• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Our new Panasonic Lumix DMC-FZ200

Doug Kerr

Well-known member
Our new Panasonic Lumix DMC-FZ200 arrived last evening about 7:00 pm (we are near the end of the Alamogordo UPS truck route).

Most of our work with it last night was devoted to functional testing (how is the EVF, various response speed issues, etc.). There was essentially no shooting that would lead to image quality assessments (that will begin this morning).

Overall we are well impressed by the machine.

It is almost exactly the size of our Canon Powershot SX20 and just a little lighter.

The user interface is well organized, and in many ways quite superior to that of the SX20.

The EVF is superb. It uses a field-sequential approach to "tricolor" display, and so there might be some color ghosting in some situations; we we have not yet seen any of this.

The display panel is very nice too. It fully articulates. When facing forward (for "selfies"), the image reverses, making it act like a mirror. I don't prefer that, but that's because I am used to dealing with orientation challenges. I think that feature cannot be canceled (but I'm not sure).

Shutter release time (an important factor here) is very short (even considering the matter of finder lag). Shot-to-shot time (reported as 0.85 s) seems to be about that. As I had suspected, by putting the machine into a burst mode (there is a dedicated button for that), one can then fire individual shots at much closer intervals.

The onboard flash recharges fairly quickly for "close" shots (although we will probably use an external flash much of the time—likely a Metz 24A-1, which comes in a version dedicated to the Olympus/Panasonic/Leica interface).

I have never liked motorized zoom control, but this machine has a pretty good one, with a nice user interface feature: in addition to the traditional swing arm around the shutter release button, there is also a slide bar on the lens mount. (In manual focus mode, one can have that be the focus control, while we zoom with the regular "swing arm".)

The rear control wheel (often reported to be "spongy") seems pretty snappy in our copy. It has a "push in" aspect, which can be used to change what parameter it sets, or to make it do exposure compensation setting.

The menu system seems pretty nice.

One review complained about the time to switch to "play" from "record" (shoot) as being about three seconds, but I suspect there is a decimal point awry there; the change is pretty snappy.

Unlike on the SX20, if one is shooting with the EVF, the default for "play" is the display panel (often most appropriate). But if we want to review the shots in the EVF, we can easily switch to that, and once we have done that a couple of times, the machine makes that the default.

The machine comes with a nice lens hood; of course it is the common "petal" design, so it is pretty useless with respect to errant light coming in from diagonal directions.

More later.

Best regards,

Doug
 

Doug Kerr

Well-known member
This is an early test shot with the DMC-FZ200 in a "plenty of light" situation:

Test_P1000135-02-S800.jpg


Douglas A. Kerr: Down the hill—DMC-FZ200

This was taken at ISO 200 (likely our "default" mode for most work with this machine).

I took the basic 4000 px × 3000 px JPG output. Then I processed it thus:

• Applied Gaussian blur (suggestion by Bart)

• Downsampled it to 800 px wide using bi-cubic interpolation.

• Sharpened with unsharp mask.

For comparison, I took the same scene just a few minutes later with the Canon EOD 40D equipped with the Canon EF 24-105 f/4L IS lens:

Test_F40549-02-C43-S800.jpg


Douglas A. Kerr: Down the hill—EOS 40D

I used a focal length such that the vertical field of view was very nearly what I had with the Lumix.

I then cropped the image so that it had about the angular extent of the Lumix image (just a tiny bit if vertical cropping).

I then downsized it to 800 px wide using exactly the same process as for the Lumix image.

(I think there might have been some small focus problem with this shot—not sure why.)

Best regards,

Doug
 

Jerome Marot

Well-known member
The legends appear to have been swapped...

On the Panasonic picture, I see strong jpeg artifacts (e.g. around the yellow traffic light).
 

Doug Kerr

Well-known member
Hi, Jerome,

Yes, when that downsized Panasonic image was saved, the JPEG compression was accidentally set very high (a result of some other testing that was going on!).

I have replaced the image in my original post.

Thanks again for the catch.

Best regards,

Doug
 

Jerome Marot

Well-known member
Now the Panasonic image appears to be slightly more pleasing, but in truth it has only a bit more sharpening. It is not possible to differentiate the camera qualities at this resolution. Or, if you prefer to see it that way, both cameras are equals at this resolution and light level.
 

Doug Kerr

Well-known member
Dear Jerome,
Now the Panasonic image appears to be slightly more pleasing, but in truth it has only a bit more sharpening.
I think it suffers less from inappropriate focusing!

It is not possible to differentiate the camera qualities at this resolution. Or, if you prefer to see it that way, both cameras are equals at this resolution and light level.
Well said.

This is why we conclude that we can use a camera like this for "most of our work" and get the advantages of less bulk and weight.

Best regards,

Doug
 

Doug Kerr

Well-known member
Interestingly enough, those two frames, ex camera, printed at 8" × 10" with Qimage (cropped as needed to fit and match), are almost indistinguishable.

Best regards,

Doug
 

Doug Kerr

Well-known member
This morning I took some interior shots (the living room at World Headquarters) with the camera on a tripod, at different ISO sensitivity settings.

The metered scene luminance was Bv -1 (what some would call, wholly-inappropriately, "Ev 4"), 1.7 cd/m².

There actually might be a small error in this determination, since the camera sensitivity settings are based on ISO SOS, while my APEX calculator is based on ISO speed.​
The camera output was in every case 4000 px × 3000 px.

Firsts, we will take three frames, shot as ISO 200, ISO 800, and ISO 3200, and present them after downsizing to 800 px wide using our new "Standard" process; there was no other postprocessing:

Test_P1000201-0200-S800.jpg


ISO 200; downsized to 800 px wide


Test_P1000203-0800-S800.jpg


ISO 800; downsized to 800 px wide


Test_P1000205-3200-S800.jpg


ISO 3200; downsized to 800 px wide​

[continued]
 

Doug Kerr

Well-known member
[part 2]

Here we see 20% x 20% crops (800 px × 600 px) of the original frame, incorporating a small decor grouping, at original camera resolution; there was no postprocessing other than the cropping:

Test_P1000201-0200-C1.jpg


ISO 200; 800 px × 600 px crop at original camera resolution


Test_P1000203-0800-C1.jpg


ISO 800; 800 px × 600 px crop at original camera resolution


Test_P1000205-3200-C1.jpg


ISO 3200; 800 px × 600 px crop at original camera resolution​

All very interesting.

Best regards,

Doug
 

Doug Kerr

Well-known member
I took the original frames of this series for ISO 200, 800, 1600, and 3200 and printed them at a size of 5" x 7" (small crop required at top and bottom), using Qimage. There was no postprocessing.

Viewed directly, I cannot distinguish these prints (maybe in the ISO 3200 I can see just the slightest degradation of detail in the brick on the fireplace). With a loupe, I can see small differences in detail, edge artifacts, and so forth.

Very impressive.

Tomorrow night I have to shoot a banquet, which I plan to do with the new machine. It will probably have to mostly be available light, as the "real" flash unit I have ordered for the machine will not be here in time (although I will certainly experiment with the onboard flash system).

This has better go well. Carla is the "producer" of the banquet (it is one component of a very complicated weekend-long charity fund-raising program), the mayor will be there, and so forth.

So far, we are very happy with our choice of a "main B camera".

A small lapse in the deign (happened a few versions back) is that there is no rubber eyecup for the EVF, sometimes a problem in operation in very bright light (which of course we have here most of the time).

In a long thread on this problem on DPR (I first thought I was in the wrong place—there was no catfight) a couple of guys reported good luck in adapting a certain rubber eyecup intended for telescope eyepieces to this machine. I have one on-the-way.

By the way, a nice feature of this machine (getting common in compacts) is that there is internal compensation for geometric distortion for the JPG outputs.

Best regards,

Doug
 

Doug Kerr

Well-known member
The Panasonic Lumix DMC-FZ200 has an internal HDR mode. If offers the user no control over its details. It takes perhaps 3-5 frames, seemingly at intervals of about 200 ms, then combines them.

Late this afternoon, I shot some clouds from the backyard, and got this (resized to 800 px × 600 px with our standard downsizing process.);

Test_P1000208-01-S800.jpg


Douglas A. Kerr: Afternoon clouds over Alamogordo

After I looked at the result, I thought, "Maybe this would be a good chance to try out the Panny's internal HDR feature." By that time, the clouds looked really nice.

So I did, and with no fuss or muss this came out of the camera:

Test_P1000213-01-S800.jpg


Douglas A. Kerr: Afternoon clouds over Alamogordo, with HDR processing

Way neat!

Best regards,

Doug
 

Doug Kerr

Well-known member
We've been having a lot of fun with the DMS-FZ200, and overall are well pleased with its behavior and performance.

AF system

The AF system (a so-called "contrast detection" system) does well, and is pretty quick. Part of that comes from a clever ploy - the system continuously tracks focus (on a leisurely basis) wherever the camera happens to be pointed. So often, as we bring the proposed subject into view, much of the AF work is done before we half-press.

Flash unit

We had ordered a Metz mecablitz 24AF1 O/P/L flash unit, which arrived late yesterday. But after about 20 fires (as I was finding out exactly how it worked with various mode settings), the trigger system died!

A new one is already on the way from Amazonia (went to UPS a little after midnight last night!).

It is a very nice unit, reasonable in size as a companion to the DMS-FZ200.

DMC-FZ200_F40582-01-C1-S800.jpg


Panasonic Lumix DMC-FZ200 with Metz mecablitz 24AF-1 O/P/L​

It has a nice tilt system for bounce, not (in the usual way) a big head that tilts but rather the "face" tilts. There are several detetented positions. There is no swivel capability.

DMC-FZ200_F40582-01-C1-S800.jpg


Panasonic Lumix DMC-FZ200 with Metz mecablitz 24AF-1 O/P/L​

I'll have more on this flash unit after one that works arrives (scheduled for tomorrow afternoon).

Best regards,

Doug
 

Doug Kerr

Well-known member
The second figure in the previous post (about the flash unit) linked to the incorrect image.

Here is the intended image (showing the "bounce" position of the flash unit face):

DMC-FZ200_F40581-01-C1-S800.jpg


Panasonic Lumix DMC-FZ200 with Metz mecablitz 24AF-1 O/P/L (90° tilt)​

Best regards,

Doug
 

Doug Kerr

Well-known member
Until recently, I had not paid very much attention to the impact of diffusion on images. But now doing much of my work with a camera with a pixel pitch of 1.54 µm, I figured I had better start paying attention.

On thing I did was shoot my famous set of lockdown shots of my living room at different apertures. Here's the full frame of one of the shots just to show the context (reduced to 800 px wide). This was shot at ISO 400 at f/2.8 (the maximum aperture available on my camera).

Test_P1000890-f2.8-S800.jpg


Doug Kerr: Diffraction test at f/2.8​

I know that brickwork as a "sharpness" test is pretty cliché, but that's what I had. Here, it is the near face of the room divider toward the left of the frame. These are 380 px × 300 px crops at original camera resolution. To allow best comparison, I have set them in a tight quad without intervening labels. The apertures were as follows for the four images:

f/2.8 f.4.0

f/5.6 f/8.0

Test_P1000890-f2.8-C1.jpg
Test_P1000892-f4.0-C1.jpg


Test_P1000894-f5.6-C1.jpg
Test_P1000895-f8.0-C1.jpg

Subjectively, I would say there is no systematic decline in the result from f/2.8 to f/4.0, a barely noticeable change from f/2.8 to f/5.6, and a fairly noticeable change from f.28 to f/8.0.

At f/8.0. the 'first minimum" diameter of the Airy figure from diffraction (at a wavelength of 550 nm) would have a diameter of 1.5 µm, almost exactly the pixel pitch.

[continued]
 

Doug Kerr

Well-known member
[Part 2]

Here, we will see, in each case reduced to 800 px wide, the frames for f/2.8 (the one first seen above) and the frame for f/8.0.

Test_P1000890-f2.8-S800.jpg

Doug Kerr: Diffraction test at f/2.8
Test_P1000895-f8.0-S800.jpg

Doug Kerr: Diffraction test at f/8.0​

Of course, here the pixel pitch is 5.6 µm.

Very interesting.

Best regards,

Doug
 

Asher Kelman

OPF Owner/Editor-in-Chief
Doug,

This is obviously a very capable camera with its long zoom lens. For much artistic work, allocating "importance" or "rank" is also important. So what about wide aperture shots to achieve limited depth of focus?

Asher
 

Doug Kerr

Well-known member
Hi, Asher,
This is obviously a very capable camera with its long zoom lens. For much artistic work, allocating "importance" or "rank" is also important. So what about wide aperture shots to achieve limited depth of focus?
An important topic. I'll do something on that.

Thanks.

Best regards,

Doug
 
Subjectively, I would say there is no systematic decline in the result from f/2.8 to f/4.0, a barely noticeable change from f/2.8 to f/5.6, and a fairly noticeable change from f.28 to f/8.0.

At f/8.0. the 'first minimum" diameter of the Airy figure from diffraction (at a wavelength of 550 nm) would have a diameter of 1.5 µm, almost exactly the pixel pitch.

Hi Doug,

2.44 * 0.550 * 8.0 = 10.736 µm, which is more than 7x the sensel pitch.

Of course the relative decline of resolution, which first manifests itself at the highest spatial frequencies by loss of micro detail contrast, also assumes perfect focus and a decently corrected lens. If some defocus or lens aberrations are included, then the relative decline is smaller, but the combined loss is greater.

For a determination of the absolute limiting resolution, one could shoot a print of my resolution test target.

The absolute diffraction limit of 550nm light at f/8 is at 1/(0.000550*8.0) = 227 cycles/mm, where the Nyquist frequency of the sensor is something like 333 cycles/mm, so diffraction at f/8 reduces resolution to roughly 2/3rd of the theoretical maximum resolution.

Cheers,
Bart
 

Doug Kerr

Well-known member
Hi, Bart,

Thanks for this concise analysis.
2.44 * 0.550 * 8.0 = 10.736 µm, which is more than 7x the sensel pitch.

Of course the relative decline of resolution, which first manifests itself at the highest spatial frequencies by loss of micro detail contrast, also assumes perfect focus and a decently corrected lens. If some defocus or lens aberrations are included, then the relative decline is smaller, but the combined loss is greater.
Very well said.

For a determination of the absolute limiting resolution, one could shoot a print of my resolution test target.
Yes, I will probably do that when I get as chance.

The absolute diffraction limit of 550nm light at f/8 is at 1/(0.000550*8.0) = 227 cycles/mm, where the Nyquist frequency of the sensor is something like 333 cycles/mm, so diffraction at f/8 reduces resolution to roughly 2/3rd of the theoretical maximum resolution.

Well, if both blurs were treated as Gaussian, we might expect the joint resolution to be about 188 cy/mm (56% of the sensor resolution).

Makes sense that by that point we would really begin to notice the decline in sharpness.

Thanks.

Best regards,

Doug
 

Doug Kerr

Well-known member
Asher has raised the matter of how the use of a "smaller-sensor" camera influences our ability to provide separation between our "major subject" and foreground or background objects through "cultivated blurring".

I will begin by discussing some background of the matter. Later, I hope to present some example images

Aspects of cultivated blurring

The matter has two main aspects:

• The "magnitude" of the blurring.

• The "nature" of the blurring. (Often this is said to be its "quality", but that term can be mistakenly thought to imply a scale of "worthiness", so I avoid it.)

Regarding the nature of the blurring, the principle attributes of interest are:

1. The radial distribution of luminance across the "blur figure" that is produced from a point source. That has two sub-aspects:

1.1 The "gross" distribution across the figure
1.2 The details of the distribution in the "border" region

2. The presence and nature of any chromatic inconsistencies, such as "color fringes".

But here I will not speak at all of the "nature" aspect, as it is not directly relatable to sensor size.

The magnitude aspect is realistically quantified in terms of the "diameter" of the blur figure. But the impact of the magnitude is perhaps most realistically quantified in terms of the diameter of the blur figure as a fraction of the overall sensor size (perhaps diagonal dimension).

So let's look into that.

The framework of comparison

Just as in the related, but separate, matter of depth of field, if we are to look into what happens when we change a parameter of our "setup" (e.g., sensor size), we must say "all other factors being equal", and then decide just what that means.

One set of definitions for that is:

• The focus distance of the camera is the same in the two cases to be compared

• The focal, lengths produce identical field of view.

• The aperture (as an f-number) is the same.

• We consider the magnitude of the blurring for an "out of focus" object at the same distance from the camera.

A full-frame 35-mm size sensor camera

We will begin by considering a camera with a 43 mm sensor (full-frame 35-mm size), focal length of 43 mm (said by some to be the "normal" focal length for that format, so let's let them have their day today) and an aperture of f/2.8 (the largest aperture I have on my smaller-sensor camera, so I will want to use that for actual testing with it). We will assume a subject at a distance of 3 m, and the camera focused on that subject. We will consider a "background" object at the times that distance, 30 m.

Then, the diameter of the blur circle (using the "classical" focus equations) will be 0.20 mm, or 0.0047 of the sensor size (about 1/213) (and thus 1/213 of the image size).

In depth of field reckoning using the "human visual acuity" outlook, we often look at blurring of 1/1400 of the image size as being "visually significant". Here, the blurring is about 6 times "as great" as that.

A smaller-sensor camera

Next we consider a camera with a 7.7 mm sensor (as found on my new Panasonic Lumix DMC-FZ200). We will assume a focal length of 7.7 mm, yielding a comparable field of view to that in our first example, and an aperture of f/2.8 (the largest available on that machine). Again, we assume focus on an object at a distance of 3 m, and consider the blurring on a "background" object at a distance of 30 m.

Then, the diameter of the blur circle (using the "classical" focus equations) will be 0.006 mm, or 0.00078 of the sensor size (about 1/1280) (and thus 1/1280 of the image size).

Here, the blurring is only about 1.2 times "as great" as the blurring we assume to be "significant" when working with depth of field.

Thus, it is unlikely that this result would be "satisfying" with respect to separation of the subject from "background" objects.

We note that with this ratio of distance to the out-of-focus object to the focus distance (10:1), the diameter of the blur figure, expressed as a fraction of the image, varies approximately proportionally to the sensor size.

Thus, as sensor size decreases, we definitely lose "leverage" in our quest to cultivate the blurring of foreground or background objects.

Best regards,

Doug
 

Doug Kerr

Well-known member
Her we have a chance to see the implications of sensor size on our ability to cultivate blur.

The scene was not ideal (it is always nice to have daisies as the background objects) but it was done in or living room (which at the high-elevation World Headquarters has to serve as our studio and indoor test range).

All images are full-frame, ex camera except for being downsized to 800 px wide for presentation here.

We start with a 27 mm sensor, on our Canon EOS 40D. The lens was a 50 mm f/1.4.

The subject (our mannikin Stella) was at a distance of 1.02 m. The background (a curio cabinet and some wall decor) was at a distance of 19.1 m.

We start with an aperture of f/16 to help show the context:

Test_OFB_F40588-f16.0-S800.jpg


Douglas A. Kerr: Stella, 27 mm sensor, f/16

Now we jump to f/2.8, the largest aperture we have on both cameras:

Test_OFB_F40589-f02.8-S800.jpg


Douglas A. Kerr: Stella, 27 mm sensor, f/2.8

Finally, we go to our largest available aperture, f/1.4:

Test_OFB_F40590-f01.4-S800.jpg


Douglas A. Kerr: Stella, 27 mm sensor, f/1.4

In the next stanza, we will see what happens with the 7.7 mm sensor on our Panasonic Lumix DMc-FZ200.

[continued]
 

Doug Kerr

Well-known member
[Part 2]

Now we will see what happens with our Panasonic Lumix DMC-SZ200 camera, with a 7.7 mm sensor.

Here, I chose a focal length (15 mm) that gave a horizontal field of view very comparable to that of the images from the EOS 40D (the two cameras have different aspect ratios, so there is no direct equivalent).

The distances to the mannikin and the background were the same as with the 40D.

We start at an aperture of f/8.0 (the smallest available) to show the context:

Test_OFB_P1000898-f08.0-S800.jpg


Douglas A. Kerr: Stella, 7.7 mm sensor, f/8.0

Now, we use f.4.0:

Test_OFB_P1000901-f04.0-S800.jpg


Douglas A. Kerr: Stella, 7.7 mm sensor, f/4.0

Finally, our best shot for this form of combat, f/2.8

Test_OFB_P1000902-f02.8-S800.jpg


Douglas A. Kerr: Stella, 7.7 mm sensor, f/2.8

As you can see, the cultivated blur is modest in magnitude. But not wholly ineffectual.

Best regards,

Doug
 

Asher Kelman

OPF Owner/Editor-in-Chief
Finally, we go to our largest available aperture, f/1.4:

Test_OFB_F40590-f01.4-S800.jpg


Douglas A. Kerr: Stella, 27 mm sensor, f/1.4



Here, Doug, one can see that APS-C sensors have the sufficient artistic blur capability to isolate a subject to the extent one desires in composing a picture. This is why there's such an acceptance of this format for DSLR's and now compact pro-level "pocketable" cameras, the NEX line being the most versatile.

Asher
 

Doug Kerr

Well-known member
Hi, Asher,

Here, Doug, one can see that APS-C sensors have the sufficient artistic blur capability to isolate a subject to the extent one desires in composing a picture. This is why there's such an acceptance of this format for DSLR's and now compact pro-level "pocketable" cameras, the NEX line being the most versatile.

Yes indeed. The 27 mm format is die neue Kleinbild.

I would love to have gone to a "compact" 27-mm sensor camera for our "B" camera, but at the moment there is no really good choice for a zoom lens with the range of reach we feel we need.

For example, I would have had a hard time doing this with a 28 mm ff35 equivalent focal length:

Relay_P1000513-01-C1-S800.jpg


Douglas A. Kerr: Carla proudly brings up the rear

This was with a ff35 equivalent focal length of 385 mm! It was shot from the opposite side of a football field.

This was the opening lap of a charity relay "race" to raise funds for the American Cancer Society. This lap featured all the actual cancer survivors in attendance. The squad was led by the most-recent survivor, and the others were arrayed in terms of their longevity, in groups each preceded by a "page" carrying a sign with the number of years (by decades after the first few). The longest period for which there was a sign was 30 years, and Carla was the only occupant of that cohort, actually 52 years after clearing from her first bout (and 7 years after clearing her third bout).​
In any case, I'm sure this isn't our last machine!

Best regards,

Doug
 

Doug Kerr

Well-known member
The saga continues.

I had reported on the grossly-anomalous behavior and then infant mortality of the Metz Mecablitz 24 A-1 (O/P/L) flash unit.

The replacement arrived timely.

I used it for the charity dinner shoot I mentioned earlier. In about one out of 5 or 10 shots, the flash fired "wide" open, no metering apparently being effective. The result was of course something like a full white frame.

We wrestled with this for a while, and concluded that either:

• The camera was defective.

• The camera was anomalous in a way that the Metz flash could not abide but that perhaps the recommended Panasonic flash would.

• The camera was fine but there was just some problem with the Metz effort to practice the Panasonic protocol. (I think that the Metz is actually attuned to the Olympus protocol, and the Panasonic protocol is believed to be "identical".)

We ordered the smallest of the Panasonic flash units, the Lumix DMW-FL220. Its ISO guide number* is 22 m.

*The ISO guide number (that is, the guide number as defined by the pertinent ISO standard) is by definition predicated on an ISO speed of ISO 100.​
It did not exhibit that type of misfire after over 500 test shots.

It is nicely built. It does not have a tilt head for bounce use, but I don't think that is really much of a problem to us. Most of our indoor flash work is in restaurants and the like where typically the ceiling is flat black and/or very high.

However, I find that the flash metering of the combo is very "irregular". For example, in a lockdown shot of a certain test subject, the delivered exposures were noticeably different over various ISO sensitivity settings of the camera (that is, the metering system seemingly does not "track" properly with ISO sensitivity).

I have no idea how the camera's TTL flash metering system works, but the exposure results are rather unpredictable. The resulting exposure for shots of subjects fairly near are consistently different from those father away.

In any case, for shots of "ordinary" indoor scene situations, I need to use a flash exposure compensation of +1/3 stop to +2/3 stop (mostly depending on distance) to get a result comparable to that with ambient illumination.

Of course, I have been used to the Canon ETTL-II system on an EOS body which, with all its warts, generally produces pretty consistent results.

But what it is is what it is.

We still like the camera a lot (warts and all).

Best regards,

Doug
 
Top