Math behind

alesh

Well-Known Member
Engauge Digitizer is free and a good alternative (IMO better) to DigitizeIt (you can't export csv in free version)...

Great tutorial, thanks!
I stole a full version of DigitizeIt so the export is not an issue. However, I agree that it's not a perfect program. I actually tried multiple pieces of software for digitizing back in the day (about 4 years ago) and none of them I'd considered good. Stuck with DigitizeIt ever since as it was the least painful to work with. At least for me.
This one seems very recent so I'd like to give it a shot but I can't get it working - missing MSVCP140.dll when I try to start it.
 

Mathay

Member
I can't get it working - missing MSVCP140.dll when I try to start it.
It's because Digitizer have been compiled without static linking for a smaller size.
You have to install VC++ redistribuable 2015

Smalls tips for Digitizer :

1°) use original image, not filtered one
2°) in configuration > export format for Y interpolation, use the third choice like following screenshot (french interface in this case but the english one is the same), and semi-colon as separator.

upload_2017-3-25_10-16-53.png
 

alesh

Well-Known Member
It's because Digitizer have been compiled without static linking for a smaller size.
You have to install VC++ redistribuable 2015
That didn't help, unfortunately. I downloaded Microsoft Visual C++ 2015 Redistributable Update 3 (from this link) but I'm still getting the same error.
 

Mathay

Member
Which Windows OS is installed on your PC?
On my Windows 7 x64, both x32 and x64 versions of Visual C++ 2015 Redistributable are installed.
x32 version of VC++ 2015 Redistribuable is required for Digitizer on an x64 Windows OS.
 

alesh

Well-Known Member
Which Windows OS is installed on your PC?
On my Windows 7 x64, both x32 and x64 versions of Visual C++ 2015 Redistributable are installed.
x32 version of VC++ 2015 Redistribuable is required for Digitizer on an x64 Windows OS.
Windows 7 x64. Didn't have the 32bit Visual C++ 2015 Redistributable, though. It's working now, thank you. Will give it a try.
 

alesh

Well-Known Member
So I did try it and I gotta say I like it a lot. Still browsing options as there's quite a few of them. Only negative would be stability, the program crashed 4 times in about 2 hours. You involved in the development @Mathay ?
 

Mathay

Member
So I did try it and I gotta say I like it a lot. Still browsing options as there's quite a few of them. Only negative would be stability, the program crashed 4 times in about 2 hours. You involved in the development @Mathay ?
No, I did not participate...
I also noticed some crashes, always in the same case (when discarding a curve).
I'll open an issue on https://github.com/markummitchell/engauge-digitizer/issues as soon as I took a look on closed issues and source code ;)
It's a promising but young project, it still needs maturity and feedback
 

alesh

Well-Known Member
I experienced crashes in more situations including the curve removal. Also some minor bugs. I'll be happy to post some feedback but I don't know how.
 

NoFucks2Give

Well-Known Member
I'm quite fallible sometimes so if you catch a mistake let me know please!
These numbers do not look correct.
The digitized numbers, are you converting them from Luminance to quantum PPFD?
Common sense says that 4 or 5 µMoles/J is way too high. 2 would be very good. 5 is like printing your own photons.

I cannot see the formulas in the spreadsheet.

These are my formulas, and tested against values measured with a radiospectrometer
PPFD = ((lumens / 683) x (wavelength x 0.00836)) / CIE
Where CIE is the Photopic Luminous Efficacy, Relative Sensitivity Curve for the C.I.E. Standard Observer, which the spreadsheet labels SPD*luminosity

Converting to Photon Radiance (µMoles) to Luminance (lm/m²/sr) commonly called lumens.
Luminance = PPFD x CIE / (wavelength x 0.00836) x 683

Where I see the problem is with the digitized numbers being treated as Luminance. The lumens in an LED datasheet are Illuminance. But that not the real issue. The SPD is Irradiance which is radiometric not luminous. So it appears you are converting W/m² (irradiance) to moles/m²/s (photon radiance) thinking you are converting lm/m²/sr (luminance) to µMoles (radiance).

When converting radiometric watts to quantum µMoles (or vise versa) you only need to remove difference in photon energy due to wavelength. Luminous efficiency and photon energy are not relevant to PPFD.

Assuming the SPD is in watts, to convert the SPD numbers to PPFD the formula would be:
PPFD (µMoles) = watts x wavelength x 0.00836

I charted the measurements in Watts and Lumens from my radiospectrometer and the watts matched the SPD perfectly.

freshFocusRedMeatSpectraWatts.jpg
freshFocusRedMeatSpectraLux.jpg
freshFocusRedMeatSpectral PowerDistributionCharacteristics.jpg
 

NoFucks2Give

Well-Known Member
Between 4 and 5 is typical at 100% efficiency
What I noticed was the math looked suspicious. I could not see the formulas in the spreadsheet, so not sure.
I did the math and the numbers in the spreadsheet did not match any of my numbers. I know my formulas are correct.
I tested measured values against calculated. Here: http://growlightresearch.com/ppfd/convert.html
There are three sections
1 converted 100 lm to Watts and PPFD for wavelengths 400-699nm
$ppfd = ((100 / 683) * ($wavelength * 0.00836)) / $cie[$wavelength];
$watts = $ppfd / 0.00836 / $wavelength;

2 converted 10 µMole to Lumens and Watts
$lux2 = 10 * $cie[$wavelength] / ($wavelength * 0.00836) * 683 ;
$watts2 = 10 / 0.00836 / $wavelength;


3 Used measured values of Watts, µMoles and Lumens (M-PPF, M-Lux, M-Watts) from a Luxeon Fresh Focus Red Meat CoB and compared to my calculated conversions from Lumens (C-Watts C-PPF)


I just measured 5 CoBs today. Measured Lux and PAR.
22,000 lm => 340 µMole Luxeon Core Range High Density 2700K 90 CRI 90 lm/W
20,000 lm => 325 µMole Luminus CLM-9 Gen3 2700K 90CRI 117 lm/W
15,000 lm => 256 µMole Luxeon Crisp White 3000K 90 CRI 93 lm/W
5,200 lm => 83 µMole Cree CXA1304 3000K 93 CRI 86 lm/W
4,700 lm => 76 µMole Cree CXA1304 2700K 93 CRI 75 lm/W


The values for $cie[$wavelength] in the conversions
Photopic Luminous Efficacy, Relative Sensitivity Curve for the C.I.E. Standard Observer
390 0.00012
391 0.0001476
392 0.0001752
393 0.0002028
394 0.0002304
395 0.000258
396 0.0002856
397 0.0003132
398 0.0003408
399 0.0003684
400 0.000396
401 0.0004774
402 0.0005588
403 0.0006402
404 0.0007216
405 0.000803
406 0.0008844
407 0.0009658
408 0.0010472
409 0.0011286
410 0.00121
411 0.001489
412 0.001768
413 0.002047
414 0.002326
415 0.002605
416 0.002884
417 0.003163
418 0.003442
419 0.003721
420 0.004
421 0.00476
422 0.00552
423 0.00628
424 0.00704
425 0.0078
426 0.00856
427 0.00932
428 0.01008
429 0.01084
430 0.0116
431 0.01274
432 0.01388
433 0.01502
434 0.01616
435 0.0173
436 0.01844
437 0.01958
438 0.02072
439 0.02186
440 0.023
441 0.0245
442 0.026
443 0.0275
444 0.029
445 0.0305
446 0.032
447 0.0335
448 0.035
449 0.0365
450 0.038
451 0.0402
452 0.0424
453 0.0446
454 0.0468
455 0.049
456 0.0512
457 0.0534
458 0.0556
459 0.0578
460 0.06
461 0.063098
462 0.066196
463 0.069294
464 0.072392
465 0.07549
466 0.078588
467 0.081686
468 0.084784
469 0.087882
470 0.09098
471 0.095784
472 0.100588
473 0.105392
474 0.110196
475 0.115
476 0.119804
477 0.124608
478 0.129412
479 0.134216
480 0.13902
481 0.14592
482 0.15282
483 0.15972
484 0.16662
485 0.17352
486 0.18042
487 0.18732
488 0.19422
489 0.20112
490 0.20802
491 0.219518
492 0.231016
493 0.242514
494 0.254012
495 0.26551
496 0.277008
497 0.288506
498 0.300004
499 0.311502
500 0.323
501 0.34033
502 0.35766
503 0.37499
504 0.39232
505 0.40965
506 0.42698
507 0.44431
508 0.46387333333333
509 0.48343666666667
510 0.503
511 0.5237
512 0.5444
513 0.5651
514 0.5858
515 0.6065
516 0.6272
517 0.6479
518 0.6686
519 0.6893
520 0.71
521 0.7252
522 0.7404
523 0.7556
524 0.7708
525 0.786
526 0.8012
527 0.8164
528 0.8316
529 0.8468
530 0.862
531 0.8712
532 0.8804
533 0.8896
534 0.8988
535 0.908
536 0.9172
537 0.9264
538 0.9356
539 0.9448
540 0.954
541 0.958095
542 0.96219
543 0.966285
544 0.97038
545 0.974475
546 0.97857
547 0.982665
548 0.98676
549 0.990855
550 0.99495
551 0.99596
552 0.99697
553 0.99798
554 0.99899
555 1
556 0.999
557 0.998
558 0.997
559 0.996
560 0.995
561 0.9907
562 0.9864
563 0.9821
564 0.9778
565 0.9735
566 0.9692
567 0.9649
568 0.9606
569 0.9563
570 0.952
571 0.9438
572 0.9356
573 0.9274
574 0.9192
575 0.911
576 0.9028
577 0.8946
578 0.8864
579 0.8782
580 0.87
581 0.8587
582 0.8474
583 0.8361
584 0.8248
585 0.8135
586 0.8022
587 0.7909
588 0.7796
589 0.7683
590 0.757
591 0.7444
592 0.7318
593 0.7192
594 0.7066
595 0.694
596 0.6814
597 0.6688
598 0.6562
599 0.6436
600 0.631
601 0.6182
602 0.6054
603 0.5926
604 0.5798
605 0.567
606 0.5542
607 0.5414
608 0.5286
609 0.5158
610 0.503
611 0.4908
612 0.4786
613 0.4664
614 0.4542
615 0.442
616 0.4298
617 0.4176
618 0.4054
619 0.3932
620 0.381
621 0.3694
622 0.3578
623 0.3462
624 0.3346
625 0.323
626 0.3114
627 0.2998
628 0.2882
629 0.2766
630 0.265
631 0.256
632 0.247
633 0.238
634 0.229
635 0.22
636 0.211
637 0.202
638 0.193
639 0.184
640 0.175
641 0.1682
642 0.1614
643 0.1546
644 0.1478
645 0.141
646 0.1342
647 0.1274
648 0.1206
649 0.1138
650 0.107
651 0.1024
652 0.0978
653 0.0932
654 0.0886
655 0.084
656 0.0794
657 0.0748
658 0.0702
659 0.0656
660 0.061
661 0.0581
662 0.0552
663 0.0523
664 0.0494
665 0.0465
666 0.0436
667 0.0407
668 0.0378
669 0.0349
670 0.032
671 0.0305
672 0.029
673 0.0275
674 0.026
675 0.0245
676 0.023
677 0.0215
678 0.02
679 0.0185
680 0.017
681 0.016121
682 0.015242
683 0.014363
684 0.013484
685 0.012605
686 0.011726
687 0.010847
688 0.009968
689 0.009089
690 0.00821
691 0.0077992
692 0.0073884
693 0.0069776
694 0.0065668
695 0.006156
696 0.0057452
697 0.0053344
698 0.0049236
699 0.0045128
700 0.004102
701 0.0039009
702 0.0036998
703 0.0034987
704 0.0032976
705 0.0030965
706 0.0028954
707 0.0026943
708 0.0024932
709 0.0022921
710 0.002091
711 0.0019866
712 0.0018822
713 0.0017778
714 0.0016734
715 0.001569
716 0.0014646
717 0.0013602
718 0.0012558
719 0.0011514
720 0.001047
721 0.0009943
722 0.0009416
723 0.0008889
724 0.0008362
725 0.0007835
726 0.0007308
727 0.0006781
728 0.0006254
729 0.0005727
730 0.00052
731 0.0004929
732 0.0004658
733 0.0004387
734 0.0004116
735 0.0003845
736 0.0003574
737 0.0003303
738 0.0003032
739 0.0002761
740 0.000249
741 0.0002361
742 0.0002232
743 0.0002103
744 0.0001974
745 0.0001845
746 0.0001716
747 0.0001587
748 0.0001458
749 0.0001329
750 0.00012
751 0.000114
752 0.000108
753 0.000102
754 9.6E-5
755 9.0E-5
756 8.4E-5
757 7.8E-5
758 7.2E-5
759 6.6E-5
760 6.0E-5
761 5.7E-5
762 5.4E-5
763 5.1E-5
764 4.8E-5
765 4.5E-5
766 4.2E-5
767 3.9E-5
768 3.6E-5
769 3.3E-5
770 3.0E-5
 
Last edited:

Rahz

Well-Known Member
Those CIE #s are slightly different than the spreadsheet but not enough to affect the results in the way you describe.

Have you tried your formula on well known SPDs? The CXB 3000K 80CRI is known to be 4.65 (400-700nm).
 
Top