Swapped one of my OGs with a v4 today. Below is a comparison of night images under similar conditions. Make sure to click the pictures for larger view.
Observations for night -
-OG is more grainy/blotchy at night, especially when viewing the full 1080P image on a computer (far more obvious than what is shown here even). This got worse on the OG with motion and I just walked around in front of the v4 and it stayed pretty clear.
-OG colors are more vivid at night however I suspect that is tied to the graininess, it is amplifying it more. The v4 is more muted. There is discussion of this in another thread. I think they should calibrate it a bit so it is somewhere between this and the OG. Just after turning off the spotlight, it is pretty much perfect but then it recalibrates back to muted, so it is capable of showing better color without adding back the graininess.
-v4 is a bit darker, which I expected due to the higher resolution sensor needing more light, but there is significant ambient light in this area so not really an issue, they’re very close in this case. But in lower light situations, from what I’ve seen, the difference would probably be more significant.
General observations:
-Interesting to note the OG saves images as .PNG where the v4 saves them as .JPEG. This makes the file size significantly smaller even though it is higher resolution. I have mixed feelings on this, these days a 4 meg picture is not a problem, so not sure why they saw the need to compress it further to 500k.
-I still prefer the OG mount. In general it looked better and was easier to adjust for me, and the two screws made it more secure vs. 1. I considered using my OG mount with a thread adapter but it would have looked strange and made the cable length issue even worse. I ended up having to angle the base to be slightly crooked to get a straight image with the camera aimed the way it is. With the OG it had enough tilt where the base could be straight. I will have to ignore my OCD as I’d rather have a straight image than base.
-This particular spot is “moderate” or even “borderline” wifi - about -65dbm on my router. Both cams have about the same signal strength. v4 hasn’t had any issue staying at 2.5K and streaming smoothly, likely due to the increased compression keeping the bandwidth about the same.
-v4 takes a bit longer to start live streaming but that’s in line with my Panv3s. The OGs got significantly faster a while back with a firmware update.
-The 5 degrees less FOV does cut off a bit but it isn’t an area that was useful (and is covered by another cam anyway). Part of this is also that I have it aimed a bit more straight forward/to the left than the OG was.
-The focus seems a bit more specific on the v4 - the leaves/lawn on the right side are not as sharp as those closer to the cam. This is evident during the day too. Perhaps due to the way I have it angled, but the OG was angled the same way.
-v4 does tend to go to 360P and stay that way when there is a poor network signal. In my case, it is when the phone’s signal is poor, not the camera. It also seems to do it if I switch out of the app without closing it then go back in. So it seems more of an app issue than a camera one.
Will post a daytime comparison tomorrow, I have the “before” but the lighting changed significantly while I was swapping it out today. Swap was not as easy as I anticipated as the USB pigtail on the v4 is several inches shorter so had to adjust my wiring some. But honestly I’m not expecting much difference during the day.
This might be the only time I remember anyone ever saying that.
I can see the image difference you are describing. The OG coloring is more vibrant. I suspect part of that is having fewer pixels with a lower resolution, but your description of calling the coloring “muted” by comparison is a good description.
I don’t use OG’s very much, so I haven’t done a lot of comparison from those. Interesting to see in your camera position though.
Just added another picture, right after turning off the spotlight, the colors are a bit more vivid and accurate. The image quality is a bit worse but I think that’s because I snapped the shot mid-calibration.
So that implies they should be able to correct it via software pretty easily. In another thread it was discussed that the most recent firmware made an adjustment for a “red hue” so I think they went a bit too far and a bit more tweaking is needed.
Even when the spotlight is on, the fence/wall across the street is a bit more color accurate, and I doubt the light is hitting that much so it is more of a setting/calibration than hardware.
I wonder why they chose PNG as it is better for flat colours just like GIF and doesn’t do that well with continuous colours. JPEG is much better at compressing continuous colours.
It doesn’t appear they’re using much (if any) compression on the PNG. It is 4 megabytes in size where the .JPEG is around 500K. Uncompressed PNG typically has better quality than a compressed JPEG. However the quality seems decent on the JPEG so guess it is a non issue. Just an oddity.
My point was that PNG as GIF do better job at compressing solid colours (no artifacts), where JPEG introduces artifacts in solid areas but does much better job at continuous colours as it does better interpolation. That’s why the JPEG file is much smaller compared to the PNG.
Yeah, understood, but both depend on the compression level assigned. I’ve made some very large JPEG files with the quality set to the highest. Generally if I want the best quality, I save as PNG (or in extreme situations, TIFF).
That is true but both file/compression types have their purpose. For example for this type of image with lots of solid colours PNG will do an excellent job. JPEG will introduce artifacts in the solid red an yellow.
OTOH. JPEG on this one will do an excellent job compressing and retaining quality, PNG not so much. Hence JPEG is better choice when it comes to cameras.
For the most part I haven’t even paid attention, I only noticed when saving one from each that they were different extensions and sizes.
I just tested the 4 models I have during the day (which will create larger images than night typically):
Pan v3 - .jpg - 1.5 megabyte 1080P
OG - .png - 4.3 megabyte 1080P
OG TEL - .png - 2.8 megabyte 1080P (not much detail in that area right now)
v4 - .jpeg (same thing, odd that they used 4 digit extension) - 1.1 megabyte 2560P
So I guess only the OGs are using .png, maybe related to the cheaper image sensor or SOC they used to keep their price down.
I’m surprised the v4 size is smaller than the Panv3, either they’re using a lot more compression on the v4 or just a difference in the complexity, my Panv3 does overlook a larger area with a lot more stuff in view.
They definitely do, the bandwidth used is about the same, but with the higher resolution that means they’re compressing more.
The OGs have a lot of ghosting and artifacts at night, the v4 handles it a lot better even with higher compression. Wonder if that ties back to the PNG vs JPG but I believe both are using MPEG4 for the actual video compression (MJPEG seems unlikely) so I doubt it, saving an image appears to choose a format and convert/re-compress within the app.
Day time comparison. Note I tried to take the same time of day on both, but obviously there are some differences with the car not being there on the OG. This time of year the sun angle also varies by the minute so it may not be a perfect apples-to-apples. The v4 does show more detail in the grass/leaves etc. Colors are more vivid which may or may not be “accurate”. The sun seems to affect the v4 with a bit more glare but nothing terrible.
Looking at these pictures makes me wonder if the sun is brighter today but I believe the conditions were pretty much identical yesterday and today, I specifically waited for a time when the sun was overhead and there wasn’t cloud cover to take the picture yesterday, and it is the same out today at the same time.
I can see “flaws” with both images. The v4 has sharper focus in some areas than others, but the OG seems to be less sharp across the whole image.
So it appears the OG overstates colors at night while the v4 probably needs some tweaks to make colors more apparent, then during the day it is basically the opposite, OG’s colors are muted and v4 are probably a bit overstated.
Personally I’m not disappointed with either camera, but comparing the two has given me a good idea of what each excels at. My other OGs will stay since the areas they’re looking at do not need the improved night clarity, and I feel the day images on both are very good even though there are some differences. They both show what I need them to during the day, and I feel the v4 will probably be better at night, especially if they can tweak the colors a bit.
I originally had v3 and OG and found the OG significantly better with the ghosting and artifacts both day and night than the v3. The OG is fine during the day, I’ve just found in “color night vision” it is somewhat grainy and blotchy. So the v4 should be a pretty huge step up over the v3. The Panv3 is pretty good also, probably better than the OG but not as good as v4, but part of that is that the camera itself is moving around.
When I got my v3s they were the top of the line, no OG at that time. I am pretty happy with their performance and image quality, but the v4 is hands down better camera. At this time I have all the cameras that I need and I am not one who changes things for sake of change. Who knows what will be available when the time comes to have them replaced.
out of it to compensate for the higher resolution.
