Menneisyys

Lives in United States United States
Joined on Sep 28, 2007

Comments

Total: 3569, showing: 1 – 20
« First‹ Previous12345Next ›Last »
In reply to:

Kimballcorson: For all those fighting the dpreview conclusions on FZ2000 sharpness visa vi the RX10M3, compare the Imatest resolution graphs for the two cameras found on Photoreview out of Australia (.com.au) which very conspicuously shows the same conclusion. The issue really isn't close. You might also check out the dpreview studio and long range resolution comparisons which show the same thing. Such rationalizations.

And traumflieger also had exactly the same conclusion (see my posts below).

Link | Posted on Jan 26, 2017 at 17:03 UTC
In reply to:

Joseph Weissman: Joe Weissman My first impressions with th FZ2500 with actual subjects and not charts were very positive. I am mostly interested in the long focal length performance.

I was concerned when I read the review in dpreview and so I ran my own comparisons with the USAF test chart. I compared with a Canon 400 mm F5.6 on an 80D (unfair, I know).

Procedure:

indoor distance of 20 feet,
tripod,
shutter delay,
ISO400,
manual focus/ live zoom.
raw converted DNG
view/PS.

The canon prime is sharper but not by more than 2 steps on the chart, a 10-20% difference. Both images were in the 50% range of the Nyquist freq.
70% to the edge og FZ2500 sensor at 480mm , it is 15% less sharp.
There is greater loss with jpeg in the Lumix than the Canon.

My conclusion is that the Panasonic lens is pretty sharp. I am wary of comparisons with jpeg or with outdoor shots during full daylight at long focal lengths. Heat shimmer varies between shots, also.

I'll post my images when i get time.

@marshwader: "it's also possible that dpreview just ended up with two bad copies for one reason or another"

I don't think so - this is a generic problem also emphasized by other reviews and comparisons. The Traumflieger folks, in their December review, have also concluded exactly the same regarding lens quality and its relative sharpness compared to both the FZ1000 and the RX10MkIII. That is, unfortunately, the FZ2000 indeed has an inherently soft lens.

Here's the review URL (German):
http://www.traumflieger.de/reports/Kamera-Tests/Panasonic-Olympus-Sony-Samsun/Panasonic-Lumix-FZ2000/Panasonic-Lumix-FZ2000-im-grossen-Traumflieger-Test::1233.html?XTCsid=9e83954328c6ae34baebea174918b39b

Link | Posted on Jan 22, 2017 at 09:59 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

nerd2: Stereo craze has passed by decade ago....we already have cheap rgbd sensors

@otto k: "On the other hand, I had no issues with LG G4 vs Note 5 in simple Fiit VR (aside from the obvious AMOLED vs LCD differences)." - the official Oculus support puts Samsung's VR implementation in a whole new dimension, compared to anything else mobile. We have a playstore with some excellent games - for example, GUNJACK, an excellent title where the, compared to desktop VR, inherently lower CPU / GPU performance don't show. We have a full-fledged game selector menu with built-in image / video viewers, Web browsers, Settings accessors etc.

All these are missing from other (non-Samsung VR-compliant) handsets.

Link | Posted on Jan 22, 2017 at 07:14 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

nerd2: Stereo craze has passed by decade ago....we already have cheap rgbd sensors

@otto k: you meant my post on the Samsung? That is, the lack of my mentioning Sony's existing 4K phones? There's a reason for that:

- complete lack of dedicated (with Samsung, even official Oculus!) support
- generally worse-quality handsets than those of Samsung - for example, the IQ of the camera.

This is why I don't even bother mentioning Sony when it comes to mobile VR.

Link | Posted on Jan 21, 2017 at 21:55 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

falconeyes: My concern with 4k VR is that resolution creates a toy-like experience.
Tried it, after first curiosity is satisfied, becomes boring very quickly. Won't take off commercially.
4k VR leaves 420px per 50mm-lens equivalent viewing width. That's very pixelated and requires 8k to be pleasurable. VR will take off with 8k though (840px or DVD like experience).
Moreover, 8k VR headsets (100 degrees FoV optics) require 5k smartphone screens (or a corresponding dedicated VR headset) to display. Both is a decade away for mass market penetration.
So, the next 3D hype will be it. Not this one. Or the one before created by the Avatar movie.

P.S.
Small 180 degree fisheye lenses resolving 4k tack-sharp edge to edge (which 8k VR requires) will be another challenge ...

"So, it would make a difference for 4k 3D content (non VR) but that line of development has silently died for some reason (motion sickness?)." - you mean the demise of traditional 3D TV's? The reasons are well-known: http://www.dailymail.co.uk/news/article-2765862/Why-3D-TV-proved-turn-Massive-investment-1-5-million-sets-sold-persuade-great-British-public-wear-silly-specs.html

I, who have watched one or two feature movies entirely in 3D on my Samsung S7 Edge + Gear VR and also tested 3D video playback on the HTC Vive (that is, using VR helmets), personally liked what I experienced. No motion sickness, no headaches. It's "only" the resolution that was problematic; this is why I stated 4k (and, in the more distant future, 6/8k) will be a big step ahead from the current 2...2.5k screens.

Link | Posted on Jan 21, 2017 at 13:46 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

nerd2: Stereo craze has passed by decade ago....we already have cheap rgbd sensors

"VR 3D leaves nothing but your jaw dropping" - definitely. As soon as (hopefully in H1 (via the Samsung S8) of this year) the resolution of the screen is bumped to at least 4K, meaning 1920*2160 for each eye, they'll become a great way to watch 3D videos / movies.

Link | Posted on Jan 21, 2017 at 10:55 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

falconeyes: On the Kickstarter page, TwoEyes claims only them are doing 3D movie and Vuze does not. According to Vuze and early CES testers, Vuze does true 4k 3D VR 30fps video too.
Is this false advertizing?
After all, Vuze and TwoEyes do the same at the same price per camera and makes obsolete this kickstarter project.

"What I meant is that TwoEyes claims Vuze is NOT doing 3D video!" - that's what I've also referred to in my very first sentence. The Vuze IS a 3D 360 camera. I've checked out their concert video sample (in a VR helmet) and it's of very good quality. And yes, there's 3D feeling in every (horizontal) direction.

Link | Posted on Jan 21, 2017 at 10:53 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

falconeyes: My concern with 4k VR is that resolution creates a toy-like experience.
Tried it, after first curiosity is satisfied, becomes boring very quickly. Won't take off commercially.
4k VR leaves 420px per 50mm-lens equivalent viewing width. That's very pixelated and requires 8k to be pleasurable. VR will take off with 8k though (840px or DVD like experience).
Moreover, 8k VR headsets (100 degrees FoV optics) require 5k smartphone screens (or a corresponding dedicated VR headset) to display. Both is a decade away for mass market penetration.
So, the next 3D hype will be it. Not this one. Or the one before created by the Avatar movie.

P.S.
Small 180 degree fisheye lenses resolving 4k tack-sharp edge to edge (which 8k VR requires) will be another challenge ...

Hopefully Samsung's new high-end models (to be launched in a few months - S8 etc.) will have 4K screens. Then, for stereo and/or 360 video consumption, they'd be passable.

Link | Posted on Jan 21, 2017 at 10:48 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

nerd2: Stereo craze has passed by decade ago....we already have cheap rgbd sensors

I really think you should give a try to some 3D 360 videos in a competent VR headset. They're absolutely great.

Link | Posted on Jan 21, 2017 at 07:22 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

falconeyes: On the Kickstarter page, TwoEyes claims only them are doing 3D movie and Vuze does not. According to Vuze and early CES testers, Vuze does true 4k 3D VR 30fps video too.
Is this false advertizing?
After all, Vuze and TwoEyes do the same at the same price per camera and makes obsolete this kickstarter project.

"On the Kickstarter page, TwoEyes claims only them are doing 3D movie and Vuze does not. ... Is this false advertizing?"

Sure it is. Physically, the Vuze is capable of far better (far more immersive) 3D 360 video as it has 3D peaks at 0, 90, 180 and 270 degrees, as opposed to the 0 and 180 degrees only with the TwoEyes.

Link | Posted on Jan 21, 2017 at 07:21 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

winkalman: I think they're about 7 years late...

https://www.engadget.com/2009/01/05/minoru-3d-webcam-ships-this-week-still-looks-freaky/

@SimenO1: "I hope Ricoh will launch a dual Theta that can replace it": all Thetas have very bad image quality compared to the alternatives. For example, the Samsung Gear 360 produces WAY better image/video.

I don't think Ricoh will fix this problem. Even their latest models, released way after the Gear 360, have as cr@ppy IQ as their previous ones.

Link | Posted on Jan 21, 2017 at 06:36 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

Turbguy1: ALL of these stereoscopic "solutions" require that you hold your gaze level (parallel with the recording interaxial axis). NONE of these will not provide true stereo if you hold your gaze at a significant angle (tilt your head to your shoulders) because that would require an almost infinite amount of data for true stereo VR imaging of a REAL scene at those (and all other) angles of gaze.

While you're theoretically right, 3D 360 video recorded with current hardware (even including the Vuze demo clips) are SIGNIFICANTLY better than 2D 360 videos recoeded by, say, the Theta or the Gear 360.

I've played around a lot with 3D 360 video using proper display (both the Gear VR and the HTC Vive) and I can assure you even the Vuze produces some absolutely astonishing 3D footage much-much better than anything 2D.

Link | Posted on Jan 21, 2017 at 06:34 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

Menneisyys: 360-degree 3D video with two cameras, even when they're spaced quite far from each other?

This won't produce very good 3D video. You'll only have true 3D vision right in front of you (and on the back). On the sides, you'll have much less or even none. In this regards, other solutions using much more cameras (e.g., the Vuze camera) are really superior.

"preorder from B&H"? That sounds good - I wouldn't pre-order directly (providing my credit card info and stuff) from a relatively unknown company but B&H is a safe bet.

Link | Posted on Jan 20, 2017 at 21:05 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

Ace of Sevens: How is this going to work? It would give you a stereoscopic view when facing the same direction as the cameras, but I don't see how it gives you the right data to construct a stereoscopic view any other direction.

@Artak Hambarian: there's at least the promise of the Vuze camera for $799 / 899eur ( http://vuze.camera/product/ ). I'm not sure whether they'll indeed be able to start shipping it in a few months, though - there have already been some long delays.

And, of course, there are the professional (much-much more expensive) stuff: the Facebook Surround 360 Camera ( https://facebook360.fb.com/facebook-surround-360/ ), the PanoCam 3D ( http://www.panocam3d.com/camera.html ) and the various GoPro-based solutions.

Link | Posted on Jan 20, 2017 at 20:32 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

Menneisyys: 360-degree 3D video with two cameras, even when they're spaced quite far from each other?

This won't produce very good 3D video. You'll only have true 3D vision right in front of you (and on the back). On the sides, you'll have much less or even none. In this regards, other solutions using much more cameras (e.g., the Vuze camera) are really superior.

@Artak Hambarian - " right in front of you, but very 3D natural": yup, it's natural. However, if you look around, the 3D feel will be reduced to zero (at 90 / 270 degrees). This only allows for very specialized shooting - for example, if you put your subjects right in front of you (and in the opposite direction). Forget placing close subjects on the sides.

Assuming the Vuze camera is indeed released (I'm VERY much eagerly waiting for it!), this camera won't have much chance, unless it's really cheap.

Link | Posted on Jan 20, 2017 at 20:19 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

winkalman: I think they're about 7 years late...

https://www.engadget.com/2009/01/05/minoru-3d-webcam-ships-this-week-still-looks-freaky/

I'm not sure if you're joking... the web camera you linked to is a plain non-360 (non-VR) solution. There have been tons of such gadgets and even P&S cameras like Fujifilm's venerable and, with 3D shooters, still very popular FinePix Real 3D W-series. The first model (W1) of which, BTW, was released almost exactly 8 years ago.

Link | Posted on Jan 20, 2017 at 20:16 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)
In reply to:

Ace of Sevens: How is this going to work? It would give you a stereoscopic view when facing the same direction as the cameras, but I don't see how it gives you the right data to construct a stereoscopic view any other direction.

It won't - it's physically impossible. This is just a cheap semi-3D alternative for folks that can not afford truly 3D 360 cameras. (The latter start at $1000.)

Link | Posted on Jan 20, 2017 at 18:54 UTC
On article TwoEyes VR stereoscopic camera simulates human vision (52 comments in total)

360-degree 3D video with two cameras, even when they're spaced quite far from each other?

This won't produce very good 3D video. You'll only have true 3D vision right in front of you (and on the back). On the sides, you'll have much less or even none. In this regards, other solutions using much more cameras (e.g., the Vuze camera) are really superior.

Link | Posted on Jan 20, 2017 at 18:53 UTC as 11th comment | 7 replies

BTW, the Traumflieger folks, in their December review, have concluded exactly the same regarding lens quality and its relative sharpness compared to both the FZ1000 and the RX10MkIII. That is, unfortunately, the FZ2000 indeed has an inherently soft lens.

Here's the review URL (German):
http://www.traumflieger.de/reports/Kamera-Tests/Panasonic-Olympus-Sony-Samsun/Panasonic-Lumix-FZ2000/Panasonic-Lumix-FZ2000-im-grossen-Traumflieger-Test::1233.html?XTCsid=9e83954328c6ae34baebea174918b39b

Link | Posted on Jan 18, 2017 at 19:33 UTC as 30th comment
On article Elevating X-Trans? Fujifilm X-T2 Review (2189 comments in total)

"Compared with the a6300 which uses a very similar sensor, you can see similar luminance noise levels but lower levels of chroma noise - suggesting it's being suppressed somewhere along the line." - as has been discussed numerous times, color noise is inherently lower with X-Trans. And this is NOT caused by any kind of RAW baking.

Link | Posted on Jan 14, 2017 at 21:23 UTC as 23rd comment | 1 reply
Total: 3569, showing: 1 – 20
« First‹ Previous12345Next ›Last »