December 19, 2014, 10:07:20 PM

Author Topic: DxO reviews Sony A7s: king of low light photography?  (Read 24308 times)

dilbert

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3289
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #105 on: August 05, 2014, 01:20:19 AM »
...
7. Save a print copy 16-bit TIFF.
8. Print from Photoshop. Sometimes I'll do a test print on a small 4x6" sheet (I usually have some boxes of that for the various types of paper I like to print on...I'm a big fan of natural fiber papers, particularly from Moab and Hahnemuhle, and the satin and metal luster papers from RedRiver.) If the 4x6" test comes out looking good from a gamut standpoint, I'll print on the full size paper, which is usually either 13x19" or 11x14".

And you do all this on windows or osx?

All on Windows 8.1. As far as Photoshop and printing is concerned, there is no reason to "need" OSX for printing anymore. That old paradigm went out over a decade ago. Windows have been top notch graphic workstations for a long time.

So you have harped on about how there is no need to have cameras with high DR because this isn't capable of it or that isn't capable of it but yet your entire workflow is in 16bit.

In various other threads you've commented that Sony are claiming too many stops of DR for the number of bits and since you're working with a 16bit workflow obviously you think you need more than 7 or 8 bits of IQ and since 1 bit equals 1 stop of DR, 7 to 8 stops of DR is not enough for you.

Elsewhere in earlier comments in this thread you've stated that print only requires 7 or 8 stops of DR.

Something doesn't add up here. You're saying that printing only delivers 7 or 8 stops of DR yet your workflow supports 16. Why not just work in an 8 bit workspace since that is all that your printing can deliver?

canon rumors FORUM

Re: DxO reviews Sony A7s: king of low light photography?
« Reply #105 on: August 05, 2014, 01:20:19 AM »

privatebydesign

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 2840
  • Ermintrude says "moo"
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #106 on: August 05, 2014, 01:37:15 AM »
So you have harped on about how there is no need to have cameras with high DR because this isn't capable of it or that isn't capable of it but yet your entire workflow is in 16bit.

In various other threads you've commented that Sony are claiming too many stops of DR for the number of bits and since you're working with a 16bit workflow obviously you think you need more than 7 or 8 bits of IQ and since 1 bit equals 1 stop of DR, 7 to 8 stops of DR is not enough for you.

Elsewhere in earlier comments in this thread you've stated that print only requires 7 or 8 stops of DR.

Something doesn't add up here. You're saying that printing only delivers 7 or 8 stops of DR yet your workflow supports 16. Why not just work in an 8 bit workspace since that is all that your printing can deliver?

It is about tonality and the subtleties of moving them around to best be represented in a smaller space. Try working an 8 bit file with large areas of subtle tones, like a blue sky, without getting posterization, or printing deep shadow detail without a curves layer to lift those few darkest tones up a touch, now do that to a 16 bit file with 14 bit information and you can push and pull it where you want with no ill effects, render that edit as an 8 bit jpeg and you are golden, send the edited 16 bit file to a printer in 16 bit or 8 bit and you have no posterization, superb tonal control in B&W etc etc.

What you are arguing is if you can't view more than 6-8 stops of DR in a print then there is no point in capturing it, that is patently not true. You capture as many stops of DR as you want and then choose where to put each of those stops in your output.

It is called photography.

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4814
  • EOL
    • View Profile
    • Nature Photography
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #107 on: August 05, 2014, 01:45:43 AM »
...
7. Save a print copy 16-bit TIFF.
8. Print from Photoshop. Sometimes I'll do a test print on a small 4x6" sheet (I usually have some boxes of that for the various types of paper I like to print on...I'm a big fan of natural fiber papers, particularly from Moab and Hahnemuhle, and the satin and metal luster papers from RedRiver.) If the 4x6" test comes out looking good from a gamut standpoint, I'll print on the full size paper, which is usually either 13x19" or 11x14".

And you do all this on windows or osx?

All on Windows 8.1. As far as Photoshop and printing is concerned, there is no reason to "need" OSX for printing anymore. That old paradigm went out over a decade ago. Windows have been top notch graphic workstations for a long time.

So you have harped on about how there is no need to have cameras with high DR because this isn't capable of it or that isn't capable of it but yet your entire workflow is in 16bit.

In various other threads you've commented that Sony are claiming too many stops of DR for the number of bits and since you're working with a 16bit workflow obviously you think you need more than 7 or 8 bits of IQ and since 1 bit equals 1 stop of DR, 7 to 8 stops of DR is not enough for you.

Elsewhere in earlier comments in this thread you've stated that print only requires 7 or 8 stops of DR.

Something doesn't add up here. You're saying that printing only delivers 7 or 8 stops of DR yet your workflow supports 16. Why not just work in an 8 bit workspace since that is all that your printing can deliver?

The workflow is actually with an original RAW file opened up in Photoshop. It's in whatever bit depth the RAW file has, which in my case is 14-bit. The saving to 16-bit TIFF occurs at the end. You also don't really seem to understand the significant differences between print and digital images. Print is an entirely different beast. In the print world, you never hear anyone talk about "dynamic range"...it's all DMax and L* and tonalities and the subtleties of color in areas of fine microcontrast. Bit depth is ultimately about tonal space, the ability to preserve fine gradients and the subtleties of detail when your pushing the white and black point around or trying to shift out of gamut colors into the gamut of the printer+ink+paper. It doesn't matter if print only has five or six stops of "dynamic range"...that simply doesn't matter. Print is different, it uses an entirely different color model and you have to be concerned about entirely different things like ink density and ink evenness and reflectivity and metamerism and viewing light and all that.

Now, even assuming I saved the file as a 16-bit TIFF up front, there are a couple things that limit the editing latitude. For one, saving a 14-bit file as a 16-bit file is no different than saving an 8-bit file as a 16-bit file. You don't suddenly gain more information out of thin air. You started with 14 bits of information. That 14 bits of information is distributed throughout the 16-bit file...it's the same amount of information, just stored in a file capable of referencing larger numbers. You don't gain anything. Your camera's dynamic range is limited by it's HARDWARE. I could save the image as a 32-bit floating point HDR TIFF if I wanted to...that still isn't going to somehow magically create more information out of nothing. The only way to take full advantage of a 32-bit HDR TIFF is to actually take multiple frames at different exposures, and do an HDR merge that combines the extended information of all of them into a single data set that is actually capable of utilizing the high dynamic range that a 32-bit floating point TIFF image can offer.

There is no magic to converting a 14-bit RAW to a 16-bit TIFF. You don't gain anything for nothing. The only reason I do it is because the only other options are either 8-bit TIFF or 32-bit integer or floating point TIFF. I don't want to lose any information, which is guaranteed if I save as 8-bit TIFF. I don't want to waste space, which is also guaranteed if I use a 32-bit TIFF without actually having enough information precision to take advantage of it. So, I save as 16-bit TIFF.

As for why work in 16-bit or 14-bit rather than 8-bit? The math used for all the various algorithms that are applied when processing is prone to introducing error. That error often ends up affecting the lower order bits most, where information is most sparse (i.e. lower midtones and shadows), although it can and will affect the entire signal. When you have only 8 bits, those errors show up more readily as artifacts. When you have more bits, you run a much lower risk of introducing processing artifacts into your images. This is one of the benefits of working in high bit depth RAW. Even the small move from 12-bit to 14-bit was actually fairly significant from the standpoint of improving the working space for all the mathematical algorithms to do their thing with RAW images without introducing artifacts. There are also other losses when moving from a true RAW format to an RGB format. Once you bind the three color channels together at each pixel, you lose a significant amount of editing latitude unless you move to a significantly higher bit depth, and even then, you can only push and pull the exposure or other information around with the basic kinds of tools in programs like Photoshop or Lightroom so much before artifacts exhibit with a vengeance. I do massive signal stretching for my astrophotography, but in order to do so, I usually work with 64-bit IEEE floating point FITS images, so I have enough precision to minimize the impact of errors introduced by algorithms. Even then, there are still limits to how far I can push those algorithms...push them too far, and your still guaranteed to get artifacts.

Sporgon

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 2079
  • 5% of gear used 95% of the time
    • View Profile
    • www.buildingpanoramics.com
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #108 on: August 05, 2014, 02:23:15 AM »

So you have harped on about how there is no need to have cameras with high DR because this isn't capable of it or that isn't capable of it but yet your entire workflow is in 16bit.


Something doesn't add up here. You're saying that printing only delivers 7 or 8 stops of DR yet your workflow supports 16. Why not just work in an 8 bit workspace since that is all that your printing can deliver?

Oh my goodness. I don't know whether to laugh or cry.

Every now and again someone on CR who is trying to sound knowledgeable and intelligent drops a real clanger - remember professor Pi - "Exposure Value ? That's a term you just made up".

Now jrista is going to patiently and conscientiously, like a seasoned old school master teaching a delinquent child, explain to you why we work in 16 bit.

Update - I see he already has !
« Last Edit: August 05, 2014, 02:25:10 AM by Sporgon »

dilbert

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3289
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #109 on: August 05, 2014, 02:26:22 AM »
...
As for why work in 16-bit or 14-bit rather than 8-bit? The math used for all the various algorithms that are applied when processing is prone to introducing error. That error often ends up affecting the lower order bits most, where information is most sparse (i.e. lower midtones and shadows), although it can and will affect the entire signal. When you have only 8 bits, those errors show up more readily as artifacts. When you have more bits, you run a much lower risk of introducing processing artifacts into your images. This is one of the benefits of working in high bit depth RAW. Even the small move from 12-bit to 14-bit was actually fairly significant from the standpoint of improving the working space for all the mathematical algorithms to do their thing with RAW images without introducing artifacts.
...

So you're all in favour of more bits and more DR and you've even argued that when it gets to printing, it's no longer about DR but tonality, etc, and that the high bit depth doesn't get wasted. i.e. if your camera gave you 14bits of DR like Sony's do, you wouldn't throw any of it away but at the same time you seem happy to argue that there's no benefit from the extra DR that Sony provide.

You obviously use and take advantage of there being many more than 8 bits of DR in your Canon images (and use imaging equipment that can and does generate images with more than 8 bits of DR) yet you've also tried to argue that those printing don't need more than 8 because it will be lost when printing.

What then was the value of saying that high DR isn't required because printers can only do 8 stops of DR?

dilbert

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3289
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #110 on: August 05, 2014, 02:32:55 AM »
...
Now jrista is going to patiently and conscientiously, like a seasoned old school master teaching a delinquent child, explain to you why we work in 16 bit.
...

I know why *I* work in 16bit but I wanted to see why *he* works in 16bit.

To me it seems like he's confused. His printer outputs 8 stops of DR yet he wants to work with 11 stops of DR from 14bit images (in a world where we equate 1bit with 1 stop of DR [people complained about Sony claiming > 15bits DR just because the files were only 14bit without seeing output] it seems that there are 3 bits being waste here.) If he doesn't need more than 8 stops of DR in his printouts then he should quit working with raw and TIFF files and just use JPEGs.

dilbert

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3289
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #111 on: August 05, 2014, 02:36:50 AM »
The whole point here being it is completely disingenuous to claim that more stops of DR are wasted because they can't be used on the screen or on paper.

They do get used and jrista has demonstrated how they do.

They don't get wasted and every bit of extra shadow, highlight and gradation helps improve the final IQ of images.

Thus there's nothing to gain from putting Sony's A7s down because it generates "too much DR." It'd be like criticising a sonata from Mozart because it has too many notes!

canon rumors FORUM

Re: DxO reviews Sony A7s: king of low light photography?
« Reply #111 on: August 05, 2014, 02:36:50 AM »

Sporgon

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 2079
  • 5% of gear used 95% of the time
    • View Profile
    • www.buildingpanoramics.com
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #112 on: August 05, 2014, 02:55:47 AM »
If he doesn't need more than 8 stops of DR in his printouts then he should quit working with raw and TIFF files and just use JPEGs.

I don't know what to say. I'd have never made a school teacher.

You're confusing theoretical DR from bit depth of A/D converter with a 16 bit RGB TIF file in post processing, or something like that.

dilbert

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3289
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #113 on: August 05, 2014, 03:17:14 AM »
If he doesn't need more than 8 stops of DR in his printouts then he should quit working with raw and TIFF files and just use JPEGs.

I don't know what to say. I'd have never made a school teacher.

You're confusing theoretical DR from bit depth of A/D converter with a 16 bit RGB TIF file in post processing, or something like that.

If you start out with a JPEG file then there's no need to ever edit in 16bit mode because your data cannot fill all of the 16bits (maybe every second bit.)

Now if you've got a 14 bit A/D that generates 11 stops of DR then you've got 11bits worth of width in the source data. Obviously there is information loss if you squeeze 11 into 8 but similarly, you cannot map the entire 11 into 16 without either creating holes or making up data. When you add in more DR you then get more bits to fill out the 16 with and hence get better IQ. Seems reasonable, yes?

Afterall, 1 stop of DR is 1 bit position in the image file, right?

Kathode-Ray

  • PowerShot G1 X II
  • ***
  • Posts: 49
  • Shoot, shoot, shoot!
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #114 on: August 05, 2014, 06:44:04 AM »
This guy really doesn't get it... it's hopeless!
6D | M || M 18-55 | M 22 2.0 || T 24-70 2.8 VC | 24-105 | 70-200 4.0 IS || SY 14 2.8 | 35 2.0 IS | 50 1.4 | Σ 85 1.4 || Metz 58 AF-2 | Røde Videomic Pro | Magic Lantern

dilbert

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 3289
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #115 on: August 05, 2014, 09:32:48 AM »
This guy really doesn't get it... it's hopeless!

Hey, don't blame me - I'm not the one asserting that extra stops of DR and information in the image files is wasted, it is others. The only problem is that it would seem that all of those bits that they say you don't need are actually used by them anyway.

If people would stop trying to belittle and put down Sony's sensors because they deliver and offer more DR then it would be a whole lot easier. This is where the problems stem from: trying to assert that what comes off the Sony sensors is no better than the Canon's.

I'm pretty sure that if I tried, I could use Google and find threads on here where various people have waxed lyrical (and received support for) about Canon producing sensors with fewer larger pixels with more DR and better IQ - especially in low light. Well guess what, Sony has done that.

But instead of accepting that and congratulating Sony on doing it, people are arguing about how all of that extra DR and IQ is not necessary. What a load of horse sh*t.

3kramd5

  • Canon 7D MK II
  • *****
  • Posts: 462
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #116 on: August 05, 2014, 09:46:43 AM »
The whole point here being it is completely disingenuous to claim that more stops of DR are wasted because they can't be used on the screen or on paper.

I've seen him say on numerous occasions that he'd like more dynamic range, and that he appreciates exmor technology. I think pretty much everyone does. Here's an example:

Well, for landscapes, I really want that extra resolution. A low resolution of 12mp really doesn't interest me much, and I was hoping the 5D III would land with 28-32mp. The D800E with 36.3mp and its incredible DR is about as close to the holy grail of landscape photography as I think it can get these days. I'd much prefer if Canon could reciprocate with their own megapixel/DR monster, though.

And another:

This kind of night photography is another area where more DR is certainly an enviable trait to have. Your the type who could probably benefit from a full 16 stops of DR with a true 16-bit sensor, even.



There have been times when he's impeached the notion of using DXO's "print DR" values to quote camera specifications, maybe that's what you're referring to?

More DR is better (up to the point that the camera DR exceeds the scene DR, which many of us use lighting, ND filters, etc to achieve), just as more resolution is better. Even if it doesn't make its way to the final format, it gives latitude to workflow.

If he doesn't need more than 8 stops of DR in his printouts then he should quit working with raw and TIFF files and just use JPEGs.

You don't honestly think that bit-depth is the only advantage to RAW, do you?
« Last Edit: August 05, 2014, 10:03:19 AM by 3kramd5 »
5D3, 5D2, 40D; Various lenses

neuroanatomist

  • CR GEEK
  • ************
  • Posts: 15213
    • View Profile
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #117 on: August 05, 2014, 10:51:53 AM »
Hey, don't blame me - I'm not the one asserting that extra stops of DR and information in the image files is wasted, it is others. The only problem is that it would seem that all of those bits that they say you don't need are actually used by them anyway.

If people would stop trying to belittle and put down Sony's sensors because they deliver and offer more DR then it would be a whole lot easier. This is where the problems stem from: trying to assert that what comes off the Sony sensors is no better than the Canon's.

I'm pretty sure that if I tried, I could use Google and find threads on here where various people have waxed lyrical (and received support for) about Canon producing sensors with fewer larger pixels with more DR and better IQ - especially in low light. Well guess what, Sony has done that.

But instead of accepting that and congratulating Sony on doing it, people are arguing about how all of that extra DR and IQ is not necessary. What a load of horse sh*t.
We blame you for your total inability to comprehend what you read.  It's generally acknowledged here (and specifically stated by the people to whom you are presumably referring) that Sony sensors offer better low ISO DR, and that it can be useful to have more DR (assuming it exists in the scenes being captured). 

The 'load of horse sh*t' is the blanket assertion that a sensor with better low ISO DR means a better camera.  For some, that might be true...but certainly not for the majority, much less everyone.


This guy really doesn't get it... it's hopeless!
This is the guy who thought a lens was a camera, and never could admit he was wrong.  His grasp of technical issues is tenuous at best, and his knowledge of facts can only aspire to become tenuous.
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

canon rumors FORUM

Re: DxO reviews Sony A7s: king of low light photography?
« Reply #117 on: August 05, 2014, 10:51:53 AM »

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4814
  • EOL
    • View Profile
    • Nature Photography
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #118 on: August 05, 2014, 12:08:03 PM »
...
As for why work in 16-bit or 14-bit rather than 8-bit? The math used for all the various algorithms that are applied when processing is prone to introducing error. That error often ends up affecting the lower order bits most, where information is most sparse (i.e. lower midtones and shadows), although it can and will affect the entire signal. When you have only 8 bits, those errors show up more readily as artifacts. When you have more bits, you run a much lower risk of introducing processing artifacts into your images. This is one of the benefits of working in high bit depth RAW. Even the small move from 12-bit to 14-bit was actually fairly significant from the standpoint of improving the working space for all the mathematical algorithms to do their thing with RAW images without introducing artifacts.
...

So you're all in favour of more bits and more DR and you've even argued that when it gets to printing, it's no longer about DR but tonality, etc, and that the high bit depth doesn't get wasted. i.e. if your camera gave you 14bits of DR like Sony's do, you wouldn't throw any of it away but at the same time you seem happy to argue that there's no benefit from the extra DR that Sony provide.

You obviously use and take advantage of there being many more than 8 bits of DR in your Canon images (and use imaging equipment that can and does generate images with more than 8 bits of DR) yet you've also tried to argue that those printing don't need more than 8 because it will be lost when printing.

What then was the value of saying that high DR isn't required because printers can only do 8 stops of DR?

You are really clueless, man. The only thing you care about is finding some way to make every discussion about dynamic range. You are UTTERLY CLUELESS, and it's clear everyone here but you sees that. Your just a waste of time, man. If you don't get it, you don't get it. It's clear your never going to get it, so I'm done trying to explain it. You can keep on believing you have everything figured out, but you really don't have the first bloody clue what your talking about.

jrista

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 4814
  • EOL
    • View Profile
    • Nature Photography
Re: DxO reviews Sony A7s: king of low light photography?
« Reply #119 on: August 05, 2014, 12:14:40 PM »
...
Now jrista is going to patiently and conscientiously, like a seasoned old school master teaching a delinquent child, explain to you why we work in 16 bit.
...

I know why *I* work in 16bit but I wanted to see why *he* works in 16bit.

To me it seems like he's confused. His printer outputs 8 stops of DR yet he wants to work with 11 stops of DR from 14bit images (in a world where we equate 1bit with 1 stop of DR [people complained about Sony claiming > 15bits DR just because the files were only 14bit without seeing output] it seems that there are 3 bits being waste here.) If he doesn't need more than 8 stops of DR in his printouts then he should quit working with raw and TIFF files and just use JPEGs.

It's really NOT about dynamic range. My average ISO is over 1000 in the majority of my 7D photos, and in the case of the 5D III, I am often shooting as high as ISO 12800! There is a difference between bit depth and dynamic range. The only real link between the two is that bit depth LIMITS dynamic range, however at higher ISO, dynamic range is limited by physics, and is usually much less than the limit imposed by bit depth. Your argument is that I have 14 stops of dynamic range just because I have a 14-bit file. No. I have a LIMIT of 14 stops of dynamic range in a 14-bit file...I usually get less than 8-10 stops because I am always at high, often very high, ISO. That would be the case with a D800, D810, or even a 50mp Medium Format Sony Exmor sensor.

Bit depth does more than limit your maximum potential dynamic range, though. It is also the factor that determines the precision of each and every value that represents a pixel. More precision, more numeric "space" within which to work when executing algorithms...such as exposure, white balance, filters like noise reduction and sharpening, etc.

PHYSICS LIMITS MY DR!! Bit depth maximizes my editing potential! Get it?!?  :o

canon rumors FORUM

Re: DxO reviews Sony A7s: king of low light photography?
« Reply #119 on: August 05, 2014, 12:14:40 PM »