New Sean Reid Article

[FastFashnReloaded]

Advertisement (gone after registration)

At this stage it seems that the the M8 /lens relationship with output leaves a bit to be desired. .... straw clutching if you think you can use a lens to its optimum capacity

 

It would be helpful if you would post some data, links to data, or just something besides your general opinion.

 

Erwin seems to disagree. Testreports

 

There are some actual charts and graphs and happy scientific stuff there. Where are yours?

[wparsonsgisnet]

The idea of having the cyan drift correction in a RAW plug in might be very useful so long as one could remember what lens he used for each picture. That's the one catch.

 

Well, sure, Sean. I figured we'd dial it in till we were satisfied. The reverse problem is also there: if the different lens brands require different cyan correction, then the in-camera solution from Leica might not be as useful as the dial-it-yourself version in post-processing sw..

 

And, I mentioned previously, Leica's functionality needs to be split up: As I understand it (not YET[!!!!!] having my M8), the user can select either no lens id with no correction, or lens id with correction. I think there should be separate settings for lens id (manual would be nice), and enable correction. That would allow us to address your point that we wouldn't later remember what lens we used. (Of course we could put a spot of paint on the edge of our lenses so we could tell which one was mounted.j)

 

Regards,

[Guest stnami]

Dana it is quite obvious from the statements made on this forum, images posted that there are problems

ps I am happy to supply buckets of sand as well

[wbesz]

Well I don't think it hurts to look around and question question before buying, and I do it all the time.

 

In the case of lenses for my Leica M cameras, I place importance for image quality at the widest aperture, and the dividing line of importance starts at 35mm.

Regarding 28mm, I would use this angle for shots that have objects at various distances from the camera, so I would probably use aperture f5.6 and smaller.

Considering the range of 28mm lenses in question, my biased choice would be the little Leica Elmarit asph. Not only is this lens a fair price, but it is small,

has coding, and most of all, it has the kind of warranty that is hard to best.

 

If I am to question longer lenses such as my Summilux-M 50 asph, or Summicron-M 75 asph, then there is no contest with other makers.

 

If I had the M8, then my choice for 28mm would probably be the the Summicron-M asph..

 

Bottom line is that to question is a good thing, ..especially when looking for wider angled lenses.

It would be fair to say these could have some competition when cost is part of the equation.

 

(I admit to having a 15mm VC )

[sean_reid]

 

1) Well, sure, Sean. I figured we'd dial it in till we were satisfied. The reverse problem is also there: if the different lens brands require different cyan correction, then the in-camera solution from Leica might not be as useful as the dial-it-yourself version in post-processing sw..

 

2) And, I mentioned previously, Leica's functionality needs to be split up: As I understand it (not YET[!!!!!] having my M8), the user can select either no lens id with no correction, or lens id with correction. I think there should be separate settings for lens id (manual would be nice), and enable correction. That would allow us to address your point that we wouldn't later remember what lens we used. (Of course we could put a spot of paint on the edge of our lenses so we could tell which one was mounted.j)

 

Regards,

 

1) That could be but my instincts (and I'm looking at the new tests right now) is that the Leica lens settings are going to be quite helpful for other lenses as well. And, of course, uncoded Leica lenses would also benefit from manual lens selection.

 

2) We should know more, firsthand, about the new firmware in early Dec.

 

Cheers,

 

Sean

[scott kirkpatrick]

There's no reason why it should. The field of view doesn't change with aperture, so the angle at which light contacts the sensor doesn't change with aperture either. Aperture is therefor not a variable.

 

If you look back at Sean's earliest vignetting studies with various WA lenses on the RD-1, you may recall that vignetting of the luminance is greatest at wide aperture, and decreases at smaller aperture. Now look in the draft article at the first series of pictures, which show vignetting with the firmware corrections disabled, and consider only the G and B values in the corners. They vignette less at f/8 then they do at f/2.8. If the Leica correction in the M8 firmware actually is aperture-dependent (using tricks like comparing the TTL exposure value with a rough value from the viewfinder "blue dot" sensor), then it will use a weaker luminance vignetting correction at small apertures. It is not clear whether the M8 is really doing anything this fancy.

 

Now look at the color-sensitive vignetting, which you can separate out by taking the ratio of the R values in the corners to the average of the G and B values. The color vignetting appears to be not dependent on aperture -- it's a decrease of about 10% (with some variations due to the small amount of data in this one experiment) for the very symmetric, small "group I" lenses and only about 5% for the more telecentric, longer and faster, "group II" lenses, at both f/2.8 and f/8.

 

It's a bit of a stretch with so little data (and it's tedious to gather lots of data like this), but I am inclined to conclude that the variations between lenses in color corrections and overall vignetting characteristics are real, but they are not very large and not that hard to predict. Sean's strategy of mapping to the nearest Leica lens looks like it may work to within something like 1/3 stop of accuracy across the frame, at least for the 28mm and maybe for the 24/25mm lenses.

 

I expect finding something Leica makes to map the CV12 and CV15 into will be tougher. They are pretty small, and sit fairly close to the imager, and the 16-18-21 TE is huge by comparison.

 

scott

[scott kirkpatrick]

Advertisement (gone after registration)

So, in a nutshell, one can try to pull shadow detail from any file but a file that begins with a longer tonal scale gives one more and better data to work with from the start. I was an exhibition printer at one point (silver process) and this was always true of the film I printed as well. *If* one wants detail at both ends of the scale (and thats a big if) nothing beats a long-scale negative (film or digital).

... years of advertising and articles that have told us that more lens contrast is always better. Sometimes it is and sometimes it isn't.

 

Sean, isn't this close to the old trick of flashing film with a bit of pre-exposure to get extra sensitivity in the shadows?

 

scott

[atournas]

I'm a LeicaNut myself, but that doesn't render the following comments less real!

 

CV and Zeiss lenses may compare with Leica well at f/8 or smaller, but what I find unbeatable with Leica performance is when I use the lens wide open. No other brand--at least to the best of my knowledge--may claim similar results to those obtained with Leica lenses at their largest apertures. Now, combine a Summicron open at f/2 plus 1/1000 speed plus a fine emulsion and you start seeing results that vibrate! Or the famous f/1.4 at 1/4 combo in dim light (sometimes known as the square-root-of-2 settings).

 

Erwin Puts is absolutely right when emphasizing that we usually do not make full exploitation of Leica lenses' unique performance.

 

Paul

[Guest guy_mancuso]

That is one of the main reasons i buy Leica in both the R and the M is the wide open performance. It is very hard to match that

[sean_reid]

If you look back at Sean's earliest vignetting studies with various WA lenses on the RD-1, you may recall that vignetting of the luminance is greatest at wide aperture, and decreases at smaller aperture. Now look in the draft article at the first series of pictures, which show vignetting with the firmware corrections disabled, and consider only the G and B values in the corners. They vignette less at f/8 then they do at f/2.8. If the Leica correction in the M8 firmware actually is aperture-dependent (using tricks like comparing the TTL exposure value with a rough value from the viewfinder "blue dot" sensor), then it will use a weaker luminance vignetting correction at small apertures. It is not clear whether the M8 is really doing anything this fancy.

 

Now look at the color-sensitive vignetting, which you can separate out by taking the ratio of the R values in the corners to the average of the G and B values. The color vignetting appears to be not dependent on aperture -- it's a decrease of about 10% (with some variations due to the small amount of data in this one experiment) for the very symmetric, small "group I" lenses and only about 5% for the more telecentric, longer and faster, "group II" lenses, at both f/2.8 and f/8.

 

It's a bit of a stretch with so little data (and it's tedious to gather lots of data like this), but I am inclined to conclude that the variations between lenses in color corrections and overall vignetting characteristics are real, but they are not very large and not that hard to predict. Sean's strategy of mapping to the nearest Leica lens looks like it may work to within something like 1/3 stop of accuracy across the frame, at least for the 28mm and maybe for the 24/25mm lenses.

 

I expect finding something Leica makes to map the CV12 and CV15 into will be tougher. They are pretty small, and sit fairly close to the imager, and the 16-18-21 TE is huge by comparison.

 

scott

 

I'll have a new table of cyan drift up tonight that is better and should make it easier for us to judge these relationships. You know that you and I are already thinking along similar lines with respect to this topic.

 

Cheers,

 

Sean

[sean_reid]

That is one of the main reasons i buy Leica in both the R and the M is the wide open performance. It is very hard to match that

 

Hi Guy,

 

Take another look at those tables of croppings. <G> Leica makes what I believe is, overall, the best set of small format lenses in the world. But there are many myths, misconceptions, etc. floating around about how these various RF lenses compare and so it's good sometimes to look at actual results (including wide open) to know what's true and what's not. The proof is in the pudding and the results aren't always what people have been lead to believe.

 

To all,

 

I don't think there's any need for a brand battle. But it is useful for photographers to know that there are actually many quite impressive RF lenses available from several manufacturers. The accomplishments of Zeiss or CV, in my mind, don't take anything away from Leica's accomplishments but it makes sense to recognize the excellences of all three. For what its worth, I've been testing RF lenses fairly intensively for several years now and my comments are based on that experience.

 

I am happy to discuss how the lenses in the test performed. It's the whole "us and them" antagonistic thing (that one sees so often on the web) that I think is a complete waste of time.

 

Best,

 

Sean

[carstenw]

Sean, I mentioned this in another thread, but here goes: when you study the MTF charts closely for the Leica 28mm f/2.0 Asph and f2.8 Asph (I know, I know, real life etc.), you will notice that just outside the 16mm mark (edge of the M8 sensor), the Summicron pulls ahead of the Elmarit by a margin. I would say that for the M8, the Elmarit is a great deal, but the Summicron should receive extra consideration for film of future FF considerations.

 

The Voigtlaender lenses notwithstanding. I just bought a CV15. Impatience...

[Guest guy_mancuso]

I have not read it yet , now i need to go do that. it was more a in general comment

[sean_reid]

I have not read it yet , now i need to go do that. it was more a in general comment

 

Hi Guy,

 

When you get to it, notice how some of those lenses do wide open. You may be surprised. The Leica lenses do beautifully, of course, but they aren't the only ones that impress wide open.

 

Cheers,

 

Sean

[sean_reid]

Sean, isn't this close to the old trick of flashing film with a bit of pre-exposure to get extra sensitivity in the shadows?

 

scott

 

Hiya Scott,

 

It's analogous to an extent.

 

Cheers,

 

Sean

[AlanG]

Hi Rob,

 

I haven't but, as you know, one can usually salvage some shadow detail in post (if desired) but as one does that he or she gets closer to the noise floor.

Sean

 

There is nothing wrong with trying to capture it in a file without going through tricks. (That's why I have so much lighting gear.) But I think you and most others may be working with one hand tied behind your back if you can't get your shadow detail looking the way you want it.

 

I have found that DxO software with its "lighting effects" can do so much to hold shadow detail and improve the look of that shadow detail that contast variation between lenses will pale by comparison. DxO can make it look as if you have a fill light in the shadows with added contrast, not just the muddy grey that one usualy gets when trying to pull detail from the shadows. You could never draw a tone curve in C1 that would come close. Plus this is way beyond what we think of as tone curves. It is intelligently applied enhancement.

 

It does this enhancement at the raw conversion stage so it has the camera's entire dynamic range to work with. Additionally it applies sophisticated noise processing where it is needed in the file so that those enhanced shadows come out really smooth. It has a similar feature for preserving highlight detail. The combination of the two effects makes it easy to use the entire dynamic range of the photograph.

 

Just to give you an idea of how significant this is, let me give an example. Often when I photograph room interiors, I would do a separate darker exposure do hold detail in the bright windows and then combine it with the normally exposed and strobe lit photograph. I generally don't have to do that now that I have DxO. I can hold the highlight detail much better.

 

Another example is photographing a North facing building that never gets sunlight. The lighting effects can make the image have a look that I've never seen before when shooting in deep shade.

 

I know I sound like some kind of evangelist, but I have been doing digital photography for more than 3 years with C1 Pro. Before that I scanned my transparencies for about about 8 years. I really think that DxO is the biggest step forward I have made in all of my digital work. It is that good. Capture One is great but it can't hold detail in the highlights and simultaneously boost detail and contrast in the shadows in high contrast scenes. DxO finally has made all of the potential of my raw files easily accessible.

 

Leica users need to encourage DxO to support the M8 and you'll be all set. I don't want to take this further on this post or force it down anyone's throat, but if anyone wants to see examples, just say so and I'll post them to my website. Here's the kind of work I do: Alan Goldstein Photography - Home Page

[Jean LeBlanc]

I'll have a new table of cyan drift up tonight that is better and should make it easier for us to judge these relationships. You know that you and I are already thinking along similar lines with respect to this topic.

 

Cheers,

 

Sean

 

Do you have a link for that?

[rvaubel]

So, Sean's results show that while the cyan colouring increases with decreasing focal length, it's also dependent on lens design. Sean, a couple of questions:

 

- Am I right that working aperture doesn't change things very much? If so, this is good news because the correction will not know the working aperture.

 

- Do you have a feel for how the cyan colouring changes with IR filter make or type? Leica will presumably be optimising for use with "their" filters and I'm wondering if the results will be different with other types of filter.

 

If using a non-Leica filter, the cyan/vignetting effect may require a different lens coding than Leica provides. Another reason to code your own lenses eith a sharpie!

 

Rex

[dlc43]

I know I sound like some kind of evangelist, but I have been doing digital photography for more than 3 years with C1 Pro. Before that I scanned my transparencies for about about 8 years. I really think that DxO is the biggest step forward I have made in all of my digital work. It is that good. Capture One is great but it can't hold detail in the highlights and simultaneously boost detail and contrast in the shadows in high contrast scenes. DxO finally has made all of the potential of my raw files easily accessible.

 

Leica users need to encourage DxO to support the M8 and you'll be all set. I don't want to take this further on this post or force it down anyone's throat, but if anyone wants to see examples, just say so and I'll post them to my website. Here's the kind of work I do: Alan Goldstein Photography - Home Page

 

Well there's the problem: They do not currently support ANY Leica cameras, including the M8. Why don't they?

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[AlanG]

Well there's the problem: They do not currently support ANY Leica cameras, including the M8. Why don't they?

 

Version 4.0, the latest version of DxO, is very new. The Mac version is still not out but should be soon. I'm not sure if the lighting effects, highlight preservation, and some other features were in earlier versions. I wasn't interested in the earlier versions because while they corrected a lot of lens flaws, I didn't think the color and exposure support in the Raw converter was very good, until now.

 

Originally, DxO was mostly concerned with correcting optical problems of lenses - distortion, chromatic aberation, vignetting, etc. This is probably not needed as much with most Leica lenses. Now they have put in all kinds of exposure and color corrections, plus you can adjust the geometry of a picture in really unusual ways. Such as volume anamorphosis correction.

 

The M8 and even the digital back for the reflex camera are pretty new. So until there is more market penetration, it probably won't be worth it to DxO to test the cameras and all of the lenses and make profiles that can be applied automatically by reading the EXIF info. But if they just allow you to open the DNG file and use the lighting effects and other controls in the raw converter, it would be very useful and should be very easy for them to do. And they'd have a bigger potential market too.

 

Originally DxO made money by charging separately for each camera and lens module that you needed. But now they just charge one price for all. So maybe they are still too hung up on their original mission to see that some people would like to use DxO as a general purpose raw converter even without automatic support. (You could just dial in the various lens corrections manually if you want them.) I think it would be very easy for them to put in corrections for cyan corners too.

 

For more:

http://dxooptics.com/