sources on 13 EV DR possible from 12 bit sensor

Started Apr 19, 2012 | Discussions
highwave
Senior MemberPosts: 1,041Gear list
Like?
sources on 13 EV DR possible from 12 bit sensor
Apr 19, 2012

I found these threads and posts from a user on DxO forums (dosdan). He gave links to them:

http://forums.dpreview.com/forums/read.asp?forum=1018&message=41071974

http://forums.dpreview.com/forums/readflat.asp?forum=1018&message=40976432&changemode=1

Apparently also the Pentax K-01 has a 12-bit sensor yet DxO gives 12.9EV dynamic range.

So the measured 13EV of the OM-D might not be as ridiculous as some here are making it out to be.

P.S. I'm a regular reader here and did not see this ever posted. If it is a repost I apologize in advance. Also thank you dosdan if you're reading this.

 highwave's gear list:highwave's gear list
Olympus OM-D E-M5 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G 14mm F2.5 ASPH Olympus M.Zuiko Digital 45mm 1:1.8 Panasonic Lumix G Vario 45-150mm F4-5.6 ASPH Mega OIS +3 more
Pentax K-01
If you believe there are incorrect tags, please send us this post using our feedback form.
boggis the cat
Veteran MemberPosts: 6,295Gear list
Like?
It isn't a per-pixel measurement
In reply to highwave, Apr 19, 2012

highwave wrote:

So the measured 13EV of the OM-D might not be as ridiculous as some here are making it out to be.

It's a statistical measurement based on an area of pixels. A 12-bit number obviously can't hold more than "12 EV" of data, so per-pixel DR is lower.

So one pixel data point just above zero in a block of 100 pixels at zero compared to two pixels out of 100 at just above zero yields a "1 EV" increase in DR. (I think. The process may be a bit more complicated...) However, at the pixel level there has been no DR change.

This confused me, too, until it was pointed out that this was what was implied by these measurements. My view was that DR is really about what a given pixel can capture and store as a data point, and I still believe this is a simpler view -- but acknowledge that DR is complicated by the issue of noise and is entirely subjective unless you use some (arbitrary) statistical base-point and assessment method.

 boggis the cat's gear list:boggis the cat's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus 12-40mm F2.8 +7 more
Reply   Reply with quote   Complain
Detail Man
Forum ProPosts: 15,270
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to highwave, Apr 19, 2012

highwave wrote:

I found these threads and posts from a user on DxO forums (dosdan). He gave links to them:

http://forums.dpreview.com/forums/read.asp?forum=1018&message=41071974

http://forums.dpreview.com/forums/readflat.asp?forum=1018&message=40976432&changemode=1

Apparently also the Pentax K-01 has a 12-bit sensor yet DxO gives 12.9EV dynamic range.

So the measured 13EV of the OM-D might not be as ridiculous as some here are making it out to be.

P.S. I'm a regular reader here and did not see this ever posted. If it is a repost I apologize in advance. Also thank you dosdan if you're reading this.

Dosdan (who I have communicated with in the past) is a very sharp and very nice fellow, indeed !

It is true that certain kinds of ADC implementations can provide DR/SNR that exceeds bit-depth. From the same thread that both of the above references link-to, see this post by bobn2 :

bobn2 wrote:

apaflo wrote:

If an analog signal is digitized the maximum dynamic range that can encoded depends on the number of bits used.

Floyd, this simply is not true. Think for instance of a sigma delta converter. A first order ΣΔ uses a one bit quantiser and within the loop encodes the signal to a one bit stream. If you wanted, you could extract the one bit stream and record that, as happens for instance in a SACD audio disc, which produces a dynamic range of 120dB.

http://forums.dpreview.com/forums/read.asp?forum=1018&message=41012206

True enough (in general). See: http://en.wikipedia.org/wiki/Delta-sigma_modulation

But let us get a little "bit" (pun) more specific. The E-M5 (the present "bone of contention") as a result of RawDigger analysis (among other possible techniques) shows the E-M5 ORF image-files to have maximum RGBG raw-levels of less than 4096 (2^12). RawDigger reports that the RMS value of the (double-sided noise-distribution, when not halved due to black-level-subtraction) in each of the RGBG raw-channels equals 1.0. That is one part out of a maximum of 4096 (or 2^12).

Thus, Anders W has calculated and published a Dynamic Range calculated at just under 12.0 EV.

When applying black-level subtraction to the same E-M5 ORF image-file (which "slides" one-half of the double-sided noise-distribution "under the radar" and into the immeasurable range of less than the minimum resolvable RAW-data), one can calculate a Dynamic Range that equals 12.308 EV - slightly greater than 12.0 EV.

But not 13 EV, or 14 EV, or 15 EV, etc. ....

The Sony NEX-7 (for example) uses a non-linear data-compression curve which effectively alters the numerical base associated with each binary bit-representation to a base that is other than the (more conventional) base 2. Such techniques do indeed allow for "data-compression" to occur ...

See: http://forums.dpreview.com/forums/readflat.asp?forum=1042&thread=39707815

Such non-linearly encoded RAW image-files require special and specific decoding when analyzed or processed (in order to apply the inverse function of the encoding function in a decoding operation). There exists zero analytical or visual (when processed using any known RAW converter/processor) evidence of this being the case where it comes to Olympus O-MD E-M5 ORF image-files. Thus, claims of E-M5 Dynamic Range that exceed 12.0 EV are either erroneous, or they involve the truncation ("throwing out") of some or all of the Read/Dark Noise distribution, period ...

There exists some evidence that (at the three highest ISO Gains only, mostly at and above ISO=3200) Olympus has deliberately designed the E-M5 so that such a "truncation" of the Read/Dark Noise does in fact occur. This likely results in improved Dynamic Range specifications ...

See: http://translate.google.com/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.optyczne.pl%2F178.7-Test_aparatu-Olympus_OM-D_E-M5_Szumy_i_jakosc_obrazu_w_RAW.html

That's all, folks,

DM ...

Reply   Reply with quote   Complain
boggis the cat
Veteran MemberPosts: 6,295Gear list
Like?
An example that may make sense
In reply to boggis the cat, Apr 19, 2012

highwave wrote:

So the measured 13EV of the OM-D might not be as ridiculous as some here are making it out to be.

Sorry, that was a rubbish attempt at explaining what is entailed. Here is an example that should make sense.

A 12-bit raw can store 2^12 or 4096 values -- 0 to 4095.

  • Assume a "perfect" sensor (no noise) with 1000 pixels. Take a photo of a bright white source, such that the raw file contains values of "4095" for each pixel -- all 12 bits are set.

  • Reduce the lighting to obtain a raw file with values of "2047" -- now 11 bits are set, so that's a -1 EV DR drop.

  • Keep reducing the light by 1 EV steps to get raw data as follows: 1023, 511, 255, 127, 63, 31, 15, 7, 3, 1 (10, 9, 8, 7, 6, 5, 4, 3, 2, 1 bits set) -- you've now dropped by 11 EV total in DR.

  • Now, reducing the light a further 1 EV results in half of the 1000 pixels dropping to "0", while half remain at "1" -- they don't all drop to "0". (This is the 12 EV point.)

  • Decrease the lighting another 1 EV and we get half of the remaining pixel data remaining at "1" (250 pixels yield "1"); then we get 125 pixels, 64 / 65, 32, 16, 8, 4, 2, 1, 0 (the 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 EV points).

  • So our theoretical sensor has 12-bit data but can reproduce 22 EV of DR -- obviously this DR is per image , and not per pixel .

So, if you have a large number of pixels you can keep dropping the DR by 1 EV until the last pixel with data ("1") finally changes to "0".

In real terms, there is always pixel noise that affects the lowest bit. So you'd shift the "half pixels for 1 EV" to a higher bit.

(I tried to include this example, but ran out of editing time for the first post.)

 boggis the cat's gear list:boggis the cat's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus 12-40mm F2.8 +7 more
Reply   Reply with quote   Complain
MAubrey
Senior MemberPosts: 1,532Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

I don't always follow your posts (I'm a linguist, not an engineer), when I do (as in this case) I greatly appreciate them.
--
--Mike

 MAubrey's gear list:MAubrey's gear list
Olympus OM-D E-M5 Sony Alpha 7 Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Voigtlander 58mm F1.4 Nokton SL II Voigtlander 35mm F1.2 Nokton +7 more
Reply   Reply with quote   Complain
Rriley
Forum ProPosts: 21,846Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

the crew that designed it are claiming they were looking for

1/3 to 1/2 and is the highlight. Shadow side has improved stage about 1/2 of itself because there was improvement ratio S / N

translations a little rough but they say 'something like' they have a 2 stage process, it looks to be each side of ISO 1600

pick it up in DC Watch

http://translate.google.com/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fdc.watch.impress.co.jp%2Fdocs%2Fnews%2Finterview%2F20120418_524068.html

-- hide signature --

Riley

any similarity to persons living or dead is coincidental and unintended
support 1022 Sunday Scapes'

 Rriley's gear list:Rriley's gear list
Sigma DP2 Merrill Canon EOS 5D Olympus E-3 Olympus E-5 Panasonic Lumix DMC-GH2 +1 more
Reply   Reply with quote   Complain
Rriley
Forum ProPosts: 21,846Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

Detail Man wrote:

But not 13 EV, or 14 EV, or 15 EV, etc. ....

of course lets be honest, we are talking in DxO terms

More realistically shot RAW against a step chart its going to be less than what is presented anyway.

but it should still look quite good in DxO as 'Print Size' wouldn't you say ...

-- hide signature --

Riley

any similarity to persons living or dead is coincidental and unintended
support 1022 Sunday Scapes'

 Rriley's gear list:Rriley's gear list
Sigma DP2 Merrill Canon EOS 5D Olympus E-3 Olympus E-5 Panasonic Lumix DMC-GH2 +1 more
Reply   Reply with quote   Complain
highwave
Senior MemberPosts: 1,041Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

Detail Man wrote:

highwave wrote:

I found these threads and posts from a user on DxO forums (dosdan). He gave links to them:

The Sony NEX-7 (for example) uses a non-linear data-compression curve which effectively alters the numerical base associated with each binary bit-representation to a base that is other than the (more conventional) base 2. Such techniques do indeed allow for "data-compression" to occur ...

See: http://forums.dpreview.com/forums/readflat.asp?forum=1042&thread=39707815

DM ...

Kind of makes you wonder, why on earth isn't this non-linear data-compression used more often. I mean if you were sitting back and thinking of the problem of a very wide dynamic range and limited hardware resources, the first thing that would pop in your head is non-linear data-compression. I mean yeah you will lose the information in between your data jumps if you could call them that. But there should be a sweet spot for it all. Or is it that the NEX really is suffering from these data jumps?

 highwave's gear list:highwave's gear list
Olympus OM-D E-M5 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G 14mm F2.5 ASPH Olympus M.Zuiko Digital 45mm 1:1.8 Panasonic Lumix G Vario 45-150mm F4-5.6 ASPH Mega OIS +3 more
Reply   Reply with quote   Complain
Leo
Leo
Veteran MemberPosts: 6,015
Like?
It may be a Revolution if...
In reply to highwave, Apr 19, 2012

if if DXO (DR Low quality) and Imaging Resources measured DR (for low, medium and high quality) less than 13V The fans would not allow it. DR of Panasonic sensors, which are small sensors would have hard time to beat a larger APS-C sensor. The more important question: Why is the DR hype if all admirers love the new Oly image quality?

I have got GH2 with 11.6EV total and around 9EV for high quality. The best sensor by Sony (K-5, D7000, D5100, NEX-5n ...) is around 13EV total and 10.3EV at the high quality. I think these are correct numbers. Again if you like the out of camera images then it should not matter. In my new GH2 I like the low ISO IQ (up to and including ISO 800) and love the camera ergonomics at any ISO :-), which for me is very important. At these ISO values I can print up to 13X19. At higher ISO I just would not print but they would be for monitors and projectors. Not a big problem.
Leo

Reply   Reply with quote   Complain
Detail Man
Forum ProPosts: 15,270
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to highwave, Apr 19, 2012

highwave wrote:

Detail Man wrote:

highwave wrote:

I found these threads and posts from a user on DxO forums (dosdan). He gave links to them:

The Sony NEX-7 (for example) uses a non-linear data-compression curve which effectively alters the numerical base associated with each binary bit-representation to a base that is other than the (more conventional) base 2. Such techniques do indeed allow for "data-compression" to occur ...

See: http://forums.dpreview.com/forums/readflat.asp?forum=1042&thread=39707815

DM ...

Kind of makes you wonder, why on earth isn't this non-linear data-compression used more often. I mean if you were sitting back and thinking of the problem of a very wide dynamic range and limited hardware resources, the first thing that would pop in your head is non-linear data-compression.

There are likely various reasons, circuit complexity, cost, speed, etc. ADCs of higher than 12 bits are (to this day) not that easy to construct (even in the form of an entirely stand-alone integrated circuit device, much less integrated near or within an image-sensor structure).

Another reason may well be that the Signal/Noise Ratio of the opto-electronic photo-sensor transducers themselves (due to Photon Shot Noise) degrades very quickly in the presence of illumination by light. Note that the "state of the art" is not much better than around 40 dB at full illumination - which equals a linear data-range of 100, or 6.64 EV (or "bits", or"stops").

Thus, there is little point in constructing data-acquisition devices the Dynamic Range (or the Signal/Noise Ratio) of which is (in most cases of application) significantly higher than that of the original transducer itself ...

Following the image-sensor photo-sites (the opto-electronic transducer) are active electronic MOSFET (N-channel and P-channel, thus called "complimentary", as in "CMOS") buffer amplifiers which may also amplify the voltage-signals derived from the photo-sites themselves. The Dynamic Range (or the Signal/Noise Ratio) of such amplifiers is on the order of around 80 dB (or 13.28 EV, or "bits", or"stops").

No (signal) chain is stronger than it's "weakest link". Note that the limitations of the active MOSFET amplification circuitry itself results in the situation where even a 14-bit ADC is inherently compromised by the noise of the amplifiers (in addition to the noise of the photo-sites themselves when illuminated, which respresents significantly more noise in nearly all situations).

bobn2 is the resident sage about this stuff - perhaps he will choose to grace us with his (much more profound and comprehensive) wisdom regarding these subjects ? The less that he is verbally jumped and ridiculed by ignoramuses and posers, the more likely it might be that he might spend his time for the benefit of our edification and increased knowledge of these interesting subjects ...

I mean yeah you will lose the information in between your data jumps if you could call them that. But there should be a sweet spot for it all. Or is it that the NEX really is suffering from these data jumps?

I have not paid much attention to the Sony technologies. Certainly, nothing good comes without a price(s), and engineering challenges to reliably implement on a large-scale and cost-effective production-level. Logarithmic amplifiers (whether implemented by anolog or semi-digital means) are a rat's-nest of potential trouble. I know (having designed and built analog log-antilog instrumentation circuitry). Integration at micro-sizes can surely help matters, but the very same sensitivities and problems remain on any scale of implementation - especially when built for relatively low production-cost for a consumer-electronics toy that has to turn profits for multi-national corporate behemoths who invoke huge development costs rivalled only by pharmaceutical manuf.s

I know little about it, but the Sony Exmor sensors appear to do some very cool stuff where it comes to ADC processes.. Maybe (if we treat him with respect), bobn2 might teach us some things about what he knows. I myself find these subjects of opto-engineering quite interesting.

Reply   Reply with quote   Complain
Raist3d
Forum ProPosts: 33,339Gear list
Like?
Sounds reasonable
In reply to highwave, Apr 19, 2012

if they did non linear mapping of bits I guess you could represent more too but tonality suffers. And it wouldn't be a trivial thing to do in a visually pleasing way.

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- "You are taking life too seriously if it bugs you in some way that a guy quotes himself in the .sig quote" - Ricardo

Reply   Reply with quote   Complain
Raist3d
Forum ProPosts: 33,339Gear list
Like?
Thanks, so Sony apparently is doing...
In reply to Detail Man, Apr 19, 2012

the non linear RAW file data mapping I was talking about.

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- "You are taking life too seriously if it bugs you in some way that a guy quotes himself in the .sig quote" - Ricardo

Reply   Reply with quote   Complain
Detail Man
Forum ProPosts: 15,270
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Rriley, Apr 19, 2012

Rriley wrote:

the crew that designed it are claiming they were looking for

1/3 to 1/2 and is the highlight. Shadow side has improved stage about 1/2 of itself because there was improvement ratio S / N

"God only knows" what they meant by that ... it seems (to me) that a most reasonable assumtion might be that they were talking in the "language of photography" - that is, "stops" (EV, powers of the base 2, "bits" in the normal binary weighted case). Why such linguistic incoherence would be construed to mean a statement about linear proportions seems (to me) to say more about the emotional hopes and dreams of the general buying public than it does about the actual likelihood ?

If they knew the truth, why would they tell you ?
- Gil Scott Heron

Reply   Reply with quote   Complain
highwave
Senior MemberPosts: 1,041Gear list
Like?
Re: It may be a Revolution if...
In reply to Leo, Apr 19, 2012

Leo wrote:

if if DXO (DR Low quality) and Imaging Resources measured DR (for low, medium and high quality) less than 13V The fans would not allow it. DR of Panasonic sensors, which are small sensors would have hard time to beat a larger APS-C sensor. The more important question: Why is the DR hype if all admirers love the new Oly image quality?

I have got GH2 with 11.6EV total and around 9EV for high quality. The best sensor by Sony (K-5, D7000, D5100, NEX-5n ...) is around 13EV total and 10.3EV at the high quality. I think these are correct numbers. Again if you like the out of camera images then it should not matter. In my new GH2 I like the low ISO IQ (up to and including ISO 800) and love the camera ergonomics at any ISO :-), which for me is very important. At these ISO values I can print up to 13X19. At higher ISO I just would not print but they would be for monitors and projectors. Not a big problem.
Leo

In terms of image quality it doesn't matter at all for me. I am way more than satisfied with the results I see from the OM-D. Heck, I was satisfied with the results of the E-P3 with that Jurassic sensor. Even that sensor was good enough for my needs.

I just posted this thread because I'm really curious about this whole fiasco of the claimed impossibility of 13EV from a 12 bit sensor. It's more of a technology curiosity than a need in IQ capability.

 highwave's gear list:highwave's gear list
Olympus OM-D E-M5 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G 14mm F2.5 ASPH Olympus M.Zuiko Digital 45mm 1:1.8 Panasonic Lumix G Vario 45-150mm F4-5.6 ASPH Mega OIS +3 more
Reply   Reply with quote   Complain
Rriley
Forum ProPosts: 21,846Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

Detail Man wrote:

Rriley wrote:

the crew that designed it are claiming they were looking for

1/3 to 1/2 and is the highlight. Shadow side has improved stage about 1/2 of itself because there was improvement ratio S / N

"God only knows"

(and one or two Japanese)

what they meant by that ... it seems (to me) that a most reasonable assumtion might be that they were talking in the "language of photography" - that is, "stops" (EV, powers of the base 2, "bits" in the normal binary weighted case).

winding back to the beginning of OMD, Oly did make some claim about an increase in DR of 1/3rd or 30%. I forget exactly what was said

Why such linguistic incoherence would be construed to mean a statement about linear proportions seems (to me) to say more about the emotional hopes and dreams of the general buying public than it does about the actual likelihood ?

I do think its meant to entertain the technocrati, not supply it hard skills

If they knew the truth, why would they tell you ?
- Gil Scott Heron

All truths are easy to understand once they are discovered; the point is to discover them.
- Galileo Galilei

-- hide signature --

Riley

any similarity to persons living or dead is coincidental and unintended
support 1022 Sunday Scapes'

 Rriley's gear list:Rriley's gear list
Sigma DP2 Merrill Canon EOS 5D Olympus E-3 Olympus E-5 Panasonic Lumix DMC-GH2 +1 more
Reply   Reply with quote   Complain
Detail Man
Forum ProPosts: 15,270
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Rriley, Apr 19, 2012

Rriley wrote:

Detail Man wrote:

Rriley wrote:

the crew that designed it are claiming they were looking for

1/3 to 1/2 and is the highlight. Shadow side has improved stage about 1/2 of itself because there was improvement ratio S / N

"God only knows"

(and one or two Japanese)

what they meant by that ... it seems (to me) that a most reasonable assumtion might be that they were talking in the "language of photography" - that is, "stops" (EV, powers of the base 2, "bits" in the normal binary weighted case).

winding back to the beginning of OMD, Oly did make some claim about an increase in DR of 1/3rd or 30%. I forget exactly what was said

Why such linguistic incoherence would be construed to mean a statement about linear proportions seems (to me) to say more about the emotional hopes and dreams of the general buying public than it does about the actual likelihood ?

I do think its meant to entertain the technocrati, not supply it hard skills

If they knew the truth, why would they tell you ?
- Gil Scott Heron

All truths are easy to understand once they are discovered; the point is to discover them.
- Galileo Galilei

Everything possible to be believed is an image of truth .
- William Blake

The truth is rarely pure and never simple .
-Oscar Wilde

Truth seems to come with its final word; and the final word gives birth to its next .
- Rabindranath Tagore

Truth is a river that is always splitting up into arms that reunite. Islanded between the arms the inhabitants argue for a lifetime as to which is the main river .
- Cyril Connolly

The truth, my friends, is not eloquent, except unspoken; its vast shadow lends eloquence to our sparks of thought as they die into it .
- George Santayana

Truth often suffers more by the heat of its defenders than from the arguments of its opposers .
- William Penn

We call first truths those we discover after all the others .
- Albert Camus

It's impossible to teach a man what he thinks he already knows .
- Epictetus

Regards,

DM ...

Reply   Reply with quote   Complain
exdeejjjaaaa
Veteran MemberPosts: 6,841Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

Detail Man wrote:

There are likely various reasons, circuit complexity, cost, speed, etc. ADCs of higher than 12 bits are (to this day) not that easy to construct (even in the form of an entirely stand-alone integrated circuit device, much less integrated near or within an image-sensor structure).

well, famous Pentax K10D had 22 bit ADC (off sensor of course) and was quite cheap... I checked my old receipt : Jan 2008 = USD $600 - $100 mail-in rebate, new, body only...

 exdeejjjaaaa's gear list:exdeejjjaaaa's gear list
Olympus E-M1 Sony Alpha 7 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Leica DG Macro-Elmarit 45mm F2.8 ASPH OIS Panasonic Lumix G 14mm F2.5 ASPH +23 more
Reply   Reply with quote   Complain
dotborg
Veteran MemberPosts: 7,652
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to Detail Man, Apr 19, 2012

Detail Man wrote:

Thus, Anders W has calculated and published a Dynamic Range calculated at just under 12.0 EV.

This is what I expect DxO will conclude as well.

-- hide signature --
Reply   Reply with quote   Complain
highwave
Senior MemberPosts: 1,041Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to dotborg, Apr 19, 2012

dotborg wrote:

Detail Man wrote:

Thus, Anders W has calculated and published a Dynamic Range calculated at just under 12.0 EV.

This is what I expect DxO will conclude as well.

I'm personally betting that DxO will publish a 13EV dynamic range.

I have no reason to doubt the results by techradar and even the tangible evidence Pekka Potka's JPEG comparison should be considered.

 highwave's gear list:highwave's gear list
Olympus OM-D E-M5 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G 14mm F2.5 ASPH Olympus M.Zuiko Digital 45mm 1:1.8 Panasonic Lumix G Vario 45-150mm F4-5.6 ASPH Mega OIS +3 more
Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 19,027Gear list
Like?
Re: sources on 13 EV DR possible from 12 bit sensor - Some Facts
In reply to dotborg, Apr 19, 2012

dotborg wrote:

Detail Man wrote:

Thus, Anders W has calculated and published a Dynamic Range calculated at just under 12.0 EV.

This is what I expect DxO will conclude as well.

It might be worth mentioning in this context that the estimate to which Detail Man refers, and which I was able to provide by means of his help, refers to the per-pixel DR (original sensor resolution), and amounts to 11.90 EV. If it were to be converted to the currency used by DxOMark for their sensor scores, where they rely on a normalized resolution of 8 MP, the figure would be about 12.4.

Both estimates are likely to be slightly higher than those eventually reported by DxOMark due to a slight difference between the formula they use for computing DR and the one used by Detail Man and me. But the practical difference between the two formulas/definitions is small so this shouldn't matter much.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads