V3, V4 etc camera shutter speeds

Hey there. I’ve checked on the internet and not found an answer to this.

I’ve read all the material I can find on the wyze cams, and have been buying them for years, and I know the frame rates, which are published.

What I haven’t been able to find is the effective shutter speed. Yes, I know they don’t have a shutter per se, but still, it’s important for some things, like catching the license number of a moving car.

I’ve been a photographer for 60+ years and know the subject well enough to have been able to fall back on camera-store jobs when all else failed.

So yeah, the question is whether Wyze sells a camera that has a reasonably fast effective shutter speed. Or perhaps one that’s quicker in broad daylight than in dim conditios. The amount of blur I get from moving objects seems to imply a shutter speed of 1/15 second or so, which would seem to mean it paints the imaging chip only once between frames. Even cheap cel phone cameras can simulate a pretty fast shutter speed. Has anyone addressed this?

Thanks!

Hdj

You can imply the shutter speed from the FPS to a certain extent. v4 is 20 day and 15 night (so 1/40 and 1/30 respectively). Of course the massive amount of compression added throws a wrench in the works of that calculation and will make it appear much less. That’s where much of the blur, or from what I’ve been told is more accurately called ghosting, comes from.

I have an old dash cam setup in both my cars, it is 720P at 20FPS. It can capture a license plate in circumstances the 1080P wyze at 20FPS could never dream of. But it also records about 10x more data to SD per minute, so significantly lower compression (especially given the lower resolution).

The only wyze cam I have that can get a license plate in motion (and it has to be moving pretty slow) is the OG-Tel just due to the fact that it has 3x optical zoom. When I calculated it, it came to something like >99 percent compression ratio on the Wyze OG and Panv3 (both 1080P cams), which is not going to be your friend when it comes to getting detail on something in motion.

The standard, non telephoto OGs and Panv3s work pretty decent on people at distances up to 10 to 20 feet or so (the v4 should improve that given the higher resolution), since they’re not moving that fast and often are moving toward the camera. Beyond that distance it starts to diminish pretty quickly.

When I compared the v3 to the OG, the ghosting and compression artifacts (blocking) on the v3 was significantly worse so I went with the OG. The v4 seems to be in line with the OG, better than the v3 from what I’ve seen, but still not going to get a plate unless it is sitting still or moving fairly slowly directly toward the camera (the higher resolution of the v4 helps, but the increased compression they added to try and keep the data rate about the same negates much of it). It gives more detail for still, or slow moving objects, but faster moving stuff especially when moving across the plane of the camera instead of toward/away from it still suffers.

There is a wishlist item to reduce the compression ratio, even if just for videos stored on the SD card (and another to give access to the SD card remotely to be able to download videos directly from it at original quality). You can vote on those but to be honest, they don’t seem to have gained any traction. Who knows if it is even possible, the compression may happen before it hits the data plane so they would not be able to tell it to upload lower resolution than is recorded to SD.

Wireless cloud cams simply have bandwidth and signal strength limitations that they have to design around (and the bandwidth and storage space on their servers also costs money so they have to minimize it as much as possible).

There are systems out there with much lower compression, they will typically be hardwired POE cameras with a dedicated NVR. If I was in an area where it was more likely to be needed, I’d invest in that (probably about $500 entry level price plus installation and wiring). But things are pretty quiet here and I’m more just looking to know when people come and go, when packages get delivered, and be alerted if someone is in my house when I’m travelling, etc, so the Wyze work well for my purposes. The one time so far I’ve needed to submit footage to the police, higher resolution would not have mattered, the kids were wearing hoodies and masks, but still comparing it to the Blink and Ring that the neighbors have, it was night and day, I can only imagine what compression those ones are using (who would have thought >99% was possible and yield anything more than a total blur).

Thanks Dave.

It has been my impression that the vid off the SD card is maybe clearer than what streams over the LAN, but perhaps I’m wrong about that. Certainly, even the V2 cams can read the plate of a parked car in the spot I’m concerned with. But I suppose the compression used is inter-frame rather than just intra-frame, so maybe the compression is the culprit.

While the FPS sets an upper limit on shutter speed, there should be nothing electronically prohibitive about having a faster shutter speed. The FPS is largely about bandwidth while the effective shutter speed is just a matter of how long it takes the sensors to react… which in principle could be 1/125 or 1/200 second if Wyze enabled that, perhaps only in daytime mode with the amount of light is 1000 or more times what is used in night imagery.

I guess, though, that nobody is asking for that, or understands how it would improve the camera’s performance, probably with no additioal hardware expense for Wyze.

What this means is that instead of buying a v3 or v4 camera for this project I’ll probably just adapt an old cel phone that allows the shutter speed to be changed, but that’s annoying because they won’t go 24/7. Trying to catch people cuttingthe curb in front of my house and almost killing people.

thanks

Depending on your LAN (assuming no bottlenecks) the live view and SD card playback should be the same as viewing directly off the SD card. However when you need to capture a recording, taking it directly off the card will be better, since the app just records it again, with your phone adding more compression (and also making the file size much bigger than it should be). So yes if you need a recording and need it to be as clear as possible, grab it off the SD card directly.

The starlight sensors in these cams tend to be slow to react so I doubt they’d be able to refresh them 100 times per second. Then on top of that the additional processing power (and buffer memory) to grab 20 of those frames only would probably require additional cost. The cameras just take what streams from the video chip and send it along with out much processing. Keep in mind how much your phone costs vs how much one of these cameras cost. And in order to take really good daytime pictures, the phone cams tend to suffer at night. They have to significantly overexpose (and you need a steady hand) to get a good nighttime picture. That’s due to the ultra high “effective shutter speed” you mention, takes good pictures during the day (and adds in features like clear slow motion, etc), but doesn’t let in much light at night. So there are trade offs either way.

Wyze cams can do a perfectly good job of recording a plate that is within 10 to 20 feet and not moving. It is motion that causes the compression artifacts (as well as motion blur due to frame rate). But 20FPS is more than capable of getting a nice clear image of a plate on even a moving vehicle, it is the compression that kills it.

So if you need to capture plates or face detail in moving vehicles, these aren’t going to be great for that. The OG-Tel might cut it, it does a pretty good job watching an intersection from about 50 feet away for me, but I would not rely on it to grab a plate or even a great face picture.

There are other standalone cams, typically hardwired, that can easily do what you’re looking for. Rigging up a cell phone to do it is going to be a hassle and unreliable. Of course you also have to keep in mind, most police will not act on a video you’ve taken, other than to possibly increase patrol of that area. Of course if someone does get hit, then they’ll use the video, but they’re not going to go issuing tickets based on your footage if nobody is injured and/or there isn’t a crash. So take that into account when deciding what to invest in this.

Heck I have this old Kodak 1080P digital video camera (door prize at some stupid work event) that before I had the wyze cams, had set up a couple times when things were going on in the neighborhood. While its night performance isn’t great, it had very clear motion and detail during the day. So you might be able to find some used thing like that which would serve your purpose.

Thanks Dave, good info.

This particular project is low-budget and will eventually be vandalized so I’d like to keep it <$50. Also it will need to be rain-proof and a bit unobtrusive… and oh yeah it’‘ll be mounted 8-10’ off the ground. This narrows my options.

I have no illusios about the cops doing anything, but I wouldn’t mind the deterrence of noting that people coming over the curb like they’re cutting corners in a grand prix are being recorded. But beyond that, the pragmatic purpose will be documenting that a danger exists so I can get the state to put in vertical guides that will prevent it. Otherwise, a video of a dangerous situation is just waiting for lawsuits when somebody gets run down. Now arguably I can show this without the plate numbers being captured, but I want them captured. A few of these people have caused my wife and me to leap at the last second to save our lives, and we’re in our 70’s now, leaping more slowly every year.

Push comes to shove, I can sit out in a lawn chair with a digital SLR and get them in 4k with a good autofocus, but that might lead to physical altercations with guys in pickup trucks demanding I delete the images and my telling them to get bent.

A fast effective shutter speed should be a tweak that one could make at the level of design… for instance, running the starlight imager in the daytime with a 1/1000 second sample from the imaging chip… Do any of the actual wyze people read the stuff in this forum?

best

Wyze concentrates on the Wyze News, Beta and Wishlist categories. Your best bet for Wyze visibility is to merge this/your topic to the existing wishlist topic that covers the ability for users to manually adjust image sensor parameters. Let us know and we’ll merge your topic to “Add Exposure and focus point/area, and brightness and contrast level controls”. Alternatively, you can simply add your use case to the wishlist topic and click the Vote button at the topic of the page. A merge would be a better action though so as not to lose the already stated, valuable discourse here.

1 Like

I don’t know what the hardware is capable of (sensor and processor), but most likely it comes from the manufacturer that way and Wyze probably doesn’t have much control over it. That’s just a guess though. Each cam also uses a different chip so v4 may be better than v3 (it certainly handles the compression better). I don’t think any of their cams do 30FPS but you would assume ones with higher frame rate would have a better effective shutter speed. I believe they all top out at 20 though. There are other more expensive brands out there, may be able to find one with 30FPS, but may put you over your budget too.

But I would not hold your breath on submitting a feature request to Wyze and having it done in any short time period, those are generally considered long term improvements and may not even get added to current cameras etc.

With a less than $50 budget, your best option may be to try an OG-Telephoto and get it lined up where it will see the plates head on. I’m not saying an OG or v4 can’t do it, but it would have to be much closer and likely you can’t mount it that close (or it would get run over/stolen/smashed). I’ll try to grab an image of a car going by on mine which is approx 30 to 50 feet away and PM it to you. Mine watches a dangerous intersection in front of my house and is also my “hopefully catches a plate” cam in case a car takes off etc.

You may not catch every plate but you may get some of them and it will at the very least give you the “proof of a problem” like you mention.

If proof of the problem is enough and you can live without the plates, then a v4 ($18 for one at Microcenter or $36 for two at Costco right now) is probably your best option. Add on a good endurance rated SD card from Samsung or Sandisk and you’re still easily under $50. The OG-T combo should be at or under $50 also.

When it comes time to gather your clips, if you can, pull the card from the camera, put it in your PC, and grab the videos directly from it, that will be the highest quality and more chance of a full or partial plate.

I really don’t understand some people. I have a 3/4 ton pickup truck, in no case would I want to drive it over a curb unless I have no other choice (which does happen from time to time and is one of the perks of a truck). Tires and shocks are not cheap.

Around me the current issue is the kids with the machine gun mufflers and rich engine tune that backfires constantly on deceleration (which they drive around speeding up and slowing down to make it do it). It eventually blows their engine but in the meantime, pointless annoyance. Then again I suppose some people say that about the 427 V8 in my car, but personally I think that is a beautiful sound :slight_smile:

One other note, none of them are going to catch a plate at night or dusk, the car headlights will wash the image out too much.

Sent you a link in private messages to a few pictures from my OG-Tel. That one may be your best bet if you want to be able to get plates (and can mount it in a similar fashion where it is somewhat straight on with the cars).

If for some reason you’re able to mount it pretty close and have it still be safe, the v4 may work for you. Not sure what distance and angle you’re working with.

As you can see the OG-Tel can get plates at up to around 100 feet when they are moving relatively slow (20-25 mph on those 3 is pretty clear). As they get faster it gets less likely, 30 would probably be where it starts to get spotty, above that and it is a crapshoot, may or may not get one.

I would guess the V4 would cut that distance to around 30 feet, maybe as much as 50. I don’t have one so can’t say for sure, just a guesstimate based on the fact that it is not 3x optical zoom, but does have a higher resolution.

With the compression they use, probably nothing is going to be able to get plates reliably above maybe 30-35mph or so.

Dave - first off, many thanks for the time you’ve taken, and clearly you’re a smart person.

I considered an OG-Tel because an oncoming car seen from the front in telephoto would be far more likely to capture the plate. Unfortunately, that view wouldn’t make it clear that the car had transgressed the safety lines and come up on my yard. Pretty much needs to be wide-angle for that. Yeah, I know, a very particular use I have in mind. IN fact, the worst and most dangerous transgressors have accelerated to about 45-50 by the time they’re cutting a foot from our mailboxm and those are the potential “grim reaper” drivers. Some driving teslas that make no noise, and me half deaf. Indeed, showing how long the car takes to traverse the known length of the safety-line area is what will in principle allow a pretty accurate documentation of the speeding in a 15mph zone.

I may well get a V4 at some point and mess around with it, despite having a drawer with never-used v2’s and V3’s. But the deals you have available on the mainland aren’t usually available in Hawaii. Ridiculous what things cost here. Microcenter would charge more to ship than the cam cost, and costco probably won’t ship. Let’s check, I’m a costco member. There’s the deal. Says free shipping, I’ll check out,

Item #1836081 cannot ship to selected destination 96734-4628.
Please review the item details for shipping restrictions. You may complete your purchase without the item or select a new delivery address.

so I’m hosed, but I tried. and on amazon and wyze.com they’re $36 each, no deals, so I needen’t jump immediately.

Best of luck with your noise issues! Living on a main road we have those too… and several harley’s with amplifying mufflers who like to zoom back and forth in front of the house at night.

anywho, many thanks for your effort here, it has given me things to think about. Moving out of Hawaii, for one.

cheers

dj

Yeah unfortunately I don’t think any Wyze cam (or any sub $50 cam) is going to pick up a plate at that speed. Who knows, maybe if you can get it mounted where it is a straight-on view but sounds like that’s not the case.

Wasn’t even paying attention to the Hawaii in your name. Yeah the Costco deal ships direct from Wyze so they probably restrict it.

You might have to just settle for capturing the problem and not the specific plates, at least that might get you some action.