Skip to content
Permalink

Comparing changes

Choose two branches to see what’s changed or to start a new pull request. If you need to, you can also or learn more about diff comparisons.

Open a pull request

Create a new pull request by comparing changes across two branches. If you need to, you can also . Learn more about diff comparisons here.
base repository: ngscopeclient/scopehal
Failed to load repositories. Confirm that selected base ref is valid, then try again.
Loading
base: dc57f4465eee
Choose a base ref
...
head repository: ngscopeclient/scopehal
Failed to load repositories. Confirm that selected head ref is valid, then try again.
Loading
compare: 2c4aec96a72c
Choose a head ref
  • 1 commit
  • 1 file changed
  • 1 contributor

Commits on Dec 13, 2020

  1. Major improvements to BaseMeasurement accuracy

    @azonenberg
    azonenberg committed Dec 13, 2020
    Copy the full SHA
    2c4aec9 View commit details
Showing with 57 additions and 29 deletions.
  1. +57 −29 scopeprotocols/BaseMeasurement.cpp
86 changes: 57 additions & 29 deletions scopeprotocols/BaseMeasurement.cpp
View file
Original file line number Diff line number Diff line change
@@ -112,14 +112,14 @@ void BaseMeasurement::Refresh()
size_t len = din->m_samples.size();

//Make a histogram of the waveform
float min = GetMinVoltage(din);
float max = GetMaxVoltage(din);
float vmin = GetMinVoltage(din);
float vmax = GetMaxVoltage(din);
size_t nbins = 64;
vector<size_t> hist = MakeHistogram(din, min, max, nbins);
vector<size_t> hist = MakeHistogram(din, vmin, vmax, nbins);

//Set temporary midpoint and range
m_range = (max - min);
m_midpoint = m_range/2 + min;
m_range = (vmax - vmin);
m_midpoint = m_range/2 + vmin;

//Find the highest peak in the first quarter of the histogram
//This is the base for the entire waveform
@@ -134,51 +134,79 @@ void BaseMeasurement::Refresh()
}
}
float fbin = (idx + 0.5f)/nbins;
float global_base = fbin*m_range + min;
float global_base = fbin*m_range + vmin;

//Create the output
auto cap = new AnalogWaveform;

float last = min;
int64_t tedge = 0;
float sum = 0;
int64_t count = 0;
float last = vmin;
int64_t tfall = 0;
float delta = m_range * 0.1;

float fmax = -99999;
float fmin = 99999;
float fmax = -FLT_MAX;
float fmin = FLT_MAX;

bool first = true;

vector<float> samples;

for(size_t i=0; i < len; i++)
{
//Wait for a falling edge
//Wait for a rising edge (end of the low period)
float cur = din->m_samples[i];
int64_t tnow = din->m_offsets[i] * din->m_timescale;

//Find falling edge
if( (cur < m_midpoint) && (last >= m_midpoint) )
tfall = tnow;

//Find rising edge
if( (cur > m_midpoint) && (last <= m_midpoint) )
{
//Done, add the sample
if(count != 0)
if(!samples.empty())
{
float vavg = sum/count;
if(vavg > fmax)
fmax = vavg;
if(vavg < fmin)
fmin = vavg;

cap->m_offsets.push_back(tedge);
cap->m_durations.push_back(tnow - tedge);
cap->m_samples.push_back(vavg);
if(first)
first = false;

else
{
//Average the middle 50% of the samples.
//Discard beginning and end as they include parts of the edge
float sum = 0;
int64_t count = 0;
size_t start = samples.size()/4;
size_t end = samples.size() - start;
for(size_t j=start; j<=end; j++)
{
sum += samples[j];
count ++;
}

float vavg = sum / count;

fmax = max(fmax, vavg);
fmin = min(fmin, vavg);

int64_t tmid = (tnow + tfall) / 2;

//Update duration for last sample
size_t n = cap->m_samples.size();
if(n)
cap->m_durations[n-1] = tmid - cap->m_offsets[n-1];

cap->m_offsets.push_back(tmid);
cap->m_durations.push_back(1);
cap->m_samples.push_back(vavg);
}

samples.clear();
}
tedge = tnow;
}

//If the value is fairly close to the calculated base, average it
//TODO: discard samples on the rising/falling edges as this will skew the results
if(fabs(cur - global_base) < delta)
{
count ++;
sum += cur;
}
samples.push_back(cur);

last = cur;
}