Camera upgrade for BlueROV2

Hi everyone,
I’d like to re-ignite an updated discussion regarding an upgrade for the USB camera of BlueROV2 and it would be great to get some feedback from users that have managed to make successful upgrades to their cameras and how they’ve done so.
I recently purchased the following camera following a recommendation here:

The camera works ok, I haven’t yet checked if it easily fits instead of the USB camera on the servo but anyway I would like to find a higher quality camera instead. The max bitrate for this camera is about 6kbps, which is not enough when you want to see fine details in the video.

I’m aware of larger cameras with higher quality but I still hope to find small cameras with high bitrate that can fit on the servo in the original Bluerov2 enclosure.

Did anyone have experience with that?

Hi @odedezra

I assemble 2 models of IP cameras to install in the customized ROVs that I design for my clients. Given the good results I have prepared these 2 models so that other people can install them in their ROVs.

The first model can be installed on the normal 4" enclouser of BR2, without any modification (except for the use of an ethernet switch), just plug and play. Includes voltage regulator for power supply, servo, aluminum housing and methacrylate mounting plate:

Features: 8mp 4k 20 fps or 3072x1728p 30 fps/ Sony CMOs imx 415 ultra low light illumination sensor/ 0.01lux color/ varifocal lens 2.8-12 mm 4xzoom/ fast focus with screw stepping motors / bit rate up to 12,000 kbps / h264 or h265

Viewing (example on the videos above):
-Via qgroundcontrol by rtsp://
-Hik protocol via hikvision free software IVMS-4200

Video image quality ( download to view in full quality):

The second model is a 12mp professional IP camera to be installed in a 400mm 4" encluser, replacing the rear spacers with shorter ones.

You also need to use an ethernet switch. Includes voltage regulator for power supply, servo, aluminum housing and HDPE mounting plates:

Features: Image Sensor 1/1.7" 12 Megapixel CMOS / Resolution: 12MP (4000x3000), 8MP/4K (3840x2160) / Video Compression Smart H.265+, Smart H.264+, H.265, H.264 y MJPEG / Stream : 12MP (1~25 ips), 1080P (1~50 ips) / 0.01 lux F1.8 (color), 0.001 lux F1.8 (B/W) / 2.7~12mm motorized optics / Viewing angle: 103°~44°(H), 75°~33°(V), 135°~55°(D) / Picture modes: BLC, HLC, Digital WDR, 3D-NR, Defog , Mirror, Manual white balance (perfect for adjusting red and green color via software to correct live underwater images without further editing), Automatic and manual gain control

-Vía dahua free software DSS-EXPRESS
-Via web browser
-Via qgroundcontrol by rtsp:// but not at full quality, this software can’t run the total quality of this camera

Video image quality ( download to view in full quality) (the video is compressed to 1080p to share it here and to be able to see it well on a smartphone. The image quality is much higher soon I will upload video with full quality):

I have been working with IP cameras in underwater equipment for years and in my opinion, I have not seen anything better.

Note: For a correct visualization of these two cameras, a good communication through tether is necessary and a computer capable of running 4k video and going smoothly with multitasking. If not, you need to reduce the video quality to have a smooth image.

People interested in these two models can contact me at

I do not manufacture in series so the two models are available in back order

Best regards



Hi @odedezra,

We’re currently looking into a few different higher quality camera options, but I don’t have a fixed timeline on when we’ll be stocking one. It’s perhaps worth noting that recent BlueOS 1.1.0-beta versions can stream some additional formats now (currently only in pirate mode).

As @Andres mentioned, IP cameras (via an ethernet switch) are another possible option.


Hi @Andres!
Thank you for the very detailed answer, I really appreciate that.
You have some very cool designs for the cameras and ROVs, looks really good.

I actually also use HIK cameras in some of my projects, we use them mostly for stationary observation systems but also attached one in the past to the ROV in a dedicated enclosure and sent the video via a switch in the main 4" enclosure. If I understood correctly you use a HIK camera on the servo? looks like a nice design. I saw that you also control the zoom/focus through a dedicated FathomX and controller? I guess it means it’s easier than working through the main controller to operate it?

I agree that those IP cameras provide a good solution although at times a better quality video would be helpful. I’ve been using higher level cameras on larger ROVs, some of them use sony cameras to my knowledge and the quality is way better. So besides upgrading the main BlueROV2 camera I would also like to have another dedicated, possibly with much higher bit rate cameras such as Sony or Tamron block camera, record the high bit rate internally and send a compressed video via IP. This might be the next project after an upgrade of the main ROV camera on the servo.

Having said that, the Dahua video you shared looks great, I’ve heard that these cameras can provide great quality but I haven’t tried them yet. Do you think they give better quality than the HIK 4K cameras?

Anyway, I will contact you directly by email regarding these two models.

Thanks again!

Hi @odedezra,

Thanks for your coments.

Camera 1, the most compact, is a modified HIK camera, with a high-speed motorized varifocal lens. The varifocal lenses that are installed in these cameras work well, but they are very slow and the zoom takes a long time to approach objects and focus them, practically losing its usefulness except for focusing on very close objects. This modification allows you to use the zoom and focus in an agile way on the move without sacrificing quality, since the focus time is considerably reduced and the zoom is very fast.

I do the zoom and focus control with dedicated software, since, on the one hand, communication with the surface via RS485 or RS232 is not necessary and thus I don’t need to use an extra twisted pair of tether. And on the other hand, the dedicated software allows zoom and focus control directly with the ROV joystick, thus avoiding extra configurations.

The dedicated software allows you to view the IP camera, record, configure camera parameters, take photos, and control zoom and focus with the joystick. But if you prefer, it can be used only to control the zoom and focus with the joystick, using a secondary stream from the IP camera, with very low image quality, so as not to consume bandwidth and view the camera through qgroundcontrol. I prefer the first option, with a minimized screen of qgroundcontrol to see the navigation parameters.

For communication with surface I use FathomX up to 200m of tether, with more tether the data rate usually drops and is not enough for decent 4k streaming.

When the distance is greater, I use an ethernet conversion board with VDSL2 protocol. For example, the images that I have shared from the DAHUA 12 mp camera are transmitted live through 500m of 2x2x24awg copper umbilical, installed in a motorized spool that I have designed for the Marine Research Center of the University of Vigo CIM. (Attached photo and vídeo.)

Regarding the DAHUA vs HIK image quality, I prefer DAHUA 12 mp. In my ROV I have the previous version of the camera that I share in my previous post, and the image is excellent. But I think the new model is much better. You can see the quality of my ROV camera in the following links (in video settings select 4k).

With respect to larger and quality cameras, there is a wide variety such as blackmagic, cinema4k, Zcam and of course the ones you mention, but the size forces large casings and makes the ROVs not very maneuverable in difficult conditions such as sea currents, etc. But it is evident that everything depends on the use that you are going to give to the equipment.

In my latest design, I have tried not to give up image quality, integrating the DAHUA 12mp camera, in a compact ROV with great maneuverability and good response in strong currents, with a wide working range at great depth, thanks to a resistant and flexible tether of 500 m, thus avoiding fiber optics. On the other hand, I have incorporated UGPS, a 14.8V 36Ah battery customized for prolonged dives and a rotating arm with a servo of my own design. I think the result is a hard-working, maneuverable ROV with very interesting image quality.

I hope I have helped.

Best regards


I’m glad to hear that people are working towards a better quality camera. The below average quality of the current BlueROV2 camera is definitely the weakest link of the system. I have had to abort a few dives because of how bad the image quality was in 6-foot visibility, but my GoPro was able to see reasonably good.

1 Like

Hi Andres,

Thanks again for explaining about your upgrades and the impressive setup, very inspiring!

I agree that the lens operation of the HIK cameras is quite slow so changing the lens to a faster one (if I understood you correctly) is a great idea. It seems like you managed to make a compact design to fit in the dome, I wonder if you separated the DSP and other parts of the camera to achieve that.

As for other cameras, I used Zcam and also mirrorless cameras to achieve very high quality images and videos indeed their size makes them less attractive for our cause here. The Dahua video the you presented is already excellent enough though. I’m glad to see this level of video on still not so high bit rate (<16Mbps).

These videos here are from our stationary cameras using HIK zoom cameras, they do a decent job as long as only a small part of the frame is changing:

It’s good to have this type of conversation here, thanks for the contribution of knowledge Andres and thanks Bluerobotics for this forum :slight_smile:

1 Like

Hi @EliotBR,
Thanks for the info!
Good to know that its something that you work on, it’s safe to say that probably all BlueROV2 users are looking forward to this upgrade. Until then we’re working on our own solutions.

Just a thought that I haven’t verified myself yet; is it possible that the current USB camera simply doesn’t sit exactly on the focal plane in regards to the dome? Its video on land seems quite good but once in the water it degrades significantly, I referred this mostly to the low light conditions underwater but I recently thought that it might has to do with its position in the dome but I have tried to move it back or fwd yet, I wonder if you guys tried this already.

Fair enough :slight_smile:

A dome has a focal point, not a focal plane. I expect the camera mount should have been designed to try to get the camera’s focal point as close as possible to the dome’s focal point, but it’s not impossible that it’s a bit off, especially since there have been multiple supported cameras over time, and the system as designed needs the camera to be able to rotate (so it may have been offset some amount to enable that - I’m not sure).

It may be worth adjusting the camera’s lens to try to improve the alignment if you notice that it’s unexpectedly poor. I found that to be helpful at the previous company I was working for.

1 Like

Will give it a try. Thanks.