images. UNmodified, Gradient Map seems to give the best results for typical images. Properly modified, Channel Blending, and various multilayer B/W stacks give the most control.

Almost all of the methods can work for some images, though it's very rare that Grayscale or Desaturate will work well.

I tend to use a modification of Russell Brown's Hue/Saturation layer stack, plus a curves layer underneath.

May 31, 2006

The two standard approaches for sharp-edged patterns is the small cookie/gobo pattern in an optical spotlight (which are available from many sources.), or a large pattern closer to the subject.

A 2x2 foot foam core board (or corrugated board) for example, with cut-out sections can be used quite well. Sharpness is adjusted by moving it closer or further from the subject.

The 53mm pattern is called "D Size", and Apollo, Rosco, and GAM all sell most of their designs in that format.

This topic has come up before; a search on photo.net for gobo/pattern/cookie will probably show a dozen or more threads.

May 31, 2006

One other advantage to the smaller units is that they're, well, smaller--smaller and lighter and easier to move around.

The biggest advantage for me, though, would be that they're less expensive, and that makes it more practical to have two--which lets you avoid the single source of failure issue. No, they don't fail often--only one of my 4 packs have actually failed, and that was a Brownline pack in pouring rain--but not having one is like driving without a spare tire.

I've got 3 800 watt-second packs and one 2400. The 2400 is used almost exclusively for an optical spotlight--or when I have to do large groups outside, and compete with the sun.

May 31, 2006

This won't answer your question directly, but it's still important.

Use the tool as it came from the box.

Adjust the built-in settings to where they work best for you.

After a while, you will discover that some things aren't as easy or as effective as you want them to be. THAT is when you look into extra products or advanced techniques.

Buying something before that is like buying a tool before you know what you need to accomplish--in fact, that's exactly what it is.

May 31, 2006

Another option, though somewhat heavier and not as extendable, is a standard grip arm. The standard length is 40" long, and the grip head slides.

It's much sturdier than any of the reflector arms I have, and I use it much more often. The only downside is that it only holds reflectors up to that length, and it's heavier. It's usually cheaper than a reflector arm, and they're so durable that getting them used is typically a safe bet.

May 31, 2006

"Shooting wide-open with a fast lens (f/1.4-f/2.8) helps keep the background from being an issue."

How is that so in relation to a heavy backlight?

It doesn't affect the backlighting. It simply "helps keep the background from being an issue."

I haven't used HDR much. It doesn't apply to what I photograph--almost exclusively people--as it does require multiple exposures--a single capture (RAW or JPEG) doesn't have enough of a dynamic range for it to operate well (if at all)--making it unsuitable for my use.

May 31, 2006

Oh: and using Adjustment layers rather than applying the Curves (or Levels, or Hue/Saturation, or any other adjustment) is highly recommended! That opens up further control via different blend modes as well as minimizing damage to the original pixels.

May 31, 2006

Curves let you choose exactly which tonal areas to brighten, darken, increase or decrease contrast. In exchange, you must choose.

Levels lets you adjust the black point, white point, and midtone gamma--essentially the same as moving the endpoints in Curves, and only adjusting a single midpoint.

Rules of thumb? Wherever the curve is steeper than 45 degrees, the contrast for that level of brightness has been enhanced; wherever it is at 45 degrees, it's unchanged, and where it's less steep than 45 degrees, it has less contrast. Where the new value is larger, that area is brighter; where it's smaller, that area is darker.

If the original is reasonably decent, it generally won't take more than a few levels of adjustment for any given area--but you may end up with 3-4-5-8-whatever different control points, each up or down by a couple of levels of brightness.

If Curves is the first operation, you can adjust your endpoints, and know that whatever increase in contrast they gave to the image can be allocated where you want to put it. Thus: control of what you "Sell" (give up) and control over what you "Buy" (gain).

The online articles at <a href="http://www.earthboundlight.com/phototips/photoshop-curves.html"> eartboundlight.com</a> and <a href="http://www.cambridgeincolour.com/tutorials/photoshop-curves.htm">CambridgeColour.com</a> provide good overviews of Curves, though I don't recommend using Levels before Curves; that reduces your ability to choose where to trade your contrast--it'll always be distributed across the image-as-a-whole; If you set the black and white points before using Curves, you only get to pick what to "Buy", not what to "Sell". (For many images, it's not critical. For critical applications, it can make a large difference.)

I strongly recommend Dan Margulis' book Professional Photoshop as a guide to mastering Curves.

May 31, 2006

See <a href="http://www.foveon.com/article.php?a=69">Foveon's X3</a> design.

May 30, 2006

I've used metallic for a handful of portraits, and had mixed results.

For high-key, or white backgrounds, none of my samples worked; the metallic sheen worked against the image.

For lower key, and/or richly colored images (jewel-tone clothing, for example), it can really pop.

I get mine from <a href="http://www.acilab.com">ACI</a>.

May 30, 2006

Okay. So I guess you could watch for those opportunities when positioning someone.

Absolutely!

Look at the eyes. See if the sockets are shaded, or if they're getting good light. If the subject was backlit, and the exposure was enough to properly expose for the face, the sky can give a great big soft catchlight, and light bouncing from the sidewalk can do so also.

A reflector makes it easier to control, but if you're familiar with the locations, or good on-the-fly at spotting light, you can often find great places in the least obvious places.

Shooting wide-open with a fast lens (f/1.4-f/2.8) helps keep the background from being an issue.

<a href="http://www.keradwc.com/photos/glamour/image/20060108_141818r1.jpg">No reflector used here</a>

May 30, 2006

While the DALSA article describes the technology accurately for the time it was written, it is also 5 years old--it was written around the same time the 2.7 CCD-based megapixel Nikon D1 came out At the time, and 6 months before Canon's 3 megapixel CMOS-sensored D30. (The penultimate paragraph shows that: "Over time, CMOS imagers should be able to advance into higher-performance applications.".)

In any event, the sensor+support electronics combination is much more important than the sensor's design itself. Consider DALSA's claim that "..CCDs still enjoy significant noise advantages over CMOS imagers..."; if that were universally true, the CMOS-based 1D Mark II and 5D Canons would have much higher noise levels than their "equivalent" CCD-based contemporaries, and the evidence shows that to not be correct.

The sensor, as important as it is, is a tiny fraction of the digital imaging system. As a photographer, I'm much more concerned about how the camera as a whole operates, with image quality being a prime issue. (Did most photographers worry about the dyes used in E6 films, or how well a given film performed?)

January 1, 2005

There's an article on <a href=http://www.photo.net/learn/studiolighting/>Making the Equipment Decision</a> that answers most of that.

Garry Edwards also wrote an excellent 5-part series on <a href=http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=008XTe>Studio Lighting & Equipment</a> which should cover pretty much everthing you've asked except for flash duration--and that's brand and model-specific. (Garry's articles talk about it, and so does mine; neither gives brand advice.)

December 31, 2004

the process in which the media is extracted seems to be my biggest hangup.

That does seem to be the sticking point.

A JPEG image straight from the camera--after some settings in camera have been chosen, much as a film stock is chosen--is no more (or less) manipulated than a similarly automatically processed slide film image. (Negative film, of course, has to be manipulated, as it doesn't directly have a 100% relationship to color or contrast.)

By choosing to label one as 'photography', and one as 'not photography', your original argument alienated many who feel their work IS photography: they light their images (or find the light) the same way; they image on the recording media the same way; and they do or don't do postprocessing.

That it's easier to manipulate digitally does not make it automatic, nor does the opposite apply. George Hurrell heavily retouched his negatives, often to the tune of multiple hours per image. Ansel Adams burned, dodged, bleached, and more his work. W. Eugene Smith did the same. Jerry Uelsmann did a lot more manipulation than most digital photographers do today.

You're certainly welcome to your opinion about Real Photography vs Digital Imagery, but it's exactly that: an opinion that, unfortunately, neglects every other transitional period in photography: wet plate; dry plate; sheet film; roll film; 35mm film; color film; digital (and many more). Each transition was labeled 'not photography' for a while...and time moved on. Those using the older forms continued or converted, and were still "photographers", making "photographs", just as those who converted made "photographs".

December 26, 2004

It all depends on what you're trying to accomplish.

Hard light, as used in the first image clearly brings out the muscles of her stomach, and emphasises the texture of her skin.

Soft light, as used in the second, emphasises smoothness of skin.

This is one of the areas where art comes into play, and personal decisions about intent, plus the ever-present luck factor.

I use hard-and-specular, soft-and-specular, and soft-and-diffuse, in various mixes, as well as different angles, depending on what I'm trying to accomplish.

This image wouldn't have been anywhere near as effective had it been soft light; the texture of the rocks, the leather, her jeans, and the sheen on her skin would all have been reduced.

October 13, 2004

Another possible solution would be to use a <a href="http://www.lensbaby.com/">lensbaby</a>, which is a commercially available lens similar in usage to Mark's PlungerCam. It uses a flexible/compressible tube to permit tilting the lens axis, and has a Holga-quality lens.

September 17, 2004

THE best answer is to check with them and find out what they want. Many 'photo editors' will want sRGB for quick viewing in a browser (which are rarely color managed, and for which sRGB is the kinda-sorta-vaguely close to normal appearance for an uncalibrated display), while others may be working in a production environment where Adobe RGB (1998) or ColorMatch is the norm.

Can you save a JPEG that's not RGB?

Yes. JPEGs can be saved in CMYK or any 8-bit RGB colorspace. (They do get saved internally as YCC or a variant for DCT purposes, but they'll decode as CMYK or RGB.)

September 9, 2004

Command-Option-Tilde (~) loads the composite (~) channel as a grayscale mask. You can do the same thing for any other channel: Command-Option-1, for example, loads either the Red or Cyan channel (RGB or CMYK modes, respectively).

It's an extension of switching to a specific channel, which is Command-1, Command-2... for the "real" channels, and Command-~ for the composite.

You can also Command-click on the channel in the Layers palette to do this.

What it does is load the RGB (or CMYK) composite channel as a selection. The effect is to create a selection with the darkest areas least selected and brightest areas most selected. It's commonly referred to as selecting luminosity, or 'load the luminosity as a selection" or similar.

August 21, 2004

If F4-F5.6 is the "best" that EF-S will ever go, then one of the key attributes of SLR's (shallow DOF) will go with the same way.

We're talking about a 10-22mm lens. How much depth of field do you get at f/2.8 anyhow? (More importantly; how quickly does the 'unsharp' areas become sufficiently blurred?)

I doubt DoF is a serious issue for the ultrawide lens, which is where "most" of the EF-S lenses are. (17-85 being the longest, and for that, yes, it would make a difference.)

August 11, 2004

Read the <a href=http://www.photo.net/learn/studiolighting/>../learn/studiolighting</a> article in the learning section, and take a look at the recent (July/August '04) Weekly Lighting Theme articles by Garry Edwards. That'll help you determine what KIND of gear you want, and the features YOU will need, before you pick specific brands and models.

August 2, 2004

Martin Evening, Barry Haynes, and Scott Kelby would be my top choices for a photographer starting out with Photoshop.

While the first PS book I bought was Deke McClelland's The Photoshop Bible (no version number, as it was the first edition), I can't recommend it for a photographer; he doesn't cover printing or color correction until after page 800 in the CS edition. (I bought it to help fill in the CS-specific information, and since I had Real World and some others of earlier vintage, I figured I'd give it a shot. The Photoshop Bible IS complete, and a good learning tool for many Photoshop users, but it's NOT designed for, nor particularly well suited for, a photographer.

To master contrast and color, Dan Margulis' Professional Photoshop, while seemingly aimed for the pre-press market, does an excellent job of explaining how to get the absolute most from any image. Much of it's overkill if the original is even half-way decent, but it covers optimising and salvaging in more depth than any other book I've seen. (It's not an easy read for most people, though.)

Sign In

Profiles

Forums

Blogs

Events

Downloads

Gallery

Store

Posts posted by kevinconnery