Sony A99 (and RX1) raw file issue? An investigative report...

Started Dec 27, 2012 | Discussions
tesilab
Senior MemberPosts: 1,992Gear list
Like?
Sony A99 (and RX1) raw file issue? An investigative report...
Dec 27, 2012
worth of unique values. Anyone can repeat this exercise for themselves, and question my method.nearly 14 bits. Parallel samples of Nikon D600 and D3 files (I found these on imaging resource), show data show less than 11 bits of actual unique sampleBoth A99 and RX1 raw files both seem to Summary:
Samples I tested on my RX1 showed no more than 1,530 unique pixel values, though they covered a 12 bit range (values 0-4095), and presumably are results of a 14 bit conversion ,which implies 0-16383 range. (Though I could accept it, if they really meant to say a 14 bit conversion that produces 12 bit output.)
I was experiencing banding artifacts in a B&W conversion that I was doing (RX1 raw file). I will admit to all kinds of possible reasons for this, related to graphics card/display issues, exposure (I purposely overexposed to get as much info in the sky as possible), in camera vignetting correction, color space choices, noise reduction, overaggressive tone tweaking, etc. But still, even when backing off of edits, and exercising care, etc., the banding didn't seem justified to me, especially since I believed I would be getting 14 bit conversions.Why I started digging:
I went into Raw Digger, to examine actual histogram data. I looked for my best exposed file (in terms of number of unique values produced) and started comparing it against a variety of raw files from A99, D600, D3.What I did:

What I found in the RX1 file (A99 is similar, but the particular file was a bit worse):

What this means and doesn't mean:
Though the file sizes are nearly the same, and this type of output should compress pretty well, the Sony raw files are close to the same size as the Nikon files, so smaller files doesn't seem to be the motivation.
I don't have proof that the Nikon files contain genuinely unique data, rather than faking the data for people like me to be impressed.
I'm aware that visually, I could not possibly comprehend all these different values, but my point is I want maximum headroom in my files for things like highlight recovery.
It is entirely possible that with exhaustive empirical visual testing of a variety of raw files that I would draw very different conclusions than what the hard numbers seem to be telling me.

Conclusion: for someone in search of the best continuous tonal range available in affordable, compact digital, I would really like a definitive explanation of what the heck is going on!

(I'm posted a virtual copy in the Cyber-shot forum for RX1 users, pardon me those who visit both forums.)
  • There are 1530 distinct values in my best raw file out of the sample set of files I tried.
  • The range of values is 0-4093 (which is twelve bits per channel output, not 14, alas)
  • The values in certain ranges are consolidated, in other words:
  • 0-800 contains 801 unique values (i.e. is continuous)
  • 801-1424 contains 320 unique values (skips 1 out of every 2)
  • 1424-1427 contains 1 unique value (skips 2 of 3)
  • 1428-2023 contains 149 unique values (skips 3 of 4)
  • 2024-2029 contains 1 unique value (skips 5 of 6)
  • 2030-4093 contains 258 unique values (skips of 8)
 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Nikon D3 Nikon D600 Sony RX1 Sony SLT-A99
If you believe there are incorrect tags, please send us this post using our feedback form.
topstuff
Senior MemberPosts: 1,165Gear list
Like?
Hyperbole is the best thing in the Universe ever !!
In reply to tesilab, Dec 27, 2012

While I can't pretend to offer an explanation ( although no doubt this will not stop plenty of people commenting on your post who pretend to know something but in reality know nothing ! ) , but I think that if there IS a problem, then the digital photo testing community would have spotted it by now.

Take DPR for example, if the A99 was NOT producing 14 bit files I am quite certain they would have told us.

So before you post potentially hysterical hyperbole ( "scandal", really??) I think it would be sensible to check, check again, and check that again. And then check it again.

Like I said, DPR have been examining A99 for a long time and have minutely examined that camera's files. And they have found nothing to corroborate your thesis.

The logical assumption to make is therefore that you got it wrong.

Still, I am sure the "Scandal" !!! headline will make this a nice long thread, full of the usual suspects.

Have fun. Knock yourself out.

Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
Hyperbolic title, but sincere and double checked content.
In reply to topstuff, Dec 27, 2012

Well, it is a little fun to get some attention on the thread, but truthfully only interested in attracting the attention of someone who really can explain this. I have a long digital imaging background, BTW, though I will own up to misunderstanding data. But Rawdigger histograms of raw files dumped into CSV files are pretty straightforward, unless they are only decoding low bits of information and the rest of the bits are hiding somewhere else where they cannot decode them. But I haven't seen anything to that effect.

I invite you to try it for yourself, and see if you have a different interpretation. There is a stark difference in the histogram data. I never claimed it would produce a noticeable visual difference in most situations. However one who wants the most tonal data to work with must explain this discreprency to his satisfaction.

I will be delighted to be corrected, since I like my camera otherwise, and just want to become master of its tone curves.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
steelhead3
Veteran MemberPosts: 4,550Gear list
Like?
Re: Hyperbolic title, but sincere and double checked content.
In reply to tesilab, Dec 27, 2012

Don't know about rx1, but the 99 only shoots 14 bit in single shot mode; all other modes are 12 bit.

 steelhead3's gear list:steelhead3's gear list
Sony a77 II
Reply   Reply with quote   Complain
VirtualMirage
Senior MemberPosts: 2,776Gear list
Like?
Re: Sony A99 (and RX1) raw file scandal? An investigative report...
In reply to tesilab, Dec 27, 2012

I think the best person to ask is Iliah Borg.  He normally resides around the Nikon FX forums but if I recall correctly he knows quite a bit about RAW files and played a part (or sole credit, not sure) in the open source LibRaw.  He seems quite knowledgeable in topics such as this and might be the best person to inquire with.

-- hide signature --

Paul

 VirtualMirage's gear list:VirtualMirage's gear list
Sony RX100 Sony a77 II Tokina AT-X Pro 11-16mm f/2.8 DX Tamron SP AF 90mm F/2.8 Di Macro Samyang 8mm F3.5 Aspherical IF MC Fisheye +20 more
Reply   Reply with quote   Complain
topstuff
Senior MemberPosts: 1,165Gear list
Like?
Re: Hyperbolic title, but sincere and double checked content.
In reply to steelhead3, Dec 27, 2012

steelhead3 wrote:

Don't know about rx1, but the 99 only shoots 14 bit in single shot mode; all other modes are 12 bit.

Could this explain it?  Presumably 14 bit is restricted in an effort to increase fps?

Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
No, use only single frame mode files.
In reply to topstuff, Dec 27, 2012

I only checked single shot drive mode files. Not interested in high FPS. Especially given I've been doing some exposures at 30 secs. (Though my definitive tests were with daytime ISO 100 shots to be safe).

I've put in an inquiry with the author of RawDigger to see what he has to say on the subject.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
Waardij
Junior MemberPosts: 49
Like?
Re: Hyperbolic title, but sincere and double checked content.
In reply to topstuff, Dec 27, 2012

It has to do with the Sony compression. The idea is there is much more resolution in the highlights than ever needed. As an example, half of all the possible values in 14 bits, are present in the highest bit (so from 011111111111111 to 10000000000000). there is no way that the human eye can distinguish such small step, not even after severe post processing. and also the noise in that last bit will be much more than smallest values. In the above 14 bit example, there are 2^13=8192 values in the range from 13 bis on to 14 bits on. The smallest step for the shadow is only 1 value. Sony equalizes this extreme unbalance  bit, saving on the amount of data that has to be stored. I do not know of an example where its is shown that there is any disadvantage in this method. in the a900 one could choose to compress or not, an I have never found an advantage in not using the compression.

Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
For single shot HDR, or strong exposure shifting it is distinguishable
In reply to Waardij, Dec 27, 2012

If one wants to accentuate local details (as in HDR) or do extreme highlight shifting into the middle tones, these values would of course make a difference. Aside from longer exposures where noise increases over time, I'm not sure it is near correct to say that this highlight information is noiser than shadows. If anything, it should be the opposite.

Also please try to explain how the  actual storage requirements of a D600 are only very slightly greater than the A99/RX1 at the same resolution. Did Sony just "forget" to compress this optimized data?

Waardij wrote:

It has to do with the Sony compression. The idea is there is much more resolution in the highlights than ever needed. As an example, half of all the possible values in 14 bits, are present in the highest bit (so from 011111111111111 to 10000000000000). there is no way that the human eye can distinguish such small step, not even after severe post processing. and also the noise in that last bit will be much more than smallest values. In the above 14 bit example, there are 2^13=8192 values in the range from 13 bis on to 14 bits on. The smallest step for the shadow is only 1 value. Sony equalizes this extreme unbalance bit, saving on the amount of data that has to be stored. I do not know of an example where its is shown that there is any disadvantage in this method. in the a900 one could choose to compress or not, an I have never found an advantage in not using the compression.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
Eric Perez
Regular MemberPosts: 110
Like?
Re: Sony A99 (and RX1) raw file scandal? An investigative report...
In reply to tesilab, Dec 27, 2012

Wow, just read the entire thread. I am glad I didn't go the sony FF route. I do a lot of underexposing and HDR, post work that needs DR. My K5 can shoot 7FPS @14 bits without a LPF, not bad for $1300. I find it odd that sony does not disclose any of this info, scandal indeed...

Reply   Reply with quote   Complain
Ralf B
Veteran MemberPosts: 7,323Gear list
Like?
Re: Sony A99 (and RX1) raw file scandal? An investigative report...
In reply to Eric Perez, Dec 27, 2012

Eric Perez wrote:

Wow, just read the entire thread. I am glad I didn't go the sony FF route. I do a lot of underexposing and HDR, post work that needs DR. My K5 can shoot 7FPS @14 bits without a LPF, not bad for $1300. I find it odd that sony does not disclose any of this info, scandal indeed...

According to DXO, DPR and Photoclubalpha, no shortage of DR in the a99 at low ISO. Your reply misses the factual point attempted to be discussed here.

You have a PM, Eric.

-- hide signature --

Cheers,
Ralf
www.ralfralph.smugmug.com

 Ralf B's gear list:Ralf B's gear list
Sony Alpha DSLR-A900 Sony Alpha DSLR-A700 Sony SLT-A55 Sony 16-35mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* Sony 24-70mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* +13 more
Reply   Reply with quote   Complain
Eric Perez
Regular MemberPosts: 110
Like?
Re: Sony A99 (and RX1) raw file scandal? An investigative report...
In reply to Ralf B, Dec 27, 2012

Ralf Bliesener wrote:

Eric Perez wrote:

Wow, just read the entire thread. I am glad I didn't go the sony FF route. I do a lot of underexposing and HDR, post work that needs DR. My K5 can shoot 7FPS @14 bits without a LPF, not bad for $1300. I find it odd that sony does not disclose any of this info, scandal indeed...

According to DXO, DPR and Photoclubalpha, no shortage of DR in the a99 at low ISO. Your reply misses the factual point attempted to be discussed here.

You have a PM, Eric.

-- hide signature --

Cheers,
Ralf
www.ralfralph.smugmug.com

Ahh ok, well I guess I missed out there. And yes thank you for the message.

Reply   Reply with quote   Complain
Ralf B
Veteran MemberPosts: 7,323Gear list
Like?
Sony A99 (and RX1) raw file issue? An interesting read ...
In reply to tesilab, Dec 28, 2012

http://www.photoclubalpha.com/2012/12/27/sonys-alpha-99-mastery-wrapped-in-dilemma/

Quote:

The 14-bit raw file has great flexibility for shadow and highlight adjustment from raw without losing colour values or subtle tones. I’d rate it as one of the best raw file formats I have worked with, at normal (100-800) ISO settings.

end quote.

The above linked review is from a team of Nikon/Canon/Sony shooters with ongoing use of the a99, D600 and 6D. Too bad but it seems they won't add the RX-1 to their line-up to see hoe that would be judged by photoclubalpha.

-- hide signature --

Cheers,
Ralf
www.ralfralph.smugmug.com

 Ralf B's gear list:Ralf B's gear list
Sony Alpha DSLR-A900 Sony Alpha DSLR-A700 Sony SLT-A55 Sony 16-35mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* Sony 24-70mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* +13 more
Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
Don't jump the gun... I'm just asking about the data
In reply to Eric Perez, Dec 28, 2012

There is an explanation for everything. I tried to state that in the end, it may make no practical difference for most purposes. The layperson's expectation of what the data should look like, doesn't mean his expectation is correct--even if that layperson is a software engineer with an imaging background.

I'm waiting for some more authoritative answer on this. I naively expect 14 bits to mean that, depending on the image and exposure, it is possible to get get samples that exercise a fair portion of the 14 bit data range. This expectation seems to be supported by other raw files from other manufacturers.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
Still there is something that needs to be understood about the data
In reply to Ralf B, Dec 28, 2012

I actually wrote this in response to one user drew a hasty conclusion, but this response makes sense here too, so sorry to repeat, but I quote myself:

I tried to state that in the end, it may make no practical difference for most purposes. The layperson's expectation of what the data should look like, doesn't mean his expectation is correct--even if that layperson is a software engineer with an imaging background.

I'm waiting for some more authoritative answer on this. I naively expect 14 bits to mean that, depending on the image and exposure, it is possible to get get samples that exercise a fair portion of the 14 bit data range. This expectation seems to be supported by other raw files from other manufacturers.

Of course, if it were not possible to deliver very high quality images that work well in practice, Sony wouldn't do what they do with the data. But it is fair to wonder if this leads to consequences with certain edge cases in processing. Single file HDR is one, other types of tone curve manipulation is another. I was experiencing some banding when manipulating a raw file--which could be for a host of unrelated reasons, but that is what prompted me to start digging into the files.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
mick232
Contributing MemberPosts: 757Gear list
Like?
Re: Don't jump the gun... I'm just asking about the data
In reply to tesilab, Dec 28, 2012

tesilab wrote:

I'm waiting for some more authoritative answer on this. I naively expect 14 bits to mean that, depending on the image and exposure, it is possible to get get samples that exercise a fair portion of the 14 bit data range. This expectation seems to be supported by other raw files from other manufacturers.

That is not a naive expectation. Generally, if I buy an audio CD, I also expect to get true samples and not some MP3 re-converted to 16-bit 44.1kHz PCM and burned onto a disc. It will take some very aggressive tests shooting all kinds of gradients to find out if the lossy RAW compression has any practical effect (posterization) or not. Only if it does, I would call it a (mini) scandal and demanding a firmware update would be in order.

 mick232's gear list:mick232's gear list
Sony Alpha NEX-5 Sony SLT-A99 Tamron SP AF 70-200mm F/2.8 Di LD (IF) MACRO Sony 75-300mm F4.5-5.6 Sony 135mm F1.8 ZA Carl Zeiss Sonnar T* +18 more
Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
Hyperbole dealt with, plus a small anecdote...
In reply to topstuff, Dec 28, 2012

And anyone else who comes late to the thread wonders, the thread title has been helpfully corrected to something more appropriate.

Just one anecdote for you:

I used to be responsible for a digital imaging technology that attracted FujiFilm's attention in the 90's. They sent their engineers, who were interested in licensing the technology. They sent very sharp people, who started to shake their heads a bit as they realized during my exposition, that certain, ahh, liberties were taken in our image impressive image compositing system that was known for exceptional speed. (The company was called Live Picture, for those who recall--we inspired Photoshop to add layers to their application!) Anyway, I saw that we were losing a sale, so I had to explain our shortcuts in an acceptable way to them, so I said:

"Look, you can go with Microsoft, and they will give you a classic computer science based approach to imaging. But we, at Live Picture, have a imaging approach to computer science!"

This little turn of phrase seemed to make all the difference. They went from shaking to nodding. We had a major sale! I ended up building a postcard creating application for them that was bundled into image cd's that Fuji used to create for their film processing customers in Japan.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
Michaels7
Contributing MemberPosts: 504Gear list
Like?
Re: Sony A99 (and RX1) raw file issue? An investigative report...
In reply to tesilab, Dec 28, 2012

From Photoalpha.com

"It took me a few days and some digging to find out that the 14-bit readout on the A99 only applies one shooting mode – single shot. Any other mode you select, including Lo 2.5fps continuous and all multishot or JPEG only modes, uses 12-bit readout from the sensor. This was already documented in the literature about the A99, but what Sony omit to say is that the 14-bit mode causes a noticeable pause between the shot being captured and the restoration of EVF viewing. This pause is around 1/10th of a second longer in single shot mode than the blackout which happens during 2.5fps or the first frame of any faster sequence, and totals 200ms or 1/5th of a second."

 Michaels7's gear list:Michaels7's gear list
Sony SLT-A99 Sigma 85mm F1.4 EX DG HSM
Reply   Reply with quote   Complain
tesilab
Senior MemberPosts: 1,992Gear list
Like?
But I use single shot mode
In reply to Michaels7, Dec 28, 2012

None of the files I checked should have been of the burst mode variety.  I don't have my own A99, just an RX1. I took the A99 samples off of imaging-resource.com.

My findings should be easily reproducible by anyone. I checked only base iso single captures at zero EV compensation (and purposely overexposed files), loaded them into Raw Digdownload savehosting full histograms as CSV files, and examined the resulting data, every blessed row.

 tesilab's gear list:tesilab's gear list
Sony RX1 Sony Alpha NEX-5 Carl Zeiss Makro-Planar T* 2/100 Sony E 50mm F1.8 OSS Sigma 19mm F2.8 EX DN +11 more
Reply   Reply with quote   Complain
Iliah Borg
Forum ProPosts: 16,313
Like?
Re: Hyperbole dealt with, plus a small anecdote...
In reply to tesilab, Dec 29, 2012

"imaging approach to computer science!"

goldene verter

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads