PDA

View Full Version : R-D 1 does a studio session (long)


jlw
03-27-2005, 20:40
Yesterday I gave my R-D 1 its first serious tryout at studio photography, in a gruelling 5-hour session of promo photos for press use and (eventually) a 4-color brochure.

In the past I've done these sessions using a Nikon D100 and a 70-200 VR Nikkor, mounted on a tripod, dumping filled memory cards into a laptop computer as I shot.

This time, I decided to try shooting the whole thing with the R-D 1 (although I did have the Nikon stuff out in the trunk of the car, just in case.) Instead of the laptop, I took along my recently-acquired Epson P-2000 photo viewer. I got the P-2000 partly because it's easier to lug around than a laptop and card reader, and partly because (until more cataloging software gets updated to support the Epson ERF format) it's the only way I can do quick onsite reviews of shots in raw format!

I'm happy to say that the experiment seemed to work well, and I did not have to go get the Nikon out of the trunk!

[To see some sample shots, click here. (http://homepage.mac.com/jlw/photo/R-D1_studio/index.html) ]

However, I encountered a few quirks and unexpected situations which I thought might interest other R-D 1 users. Here's how it went:



Basic setup: The client (I'll call them a client even though I volunteer my time for this group) wanted the shots done in a very simple environment with a plain white background. (Yes, the white background can be a bit drab, but newspapers love it because it helps conceal how lousy newspaper photo reproduction usually is!)

To get this effect, I used a roll of 9-foot-wide white seamless paper, lit separately from the subjects with three electronic flash heads. The subjects were lit with a main light in an octagonal softbox, positioned at a 45-degree angle, and a less-powerful fill light with a shoot-through umbrella over it, positioned at about a 45-degree angle on the opposite side.

The way you get the white-background effect is to first take incident flash meter readings at the backdrop, adjusting the heads until you get even lighting across the backdrop. Note the metered reading, and then open up 2-1/2 stops (no more, no less.) Adjust the power of the lights on the subject to give correct exposure at this opened-up setting. If you get everything balanced properly, this will give a correctly-exposed subject with a clean white background and an evenly gradated horizon in between.

Equipment: I used the R-D 1 with a 50mm f/1.4 Canon lens. I've learned from past experience that with the lighting setup described above, you need an effective lens hood to keep the background lighting from scattering into the lens, reducing contrast. Fortunately, I had the correct clamp-on hood, acquired some time ago in a swap from a friendly RFF member.

I took three batteries for the R-D 1. During the session, I ran down one battery all the way, and the second battery about halfway. I don't have a spare battery for the P-2000; its battery made it all the way through the shooting (about 3-1/2 hours) but began to poop out afterward while we were reviewing the shots (aboutr 1-1/2 hours.) Fortunately, I had its AC adapter, so it was no problem to plug it in and continue. In a "field" shoot, though, I would have run out of juice.

Procedure: I shot all the pictures in raw format. I have two 512-mb memory cards for the R-D 1; each of these holds somewhat less than 50 raw-format shots, so I wouldn't have been able to get through the whole session (about 280 shots before culling) without dumping them onto the P-2000.

In practice, I'd shoot a group of poses on one card -- maybe 20-30 shots -- then take out that card, put it in the P-2000, put the other card in the camera, and continue shooting. I kept alternating, shooting on one card while the P-2000 dumped the other; this gave a very fluid pace to the shooting.

Once we had finished the shots, I plugged the P-2000 into a handy TV set via its A/V output, so the directors and I could review the shots and select the "keepers" to be converted in Photoshop once I got back home. We made these selections by using the P-2000's "album" function, creating a new album and copying the selected images into it. More on this later.

Overall impression: Shooting the session with the R-D 1 worked out really well. It accomplished what I had hoped it would accomplish: making me more productive by letting me work handheld (which saves time when responding to changing shot situations) and helping me feel confident (it's a huge advantage to be able to watch the flashes fire while looking through the viewfinder, especially when trying to catch action shots.)

Now some problems, unexpected results, and "gotchas" --

Flash gotcha: One unexpected problem I've discovered using the R-D 1 with studio flash equipment: its PC socket is polarized. That means that if you use flash gear that has a standard household-type "H" connector for the sync cord, the flash may work when the connector is plugged in one way but NOT the other! When I first plugged in the camera, the flashes wouldn't fire; at first I thought I had a bad sync cable, until it occurred to me to try it again with the connector turned around.

Previewing problems: The big advantage of digital photography vs. film is the ability to check your results on the spot. But I had an unusual amount of trouble with this --

-- For some reason, I found it really hard to evaluate the lighting on the R-D 1's display. Even when my meter readings told me the foreground/background lighting balance was correct, it didn't look correct on the LCD. The same shots viewed on the P-2000 looked okay, so I decided to grit my teeth and proceed, and fortunately they did turn out to be OK. I'm still not sure why this was happening; the very bright background may have had something to do with it, but I can't think what.

-- All the digital-photography articles say that if you want top quality, you must shoot in raw format. I understand the reason for that, and I endorse the theory -- but in practice, I'm beginning to waver. The problem is that there's no way to field-review raw images without going through the time-consuming process of rendering them out in Photoshop; otherwise, previewing -- whether using software or hardware such as the P-2000 -- only shows you the image's embedded JPEG preview image.

What you can do with these preview images on the P-2000 is pretty limited. You can't rotate them; when we were reviewing the images, I got around this problem by tipping the TV up on its side! And you can't zoom in on them -- which would have been very handy when we were trying to make sure people had their eyes open and didn't have goofy facial expressions. I'm still undecided on whether it's better to have the extra image quality of raw format, or the extra reassurance of zoomable, previewable images in high-quality JPEG format.

P-2000 frustrations: Speaking of the P-2000, it's a very slick little piece of hardware. But its software, particularly the file system, is abysmal!

Imagine if, when you downloaded or copied a file onto your PC, the PC operating system immediately assigned it a new, arbitrary filename. Now imagine that if you then copied this file into another folder, the operating system would give it a different filename. And suppose that you weren't allowed to edit these names! There's no way you'd put up with that -- yet it's exactly how the P-2000 works.

Another pet peeve about albums: You can name them on the P-2000, but once it's hooked up to your computer, they just show up as ALBUM001, ALBUM002, etc.

I still feel the P-2000 is a good device for dumping memory cards and handling ERF files, but I need to do some more thinking about an efficient way to manage files on it.

Image Quality Conundrum: Generally I was very satisfied with the quality of the images I got from the R-D 1; they seemed at least as good as the photos I've shot under similar conditions with the Nikon D-100.

If you look at the samples, though, you might notice that many of them seem to have a bluish cast in the areas away from the main lights, producing somewhat cold-looking skin tones. I had noticed this in my D-100 shots as well, but the effect seems more pronounced with the R-D 1.

I tried a lot of tricks for removing this bluish cast during raw conversion and in Photoshop, but didn't find a completely satisfactory solution. (I'm sure it won't bother the client, but I'm a little pickier...)

The only theory I can think of for this is that it's caused by UV radiation -- either directly affecting the camera's imager, or causing white materials in the shot to fluoresce in a way that the imager captures differently than film would do.

So next time, I think I'll try using a UV filter over the lens. (My 70-200 Nikkor does have a UV filter on it, so this may be why the effect was less noticeable with it.) That should reduce the effect if the problem is the direct action of UV on the camera imager.

Unfortunately, if the effect is caused by fluorescence of the materials in the shot, a UV filter won't help -- the fluorescence would be in the visible spectrum, not UV, so a filter wouldn't eliminate it. My flash heads do not have UV suppression -- and I can't afford to replace them with ones that do -- so if that's the problem, I'll just have to live with it. (I'd prefer to eliminate it by shooting in black-and-white, but the clients won't always sit still for that!)



Well, that's my "field report" on doing a studio shoot with the R-D 1! If anyone has any suggestions for the problems I encountered, I'd appreciate hearing about them!

sljm
03-28-2005, 00:37
Whao great shots !!! Gd job showing off the capabilities of the RD-1, but just a question. Is there any other camera with the PC socket polarised ? for i don't seem to remember any other camera that does it.

Sean Reid
03-28-2005, 03:18
Hi jlw,

How did you set the white balance in RAW conversion? Were you sampling from a neutral grey card? Michael Tapes came up with a good set of cards for this purpose, see http://www.rawworkflow.com

I've found that highlights sometimes look like they're over-exposed on the R-D1 LCD when, in fact, they're actually fine. The LCD seems to show a higher contrast than the actual RAW file has. The histogram seems to be accurate though, so I trust that much more than the LCD image. This has been true of many digital cameras I've used.

Cheers,

Sean

tamerlin
03-28-2005, 04:21
Not all of the issues you had with RAW files are endemic to RAW files, though. I use RAW on
my DRebel, and it actually rotates images shot vertically automatically (unless you turn off
that option). Of course, my FlashTrax does not, but that's another story. I think it's a limitation
of the software on the RD-1, rather than of the format. Photoshop lets me rotate RAW images
also.

Nice report, and thank you also for your explanation on uing white backdrops.

jlw
03-28-2005, 04:54
Whao great shots !!! Gd job showing off the capabilities of the RD-1, but just a question. Is there any other camera with the PC socket polarised ? for i don't seem to remember any other camera that does it.

Lots of modern digitally-controlled cameras have polarized sync connections (hot shoe, PC socket, or both.) I know the Nikon D100 does, and so did some of my previous Minolta film SLRs. (On the top-of-the-line Maxxum 9, they went to the extra expense of adding additional circuitry so the flash circuit would not be polarized.) If I recall correctly, this situation first appeared when the Nikon F3 camera came out, so it's been around for quite a long time!

Going by what I've read, this behavior is caused by the use of a transistor, rather than a mechanical switch, to fire the sync circuit. The advantage of the transistor is that it isn't subject to mechanical wear; the disadvantage is that the flash's trigger current will only flow through the transistor gate in one direction (polarity.)

It seems that camera and flash manufacturers have more or less standardized a polarity for hot-shoe-mounted devices -- but if you're using a studio flash with an H connector, the polarity changes depending on which way you plug in the connector. Of course, if you guess wrong, it's easy to turn over the connector... as long as you're aware that's the problem!

Some studio flash units use a coaxial (headphone-type; also known as a Bowens connector because it's used on this brand) sync connector, and wrong polarity can be more of a problem with these because you can't just reverse the connector. Bowens used to offer a 'sync reverser' cable, and this may still be available.

Another option is to use a 'sync isolator' such as the Wein SafeSync. This is essentially an electronic relay that separates the camera's sync circuit from the flash's trigger circuit. Aside from eliminating polarity problems, it also protects the camera's sync transistor from damage that might be caused by the high trigger voltages in very old flash units; some of these are high enough to "blow the gate" in the transistor, permanently disabling flash sync!

(I don't know how valid this is, and if you're in any doubt it's safer just to buy a SafeSync and use it... but I was told by an old-timer that a crude way of testing the trigger current is to fire the flash by shorting its PC cord contacts with the tip of an ordinary lead pencil. If you see a spark at the pencil tip when you do this, your flash has a high trigger current; no spark means it should be safe. The old-timer said that some early studio flash power packs -- designed for the big, heavy sync switch contacts in Kalart-type external synchronizers -- had such a high trigger current that you'd see a big, bright blue spark like the one you'd see from a car's spark plug!

jlw
03-28-2005, 05:35
Hi jlw,

How did you set the white balance in RAW conversion? Were you sampling from a neutral grey card? Michael Tapes came up with a good set of cards for this purpose, see http://www.rawworkflow.com

Hmmm, interesting point. I was sampling from a shadowed area on the white floor. Normally this seems to work -- if the white is neutral, then its shadows should be neutral as well.

But if the sensor is responding to UV, then I may be getting "skew" because the shadows are more UV-rich than the main subject -- the main lights hitting the subject are going through fabric diffusers (which probably reduce their UV content, same way your clothes protect you from sunburn by blocking UV) while the background lights, which are the only ones filling the shadow areas, don't have any diffusers over them.

If that's the case, then if I can't get rid of the extra UV, I should be sampling white balance from a neutral area on the main subject, as you suggest.

I've found that highlights sometimes look like they're over-exposed on the R-D1 LCD when, in fact, they're actually fine. The LCD seems to show a higher contrast than the actual RAW file has. The histogram seems to be accurate though, so I trust that much more than the LCD image. This has been true of many digital cameras I've used.

Agreed. What's tricky in this case is that to get a clean white background, but avoid having it so bright that it "bleeds around" the subjects, I have to get it right on the threshold of being overexposed. At the time, I wasn't sure what the histogram for this should look like.

Now I think I'll pick some of the most successful exposures, load their raw files back onto a memory card, put the card back into the R-D 1, and look at their histograms so I'll know what curve shape I should aim for.

Thanks for the suggestions!

Sean Reid
03-28-2005, 08:16
My pleasure. I'd grab that card from Michael Tapes and use it with each session. http://www.rawworkflow.com/products/whibal/index.html As for the white, why not move the histo over to the left a little and bring the background back up to pure white (if needed) in PS CS?

Cheers,

Sean

jlw
03-28-2005, 09:22
My pleasure. I'd grab that card from Michael Tapes and use it with each session. http://www.rawworkflow.com/products/whibal/index.html As for the white, why not move the histo over to the left a little and bring the background back up to pure white (if needed) in PS CS?


That's what I wound up doing. The reason it's ticklish is that if you've overexposed the original file, the subject highlights will be close enough to the backgroundhighlights that you start blowing out the subject as well.

I got away with it most of the time, but there were a few in which the tones were too close together, and I had to give up on those. Now that I know which original exposures produced separable highlights, I want to go back and memorize their in-camera histograms so I'll be able to reproduce them again.

Those cards look like a great idea! I'll have to check 'em out. I see that he does specifically reference UV reflectivity in his writeup, so evidently I'm not the first person to encounter this sort of thing...

jlw
03-28-2005, 10:20
Not all of the issues you had with RAW files are endemic to RAW files, though.

I know -- they weren't really R-D 1 issues, either. More of a problem of other devices/software not yet supporting the ERF format very well. I should have made that clearer in my original writeup; thanks for pointing it out.

Larry Kellogg
03-28-2005, 12:56
Good work! How do you like that Canon 50 f/1.4 lens? I have one coming to me in the mail and I can't wait to try it out. What do you think about it on the R-D1?

Regards,

Larry

Sean Reid
03-28-2005, 14:06
jlw,

Are you converting from RAW in CS or PhotoRAW?

Sean

jlw
03-28-2005, 19:25
Good work! How do you like that Canon 50 f/1.4 lens? I have one coming to me in the mail and I can't wait to try it out. What do you think about it on the R-D1?

I like it a lot. I've owned one for years and it has always worked well for me. I would describe it as a lens without any specific "character traits" -- just an all-around solid performer that gets the job done. It's possible that the 50/1.8 might be a little sharper -- but the difference (if any) isn't dramatic, and the 50/1.4 is so nice and small that you pay hardly any size/weight penalty for the extra 2/3 stop of maximum aperture.

I've been surprised at how well the 50mm focal length suits me on the R-D 1. I've always shot a lot with a 50 on film, so logically I'd be leaning toward the 35mm focal length on the R-D 1 (since it's roughly equivalent to about 55mm) but it seems that I use the 50 more. That may be because my preference is to stand off a bit and crop tightly.

I'm looking forward to Sean's forthcoming review of high-speed lenses on the R-D 1, since he's going to include these Canon oldies along with several current 50s. I'm sure the latest $3,000 aspherical Summilux will outperform the old reliable Canon... but it will be interesting to see how much, if only because there's no way I can afford the Summilux, and it should be fun to know what I'm missing!

jlw
03-28-2005, 19:31
jlw,

Are you converting from RAW in CS or PhotoRAW?

Sean

I'm using the Epson plug-in, which is the only option available to me using a Mac and Photoshop 7.

Although I use CS at the office, I haven't updated at home... and I'm not going to, since it was just leaked today that Photoshop CS2 will ship in May and I'd rather wait and spend my upgrade bucks for that. Among its new features (per AppleInsider): "Meanwhile, the new Camera Raw 3.0 workflow will support batch processing of raw files in the background and without forcing the user to launch the main Photoshop application. Additionally, the new version will allow settings for multiple raw files to be simultaneously modified, while new non-destructive cropping and straightening controls will allow raw files to be easily prepared for final output."

You probably knew all that, but most likely you were NDA'ed, and I ain't!

Doug
03-28-2005, 19:54
I see that B+W is offering a "stong UV" filter that removes almost all UV present. Might be useful for this situation? http://www.bhphotovideo.com/bnh/controller/home?O=productlist&A=details&Q=&sku=98957&is=REG

jlw
03-28-2005, 20:19
Thanks for the pointer! Might be worth the investment, if only to try to localize the problem.

I wonder if anyone has done any research into distinctive effects of UV (if any) on digital camera imagers...?

[Follow-up: Yes, I guess someone has -- click here (http://www.rit.edu/~andpph/text-reflected-uv.html) to read about it.

For the record, this experimenter says that a digital camera imager's UV sensitivity is so low that you have to go to great lengths to get it to register at all; I guess that shoots my theory that the bluish cast in my results might be caused by the direct action of UV on the R-D 1's imager. Fluorescence is still a possibility, but I don't know how I would test for that...]

RML
03-28-2005, 20:35
I've been surprised at how well the 50mm focal length suits me on the R-D 1. I've always shot a lot with a 50 on film, so logically I'd be leaning toward the 35mm focal length on the R-D 1 (since it's roughly equivalent to about 55mm) but it seems that I use the 50 more. That may be because my preference is to stand off a bit and crop tightly.

I've noticed the same thing with my Eos 300D and my 50/1.8. I shoot 50mm on my Bessa R (sorry, no R-D1 yet) and at first I was reluctant to shoot the 50 on the Eos. But now, a few months later, I find it suits me well. I'm not going over to an 85mm on the Bessa (though I do have one, which I occassionally use) but I doubt I'll shell out any money for a 35mm for the Eos.

Doug
03-28-2005, 21:15
For the record, this experimenter says that a digital camera imager's UV sensitivity is so low that you have to go to great lengths to get it to register at all; I guess that shoots my theory that the bluish cast in my results might be caused by the direct action of UV on the R-D 1's imager. Fluorescence is still a possibility, but I don't know how I would test for that...] Interesting... I suppose if it IS UV fluorescence, that puts the emphasis on removing UV from the light sources.

Sean Reid
03-29-2005, 00:01
Odds are that nailing the WB in RAW conversion (using those card Michael Tapes came up with) will do the trick.

Cheers,

Sean

Sean Reid
03-29-2005, 00:43
I've been surprised at how well the 50mm focal length suits me on the R-D 1. I've always shot a lot with a 50 on film, so logically I'd be leaning toward the 35mm focal length on the R-D 1 (since it's roughly equivalent to about 55mm) but it seems that I use the 50 more. That may be because my preference is to stand off a bit and crop tightly.

You know, one thing that's interesting about that is that a 50 on the R-D1 is still a 50. People sometimes think of it as a 75 but it's really a cropped 50 and I make that distinction because a 75 has a different kind of spatial compression, a different abrubtness of transition from sharp areas to OOF areas, etc.. So the frame edges come in closer, of course, but the "drawing" is a 50mm kind of drawing, not a 75mm kind of drawing. I need to remember to talk about this in the fast lenses article. Erwin Puts talked about this aspect in his R-D1 review but I don't think he's yet made this distinction as clear as he could. It's an important point though. Wide lenses of course, expand the space in a picture (not just FOV but the sense of space as the eye works across the depth of the picture) whereas telephoto lenses, of course, compress it. So that tension or balance between expansion and compression of space that is inherent in a 50mm (and that I think appealed so much to Cartier Bresson, for example) remains the same when that lens is mounted on a body that crops its FOV. So, while a 35 on the R-D1 provides a similar FOV to a 50 because of the sensor's cropping effect, it still draws like a 35. In that sense, it can't replace a 50 in terms of many of the subtle cues a picture gives us about space as seen by a lens.

Some might see this as a drawback of sensors that are less than FF. I think it just opens some interesting doors.

Cheers,

Sean

RML
03-29-2005, 00:56
Sean, it can never hurt to make this distinction clear.

Ed Schwartzreic
03-29-2005, 03:08
I'd like to clarify what Sean is saying, so if he thinks this is not correct, he should definitely respond.

Any 50mm lens, for whatever format, will behave similarly with respect to DOF and what Sean calls "drawing." Therefore a 50/4 Distagon for Hasselblad and a 50/2.8 for a Pentax 110 have similar characteristics in this regard, despite their major differences in image circle / coverage. The 80/2.8 Planar on my Rolleiflex and my 75/1.4 Summilux will likewise be similar.

Ed

jlw
03-29-2005, 05:12
You know, one thing that's interesting about that is that a 50 on the R-D1 is still a 50. People sometimes think of it as a 75 but it's really a cropped 50 and I make that distinction because a 75 has a different kind of spatial compression, a different abrubtness of transition from sharp areas to OOF areas, etc.. So the frame edges come in closer, of course, but the "drawing" is a 50mm kind of drawing, not a 75mm kind of drawing. I need to remember to talk about this in the fast lenses article.

True, although to be pettifoggingly academic, the spatial effects of perspective depend on the relationship between the lens' viewing angle of the scene and the eye's viewing angle of the reproduction.

There's a longish but very clear explanation in one of Kingslake's books: Imagine that you're viewing a scene with your naked eye, which by definition produces "normal" perspective. Now hold up an 8x10-inch sheet of glass at normal reading distance, and use a grease pencil to trace the outlines of objects in the scene.

This tracing will reproduce the spatial relationships of objects in the scene, and consequently also will have "normal" perspective -- as long as you continue to view the sheet of glass at the same distance.

If you hold the glass at a shorter or longer distance, the outlines will have a different spatial relationship to the scene, and an "unnatural" perspective. This represents what happens when you use a wide or long lens instead of a "normal" lens.

Note, however, that if you re-traced the scene with the glass held at a shorter or longer distance, the perspective would still appear "natural" when the glass was viewed at that distance. This corresponds to the fact that the spatial relationships of a picture taken with a super-wide-angle lens appears natural if you view the print from a very close distance, and the spatial relationships of a super-tele picture appear natural if you view it from a very long distance.

In other words: As long as the angle subtended by the eye when viewing an object in the reproduction is the same as the angle subtended by the eye when viewing the object in its original scene, the spatial relationships will appear "natural."

This concept also can be used to illustrate Sean's point -- that using a 50mm lens on a small-sensor digital camera doesn't change the spatial relationships compared to using the same lens on a 35mm camera.

To follow Kingslake's analogy, you're still making the sketch while holding the glass at the same distance; the only difference is that you're sketching on a smaller piece of glass.

Does that make sense to anyone besides me?

Sean Reid
03-29-2005, 10:40
Well...to clarify further (and this all is good practice for writing a final version on this topic in the fast lenses review). A given 50mm lens will produce the same optical image no matter what camera it is mounted on. I do mean image, not picture, although the former term is often (mistakenly, I think) used as a synonym for the latter. DOF is a measurement that can only be expressed in relation to a given print size with defined circle of confusion limits. An 8" x 12" print from the R-D1 involves greater magnification than the same size print from a frame of 35mm film. As such, in theory, the CoC requirements are more stringent for a given area to be defined as "in focus". For that reason, the effective DOF of a given 50mm lens (mathematically) decreases as the size of the capture medium decreases and/or as the size of the final print increases. The smaller the capture medium (or the larger the final print), the greater the magnification and thus the greater the demands placed on focus. So DOF isn't an independent property of the lens itself and, by definition, can't be. In fact, perceived DOF can also be affected by the way in which the digital file is sharpened in post-processing. It's a muddy business. <G>

The way in which a given lens "draws", that is to say the way it converts a three-dimensional world into a two-dimensional projected image, varies from lens to lens and sometimes even among examples of a given lens. The ways in which lenses can vary in their drawing is almost infinite which is why MTF numbers, etc. can't even begin to describe a lens accurately (any more than one could describe the taste of a specific wine using the number "seven"). But...a 50mm lens (as a category) has a specific way in which it portrays objects in relation to each other in space. It's somewhere between the expanded sense of depth created by a wide angle lens and the compressed sense of space created by telephoto lenses. To my eye, the longer the focal length, the more the drawing of the lens starts to approximate a collage.

jlw wrote:

"To follow Kingslake's analogy, you're still making the sketch while holding the glass at the same distance; the only difference is that you're sketching on a smaller piece of glass."

Exactly, beautifully said. That is exactly true. And yes, of course, your post makes sense.

The practical relevance of this, methinks, is that those of us who like the way a certain focal length (say 50mm) draws objects in space, may still prefer to use that lens on the R-D1 even though it now has a 75mm FOV. The borders of the picture may match those made with a 75 but the content inside those borders is still pure 50mm drawing.

Cheers,

Sean

LCT
03-29-2005, 14:01
Yes but perspective depends on the position of the lens w.r.t. the subject.
If i use a 50mm lens on the R-D1, i'll have the same FoV but also the same perspective as a 135 film user with a 75mm lens because both of us will shoot from the same place approximatively.
Then my pic taken with the R-D1 will not be a pure 50mm drawing but a drawing with 50mm DoF and 75mm FoV and perspective IMHO.
Best,
LCT

Ed Schwartzreic
03-29-2005, 14:20
Aren't we confusing two things here? A 50 on an R-D1 and a 75 on an M6, shot of the same subject at the same distance, and both enlarged full frame to an 8x12 size and viewed from the same distance should show the same perspective.

On the other hand, all 50mm lenses ought to behave more or less the same with regard to how they portray objects in relation to each other in space, and all 75mm lenses should more or less behave the same, but different from the 50's.

Ed

Sean Reid
03-29-2005, 15:03
"On the other hand, all 50mm lenses ought to behave more or less the same with regard to how they portray objects in relation to each other in space, and all 75mm lenses should more or less behave the same, but different from the 50's."

Agreed.

"Aren't we confusing two things here? A 50 on an R-D1 and a 75 on an M6, shot of the same subject at the same distance, and both enlarged full frame to an 8x12 size and viewed from the same distance should show the same perspective."

What do you mean by perspective? Field of view?

Sean

Sean Reid
03-29-2005, 15:06
Yes but perspective depends on the position of the lens w.r.t. the subject.
If i use a 50mm lens on the R-D1, i'll have the same FoV but also the same perspective as a 135 film user with a 75mm lens because both of us will shoot from the same place approximatively.
Then my pic taken with the R-D1 will not be a pure 50mm drawing but a drawing with 50mm DoF and 75mm FoV and perspective IMHO.
Best,
LCT

Hi LCT,

What do you mean by "perspective"? Are you using this in the traditional sense of vanishing points and the like?

Cheers,

Sean

Ed Schwartzreic
03-29-2005, 15:50
Yes, FOV and vanishing points, as well as DOF if the 75mm lens is closed down 50% relative to the 50.

LCT
03-30-2005, 02:12
Yes even the DoF won't be exactly a 50mm one due to the different circles of confusion.
It's rather a 50mm DoF at one f stop wider in practice.
For example the same 50mm lens gives approximately the same hyperfocal distance @ f/8 on the R-D1 as f/5.6 on a 35mm film camera.
Best,
LCT

Sean Reid
03-30-2005, 03:12
I talked about the DOF factor above. I wouldn't interchange the term "perspective", as its used in drawing for example, with the aspect we're discussing. One's vantage point and distance from the subjects might be the same with a 75mm and with a cropped 50mm but the compression of space each lens creates will not be the same. Of course we're not talking about space side to side in the frame or top to bottom, that's defined by the picture edges. That kind of space, field of view, will be the same between the 75 and the cropped 50.

The space that I'm talking about, and I think some may have misunderstood this, is the space across the depth of the picture from foreground objects to background objects. Wide angle lenses will make foreground and background objects seem further apart (they expand the drawing of the space), telephoto lenses make them seem closer together (they compress the space). This quality doesn't change when one crops the image circle in tighter (as the R-D1 does relative to a 35mm camera). One might well stand in the same place with an R-D1/50mm lens in one hand and an M6/75mm lens in the other. Both combinations would yield the same field of view and could have the same picture edges. But in pictures made with the latter combination, foreground and background objects will appear to be closer to each other. This isn't because of perspective, per se, it's because of the differences in the ways these two lenses draw space. It's an aspect of lenses I seldom see discussed.


The longer a lens is, the more its image appears to compress the picture space - the more the drawing of the lens begins to approximate a collage. That's not quite the same as perspective. One can crop the image from a 50mm lens to show the same picture edges as a 75mm lens (which is what's happening with the 1.5X sensor mag) but that doesn't mean the 50mm will compress the picture space in the same way as the 75.

I think it might be good to illustrate this point with pictures which I'll try to do as soon as time allows.

Cheers,

Sean

Ed Schwartzreic
03-30-2005, 03:31
Posting pictures will be really helpful, Sean. I'm torn on this subject between Kingslake's text, which supports exactly what you say, and the material in Scheerer's (?sp) and Makovitch's Leica book which shows a 21mm view of the Norwegian town of Alesund, with marked-off areas for 400mm, 200mm, 135mm, 90mm, and 50mm fields of view, then also the shots taken with these lenses. I'll try to post this picture. I think the problem may be that at infiinity this all works, but we are talking about images made of objects in space at various distances.

Ed

Sean Reid
03-30-2005, 03:46
"I think the problem may be that at infiinity this all works, but we are talking about images made of objects in space at various distances."

Exactly. I wish I could make illustrations for this today but I'm behind on several projects right now and never seem to be able to catch up fully. Does anyone else want to make some picture experiments and post them?

Cheers,

Sean

pfogle
03-30-2005, 04:40
Maybe a silly question, but we never had all this concern when moving between medium format and 35mm, which presents exactly the same issues; we just know that the lenses give different results. Seems to me the same thing applies here. Without specifying the degree of enlargement, the arguments about DOF etc are purely academic.

Also, perspective is a function of the point of view only; none of the other factors (focal length, format etc) have any effect.

just my 2c worth, sorry to interrupt! ;)

Phil

Ed Schwartzreic
03-30-2005, 05:36
Here is the 21mm Sheerer picture with the FoV of several lenses marked out upon it. Here also are the 200mm and 135mm segments, taken with their respective lenses.

Edhttp://www.rangefinderforum.com/forums/attachment.php?attachmentid=5037&stc=1
sheerer1.jpg

LCT
03-30-2005, 07:40
Maybe a silly question, but we never had all this concern when moving between medium format and 35mm, which presents exactly the same issues...

Right but we can hardly view the results of a Leica lens on a MF camera can't we ;-)
To some Leica users at least the R-D1 has become a digital back for M lenses sort of and it is interesting to check if the qualities, flaws and 'signature' of such lenses are maintained in the digital field.
Hence those comparisons i guess.
Should the 'cron 50mm be treated as a 75mm lens for example or would it be usefull to purchase the new 'cron 75mm for R-D1 users?
Would the 'cron 28mm give the same results on a R-D1 as a 'cron-C 40mm on a Leica M?
... and so on.
Best,
LCT

pfogle
03-30-2005, 08:37
good point LCT :)

Sean Reid
03-30-2005, 10:45
Those of using the R-D1 and eager to share information about it are a very small group. But I'd venture a guess that among that group are many people who are very aware of (and who care about) the way various lenses render the world. Some of us are professional photographers and some aren't but most of us seem to be quite thoughtful and intentional in our photography. That being the case, it's important to know that a 28 on the R-D1 definitely does not draw like a 40 on a Leica M. They'll show about the same FOV but are distinctly different animals. I think this in part explains why people seem to like the 50 on the R-D1 even though they didn't work much at the 75-80mm range before with film. The 50 is cropped but its still rendering like a 50, not like a 75. In fact, I think the juxtaposition is really interesting. For a given FOV, what we now have, with these APS-C sensor cameras, are lenses that show more effective DOF and less spatial compression. For example, if someone is using a 35 on the R-D1 in lieu of a 50 on a Leica M he or she now has the combination of 35mm (focal length) drawing and most of the 35mm DOF potential with the FOV of about a 50. It's a fascinating kind of combination and I expect that, over time, very good photographers will make use of it just as Cartier-Bresson made use of the space compressed/space non-compressed middle ground of the 50mm lens.

The point made about medium format vs. small format film is true but kind of a horse of a different color. One generally used a given lens on either an MF or an SF camera, not both usually. If one got a feel for a certain 80mm MF lens, that feel probably came from using it on an MF body. One would be unlikely to mount it on a 35mm body (even if that were possible).

But it's different with these digital bodies that take 135mm format lenses. Most of us have been at this for awhile and have, consciously or unconsciously, come to know the drawing of various 135mm format lenses, whether they're from Leica, CV, Canon, Nikon, etc.. I have a friend, for example, who does much of his professional work with the 50 Noctilux. He really knows that lens and is very familiar with the way it translates the world. When he puts that lens on an R-D1 (as he is doing right now, in fact) the edges of the picture are all going to move in a bit but the lens will still behave like itself.

So, since we are now using lenses, that we know from 35mm film use, on the R-D1 it really is useful to know that a 28 doesn't really become a 42, etc.. The image is just being cropped.

Cheers,

Sean

LCT
03-30-2005, 12:22
... For a given FOV, what we now have, with these APS-C sensor cameras, are lenses that show ... less spatial compression ...

Interesting indeed Sean.
I've never heard of such spatial compression effect until now.
Is there some litterature on the subject?
Best,
LCT

Sean Reid
03-30-2005, 13:09
Hi LCT,

I'll look but you've no doubt experienced this effect without calling it by that name. If I come across a good link on this I'll post it. It's one reason that you're able to recognize a telephoto picture as such even without knowing how close the person was to the subject. It's such a familiar effect that many of us, I think, never think about it.

Cheers,

Sean

Sean Reid
03-30-2005, 13:20
LCT,

Erwin Puts starts to talk about this in his review of the R-D1 at: http://www.imx.nl/photosite/japan/epsonrd1/epsonrd1.html

One quote from that article which is relevant: "You may thus conclude that the correction factor of 1.53 gives you the same magnification of the subject as the original 50mm lens, but it will simply show less image area. When you say that that the original 50mm lens will become a 75mm lens, this is only true for the scene framing, not for the focal length. There is no magnification of the scene as will happen when you replace the 50mm lens by a 75mm lens."

It's that magnification that creates the compression of space and it also goes hand in hand with decreasing depth of field.

Cheers,

Sean

jlw
03-30-2005, 13:30
I'm going to start a new, separate thread (under 'And Now for Something Completely Different,' since it's not specifically RF-related) with some visual examples which I hope will clear all this up a bit!

Sean Reid
03-30-2005, 13:30
Here's a quick one: http://bj.canon.co.jp/english/photoshooting/technihsc/howtophotograph/howtophotograph08.html

Sean

LCT
03-30-2005, 22:26
Yes of course.
Forgot what i've learnt in my youth.
Thanks Sean.
Best,
LCT

pfogle
03-31-2005, 01:01
Sean... your quote from Puts puts (no pun intended ;)) it well. I think that some of the confusion comes from the fact that the 'magnification' he mentions arises from the enlarging/printing process as well as from the lens. However, DOF does not scale the same way in enlarging/printing as it does in changing the focal length of the lens.

I guess the upshot of all this, and the point I was (not very well) trying to make, is that what we are interested in is the characteristics of a *system* ie the whole shebang up to the finished print, and the R-D1 is a different system, in the same way that MF is a different system. So a 50mm on a 35 body is a different system to a 50mm on an R-D1, and will have it's own unique character.

I'd just like to say at this point how much I appreciate your work, and the input of all the other people who contribute to this forum - it really is a special place!

cheers
Phil

Sean Reid
03-31-2005, 04:31
jlw posted a separate thread here: http://www.rangefinderforum.com/forums/showthread.php?t=5121 and his summary point is: "But it's the distance change, not the lens or format change, that causes the perspective relationship to change." I think that statement points out that, for the sake of precision, we should be talking about the "perceived" or "effective" spatial compression of longer lenses. It's a similar distinction I always try to make when writing about DOF (and now need to make with respect to spatial compression). DOF, in a pure sense, is exactly the same for all lenses (at a given aperture) regardless of focal length *if* the primary subject is kept exactly the same size on the capture medium. That means, of course, coming closer to the subject with a wide lens and further from the subject with a long lens. So that's an absolute property of DOF. Yet, we generally experience DOF much differently because we do not always keep the subject the same size on the sensor or film. Our actual experience of lenses, usually, is that we see more DOF in pictures we make with wide angle lenses than we do with pictures made with telephoto lenses based on the way we actually work with these lenses. In absolute terms, a 12mm lens has exactly the same DOF (at a given aperture) as a 300mm lens (again, if the subject size is kept exactly the same on the capture medium). In practice, though, few of us would be comfortable zone-focusing a 300mm whereas we might readily do that with the 12mm (in our normal picture-making). The way we tend to actually use these lenses leads us to perceive that the "effective" DOF is greater with the wider lenses. For example, the wide angle end of a Digilux 2 lens is 7mm. If subject size on the sensor was kept absolutely constant, that 7mm lens would have exactly the same DOF as a 300mm lens. But to accomplish that we'd need to be almost pressing that 7mm lens up against the subject's nose (for example). It's a property that's true in an absolute sense, but not very true to our experience. That's why it seems unbelievable to many of us that all focal length lenses could have exactly the same DOF.

I think the distinction is similar here for the relationship between spatial compression and focal length. In actual practice, we tend to use wider lenses closer to our subjects and longer lenses further from them. A concrete example: When Garry Winogrand was working in the streets of NYC, for example, he was often working with a 28mm lens about 6-10 feet from his primary subjects. His framing is often such that the height of the pictures (made at that distance) is defined largely by the height of his primary subjects. His work was mostly made with 28mm and 35mm lenses, working fairly close to his primary subjects. Now imagine a young photographer who is interested in Winogrand's subject (nominally people) but feels too shy to work close in to people. So he starts working from much further away with a 135mm lens. The figures perhaps appear the same height in his frame that they did in Winogrand's pictures but the spatial compression seen in the latter pictures is vastly different (as is the drawing of the figures themselves). In general, we use longer lenses when we stand further back and as such the truth of our experience (generally) using these longer lenses is that space is compressed, just as our experience of using wide angle lenses (at the distances we normally work with them) is that they have a greater DOF.

Cheers,

Sean

LCT
03-31-2005, 06:46
You're opening new perspectives in my FoV Sean ;-)
Always a pleasure to read you.
Best,
LCT

Sean Reid
03-31-2005, 07:14
Thanks LCT.

Cheers,

Sean

Phil_Hawkes
07-01-2005, 00:46
Hi all,

I've been thining about DOF issues. I gather that (without considering the field of view) the depth of field on the R-D1 using focal length f at aperture x would be the same as on a M6 using focal length f at aperture x/1.5.So the photo I take with 50/1.0 on the R-D1 has the same depth of field as a 50/0.7 on M6. A few comments I've seen on this forum seem to confirm that this is the accepted view.

I was reading a few threads on this, and came across Sean's comment...

jlw posted a separate thread here: http://www.rangefinderforum.com/forums/showthread.php?t=5121 and his summary point is: "But it's the distance change, not the lens or format change, that causes the perspective relationship to change." I think that statement points out that, for the sake of precision, we should be talking about the "perceived" or "effective" spatial compression of longer lenses. It's a similar distinction I always try to make when writing about DOF (and now need to make with respect to spatial compression). DOF, in a pure sense, is exactly the same for all lenses (at a given aperture) regardless of focal length *if* the primary subject is kept exactly the same size on the capture medium. <snip>

What stood out to me was this..

DOF, in a pure sense, is exactly the same for all lenses (at a given aperture) regardless of focal length *if* the primary subject is kept exactly the same size on the capture medium. <snip>

I haven't heard that comment before. I was wondering if there are some references/links to some diagrams or text that explains this. I'm intrigued!!!

Cheers
Phil

Doc William
07-02-2005, 13:08
Odds are that nailing the WB in RAW conversion (using those card Michael Tapes came up with) will do the trick.

Cheers,

Sean

Sean,

I would only warn those people that are converting in CS2 that the new Adobe RAW converter has a white balance picker. Bruce Fraser in his recent book "Real-World Camera Raw with Adobe Photoshop CS2" warns that you should pick a white in your image that is not blown out rather than the neutral grey card we are all used to, to get everything into balance. Again, that is with the new Adobe RAW converter in CS2.

Peace,
bill

Doug
07-02-2005, 14:41
Phil, in your first paragraph, there are two necessary elements not considered... First is subject distance or magnification, and second is the user's choice of "circle of confusion" that are both used in the computation. If the focal length is the same, the aperture diameter is the same, the subject magnification is the same, and you choose the same circle of confusion, then Depth of Field is the same regardless of format dimensions or cropping, etc.

The thing in comparing DoF with different sensor/film sizes is that the enlargement ratio is likely to be different, and for that reason CoC would likely be chosen differently. You blow something up larger, it gets fuzzier before and behind the point of focus, so apparent DoF is less... the CoC has been magnified as well. To keep the apparent DoF the same, the CoC must be kept smaller at the sensor, requiring a smaller aperture...

Since there are a number of related factors, and it all boils down to human perception in the end anyway, it gets difficult! There are a few DoF calculator web pages online that are fun to play with...
http://www.dudak.baka.com/dofcalc.html
Here's one in the form of an Excel spreadsheet (the blur circle mapper)
http://www.smartgroups.com/vault/lens_grading/Public/

MarkEDavison
07-02-2005, 18:43
Let me suggest an actual experiment which can clarify some of these matters, and which is fun to conduct (I just now did it myself).

Take your R-D1 and mount it on a tripod facing a scene like a room with furniture where there are plenty of features which will provide persepective cues. Without moving the camera, photograph the scene with, say, a 15mm, 35mm, and 50mm lens, stopped way down so in each case the entire scene appears to be in focus. Pull the images into Photoshop. Find a pair of distinguishing points (which appear in all three images) where you can easily measure the distance between the points using the ruler tool. Use the resulting measurements to figure out the relative scale difference between the three scenes. Now shrink the 35mm and 50mm scenes by the appropriate scales, and overlay the shrunken images on the 15mm scene. You will find that they match almost exactly. That is, in terms of geometric perspective, the images produced by the 35mm and 50mm lens are identical to crops from the 15mm image. Just such an overlay is attached below. The 35mm image overlay is colored yellow, the 50mm overlay sort of grayish blue. I admit that the images don't align exactly perfectly, but they are very very close.

Beware if you do this experiment yourself--some lenses display differing amounts of barrel or pincushion distortion. These are departures from ideal rectilinear perspective which will cause the images not to align.

The academic conclusion is that if we just had a good enough 15mm lens (with infinite resolution) and a digital sensor with high enough resolution we could dispense with carrying lenses of different focal lengths and simply carry the 15mm, and magnify crops from images produced with that lens.

But in point of fact, we do not have infinite resolution in either lenses, or sensors. We use different lenses because they draw the fine details differently, not because there is some generic difference in the way they render geometric perspective.

Now why do photographers insist that telephoto lenses compress space, and wide-angle lenses foreshorten in? Because we typically create prints that are all the same size, and then view the prints from the same distance. With telephoto lenses, we are viewing the prints from a point which is closer to the print than the center of perspective, or natural viewpoint, and this creates compression. For wide angle lenses we are viewing the prints at a position which is further from the print than the natural viewpoint, and this produces foreshortening.

I think you will find that if you take an R-D1 image with a 50mm, and then take a 35mm image with a 75mm lens, the two images will be identical in terms of geometric rendering, but different in terms of rendering of fine details and depth of field effects.