Pages: 1 [2] 3   Go Down

Author Topic: The "Joel Grimes Method".  (Read 4779 times)

rabanito

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1482
Re: The "Joel Grimes Method".
« Reply #20 on: July 07, 2021, 01:40:14 pm »

"Quod gratis asseritur, gratis negatur."   :)

Hmmmm...
I would rather say
"Do what works for you"
Logged

Alan Klein

  • Sr. Member
  • ****
  • Online Online
  • Posts: 13957
    • Flicker photos
Re: The "Joel Grimes Method".
« Reply #21 on: July 07, 2021, 02:15:06 pm »

Alan, I don't think it is noise... 32-bit allows you to make more aggressive adjustments on the colour sliders, especially the blue. If you have ever been too heavy on dragging the blue slider too far to the left you will know what happens to the sky.
What you are seeing - I think - is those artefacts which the 16-bit can't handle.

I will gladly bow to a superior theory!
Just as uprezing 2K to 4K only gives you an illusion of 4K, doesn't overlying 16-bit images give you an illusion of 32 bits or 48 bits rather than 16 bits. All you're doing is blending three different exposure levels, each of which have only 16 bits.  You can't make a silk purse from a sow's ear.
 

NikoJorj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1071
    • http://nikojorj.free.fr/
Re: The "Joel Grimes Method".
« Reply #22 on: July 07, 2021, 04:21:42 pm »

I would agree with Alan... 16 bits should be way more than enough to process an image destined to be displayed on a 8- or 10-bit device.
That's 6 bits of leeway, or roughly the ability to multiply the contrast by 64.
32bits is not that far from overkill.

What we may see here (if it's not a bug) is rather the effect of the merge to HDR routine : each image section benefits from the maximal S/N ratio in any of the 3 captures (I'd personally like to space them a bit more, -3/0/+3EV eg).
See http://www.guillermoluijk.com/article/virtualraw/index_en.htm written by Guillermo 13 years ago, many thanks to him!
Logged
Nicolas from Grenoble
A small gallery

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 17756
  • When everyone thinks the same, nobody thinks
    • My website
Re: The "Joel Grimes Method".
« Reply #23 on: July 08, 2021, 06:03:49 am »

... All I did in the above pics of the church was as described in the process above then I selected the 0.0EV photograph and synced the adjustments (with the 32-bit DNG) and ended up with the two examples I posted...

Ok... correct me if wrong... but it seems to me that the noisier image is a result of just one image (0.0EV) adjusted in Lightroom. If so, you are losing the benefits of 1. the HDR process selecting the +1.0 image (more light, less noise) for a good part of the final image and 2. averaging noise when blending the three images.

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 17756
  • When everyone thinks the same, nobody thinks
    • My website
Re: The "Joel Grimes Method".
« Reply #24 on: July 08, 2021, 06:05:19 am »

Hmmmm...
I would rather say
"Do what works for you"

Perhaps, but I was simply translating William's signature phrase to Latin (where it originated) :)

William Walker

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1076
    • William Walker Landscapes
Re: The "Joel Grimes Method".
« Reply #25 on: July 08, 2021, 06:17:19 am »

Ok... correct me if wrong... but it seems to me that the noisier image is a result of just one image (0.0EV) adjusted in Lightroom. If so, you are losing the benefits of 1. the HDR process selecting the +1.0 image (more light, less noise) for a good part of the final image and 2. averaging noise when blending the three images.

Try doing an HDR in Lightroom - which is 16-bit - and compare. You get the same "16-bit" result.
No question (in my mind - and apparently everyone else who has actually done the experiment) that the 32-bit is better.

As an aside: There is no noise in the 0.0EV image. I took it at midday in good light @ 100 ISO.
Logged
"What can be asserted without evidence can also be dismissed without evidence." Christopher Hitchens

William Walker

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1076
    • William Walker Landscapes
Re: The "Joel Grimes Method".
« Reply #26 on: July 08, 2021, 06:22:12 am »

Perhaps, but I was simply translating William's signature phrase to Latin (where it originated) :)

So, you aren't just a pretty face, I see?  8)
Logged
"What can be asserted without evidence can also be dismissed without evidence." Christopher Hitchens

rabanito

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1482
Re: The "Joel Grimes Method".
« Reply #27 on: July 08, 2021, 06:44:38 am »

Perhaps, but I was simply translating William's signature phrase to Latin (where it originated) :)
I knew Hitchens Razor but I didn't notice that that is William's signature (after the wording)
I know, it's all there... :-[
Chapeau, Slobodan
I'll put more attention next time
Logged

NikoJorj

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1071
    • http://nikojorj.free.fr/
Re: The "Joel Grimes Method".
« Reply #28 on: July 08, 2021, 12:21:42 pm »

Ok... correct me if wrong... but it seems to me that the noisier image is a result of just one image (0.0EV) adjusted in Lightroom. If so, you are losing the benefits of 1. the HDR process selecting the +1.0 image (more light, less noise) for a good part of the final image and 2. averaging noise when blending the three images.
Well spotted : the 32bit HDR should be compared with a 16bit HDR (made directly in LR, eg), a single frame will be noisier than both.
Logged
Nicolas from Grenoble
A small gallery

SrMi

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 292
Re: The "Joel Grimes Method".
« Reply #29 on: July 08, 2021, 03:10:39 pm »

I tried Joel's technique using his files and I got a really strange result.


Has anyone replicated his steps with the latest PS?
Logged

William Walker

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1076
    • William Walker Landscapes
Re: The "Joel Grimes Method".
« Reply #30 on: July 09, 2021, 02:44:03 am »

Well spotted : the 32bit HDR should be compared with a 16bit HDR (made directly in LR, eg), a single frame will be noisier than both.

Try doing an HDR in Lightroom - which is 16-bit - and compare. You get the same "16-bit" result.
No question (in my mind - and apparently everyone else who has actually done the experiment) that the 32-bit is better.

Logged
"What can be asserted without evidence can also be dismissed without evidence." Christopher Hitchens

William Walker

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1076
    • William Walker Landscapes
Re: The "Joel Grimes Method".
« Reply #31 on: July 09, 2021, 02:44:57 am »

I tried Joel's technique using his files and I got a really strange result.


Has anyone replicated his steps with the latest PS?

I agree, I gave up working on his files and used my own...
Logged
"What can be asserted without evidence can also be dismissed without evidence." Christopher Hitchens

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 17756
  • When everyone thinks the same, nobody thinks
    • My website
Re: The "Joel Grimes Method".
« Reply #32 on: July 09, 2021, 04:13:25 am »

Ok... will go out on a limb and say it: you can not possibly get so much cleaner and sharper results just by going 32 vs. 16. There must be an error in your comparison workflow.

mcbroomf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1231
    • Mike Broomfield
Re: The "Joel Grimes Method".
« Reply #33 on: July 09, 2021, 05:06:27 am »

Ok... will go out on a limb and say it: you can not possibly get so much cleaner and sharper results just by going 32 vs. 16. There must be an error in your comparison workflow.

The video shows and compares processing of files; 32 bit with HDR, 16 bit with HDR and 16 bit single shot.  So the discussion does need to be clear about what is being compared.
Logged

William Walker

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1076
    • William Walker Landscapes
Re: The "Joel Grimes Method".
« Reply #34 on: July 09, 2021, 06:01:32 am »

Ok... will go out on a limb and say it: you can not possibly get so much cleaner and sharper results just by going 32 vs. 16. There must be an error in your comparison workflow.

Well Slo, there's only one way to find out. Try it yourself, it will take you less than the time you've spent thinking and questioning all this!   8)

P.S. What is the Latin translation for "What can be asserted with evidence cannot be dismissed without evidence."
« Last Edit: July 09, 2021, 06:05:59 am by William Walker »
Logged
"What can be asserted without evidence can also be dismissed without evidence." Christopher Hitchens

Alan Klein

  • Sr. Member
  • ****
  • Online Online
  • Posts: 13957
    • Flicker photos
Re: The "Joel Grimes Method".
« Reply #35 on: July 09, 2021, 10:13:45 am »

Are the results of shooting three shots for every picture worth the time and effort and processing?  Or is this just a case of pixel peeping?

rabanito

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1482
Re: The "Joel Grimes Method".
« Reply #36 on: July 09, 2021, 10:39:30 am »

Are the results of shooting three shots for every picture worth the time and effort and processing?  Or is this just a case of pixel peeping?
If you get the results you want in this way, then it is worth it.
If you are satisfied with less, probably not.
Logged

William Walker

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1076
    • William Walker Landscapes
Re: The "Joel Grimes Method".
« Reply #37 on: July 09, 2021, 11:35:37 am »

Are the results of shooting three shots for every picture worth the time and effort and processing?  Or is this just a case of pixel peeping?

If you make large prints - which I do - it is worth it.
Logged
"What can be asserted without evidence can also be dismissed without evidence." Christopher Hitchens

Alan Klein

  • Sr. Member
  • ****
  • Online Online
  • Posts: 13957
    • Flicker photos
Re: The "Joel Grimes Method".
« Reply #38 on: July 09, 2021, 09:12:52 pm »

If you make large prints - which I do - it is worth it.
What size is where you can see the difference with three shots?

fdisilvestro

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1716
    • Frank Disilvestro
Re: The "Joel Grimes Method".
« Reply #39 on: July 10, 2021, 03:47:51 am »

Photoshop does behave differently with 32 bit images. Iím not sure if this is due more to the fact that 32bit uses floating point math vs 8/16 bitís integer math.

Anyway, a good demonstration of the differences between the editing capabilities of 32 bit images and 8/16 bit in photoshop can be seen in the following video:

Photoshop 32 vs 16 bits
Pages: 1 [2] 3   Go Up