How wide are these?

Started May 12, 2012 | Discussions
Hen3ry
Senior MemberPosts: 8,608Gear list
Like?
An everlasting life full of nothing questions about angles of view?
In reply to Adventsam, May 14, 2012

That could be a definition of hell.

Adventsam wrote:

John 6:47 Verily, verily, I say unto you, He that believeth on me hath everlasting life

Enjoy, Adventsam -- hope you can take your GH2 with you.

Cheers, geoff
--
Geoffrey Heard
http://pngtimetraveller.blogspot.com/2011/10/return-to-karai-komana_31.html

 Hen3ry's gear list:Hen3ry's gear list
Panasonic Lumix DMC-G6 Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Lumix G 14mm F2.5 ASPH Panasonic Lumix G Vario 100-300mm F4-5.6 OIS Olympus M.Zuiko Digital 45mm 1:1.8 +5 more
Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: Not as wide as these ;)
In reply to papillon_65, May 14, 2012

papillon_65 wrote:

I don't need to rationalise anything,

So why do you still do it?

you raised it remember,

Raised what?

when are you going talk from experience and not cling on to pointless 100% crops of charts?

I am always talking from experience, just a wider and more intersubjective one than yours.

That's a sarcastic way of saying you're no expert at all, I guess you missed it lol.

No I didn't miss it. Taking sarcasms literally is an effective way of disarming them.

No, I take photographs, photography for me has nothing to do with "testing optics", it's about taking photographs and a lens is just a tool to make that happen.

Reasonably serious photographers look at tests and/or perform them themselves to know they buy the right stuff and to know the strong and weak side of their own equipment. Apparently, you don't belong to that group.

It doesn't have to have class leading "edge to edge" performance or impress anyone, that's the point you're clearly missing.

Can't force you to raise your standards. If you're happy to use inferior stuff, feel free to do so.

The facts are that you are clearly all about charts and MTF's and nothing about creativity and real photography. Even Ken Rockwell gets it, you should read this, it pretty much sums up everything about the current state of play and why people like you are so clueless. Love him or hate him he's nailed your ilk totally in this piece.

http://www.kenrockwell.com/tech/lens-sharpness.htm

So Ken Rockwell is your kind of guy? Can't say I am surprised.

Speaking of lens tests, these are some pics I caught last week for the purpose of testing my 100-300. What do you think?

 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
Steen Bay
Veteran MemberPosts: 6,188
Like?
Re: Not as wide as these ;)
In reply to Adventsam, May 14, 2012

Adventsam wrote:

papillon_65 wrote:

Adventsam wrote:

Shot at F2.8 and ISO 1600, yes the bottom corners are soft, it's at F2.8 (That's F1.8 in terms of dof on m4/3's and outside the dof for the focus point so it's normal ). You see, in the real world where I shoot it's not important, the shot is sharp where it should be and not so sharp where I would expect it to be.

2.8 is what in m43? 2.8 is 4.2! FF therefore 2.1, @ these angles it meaningless to compare dof!

Everything else being equal if you shot with m4/3's you be shooting at F1.8 for the same dof and hence your corners would undoubtedly be flaky, what's your point caller?

Nex is 1.5, m43 2 crop! you cant compare m43 to aps-c, you need to base to FF!

Sure we can compare m43 and APS-C. The crop factor between m43 and Nex is 1.33 (2/1.5 = 1.33), so f/2.8 on Nex is DoF/diffraction equivalent to f/2.1 on m43 (2.8/1.33 = 2.1).

Reply   Reply with quote   Complain
papillon_65
Forum ProPosts: 17,062Gear list
Like?
Re: Not as wide as these ;)
In reply to Anders W, May 14, 2012

Anders W wrote:

papillon_65 wrote:

I don't need to rationalise anything,

So why do you still do it?

you raised it remember,

Raised what?

when are you going talk from experience and not cling on to pointless 100% crops of charts?

I am always talking from experience, just a wider and more intersubjective one than yours.

Clearly not if your opining about a lens you've never used ;).

That's a sarcastic way of saying you're no expert at all, I guess you missed it lol.

No I didn't miss it. Taking sarcasms literally is an effective way of disarming them.

Well that didn't work either :).

No, I take photographs, photography for me has nothing to do with "testing optics", it's about taking photographs and a lens is just a tool to make that happen.

Reasonably serious photographers look at tests and/or perform them themselves to know they buy the right stuff and to know the strong and weak side of their own equipment. Apparently, you don't belong to that group.

Well taking photographs is testing your equipment, I don't need a brick wall or a chart for that because they don't represent real world usage, unless you regularly photograph brick walls or charts of course, and I'm guessing you probably do.

It doesn't have to have class leading "edge to edge" performance or impress anyone, that's the point you're clearly missing.

Can't force you to raise your standards. If you're happy to use inferior stuff, feel free to do so.

As Ken says, all modern lenses are superior to what the masters of old used so I'm in good company.

The facts are that you are clearly all about charts and MTF's and nothing about creativity and real photography. Even Ken Rockwell gets it, you should read this, it pretty much sums up everything about the current state of play and why people like you are so clueless. Love him or hate him he's nailed your ilk totally in this piece.

http://www.kenrockwell.com/tech/lens-sharpness.htm

So Ken Rockwell is your kind of guy? Can't say I am surprised.

Well he clearly knows more about photography than you do, but then so does photoperzon.

Speaking of lens tests, these are some pics I caught last week for the purpose of testing my 100-300. What do you think?

I think you need a tripod, you've succeeded in making a good lens look worse than it actually is, which is kind of ironic considering how much store you put in pointless lens tests lol.
--
It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/

 papillon_65's gear list:papillon_65's gear list
Sigma DP1 Merrill Sigma DP2 Merrill Fujifilm XF1 Fujifilm X-T1 Fujifilm XF 18-55mm F2.8-4 R LM OIS +2 more
Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: Not as wide as these ;)
In reply to papillon_65, May 14, 2012

papillon_65 wrote:

(sigh) Sony APS-C has approx 1 stop more dof than FF and m4/3's has approx 2 stops for a given aperture (all else being equal) this equates to approx 1 stop difference between APS-C and m4/3's. In the shot I posted you would be at F1.8 shooting with m4/3's if you wanted to replicate the dof of the Sony sensor in that shot (all else being equal). Therefore it is safe to conclude that an F2.8 lens on an APS-C sensor is not going to be sharp edge to edge in much the same way as you would not expect an F1.8 lens on m4/3's to be sharp edge to edge, all else being the same, could I make it any clearer than that?

Although you have expressed your lack of interest in anything technical, you might still want to keep in mind that 2.8 on Sony APS-C corresponds to 2.1, not 1.8, on MFT. Do you want me to explain how to get there?

 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
papillon_65
Forum ProPosts: 17,062Gear list
Like?
Re: Not as wide as these ;)
In reply to Anders W, May 14, 2012

Anders W wrote:

papillon_65 wrote:

(sigh) Sony APS-C has approx 1 stop more dof than FF and m4/3's has approx 2 stops for a given aperture (all else being equal) this equates to approx 1 stop difference between APS-C and m4/3's. In the shot I posted you would be at F1.8 shooting with m4/3's if you wanted to replicate the dof of the Sony sensor in that shot (all else being equal). Therefore it is safe to conclude that an F2.8 lens on an APS-C sensor is not going to be sharp edge to edge in much the same way as you would not expect an F1.8 lens on m4/3's to be sharp edge to edge, all else being the same, could I make it any clearer than that?

Although you have expressed your lack of interest in anything technical, you might still want to keep in mind that 2.8 on Sony APS-C corresponds to 2.1, not 1.8, on MFT. Do you want me to explain how to get there?

I'd prefer it if you could show me the difference visually so I can understand just how much that 0.3 is actually relevant in the real world and how it would impact the photograph in visual terms.
--
It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/

 papillon_65's gear list:papillon_65's gear list
Sigma DP1 Merrill Sigma DP2 Merrill Fujifilm XF1 Fujifilm X-T1 Fujifilm XF 18-55mm F2.8-4 R LM OIS +2 more
Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: Not as wide as these ;)
In reply to papillon_65, May 14, 2012

papillon_65 wrote:

Clearly not if your opining about a lens you've never used ;).

I have some basis for my judgment. You obviously have none.

Well that didn't work either :).

What makes you think I care about your opinion? What works or not is obviously not for you to decide.

Well taking photographs is testing your equipment, I don't need a brick wall or a chart for that because they don't represent real world usage, unless you regularly photograph brick walls or charts of course, and I'm guessing you probably do.

Yes, I have understood that you don't think you need very good equipment.

As Ken says, all modern lenses are superior to what the masters of old used so I'm in good company.

Right. I am sure Ken shoots with lenses that matches the quality standards of the 30s too, just like the 16/2.8.

Well he clearly knows more about photography than you do, but then so does photoperzon.

Ha ha. Why I on earth do you think I would care about your opinion about my photos? I asked merely to reveal you for what you are. That's now clear to everyone.

Speaking of lens tests, these are some pics I caught last week for the purpose of testing my 100-300. What do you think?

I think you need a tripod, you've succeeded in making a good lens look worse than it actually is, which is kind of ironic considering how much store you put in pointless lens tests lol.

Ha ha. Interesting that you are ready to pass judgment on these images in relation to the performance of the lens without knowing more than I told you. The amount of cropping and thus the relation to 100 percent view for example. Or the extent to which they were sharpened.

 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
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: Not as wide as these ;)
In reply to papillon_65, May 14, 2012

papillon_65 wrote:

Anders W wrote:

papillon_65 wrote:

(sigh) Sony APS-C has approx 1 stop more dof than FF and m4/3's has approx 2 stops for a given aperture (all else being equal) this equates to approx 1 stop difference between APS-C and m4/3's. In the shot I posted you would be at F1.8 shooting with m4/3's if you wanted to replicate the dof of the Sony sensor in that shot (all else being equal). Therefore it is safe to conclude that an F2.8 lens on an APS-C sensor is not going to be sharp edge to edge in much the same way as you would not expect an F1.8 lens on m4/3's to be sharp edge to edge, all else being the same, could I make it any clearer than that?

Although you have expressed your lack of interest in anything technical, you might still want to keep in mind that 2.8 on Sony APS-C corresponds to 2.1, not 1.8, on MFT. Do you want me to explain how to get there?

I'd prefer it if you could show me the difference visually so I can understand just how much that 0.3 is actually relevant in the real world and how it would impact the photograph in visual terms.

Sorry. Can't help you with everything. You've got to make at least a minimum effort yourself.

 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
papillon_65
Forum ProPosts: 17,062Gear list
Like?
Re: Not as wide as these ;)
In reply to Anders W, May 14, 2012

Anders W wrote:

papillon_65 wrote:

Anders W wrote:

papillon_65 wrote:

(sigh) Sony APS-C has approx 1 stop more dof than FF and m4/3's has approx 2 stops for a given aperture (all else being equal) this equates to approx 1 stop difference between APS-C and m4/3's. In the shot I posted you would be at F1.8 shooting with m4/3's if you wanted to replicate the dof of the Sony sensor in that shot (all else being equal). Therefore it is safe to conclude that an F2.8 lens on an APS-C sensor is not going to be sharp edge to edge in much the same way as you would not expect an F1.8 lens on m4/3's to be sharp edge to edge, all else being the same, could I make it any clearer than that?

Although you have expressed your lack of interest in anything technical, you might still want to keep in mind that 2.8 on Sony APS-C corresponds to 2.1, not 1.8, on MFT. Do you want me to explain how to get there?

I'd prefer it if you could show me the difference visually so I can understand just how much that 0.3 is actually relevant in the real world and how it would impact the photograph in visual terms.

Sorry. Can't help you with everything. You've got to make at least a minimum effort yourself.

Well you've helped me to understand that even though you agonise over test charts about edge sharpness you shoot birds where edge sharpness is totally irrelevant and you do it badly so that your shots are soft and blurred.

So in the interests of harmony I would suggest that you leave the charts alone (as it clearly doesn't work for you) and you use a tripod ;).
And with that sage advice I bid you a good night :).

-- hide signature --

It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/

 papillon_65's gear list:papillon_65's gear list
Sigma DP1 Merrill Sigma DP2 Merrill Fujifilm XF1 Fujifilm X-T1 Fujifilm XF 18-55mm F2.8-4 R LM OIS +2 more
Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: Not as wide as these ;)
In reply to papillon_65, May 15, 2012

papillon_65 wrote:

Anders W wrote:

papillon_65 wrote:

Anders W wrote:

papillon_65 wrote:

(sigh) Sony APS-C has approx 1 stop more dof than FF and m4/3's has approx 2 stops for a given aperture (all else being equal) this equates to approx 1 stop difference between APS-C and m4/3's. In the shot I posted you would be at F1.8 shooting with m4/3's if you wanted to replicate the dof of the Sony sensor in that shot (all else being equal). Therefore it is safe to conclude that an F2.8 lens on an APS-C sensor is not going to be sharp edge to edge in much the same way as you would not expect an F1.8 lens on m4/3's to be sharp edge to edge, all else being the same, could I make it any clearer than that?

Although you have expressed your lack of interest in anything technical, you might still want to keep in mind that 2.8 on Sony APS-C corresponds to 2.1, not 1.8, on MFT. Do you want me to explain how to get there?

I'd prefer it if you could show me the difference visually so I can understand just how much that 0.3 is actually relevant in the real world and how it would impact the photograph in visual terms.

Sorry. Can't help you with everything. You've got to make at least a minimum effort yourself.

Well you've helped me to understand that even though you agonise over test charts about edge sharpness you shoot birds where edge sharpness is totally irrelevant and you do it badly so that your shots are soft and blurred.

Ha ha. Caught with your pants down again. You really don't have a clue about optical performance characteristics, do you? So you think the tests we have been talking about measure edge sharpness (aka acutance) rather than resolution? First have a look here

http://www.cambridgeincolour.com/tutorials/sharpness.htm

and then go check up how MTF-50 relate to the two.

But I thought you didn't give a damn about either since it doesn't matter for all the interesting real-life subjects you shoot. Why this sudden concern for things like softness and blurriness?

So in the interests of harmony I would suggest that you leave the charts alone (as it clearly doesn't work for you) and you use a tripod ;).

Mmm. Just look at how well it can be done with a real master behind the camera. That sparrow is clearly superior to my own lame attempts at "birding", isn't it?

Source: Previously posted here on DPR in
http://forums.dpreview.com/forums/read.asp?forum=1041&message=39698307

And with that sage advice I bid you a good night :).

Same to you. Why not use the 16/2.8 as a pillow. Soft corners you know.

 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
Detail Man
Forum ProPosts: 14,950
Like?
DMC-G1 + LGV 100-300mm Analysis, Diagnosis, and Recommendation
In reply to Anders W, May 15, 2012

Anders W wrote:

Speaking of lens tests, these are some pics I caught last week for the purpose of testing my 100-300. What do you think?

Cute little critter ! LGV 100-300mm Mega OIS, FL=300mm. F=5.6, T=1/320, ISO=100, EV=11.64.

Here are some mere suggestions from Doctor Detailia's Institute of Digital Image Signal Processing. Forensic analysis by my fiercely loyal team of idiot savants reveals your ACR 7.1 settings used were:
.

Detail Palette Control Settings :

Sharpen Detail= 100 - [Relying exclusively upon decon-deblur is a prescription for unsightly artifacts]

Sharpen Radius=+1.0 - [A rather wide Radius to use when relying exclusively upon decon-deblur]

Sharpen Edge Masking= 0 - [Using a zero sharpening threshold value is asking for noise artifacts]

Sharpness=50 - [Thank heaven for the ability to reduce output-weighting of the above settings]

Luminance Smoothing= 25 - [A rather strong amount of Lumi NR is required to hide said artifacts]
Luminance Noise Reduction Contrast=0
Luminance Noise Reduction Detail=50

Color Noise Reduction= 25 - [A rather strong amount of Chroma NR required to hide said artifacts]
Color Noise Reduction Detail=50
.

JPEG conversion parameters :

Compression stats ... [Large amounts of Quantization-compression required to hide said artifacts]
Compression Ratio: 17.07:1
Bits per pixel: 1.41:1

YCbCr color-space Luminance channel quantization data-compression division coefficients:

DQT, Row #0: 3 2 2 3 4 6 8 10
DQT, Row #1: 2 2 2 3 4 9 10 9
DQT, Row #2: 2 2 3 4 6 9 11 9
DQT, Row #3: 2 3 4 5 8 14 13 10
DQT, Row #4: 3 4 6 9 11 17 16 12
DQT, Row #5: 4 6 9 10 13 17 18 15
DQT, Row #6: 8 10 12 14 16 19 19 16
DQT, Row #7: 12 15 15 16 18 16 16 16
Approx quality factor = 91.86

YCbCr color-space Chrominance channels quantization data-compression division coefficients:

DQT, Row #0: 3 3 4 8 16 16 16 16
DQT, Row #1: 3 3 4 11 16 16 16 16
DQT, Row #2: 4 4 9 16 16 16 16 16
DQT, Row #3: 8 11 16 16 16 16 16 16
DQT, Row #4: 16 16 16 16 16 16 16 16
DQT, Row #5: 16 16 16 16 16 16 16 16
DQT, Row #6: 16 16 16 16 16 16 16 16
DQT, Row #7: 16 16 16 16 16 16 16 16
Approx quality factor = 91.90

Chroma subsampling: 1x1 - [No Chroma Sub-Sampling used ... which would have been a better data-compression alternative to the use of heavy amounts of quantization data-compression - were it not for the necessity to low-pass spatial-frequency filter image-data to attenuate said artifacts] .
.

Diagnosis : The potential "sharpness" (from the standpoint of the high spatial-frequency response of the composite modulation-transfer-function) is being unecessarily compromised by the relatively heavy amounts of Luminance NR as well as JPEG Quantization Compression that were used in an effort to attenuate the significant sharpening artifacts that Adobe's deconvolution-deblurring is known to inject into image-data (especially when the "Detail" control-slider is set to 100, which causes the Adobe Sharpening tool to use deconvolution-deblurring exclusively, and to not use any USM-related edge-enhancement techniques whatsoever in generating it's composite output-data).
.

Recommendation :

DxO Optics Pro 7.23, which implements fully automatic (and optional manual) optical corrections (including DxO's proprietary and truly excellent "Lens Softness" corrections, described at this link):

http://www.dxo.com/us/photo/dxo_optics_pro/features/optics_geometry_corrections/lens_softness

... fully supports the Panasonic Lumix G Vario 100-300mm F/4.0-5.6 on the DMC-G1 camera body. Following de-mosaicing and optical corrections, DxO can export 16-bit TIFFs to any editor of choice

A fully functional 31-day trial version (which I am using with good success) can be downloaded at:

http://www.dxo.com/us/photo/free_trial_version

Full disclosure of financial relationships: I have paid $99 USD (for a DxO Optics Pro 6.x license), and will soon be paying an additional $49 (for a DxO Optics Pro 7.x upgrade license) to DxO Labs Inc. This information is being provided out of (fanboy) enthusiam, without any compensation involved.

Full disclosure of psychological relationships: No particular emotional need exists on the part of my ego to "correct" any "errors" in the present ways of the "Adobe Faithful". No value judgments ...
.

Public Notice regarding "Acutance" vs "Resolution" :

Ha ha. Caught with your pants down again ...

Let it be here known that Dr Detailia has no intention of "dropping his pants" anytime soon ...

... the subjective sharpness of a print corresponds to the area under the MTF curve between the spatial frequencies of (0.5 x magnification) and (2 x magnification) when spatial frequency is plotted on a logarithmic scale .

... if we make an 8x10 (or 8x12) print from a 35mm negative we have to magnify the negative by a factor of 8 (since the negative is approximately 1" x 1.5"). So as far as the SQF is concerned, the area under the MTF curve of the lens between 4 and 16 cycles/mm is then what really counts .

http://www.bobatkins.com/photography/technical/mtf/mtf4.html

See also: http://www.imatest.com/docs/sqf/#csf

"The Pinhole Camera Revisited or The Revenge of the Simple-Minded Engineer" is quite interesting:

http://www.biox.kth.se/kjellinternet/Pinhole.pdf

Check out the "Pinhole Camera MTF (modulation transfer function)" graph, and check out the simulated images that follow in the referenced paper.

DM ...

Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: DMC-G1 + LGV 100-300mm Analysis, Diagnosis, and Recommendation
In reply to Detail Man, May 15, 2012

Detail Man wrote:

Anders W wrote:

Speaking of lens tests, these are some pics I caught last week for the purpose of testing my 100-300. What do you think?

Cute little critter ! LGV 100-300mm Mega OIS, FL=300mm. F=5.6, T=1/320, ISO=100, EV=11.64.

Here are some mere suggestions from Doctor Detailia's Institute of Digital Image Signal Processing. Forensic analysis by my fiercely loyal team of idiot savants reveals your ACR 7.1 settings used were:

As I said, these are test pictures and I am using them (among other things) to experiment a bit with my newly purchased LR 4. I've finally taken the plunge after five or so years with Silkypix 3.x. Having a lot of fun so far.

Luminance Smoothing=25 - [A rather strong amount of Lumi NR is required to hide said artifacts]

Pants half down now. That's just a small step up from default.

Color Noise Reduction=25 - [A rather strong amount of Chroma NR required to hide

And fully down. That is the default.

Diagnosis : The potential "sharpness" (from the standpoint of the high spatial-frequency response of the composite modulation-transfer-function) is being unecessarily compromised by the relatively heavy amounts of Luminance NR as well as JPEG Quantization Compression that were used in an effort to attenuate the significant sharpening artifacts that Adobe's deconvolution-deblurring is known to inject into image-data (especially when the "Detail" control-slider is set to 100, which causes the Adobe Sharpening tool to use deconvolution-deblurring exclusively, and to not use any USM-related edge-enhancement techniques whatsoever in generating it's composite output-data).

Well, I adjusted it a bit. Better now?

Recommendation :

DxO Optics Pro 7.23

Thanks but no thanks. One change of RAW converter for every five years or so suffices.

 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
Detail Man
Forum ProPosts: 14,950
Like?
Re: DMC-G1 + LGV 100-300mm Analysis, Diagnosis, and Recommendation
In reply to Anders W, May 15, 2012

Anders W wrote:

Detail Man wrote:

Anders W wrote:

Speaking of lens tests, these are some pics I caught last week for the purpose of testing my 100-300. What do you think?

Cute little critter ! LGV 100-300mm Mega OIS, FL=300mm. F=5.6, T=1/320, ISO=100, EV=11.64.

Here are some mere suggestions from Doctor Detailia's Institute of Digital Image Signal Processing. Forensic analysis by my fiercely loyal team of idiot savants reveals your ACR 7.1 settings used were:

As I said, these are test pictures and I am using them (among other things) to experiment a bit with my newly purchased LR 4. I've finally taken the plunge after five or so years with Silkypix 3.x. Having a lot of fun so far.

Nearly anything would be "up" from Silkypix SE 3.x, it can only get better from there !

Have you checked out the excellent RAW Therapee 4.x ? (IMO) a bit weak in the NR department, but the best UI around , and lots of good stuff. Ever-updated, and not by $-grubbers

Luminance Smoothing=25 - [A rather strong amount of Lumi NR is required to hide said artifacts]

Pants half down now. That's just a small step up from default.

My, aren't we "frisky" in the mornings !

That's interesting. LR 3.x's default Luminance NR setting is Zero. Wonder what they are hiding ?

Color Noise Reduction=25 - [A rather strong amount of Chroma NR required to hide

And fully down. That is the default.

The default in my LR 3.6 is 25 itself. My thought was that at ISO=100 with your G1, and with a fairly robust Radiometric Exposure, that you might have been able to reduce the Color NR a tad and live to tell the tale (as it doesn't itself affect fine-detail very much, or act as a LP spatial-frequency filter to cover Sharpening artifacts). It is the Luminance NR that is squashing fine-detail to cover the Sharpening artifacts. Take a close look at these lovely artifacts generated by LR 4.x:

And that is with the "Detail" control-slider at settings that are less than the maximum of 100 ...

Diagnosis : The potential "sharpness" (from the standpoint of the high spatial-frequency response of the composite modulation-transfer-function) is being unecessarily compromised by the relatively heavy amounts of Luminance NR as well as JPEG Quantization Compression that were used in an effort to attenuate the significant sharpening artifacts that Adobe's deconvolution-deblurring is known to inject into image-data (especially when the "Detail" control-slider is set to 100, which causes the Adobe Sharpening tool to use deconvolution-deblurring exclusively, and to not use any USM-related edge-enhancement techniques whatsoever in generating it's composite output-data).

Well, I adjusted it a bit. Better now?

Oh, yes, that looks so much better now. That Lightroom 4.x must be some kind of magical beast !

Thanks but no thanks. One change of RAW converter for every five years or so suffices.

Yes - why one might want these clearly inferior DxO Lens Softness corrected results is beyond me:

DM ...

Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,928Gear list
Like?
Re: DMC-G1 + LGV 100-300mm Analysis, Diagnosis, and Recommendation
In reply to Detail Man, May 15, 2012

Detail Man wrote:

Nearly anything would be "up" from Silkypix SE 3.x, it can only get better from there !

That's for sure. Nothing seriously wrong with Silkypix, but that version is getting really old by now (although it was considered pretty advanced when I first started using it in 2007).

Have you checked out the excellent RAW Therapee 4.x ? (IMO) a bit weak in the NR department, but the best UI around , and lots of good stuff. Ever-updated, and not by $-grubbers

No, haven't really given it serious consideration. Too much of an "enthusiast" program for me I am afraid.

Color Noise Reduction=25 - [A rather strong amount of Chroma NR required to hide

And fully down. That is the default.

The default in my LR 3.6 is 25 itself.

Yes, same in LR 4 and that's what I used. So calling it "a rather strong amount of Chroma NR" is ... ahem ... a slight exaggeration, wouldn't you say?

My thought was that at ISO=100 with your G1, and with a fairly robust Radiometric Exposure, that you might have been able to reduce the Color NR a tad and live to tell the tale (as it doesn't itself affect fine-detail very much, or act as a LP spatial-frequency filter to cover Sharpening artifacts).

Well, what I noticed already when I played with a trial version of LR 3.x earlier is that the chroma NR slider should usually just be left where it is. It doesn't do much harm but is very efficient at 25. Increasing it further doesn't seem to yield any tangible benefits in most cases.

It is the Luminance NR that is squashing fine-detail to cover the Sharpening artifacts.

Yes, I think you are right about that. In line with your advice, I reduced "Detail" a bit (although I don't understand how could you suggest something like that; seems contrary to your nature ;)) and used a bit of masking. Didn't need any Luminance NR after that.

Take a close look at these lovely artifacts generated by LR 4.x:

Yes, that looks horrible.

And that is with the "Detail" control-slider at settings that are less than the maximum of 100 ...

Oh, yes, that looks so much better now.

Hope you are serious about that. Well, at least I am happier with the result.

That Lightroom 4.x must be some kind of magical beast !

In some ways the program is really good. The highlight and shadow sliders can work miracles when used in conjunction with the other four light controls (exposure, contrast, whites, blacks). I find it very easy in most cases to come close to the tone curve I like (which was quite difficult in Silkypix). The NR is also very good when you need it. And the tools available for selective editing without leaving the RAW converter are great too (although I haven't really had the time to play with those yet).

Yes - why one might want these clearly inferior DxO Lens Softness corrected results is beyond me:

OK. Those clearly look better than the previous version. But what are you really trying to tell me here. I am willing to hear you out. But please be brief and to the point. I'll let you know when/if I am ready for a second chunk of your wisdom.

 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
Detail Man
Forum ProPosts: 14,950
Like?
Re: DMC-G1 + LGV 100-300mm Analysis, Diagnosis, and Recommendation
In reply to Anders W, May 15, 2012

Anders W wrote:

Detail Man wrote:

Nearly anything would be "up" from Silkypix SE 3.x, it can only get better from there !

That's for sure. Nothing seriously wrong with Silkypix, but that version is getting really old by now (although it was considered pretty advanced when I first started using it in 2007).

Color Noise Reduction=25 - [A rather strong amount of Chroma NR required to hide ...

And fully down. That is the default.

The default in my LR 3.6 is 25 itself.

Yes, same in LR 4 and that's what I used. So calling it "a rather strong amount of Chroma NR" is ... ahem ... a slight exaggeration, wouldn't you say?

Not in the case of your ultra low image-noise G1 at ISO=100 with fairly "robust" exposure, no ...

My thought was that at ISO=100 with your G1, and with a fairly robust Radiometric Exposure, that you might have been able to reduce the Color NR a tad and live to tell the tale (as it doesn't itself affect fine-detail very much, or act as a LP spatial-frequency filter to cover Sharpening artifacts).

Well, what I noticed already when I played with a trial version of LR 3.x earlier is that the chroma NR slider should usually just be left where it is. It doesn't do much harm but is very efficient at 25. Increasing it further doesn't seem to yield any tangible benefits in most cases.

In LR 3.x, past around 50, some color-desaturation begins (depending somewhat on the image itself). I begin at that point to use some Vibrancy to offset that color de-saturation (accordingly).

It is the Luminance NR that is squashing fine-detail to cover the Sharpening artifacts.

Yes, I think you are right about that. In line with your advice, I reduced "Detail" a bit (although I don't understand how could you suggest something like that; seems contrary to your nature ;)) and used a bit of masking. Didn't need any Luminance NR after that.

Sharpening artifacts are not my idea of "image-details" ...

Truth be told, I consider Adobe's implementation of (what I have read may be single-pass only) deconvolution-deblurring (similar to that reputed to be used in PS) to be (IMO) a horrendous abomination which I have been "dissing" and "decrying" and "bemoaning" ever since the release of LR 3.x nearly 2 years ago.

Take a close look at these lovely artifacts generated by LR 4.x:

Yes, that looks horrible.

I checked the meta-data for that GH2 RW2 at ISO=160 processing using LR 4.x, and it used exactly the same Sharpening settings as yours - except the "Detail" control was set to 80 (not at 100, as in your case). As far as the NR controls, the Color=25, and the Luminance=50. JPEG encoding parameters show no Chroma Sub-Sampling (as in your case), and an incredibly low Quality Factor of 77.39% (YCbCr Luminance channel) and 87.15% (YCbCr Chrominance channels).

Thus, it seems clear that neither a Luminance NR setting of 50, or very hard-core Quantization data-compression in the JPEG encoding process were able to sufficiently reverse the sharpening artifacts that were generated at a "Detail" setting of (merely) 80 . (IMO), dissappointing, and just a bit "scary", indeed ...

And that is with the "Detail" control-slider at settings that are less than the maximum of 100 ...

Oh, yes, that looks so much better now.

Hope you are serious about that. Well, at least I am happier with the result.

Well, since you did not post any images showing your revisions, it's hard for me to see it ...

That Lightroom 4.x must be some kind of magical beast !

In some ways the program is really good ...

The highlight recovery seems impressive, indeed (and a lot more "potent" than LR3.x's similar control). However, I have noted that nothing comes "free", and heavy application of the LR 4.x highlight-recovery (at least in the case of Sam's solidly clipped in all RGBG RAW-channels GH2 RW2 image-files) results in a massive peaking in the histogram of highlight tones that have been "schmushed" downwards into the mid-tones ...

My OS being XP Pro, I am not able to partake (directly) in all the dizzying LR 4.x fun, however ...

Yes - why one might want these clearly inferior DxO Lens Softness corrected results is beyond me:

OK. Those clearly look better than the previous version. But what are you really trying to tell me here. I am willing to hear you out. But please be brief and to the point. I'll let you know when/if I am ready for a second chunk of your wisdom.

Already stated. That I find that DxO's Lens Softness (RAW-level) optical corrections (for a supported camera body and lens combination) appears to my eyes to be vastly superior to anything that my LR 3.6 can do in the "sharpening department" (or, it seems, much of anything that various other people using LR 4.x could do, either). It matters not to me what you may like .

Here is a recent GH2 RW2 processed using DxO Optics Pro 7.23 (which now supports the LGV 14-45mm, 100-300mm, and 7-14mm lenses on the GH2 as well as on your G1, and which I am rather liking more than I had anticipated, relative to my previously used DxO Optics Pro 6.60):

Mono-pod stabilized shot, OIS Mode 2, Spot Auto Focus, focus-distance approximately 18 Meters. FL=18mm, F=4.0, ISO=320, T=1/20, UniWB. Clear skies, with indirect evening light, EV=6.64. 12.258 Mpixel crop. DxO 7.23 using fully automatic rectilinear, chromatic aberration, vignetting, and (moderate levels of) Lens Softness corrections, 16-bit Lanczos resizing, very mild USM, loss-less JPG:

1600x1200 version :

.

The full sized processed image, exported from DxO without resizing. Mild USM applied, converted to JPG (2x1 Chroma Sub-sampling, 100% Quality Factor). Note the absence of sharpening artifacts:

4043x3032 version :

Download the original at: http://www.dpreview.com/galleries/4464732135/download/1959273

DM ...

Reply   Reply with quote   Complain
Detail Man
Forum ProPosts: 14,950
Like?
Re: DMC-G1 + LGV 100-300mm Analysis, Diagnosis, and Recommendation
In reply to Detail Man, May 15, 2012

Detail Man wrote:

Anders W wrote:

Oh, yes, that looks so much better now.

Hope you are serious about that. Well, at least I am happier with the result.

Well, since you did not post any images showing your revisions, it's hard for me to see it ...

Hmmm, it occured to me that since your post links outside of DPR to a web-link, that you are able to modify your displayed images. That's a nice advantage compared to having a post linked for all time to a specifically numbered DPR Gallery image upload (which cannot be changed, only deleted)

I had stored your originally linked "TreeSparrow01.jpg", and saved (what I thought might be) the newer version ? However, I cannot see any difference between them. Further, the byte-size of the JPG linked-to has not changed by one single byte - which seems highly improbable (if not impossible) if you had managed to modify the linked-to JPG image via a tweak of the LR 4.x Luminance NR control-slider. So, I am a bit confused. Please advise me as to what you meant, then

Reply   Reply with quote   Complain
papillon_65
Forum ProPosts: 17,062Gear list
Like?
Anders...relax....
In reply to Anders W, May 15, 2012

It's photography not Particle Physics
--
It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/

 papillon_65's gear list:papillon_65's gear list
Sigma DP1 Merrill Sigma DP2 Merrill Fujifilm XF1 Fujifilm X-T1 Fujifilm XF 18-55mm F2.8-4 R LM OIS +2 more
Reply   Reply with quote   Complain
Steen Bay
Veteran MemberPosts: 6,188
Like?
Re: Not as wide as these ;)
In reply to Anders W, May 15, 2012

Anders W wrote:

Mmm. Just look at how well it can be done with a real master behind the camera.

A bit OT, sorry, but speaking about real masters.. have you ever seen the results that Kenn Threed achieves with his SX30 'superzoom' (JPEGs, handheld at 840mm, f/32 equivalent)?

http://forums.dpreview.com/forums/read.asp?forum=1010&message=39664875

Reply   Reply with quote   Complain
papillon_65
Forum ProPosts: 17,062Gear list
Like?
Re: Not as wide as these ;)
In reply to Steen Bay, May 15, 2012

Steen Bay wrote:

Anders W wrote:

Mmm. Just look at how well it can be done with a real master behind the camera.

A bit OT, sorry, but speaking about real masters.. have you ever seen the results that Kenn Threed achieves with his SX30 'superzoom' (JPEGs, handheld at 840mm, f/32 equivalent)?

http://forums.dpreview.com/forums/read.asp?forum=1010&message=39664875

These are great examples of how I see it, there aren't really any bad choices in terms of gear now, any enthusiast level gear is capable of excellent results, lenses or bodies. I don't choose charts or mtf's, I chose life :).
--
It's a known fact that where there's tea there's hope.
Tony
http://the-random-photographer.blogspot.com/

 papillon_65's gear list:papillon_65's gear list
Sigma DP1 Merrill Sigma DP2 Merrill Fujifilm XF1 Fujifilm X-T1 Fujifilm XF 18-55mm F2.8-4 R LM OIS +2 more
Reply   Reply with quote   Complain
tt321
Senior MemberPosts: 3,126Gear list
Like?
Re: Not as wide as these ;)
In reply to Anders W, May 15, 2012

Anders W wrote:

I am sure it can turn out decent although not really good shots once stopped down to f/5.6 or so. But at that point I wouldn't be interested. Personally, I shoot primes for speed. Otherwise I use a good zoom. Others shoot primes for optimal optical quality, no matter what the aperture. The 16/2.8 can't deliver in either respect. Hopeless lens, just like the new 30/3.5 macro.

In the case of the 16mm, it's the only way of getting that wide and wider on the NEX system apart from using non-AF lenses. So in that regard you don't necessarily shoot that particular prime for speed, you probably more likely shoot it for the AoV and the ability to go wider and even fisheye with adapters.

Also, some of the "corner softness" could be due to decisions made on not sticking to a flat-field design in order to achieve compactness (pure speculation on my part - I have no data to back it up). With a non-flat field lens, if you are not often dealing with flat targets shot perpendicularly, you could always tune your focus point as you like and the corners could be less soft if properly focused for. Most simplistic tests published online tend not to expose these issues. Probably it would be too time consuming to investigate in every case. The mechanistic algorithmic test flow of setting camera at a fixed distance towards a fixed target and rattling off a number of shots to be analysed in software would probably take half an hour. Exploring corners properly and finding what the lens does at infinity would cost too much labour...

Then this process could feed back into the lens designing and making process and we could see future items optimized for these popular tests only

WA lenses are not traditional portrait lenses and should be a jack of all trades, i.e. able to do well from closest focus to infinity, esp. wide open capable of being sharp at somewhere in the frame (people who do document copying usually can stop down a little) and stopped down sharp everywhere. I have yet to see a test specifically targeting WA lenses covering all these issues.

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads