On the way from FX7 to AX2000E

HDR-AX2000 / HXR-NX5 (2010).
Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 01 Jun 2010 15:07

Really? I never noticed a thing.

I tell you more, after experienced that resolution fall with the new AX2000 I checked my old screenshots taken from various footages made with the FX7. Sharpness wise, whatever the aperture is, they all are alike.

Perhaps, it was something that hindered me to observe loss in detail (or just eclipsed it).

User avatar
Stephan
Site Admin
Posts: 592
Joined: 20 Mar 2010 18:51
Location: Paris, France

Re: On the way from FX7 to AX2000E

Postby Stephan » 01 Jun 2010 22:31

Arkady Bolotin wrote:Therefore, it can be readily assumed that in their new cameras Sony finally shifted the default sharpness level toward the soft end. The question is why they did this.
I second that assumption.

I can imagine a reason for this: remember the wild debate about 25p with the HVR-V1 - 'marching ants', 'flying mosquitoes'? Sony being called names... because people hadn't yet figured out that, the same as you need a vertical low-pass filter with interlaced footage to prevent flicker, you also need a low-pass filter with progressive to prevent aliasing. You cannot really expect progressive footage to give you more resolution while avoiding aliasing. So it came out that people had to turn down sharpness a couple of notches on the V1 to be able to shoot 25p without all the 'marching ants'.

Maybe Sony didn't want to stir the same debate, esp. if the AX2000 is targeted at unaware (upscale) consumers, so they decreased sharpness upfront by default out of the factory. Just speculating.

Arkady Bolotin wrote:I always knew that closing the camera’s aperture could cause loss of picture detail, but with my FX7, I never observed it. I freely changed the aperture usually keeping it at f5.6 (under strong light) and never saw any perceptible resolution fall.
Same here.

Steve, you're absolutely correct. With 1/3'' sensors, the AX2000 should be less prone to diffraction.

And yet, I experienced the same thing as Arkady with the FX7. I wanted to check how far I could close the iris on my FX7 until I ended up in trouble with blurred detail. I remember that user 'megabit' (Piotr) showed some footage that seemingly exhibited diffraction blur, but I performed a full test and was not able to reproduce the issue even at f8.

Arkady Bolotin wrote:No such luck with AX2000… Yesterday’s morning I recorded a couple of shots with two different settings: iris = f4 and shutter = 1/50p against iris = f5.6 and shutter = 1/25p. To my surprise, the pictures produced with the aperture at f4 were noticeably sharper...

I wonder, does this mean the AX2000 can be a school exhibit demonstrating physic phenomenon of diffraction? On the other hand, maybe, the shutter speed has something to do with it.
Shutter speed, probably not. Never heard or read of such thing.

How did you take care of focus? Rule #1 I painfully learned with my FX7: never, ever, change zoom factor or even maybe iris settings while in manual focus. In all those so-called 'varifocal lenses', the focus plane shifts when you change optical settings, and you basically lose focus. This is supposedly taken care of by automatic behind-the-scenes back-focus adjustment (as I think I understood the issue), and sometimes it just doesn't work so well.

Did you make that test staying in autofocus?

Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 02 Jun 2010 16:27

Stephan,

Your explanation on the subject of lesser sharpness (detail) midpoint level in Sony new camera AX2000 is logic and sound. Well, it seems very believable that to avoid all that fuss about aliasing they (Sony) came out with the simple anti-aliased technique: turn the sharpness down. Ha!

Regarding the aperture values and their impact on the AX2000 pictures, I did my tests with autofocus entirely (making exception, of course, for shooting resolution charts).

I thought that some softness in my clips I mentioned in the previous post could be attributed to the difference in bit-rate during different recording conditions (if in the FX mode the maximum bit-rate is 24 Mb per second, it means that it could be much lower in some scenes, right?).

Then I did the tests again keeping now the iris at f4 and compensating by faster shutter 1/50 (25 fps). The pictures appeared pin-sharp, almost touchable. When I was watching them last night, a little bug flew in through the open window and landed on my TV screen; so, it took me a while to realize it was real and behind the TV glass and not inside the picture. It was an almost surrealistic sensation to see it…

By the way, as Adam Witt assessed, the new Sony G lens has the sweet spot at f2.8, therefore one should maintain aperture wider than f4.8 to preserve the picture detail level. Frankly, I never paid much attention to such statements, now I’ve learned my lesson…

I made another test to compare the resolution between FX7 and AX2000 under normal light condition. Actually, I shot ISO-12233 resolution chart beamed on my PC LCD monitor (24', 350 lumens). Calculating the area (the display screen square, 0.58 x 0.21 m) over which the luminous flux is spread (i.e. 350 lumens) gives us luminance of the chart: 350/(0.58x0.21) = 2900 lux, which is approx. corresponds to overcast day light condition.

The fragments of the chart I recorder are following, please observe. In the pictures made from AX2000 footage you can see grains of the computer screen matrix.
Attachments
AX2000 Test 2.jpg
AX2000
AX2000 Test 2.jpg (47.8 KiB) Viewed 5409 times
FX7 Test 2.jpg
FX7
FX7 Test 2.jpg (46.49 KiB) Viewed 5409 times

Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 02 Jun 2010 16:30

And another couple of the fragments...
Attachments
AX2000 Test 1.jpg
AX2000
AX2000 Test 1.jpg (41.67 KiB) Viewed 5409 times
FX7 Test 1.jpg
FX7
FX7 Test 1.jpg (47.6 KiB) Viewed 5409 times

User avatar
Stephan
Site Admin
Posts: 592
Joined: 20 Mar 2010 18:51
Location: Paris, France

Re: On the way from FX7 to AX2000E

Postby Stephan » 02 Jun 2010 23:58

Interesting. Being careful about iris aperture absolutely makes sense, and we'll never know why the FX7 was so tolerant about it (or why maybe we both failed with our tests at that time). Anyway, what's important is that you found the right setting for your AX2000 now. Enjoy!

Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 03 Jun 2010 17:50

I did some tests of the dynamic range of the camcorders AX2000 vs. FX7.

Sure, it is not easy task to do, and honestly speaking without suitable measurement devices is quite difficult to compare the limits of luminance range which two different cameras can capture.

Therefore, for a refer point in such comparison I decide to take 100% zebra pattern so that recording the same scene with both cameras, I preserved the same amount of the strip pattern displayed over the same parts of the image on the cam’s LCD.

Please look at the screenshots I downloaded.

In both camcorders, the strip pattern was displayed over the stonework of the metal fence indicating that this part of the image was higher than the preset 100% brightness level.

As you can see, in this situation the FX7 recorded overexposed picture, whilst the AX2000’ picture has no clipping of any kind.
Attachments
AX2000.jpg
(62.41 KiB) Not downloaded yet
FX7.jpg
(71.37 KiB) Not downloaded yet

User avatar
Doughie
Global Moderator
Posts: 452
Joined: 22 May 2010 16:57
Location: Mexico

Re: On the way from FX7 to AX2000E

Postby Doughie » 05 Jun 2010 17:29

thanks for doing that Arkady - this is all good information and for sure interesting comparing the latest-generation of HD Sony machines (AX2000 running AVCHD codec) and a previous-gen cam like the FX7 running HDV codec.

Appreciate all your work ! keep going!

Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 06 Jun 2010 13:20

Thank you Doughie…

What the new camera checks can do without the famous (or infamous) “Burning candle” test and shallow focus pictures?

So, I did them all, here are some screenshots, see for yourselves.
Attachments
Shallow Focus.jpg
AX2000's Shallow Focus (distance to the nearest bush branch ~ 2 m, the aperture =f/3.4, shutter speed = 50)
(91.89 KiB) Not downloaded yet
Candle (overhead light).jpg
AX2000: Picture taken under a single overhead light 60W
(64.88 KiB) Not downloaded yet
Candle.jpg
AX2000: Picture taken under the light with a luminous intensity of roughly 1 candela
(69.91 KiB) Not downloaded yet

Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 09 Jun 2010 19:14

Actually, I have never been fond of night shooting (probably knowing that no consumer camera could ever record subjects clearly in light as low as ordinary nighttime street lighting). Sure, I had a couple of happenstances when I had some acceptable night shots of Virgin Megastore and Kodak Theater in Hollywood, but that place was lit brighter than a photography studio.

So, yesterday’s night I brought my new AX2000 to the middle of a mundane town street lighted only by commonplace lamp posts.

As I found, shooting at night was not that easy.

First, focusing, I did not trust autofocus; therefore, I switched focus into manual mode. However, it was so dark that sometimes I could barely see what I was focusing.

Second, street cats (Beersheba has a lot of them), they probably decided that I was there to feed them (good thinking: otherwise what could a human possibly do in the street in 2 o’clock in the morning?) and surround me disturbing my cinematographic activity asking for food and some sympathy.

Nevertheless, the footage was worth having: what I finally got was above all my expectations! Look at picture I’ve uploaded, of course, JPEG compressing introduced lots of artifacts, but still, you may have a taste of the AX2000 low-light capabilities.
Attachments
Night street (2).jpg
Iris = f2, shutter speed = 1/25(p), gain = -3db
(93.86 KiB) Not downloaded yet
Night-street-(1).jpg
Iris = f1.7, shutter speed = 1/25(p), gain = -3db
(88.78 KiB) Not downloaded yet

Arkady Bolotin
Posts: 60
Joined: 24 May 2010 16:46
Location: Beersheba, Israel

Re: On the way from FX7 to AX2000E

Postby Arkady Bolotin » 11 Jun 2010 16:07

Yesterday I tested `back focus’ on my camcorders.

I hanged up a piece of newsmagazine on my kitchen window and taking a camera on a tripod from the distance of about 4 meters I moved the camera zoom ring towards telephoto end, adjusted focus manually, and then moved zoom towards wide angle. After making a shot, I switched the focus back to auto mode and made another shot.

Having transferred the media files to a PC I compared them side-to-side: I have the results I anticipated. My FX7 has a severe `back focus’ issue (I knew it long time ago but never had the guts to confirm that in a test), but the AX2000 did not flinch: both clips (manually adjusted at telephoto and autofocused) came out identical.
Attachments
Auto-focus.jpg
FX7 autofocus mode
(44.71 KiB) Not downloaded yet
Manual-focus.jpg
FX7 manual focus
(37.66 KiB) Not downloaded yet


Return to “Sony AVCHD - High-end / 3-sensor camcorders - NXCAM”