Screenshots: IP megapixel vs Analog

[alpine0000 0]

Had these on my computer and figured I'd share them. I recently replaced my Panasonic WV-CW484S analog dome camera with an Axis P3346-VE megapixel IP camera and had some day and night shots of each. Admittedly, the Panasonic definitely could've used a bit of tweaking and adjustment, but I didn't want to fuss around with it before uninstalling it.

 

Enjoy the screenshots anyways. You can click on them for the full resolution to see the real difference.

 

Day time Panasonic:

 

Day time Axis:

 

Night time Panasonic:

 

Night time Axis:

[Birdman Adam 0]

Great shots!

 

The Axis is clearly much better, night and day! This is a very good example of what each side has to offer, considering price as well. What kind of frame rate do you get with the Axis cameras? Is it just me or does the Axis seem out of focus at night?

[mike_va 0]

Nice, I have a P3346 as well on the driveway.

 

There is a lot you can do with them, send motion detect ftp to NAS, set a schedule to switch day night etc. Not the "best" with low light but if you have enough light they work well. Plus they have a mic built in. I keep the shutter at 1/30-1/50 min otherwise you'll have a lot of blur. Have fun!

 

They say 20fps for 3MP, 30fps for 2MP. I run mine I think around 10fps. Note that pic is probably not 3MP, given the aspect ratio. Should be better though at ~6-10x improvement in resolution compared to analog.

 

You can also set up a view area and send that stream somewhere else. Work well with multiple streams also.

 

Edit: to get the shutter speeds first select motion priority on the exposure otherwise they remain hidden. That took me a while to figure out just thought I'd mention.

Edited March 11, 2012 by Guest

[alpine0000 0]

The Axis is clearly much better, night and day!

Yea, one of my biggest concerns was that if I went megapixel, my night time performance would get worse. I was shocked to see that it is now MUCH better, night and day.

 

 

What kind of frame rate do you get with the Axis cameras?

It will do up to 30, but I think I have it set on 15.

 

 

Is it just me or does the Axis seem out of focus at night?

It could use some adjustment, but that post light in the front yard really messes things up for me and makes it difficult. I need to do something about that. The image is still very usable, though.

[mike_va 0]

Is it just me or does the Axis seem out of focus at night?

It could use some adjustment, but that post light in the front yard really messes things up for me and makes it difficult. I need to do something about that. The image is still very usable, though.

You can also set up focus and exposure windows on the Axis.

[fa chris 0]

From my experience the whole "analog is better at night" thing only applies to 1st generation and low end MP cameras, and is way overplayed. These are some great comparison shots.

[bike_rider 0]

While I appreciate the time everyone talking to post shots, I would find the information much more useful is people would share a some settings on the camera. For example, shutter speed is critical for evaluating night shots. Thanks

[SEANHAWG 1]

I think most generally IP MP cameras still see better at night time compared to Analog when you look at the "whole picture". But there is a drastic change in performance between a daytime MP picture and a Nighttime MP picture picture. With Analog, the performance difference is not so drastic. So this gives the perception that analog is better at night when in some cases its not. Thats how it is with my experience though, I havent played with the old school stuff.

[Soundy 1]

From my experience the whole "analog is better at night" thing only applies to 1st generation and low end MP cameras, and is way overplayed.

Agreed. There is a solid reason why this is THEORETICALLY the case - more pixels in a given sensor size means smaller pixels with less ability to collect light. However, as is the case with analog, there's more to low-light performance than just the pixels; otherwise a $30 eBay cam would look just as good at night as a Monalisa cam.

[shockwave199 0]

To me, if you're comparing differences without a person in different spots in sample picures, you're only comparing how nice a picture looks- especially at night. I'd guess a good MP cam would provide an improved ID shot at night as well, provided the lighting is appropriate. But that's only a guess. I only ever see pictures comparing scenic views, not what any one camera will provide for ID'ing purposes. That requires humans photograped by the cameras at different points in the field.

[SpyGuy10 0]

I only ever see pictures comparing scenic views, not what any one camera will provide for ID'ing purposes. That requires humans photograped by the cameras at different points in the field.

True, but even a human test is somewhat subjective. Besides having to take into account the various angles at which a person's face can be viewed, you must also consider that the human brain must perform very complex analysis in order to recognize a face, and that analysis is influenced by a number of factors. Obviously, if the test subject is known to the viewer, then that subject will be MUCH more readily, positively, and quickly identified than if the subject is a total stranger. Also, it is well-known (at least among those whom are intellectually honest) that people are better able to identify members of their same race.

 

While I agree that human subject testing is very useful for comparing CCTV systems, it is not as objective as using a test target. I think license plates tend to provide a useful metric, at least for outdoor cams, as they are standard in size and composition (within a country) and they are universally available.

[alpine0000 0]

While I appreciate the time everyone talking to post shots, I would find the information much more useful is people would share a some settings on the camera. For example, shutter speed is critical for evaluating night shots. Thanks

 

resolution: 1920x1080

compression: 5

15fps

h.264

 

everything else is still on default (except the focus/zoom), as these screenshots were taken *right* after the camera install before any major tweaking took place.

however, i don't believe you can set a specific shutter speed on this camera. there is an "exposure priority" option, and you can choose 'default', 'motion', or 'noise' from a drop down menu. mine is set on default, but there is no way for me to know the actual numeric value of my shutter speed I guess, unless Axis releases that info for each of their pre-determined modes?

 

 

EDIT:

 

I just found this in the users manual for the camera:

 

'Exposure Settings' below allow you to configure the exposure settings to suit the image quality requirements in relation to lighting, frame rate and bandwidth

considerations.

 

Exposure value - Click in the bar to fine-tune the exposure.

Exposure control - These settings is used to adapt to the amount of light used. Automatic is the default settings can be used in most situations. The shutter speed is automatically set to produce optimum image quality. Flicker-free 50 or 60 Hz is used to remove flicker which can be caused by fluorescent and other light sources. The Hold current option locks the current exposure settings.

Enable Backlight compensation - Enable this option if a bright spot of light, for example a light bulb, causes other areas in the image to appear too dark.

Exposure zones - This settings determines which part of the image is used to calculate the exposure. For most situations, the Auto setting can be used. For particular requirement, select a predefined area.

Exposure priority - When Motion is prioritized and maximum Shutter time is set to a small value, motion blur in the image is minimized. This can be useful for recognition of moving objects such as people and vehicles. However, prioritizing motion may cause an increase in image noise, especially in low light situations. When Low noise is prioritized and Gain is set to a small value, image noise is minimized. The file size is reduced, which can be useful if storage space

[mike_va 0]

Exposure set a window using "defined" then select "edit". You can draw a window on whatever you want the exposure set on.

[shockwave199 0]

I only ever see pictures comparing scenic views, not what any one camera will provide for ID'ing purposes. That requires humans photograped by the cameras at different points in the field.

True, but even a human test is somewhat subjective. Besides having to take into account the various angles at which a person's face can be viewed, you must also consider that the human brain must perform very complex analysis in order to recognize a face, and that analysis is influenced by a number of factors. Obviously, if the test subject is known to the viewer, then that subject will be MUCH more readily, positively, and quickly identified than if the subject is a total stranger. Also, it is well-known (at least among those whom are intellectually honest) that people are better able to identify members of their same race.

 

While I agree that human subject testing is very useful for comparing CCTV systems, it is not as objective as using a test target. I think license plates tend to provide a useful metric, at least for outdoor cams, as they are standard in size and composition (within a country) and they are universally available.

Specifically talking about the night shots here, no matter what camera, and for all the fiddling, it's the difference between setting it up for a good picture or setting it up for the best shot at ID'ing, imo. I'd simply walk out to those cars and act like I'm lurking and looking in them. Pull on the door handles. Walk around them just as someone would do who might be looking to bust into them. What you see [or don't see] would tell you where your lighting deficiencies are, that's all. I don't discount myself from this either. My analog camera pictures aren't all perfect for ID'ing yet either. But all too often I see noodling to dial in good pictures. How well that picture catches a thief could be another story. Dial in settings for the targets is all I'm suggesting.

 

Btw- I think the panisonic day shot was very good but the night shot was not good at all. I would have liked to see what that camera produced with a little fussing. The Axis is fantastic in the day shot and good at night. What anyone would look like when they peer into the passenger window at 3am with their backs to the lawn lamp remains to be seen though. Damn lamps are tough to deal with. I deal with one too. I've tried to just keep it out of the frame for the time being, but something will have to be done in the end.