Running a random set of colored LEDs from the Basement Laboratory Parts Warehouse Wing through the LED Curve Tracer produced this pleasant plot:
The white LED doesn’t match up with either the blue or the UV LED. Perhaps the blue LED uses a completely different chemistry that shoves further to the right than seems proper? I suppose I should run a handful of white, blue, and UV LEDs through the thing just to see what’s going on…
The Bash / Gnuplot source code:
#!/bin/sh
numLEDs=8
#-- overhead
export GDFONTPATH="/usr/share/fonts/truetype/"
base="${1%.*}"
echo Base name: ${base}
ofile=${base}.png
echo Input file: $1
echo Output file: ${ofile}
#-- do it
gnuplot << EOF
#set term x11
set term png font "arialbd.ttf" 18 size 950,600
set output "${ofile}"
set title "${base}"
set key noautotitles
unset mouse
set bmargin 4
set grid xtics ytics
set xlabel "Forward Voltage - V"
set format x "%4.1f"
set xrange [0.5:4.5]
#set xtics 0,5
set mxtics 2
#set logscale y
#set ytics nomirror autofreq
set ylabel "Current - mA"
set format y "%3.0f"
set yrange [0:35]
set mytics 2
#set y2label "right side variable"
#set y2tics nomirror autofreq 2
#set format y2 "%3.0f"
#set y2range [0:200]
#set y2tics 32
#set rmargin 9
set datafile separator whitespace
set label 1 "IR" at 1.32,32 center
set label 2 "R" at 1.79,32 center
set label 3 "O" at 2.10,32 center
set label 4 "Y" at 2.65,32 center
set label 5 "G" at 2.42,32 center
set label 6 "B" at 4.05,32 center
set label 7 "UV" at 3.90,32 center
set label 8 "W" at 3.25,32 center
#set arrow from 2.100,32 to 2.125,31 lt 1 lw 2 lc 0
plot \
"$1" index 0 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "red" ,\
"$1" index 1 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "orange" ,\
"$1" index 2 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "dark-yellow" ,\
"$1" index 3 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "green" ,\
"$1" index 4 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "blue" ,\
"$1" index 5 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "purple" ,\
"$1" index 6 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "magenta" ,\
"$1" index 7 using (\$5/1000):(\$2/1000) with linespoints pt 1 lw 2 lc rgb "dark-gray"
EOF
And the raw data file:
# LED Curve Tracer # Ed Nisley - KE4ZNU - March 2013 # VCC at LED: 4897 mV # Bandgap reference voltage: 1041 mV # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 1 0 0 4892 3889 1002 0 0 0 3889 5 4613 4892 3264 1627 1990 48 1942 3216 10 10148 4892 3216 1675 2092 106 1985 3109 15 15223 4892 3182 1709 2199 159 2039 3022 20 19836 4892 3148 1743 2271 208 2063 2940 25 24910 4897 3129 1767 2354 261 2092 2867 30 30446 4897 3104 1792 2431 319 2111 2785 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 2 0 0 4892 3884 1007 0 0 0 3884 5 4613 4892 3124 1767 1985 48 1937 3075 10 9687 4897 3037 1860 2111 101 2010 2935 15 14761 4897 2964 1932 2189 155 2034 2809 20 19836 4897 2906 1990 2271 208 2063 2697 25 24910 4897 2848 2048 2349 261 2087 2586 30 30446 4892 2794 2097 2431 319 2111 2475 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 3 0 0 4892 3826 1065 0 0 0 3826 5 4613 4897 2862 2034 1990 48 1942 2814 10 10148 4897 2688 2208 2097 106 1990 2581 15 15223 4897 2552 2344 2194 159 2034 2392 20 19836 4892 2436 2455 2276 208 2068 2228 25 24910 4897 2349 2547 2354 261 2092 2087 30 29985 4897 2257 2639 2426 314 2111 1942 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 4 0 0 4892 3734 1157 0 0 0 3734 5 5074 4892 2935 1956 1976 53 1922 2882 10 10148 4897 2823 2073 2102 106 1995 2717 15 15223 4892 2722 2170 2199 159 2039 2562 20 20297 4897 2649 2247 2276 213 2063 2436 25 24910 4897 2567 2329 2349 261 2087 2305 30 29985 4897 2489 2407 2426 314 2111 2174 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 5 0 0 4892 4485 406 0 0 0 4485 5 4613 4897 1724 3172 1990 48 1942 1675 10 10148 4892 1443 3448 2097 106 1990 1336 15 15223 4897 1249 3647 2199 159 2039 1089 20 19836 4892 1099 3792 2276 208 2068 891 25 24910 4897 983 3913 2354 261 2092 721 30 29985 4892 862 4030 2426 314 2111 547 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 6 0 0 4892 4165 726 0 0 0 4165 5 5074 4892 1448 3443 1985 53 1932 1395 10 10148 4897 1322 3574 2102 106 1995 1215 15 15223 4892 1220 3671 2194 159 2034 1060 20 20297 4892 1147 3744 2276 213 2063 934 25 25372 4892 1075 3816 2354 266 2087 808 30 29985 4892 1002 3889 2426 314 2111 687 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 7 0 0 4892 4247 644 0 0 0 4247 5 5074 4892 3647 1244 1981 53 1927 3594 10 10148 4892 3618 1273 2107 106 2000 3511 15 14761 4892 3603 1288 2170 155 2015 3448 20 20297 4892 3584 1307 2271 213 2058 3371 25 25372 4892 3574 1317 2354 266 2087 3308 30 29523 4892 3565 1327 2412 310 2102 3255 # Insert LED, press button 1 to start... # INOM ILED VccLED VD VLED VG VS VGS VDS <--- LED 8 0 0 4892 4945 -53 0 0 0 4945 5 5074 4892 2160 2731 1985 53 1932 2107 10 10148 4892 2034 2857 2097 106 1990 1927 15 15223 4897 1927 2969 2194 159 2034 1767 20 19836 4892 1826 3066 2271 208 2063 1617 25 25372 4897 1734 3162 2349 266 2082 1467 30 29523 4892 1666 3225 2412 310 2102 1356 # Insert LED, press button 1 to start...
The Smell of Molten Projects in the Morning 
Is it odd that these appear roughly in spectrum order or is that just a coincidence?
The forward drop roughly depends on the oomph required to toss photons out of the junction, so it’s inversely proportional to wavelength… plus a host of other factors that account for the variations & reversals in that plot.
Sooo, although the white LED should be close to either the blue or UV LED, it ain’t!
Assuming the Wikipedia light-emitting-diode article is accurate (looks right, but LEDs aren’t my forte) there’s a fairly wide range for Vf in blue LEDs. There’s a bit of wiggle-room in the chemistry, and the white example you have is within the range claimed for a blue one, and the blue example is also in the range. As I recall, it was hell for the early green LEDs, because small changes in process variables tweaked the color and Vf, thus making for problems in matching. Not sure how tricky blues are, but I doubt that white LEDs care much about the color of the forcing light.
I vaguely recall some numeric (?) LEDs with a two-axis binning scheme sorting them by color and intensity-at-current. Fortunately, I didn’t really care, but if you were building a big multi-digit display, you’d care a lot.