Velocity Reviews

Velocity Reviews (http://www.velocityreviews.com/forums/index.php)
-   Digital Photography (http://www.velocityreviews.com/forums/f37-digital-photography.html)
-   -   Re: This is not an advertisement ... (http://www.velocityreviews.com/forums/t955772-re-this-is-not-an-advertisement.html)

otter 12-23-2012 01:57 PM

Re: This is not an advertisement ...
 
On Dec 21, 3:18*am, Eric Stevens <eric.stev...@sum.co.nz> wrote:
> This is not an advertisement but you should know that DXO has a
> special deal running up to 25th December.
>
> Why am I telling you this?
>
> Well, I was sniffing around the DXO site in the process of evaluating
> lenses when I came acrosshttp://www.dxo.com/intl/photoand similar.
>
> "Not another hot deal I said to myself" (said I) when I noticed the
> existence of 31 days free trial. So I read the blurb, thought a while,
> and then downloaded the 'Standard' edition of DXO Optics Pro 8. After
> all, if I don't like it I can give it back at no cost to me.
>
> I played around with it and then sent in my US$99. I've continued to
> play around with it and have to say that I am generally blown away.
>
> It's got a plethora of adjustments, most of which include an
> 'Automatic' setting. At first I twiddled everything and now I tend to
> leave most of it alone. I won't say it extracts detail and texture
> which no other software can match, cos it doesn't. I will say it
> extracts detail and texture with an ease which no other software that
> I have used can match. When I started I used to fiddle with every
> accessible setting but now I'm finding that I can usually leave DXO to
> look after most of its own settings.https://dl.dropbox.com/u/31088803/_DSC3087_DxO.tifis an example which
> is much cleaner and crisper than the example I posted in [SI] a few
> months ago.
>
> All I'm suggesting that if you think you might be interested you
> should download and try it now before US$99 deal runs out. Have fun.
> --
>
> Regards,
>
> Eric Stevens


Your example picture does look nice. Unfortunately, I'd have to buy
the $199 elite version, and they don't support 2 of my 3 lenses.

I noticed one feature was DNG output. Does that work OK? Do you
really get a RAW file that you can then feed into Lightroom?

Robert Coe 12-24-2012 07:56 AM

Re: This is not an advertisement ...
 
On Mon, 24 Dec 2012 11:35:44 +1300, Eric Stevens <eric.stevens@sum.co.nz>
wrote:
: On Sun, 23 Dec 2012 05:57:07 -0800 (PST), otter
: <bighorn_bill@hotmail.com> wrote:
:
: >On Dec 21, 3:18*am, Eric Stevens <eric.stev...@sum.co.nz> wrote:
: >> This is not an advertisement but you should know that DXO has a
: >> special deal running up to 25th December.
: >>
: >> Why am I telling you this?
: >>
: >> Well, I was sniffing around the DXO site in the process of evaluating
: >> lenses when I came acrosshttp://www.dxo.com/intl/photoand similar.
: >>
: >> "Not another hot deal I said to myself" (said I) when I noticed the
: >> existence of 31 days free trial. So I read the blurb, thought a while,
: >> and then downloaded the 'Standard' edition of DXO Optics Pro 8. After
: >> all, if I don't like it I can give it back at no cost to me.
: >>
: >> I played around with it and then sent in my US$99. I've continued to
: >> play around with it and have to say that I am generally blown away.
: >>
: >> It's got a plethora of adjustments, most of which include an
: >> 'Automatic' setting. At first I twiddled everything and now I tend to
: >> leave most of it alone. I won't say it extracts detail and texture
: >> which no other software can match, cos it doesn't. I will say it
: >> extracts detail and texture with an ease which no other software that
: >> I have used can match. When I started I used to fiddle with every
: >> accessible setting but now I'm finding that I can usually leave DXO to
: >> look after most of its own settings.https://dl.dropbox.com/u/31088803/_DSC3087_DxO.tifis an example which
: >> is much cleaner and crisper than the example I posted in [SI] a few
: >> months ago.
: >>
: >> All I'm suggesting that if you think you might be interested you
: >> should download and try it now before US$99 deal runs out. Have fun.
: >> --
: >>
: >> Regards,
: >>
: >> Eric Stevens
: >
: >Your example picture does look nice. Unfortunately, I'd have to buy
: >the $199 elite version, and they don't support 2 of my 3 lenses.
: >
: >I noticed one feature was DNG output. Does that work OK? Do you
: >really get a RAW file that you can then feed into Lightroom?
:
: On rereading your article I realised I had misunderstood the question
: of your last paragraph. DXO Optics does not pass on a raw file to
: Lightroom. The output of DXO Optics is an image file of the kind you
: have selected (JPG, TIFF, DNG) with the variations you have selected
: and it is this which is passed on to lightroom. In other words, DXO
: Optics does the raw conversion subject to all the bells, whistles,
: levers and controls embedded with DXO Optics.

But isn't the point of DNG that it serves as a sort of vendor-independent RAW
mode? I interpreted Otter's question to mean, "Is DXO's conversion to DNG good
enough that Lightroom will accept a DXO-generated DNG file as a RAW file for
processing purposes?" To which I understood your answer to be, "That appears
to be the intent, although I haven't tried it myself." Is that about right, or
do I misunderstand the purpose of DNG?

Bob

otter 12-26-2012 04:45 PM

Re: This is not an advertisement ...
 
On Dec 25, 5:18*pm, Eric Stevens <eric.stev...@sum.co.nz> wrote:
> On Tue, 25 Dec 2012 19:03:11 +1300, Eric Stevens
>
>
>
>
>
>
>
>
>
> <eric.stev...@sum.co.nz> wrote:
> >On Mon, 24 Dec 2012 18:41:58 -0800, Savageduck
> ><savageduck1@{REMOVESPAM}me.com> wrote:

>
> > *--- snip ---

>
> >>Even stranger, when I checked the EXIF of that 536x356 DNG it is
> >>supposed to be a 2848x4288 12.2 MP DNG.

>
> >>I am trying to get my head wrapped around just what might be happening with DB.

>
> >Me too!

>
> >I find I start of with (an image size of) 4288 x 2848 as a NEF file,
> >1072 x 712 as a DNG and 4000 x 2656 as a TIFF.

>
> >DXO offers the ability to change the size of an image when saving but
> >not with DNG.

>
> >I will write to DNG and ask them to explain what is going on.

>
> >I note that in their manual they say of DNG:

>
> > * *"The DNG format is designed for files that you intend to archive
> > * * or that are going to undergo further post-processing with
> > * * DNG compatible software such as Adobe Camera
> > * * Raw/Photoshop/Lightroom. The DNG generated by DxO Optics Pro isa
> > * * 16-bit linear DNG format, which is only available for original
> > * * images shot in RAW.

>
> > * * With linear DNG, the three color channels have been individually
> > * * calculated, which means that the file is three times larger
> > * * than the original RAW file. This format offers the same level of
> > * * color and exposure control as a camera-generated RAW file.

>
> > * * For the other options, you can treat DNG files in the same way as
> > * * TIFF files — except for image size, as files intended for
> > * * archiving and/or further post-processing should not be resized
> > * * (so as to maintain maximum potential)."

>
> I haven't yet written to DXO but I have done a little more digging.
> When choosing a file type to save as there is a box "Allow
> resampling". If this is ticked then up pops a window which allows the
> maximum image dimension to be set in pixels. I found these to be all
> set to different dimensions. I set them all to 4000 pixels and both
> TIFF and JPG produced 4000 x 2656. However DNG (according to windows
> properties) was 1000 x 664: as previously, this was one quarter of the
> window size reported by others.
>
> Being yesterday surrounded by family, Christmas cheer and hurly
> burley, I wasn't thinking very clearly and it only belatedly dawned on
> me that Paint Shop Pro X5 claims the ability to import DNG files. I
> thought I would try it on the DXO fruit bowl DNG.
>
> Lo and behold it works perfectly. It read the DXO file as a RAW file
> and displayed it's contents as 4000 x 2656 image, exactly what it
> should have been. It appears as though the small size reported by
> Windows is a Windows problem, not a DXO problem.
>
> All of this has caused me to go hunting Linear DNG files. I found URLs
> such ashttp://www.barrypearson.co.uk/articles/dng/linear.htm*andhttp://dpbestflow.org/file-format/raw-file-formats
>
> There are a great many more that tend to add more confusion than
> enlightenment. A great part of the problem is that people tend not
> understand the fundamentals of how things work (and that includes me
> on this subject) and try to explain the topic in terms of some other
> type of file - which often they don't seem to properly understand
> either.
>
> One of the points which emerged is that the DXO output DNG is a proper
> file format which can be read by all the software which has been
> worrying people up to now.
> --
>
> Regards,
>
> Eric Stevens


You guys have been busy! Thanks for the research. What I got from
this, though, is that there seems to be no advantage to using DxO
generated DNG files to feed to LR over using a TIFF, is that right?

Well, it is academic to me, anyway, until DxO supports the lenses I'm
interested in.

otter 12-27-2012 06:10 AM

Re: This is not an advertisement ...
 
On Dec 26, 6:05*pm, Eric Stevens <eric.stev...@sum.co.nz> wrote:
> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>
>
>
>
>
>
>
>
>
> <savageduck1@{REMOVESPAM}me.com> wrote:
> >On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stev...@sum.co.nz> said:

>
> >> On Wed, 26 Dec 2012 09:33:16 -0800, Savageduck
> >> <savageduck1@{REMOVESPAM}me.com> wrote:

>
> >> * *--- snip ---

>
> >>>> You guys have been busy! *Thanks for the research. *What I got from
> >>>> this, though, is that there seems to be no advantage to using DxO
> >>>> generated DNG files to feed to LR over using a TIFF, is that right?

>
> >>> That is the way it appears to me. The only difference would be file size.
> >>> This could be problematic for folks receiving a DXO generated DNG with
> >>> the implied promise the would have a RAW equivalent or "digital
> >>> negative" to work with. Better to have the actual RAW file.

>
> >>>> Well, it is academic to me, anyway, until DxO supports the lenses I'm
> >>>> interested in.

>
> >>> I am not in the market for DXO.

>
> >>> I am quite happy with LR4 and ACR, and I am always able to find lens &
> >>> camera profiles for the Adobe products. If the lens profile isn't
> >>> available, as was the case when I first got my Tokina 11-16mm f/2.8, it
> >>> is easy enough to create your own with Adobe Lens Profile Creator.
> >>> For Mac:
> >>> <http://www.adobe.com/support/downloads/detail.jsp?ftpID=5489>

>
> >>> ...and for Windows:
> >>> <
> >>>http://www.adobe.com/support/downloa...uct=193&platfo...

>
> >> I can't speak about the comparitive quality of the respective
> >> corrections but one thing tha DXO does that Adobe seems to lack is
> >> described in the DXO manual is:

>
> >> * "DxO Lens Softness
> >> * *About lens softness
> >> * *The exclusive DxO Lens Softness tool is one of the major strengths
> >> * *of DxO Optics Pro. Lens softness is an optical aberration in
> >> * *which a point is transformed by the lens into a small blurred
> >> * *circle. (This should not be confused with out-of-focus or motion
> >> * *blur, which DxO Optics Pro is unable to correct). DxO Optics
> >> * *Modules have been created by measuring the amount of blur at
> >> * *every point in the image area for each camera body and lens
> >> * *combination.

>
> >> * *Matching the shooting parameters contained in the EXIF data
> >> * *(aperture, focal length, etc.) and the map provided by the
> >> * *DxO Optics Module, DxO Optics Pro can apply precisely-tailored
> >> * *corrections according to the position of each pixel in the
> >> * *image field. This correction is not the same over the whole image,
> >> * *since lenses are sharper in the center; thus central pixels
> >> * *need less correction."

>
> >While that is an interesting take on DXO's proprietary solution to a
> >"lens softness" issue, and their self critique and praise of their
> >system. As some have also claimed, such "lens softness" can also be
> >attributed to a need for said lens to be calibrated with the camera on
> >which it is to be used.

>
> 1. *DXO _does_ calibrate a lens and camerea combinations.
>
> 2. *The lens softness is not a constant over the image but varies from
> point to point. This is mapped by DXO and applied accordingly. This
> generally enables more detail to be recovered at the edges of the
> image than at the centre.
>
>
>
> >I suspect they are trying to differentiate their product from those
> >produced by Adobe and some camera manufacturers.

>
> I think they are just describing what they do.
>
>
>
> >One issue is, not all CS3/4/5/6 and LR users even bother to use lens
> >profiles in their workflow. So they miss out on the benefits of
> >correcting many of the issues in a particular lens's DNA. "Lens
> >Softness" may well be one of those issues, but Adobe does not speak of
> >it in the same terms as DXO.

>
> As far as I can tell Adobe does not speak of it at all.
>
>
>
> >> As I remarked in passing the other day, the DXO implementation of this
> >> correction enabled me to recover detail from my old D70 + 18~70 RAW
> >> files which made them comparable in quality with images produced by my
> >> D300 with the 16~80. The visual effect is quite remarkable.

>
> >...and you might well have experienced a similar benefit had you been
> >using LR4 or CS4/5/6 and made sure you had the 18-70mm + D70 (RAW) lens
> >profile installed.
> ><https://dl.dropbox.com/u/1295663/FileChute/screenshot_87.jpg>
> >BTW: The 18-70mm is a particularly good lens for the money, and I would
> >not describe it as being afflicted with "lens softness". I still have
> >my D70 & 18-70mm

>
> Virtually all lenses have softness somewhere in their images. Some
> might have less than others but they will still haved it.
>
>
>
> >You might gain a similar benefit if you checked the calibration of your
> >lens/camera pairings using a tool such as LensAlign.
> ><http://michaeltapesdesign.com/lensalign.html>

>
> That's a an auto-focus checking tool. I've already done that. I've
> been waiting (literally for months) for the weather to clear
> sufficiently to enable me to carry out more sophisticated tests on the
> 7~200.
>
> [You wouldn't believe it: we have had strong winds, heavy rain, thick
> clouds, a near miss (less than 100 yds) from a a near tornado and now
> we have the residue of a tropical depression sitting on us giving with
> low clouds, high humidity and drizzle. We have had three clear days in
> the last three months and I was committed elsewhere at the time.]
>
>
>
> >DXO is just another solution for some issues, and as much as any user
> >of a particular piece of software is going to praise the perceived
> >features/benefits of whatever they are using, and are comfortable with
> >the results, they will be happy.
> >That is true for me with LR4 & ACR+CS5, as it was for you with the
> >Nikon software, as it is for those happy with UFRaw & GIMP, or the
> >Canon shooters using DPP.

>
> I wasn't putting up DXO in a comparitive or competitive sense. In fact
> I wasn't putting it up at all. What I was doing was advising those who
> might be interested was that until the 25th December DXO was available
> at a considerable discount.
>
>
>
> >Now you have been exposed to something different to the Nikon offering
> >with DXO and you have experienced an improvement over your earlier
> >experience, and that is a good thing. There is little point in pushing
> >Adobe products on you, mainly due to the punitive antipodean pricing.

>
> I may be able to deal with that in the forseeable future. I have a
> friend who is a pilot and he expects to be switching from London
> routes to LA. We shall see.
>
>
>
> >The important thing is you have DXO in hand, and you are satisfied with
> >the results. In your situation I would probably be just as happy.

>
> From what I presently know, until Adobe deals with the lens softness
> issue, I will use DXO with _all_ its lens corrections as a
> preprocessor/raw converter for CR6/LR.
>
> Free trials of DXO are still available and I would like to hear from
> someone who normally uses lens corrections in CR6/LR.
> --
>
> Regards,
>
> Eric Stevens


Yeah, DxO seems to be able to add extra resolution to a lens in a way
that is different from normal "sharpening". At least, some lenses,
that is. The DLO feature in DPP appears to do a similar thing. I've
seen some impressive results from some people who have used it. It
goes beyond the distortion, CA, and vignetting correction that you get
with LR/ACR lens profiles.

It seems to do better with lenses that are less sharp to begin with,
though. I can't tell the difference with the 24-70II or the 70-200 f/
2.8 IS II.

Wolfgang Weisselberg 12-27-2012 06:59 PM

Re: This is not an advertisement ...
 
Eric Stevens <eric.stevens@sum.co.nz> wrote:
> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>>On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stevens@sum.co.nz> said:


>>> "DxO Lens Softness
>>> About lens softness

[...]
>>> Matching the shooting parameters contained in the EXIF data
>>> (aperture, focal length, etc.) and the map provided by the
>>> DxO Optics Module, DxO Optics Pro can apply precisely-tailored
>>> corrections according to the position of each pixel in the
>>> image field. This correction is not the same over the whole image,
>>> since lenses are sharper in the center; thus central pixels
>>> need less correction."


>>While that is an interesting take on DXO's proprietary solution to a
>>"lens softness" issue, and their self critique and praise of their
>>system. As some have also claimed, such "lens softness" can also be
>>attributed to a need for said lens to be calibrated with the camera on
>>which it is to be used.


> 1. DXO _does_ calibrate a lens and camerea combinations.


But not your lens to your camera. If all cameras and lenses were
identical, we'd not need microfocus settings.

All that's needed is that the bayonet not be perfectly parallel to
the sensor, and there you go: one side is different from the other.
With luck DXOs version is tilted slightly just the other way.

> 2. The lens softness is not a constant over the image but varies from
> point to point. This is mapped by DXO and applied accordingly. This
> generally enables more detail to be recovered at the edges of the
> image than at the centre.


And you assume that for good lenses the other errors don't
outweight that change?


>>I suspect they are trying to differentiate their product from those
>>produced by Adobe and some camera manufacturers.


> I think they are just describing what they do.


What you think doesn't matter.

What matters is if what they do (not what they claim to do)
makes a noticable difference for the *user's* gear.


>>One issue is, not all CS3/4/5/6 and LR users even bother to use lens
>>profiles in their workflow. So they miss out on the benefits of
>>correcting many of the issues in a particular lens's DNA. "Lens
>>Softness" may well be one of those issues, but Adobe does not speak of
>>it in the same terms as DXO.


> As far as I can tell Adobe does not speak of it at all.


Maybe because it's mostly a non-issue for it's clientel?

[...]
> Virtually all lenses have softness somewhere in their images. Some
> might have less than others but they will still haved it.


Virtually all photos are somewhat misfocussed. Some might have
less misfocus than others bit they will still haved[sic] it.

>>You might gain a similar benefit if you checked the calibration of your
>>lens/camera pairings using a tool such as LensAlign.
>>< http://michaeltapesdesign.com/lensalign.html >


A better benefit. Unless your lens is perfectly calibrated
to your body, well, DXO doesn't fix misfocussing.

And I don't think just doing microfocus adjustment is enough.

Best start with having at least the centre always and consistenly
sharp before worrying about the corners.

-Wolfgang

Wolfgang Weisselberg 01-10-2013 05:30 PM

Re: This is not an advertisement ...
 
Eric Stevens <eric.stevens@sum.co.nz> wrote:
> On Thu, 27 Dec 2012 19:59:08 +0100, Wolfgang Weisselberg
> <ozcvgtt02@sneakemail.com> wrote:


>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>>>>On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stevens@sum.co.nz> said:


>>>>> "DxO Lens Softness
>>>>> About lens softness

>>[...]
>>>>> Matching the shooting parameters contained in the EXIF data
>>>>> (aperture, focal length, etc.) and the map provided by the
>>>>> DxO Optics Module, DxO Optics Pro can apply precisely-tailored
>>>>> corrections according to the position of each pixel in the
>>>>> image field. This correction is not the same over the whole image,
>>>>> since lenses are sharper in the center; thus central pixels
>>>>> need less correction."


>>>>While that is an interesting take on DXO's proprietary solution to a
>>>>"lens softness" issue, and their self critique and praise of their
>>>>system. As some have also claimed, such "lens softness" can also be
>>>>attributed to a need for said lens to be calibrated with the camera on
>>>>which it is to be used.


>>> 1. DXO _does_ calibrate a lens and camerea combinations.


>>But not your lens to your camera. If all cameras and lenses were
>>identical, we'd not need microfocus settings.


> --- snip ---


> DXO doesn't fix people who are determined to argue, no matter what.


>>Best start with having at least the centre always and consistenly
>>sharp before worrying about the corners.


> http://news.cnet.com/8301-11386_3-57...ns-latest-slr/
> is an interesting read about DXO.


Please note carefully where they do NOT say that they test a
couple copies of each camera and each lens. Much less a dozen
or more copies. Then head over to the lensrentals blog and
read how far 'identical' lenses and cameras can differ from
each other when it comes to the region of improvements where
DxO claims to work with. (And that's not even including bad
lenses and cameras.)

As I said: if the cameras and lenses really were identical,
there'd be no need for microfocus settings and photographers
wouldn't need to match lenses to cameras and wouldn't at the
high end send them in together to make them matched pairs.
This should be *blindingly* *obvious* even to rabid DxO fans.

So you tell me why a random lens copy with a random camera
copy, each having random variations from the ideal model is
enough to tell how *your* copies with *their own*, *different*
random variations behaves? If it's a bad lens design, then
sure: the bad design overshadowes all the variations (unless
the lens/camera is broken). But people using that would use
a good lens in the first place, and there the lens-to-lens
and camera-to-camera variations are not tiny compared to the
systematic lens defects!

-Wolfgang

Wolfgang Weisselberg 01-10-2013 05:54 PM

Re: This is not an advertisement ...
 
Eric Stevens <eric.stevens@sum.co.nz> wrote:
> On Fri, 28 Dec 2012 11:35:58 +1300, Eric Stevens
>>On Thu, 27 Dec 2012 19:59:08 +0100, Wolfgang Weisselberg
>>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>>> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>>>>>On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stevens@sum.co.nz> said:


>>>>>> "DxO Lens Softness
>>>>>> About lens softness
>>>[...]
>>>>>> Matching the shooting parameters contained in the EXIF data
>>>>>> (aperture, focal length, etc.) and the map provided by the
>>>>>> DxO Optics Module, DxO Optics Pro can apply precisely-tailored
>>>>>> corrections according to the position of each pixel in the
>>>>>> image field. This correction is not the same over the whole image,
>>>>>> since lenses are sharper in the center; thus central pixels
>>>>>> need less correction."


> I've also discovered
> http://www.computer-darkroom.com/blo...tion-profiles/ which
> deals with the creation of Adobe lens profiles. It sounds similar to
> to the DXO method but not identical.


| This application characterises three common types of lens
| aberrations, namely geometric distortion, lateral chromatic
| aberration, and vignetting.

Reading THAT should have told you that there's not a single
bit of "lens softness correction" in there. A bit of
research would have told you that it's also an old hat.

> The wording of the article
> suggests that profiles may not all be created by Adobe.


| If you’re into making your own lens profiles the Lens Profile
| Creator application can be download from Adobe Labs.

It plain says it. No suggestions here.

> The procedure only "characterises three common types of lens
> aberrations, namely geometric distortion, lateral chromatic
> aberration, and vignetting". There is no mention of the measurement of
> what DXO calls lens 'softness'. Nor is there any mention of color.


Look up "chroma" and "chrominance". Contrast to "luminance"

> In particular, unlike DXO, what is being measured appears to be the
> behaviour of a particular lens/camera combination.


You meant to say "exactly like DxO, however, unlike DxO it'll
be your camera and your lens you test".

> so you guys with
> (say) a D800, how do you feel about relying on a lens profile created
> with (say) a Canon Rebel?


The guys with a D800 will of course use lens profiles generated
with *their* D800 and *their* lenses. People who don't care
so much --- i.e. not D800 owners --- will download one created
with a compatible type (e.g. Nikon 1.5x crop) and the same lens
(different copy, though).

If you had the slightest idea of the whole process --- i.e.
spend less than 30 minutes reading --- you'd know that your
scenario would only happen when someone took the same lens,
mounted it on a camera where it didn't fit natively, extended
the Rebel's sensor to the larger crop of Nikon and then measured
the result. Which would be quite a feat.

-Wolfgang

Wolfgang Weisselberg 01-20-2013 05:41 PM

Re: This is not an advertisement ...
 
Eric Stevens <eric.stevens@sum.co.nz> wrote:
> On Thu, 10 Jan 2013 18:54:12 +0100, Wolfgang Weisselberg
>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>> On Fri, 28 Dec 2012 11:35:58 +1300, Eric Stevens
>>>>On Thu, 27 Dec 2012 19:59:08 +0100, Wolfgang Weisselberg
>>>>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>>>>> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>>>>>>>On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stevens@sum.co.nz> said:


>>>>>>>> "DxO Lens Softness
>>>>>>>> About lens softness
>>>>>[...]
>>>>>>>> Matching the shooting parameters contained in the EXIF data
>>>>>>>> (aperture, focal length, etc.) and the map provided by the
>>>>>>>> DxO Optics Module, DxO Optics Pro can apply precisely-tailored
>>>>>>>> corrections according to the position of each pixel in the
>>>>>>>> image field. This correction is not the same over the whole image,
>>>>>>>> since lenses are sharper in the center; thus central pixels
>>>>>>>> need less correction."


>>> I've also discovered
>>> http://www.computer-darkroom.com/blo...tion-profiles/ which
>>> deals with the creation of Adobe lens profiles. It sounds similar to
>>> to the DXO method but not identical.


>>| This application characterises three common types of lens
>>| aberrations, namely geometric distortion, lateral chromatic
>>| aberration, and vignetting.


>>Reading THAT should have told you that there's not a single
>>bit of "lens softness correction" in there.


> It did.


So you should have known that it's similar only in the way
that it applies mathematical operations on digital images.


>>A bit of
>>research would have told you that it's also an old hat.


>>> The wording of the article
>>> suggests that profiles may not all be created by Adobe.


>>| If youÂ’re into making your own lens profiles the Lens Profile
>>| Creator application can be download from Adobe Labs.


>>It plain says it. No suggestions here.


> Have you considered the possibility that some of the independently
> created profiles may be fed back to Adobe?


Sure.

> I know of nothing to
> indicate where individual profiles supplied by Adobe may have come
> from.


| suggests that profiles may not all be created by Adobe.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Your point was?


>>> The procedure only "characterises three common types of lens
>>> aberrations, namely geometric distortion, lateral chromatic

^^^^^^^^^^
>>> aberration, and vignetting". There is no mention of the measurement of
>>> what DXO calls lens 'softness'. Nor is there any mention of color.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

>>Look up "chroma" and "chrominance". Contrast to "luminance"

^^^^^^^^^^^^^^^^

> None of which are mentioned in connection with the Adobe profile
> procedure or data.


I've underlined the parts you should reread.

I've also noticed that you seem to have misplaced your
| It sounds similar to
| to the DXO method but not identical.
claim.

>>> In particular, unlike DXO, what is being measured appears to be the
>>> behaviour of a particular lens/camera combination.


>>You meant to say "exactly like DxO, however, unlike DxO it'll
>>be your camera and your lens you test".


> No. Unlike DXO.


Really?

> DXO will test a lens on a number of different cameras They will also
> test a camera on a number of different lenses.


If I test 2 cameras I borrowed and 3 lenses I rented in each
combination, I'm doing it exactly like DxO (cause I'm not
testing my own stuff and most of the tests don't help me for
a lack of having that gear), but if I test my own camera(s)
to my own lens(es), it's different from DxO?

> By a suitable choice of
> combinations they are able to characterise to a sufficient degree of
> accuracy the properties of the cameras and the properties of the
> lenses. (Look up the design of experiments if you want to know more.)


Eric, you're out of your mind if you believe you can push
providing proof for your claims on me. You're twice out
of your mind if you think the *design* of experiments tells
anything about a *need* for the contortions someone will take
with their design. Superstition is a prime example where one
designs something (e.g. a behavior, a building (Feng Shui),
and so on) and may genuinely believe the contortions are indeed
necessary, when they in fact are not at all.


> With this information they are able assess the performance of
> combinations of lenses and cameras which they have not actually
> tested.


Or so they claim. To a limited degree they may be even right
--- and by the same degree anyone with a bit of knowledge in
the right fields can do that.

> This class of research is way above the level of photographing
> a brick wall.


Eric, if my photo of a brick wall shows strong distortions, I
don't care if DxO labs says about that lens-camera combination
that there's no distortion.

What DxO does is akin to publishing a generic profile for a
printer and paper (tested on a single copy, which may or may
not be close to the average copy), whereas I can measure my
*specific* printer, with my *specific* batch of ink and my
*specific* batch of paper.

Now, for those who won't or can't create a profile on their
own a generic profile is better than nothing ...



>>> so you guys with
>>> (say) a D800, how do you feel about relying on a lens profile created
>>> with (say) a Canon Rebel?


>>The guys with a D800 will of course use lens profiles generated
>>with *their* D800 and *their* lenses.


> Agreed, but they can only do that to within the limits of the Adobe
> test procedures.


Because there's only Adobe or the road?

> For example, I don't think Adobe can embed
> corrections for colour innaccuracy in their profiles. DXO most
> certainly does.


Please show the cite where DxO does that.

And (assuming they do that at all) that helps you only if you
shoot studio and log the runtimes of your lamps and flashes
and have a very stable power source for them --- no, wall
sockets won't do.
In which case you already *have* a proper colour target or
five and can apply colour correction to your specific lighting
situation, your different cameras and lenses and --- of course
--- your RAW converter's settings.

In any other setting the light source and environmental
reflections are too unstable --- unless you maybe shoot film
and digital at the same time.


>>People who don't care
>>so much --- i.e. not D800 owners --- will download one created
>>with a compatible type (e.g. Nikon 1.5x crop) and the same lens
>>(different copy, though).


>>If you had the slightest idea of the whole process --- i.e.
>>spend less than 30 minutes reading --- you'd know that your
>>scenario would only happen when someone took the same lens,
>>mounted it on a camera where it didn't fit natively, extended
>>the Rebel's sensor to the larger crop of Nikon and then measured
>>the result. Which would be quite a feat.


> Especially when you try to take into account the effect of the
> differences in the sensor grid between a Rebel and a D800.


"A person would have to lift a multi-gigaton sky scraper and
carry it through the bog to archive that." -- "Especially when
they'd have to wear flip-flops."

-Wolfgang

Wolfgang Weisselberg 01-20-2013 06:02 PM

Re: This is not an advertisement ...
 
Eric Stevens <eric.stevens@sum.co.nz> wrote:
> On Thu, 10 Jan 2013 18:30:31 +0100, Wolfgang Weisselberg
> <ozcvgtt02@sneakemail.com> wrote:


>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>> On Thu, 27 Dec 2012 19:59:08 +0100, Wolfgang Weisselberg
>>> <ozcvgtt02@sneakemail.com> wrote:

>>
>>>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>>>> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>>>>>>On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stevens@sum.co.nz> said:

>>
>>>>>>> "DxO Lens Softness
>>>>>>> About lens softness
>>>>[...]
>>>>>>> Matching the shooting parameters contained in the EXIF data
>>>>>>> (aperture, focal length, etc.) and the map provided by the
>>>>>>> DxO Optics Module, DxO Optics Pro can apply precisely-tailored
>>>>>>> corrections according to the position of each pixel in the
>>>>>>> image field. This correction is not the same over the whole image,
>>>>>>> since lenses are sharper in the center; thus central pixels
>>>>>>> need less correction."

>>
>>>>>>While that is an interesting take on DXO's proprietary solution to a
>>>>>>"lens softness" issue, and their self critique and praise of their
>>>>>>system. As some have also claimed, such "lens softness" can also be
>>>>>>attributed to a need for said lens to be calibrated with the camera on
>>>>>>which it is to be used.

>>
>>>>> 1. DXO _does_ calibrate a lens and camerea combinations.

>>
>>>>But not your lens to your camera. If all cameras and lenses were
>>>>identical, we'd not need microfocus settings.

>>
>>> --- snip ---

>>
>>> DXO doesn't fix people who are determined to argue, no matter what.

>>
>>>>Best start with having at least the centre always and consistenly
>>>>sharp before worrying about the corners.

>>
>>> http://news.cnet.com/8301-11386_3-57...ns-latest-slr/
>>> is an interesting read about DXO.

>>
>>Please note carefully where they do NOT say that they test a
>>couple copies of each camera and each lens. Much less a dozen
>>or more copies. Then head over to the lensrentals blog and
>>read how far 'identical' lenses and cameras can differ from
>>each other when it comes to the region of improvements where
>>DxO claims to work with. (And that's not even including bad
>>lenses and cameras.)


> It's inconceivable that an organization such as DXO would test any
> lens without making certain that it was working properly.


"[inconceivable] You keep using that word. I do not think it
means what you think it means."

Even if DxO would not test a clearly bad lens, they have *no*
way of knowing if it's at the low end of OK or at the high end.
I say it again: lensrentals.com blog. Go there. Get educated.

> Nor would
> any lens or camera manufacturer supply gear for test without
> confirming that it was working properly.


And there is no way a lens can be damaged during transport,
of course. Again, head over to the lensrentals.com blog.

> Had you not been in such a
> hurry to have a wack at me you might have found articles like
> http://news.cnet.com/8301-30685_3-20008399-264.html which, among
> other things, says:


> "The tests show results for a single copy of each lens, with serial
> number provided, but DxO factors in variability across multiple
> copies, Touchard said. Specifically, he said the score numbers are
> rounded to not show a misleading level precision. Lower-end lenses
> show more variability, he added, and typically only in resolution
> but not chromatic aberration, vignetting, and distortion."


> Clearly DXO are not relying on tests based on just the one example of
> anything.


Clearly they test *a* *single* *lens* (with serial number
provided!) and (simply) "factors in variability": they round
score numbers, more for lower-end lenses' resolutions.

It says so right in the snippet you quoted. Is reading a
lost art nowadays?

I'll not even ask (because it doesn't even factor in here) if
you believe everything you read and never found out journalists,
even journalists focussing on the topic they write about,
make mistakes and mishear stuff. 50% of the journalists which
wrote about press releases from a company where I used to work
didn't even get the name right, for goodness sake!, even when
it was right in front of them --- in writing! In general I
found journalists know their stuff unless *I* have a modicum
of knowledge about the topic or event they write about.
But again, that's not even a factor here ...


>>As I said: if the cameras and lenses really were identical,
>>there'd be no need for microfocus settings and photographers
>>wouldn't need to match lenses to cameras and wouldn't at the
>>high end send them in together to make them matched pairs.


> Buy a Nikon. You can do it yourself.


I'm afraid Nikons over here don't come with optical workshops
and tools and gear (worth several $100,000) and instructions how
to change shims and adjust single element positions and tilts
and so on to make a lens really shine with a single camera.

OTOH the prices are below several $100,000 for a Nkon over here.


>>This should be *blindingly* *obvious* even to rabid DxO fans.
>>
>>So you tell me why a random lens copy with a random camera
>>copy, each having random variations from the ideal model is
>>enough to tell how *your* copies with *their own*, *different*
>>random variations behaves?


> You would hhave to be feeble minded to think that's what it does. At
> best, it tells you how your gear ought to behave.


Well, since you bring it up: you seem to believe DxO manages
*exactly* that. Though I'd call you misled, not feeble minded.

-Wolfgang

Wolfgang Weisselberg 01-22-2013 10:12 PM

Re: This is not an advertisement ...
 
Eric Stevens <eric.stevens@sum.co.nz> wrote:
> On Sun, 20 Jan 2013 18:41:15 +0100, Wolfgang Weisselberg
>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>> On Thu, 10 Jan 2013 18:54:12 +0100, Wolfgang Weisselberg
>>>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>>>> On Fri, 28 Dec 2012 11:35:58 +1300, Eric Stevens
>>>>>>On Thu, 27 Dec 2012 19:59:08 +0100, Wolfgang Weisselberg
>>>>>>>Eric Stevens <eric.stevens@sum.co.nz> wrote:
>>>>>>>> On Wed, 26 Dec 2012 15:16:34 -0800, Savageduck
>>>>>>>>>On 2012-12-26 14:22:39 -0800, Eric Stevens <eric.stevens@sum.co.nz> said:


>>>>>>>>>> "DxO Lens Softness
>>>>>>>>>> About lens softness
>>>>>>>[...]
>>>>>>>>>> Matching the shooting parameters contained in the EXIF data
>>>>>>>>>> (aperture, focal length, etc.) and the map provided by the
>>>>>>>>>> DxO Optics Module, DxO Optics Pro can apply precisely-tailored
>>>>>>>>>> corrections according to the position of each pixel in the
>>>>>>>>>> image field. This correction is not the same over the whole image,
>>>>>>>>>> since lenses are sharper in the center; thus central pixels
>>>>>>>>>> need less correction."


>>>>> I've also discovered
>>>>> http://www.computer-darkroom.com/blo...tion-profiles/ which
>>>>> deals with the creation of Adobe lens profiles. It sounds similar to
>>>>> to the DXO method but not identical.


>>>>| This application characterises three common types of lens
>>>>| aberrations, namely geometric distortion, lateral chromatic
>>>>| aberration, and vignetting.


>>>>Reading THAT should have told you that there's not a single
>>>>bit of "lens softness correction" in there.


>>> It did.


>>So you should have known that it's similar only in the way
>>that it applies mathematical operations on digital images.


> DxO appears to share only some mathematical operations with Adobe.


I doubt they "share". They use the same external
knowledge base: Scientific papers and the like from third
parties. (And they likely try to improve upon them in
secret, mostly by finding better values to plug into the
algorithms and make them fast, I guess.)

>>>>A bit of
>>>>research would have told you that it's also an old hat.


>>>>> The wording of the article
>>>>> suggests that profiles may not all be created by Adobe.


>>>>| If you?re into making your own lens profiles the Lens Profile
>>>>| Creator application can be download from Adobe Labs.


>>>>It plain says it. No suggestions here.


>>> Have you considered the possibility that some of the independently
>>> created profiles may be fed back to Adobe?


>>Sure.


> So, as I said, Adobe "profiles may not all be created by Adobe."


"History may not all be in the present."

>>> I know of nothing to
>>> indicate where individual profiles supplied by Adobe may have come
>>> from.


>>| suggests that profiles may not all be created by Adobe.
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>Your point was?


> Unlike DxO, there is not a common methodological base underlying all
> Adobe profiles.


And you know that because at DxO there's only one single guy
creating their profiles? Or are you saying you can only follow
identical instructions and use identical software if you're
employed by DxO or Adobe?


>>>>> The procedure only "characterises three common types of lens
>>>>> aberrations, namely geometric distortion, lateral chromatic

>> ^^^^^^^^^^
>>>>> aberration, and vignetting". There is no mention of the measurement of
>>>>> what DXO calls lens 'softness'. Nor is there any mention of color.

>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


>>>>Look up "chroma" and "chrominance". Contrast to "luminance"

>> ^^^^^^^^^^^^^^^^


>>> None of which are mentioned in connection with the Adobe profile
>>> procedure or data.


>>I've underlined the parts you should reread.


> With what end in mind?


There was a slim chance you'd recognize that your claim is
true only if you meant the literal word "color", spelled
exactly like that.


>>I've also noticed that you seem to have misplaced your
>>| It sounds similar to
>>| to the DXO method but not identical.
>>claim.


>>>>> In particular, unlike DXO, what is being measured appears to be the
>>>>> behaviour of a particular lens/camera combination.


>>>>You meant to say "exactly like DxO, however, unlike DxO it'll
>>>>be your camera and your lens you test".


>>> No. Unlike DXO.


>>Really?


>>> DXO will test a lens on a number of different cameras They will also
>>> test a camera on a number of different lenses.


>>If I test 2 cameras I borrowed and 3 lenses I rented in each
>>combination, I'm doing it exactly like DxO (cause I'm not
>>testing my own stuff and most of the tests don't help me for
>>a lack of having that gear), but if I test my own camera(s)
>>to my own lens(es), it's different from DxO?


No answer?

>>> By a suitable choice of
>>> combinations they are able to characterise to a sufficient degree of
>>> accuracy the properties of the cameras and the properties of the
>>> lenses. (Look up the design of experiments if you want to know more.)


>>Eric, you're out of your mind if you believe you can push
>>providing proof for your claims on me. You're twice out
>>of your mind if you think the *design* of experiments tells
>>anything about a *need* for the contortions someone will take
>>with their design. Superstition is a prime example where one
>>designs something (e.g. a behavior, a building (Feng Shui),
>>and so on) and may genuinely believe the contortions are indeed
>>necessary, when they in fact are not at all.


> A technician you may be, a scientist you are not.


"He who makes a claim has to provide proof" is not valid in
science?

The existence of superstition is not accepted in science?
http://www.jstor.org/discover/10.230...21101565427673
http://www.ingentaconnect.com/conten...20003/art00001
http://www.northlandprep.org/proctor...20Behavior.pdf

The existence of unneccessary contortions (see above) believed
to be necessary by the performer are not accepted in science?

> Have you *never*
> encountered the theory behind the design of experiments? It sounds as
> though you haven't the faintest idea of what I am talking about.


I've not the faintest idea why you think that the design of
experiments has anything to do with making a vague ("sufficient
degree" --- for what?) claim you want *me* to provide proof for.

If that's the scientific method, I'll read entrails, that's
more reliable.


>>> With this information they are able assess the performance of
>>> combinations of lenses and cameras which they have not actually
>>> tested.


>>Or so they claim. To a limited degree they may be even right
>>--- and by the same degree anyone with a bit of knowledge in
>>the right fields can do that.


> Only if they have the data to an acceptable degree of accuracy.


Since the 'acceptable degree' is rather low and the resolution
of cameras is nowadays very high ...


>>> This class of research is way above the level of photographing
>>> a brick wall.


>>Eric, if my photo of a brick wall shows strong distortions, I
>>don't care if DxO labs says about that lens-camera combination
>>that there's no distortion.


> Then you need DxO software to rectify the distortion, don't you?


Epic fail. Facepalm.

Slow down, stop knee jerking, start thinking.

If DxO says: "no distortion", all they can correct is "no
distortion", not "strong distortions", since they don't have
the data in first place they need for rectifying.

On the other hand, I can just use Adobe's method (or use hugin,
for heaven's sake) and correct the distortion really observed
with my specific gear. I don't need DxO at all for that.


>>What DxO does is akin to publishing a generic profile for a
>>printer and paper (tested on a single copy, which may or may
>>not be close to the average copy), whereas I can measure my
>>*specific* printer, with my *specific* batch of ink and my
>>*specific* batch of paper.


> I have already given a URL to a site where DxO has explained how they
> use multiple lenses and camera bodies in their work.


Sure, just like many amateurs use multiple lenses and bodies:
ONE telezoom, ONE UWW-zoom, ONE AL-prime, ONE macro prime,
ONE 1.4x extender, ONE 2x extender, ..., ONE main body,
ONE smaller backup body.
Yes, they use multiple lenses and camera bodies ... NO, THEY DO
NOT USE MULTIPLE COPIES OF THE SAME LENS AND OF THE SAME BODY.
And that's what that URL of yours said.


>>Now, for those who won't or can't create a profile on their
>>own a generic profile is better than nothing ...


>>>>> so you guys with
>>>>> (say) a D800, how do you feel about relying on a lens profile created
>>>>> with (say) a Canon Rebel?


>>>>The guys with a D800 will of course use lens profiles generated
>>>>with *their* D800 and *their* lenses.


>>> Agreed, but they can only do that to within the limits of the Adobe
>>> test procedures.


>>Because there's only Adobe or the road?


Hmmm?


>>> For example, I don't think Adobe can embed
>>> corrections for colour innaccuracy in their profiles. DXO most
>>> certainly does.


>>Please show the cite where DxO does that.


> It's in their software. You can select camera makes and models and see
> the effects.


I see the same effects when I use a different calibration curve.
I see the same effects when I switch off the monitor profile.
I see the same effects when I switch to a linear colourspace.
I see the same effects when I move the RGBCMY_sliders around.

The only way to get accurate colours is to use a lightsource
that gives the same light from shot to shot, avoid all coloured
reflections in the environment (say, green leafes or blue sky),
and use a colour chart to profile your camera and lens for
the specific light situation and RAW converter (including the
inbuild JPEG engine) settings. Which also means no auto-WB,
for example.

Everything else will not get you "corrections for colour
inaccuracy" --- at best you'll get one specific light situation
that may be correct, but which you'll never have.

Wake up and smell the coffee.


>>And (assuming they do that at all) that helps you only if you
>>shoot studio and log the runtimes of your lamps and flashes
>>and have a very stable power source for them --- no, wall
>>sockets won't do.
>>In which case you already *have* a proper colour target or
>>five and can apply colour correction to your specific lighting
>>situation, your different cameras and lenses and --- of course
>>--- your RAW converter's settings.


>>In any other setting the light source and environmental
>>reflections are too unstable --- unless you maybe shoot film
>>and digital at the same time.


> You are getting more than a little ridiculous.


So where do you think I am wrong? Unless you can point that
out, you're ridiculous and in denial.


>>>>People who don't care
>>>>so much --- i.e. not D800 owners --- will download one created
>>>>with a compatible type (e.g. Nikon 1.5x crop) and the same lens
>>>>(different copy, though).


>>>>If you had the slightest idea of the whole process --- i.e.
>>>>spend less than 30 minutes reading --- you'd know that your
>>>>scenario would only happen when someone took the same lens,
>>>>mounted it on a camera where it didn't fit natively, extended
>>>>the Rebel's sensor to the larger crop of Nikon and then measured
>>>>the result. Which would be quite a feat.


>>> Especially when you try to take into account the effect of the
>>> differences in the sensor grid between a Rebel and a D800.


>>"A person would have to lift a multi-gigaton sky scraper and
>>carry it through the bog to archive that." -- "Especially when
>>they'd have to wear flip-flops."


> You don't understand the problem, do you?


You don't understand, full stop. The differences in the sensor
grid are no real problem, in fact, you've claimed DxO handles
them routinely with no problems with "sufficient accuracy".

Oh, BTW, anything Canon crop format over 14 MPix has denser
pixels than the D800.

-Wolfgang


All times are GMT. The time now is 11:22 PM.

Powered by vBulletin®. Copyright ©2000 - 2014, vBulletin Solutions, Inc.
SEO by vBSEO ©2010, Crawlability, Inc.