Since there are lots of posts relating to the "Zeiss look" or "Leica look" ect. I got to thinking about this. What are the generalizations of style when considering a lens from several various quality brands. For example what really differentiates Canon from Zeiss, from Leica, from any of the other Alt options?

To put it simply, On a picture taken with a Zeiss lens, items are more differentiated (by colour, texture, volume, spatial position) from other items on the same picture. This differentiation is what give Zeiss its reputation in its ability to produce a "3D" effect, and in its reproduction of colours.

Color Rendering, macro and mIcro contrast, bokeh rendering, distortion properties and resolving power to name a few.

There are various differences, but the biggest one to me is color rendering. Most companies try to keep some sort of consistency with the way they apply their coatings to get the same kinds of color results for all of their lenses. Color is probably the difference that is most consistent among a particular line of lenses.

I haven't used Minolta glass myself, but I've read that Minolta was particularly dedicated to giving all of their lenses the same look.

Other companies, like Canon, have different sets of looks for different lenses. The F4L zooms look pretty different from the fast L primes. Their 2.8L zooms seem closer to the primes in terms of colors and all that. At least that's how it seems to me.

I think it's usually a mistake to group all lens designs from a company together. Designing a lens is a series of compromises and each designer will choose a different balance. Different lenses from a particular company are not all designed by the same person. There was a "Mandler Era" for Leica lenses, but you could hardly say that modern Leica glass has the same look as those.

for me it is hard to explain. You just have to see the pictures. every lens has a certain "look". I found that Pentax, Leica and Zeiss often give me a color and sharpness that I can obtain with my Canon lenses but I must use some extra Photoshop steps. plus it is just fun to play with different lenses.

hope this is helpful,

steve

Modern Leica lenses generally are sharp from wide open, usually right into the corners. They have very low CA, often practically non-existent, and great, but relatively strongly blurred boke. They do not generally have 3D in the way that Zeiss lenses do. Leica lenses are often very fast (f/1, f/1,4, f/2). They exhibit great subject isolation, wide open. They have very good micro-contrast, and many Leica lenses >50mm are APO designated. Leica has many patents on aspherical element manufacture, which helps give them their edge in this field. The colours are often very beautiful, but not necessarily always totally accurate. Even on digital there is such a thing as Leica colours.

I am a Leica fan, so take everything with a grain of salt, although I have tried not to exaggerate.

Leica colors:
[60/2.8 macro-Elmarit on Canon 5D]

Ken Rockwell just posted an M9 review-ish thing where he has some side by side examples of M9 shots and Velvia shots. It's interesting to see how strong of an effect on the color Velvia has, far more than the lenses have. Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots.

kidtexas wrote:
Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots.

I honestly doubt it. Kodak and Fuji's colour pallettes are the product of decades of experience and they are very delicately balanced. I have seen very few people who are talented enough in colour management to just move a few sliders and get something equally vibrant from a digital file without completely destroying it.

I can't get the great Kodak SLR/c colors out of my Canon 5DII no matter what I try (RAW, converters, settings)...

Spyro P. wrote:
kidtexas wrote:
Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots.

I honestly doubt it. Kodak and Fuji's colour pallettes are the product of decades of experience and they are very delicately balanced. I have seen very few people who are talented enough in colour management to just move a few sliders and get something equally vibrant from a digital file without completely destroying it.

pdmphoto wrote:
I can't get the great Kodak SLR/c colors out of my Canon 5DII no matter what I try (RAW, converters, settings)...

Spyro P. wrote:
kidtexas wrote:
Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots.

I honestly doubt it. Kodak and Fuji's colour pallettes are the product of decades of experience and they are very delicately balanced. I have seen very few people who are talented enough in colour management to just move a few sliders and get something equally vibrant from a digital file without completely destroying it.


Same here with my A900 colors. No matter what I do with my previous canon files.

carstenw wrote:
Modern Leica lenses generally are sharp from wide open, usually right into the corners. They have very low CA, often practically non-existent, and great, but relatively strongly blurred boke. They do not generally have 3D in the way that Zeiss lenses do. Leica lenses are often very fast (f/1, f/1,4, f/2). They exhibit great subject isolation, wide open. They have very good micro-contrast, and many Leica lenses >50mm are APO designated. Leica has many patents on aspherical element manufacture, which helps give them their edge in this field. The colours are often very beautiful, but not necessarily always totally accurate. Even on digital there is such a thing as Leica colours.

I am a Leica fan, so take everything with a grain of salt, although I have tried not to exaggerate.

I am not a big Leica lens fan, because many Leica lenses seem to lack 3D, but I agree 100% with Carsten's assessment. In particular, Leica lens colours tend to be very rich, balanced, and beautiful, and the bokeh is typically soft (too soft for my taste) and almost apochromatic (i.e. no LoCA). They are superb lenses in most measurable respects, I just don't find them very convincing at making photorealistic pictures because of the 3D issue.

There are other aspects to brand-styles beyond optical considerations, like build-quality, size/weight and ergonomics. For instance, AF lenses tend to have rattly focus rings which are not easily reached and which don't have an appropriate rate of rotation for accurate manual focus. EOS lenses don't have an aperture ring, which is a major ergonomic blunder, as changing aperture with a small dial on the camera itself is less intuitive because the aperture is in the lens, and the functions of the right hand are overloaded.

pdmphoto wrote:
I can't get the great Kodak SLR/c colors out of my Canon 5DII no matter what I try (RAW, converters, settings)...

Spyro P. wrote:
kidtexas wrote:
Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots.

I honestly doubt it. Kodak and Fuji's colour pallettes are the product of decades of experience and they are very delicately balanced. I have seen very few people who are talented enough in colour management to just move a few sliders and get something equally vibrant from a digital file without completely destroying it.



With skill you could match Velvia, but it would be a labour of love. I agree that the 5D2 has a problematic colour palette, with an excess of red and green and an absence of blue. It is very hard to get natural looking colours out of a 5D2 in many circumstances, even if you shoot with Leica lenses.

brainiac wrote:
I am not a big Leica lens fan, because many Leica lenses seem to lack 3D, but I agree 100% with Carsten's assessment. In particular, Leica lens colours tend to be very rich, balanced, and beautiful, and the bokeh is typically soft (too soft for my taste) and almost apochromatic (i.e. no LoCA). They are superb lenses in most measurable respects, I just don't find them very convincing at making photorealistic pictures because of the 3D issue.

I can't argue with this. I have seen 3D from Leica lenses, but with nowhere near the regularity that I have seen it from Zeiss lenses, and a few other specific lenses from various manufacturers. More common with Leica is extreme subject isolation, at the cost of 3D.

Anyway, I am beginning to consider owning a more varied lens stable (I currently have only Leica lenses in 35mm sizes, although I have some Zeiss lenses for Hasselblad V and Contax 645 cameras), and I am wondering if the Zeiss lenses for the Leica M bayonet (ZM) also have this 3D look. Does anyone know?

Spyro P. wrote:
I honestly doubt it. Kodak and Fuji's colour pallettes are the product of decades of experience and they are very delicately balanced. I have seen very few people who are talented enough in colour management to just move a few sliders and get something equally vibrant from a digital file without completely destroying it.


You're right. It was completely off.

I'd post the comparisons myself, but they aren't my photos, so I won't. They were pretty close. Remember, I said pretty close. I'm not sure I could do it if I didn't have the Velvia shot in front of me to work from, but with a comparison shot, it took about 1 min in LAB to match; it was pretty easy.

Eight shots down this page, the redwoods. I tweaked the L curve a tiny bit in the shadows and in the highlights to give them a bit more contrast, while retaining the midtone contrast. I also added a hair of large radius unsharp masking to the L channel. Then, steepened the a and b curves a bit and moved the centers of each curve a bit until the color of the picture was a closer match.

Heck, you can get REALLY close with no work at all. Just use the Match Color command. But, download the two tree images (the small ones are fine), convert the digital one to LAB, and run this this curve on it. Like I said, not exact, but it gets a lot of the warmth back in the colors of the tree and sky. The sharpening step helped some the contrast, but the curve gives you the idea on the color. My tree ended up a hair yellower and the contrast is still a bit different.

I do hope you guys try. It'd be interesting to see if you think I completely destroyed the image or got a reasonable match. It's funny how you all think it's either impossible or at least very hard to do.

kidtexas wrote:
Spyro P. wrote:
I honestly doubt it. Kodak and Fuji's colour pallettes are the product of decades of experience and they are very delicately balanced. I have seen very few people who are talented enough in colour management to just move a few sliders and get something equally vibrant from a digital file without completely destroying it.


You're right. It was completely off.

I'd post the comparisons myself, but they aren't my photos, so I won't. They were pretty close. Remember, I said pretty close. I'm not sure I could do it if I didn't have the Velvia shot in front of me to work from, but with a comparison shot, it took about 1 min in LAB to match; it was pretty easy.

Eight shots down this page, the redwoods. I tweaked the L curve a tiny bit in the shadows and in the highlights to give them a bit more contrast, while retaining the midtone contrast. I also added a hair of large radius unsharp masking to the L channel. Then, steepened the a and b curves a bit and moved the centers of each curve a bit until the color of the picture was a closer match.

Heck, you can get REALLY close with no work at all. Just use the Match Color command. But, download the two tree images (the small ones are fine), convert the digital one to LAB, and run this this curve on it. Like I said, not exact, but it gets a lot of the warmth back in the colors of the tree and sky. The sharpening step helped some the contrast, but the curve gives you the idea on the color. My tree ended up a hair yellower and the contrast is still a bit different.

I do hope you guys try. It'd be interesting to see if you think I completely destroyed the image or got a reasonable match. It's funny how you all think it's either impossible or at least very hard to do.

I went through all this myself about two years ago but in my case, I played around with matching the look of an Olympus lens to a Zeiss lens I owned at the time. You probably could not come up with two lens brands with dramatically different rendering. What I found was that if I shot the same scene with both lenses, side by side, I could pretty much make the Olympus look like the Zeiss with a lot of work. The way most Zeiss lenses render red in particular is pretty unique. Problem was that if I then saved that particular recipe, it never worked with different scenes of different colors in different light. I would bet trying to match film in different situations would be the same. I know all the canned film looks in various plug-ins are decent but really not that close to how film reacts in different situations.

My experience using Match Color has not been the same as yours at all. I find it fairly useless.

My point wasn't about matching subtle differences between lenses. My point was that some choices, such as filtering, film choice, film or digital, post processing, etc., often overpower differences in color characteristics between lenses. I also did state that I was matching color between the two images, and not conjuring up a Velvia simulation on any given picture. I also clearly stated I wasn't going for an exact match, just 'in the neighborhood'. Regardless, I got called out for it.

Match Color seems to work reasonably well when you have to shots of the same subject, as in the example I was using. It of course won't work in all cases. It's a pretty simple tool that needs to be used in certain cases, though you modify its behavior some using selections.

Did you try the curves file I saved? Try it. Save the two images I referred to, convert the digital one to LAB, and run the curves file on it. Try Match Color while your at it.

kidtexas wrote:
My point wasn't about matching subtle differences between lenses. My point was that some choices, such as filtering, film choice, film or digital, post processing, etc., often overpower differences in color characteristics between lenses.


Did you try the curves file I saved? Try it. Save the two images I referred to, convert the digital one to LAB, and run the curves file on it. Try Match Color while your at it.

On the first point, I would disagree in the case of the example I gave of Zeiss and Olympus. The difference was not subtle by any means.

I will play with your curves tweaks

Certainly the differences in lens colors were more subtle than a straight out of camera jpg and Velvia?

Maybe not though. Though I did say 'often overpower'

kidtexas wrote:

Did you try the curves file I saved? Try it. Save the two images I referred to, convert the digital one to LAB, and run the curves file on it. Try Match Color while your at it.

Ok, I tried your curves adjustment on the digital version of the redwood tree in LAB mode. It looks nothing like the Velvia Redwood on my high end NEC monitor as the redwood tree is now suffering from an extreme case of Jaundice! Match color will not get you there either. I guess it might take more than 1 minute of tweaking, eh?

Alright you win. The tree was a bit more yellow than the velvia version as I said previously, but much closer than the digital version over here. I was looking at overall balance of the skies colors, the greens, and the tree's colors, and how they matched up with the Velvia shot. Personally, I think the Velvia looks pretty unreal and overly orange, but that's the way it is. With 30 seconds of tweaking. If you really wanted to remove some of the yellow, I'm sure you could, it might take a bit longer. Adjust it on your super fancy monitor. I threw up a quick curve on my laptop to get it in the neighborhood, if it doesn't work, maybe try some of your own adjustments

You guys just love to nitpick. My original statement was:
"Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots. "

IN THE NEIGHBORHOOD. Not an exact simulation. Later.

kidtexas wrote:
Alright you win. The tree was a bit more yellow than the velvia version as I said previously, but much closer than the digital version over here. I was looking at overall balance of the skies colors, the greens, and the tree's colors, and how they matched up with the Velvia shot. Personally, I think the Velvia looks pretty unreal and overly orange, but that's the way it is. With 30 seconds of tweaking. If you really wanted to remove some of the yellow, I'm sure you could, it might take a bit longer. Adjust it on your super fancy monitor. I threw up a quick curve on my laptop to get it in the neighborhood, if it doesn't work, maybe try some of your own adjustments

You guys just love to nitpick. My original statement was:
"Of course, you can tweak digital pictures colors pretty quickly to get in the neighborhood of the Velvia shots. "

IN THE NEIGHBORHOOD. Not an exact simulation. Later.

I guess we have different criteria for what "In the Neighborhood" means. Yes, yours are ON THE SAME PLANET! The issue between the two trees goes beyond just the yellow color. If you look at the texture of the bark, you will notice there is more detail/microcontrast in the Velvia example. Anyway, color accuracy is my business so "on the same planet" might be fine for some but...

Nitpicking or, as some would refer to it, the subtleties of photography, is what this forum is all about.

kidtexas wrote:
I threw up a quick curve on my laptop to get it in the neighborhood...

...and, there it is!

I'm sure if I was on my other monitor the results migt be more acceptible. As far as microcontrast and the such goes, I'm sure you read the part of my original post where I said I also applied some sharpening in the L channel, specifically while watching the tree. The velvia shot also had a warming filter if I recall.

More importantly, who cares? In the abscence of a direct velvia comparison, the digital shot as put up looks boring. I think most people would have adjusted the color and saturation on that shot until they liked the results. If it doesn't match the color 100% of xxx film or reality, so? Which gets back to the original point I made in passing: choice in film, sensor, lighting, filters, and post processing CANmake much more of a difference in color than lens choice. And most photographers I'm aware of tweak the color in post processing nowadays one way or another.

kidtexas wrote:

More importantly, who cares? In the abscence of a direct velvia comparison, the digital shot as put up looks boring. I think most people would have adjusted the color and saturation on that shot until they liked the results. If it doesn't match the color 100% of xxx film or reality, so?

Well, you cared enough to make the glib statements and claims below which are simply not true and which you are now backpedaling on:

"I'm not sure I could do it if I didn't have the Velvia shot in front of me to work from, but with a comparison shot, it took about 1 min in LAB to match; it was pretty easy."

"Heck, you can get REALLY close with no work at all. Just use the Match Color command."

"It's funny how you all think it's either impossible or at least very hard to do."