Camera Video Display Problems Tested

Published Mar 18, 2016 04:00 AM

** *** *** ****, ********** ***** was ******. **** ******* *** ******** were ********* *** **** ****** *****, so ** *** ****** ****** ** display ***** ******* *** ********.

***, ******, **** ******* ****** ****** (16:9 ********, *:* *******, **:* *******, fisheye *******, *****-*******, ***.), **** ** becoming **** **** *********.

******, **** *** ******* *** *** defaulting ** ******** ** ******* ***** by *******.

** **** ******, ** **** ** this ******* *** ***** ******** ** how ****, ***** *** ********* ** this.

Cropping *******

***** *** *** *** ******** ****, cropping *** **********.

******* *** *** ****** *****, ****** how **** ** '*****' *** ****, the *** *** ****** *** *******:

**** ********* ***** ***** ** ** action.

Stretching *******

******* *** *** ****** *****, *** narrower *** ** * **:* ****** ratio ****** *** *** *** ******** it ***** / *********:

** *** ********** *****, *** *** see *** *** *** ** **********:

*********

********* *** ****** ****** *** ****** aspect *****. ** *******, ********* *******, DVRs/NVRs, *******-**** ****** ******* ***** ** *** ******, regardless ** ****** *****. *** ****** image ** ******* (****** ****, *****, which ***** ***** ** *** *****), but ********* ** **** *****.

**** *** ** ******* ** ******* aspect *****, *** *** ****** ** buried ** ******** (**** > ****** Configuration > ***** > **** ***** > ******** **********). *** ***** ***** shows *** **** ******* ** *****, all **:*, ** ***** ****** ******.

Axis ****** *******

** ******* ****** ******* * ****, ******* *** *** ** "****" the **** ** *******. **** *********** means **** *** ********** ***** ** the ***** ** *** ** *** width ** *** ******* ****. *******, areas ** *** *** *** ****** of *** *** *** *** ***.

**** *** ** ******* ** ************* to *** *** **** ***** ** the ****, ***** ******** *** **** FOV, *** ******* ** ******* ****** of ***** ***** ** ****** ****, seen *****. *******, ***** **** ******, at ***** *** **** *** ** visible *** ****** ***** ** *********.

*****

** ***** *** **********, ****** ***** is *** *********, ******** ****** ******** defaults ** "******** ****** *****", **** options ** ******* ** ****. **** can ****** ** ******* ********** ********* on *** ****** ****** ***** *** size.

*** *******, *** ********** ***** *** taken **** *** *** ********* ** a***** ** *****. ********** ** *** ****** ******, video ** ***** ********* ** **** the ***** ** *******, ********* ** squashed ** ********* *****. ***** *** controls ** **** ***** ** *** original ******, *** **** *** *** well ******* *** *** ***** *** not ******* **** *** **** *****.

The ********

****** ***** **** ****** ** ********* filled, *** *** *******:

  1. Monitors *** **:*, *** **** ******* *** ***: Though many would like to proclaim HD and 16:9 formats as the way to go, 4:3 models are still commonly used (and arguably provide more useful images), including 3 MP, 5 MP, and now 12 MP resolutions. It will be years before 4:3 is no longer available.
  2. UI ******** **** ** *****: Toolbars, Windows and Mac task bars, event lists, and other UI elements take up screen space, meaning even if the VMS client is run full screen on a 16:9 monitor with 16:9 cameras in a split view, these controls may take up space and skew the aspect ratio of viewing windows.

****** ***, *** ******** ** ** stop ********** ****** ****** *****. ****** some *** ******** **** ******* ******* "prettier", **** ********** ******* **** ****** usability *** ** *** ****** *****. Blank ***** ******* ***** ** * better ******** **** ******* ********* *****.

Comments (7)
U
Undisclosed #1
Mar 18, 2016
IPVMU Certified

Though some may consider full viewing windows "prettier", they ultimately detract from system usability due to the issues above. Blank space between panes is a better solution than viewing distorted video.

I would disagree.

First, I don't consider stretched windows prettier, just the opposite, wide faces, bodies etc.

Second, this is not Avatar in 3D. The primary goal is not photographic realism, it's effective information transfer.

For me, stretching the screen to fit the size makes it easier to see movement when not looking directly at the pane.

In any event it does not mislead me in any way, since I know the true aspect ratio.

On the other hand, the difference/distortion of the color pallete by even cameras of the same mfr/model, is something that could materially affect an investigation. Do you agree about color distortion?

Maybe a poll?

"Do you prefer stretch to fit windows over letterboxing?"

(1)
Avatar
Ryan Hulse
Mar 21, 2016

I can understand the advantage you described of filling all the pixels. I do think the correct aspect ratio does improve identification to some degree. The ability to discern someone's "build", is easier if the aspect ratio is correct.

(1)
(1)
Avatar
Jon Dillabaugh
Mar 18, 2016
Pro Focus LLC

I hate it when windows are stretched and distorted. It's difficult to evaluate things like focus and other criteria. It is one of my pet peaves with Dahua's web interface for IP cams in the live view tab. The first thing I click is the "Original" instead of "Adaptive" aspect ratio.

(1)
(1)
U
Undisclosed #1
Mar 18, 2016
IPVMU Certified

It's difficult to evaluate things like focus...

Absolutely. But is this an installer concern more than a user concern?

For instance given a 1080p 50" that is part of a video wall, which can you discern important details better from across the room with?

This:

or

Things like "is someone at the end of the hall?" are easier to make out because there are more pixels displaying them. The fact that the chairs are really taller than they appear is unimportant IMHO.

And what is the downside to the user exactly?

Unintended scene cropping, on the other hand is an abomination, and deserving of criticism.

(1)
Avatar
Michael von Hauff
Mar 18, 2016

I have a difficult time understanding how stretching the video feed to fit the window size has any impact on security monitoring (or is even worth writing about really). If someone is watching the video feed in realtime (all the time), their brain will adjust what they see. If they aren't watching realtime (all the time), then post incident analysis is the only situation you can assume. In a post incident analysis situation, manipulating the aspect ratio is reasonable if in fact it really bothers the investigator.

In any case, cropping is obviously the worse of the two. Destroying information is worse than transforming information (there is no loss of information when you distort the image to fit the window - you can transform it back).

UM
Undisclosed Manufacturer #2
Mar 18, 2016

I think it depends on who is watching. A security professional wants to keep the true nature of the video and the aspect ratio.

A typical "end user" has been told that black bars are bad. That is why you had to go out and replace your CRT 4:3 TV with a new 16:9 expensive HDTV. They don't want to see "wasted space" on their VMS display. Even if the cameras are higher resolution - 3mp vs 2 mp, they will question it. I think that it takes educating the end user, and ultimately giving them the option in the settings, as long as the true video is still preserved as evidence...

Of course, now movies are in newer aspect ratios, and there are still black bars. And many TV shows are stretched in different amounts from the center out.

(2)
(1)
UM
Undisclosed Manufacturer #3
Mar 31, 2016

Interesting topic! I think perspectives about this vary a lot, so I guess there can be no clear winner. Personally, I don't like cropping at all, and I'm surprised Axis ACS does it. It is common in most Chinese cameras' Web interfaces, though... Do many of the major VMS' clients (Genetec, Milestone, Exacq...) crop too?

About VMS clients now warping by default, I expected the tendency to be the opposite, and anyhow, changes being uncommon since users will probably be used to what they have, so manufacturers will be reluctant to make big changes and disappoint them... Do you know of any important change in this aspect, from any of the main players?

Thanks for the article!