Continuous shooting is lower quality...

Yes, in LENR, bulb, continous shooting and silent shooting your DR is 12bit instead of 14bit.

Professional information around the 12bit-14bit issue is found in Jim Kassons blog around August 2015:

http://blog.kasson.com/?m=201508

(and maybe also later)

--
flickr
 
Last edited:
Yes, nothing new.
 
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)

All the best,
Pedro
 
Last edited:
Thanks for pointing this out. I was aware of some of those modes going to 12bit but not LENR. I sometimes use that and other times not. I think I will not use that anymore and simply use a dark to subtract noise in nightscapes from now on.

Greg.
 
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)
I generally avoid 12-bit modes, but I'll admit to being less-than-informed about how much difference it really makes.

My speculation is that the roughly 15% reduction in going from 14-bit to 12-bit works out to a non-proportional loss of real-world IQ ... perhaps on the order of 5% or even less.

However, the explanations I've seen have tended to be "over my head".

Also, am I the only hobby'ist who wants to get every iota of IQ that the budget-busting a7Rii is capable of?

But .. am I "majoring in minor"? There are times when "Fast Continuous" or "Slow Continuous" would be very useful to capture a fleeting moment, for example.
 
Last edited:
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)
I generally avoid 12-bit modes, but I'll admit to being less-than-informed about how much difference it really makes.

My speculation is that the roughly 15% reduction in going from 14-bit to 12-bit works out to a non-proportional loss of real-world IQ ... perhaps on the order of 5% or even less.

However, the explanations I've seen have tended to be "over my head".

Also, am I the only hobby'ist who wants to get every iota of IQ that the budget-busting a7Rii is capable of?

But .. am I "majoring in minor"? There are times when "Fast Continuous" or "Slow Continuous" would be very useful to capture a fleeting moment, for example.
Most time we will loose more colorinformation with not-ETTR, than with any of these modes. But if you shoot a landscape at base ISO and do ETTR, it may be good to avoid those modes. But with a spontanous family shot at ISO1600, you will never notice any difference.
 
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)
I generally avoid 12-bit modes, but I'll admit to being less-than-informed about how much difference it really makes.

My speculation is that the roughly 15% reduction in going from 14-bit to 12-bit works out to a non-proportional loss of real-world IQ ... perhaps on the order of 5% or even less.

However, the explanations I've seen have tended to be "over my head".

Also, am I the only hobby'ist who wants to get every iota of IQ that the budget-busting a7Rii is capable of?

But .. am I "majoring in minor"? There are times when "Fast Continuous" or "Slow Continuous" would be very useful to capture a fleeting moment, for example.
I believe the important thing is to be aware of it - Using it or not depends, IMO, greatly on the final work pretended (or needed) IQ; for display/see on a monitor, even within today's state-of-the-art availability, there's no visible difference... :)

All the best,
Pedro
 
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)
I generally avoid 12-bit modes, but I'll admit to being less-than-informed about how much difference it really makes.

My speculation is that the roughly 15% reduction in going from 14-bit to 12-bit works out to a non-proportional loss of real-world IQ ... perhaps on the order of 5% or even less.

However, the explanations I've seen have tended to be "over my head".

Also, am I the only hobby'ist who wants to get every iota of IQ that the budget-busting a7Rii is capable of?

But .. am I "majoring in minor"? There are times when "Fast Continuous" or "Slow Continuous" would be very useful to capture a fleeting moment, for example.
Most time we will loose more colorinformation with not-ETTR, than with any of these modes. But if you shoot a landscape at base ISO and do ETTR, it may be good to avoid those modes. But with a spontanous family shot at ISO1600, you will never notice any difference.
 
Would love one day to see some real world examples instead of lab tests and graphs to see if one can see any differences between a post processed photo that was done say with LENR on and one with it off but using noise reduction in post.

Or maybe a milky way photographer that does both methods and counts the stars visible in the photo to see if LENR eats any of them? I know this is a daunting task on such large files sizes.

Similarly say a 120 second bulb shoot vs. the same done with 4 separate 30 second non-bulb shoots stacked via the mean method in PS; both at base ISO and higher ISO's

In the end does it matter and are we avoiding these in camera conveniences out of fear that may never make a real world difference?

I know Jim Kasson on this forum and in his excellent blog has been a pioneer with his testing and give him all of the credit and respect in the world for bringing many of these issues to light to many of us and providing the technical experience. Hopefully at some point either he or another personal with similar expertise will test some of these lab tests with outside real world imagery. Not brick walls and average sky stuff but real scenes in beautiful light.

There are plenty of professional photographers that are likely unaware of these 12 bit modes and use their Sony cameras much as they would their other brands that may not have these downgrades. If they are producing excellent images in 12 bit modes, should we care?

I know some of us are more technical and mathematical while others are visual and are blind to graphs and lab results. I fall somewhere in the middle myself.

Continuous shooting in my view: when I need it, are situations where absolute greatest dynamic range are not as important so I am less likely to care about any potential loss.

For taking photos of active toddlers, birds in flight, action sports and countless other things you really do need continuous shooting and I cannot imagine the difference in files is going to matter. These situations I would also be in Compressed Raw as I don't see the advantages of uncompressed coming into play.

Low light landscapes and night astro stuff is where I am not so sure and would love to see some clarity on the subject.

Some Sony Artisan should take it on.

As others have mentioned I believe it sounds like one potential issue is it could result in you overexposing as the histogram will not be as accurate in 12 bit mode. if that is the only negative I can live with and work around that as long as I am knowledgeable of that.
 
Yes, in LENR, bulb, continous shooting and silent shooting your DR is 12bit instead of 14bit.

Professional information around the 12bit-14bit issue is found in Jim Kassons blog around August 2015:

http://blog.kasson.com/?m=201508

(and maybe also later)
Here's a link to both graphs and pictures that show how the change in precision and increase in read noise affects photography:


No effect at all except at fairly low ISO settings.

Jim
 
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)
If you use compressed mode, the reduction in precision is from 13 to 12 bits, not 14 to 12 bits. However, there is also an increase in read noise.

Jim
 
Would love one day to see some real world examples instead of lab tests and graphs to see if one can see any differences between a post processed photo that was done say with LENR on and one with it off but using noise reduction in post.
Lotta options there. You've got an a7x camera, why not produce exactly what you think is dispositive and post it here? I'm sure we'd all love to see it.
Or maybe a milky way photographer that does both methods and counts the stars visible in the photo to see if LENR eats any of them? I know this is a daunting task on such large files sizes.

Similarly say a 120 second bulb shoot vs. the same done with 4 separate 30 second non-bulb shoots stacked via the mean method in PS; both at base ISO and higher ISO's

In the end does it matter and are we avoiding these in camera conveniences out of fear that may never make a real world difference?

I know Jim Kasson on this forum and in his excellent blog has been a pioneer with his testing and give him all of the credit and respect in the world for bringing many of these issues to light to many of us and providing the technical experience. Hopefully at some point either he or another personal with similar expertise will test some of these lab tests with outside real world imagery. Not brick walls and average sky stuff but real scenes in beautiful light.

There are plenty of professional photographers that are likely unaware of these 12 bit modes and use their Sony cameras much as they would their other brands that may not have these downgrades. If they are producing excellent images in 12 bit modes, should we care?

I know some of us are more technical and mathematical while others are visual and are blind to graphs and lab results. I fall somewhere in the middle myself.

Continuous shooting in my view: when I need it, are situations where absolute greatest dynamic range are not as important so I am less likely to care about any potential loss.

For taking photos of active toddlers, birds in flight, action sports and countless other things you really do need continuous shooting and I cannot imagine the difference in files is going to matter. These situations I would also be in Compressed Raw as I don't see the advantages of uncompressed coming into play.

Low light landscapes and night astro stuff is where I am not so sure and would love to see some clarity on the subject.

Some Sony Artisan should take it on.

As others have mentioned I believe it sounds like one potential issue is it could result in you overexposing as the histogram will not be as accurate in 12 bit mode. if that is the only negative I can live with and work around that as long as I am knowledgeable of that.
 
Lotta options there. You've got an a7x camera, why not produce exactly what you think is dispositive and post it here? I'm sure we'd all love to see it.
I wish I thought of it earlier, I just did some Milky Way shooting this weekend and kept LENR on.

Personally I don't see any concerns with the pics. Took some foreground shots that were multiple minutes to blend (ISO 640 and ISO 2000 location dependent) with my 25 second sky photos (ISO 3200 and 6400)

Guess I am just curious if it would make a visual difference at all.

I knew about the 12 bit LENR from your posts and others months ago, but then as time passed I forgot and went I went to make the magic happen I put it on LENR and should have but didn't try even one shot with it off.

I would have left it on for my long bulb times but wish I experimented on and off for the 25 second skies not taken in bulb mode.

Had both the A7r and A7rii with me and kept both with LENR on. I wanted to try a few different things and just never had time for fiddling in the pitch black night, the 3 hour or so window I had with a couple of planned stops ate up all of the time

My point is it is easier to shoot without having to think in the back of your head, if I do this it is 12 bit, that is 14 bit. In the end does it matter to the finished image and are you better off spending your energy getting the best composition and not worrying about the other things.

I am not a good tester!
 
Just read this: http://helpguide.sony.net/ilc/1520/v1/en/contents/TP0000818693.html?search=Silent

So, you get lower quality with continuous shooting. Hmmmmm...

Continuous autofocus or a different drive mode?
Undoubtedly true, but you have to see if going from 14 to 12 bit is noticeable regarding the quality of your workflow/final work; basically you're reducing tonality/shades from 16384 to 4096 steps for each R,G,B, i.e, from 16384x16384x16384 to 4096x4096x4096 available "colours"... :)
If you use compressed mode, the reduction in precision is from 13 to 12 bits, not 14 to 12 bits. However, there is also an increase in read noise.

Jim
 
Here's a link to both graphs and pictures that show how the change in precision and increase in read noise affects photography:

http://blog.kasson.com/?p=8586

No effect at all except at fairly low ISO settings.
Thanks for the follow-up, but I do have several questions after re-reading the blog article (some of which was over my head):
  • In this early 2015 article before uRAW, were you able to measure/isolate how much of this was related to cRAW? Is the problem less with uRAW since it was made available many months after the blog article?
  • > ISO 100 Continuous
    > ISO 100 Single Shot

    TLW: > There is quite a bit more noise in the continuous case [at base ISO 100 .. emphasis added]
    ...
  • Am I the only one whose reaction is something like "that is no small matter"?
  • Much of this retired hobby'ist justification for semi-defecting from my otherwise very usable Canon full frames was base ISO 100 performance.
  • I infer the "quite a bit more noise at low ISO" in 12-bit continuous also applies to other 2-modes like BULB, bracket'ing, Silent, etc. Or not?
  • Maybe I'm mis-reading, but my interpretation is a strengthening of my practice to stay away from 12-bit modes ... with an possible exception of continuous at ISO 640 to attempt to capture fleeting moments.
  • My general use of the a7Rii is typically on a tripod at low ISO to try to get every iota of IQ that its world class sensor is capable of.
  • In many other situations, my Canon 6d is more than adequate, and even after nearly 60k shutter activations on the a7Rii, still easier to use and get a good proportion of keepers. Some of that is no doubt operator error.
Hmmmm ... I just checked with Bill Claff's excellent PhotosToPhotos website if he had PDR charts for 12-bit vs 14-bit for my a7Rii (and other a7x cameras).

Nope.

I'll check with him about whether he already has that info (not published) or would be interested in the files to analyze and graph that.
 
Last edited:
Here's a link to both graphs and pictures that show how the change in precision and increase in read noise affects photography:

http://blog.kasson.com/?p=8586

No effect at all except at fairly low ISO settings.
Thanks for the follow-up, but I do have several questions after re-reading the blog article (some of which was over my head):
  • In this early 2015 article before uRAW, were you able to measure/isolate how much of this was related to cRAW? Is the problem less with uRAW since it was made available many months after the blog article?
I have seen no difference in read noise measurements between compressed and uncompressed images.

  • > ISO 100 Continuous
    > ISO 100 Single Shot

    TLW: > There is quite a bit more noise in the continuous case [at base ISO 100 .. emphasis added]
    ...
  • Am I the only one whose reaction is something like "that is no small matter"?
Probably not. I don't use continuous mode at ISO 100 myself.
  • Much of this retired hobby'ist justification for semi-defecting from my otherwise very usable Canon full frames was base ISO 100 performance.
  • I infer the "quite a bit more noise at low ISO" in 12-bit continuous also applies to other 2-modes like BULB, bracket'ing, Silent, etc. Or not?
It does apply, modulo gratuitous spatial filtering in bulb mode.
  • Maybe I'm mis-reading, but my interpretation is a strengthening of my practice to stay away from 12-bit modes ... with an possible exception of continuous at ISO 640 to attempt to capture fleeting moments.
  • My general use of the a7Rii is typically on a tripod at low ISO to try to get every iota of IQ that its world class sensor is capable of.
  • In many other situations, my Canon 6d is more than adequate, and even after nearly 60k shutter activations on the a7Rii, still easier to use and get a good proportion of keepers. Some of that is no doubt operator error.
Hmmmm ... I just checked with Bill Claff's excellent PhotosToPhotos website if he had PDR charts for 12-bit vs 14-bit for my a7Rii (and other a7x cameras).

Nope.

I'll check with him about whether he already has that info (not published) or would be interested in the files to analyze and graph that.
 
Hmmmm ... I just checked with Bill Claff's excellent PhotosToPhotos website if he had PDR charts for 12-bit vs 14-bit for my a7Rii (and other a7x cameras).

Nope.

I'll check with him about whether he already has that info (not published) or would be interested in the files to analyze and graph that.
OK ... done. No reason not to dig deeper on this 12-bit vs 14-bit issue, at least regarding PDR.

My perhaps flawed understanding is that PDR and sensor noise are very highly correlated ... almost reciprocals?
 
Hmmmm ... I just checked with Bill Claff's excellent PhotosToPhotos website if he had PDR charts for 12-bit vs 14-bit for my a7Rii (and other a7x cameras).

Nope.

I'll check with him about whether he already has that info (not published) or would be interested in the files to analyze and graph that.
OK ... done. No reason not to dig deeper on this 12-bit vs 14-bit issue, at least regarding PDR.

My perhaps flawed understanding is that PDR and sensor noise are very highly correlated ... almost reciprocals?
You're thinking of EDR. The main factors influencing PDR are FWC, QE, and read noise.

Jim

--
http://blog.kasson.com
 
Last edited:

Keyboard shortcuts

Back
Top