Well, it might interest some...

[gfspencer]

Advertisement (gone after registration)

Thank you, Chris. I compared the M8 to the 50D. Very similar.

[tomasis7]

"It doesn't recover information, it moves the shadow information further from the noise floor. That, with digital capture, can help to preserve shadow detail and/or allow less exposure so that more highlight detail is preserved"

 

Why are the dark areas with a low-contrast lens brighter? Where does the light come from?

Even if the dark areas are further moved from the noise floor because of low-lens-contrast, it isn't real information from the dark areas!

That's at least my understanding, but I also know which effect you mean, but I think it's an illusion.

 

you need think about other effects like veiling flare which reduces "contrast" of scene so you get more usable Raw file to edit than having clipped highlights when you have shot a very contrasty scene to example. It is matter of filling between 0-255 at brightness level. The contrast thing can affect the range depending how it is exposed (under or over).

 

If you expose the same scene under one stop, so you have recover Zone1 back to 2,3 by pushing a stop or half through processing. Thus noise increases if you push and it looks "worse" for DR. But this metod has to work for preventing clipped higlights. I have experimented with underexposing 4 stops and pushing 4 stops further. There digital does nice things while films are hardly recoverable 1 stop for underexposing.

[preston]

It might help to clarify the DR discussion to distinguish between Alan's claim that a low-contrast lens *clips* the shadow tones, and Sean's claim that it *compresses* the shadow tones.

 

Clipping means cutting off shadow tones, and thereby irreversibly eliminating tone difference information. Compressing means remapping the shadow tones upwards, retaining some of the difference information; the tones can be then be (partially) restored by expansion in postprocessing.

 

For those old enough to remember audio, it's like the difference between clipping an excessively wide dynamic range signal, as opposed to compressing / decompressing the signal, a la Dolby NR on a cassette tape. This precompression / postexpansion allowed a limited dynamic range medium (cassette tape) to record and play back a wider dynamic range signal than the medium could otherwise capture.

 

The camera sensor's limited DR is a similar bottleneck, and optical tone compression could definitely be useful in contrasty scenes... if we are really talking compressing and not clipping.

[sean_reid]

They must not my friend. You should try Mandler R lenses with your Canons. They are inexpensive nowadays. Mine are the 35/2, 50/2, 90/2, 135/2.8 and 180/3.4 but the 19/2.8 and 80/1.4 are said to be good as well. Tina must have used a 19/2.8 vers. 1 or 2 if i remember well.

 

Sure, I have several CZ lenses I use on the Canons and (as you may not know) I've tested R lenses on EOS bodies. But one loses auto-aperture stop down which is very important feature. The new ZE lenses are fully compatible.

 

Cheers,

 

Sean

[sean_reid]

TBut this links in with other posts on the need (or not) to fo for the M8.2, Sean's thoughts on obsolesence...

 

For those who haven't read that article (which was written long before the stock market took a nose dive) it is about resisting the idea of obsolescence. I thought I'd clarify.

 

Cheers,

 

Sean

[mjh]

Why are the dark areas with a low-contrast lens brighter? Where does the light come from?

From other, brighter parts of the image, obviously.

 

Even if the dark areas are further moved from the noise floor because of low-lens-contrast, it isn't real information from the dark areas!

The additional light doesn’t carry any information; in fact that’s the point. If it did, it would interfere with the real information. All it does is to reduce the contrast, translating the original subject contrast into a smaller image contrast that the sensor can handle. All the tonality is preserved, only compressed to fewer EV steps.

 

This effect is quite real. There is a downside in that you cannot have a low contrast at extremely low spatial frequencies (which is sometimes beneficial) without some less desirable loss of contrast at higher spatial frequencies. For some part it is a trade-off between spatial resolution and dynamic range. But that doesn’t render it illusory.

[sean_reid]

Advertisement (gone after registration)

"It doesn't recover information, it moves the shadow information further from the noise floor. That, with digital capture, can help to preserve shadow detail and/or allow less exposure so that more highlight detail is preserved"

 

Why are the dark areas with a low-contrast lens brighter? Where does the light come from?

Even if the dark areas are further moved from the noise floor because of low-lens-contrast, it isn't real information from the dark areas!

That's at least my understanding, but I also know which effect you mean, but I think it's an illusion.

 

It's not a question of *adding* shadow information. Many people get confused on this point. What happens is the that shadow information is recorded with a higher exposure level (more light, brighter tone). That moves it away from black and away from the noise floor. The S/N ratio increases.

 

By analogy, the same people (shadow details) enter the elevator but now instead of getting on at the basement, they get on at the first or second floor. In the basement, there's loud music playing so it's hard to hear what they are saying. But at the higher floors the conversations of these same people become more intelligible. We don't add more people but they move away from the noise.

 

Now some will argue that the generalized light that mixes with the shadow information partially obscures that information. To whatever degree that is or is not true, the S/N advantage those shadow areas gain (at the sensor) from coming into the camera at a higher light level more than offsets what it might lose (based on my observations and testing for several years now).

 

Is that any clearer? The lower contrast lens is not adding shadow detail per se (which it couldn't possibly do simply via a lowering of contrast) but rather it is lifting the tones of the existing shadows so that they are recorded on the sensor at a higher S/N.

 

Have you read "Exposing to the Right" on LL? I can find the link if you haven't. It is relevant to understanding this.

 

Cheers,

 

Sean

[jaapv]

"It doesn't recover information, it moves the shadow information further from the noise floor. That, with digital capture, can help to preserve shadow detail and/or allow less exposure so that more highlight detail is preserved"

 

Why are the dark areas with a low-contrast lens brighter? Where does the light come from?

Even if the dark areas are further moved from the noise floor because of low-lens-contrast, it isn't real information from the dark areas!

That's at least my understanding, but I also know which effect you mean, but I think it's an illusion.

A thought experiment

 

Take a subject that shows zone 0 through zone 10

Take a camera-lens combination that renders the chart in exactly those zones. Assume it will have 100% noise in zone 0 and 0% noise in zone 10, thuse 18% noise in zone 5.

 

Replace the lens by a lens that shows no contrast.

The whole image will be exposed in zone 5 with 18% noise. Now take the part of the image that should be in zone 0 down to that zone 0 in postprocessing. It will have 18% noise instead of 100% noise.

 

For those puzzled... zone 0 is absolute black, zone 10 is absolute white and zone 5 is 18% middle gray.

[sandymc]

The additional light doesn’t carry any information; in fact that’s the point

 

AKA, actually, its noise - thing is, there is no free lunch; any light that the lens redistributes across the image does indeed reduce contrast, but the cost is that information is lost in the highlights.

 

Michael's spatial frequency point is important; mostly when people talk about low contrast lenses, they actually mean lower contrast at high spatial frequencies - contrast is reduced in fine details, but not if you have large features. So a big black patch will typically remain a big black patch.

 

Sandy

[sean_reid]

It might help to clarify the DR discussion to distinguish between Alan's claim that a low-contrast lens *clips* the shadow tones, and Sean's claim that it *compresses* the shadow tones.

 

Clipping means cutting off shadow tones, and thereby irreversibly eliminating tone difference information. Compressing means remapping the shadow tones upwards, retaining some of the difference information; the tones can be then be (partially) restored by expansion in postprocessing.

 

For those old enough to remember audio, it's like the difference between clipping an excessively wide dynamic range signal, as opposed to compressing / decompressing the signal, a la Dolby NR on a cassette tape. This precompression / postexpansion allowed a limited dynamic range medium (cassette tape) to record and play back a wider dynamic range signal than the medium could otherwise capture.

 

The camera sensor's limited DR is a similar bottleneck, and optical tone compression could definitely be useful in contrasty scenes... if we are really talking compressing and not clipping.

 

That's a very, very interesting analogy. I'm not an audio expert but the purpose of Dolby is, of course, noise reduction and what we're talking about here is moving shadow information away from the noise floor.

 

And yes, by definition a lower contrast lens will compress the tonal range.

 

Again, that's really an interesting analogy - one of the most interesting posts I've read on the web in a while.

 

Cheers,

 

Sean

[sean_reid]

From other, brighter parts of the image, obviously.

 

 

The additional light doesn’t carry any information; in fact that’s the point. If it did, it would interfere with the real information. All it does is to reduce the contrast, translating the original subject contrast into a smaller image contrast that the sensor can handle. All the tonality is preserved, only compressed to fewer EV steps.

 

This effect is quite real. There is a downside in that you cannot have a low contrast at extremely low spatial frequencies (which is sometimes beneficial) without some less desirable loss of contrast at higher spatial frequencies. For some part it is a trade-off between spatial resolution and dynamic range. But that doesn’t render it illusory.

 

Hello Michael!

 

I'm very glad that you dropped into the thread because I know that you'll be able to help me explain this phenomenon to those for whom it is confusing. And yes, as I mentioned earlier (and in various articles) there are some losses with lower contrast lenses but the advantages, with higher contrast subjects, are indeed quite real.

 

I first experienced this (visually) about four years ago and presented the ideas in an article on lenses for the Epson R-D1. To this day, I still periodically hear from people who insist it can't be true. And, at the same time, many photographers have seen the differences first hand.

 

To all,

 

Those who've been reading my article about this subject will know the following but for those who haven't I'll note the following. None of the this comes from theory but instead from several years of close observation of what lenses do on digital cameras.

 

1. To be sure, there are strong advantages to the flare resistance of high contrast lenses like the Zeiss ZMs. They present a "purer" image to the sensor and they're especially wonderful when shooting into the light (unless one is specifically looking for flare). They also provide certain kinds of tonal definition that can be lost with lower contrast lenses. In most cases, optical designers (Mr. Karbe included) aim for high contrast designs.

 

2. These darn filters we need to use with the M8 add reflective surfaces and the combination of lower contrast lenses, filters and shooting towards a light source can be disastrous. All else being equal, higher contrast lenses do better with these filters when shooting into the light.

 

3. If a subject is high contrast (street, noon, sunny day) I tend to prefer more moderate contrast lenses. But, even then, if I'm shooting into the light with filters, I prefer higher contrast lenses unless I'm intentionally provoking flare.

 

4. In lower contrast light, I prefer higher contrast lenses.

 

When I first put some of these ideas forward in 2004, they were controversial (and it seems they still are). It would be interesting to hear from more photographers about when and why they tend to use higher or lower contrast lenses. This, to my mind, is more interesting a topic than much of what occupies space on forums.

 

Cheers,

 

Sean

[sean_reid]

there is no free lunch

Sandy

 

Absolutely...

 

Cheers,

 

Sean

[pgk]

But, even then, if I'm shooting into the light with filters, I prefer higher contrast lenses unless I'm intentionally provoking flare. Sean

Now that is an interesting comment!

 

For a start I find few photographers who actually admit to utilising flare characteristics as part of their toolset.

 

Secondly I remember having a Zeiss Zoom on a Contax (NX I think) - a stunningly sharp, high contrast lens, with great flare resistance - except that when flare spots did occur they were very distinct, hard-edged, white circles which I found totally unacceptable in the images - I actually like the 35 Summicron v.4 on my M8 because of how it renders oof areas (bokeh) and also because of its contrast and flare characteristics.

 

I've also tried a 'rough' comparison of a pre-war 90mm f/4 Elmar and a 90/2.8 Elmarit (last version) on a very contrasty day and was actually surprised at how well the old lens performed under such conditions - exposure and raw conversion both had to be tweeked but the images that the old lens produced were very acceptable (I think one actually got used in the end as it was taken on a shoot!).

[Tina Manley]

Hi Tina:

Did you have to shave the mirror or the lens shroud? I don't have the 5D but have a few R lenses including the R19 and the Lux 35 and am considering getting the 5DII.

 

Thanks,

Alan

 

Hi, Alan - No, I have the first version of the R19 and it fits fine with a Novoflex adapter. It's only the newest 19 that won't work. The old R19 and the R100 Macro are my favorite lenses on the Canons.

 

Tina

[AlanG]

Consider this for a minute. DXO has designed and markets systems for testing cameras and lenses. DXO has tested and profiled hundreds of lenses for DXO Optics alone. And probably has tested many more for other applications. If the choice of lens used in DXOmark sensor measurements would make a significant difference in the sensor measurement results, don't you think they would know that?

[Isabelle Lenatio]

Consider this for a minute. DXO has designed and markets systems for testing cameras and lenses. DXO has tested and profiled hundreds of lenses for DXO Optics alone. And probably has tested many more for other applications. If the choice of lens used in DXOmark sensor measurements would make a significant difference in the sensor measurement results, don't you think they would know that?

 

Alan, I would think so too .... but keep the LUF audience in mind, some of them would go to the lengths of theorizing the amount, distribution and effect of unforeseen dust particles on an inner lens to explain why any leica gear might be rated lower than alternative systems.

 

Not bitching here but all this measuring, philosophizing and what if when x does y compared to a versus b on a scale from one to ten reflected on extrapolating the total effect on a 5-20 scale, then the result will be , etc, etc.... Boring, unnecessary and at some points overly childish.

 

And then Tina Manley comes in, goes into great length on the quality of lens x compared to y and why z is better .... obviously getting a lot of responses.... subsequently telling the followers to push off and go take some pictures....

 

Sorry for the outburst, don't mean any harm but try to accept that the ace of spades will never turn into a queen of hearts.

[sean_reid]

Consider this for a minute. DXO has designed and markets systems for testing cameras and lenses. DXO has tested and profiled hundreds of lenses for DXO Optics alone. And probably has tested many more for other applications. If the choice of lens used in DXOmark sensor measurements would make a significant difference in the sensor measurement results, don't you think they would know that?

 

I can't speak for others here but my focus is more on the broader topic of lens contrast rather than DXO per se. Who knows what they've considered? I wouldn't assume either way. If they're using high contrast lenses on all the cameras, the point may be moot with respect to their DR tests. But lenses are a potential variable in that kind of testing.

 

That said, I'm not sure we're talking about DXO anymore. It think it ended up being a jumping point for a broader topic.

 

Cheers,

 

Sean

[sean_reid]

Boring, unnecessary and at some points overly childish.

 

I guess we all have our own perspectives. I find parts of this thread more interesting than much of what is argued on the forum. But if this thread bores and annoys you, why not skip it?

 

I'm not a scientist but without science we don't have these wonderful cameras and lenses that are needed for our work. So science has its place in the discussion of photography. That said, no one is being forced to read this thread. Some may find that parts of it help them think through lens choices and use, exposure, etc.

 

Lastly, I haven't seen much argument about the M8's performance in those tests at all. Some of us are just working some related ideas through.

 

Cheers,

 

Sean

[sean_reid]

Now that is an interesting comment!

 

For a start I find few photographers who actually admit to utilising flare characteristics as part of their toolset.

 

Secondly I remember having a Zeiss Zoom on a Contax (NX I think) - a stunningly sharp, high contrast lens, with great flare resistance - except that when flare spots did occur they were very distinct, hard-edged, white circles which I found totally unacceptable in the images - I actually like the 35 Summicron v.4 on my M8 because of how it renders oof areas (bokeh) and also because of its contrast and flare characteristics.

 

I've also tried a 'rough' comparison of a pre-war 90mm f/4 Elmar and a 90/2.8 Elmarit (last version) on a very contrasty day and was actually surprised at how well the old lens performed under such conditions - exposure and raw conversion both had to be tweeked but the images that the old lens produced were very acceptable (I think one actually got used in the end as it was taken on a shoot!).

 

Flare can be lovely, more later. There's a picture I want to track down for you.

 

Cheers,

 

Sean

[Advertisement]

Hallo,

Look here: Leica M

[Isabelle Lenatio]

I guess we all have our own perspectives. I find parts of this thread more interesting than much of what is argued on the forum. But if this thread bores and annoys you, why not skip it?

 

I'm not a scientist but without science we don't have these wonderful cameras and lenses that are needed for our work. So science has its place in the discussion of photography. That said, no one is being forced to read this thread. Some may find that parts of it help them think through lens choices and use, exposure, etc.

 

Lastly, I haven't seen much argument about the M8's performance in those tests at all. Some of us are just working some related ideas through.

 

Cheers,

 

Sean

 

I agree, fully agree actually .... the M8 is good, I enjoy it, I enjoy using it and I like the images it produces. There are better alternatives this is why overall the M8 is rated lower than most modern day cameras, I can live with that it's just the fact of turning the bottom stone upward to try and argue the opposite and this with arguments which might appeal to wall and newspaper photographing enthusiasts, but which actually have very little to do with actual photography, as in taking, pictures. And these specs are not the absolute reason for these discussions are they now..... take the human psyche into account, as you as a clever man will well know, and you will find these arguments tending to "defend" as opposed to "add". This is actually what I dislike about this whole discussion. It's not only this thread but any criticism on any Leica part or piece will at one point amount to lengthy and boring posts such as this one will undoubtedly turn into.

Why not just admit that the M8 is not as good as x, y or z, because it isn't but admit to the fact that you enjoy using it and that is the sole reason for spending/wasting your money .... I can easily admit that.... but all of this is probably a "man" thing!

 

don't get pee'ed off with me Sean I'm sure you know what I mean even if I am not able to communicate it well enough. ;-)