Why did the DxOmark thread get locked?

[eronald]

We had this nice thread going with lots of links to reference material - what happened here?

I'm playing with my D4. So far, so good.

Edmund

[BJL]

Yes, some good discussions, but a lot of it way out of place in the "MF, etc." forum. Maybe some topics should be revived in new threads, probably under "Cameras, Lenses and Shooting Gear".

I was going to explain the reasons why compact cameras often have a lower base ISO speed than most DSLRs, which is mainly due to their designs opting for higher exposure levels and thus less highlight headroom in exchange for less shadow noise. Hardly worth a new thread just for that though.

[jduncan]

We had this nice thread going with lots of links to reference material - what happened here?

I'm playing with my D4. So far, so good.

Edmund

I believe that it is a defensive move. We have a general respectful and informed tone here. Some people don't. I was loving the thread and ignoring the provocateur comments. Nevertheless, after what happened to the Rob Galbraith forums, I prefer the Luminous Landscape "err on the side of caution" approach.

Best regards,

James

[Kirk Gittings]

Hmmmm....on the forum where I am a moderator we would at least post a brief explanation when locking a thread.

[eronald]

I was going to explain the reasons why compact cameras often have a lower base ISO speed than most DSLRs, which is mainly due to their designs opting for higher exposure levels and thus less highlight headroom in exchange for less shadow noise. Hardly worth a new thread just for that though.

Well, anyway, I have found that whether it is the Phase or the D4, high ISO use is in practice impacted much more by sensor non-uniformities and correlated noise than by gaussian noise.

If I may be allowed the joke, if someone were to watermark a night picture with a few light dots writing his name in a dark picture zone, it would be visible although the global S/N measure would be hardly modified.

In the same way sensor non-uniformities often impact night shots or indoors shots where dark backgrounds or dark clothes get streaked well before the main subject is impacted. The correlated noise is a problem, although the noise measure would be negligible.

The "not a cat" image works so well precisely because there are no really dark areas.

Of course, once people's attention is drawn to the issues of non-uniformities and streaking, software can be written to test for these issues. I am surprised that standard tests do not already incorporate these questions.

Edmund

BTW, re. the compacts - do they do highlight recovery to generate the Jpegs?

[BJL]

re. the compacts - do they do highlight recovery to generate the Jpegs?

As fas I know, compacts typically place the metered mid-tones at a level that is about 18/106 of maximum, whereas larger format cameras (and the saturation based ISO definition of sensitivity, a.k.a. "base-ISO speed") place that metered level at about 18/140 or a bit lower. Thus even with the same saturation based sensitivity (base ISO speed), the compact is choosing a longer shutter speed to push the wells closer to full, and thus a lower Exposure Index. This was indicated in a Kodak document years ago. And this is completely allowed by the definitions of sensitivity in the ISO12232:2006 standard that the Japanese industry association CIPA allows its members to use, in which there is a choice of using Standard Output Sensitivity (SOS) or Recommended Exposure Index (REI). The latter is indeed a recommendation, so the a camera can recommend a lower Exposure Index and thus using a longer exposure time under the same lighting, giving less highlight headroom than is typical with a larger format, for the sake of less shadow noise. That is, an exposure index (REI) lower than the sensitivity (SOS), the latter measured from how 18% gray input is rendered in sRGB output with gamma=2.2, typically JPEG, rather then from raw files.

As to highlight recovery, I believe that the lower EI rating is used to choose a lower shutter speed, so that photo-sites under the mid-tones get an exposure closer to full well capacity, rather than there being a "push" in raw-to-JPEG conversion, because that is the main way to reduce shadow noise.
If I am right on this, the DxO "headroom based" measurement of EI values would rate such a compact well above the camera's stated rating, but the camera would give shutter speeds as indicated by its own rating, not the DxO number.

[bjanes]

Hmmmm....on the forum where I am a moderator we would at least post a brief explanation when locking a thread.

It would seem as if the locking of the thread has something to do with a slight dust-up between one of Michael's buddies, Yair Shahar, with Emil Martinec where Yair replied in a dismissive way to a post made by Emil (who is well known on LuLa) and suggested to Emil that "perhaps it is time for you to step out of the shadows and tell us who you are and what you do? It seems silly to talk to a nickname...". This is somewhat ironic, since Yair goes by the screen name of YaYa, although he does give his name as a footnote to his posts. Emil is a professor of physics at the University of Chicago and the discussion devolved to how a PhD in physics pertains to photographic expertise and then to a less than well received article on LuLa by another PhD.

I agree with Edmund that the thread did cover some interesting topics. Discussion of high end dSLRs is not really off topic in a MFDB forum, since many prospective users are interested in how the two formats compare. Referring to the Nikon D7000 sensor as thumbnail sized shows condescension and did not add to the demeanor of the discussion.

Regards,

Bill

[jeremypayne]

Hmmmm....on the forum where I am a moderator we would at least post a brief explanation when locking a thread.

isn't it possible that the thread was locked by the OP and not the mods?

[Kirk Gittings]

I don't think so. Looking back at threads I have started, it seems I can remove a thread but not lock it by myself.

[ErikKaffehr]

Hi,

No, an OP can lock a thread, I think. I never did it myself. There is a checkbox under modify to do that.

Best regards
Erik

I don't think so. Looking back at threads I have started, it seems I can remove a thread but not lock it by myself.

[yaya]

It would seem as if the locking of the thread has something to do with a slight dust-up between one of Michael's buddies, Yair Shahar, with Emil Martinec where Yair replied in a dismissive way to a post made by Emil (who is well known on LuLa) and suggested to Emil that "perhaps it is time for you to step out of the shadows and tell us who you are and what you do? It seems silly to talk to a nickname...". This is somewhat ironic, since Yair goes by the screen name of YaYa, although he does give his name as a footnote to his posts. Emil is a professor of physics at the University of Chicago and the discussion devolved to how a PhD in physics pertains to photographic expertise and then to a less than well received article on LuLa by another PhD.

I agree with Edmund that the thread did cover some interesting topics. Discussion of high end dSLRs is not really off topic in a MFDB forum, since many prospective users are interested in how the two formats compare. Referring to the Nikon D7000 sensor as thumbnail sized shows condescension and did not add to the demeanor of the discussion.

Regards,

Bill

Bill, some pointers for you if I may

I am not "one of Michael's buddies". We know each-other since 2006 and we've had professional dealings (including my visit to mexico last year) over the years. If you have anything to say to me you can contact me directly 24/7/365 as you have all my contact details (inc. mobile number) and full name in my signature

I did not know who Emil was until he cleared it up. I should have apologised for not looking him up. I only ever visit the MF forum so I did not know him from anywhere else

I'm a happy Nikon user of a camera with a thumbnail-size sensor. I like the camera but I'm also aware of its limitations. I dismissed the images that were used by Emil since I did not find them very informative. This was my opinion and I still stand behind it. I still think the DxO's reports only tell half the story of a camera (or less) hence why I was joking about the differences between my F5100 and the D7000 in their comparison...

I have no science degree which is why you will never catch me waving around scientific terms. I speak photography to photographers regardless of their genre, background or interest. I'm here to share my experience and to help those who want to learn about digital medium format photography

Edmund and I have been around the block a few time and I'd like to believe that we know each other's crooked sense of humour by now. I think we also share the belief that an image is made out of light, subject, colour, texture etc. and not out of photons and microns...

Anyway it's getting late here and I had a long day travelling to another country and back.

Again feel free to contact me at any time if you need any further clarifications

Yair

(PS I got "yaya" as a nickname since the army days and I've been using it as a screen name for many years but always attached to my real name)

[david distefano]

i didn't lock it. i thought it was something i said in my last post. or i thought it was locked because i used the screen name instead of the correct name and for that,  being new, if that is a no no i apologize. i don't capitalize anything so i hope it wasn't because i didn't do that to a proper name. if i went off tangent maybe it could have been moved to another category?

[eronald]

I did not know who Emil was until he cleared it up. I should have apologised for not looking him up. I only ever visit the MF forum so I did not know him from anywhere else

Well Guys, consider yourselves introduced

Edmund and I have been around the block a few time and I'd like to believe that we know each other's crooked sense of humour by now. I think we also share the belief that an image is made out of light, subject, colour, texture etc. and not out of photons and microns...

(PS I got "yaya" as a nickname since the army days and I've been using it as a screen name for many years but always attached to my real name)

Absolutely. I'm sure Yair agrees that dissent contains a lot of information, while anything that confirms your beliefs is mostly noise (to paraphrase Shannon).

Let's chalk this up to accident, and let's resume talking about MF (and other) sensors.

Edmund

PS. Yair's remark on texture is interesting. Have you guys seen Image Engineering's paper on texture loss?
http://www.image-engineering.de/index.php?option=com_content&view=article&id=570&Itemid=210

I find that the D4 has exceptionally good skin texture rendering when used indoors in available light.

[bjanes]

Absolutely. I'm sure Yair agrees that dissent contains a lot of information, while anything that confirms your beliefs is mostly noise (to paraphrase Shannon).

Let's chalk this up to accident, and let's resume talking about MF (and other) sensors.

+1 on this.

PS. Yair's remark on texture is interesting. Have you guys seen Image Engineering's paper on texture loss?
http://www.image-engineering.de/index.php?option=com_content&view=article&id=570&Itemid=210

I find that the D4 has exceptionally good skin texture rendering when used indoors in available light.

The article to which you gave the link is quite interesting, and the term "dead leaves" is poetic. In that article, the Pentax K5 comes out much better in the colored dead leaves model than the Nikon D7000, which is a bit surprising since they both use the same or closely related sensors. On the monochrome model, they are more or less equivalent. The variables accounting for this difference would be most illuminating. A quick Google Scholar search using the search term "dead leaves target" turned up this more rigorous treatise as the most cited.

It would be interesting to apply this model to images taken with the top dSLRs and top MFDBs. DXO does not test resolution, much less texture.

Regards,

Bill

[eronald]

When digital was yet in infancy, an AD friend once told me: You have the unique ability to render women as porcelain dolls.
I took it as a compliment

Edmund

+1 on this.

The article to which you gave the link is quite interesting, and the term "dead leaves" is poetic. In that article, the Pentax K5 comes out much better in the colored dead leaves model than the Nikon D7000, which is a bit surprising since they both use the same or closely related sensors. On the monochrome model, they are more or less equivalent. The variables accounting for this difference would be most illuminating. A quick Google Scholar search using the search term "dead leaves target" turned up this more rigorous treatise as the most cited.

It would be interesting to apply this model to images taken with the top dSLRs and top MFDBs. DXO does not test resolution, much less texture.

Regards,

Bill

[EricWHiss]

MD PhD

[LKaven]

A quick Google Scholar search using the search term "dead leaves target" turned up this more rigorous treatise as the most cited.

Thank you for supplying that interesting reference.