@TE-KarlKomierowski Thanks for the advice Karl. It will likely not be with color.. I just used these beads as examples .. to see if it can be done easily. thought that if the pellets are changed to other items, neural networks can be used to train new objects and models can be replaced to identify new items. .
Unfortunately there is no built in support for recording to a video file on the SD card. There are two camera data streams available: one for taking still pictures and one for showing a preview video like you have in a "regular" digital camera. No support for actual video recording.
Now, it doesn't say exactly what your project is supposed to do or what you want to use the video for, but perhaps it would be possible to generate the actual video file as a post processing step? This might work If you don't intend to use the actual video file on the Spresense board itself after recording. In that case you could save the frames from the preview video as separate image files on the SD card instead of displaying them on a display etc. Then, when the recording is finished, plug the SD card into your PC instead and combine them into a video file using an external tool such as ffmpeg.
This workaround might not work for you - it depends on what you want to accomplish - but it is the simplest solution I can come up with.
Nice to hear about your ideas! Unfortunately I am not able to help out with the actual project design so I will just give some generic recommendations to help you get started
As a first step I'd suggest you have a look at the sample applications available in the Spresense SDK and/or Arduino library if you haven't already. Getting an idea of how audio works in Spresense and what it can do might help you decide the scope and direction of your project. The audio system might seem a bit complicated at first, but start playing around a bit with the samples to get a feel for it. There are several audio related apps in both the Arduino and SDK development environments, but remember that they are just examples. Don't let their scope limit you!
Then you can start to think about what sampling devices you think are suitable for your project and figure out what adaptations might be needed to connect them to the Spresense mic inputs.
@TE-KarlKomierowski yes you are totally right, it would be a lot easier, but the company I work in, want everything in the cheapest way !!!!
Ok, I understand.
I don't know if it is of any help but the schematics for the extension board is open and free to download. So if you prototype with the extension board you should be able to copy the parts of the design that suits your needs. It could save time doing so (and money) for your company.
Just a comment, you can ignore this if it is not feasible for you.
Sorry, has taken me time to get back to this problem, once I make the suggested change the error I got previously no longer appears when I compile the code. I now get other issues but that appears to be due to board differences.
Hello....The OmniVision OV48C is a 48MP picture sensor with enormous 1.2-micron pixel size. This makes it atypically huge as the various 48MP picture sensors have a 0.8-micron pixel size. The higher pixel size is on the grounds that the sensor is truly greater. It has a 1/1.3″ sensor size, while the IMX586 and the Samsung GM1/GM2 have a 1/2″ sensor size.
@TE-KarlKomierowski That would be awesome! Yes anything that you may have will be helpful. i have been trying to port over a driver i wrote using the AT-cmd interface and some parsing ect. But my main struggle is integrating into nutx of course..
I am just having trouble with getting the driver i made to function. my hope was i could see how the code you had used worked and maybe clue me in on my mistakes.
My primary issue is getting the serial pins to communicate with the ESP. no matter what i include from the nutx library i cannot get serial communication between the AT and the ESP.
I'v been using standard 115200 8 none 1 settings. And i've noticed the issue both on native MacOs (using screen) and with Ubunutu via Parallells (using minicom). I originally thought it was a parallels issue but it seems its "my mac". I'll see if I get different results on another machine with USB 3. If this has not been reported by anyone else on a recent USB 3 machine then I guess it's something local to my setup, or some weird incompatibity with the Mac USB devices. (I have experienced issues with USB 2 devices requiring firmware upgrades to work with USB 3).
You can capture 8 channels digital simultaneously at 19khz 24bit. But if you want to record the sounds, please watch the bus bandwidth to the storage. When you record 8 channels of 192khz 24bit, you need to have the bandwidth of 192kHz x 32bit x 8channels = 6.144 MBytes/sec to the storage. In many cases, the bandwidth of SD cards is not stable to keep the performance due to the mechanism of the file system. If you want to capture stable sounds, you need to measure the actual performance of the storage that you are using. Please check the throughput from microphones to your storage beforehand.
It depends on the system. If you will use the same characteristic of digital microphones, I believe it can keep the phase coherence.
Please see the connection of digital microphones and Spresense below:
As you see, Spresense provides the same clock to digital microphones. Please see the capture timing of microphones below:
The data at the same timing from microphones are serialized at the same time. It means if the outputs from the digital microphones are generated at the same time based on the clock provided from Spresense, it will keep the phase between the microphones.