Good test Ethan. Interesting seeing all the results of the big three we're seeing.
Testing Panoramic Cameras Outdoors
Published May 12, 2013 04:00 AM
********, **** **** ** ** * favorite *** ******** ***** **** ***** but **** **** *** ******** **** are ********* ******* **. ********** *** an ******** *********** *** *** **** do **** ******* *** **** **** of ***** ******* ** **** *******?
********* ***************** ********* ****** ********, ** **** * ***** ** panoramic ******* ** * ********** **** intersection ***** ** **** ***** *********** viewing *********** *** ******* ******* ** various *********, ***** ** **** ******** view:
*** ******* **** ******* ** *** portable ******, ***** **' *** *** ground:
********** ********* ********* **** **:
**** *** *** *** ******** **** the ****:
- **** ** * **** *** ******* scene, *********** ** * ***** ******* is ******** **** ** **** ***** range, *.*., * ****** ******* ** cross *** ******. ** ** ***** in *** ****** ********, ********* ******* than *-** **** ******* ****** ** the *** ** ************ *******.
- **** ** *** ******** ***** ********, with * *** **** **** *** cameras, ******* ****** **** *** ********. *******, make, *****, *** ***** ** ******** is ****** ********.
- *********** ****** ****** ** *****. *** Axis ***** ******** *** ***** ******* better ** ****** *********, *** ******** from ******* ****** *** ******* ******* lights ******. ********* ******* ****** ****, but *** ***** ******* ***** *** to * *** *****. *******, *** Vivotek ******* ******** *** ********* *****, but ***** *** ***** ***** ******** made *** ******* ********* ** **** and ******** ********* ** *********.
- ** *****, *********** **** ** ***** distances ** *********, *** ** ***** and ****** ******. ********* ** * subject ** ****** ** **** ****** (50'+) **** ******* *********. ******* *********** becomes **** ********* ** ******, *** poor ******** ** **********, *****, *** dim ******. **** *** ***** **** not ******** ** ********* ** **** cases, *** ***** *** *********** ********** to *********.
****** *** ***** **** *******, ** believe *** **** *** *** ********* cameras ******** ** *********** ********* ****. Users *** **** **** ********* ** conjunction **** ***** *******, ** ******* an ******** ** ********** ** *** scene, *** ******* *** ******* ** follow ******** ** ******** ** **** move ********** *** *****. ** ** not ********* ***** ********* ******* ** any ***** ** ************** ** * requirement, ***** **** ******* **** *** model ** **** *** ******.
Daytime ***********
** ***** ** **** **** ******* common ** *************, ** *** *** subject **** **** ************* ***' ** near *** ********* ******, ***** *' from *** *******. ******** **** **** overview *****, ** *** *********** *** differences ** ***** *** ********** ***** the ***** *******. **** ** **** may ** ********* *** ** ********* white ******* ****, *** *** **** left ** ********* *** *** *****.
*** ***** ******* ********** ** *** much ** *** **** **** ****** uses. ** *** *** *** ***** edge ** *** ******* ** *** M3007, ***** ********* ***** **** ***. Vivotek **** *** **** **, *** does *** *** *** **** *** up *** ****** ** *** **. This *** **** ** ****** ** image *******, ** ********** ******* *** further *** **** ******* *** **** of *** *****, ******* ** *** blurring ** *** **** *****. *******, eliminating *** **** ****** *** ******* reduces *** ******* ******** ** *** camera. *** ******* ** ** ******* of ****, ** **** ***** ** panned ** *** ** ** ********, but *** ***** **** *** ****** reaching *** *******, ******* ** *** other *** ******, ******* *** ******* being ******* ** *** **** ****** and *******.
** ***', *** ******* ** *******, without *** **** ** **** *******. Light ***** *** (** ****) **** jeans ** *** **** *** ** made ***.
****** ** ***** **', ** *** crosswalk ****** ******** *** ******, ******** more ****** ** ********* ** **** cameras, ****** *** ****. ** **** point, ** *** **** **** *********** in *********, **** *** ******* ********* squat ** *** **** *****, *** approximately ************* ** *** ***** *******. This ***** ** ** ***** ** trying ** ***** ****** ** * subject.
** **', ** *** ***** ** make *** ****** ********. ** **** point *** ******* *** ** ************ if **'* *****, *** ********* ***** be **** ** ******** **** **** resolution. *** ******* ***** ******* ***** using *** **** ****** **** ******.
*******, ** *', *** ******* ** recognizable, **** ******* *** **** *******. A **** ** ****** ******* ** his *****, ** ****.
**** *** ***** ** ******** ****** through *** ************ **** *** ********* to *********, *** **** ** *** closest ******** ***** (~**'), ******* ****** were *** ***********:
**** **** ***** ******* ** ****** for **********. ***** *********** **** ***** ****, ** ****. **** **** **** *** ***** and **-***** ******** ** *** ******, so *** **** *** *** ** changed. *** ******* ******** ********* *** immersive ********, *******, ** ***** *** view *** ***** ***** ***** **** archive ****** ********.
Nighttime ***********
*********** ** *** **** **** ** night ******. *** ******** ** **** scene ****** ******* ************* *-** ***, depending ** *** ***** *** ****** was ** ****** *** ******* ******, and *********** *** ******* ** **********. All ******* ********* ******** ****** ******* to ******* **********, ****** ******* ** either *** ***** ******* ** ******** were ********* ** ******. *** **** M3007 ******** *** ******** ***** **** the ****** ***** *** ***** *****. The ********* **-***** ******* ********** *** taillights ****, *** *** ******* *** dimmest ** *** *******, **** *** subject ********* ** ****** ** *****. Finally, ****** *** ******* ** *** only **** ***/***** ****** ** **** test, *** ** ******** ****** ******, it ******* ********** *** ********** ******, and *** *********** ***** ******** ** the ***** ***.
** ***', *** ******* ** ****** detectable, ** ****:
** **', *** ******* ** *********** on *** *****, ****** ******* *** detail. ** ******* ********* ** **** in *** ***** *** *******, ** to ***** ****** ***/** *****:
** **', ******** *** ** ** least ********* ********** ***** *** *****, and ** ** ****** **********.
*******, ** ***** *****, ******** ******* easy ** *********, *** **** ****** detail ** *********.
*** ***** ***** **** "*****" ** "dark" *** ********* ** ********* *** to ***** **** ********** *** **********, making ***** * **** *****. **** and ***** **** *********** ********** ** determine, ** ****, *** ** *****, glare, ** ****** ****.
**** ***** ***** ***** ******* ** the ********* ******* *****. ***** *** also******** * **-****** ***** ****.
Test *******
** ****** ***** ******* **** *** ********* ********:
- **** *****-** -- *** ******, ******* to ******* **** ****, ******** ** ******
- ********* **-*****: *.*** ******, ******* ** 1.3MP **** ****, ******** ** ******
- ******* *******: *** ******, ******** ****** ****
*** ******** **** **** *********, ****** shutter *****, ***** *** ************ ** 1/30s. ********** **** ** *** ******'* highest ******** ***** **** *** **********.
Comments (15)
Nathan Wheeler
May 13, 2013MW
Michael Wright
May 14, 2013Nthan...great work! However assuming that the images were time synchronized, why are the images from the Vivotek camera always lagging behind the others?
Ethan Ace
May 14, 2013Hi Michael, the cameras were synced to the same time server, and they're double checked before each test. There may be very slight float in the time (less than a second), but they generally show the same time. I suspect the lag we see is due to encoding latency, more than anything, as I can see it even in live video.
MW
Michael Wright
May 14, 2013Ethan,
As I expected...I knew you would have them time synch'd! My comment was specifically directed to the encoding latency issue...and wondered why it's so noticeable w/ Vivotek and apparently nonexistent in the other 2?
Ethan Ace
May 14, 2013Hey Michael, the Vivotek actually showed the least lag (all the way on the right in the embedded clips above), followed by Panasonic about a half a second (or less) behind it, then Axis about 1.5 seconds back.
I suspect it may have to do with how many streams the M3007 sends out, as we've noticed other issues because of this, too. I've already asked Axis for comment, and will circle back with whatever they tell us.
MW
Michael Wright
May 14, 2013Ethan...yes, it was Axis...I was goijg from memory as I had closed the window long before I added the comment. It will be interesting to see what their response is...and it it's related to the number of streams, I wonder if shutting some of the added streams off would improve the performance?
U
Undisclosed
May 15, 2013Hello Ethan,
did u done any test on arecont vision 360/180 deg camera in indoor & outdoor application>
JH
John Honovich
May 15, 2013We have done a test of the 20MP 180 degree model.
We plan to test the upcoming 12MP 360 head to head against these models above.
BC
Ben Clay
May 15, 2013Do you plan to test the 180 panoramics, such as those from scallop imaging? Would be interesting to see how they compared to 360 fisheye cameras, particularly with their ultra low light unit for night time.
JH
John Honovich
May 15, 2013Ben, we will do one general comparison between 180s and 360s to show fundamental tradeoffs but we are not going to 'shoot' them out head to head as it would not be apples to apples (dramatically different field of views).
As for Scallop Imaging, the issue we have is the max 1fps frame rate for full resolution. Regardless of its low light capability, at 1fps, that's a huge limitation. Our understanding is a new generation will come out (sometime) that will improve this and when that happens, we definitely test Scallop.
JE
Jim Elder
May 15, 2013I would like to see how Immervision does in this comapriason. Are you planning to run include Immervision any time soon?
JH
John Honovich
May 15, 2013Jim, we included Immervision 1.3MP in the overall panoramic shootout, so you can see that comparison. Immervision image quality was in the middle of the pack there. As such, we would expect the same here.
More importantly, soon Immervision has their 5MP lens coming out and the first 3rd party minidome embedding their lenses. We plan to test both and would assume better performance from those new ones vs the old lens they have had out for years.
JE
Jim Elder
May 16, 2013Right. I read that post with great interest some time ago. Awsome job by the way... I was also at the ISC when you did the live comparison. What i meant was that i wanted to see in in the 'street corner' version of the shootout. Should have explained myself better.
Also, that 5 mp Immervision lens was to have shipped under the Fuji name this month. I wanted a couple for a job I'm doing now, but they are not yet selling them. You are correct about the form factor issue. So i am sticking an Axis camera in a cylindrical light fixture (see below) that matches with other fixtures in the space. But after your shoot out, i am having second thoughts and thinking more about Arecont.
I am very interested in your test of the Brickcom camera (which imbeds the 5mp lens in a dome camera). I like the form factor and the fact that i don't have to worry about the VMS support issue; but i know nothing about the company. Tell me more....
JH
John Honovich
May 16, 2013Jim, thanks for the feedback and the diagram. Nice approach.
It's too bad that the 5MP lens keeps getting delayed. It's been going on for a very long time.
As for Bricksom, they are a Taiwanese manufacturer who first entered the surveillance market a few years ago. They are a subsidiary of Gemtek, a wireless broadband networking provider. Brickcom has a fairly standard line of surveillance cameras, with their key differentiator, not surprisingly being for wireless / cellular integrated options.
Immervision says they are going to embed their M mount lenses into many other minidomes over the next year so you can pick from Brickcom or later on, see who else Immervision adds.
SU
Stanislav Utochkin
Sep 30, 2014Very interesting testing, thank you. I have implemented modeling panoramic cameras in VideoCAD8 and compared results of modeling with these tests. 
