@Alex-Kushleyev alright, thanks for the help!

@jimbow77 and @oasisartisan ,

TOF V1 (which is now EOL, so you don't have it on Starling 2 Max) was set up the following way:

The calibration data was downloaded the first time the sensor was used and stored in /data/misc/camera/, then re-used for next time calibration files include pmd.spc, tof_cal_eeprom.bin and irs10x0c_lens.cal - the latter likely containing the intrinsics calibraiton data (although i have not checked explicitly)

TOF V2 (what is shipping in Starling 2 (and Max) ), is set up so that the TOF Library downloads (via i2c) the calibration data each time the voxl-camera-server starts. This is done internally to the TOF libraries. I don't know exactly why there is a difference between how the calib data is handled for two sensors by the TOF software, but this is how our sensor vendor helped us set it up. The TOF processing software downloads the calibration from the sensor each time voxl-camera-server starts and applies it to generate the point cloud. It may be possible to get the TOF intrinsics using the TOF library API, but we have not checked.

Do you have a specific reason why the TOF intrinsics are needed?

If you wanted to calibrate the TOF sensor yourself, then you would either need a checkerboard some material that is IR reflective ("white") and non-reflective ("black"). Alternatively, you could use an array of IR leds (or IR reflective dots + IR light) and use dot pattern detector as opposed for checkerboard for calibration.

Regarding Hi-res cameras, the intrinsics are not calibrated and not included. It can be calibrated using our camera calibration app (voxl-camera-calibration), but you should do it at half resolution or smaller (not full size like 4056x3040), otherwise the app runs very slowly. If you need more details how to calibrate at high resolution, let me know (involves downscaling exactly by a factor of 2 and then upscaling the resulting intrinsics)

Alex

Hi @Haxerus ,

I will double check regarding enabling J10 in spi mode on VOXL2 Mini.

Can you please confirm that you did all 3 changes:

the DTS changes you listed below used the older devcfg.mbn file from SDK 1.0.0 configure the level shifters to SPI mode using voxl-gpio -w 67 1

Can you please clarify which second i2c channel was used in your test for the 2nd lepton?

Thanks!

Alex

Carson—

I'm trying to do the same: create a Kicad symbol for VOXL 2 as a module (ie, no internal details, just interfaces). Have you had any luck in locating something?

Best,

—John

@Cliff-Wong the drone is on SDK 1.5.0, I switched to open vins, and the vio portal reflected the movement, but when i toggled my controller from manual mode to position mode, the drone crashed for the first fight, on the second and third trial the drone kept landing, when i tried the fourth time, the drone finally maintained the height and hovered. Why do I have that unpredictable behavior ? I have attached images of my controller to show the toggling i meant.

Kindly advice.
Thanks. WhatsApp Image 2026-02-13 at 10.53.46 AM.jpeg WhatsApp Image 2026-02-13 at 10.53.46 AM (1).jpeg

@Haxerus , that's great to hear! Can you clarify what you did? I am guessing the following:

modify voxl-lepton-server to run multiple instances disable i2c interface for the second lepton

Also, the image on the left seems noisy (horizontal lines)? I am not sure what it is, something to check.. (is it always present, what about when only single lepton is running).

Do you need anything else for any next steps?

Alex

Is more information needed? Is there an issue with my request?

@Zachary-Lowell-0

Before factory reset, I am wondering if you have any comment...

@zauberflote1 I am no longer having this issue. I have been flying with OpenVins for some time, and I was sometimes seeing drift in the vio estimate, and I was wondering what are the best configs to use. for vio_cams.conf and voxl-open-vins-server.conf?

@Zachary-Lowell-0 I would first like to diagnose what is causing the yolov8 model to not work on the voxl 2 for me. Will it only work when you train the model/ export it to tflite on a TPU? I am getting the issue even if I train the yolov8 model on an open source dataset/ use the pretrained yolov8n.pt model downloaded from ultralytics. I want to make sure I can train a yolov8 model on an open source dataset from scratch works on the voxl, so that I can move on to using my custom dataset.

In case there is no other workaround then I should be able to send you the dataset I am using.

Thanks!

@Cliff-Wong. For completeness below is the output of the start of voxl-vision-hub --debug_offboard. I trimmed the entire output above until the drone started to move from the plan_to command.

voxl2:~$ voxl-vision-hub --debug_offboard VOXL FLIGHT CONTROLLER (VFC)
always_on 0
vfc_rate 100.000000
vfc_rc_chan_min 980
vfc_rc_chan_max 2020
vfc_thrust_ch 3
vfc_roll_ch 1
vfc_pitch_ch 2
vfc_yaw_ch 4
vfc_submode_ch 6
vfc_backtrack_seconds : 60
vfc_backtrack_rc_chan : 10
vfc_backtrack_rc_thresh : 1500
vfc_alt_mode_rc_min 0
vfc_alt_mode_rc_max 0
vfc_flow_mode_rc_min 1700
vfc_flow_mode_rc_max 2100
vfc_hybrid_flow_mode_rc_min 0
vfc_hybrid_flow_mode_rc_max 0
vfc_position_mode_rc_min 1300
vfc_position_mode_rc_max 1700
vfc_traj_mode_rc_min 0
vfc_traj_mode_rc_max 0
vfc_yaw_deadband 30
vfc_vxy_deadband 50
vfc_vz_deadband 150
vfc_min_thrust 0.000000
vfc_max_thrust 0.800000
vfc_tilt_max 0.436000
vfc_yaw_rate_max 3.000000
vfc_thrust_hover 0.250000
vfc_vz_max 1.000000
vfc_kp_z 5.000000
vfc_ki_z 0.500000
vfc_ki_z_max 0.300000
vfc_kd_z 4.000000
vfc_vxy_max 3.000000
vfc_kp_xy 0.640000
vfc_kd_xy 2.560000
vfc_kp_z_vio 5.000000
vfc_ki_z_vio 0.700000
vfc_kd_z_vio 5.000000
vfc_kp_xy_vio 6.000000
vfc_kd_xy_vio 4.000000
vfc_w_filt_xy_vio 10.000000
vfc_w_filt_xy_flow 3.000000
vfc_vel_ff_factor_vio 0.500000
vfc_xy_acc_limit_vio 3.000000
vfc_max_z_delta 3.000000
vfc_att_transition_time 0.500000
vfc_stick_move_threshold 30.000000
vfc_flow_transition_time 1.000000
vfc_q_min 15
vfc_points_min 0
vfc_en_submode_announcement 1
vfc_en_backtrack_heading_hold 0
vfc_disable_fallback 0 loading our own config file

Parameters as loaded from config file:
config_file_version: 1

MAVROS / MAVSDK
en_localhost_mavlink_udp 0
localhost_udp_port_number: 14551

VIO
en_vio: 1
vio_pipe: qvio
secondary_vio_pipe: ov
en_reset_vio_if_initialized_inverted: 1
vio_warmup_s: 3.000000
send_odom_while_failed: 1

MISC FEATURES
horizon_cal_tolerance: 0.500000
en_hitl: 0
OFFBOARD MODE
offboard_mode: trajectory
follow_tag_id: 0
figure_eight_move_home: 1
tracking_trigger_ch: 12
tracking_trigger_ch_thresh_temp: 1200
tracking_trigger_ch_thresh_new: 1800
wps_move_home: 1
wps_timeout: 0.000000
wps_damp: 1.000000
wps_vfc_mission: true
wps_vfc_mission_loop: false
wps_vfc_mission_to_ramp: 25.000000
wps_vfc_mission_cruise_speed: 1.000000
wps_vfc_mission_to_kp: 0.100000
robot_radius: 0.300000
collision_sampling_dt: 0.100000
max_lookahead_distance: 1.000000
FIXED FRAME RELOCALIZATION
en_tag_fixed_frame: 0
fixed_frame_filter_len: 5
en_transform_mavlink_pos_setpoints_from_fixed_frame:0

COLLISION PREVENTION (VOA)
en_voa: 1
voa_upper_bound_m: -0.150000
voa_lower_bound_m: 0.150000
voa_memory_s: 1.000000
voa_max_pc_per_fusion: 100
voa_pie_min_dist_m: 0.250000
voa_pie_max_dist_m: 20.000000
voa_pie_under_trim_m: 1.000000
voa_pie_threshold: 3
voa_pie_slices: 36
voa_pie_bin_depth_m: 0.150000
voa_send_rate_hz: 20.000000

voa_input #0
enabled: 1
type: point_cloud
input_pipe: dfs_point_cloud
frame: stereo_l
max_depth: 8.000000
min_depth: 0.300000
cell_size: 0.080000
threshold: 4
x_fov_deg: 68.000000
y_fov_deg: 56.000000
conf_cutoff: 0
voa_input #1
enabled: 1
type: point_cloud
input_pipe: stereo_front_pc
frame: stereo_front_l
max_depth: 8.000000
min_depth: 0.300000
cell_size: 0.080000
threshold: 4
x_fov_deg: 68.000000
y_fov_deg: 56.000000
conf_cutoff: 0
voa_input #2
enabled: 1
type: point_cloud
input_pipe: stereo_rear_pc
frame: stereo_rear_l
max_depth: 8.000000
min_depth: 0.300000
cell_size: 0.080000
threshold: 4
x_fov_deg: 68.000000
y_fov_deg: 56.000000
conf_cutoff: 0
voa_input #3
enabled: 1
type: tof
input_pipe: tof
frame: tof
max_depth: 6.000000
min_depth: 0.150000
cell_size: 0.080000
threshold: 3
x_fov_deg: 106.500000
y_fov_deg: 85.099998
conf_cutoff: 125
voa_input #4
enabled: 1
type: rangefinder
input_pipe: rangefinders
frame: body
max_depth: 8.000000
min_depth: 0.300000
cell_size: 0.080000
threshold: 4
x_fov_deg: 68.000000
y_fov_deg: 56.000000
conf_cutoff: 0

loading extrinsics config file
loading horizon cal file
existing instance of voxl-vision-hub found, attempting to stop it
starting geometry module
starting autopilot monitor
starting mavlink IO
Waiting to connect to voxl-mavlink-server
Connected to voxl-mavlink-server
starting fixed pose input
starting vio manager
Connected to voxl-mavlink-server
Connected to VIO pipe: ov
Geometry module updating to use imu: imu_apps_body for VIO
ERROR in vcc_fetch_extrinsic, failed to find desired extrinsic relation in file
Likely you need to run voxl-configure-extrinsics to make a new file
ERROR: /etc/modalai/extrinsics.conf missing body to imu_apps_body, sticking with identity for now
Detected Autopilot Mavlink SYSID 1
requesting autopilot_version
Detected autopilot version: 1.14.0
starting tag manager
starting voa manager
ERROR in pipe_client_open, channel 11 already running
Connected to VOA input pipe: tof
starting horizon cal module
starting imu manager
starting state manager
starting offboard trajectory
Connected to voxl-mapper
voxl-mapper pipe size is: 65536 bytes
Init complete
Trajectory Monitor connected to voa pointcloud
commanding: XYZ -1.0 -0.1 -2.3 yaw: 0.1
WARNING in VOA manager, no attitude data from px4

@joshuagaus Sorry, but SBUS is not supported. That option will be hidden in future SDK releases to prevent confusion.

@Myles-Levine Nevermind, I got it working through the voxl interface instead of Qgroundcontrol. Thank you.

@ramp I have the same issue. Can I ask you how did you solve it ?

@Alex-Kushleyev

The drone is wired up according to the diagram, could you explain what you mean by the pin1 location?

Here is the output for voxl-camera-server -l

starling (D0005):~$ voxl-camera-server -l existing instance of voxl-camera-server found, attempting to stop it DEBUG: Attempting to open the hal module DEBUG: SUCCESS: Camera module opened on attempt 0 DEBUG: ----------- Number of cameras: 1 DEBUG: Cam idx: 0, Cam slot: 3, Slave Address: 0x00E2, Sensor Id: 0x7750 DEBUG: Note: This list comes from the HAL module and may not be indicative DEBUG: of configurations that have full pipelines DEBUG: Number of cameras: 1 ====================================

Here is the contents of /usr/lib/camera

starling (D0005):~$ ls -l /usr/lib/camera total 6612 -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.eeprom.cmk_imx577.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.ar0144.so -rwxr-xr-x 1 root root 5808 Oct 22 03:46 com.qti.sensor.boson.so -rwxr-xr-x 1 root root 5816 Oct 22 03:46 com.qti.sensor.cmk_imx577.so -rwxr-xr-x 1 root root 5816 Oct 22 03:46 com.qti.sensor.cmk_imx577_flip.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.imx214.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.imx214_flip.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.imx335.so -rwxr-xr-x 1 root root 5808 Oct 22 03:46 com.qti.sensor.imx377.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.imx412_fpv.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.imx664.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.imx678.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.imx678_flip.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.imx678_mod.so -rwxr-xr-x 1 root root 5808 Oct 22 03:46 com.qti.sensor.irs1645.so -rwxr-xr-x 1 root root 5808 Oct 22 03:46 com.qti.sensor.irs2975c.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.ov64b40.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.ov7251.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov7251_front_left.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov7251_front_left_flip.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov7251_front_right.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov7251_front_right_flip.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov7251_fsin.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov7251_fsout.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov7251_rear_left.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov7251_rear_left_flip.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov7251_rear_right.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov7251_rear_right_flip.so -rwxr-xr-x 1 root root 9904 Oct 22 03:46 com.qti.sensor.ov9782.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov9782_front_left.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov9782_front_left_flip.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov9782_front_right.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov9782_front_right_flip.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov9782_rear_left.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov9782_rear_left_flip.so -rwxr-xr-x 1 root root 9912 Oct 22 03:46 com.qti.sensor.ov9782_rear_right.so -rwxr-xr-x 1 root root 9920 Oct 22 03:46 com.qti.sensor.ov9782_rear_right_flip.so -rw-r--r-- 1 root root 149272 Feb 10 17:39 com.qti.sensormodule.imx214_2.bin -rw-r--r-- 1 root root 674736 Feb 10 17:39 com.qti.sensormodule.irs1645_0.bin -rw-r--r-- 1 root root 142472 Feb 10 17:39 com.qti.sensormodule.ov7251_3.bin -rw-r--r-- 1 root root 957344 Oct 22 03:46 com.qti.tuned.cmk_imx577.bin -rw-r--r-- 1 root root 835464 Oct 22 03:46 com.qti.tuned.cmk_ov9282.bin -rw-r--r-- 1 root root 931864 Oct 22 03:46 com.qti.tuned.default.bin -rw-r--r-- 1 root root 957344 Oct 22 03:46 com.qti.tuned.imx412_fpv.bin -rw-r--r-- 1 root root 793816 Oct 22 03:46 com.qti.tuned.ov9782.bin -rw-r--r-- 1 root root 882376 Oct 22 03:46 com.qti.tuned.sony_imx335.bin drwxr-xr-x 2 root root 4096 Oct 22 04:04 components starling (D0005):~$

@jameskuesel Thanks! Merged