Which has the highest CRI? (Just for fun)

22 posts / 0 new
Last post
clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia
Which has the highest CRI? (Just for fun)
1
13% (9 votes)
2
20% (14 votes)
3
20% (14 votes)
4
47% (33 votes)
Total votes: 70
Edited by: clemence on 01/09/2019 - 13:18
clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia
raccoon city
raccoon city's picture
Offline
Last seen: 25 min 16 sec ago
Joined: 10/06/2010 - 02:35
Posts: 13315
Location: रॅकून सिटी Palm Desert CA USA

Why link to a post when you can just copy it so no one has to follow the link?  ;)

clemence wrote:

Just for fun Quiz!

Which one has the highest CRI?
Only geometry correction, NO color editing whatsoever simply RAW to JPEG conversion, 5600K WB balance

My quess, nobody would see the differences since our monitors are still limited to RGB Amoled or LED. In the real world it’s so much easier to discern.
I’ll give the answer later…. happy guessing!

To prevent anyone think I would cheat, here’s the link to the answers: https://drive.google.com/open?id=1wWGvSrdfIve9Q4EohP3B6sFzbal9w_q4
Will give the open password later.

*Please vote HERE

- Clemence

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

1. Didn’t plan this quiz. Already started there. The comments and opinions are very interesting and spark interesting topics.
2. I want all the discussions grouped in there. To make new visitors know what we were discussing

- Clemence

raccoon city
raccoon city's picture
Offline
Last seen: 25 min 16 sec ago
Joined: 10/06/2010 - 02:35
Posts: 13315
Location: रॅकून सिटी Palm Desert CA USA

Who can explain to me which rule was broken in post #3?  :FACEPALM:

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

raccoon city wrote:

Who can explain to me which rule was broken in post #3?  alt=:" />


Please tell me RC

- Clemence

raccoon city
raccoon city's picture
Offline
Last seen: 25 min 16 sec ago
Joined: 10/06/2010 - 02:35
Posts: 13315
Location: रॅकून सिटी Palm Desert CA USA

clemence wrote:
raccoon city wrote:

Who can explain to me which rule was broken in post #3?  :FACEPALM:

Please tell me RC - Clemence

Someone reported my post, and I don't have the foggiest idea why.

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

raccoon city wrote:

clemence wrote:
raccoon city wrote:

Who can explain to me which rule was broken in post #3?  alt=:" />


Please tell me RC – Clemence

Someone reported my post, and I don’t have the foggiest idea why.


Definitely not me RC because I knew it was a good intention. I rarely use that button. I usually just post my expression or simply use PM. You can ask SB.
Few days ago I accidentally clicked that button while tried to push the post number (“#xxx”) at far right corner of the post box. That could be your case too

- Clemence

raccoon city
raccoon city's picture
Offline
Last seen: 25 min 16 sec ago
Joined: 10/06/2010 - 02:35
Posts: 13315
Location: रॅकून सिटी Palm Desert CA USA

clemence wrote:

Definitely not me RC because I knew it was a good intention. I rarely use that button. I usually just post my expression or simply use PM.

I'm glad it wasn't you.

clemence wrote:

Few days ago I accidentally clicked that button while tried to push the post number ("#xxx") at far right corner of the post box. That could be your case too - Clemence

I don't think so.

After clicking the report button, you have to confirm that you really want to report the post.

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

Well, I clicked the spam button. Off course I clicked back button immediately. I don’t know if that counts or not.

- Clemence

mdeni
mdeni's picture
Offline
Last seen: 3 days 22 hours ago
Joined: 08/06/2016 - 05:17
Posts: 271
Location: Europe

Do we need more votes? The test is very onesided.

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

mdeni wrote:
Do we need more votes? The test is very onesided.
Agree, looks like more votes wouldn’t yield different result. I’ll reveal the answer once the electricity goes back. It’s been 30 minutes out so far.

- Clemence

ToyKeeper
ToyKeeper's picture
Offline
Last seen: 1 hour 39 min ago
Joined: 01/12/2013 - 14:40
Posts: 9992
Location: (469219) 2016 HO3

I was really hoping #4 would turn out to be a low-CRI light. It’s the obvious “correct” answer because the white shades in its image are actually white with equal values for R, G, and B. But the other three have some tint in their white areas — #1 is greenish while #2 and #3 are rosy.

But looking at the revealed answers, it doesn’t really fit the photos. Based on the measured Duv, #2 should be the greenest by far, but its image instead has a pink tint. And in person, #1 should look blue, but in the photo it doesn’t.

Usually when I photograph low-CRI lights with auto white balance, it comes out pure white. And when I photograph high-CRI lights with auto white balance, they tend to have slightly off-white shades. The result says more about the white balance algorithm than it says about the emitters.

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

ToyKeeper wrote:
I was really hoping #4 would turn out to be a low-CRI light. It’s the obvious “correct” answer because the white shades in its image are actually white with equal values for R, G, and B. But the other three have some tint in their white areas — #1 is greenish while #2 and #3 are rosy.

But looking at the revealed answers, it doesn’t really fit the photos. Based on the measured Duv, #2 should be the greenest by far, but its image instead has a pink tint. And in person, #1 should look blue, but in the photo it doesn’t.

Usually when I photograph low-CRI lights with auto white balance, it comes out pure white. And when I photograph high-CRI lights with auto white balance, they tend to have slightly off-white shades. The result says more about the white balance algorithm than it says about the emitters.

Wait, I’ll publish the RAW files for anyone to check it. I used DxO Optics software for the conversion along with it’s built-in color rendering index profile. Stitching done with ACDsee.
EDIT: pic #2 is not rosy, it’s brownish actually

- Clemence

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

Why I can’t post google drive links in here? I always returned back to blank post

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia

Ah, no more than a single link

Here you go TK:
https://drive.google.com/open?id=1mge2dx6qvCSkYAWcQ6ER1GLScdYjawxC

- Clemence

Boaz
Boaz's picture
Offline
Last seen: 4 hours 29 min ago
Joined: 11/07/2010 - 09:31
Posts: 7158
Location: Birthplace of Aviation

I'm voting for all the above .. because that would just be sneaky .

 

#5

       καὶ τὸ φῶς ἐν τῇ σκοτίᾳ φαίνει καὶ ἡ σκοτία αὐτὸ οὐ κατέλαβεν

                            

       Dc-fix diffuser film  >…  http://budgetlightforum.com/node/42208

mdeni
mdeni's picture
Offline
Last seen: 3 days 22 hours ago
Joined: 08/06/2016 - 05:17
Posts: 271
Location: Europe

This was more of a TINT question since the CRI is very high and indistinguishable.

But the difference in tint is very very evident. Ant that is why the optisolis won hands down. That is some amazing tint neutrality. Amazing LED.

ToyKeeper
ToyKeeper's picture
Offline
Last seen: 1 hour 39 min ago
Joined: 01/12/2013 - 14:40
Posts: 9992
Location: (469219) 2016 HO3
clemence wrote:
pic #2 is not rosy, it’s brownish actually

Using the white section in the upper-right corner of the image, blurred to get an average color, the values I’m seeing are:

  1. R,G,B = 233, 238, 232
  2. R,G,B = 246, 243, 238
  3. R,G,B = 243, 239, 234
  4. R,G,B = 233, 233, 233

Normalized to 100%, the values are:

  1. 97.9, 100, 97.4 (greenish)
  2. 100, 98.8, 96.7 (pinkish)
  3. 100, 98.4, 96.2 (pinkish)
  4. 100, 100, 100 (pure white)

This is why I assumed #4 would get the most votes… because it’s actually white. However, I also guessed that #2 or #3 would be the ones with the highest CRI because lights with high R9 values usually show up a bit pink in pictures.

If #2 has a +0.0038 Duv and #3 has a -0.0011 Duv, but they have nearly identical pink-ish tints in photos, that basically means there was an error on the camera or in post-processing. Accurate color temperature correction would have produced different results for #2… something like this:

  1. 97.9, 100, 97.4 (greenish)
  2. 94.0, 100, 94.0 (green)
  3. 100, 98.4, 96.2 (pinkish)
  4. 100, 100, 100 (pure white)

… and in this case, the 92.5 CRI emitter in photo #2 would produce worse photos than the 74.8 CRI emitter in photo #1, because its tint is more green.

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia
ToyKeeper wrote:
clemence wrote:
pic #2 is not rosy, it’s brownish actually

Using the white section in the upper-right corner of the image, blurred to get an average color, the values I’m seeing are:

  1. R,G,B = 233, 238, 232
  2. R,G,B = 246, 243, 238
  3. R,G,B = 243, 239, 234
  4. R,G,B = 233, 233, 233

Normalized to 100%, the values are:

  1. 97.9, 100, 97.4 (greenish)
  2. 100, 98.8, 96.7 (pinkish)
  3. 100, 98.4, 96.2 (pinkish)
  4. 100, 100, 100 (pure white)

This is why I assumed #4 would get the most votes… because it’s actually white. However, I also guessed that #2 or #3 would be the ones with the highest CRI because lights with high R9 values usually show up a bit pink in pictures.

If #2 has a +0.0038 Duv and #3 has a -0.0011 Duv, but they have nearly identical pink-ish tints in photos, that basically means there was an error on the camera or in post-processing. Accurate color temperature correction would have produced different results for #2… something like this:

  1. 97.9, 100, 97.4 (greenish)
  2. 94.0, 100, 94.0 (green)
  3. 100, 98.4, 96.2 (pinkish)
  4. 100, 100, 100 (pure white)

… and in this case, the 92.5 CRI emitter in photo #2 would produce worse photos than the 74.8 CRI emitter in photo #1, because its tint is more green.

Thanks for the thorough inspection TK. I did nothing but take the pictures, convert, stitch, and upload. You can check the EXIF data to double check if it’s the camera or the post processing that cause errors. I took each picture twice using the camera built in double exposure. This is the only quick way to get even lighting without making a dedicated lightbox. I only use 5600K WB because this is what I’ve been using since the first time I uploaded pictures in BLF. I like consistency
If anyone think I must redo the supposed to be “Just for Fun” quiz, I’ll do it no problem. Didn’t mean to misled anyone, it is what it is. Don’t forget to include monitor limitation into consideration.

- Clemence

ToyKeeper
ToyKeeper's picture
Offline
Last seen: 1 hour 39 min ago
Joined: 01/12/2013 - 14:40
Posts: 9992
Location: (469219) 2016 HO3
clemence wrote:
I only use 5600K WB because this is what I’ve been using since the first time I uploaded pictures in BLF. I like consistency

… those all used 5600K white balance?

The differences should look pretty dramatic unless they were re-balanced at some point. There’s no way a white sheet actually turns out pure white when the light source is 4852K and the camera is 5600K, even if the CRI is 100.

clemence wrote:
If anyone think I must redo the supposed to be “Just for Fun” quiz, I’ll do it no problem.

No, of course not. Smile

It’s just kind of a recurring topic with a long history of debate. Some people insist CRI can only be seen in person, others insist it can be seen even after passing through a camera, a 3-channel image file, the internet, and a 3-channel screen.

Me, I think CRI is pretty noticeable in person but basically impossible to represent accurately onscreen.

Er, unless a spectrometer is used, I mean. That allows us to get around the 3-channel limit.

clemence
clemence's picture
Offline
Last seen: 12 hours 13 min ago
Joined: 07/12/2015 - 02:58
Posts: 2340
Location: Bali - Indonesia
ToyKeeper wrote:
clemence wrote:
I only use 5600K WB because this is what I’ve been using since the first time I uploaded pictures in BLF. I like consistency

… those all used 5600K white balance?

The differences should look pretty dramatic unless they were re-balanced at some point. There’s no way a white sheet actually turns out pure white when the light source is 4852K and the camera is 5600K, even if the CRI is 100.

clemence wrote:
If anyone think I must redo the supposed to be “Just for Fun” quiz, I’ll do it no problem.

No, of course not. Smile

It’s just kind of a recurring topic with a long history of debate. Some people insist CRI can only be seen in person, others insist it can be seen even after passing through a camera, a 3-channel image file, the internet, and a 3-channel screen.

Me, I think CRI is pretty noticeable in person but basically impossible to represent accurately onscreen.

Er, unless a spectrometer is used, I mean. That allows us to get around the 3-channel limit.

That’s exactly what I said in the beginning of the quiz. I was kind of surprised the picture showed the real highest CRI. Because to me it’s the tint that made it stands out, not the color vividness. Only #1 pic easy to guess, the rest are near identical except for the tint. Yes, all shot in 5600K WB. The information can be found in the EXIF RAW files like this:

JFIFVersion – 1.02
X-Resolution – 1 dpi
Y-Resolution – 1 dpi
Image Description – OLYMPUS DIGITAL CAMERA
Make – OLYMPUS IMAGING CORP.
Orientation – Horizontal (normal)
Software – ACDSee Ultimate 8
Date and Time (Modified) – 2019:01:09 22:49:24
YCbCr Positioning – Centered
ISO Speed – 200
Sensitivity Type – Standard Output Sensitivity
Exif Version – 0230
Date and Time (Original) – 2019:01:09 22:27:34
Date and Time (Digitized) – 2019:01:09 22:27:34
Exposure Bias – 0 EV
Max Aperture Value – 3.5
Subject Distance – 0.24 m
Metering Mode – Multi-segment
Light Source – Other
Sub Sec Time – 523
Flashpix Version – 0100
Color Space – sRGB
File Source – Digital Camera
CFAPattern – [Red,Green][Green,Blue]
Custom Rendered – Custom
Exposure Mode – Manual
White Balance – Manual
Digital Zoom Ratio – 1
Scene Capture Type – Standard
Gain Control – Low gain up
Contrast – Low
Saturation – Normal
Sharpness – Soft
Lens Info – 12-50mm f/3.5-6.3
Lens Model – OLYMPUS M.12-50mm F3.5-6.3
Special Mode – Normal, Sequence: 0, Panorama: (none)
Camera ID – OLYMPUS DIGITAL CAMERA
Equipment Version – 0100
Camera Type2 – E-M5
Serial Number – BEL201385
Internal Serial Number – 4138309002686001
Focal Plane Diagonal – 21.6 mm
Body Firmware Version – 2.000
Lens Type – Olympus M.Zuiko Digital ED 12-50mm F3.5-6.3 EZ
Lens Serial Number – ABP383943
Lens Model – OLYMPUS M.12-50mm F3.5-6.3
Lens Firmware Version – 1.102
Max Aperture At Min Focal – 3.5
Max Aperture At Max Focal – 6.3
Min Focal Length – 12
Max Focal Length – 50
Max Aperture – 6.2
Lens Properties – 0xc140
Extender – None
Extender Firmware Version – 0
Flash Type – None
Flash Model – None
Flash Firmware Version – 0
Camera Settings Version – 0100
Preview Image Valid – Yes
Preview Image Start – 49801
Preview Image Length – 804837
Exposure Mode – Manual
AELock – Off
Metering Mode – ESP
Exposure Shift – 0
Macro Mode – Off
Focus Mode – MF; MF
Focus Process – AF Not Used; 0
AFSearch – Not Ready
AFAreas – none
AFPoint Selected – undef undef undef undef
AFFine Tune – Off
AFFine Tune Adj – 0 0 0
Flash Mode – Fill-in
Flash Exposure Comp – 0
Flash Remote Control – Off
Flash Control Mode – Off; 0; 0; 0
Flash Intensity – undef undef undef undef
Manual Flash Strength – undef undef undef undef
White Balance2 – Custom WB 5400K
White Balance Temperature – 5600
White Balance Bracket – 0 0
Custom Saturation – 0 (min -5, max 5)
Modified Saturation – Off
Contrast Setting – -1 (min -5, max 5)
Sharpness Setting – -1 (min -5, max 5)
Color Space – sRGB
Scene Mode – Standard
Noise Reduction – (none)
Distortion Correction – Off
Shading Compensation – Off
Gradation – High Key; User-Selected
Picture Mode – Muted; 2
Picture Mode Saturation – 0 (min -2, max 2)
Picture Mode Contrast – 0 (min -2, max 2)
Picture Mode Sharpness – 0 (min -2, max 2)
Picture Mode BWFilter – n/a
Picture Mode Tone – n/a
Noise Filter – Off
Art Filter – Off; 0; 0; 0
Picture Mode Effect – Standard
Tone Level – -31999; Highlights 0; -7; 7; -31998; Shadows 0; -7; 7; 0; 0; 0; 0
Art Filter Effect – Off; 0; 0; 0; No Effect; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0
Drive Mode – Single Shot
Panorama Mode – Off
Image Stabilization – On, Mode 1
Manometer Pressure – 0 kPa
Manometer Reading – 0 m, 0 ft
Extended WBDetect – Off
Raw Dev Version – 0100
Raw Dev Exposure Bias Value – 0
Raw Dev White Balance Value – 0
Raw Dev WBFine Adjustment – 0
Raw Dev Gray Point – 0 0 0
Raw Dev Saturation Emphasis – 0 0 0
Raw Dev Memory Color Emphasis – 0
Raw Dev Contrast Value – 0 0 0
Raw Dev Sharpness Value – 0 0 0
Raw Dev Color Space – sRGB
Raw Dev Engine – High Speed
Raw Dev Noise Reduction – (none)
Raw Dev Edit Status – Original
Raw Dev Settings – (none)
Image Processing Version – 0112
WB_ RBLevels – 256 256 256 256
WB_ RBLevels3000 K – 158 456 256 256
WB_ RBLevels3300 K – 188 362 256 256
WB_ RBLevels3600 K – 200 332 256 256
WB_ RBLevels3900 K – 210 308 256 256
WB_ RBLevels4000 K – 250 360 256 256
WB_ RBLevels4300 K – 226 298 256 256
WB_ RBLevels4500 K – 240 298 256 256
WB_ RBLevels4800 K – 242 274 256 256
WB_ RBLevels5300 K – 256 256 256 256
WB_ RBLevels6000 K – 280 230 256 256
WB_ RBLevels6600 K – 288 242 256 256
WB_ RBLevels7500 K – 304 200 256 256
WB_ RBLevels CWB1 – 200 400 256 256
WB_ RBLevels CWB2 – 248 294 256 256
WB_ RBLevels CWB3 – 196 218 256 256
WB_ RBLevels CWB4 – 196 218 256 256
Color Matrix – 392 -108 -28 -60 372 -56 6 -104 354
Black Level2 – 0 0 0 0
Gain Base – 256
Valid Bits – 12 0
Crop Left – 8 0
Crop Top – 8 0
Crop Width – 4608
Crop Height – 3456
Sensor Calibration – 3706 256
Noise Reduction2 – (none)
Distortion Correction2 – Off
Shading Compensation2 – Off
Multiple Exposure Mode – On (2 frames); 1
Aspect Ratio – 4:3
Aspect Frame – 0 0 4607 3455
Faces Detected – 0 0 0
Face Detect Area – (Binary data 383 bytes, use -b option to extract)
Max Faces – 8 8 0
Face Detect Frame Size – 0 0 0 0 0 0
Face Detect Frame Crop – 0 0 0 0 0 0 0 0 0 0 0 0
Focus Info Version – 0100
Scene Detect – 0
Zoom Step Count – 27
Focus Step Count – 4335
Focus Step Infinity – 0
Focus Step Near – 0
Focus Distance – 0.24 m
AFPoint – Left (or n/a)
External Flash – Off
External Flash Bounce – Bounce or Off
External Flash Zoom – 0.9310298102
Internal Flash – Off
Manual Flash – Off
Macro LED – Off
Sensor Temperature – 43 C
Compression – JPEG (old-style)
Thumbnail Offset – 11383
Thumbnail Length – 8564
Camera ID – 72157629249027061
Camera Type – Mirrorless Camera