Color space settings for the Lightroom user

[bjanes]

Not that this 'fact' was backed up with a lick of colorimetry. Just the opposite, mostly disproven with colorimetry. Working on a video to show how and why.
But the problem is, people who can't back any of this up keep writing such text, others read it and believe it. Then spew  the same nonsense onto others.

Andrew,

I noted that your sRGB video was published on Feb 9. Your investigations do demonstrate that there is no significant difference in color values when rendering into sRGB or ProPhotoRGB using you test parameters. Towards the end of the presentation you recommended rendering into 16 bit ProPhoto but it was not clear to me what bit depth you used for the bulk of your experiments. With 8 bit ProPhoto one would expect more rounding errors since the color steps between one bit increments in rendered values are greater in ProPhotoRGB than sRGB.

Regards,

Bill

[digitaldog]

Towards the end of the presentation you recommended rendering into 16 bit ProPhoto but it was not clear to me what bit depth you used for the bulk of your experiments. With 8 bit ProPhoto one would expect more rounding errors since the color steps between one bit increments in rendered values are greater in ProPhotoRGB than sRGB.

All the colorimetric data in the video was done using 16-bit output. I did do a test using 8-bit per color and have to dig up the dE report but it wasn't much different if memory serves.

[JRSmit]

Andrew where to find your video?

[adias]

Andrew where to find your video?

Here: http://digitaldog.net/files/sRGBMyths.mp4

[digitaldog]

With 8 bit ProPhoto one would expect more rounding errors since the color steps between one bit increments in rendered values are greater in ProPhotoRGB than sRGB.

Yes, after the initial rendering to each, from high bit data. I still have to dig up the tests and dE reports doing this when setting LR to save out 24-bit data. But I suspect it will produce the same results (unless Dither is added which Photoshop would do and LR doesn't with high bit export). After the rendering, yes, you would ideally want to edit the wider gamut, any gamut data, in high bit.

[JRSmit]

Andrew a through test procedure in your video. It concurrentie with a us museums initiative study much like the dutch metamorfoses.  Their Conclusion was that the encoding in any of the 3 usually color spaceshuttle was insignificant for the quality. The engineering used however does make a significant difference.

[adias]

Here: http://digitaldog.net/files/sRGBMyths.mp4

Great presentation!

[orc73]

+1 on this and it goes for everything in life and also in realy life advise:
https://www.youtube.com/watch?v=Z1Tmhit277c

The thing is, stupid people don't know they are stupid.

I have been retired for a few years and have devoted a significant amount of that time focused on learning what it means to be a photographer compared to the younger me who recorded life events with his camera and occasionally scored a 'hit' with a great sunset.

A lot of my learning frustration has been associated with the topic of this thread - garbage advise from self-proclaimed 'experts'.  It is one of the great unintended consequences of the internet (adding confusion rather than clarity).  Early on, I was pretty naive about accepting internet advice.  The advise confused me cause I couldn't separate the wheat from the chaff. It would send me in circles and added significant time and frustration to the very steep learning curve I have been trying to climb.

Today, I am extremely choosy about where I go and who I listen to on the internet.  A lot of my leads have come from Adobe and this forum (which I discovered by accident).  I then determine if the lead has the KSE to attract a publisher.  These authors are the people I rely upon to get me further up the curve.  Still steep, still a long way to go, but now I have a greater sense of confidence that the routes I am following are technically sound.