OPT8241 Voxel Viewer calibration how to do it

I just realized you are already using a conf file, and your measured data looks correct. Your last two values corresponding to 4.2 and 5.0 metres may be wrong due to low returned signal. What is the amplitude at these points?

I'm sorry but these measures are referred to another profile (F1 = 50 MHz and F2 = 60 MHz) that I created.
I made the step of phase measuring by using your config file, dividing the measures as follows:

d [m] phase 1 phase 2
0,34 4095 4095
0,629833333 3634 1942
0,919666667 3871 2422
1,2095 21 2904
1,499333333 265 3380
1,789166667 495 3855
2,079 745 261
2,368833333 993 765
2,658666667 1218 1197
2,9485 1478 1703
3,238333333 1719 2178
3,528166667 1921 2678
3,818 2216 3163
4,107833333 2455 3642
4,397666667 2705 65
4,6875 2973 575

I put this in the voxel wizard and then restart the camera and voxel view, but the measure are wrong. The config file after this calibration is:

[global]
id = 143
name = Dealiased1
parent = 130

[defining_params]
unambiguous_range = 5
mix_volt = 1500
op_clk_freq = 0
sub_frame_cnt_max = 2
quad_cnt_max = 6
frame_rate = 30.0000743866
mod_freq1 = 30.0
mod_freq2 = 60.0
pix_cnt_max = 133333
ma = 1
mb = 2
ka = 3
kb = 1
freq_ratio = 2048
dealiased_ph_mask = 0
intg_duty_cycle = 20
illum_power_percentage = 100
delay_fb_corr_mode = 0
delay_fb_dc_corr_mode = 0
delay_fb_coeff_1 =
disable_temp_corr = 1
intg_time = 31

[calib]
freq_corr_at = 48.0
freq_corr = 1.0
actual_distance = 0.65
tillum_calib = 61
phase_corr_2 = 1973
phase_corr_1 = -1412
tsensor_calib = 56
phase_lin_coeff1 = 4095 4095 4095 3138 1975 2192 2450 2710 2968 3226 3482 3738 2809 880 446 718
phase_lin_coeff0 = 4095 4095 3641 3647 54 315 568 836 1093 1354 1623 1856 2144 2415 2683 2970
phase_lin_corr_period = 2


These are the measure after do the caibration
d real d voxel dealiased phase amplitude
0,6 3,598 182 315
1,2 3,42 175 95
1,8 2,43 146 45
2,4 3 157 24
3 4,13 214 13
3,6 3,55 183 10
4,2 79 4090 6

The distances of the 16 points choosen to do the calibration are correct?
Why I get 79 meters if my unambiguous range is of 5 meters?

I deleted this three line in the conf file:
phase_lin_coeff1 = 4095 4095 4095 3138 1975 2192 2450 2710 2968 3226 3482 3738 2809 880 446 718
phase_lin_coeff0 = 4095 4095 3641 3647 54 315 568 836 1093 1354 1623 1856 2144 2415 2683 2970
phase_lin_corr_period = 2

I do the phase offset calibration manually.
Before start the phase offset calibration I deleted the previously vaues in the parameters phase_corr_2 = 1973 and phase_corr_1 = -1412
I put an object at 0.65m and measure the two phases, the values are:

actual_distance = 0.65
phase_corr_2 = 1029
phase_corr_1 = 774

and the results are the follow:

d real d voxel phase amplitude
0,6 77,9 4024 108
1,2 18,02 931 28
1,8 37,47 1915 12
2,4 57,19 2947 6,5

But I do not understand why we must not make the non-linear calibration phase. The unambiguous range of LongRange profile is 8m, and this one is 5m, and we also changed the two frequencies.

what is wrong in this procedure?

Thank you to help me.

I'm sorry, I forgot to subtract the theoretical value.
Before measure the two fases I deleted the previous phase corr, is it correct?

without delte the previous phase correcct
distance = 0,65 m
phaseF1 = 2920 ----> phase_corr_1 = 2386
phaseF2 = 4068 ----> phase_corr_2 = 3002

If I set the two phase_corr before measure the offset I obtain:
distance = 0,65 m
phaseF1 = 1527 ----> phase_corr_1 = 993
phaseF2 = 1973 ----> phase_corr_2 = 907

with the secon values ( phase_corr_1 = 993 and phase_corr_2 = 907 ):

d real d voxel (d_voxel - 3.45 m) (d_voxel - 3.45 m) - d real deal phase
0,6 4,05 0,6 0 214
1,2 4,68 1,23 0,03 245
1,8 0,37 1,92 0,12 20
2,4 0,97 2,52 0,12 50
3 1,59 3,14 0,14 82
3,6 2,2 3,75 0,15 114
4,2 2,89 4,44 0,24 148
4,62 3,28 4,83 0,21 169

This point of the calibration procedure is correct?

How can I delete this offset of 3.45m?
The second thing is after correcting this initial offset (3.45m) I can correct the relative error ( (d_voxel - 3.45 m) - d real ) with the non linear calibration right?

Is exactly what I made. This is the conf file before do the offset calibration:

[global]
id = 148
name = 3060LAST
parent = 130

[defining_params]
unambiguous_range = 5
mix_volt = 1500
op_clk_freq = 0
sub_frame_cnt_max = 2
quad_cnt_max = 6
frame_rate = 30.0
mod_freq1 = 30.0
mod_freq2 = 60.0
pix_cnt_max = 133333
ma = 1
mb = 2
ka = 3
kb = 1
freq_ratio = 2048
dealiased_ph_mask = 0
intg_duty_cycle = 20
illum_power_percentage = 100
delay_fb_corr_mode = 0
delay_fb_dc_corr_mode = 0
delay_fb_coeff_1 =
disable_temp_corr = 1

[calib]
freq_corr_at = 48.0
freq_corr = 1.0
tillum_calib = 61
tsensor_calib = 56
#phasecorrection = file:TintinCDKCameraDealiased18mPhaseOffset.bin

I'm going to explain all the steps to do this.
After open Voxel viewer I see the phase_corr_1 and phase_corr_2 to 0 and then set the disable_offset_corr to true and finally set IND_FREQ_DATA_EN = 1 and IND_FREQ_DATA_SEL = 0/1

The two phases are:
phaseF1 phaseF2
202 3962

But when I try to apply this two coefficient:

phase_corr_1 = -332
phase_corr_2 = 2896

I got an error to the phase_corr_2:
DepthCamera: Could not set value 2896 for parameter TintinCDKCamera(0::0451:9105::xxxxxxxx).phase_corr_2
Can you help me? I guess that we are near the solution

wow, it's working many many thanks.

These are the measures:

d real d voxel
0,65 0,66
1,2 1,2
1,8 1,81
2,4 2,38
3 2,98
3,6 3,61
4,2 4,25

This is the solution, because I tried with a different set of F1 and F2 and all this profile are working. My problem was to do the calibration phase of the automatic offset through voxel viewer.

Now if I want to compensate for differences along the range I perform the non-linear phase calibration right?

Thank you. Today I will try to perform the Common Phase Offset calibration through the Calibration wizard and I will write the results.

After doing the manual phase offset calibration I tried with a voxel wizard, and I have to do this for 5 or 6 times to obtain acceptable phase compensation values, I don't know why, but finally I can considering this problem solved.
When I try to see the depth map I get some problem when I have opaque object (see the allegate picture dl.dropboxusercontent.com/.../Problem.png ) in my scenario. The measurement by means of software is greater than 8 meters while the real measure is 3 meters. But if I get close to this object under 2 meters I get the right distance. Can I solve this problem? or maybe is a limit of this board?

thank you

Dear Gianluca&Anand,

how can i computed the phase_corr1/2 theoretical value? Can you give me example?

I think i am almost near the solution.

thanks&regards

Dear Gianluca,

thanks for your reply.

You explain to me help is very big, very detailed.

But I have another question I hope you can help me, how to How to manually calculate Phase_corr1/2.

Assume that the second step measured phase as Phase_measured_F1, the Third step measured phase as Phase_measured_F2,

actual_distance‘s theoretical phase as Phase_theoretical , so

Phase_corr1=Phase_measured_F1 -Phase_theoretical;
Phase_corr2=Phase_measured_F2 -Phase_theoretical;

I want to know whether my understanding is correct?Because I am in this way to calculate the Phase_corr1/2,my d_voxel is not right. Looking forward to get your guidance.

Thanks&Regards

Dear Gianluca

thanks for you information.

I wonder why not Phase theoretical = actual_distance*ma*mb * 2^12 / (Unambiguous_range*2^(5-dealiased_ph_mask) ,because it is working on aliased stated?

thanks&regards

Dear Gianluca,

thanks for you reply.

maybe you are right ,because in my formula the d real is not equal to d voxel .but you data looks like right.

thanks&regards.

None

perfect.

Gianluca

Dear Gianluca,

"For my opinion I suggest you to set the frame statistic mean to 100 frame or upper."

1、Can you tell me how can i get statistic mean to 100 frame or upper data by voxel viewer?

2、it should be right：Phase_corr1/2=Phase_theoretical - Phase_measured_F1/2 ; please check it?

thanks&regards.

None

Dear None,

for the first question you just need to set in settings check the temporal statistics and set 100 measures or more.

For the second is it correct for me. I do extactly this formula

Dear Gianluca,

Phase_corr1/2=Phase_theoretical - Phase_measured_F1/2 not Phase_measured_F1/2 - Phase_theoretical , maybe it I didn't say clear.

I will get the 200 frame data by myself not voxviewer.

Thanks&Regards.

None

Dear Gianluca,

Today I did experiments, obtain a 200 frames of data by my self.

this is conf file:
[global]
id = 130
parent = 128
name = Long Range

[defining_params]
unambiguous_range = 5
mix_volt = 1500
op_clk_freq = 0
sub_frame_cnt_max = 2
quad_cnt_max = 6
frame_rate = 30.0
mod_freq1 = 30.0
mod_freq2 = 60.0
pix_cnt_max = 133333
ma = 1
mb = 2
ka = 3
kb = 1
freq_ratio = 2048
dealiased_ph_mask = 0
intg_duty_cycle = 20
illum_power_percentage = 100
delay_fb_corr_mode = 0
delay_fb_dc_corr_mode = 0
delay_fb_coeff_1 =
disable_temp_corr = 1

[calib]
freq_corr_at = 48.0
freq_corr = 1.0
actual_distance = 1.812

tillum_calib = 43
disable_temp_corr = 0
disable_offset_corr = 0
phase_corr_1 = -40
phase_corr_2 = -733
tsensor_calib = 51

the Phase1_measure=1524.67;
the Phase2_measure=2217.671;

the acutal distance=1.812, Phase_theoretical=1.812*2^12/5=1484.39. So I get

the Phase_corr1=1484.39-1524.67=-40;
the Phase_corr2=2217.671-1524.67=-733;

After this I get the distance calibration value;

d_real d_measure
1.052 0.958
1.384 1.286
1.758 1.707
2.19 2.13
2.482 2.413
2.982 2.909

I want to know what went wrong? Can you help me ?

Thanks&Regards

None

after porforming crosstalk calibration incorretly, i find the data bad ,what can i do to recover the original set???