Jump to content

What about a BW-M8?


eckart

Recommended Posts

Advertisement (gone after registration)

You did not mention a cheap camera with 14 MP, but I did.

What I tried to declare with my example is, that the size of the file says nothing about the details.

The amount of the pixels, in your example, stay the same, nobody deny that, 10,3 MP stay 10,3 MP, but the possibilities of describing details are changing with the different use of them.

 

Now we are getting somewhere. You are absolutely correct, a file’s size is not a measure of its quality; it is only an indicator of how large it can be output without further interpolation. A persistent misconception with digital photography is that resolution (pixel count) is the ultimate indicator of image quality. Camera makers are more than happy to take advantage of that misconception by cramming as many photosites as they can on sensors smaller than your fingernail (much smaller in cell phones and some P&S cameras).

 

I was making a point that a sensor's resolution does not change if the Bayer filter is removed, and as you just agreed with me, 10.3mp will still be 10.3 mp after you remove the filter. This was my point of disagreement with Michael, who stated the Bayer filter drops the resolution of the sensor from 10.3mp to 5 or 2.5mp "effective" resolution. It is easy to throw around a lot of numbers, but the end result that counts is what hangs on your wall. If you can't see a difference in the print, why even discuss this?

 

With digital images, resolution is measured by the number of points or pixels your printer can place on a sheet of paper measured in ppi (I am ignoring displays because by nature they are low resolution). That ppi number is directly related to the number of pixels the camera's sensor has onboard. The key is to separate resolution (pixel count) from image quality, which is more influenced by physical pixel size, pixel spacing, AA filter, electronics and software.

 

I frequently make large prints in color and B&W from my digital files – often from the M8. Even under close scrutiny, the prints bear no resemblance to any 2.5 or 5mp image files that I have ever worked with. My first digital camera was actually under 2mp, and I have used most of the pro digital cameras (including MF) from then to the present including some that far exceed the M8's 10.3mp. I have the ability to critically compare images from film, scanned film, and numerous digital camera files of various resolutions.

 

Yes, and when you take a photograph with a sub-standard lens, don’t focus correctly, or get motion blur from camera shake, what do these 10.3 million pixels buy you in terms of resolution? What if you trade your M8 for some cheap compact digicam that also happens to be equipped with a 10.3 megapixel sensor? Do you think 10.3 megapixel is 10.3 megapixels, resolution-wise?

 

Effective resolution is the power to resolve fine detail and it is measured in lines or line pairs per image height (or mm), not pixels. That’s where different cameras and lenses differ, even when their megapixel count is the same. If you do not care for effective resolution, what do you care for?

 

Michael,

 

With all due respect, you are now starting to babble.

1. I do not shoot with sub-standard lenses, I use hand-picked primes even with my dSLR cameras.

2. I learned to focus rangefinders and SLRs over 40 years ago and still have not forgotten how. Incidentally, the last 40 years were spent shooting advertising and commercial assignments in all formats up to and including 8x10 film. Ever focus on multiple groundglass points through an f5.6 or f8 lens?

3. I do not worry about motion shake or blur because when necessary, I do use a tripod!

And finally,

4. I never said all camera sensors having the same pixel count will produce the same quality output. Your mention of me trading the M8 for a digicam brings this entire discussion to a level of silliness that is just not worth my time to prolong.

I tried to be polite in my responses and willing to agree to disagree, but you are bringing this down to a level of childishness I don’t wish to participate in. If you wish to continue, do so without addressing me and place me on the ignore list as I will do with you…

Link to post
Share on other sites

  • Replies 74
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

With all due respect, you are now starting to babble.

1. I do not shoot with sub-standard lenses, I use hand-picked primes even with my dSLR cameras.

2. I learned to focus rangefinders and SLRs over 40 years ago and still have not forgotten how. Incidentally, the last 40 years were spent shooting advertising and commercial assignments in all formats up to and including 8x10 film. Ever focus on multiple groundglass points through an f5.6 or f8 lens?

3. I do not worry about motion shake or blur because when necessary, I do use a tripod!

And finally,

4. I never said all camera sensors having the same pixel count will produce the same quality output. Your mention of me trading the M8 for a digicam brings this entire discussion to a level of silliness that is just not worth my time to prolong.

I tried to be polite in my responses and willing to agree to disagree, but you are bringing this down to a level of childishness I don’t wish to participate in. If you wish to continue, do so without addressing me and place me on the ignore list as I will do with you…

I am slowly getting the impression that you are not even trying to understand what I have written. Be that as it may, I think everyone else will have understood me just fine, so guess the time I have put into this wasn’t entirely wasted.

Link to post
Share on other sites

I am slowly getting the impression that you are not even trying to understand what I have written. Be that as it may, I think everyone else will have understood me just fine, so guess the time I have put into this wasn’t entirely wasted.

 

Sometimes reading one of these infra-red* threads I find myself thinking "I really hope Hußmann comes and posts something sensible."

 

Thank you - and please keep it up!

 

 

 

 

* Infra-red: more heat than light.

Link to post
Share on other sites

I am slowly getting the impression that you are not even trying to understand what I have written. Be that as it may, I think everyone else will have understood me just fine, so guess the time I have put into this wasn’t entirely wasted.

 

Sorry Michael, but I completely understand what you are saying, it's not that difficult to follow. Because I do not embrace your views does not mean I am incapable of understanding your comments.

 

I have worked with resolution charts, am quite familiar with lens MTF curves, used the Zone System and plotted H&D curves, and processed all of my film in my studio well before digital became my main media. So please, don't be so condescending. Although I replaced densitometers with spectrophotometers, it does not mean I am any less proficient with digital technology.

 

One of the things I discovered long ago about lens MTF curves and other specifications is that empirical results often contradicted them. Only when you physically test the actual lenses on actual cameras shooting real subjects with actual film (or sensor) and evaluate the results do you have that empirical data.

 

When you get your hands on an M8 with a sensor sans the Bayer array, and have tested your assumptions with image captures, perhaps you will be in a better position to convince me. You certainly will not convince me hypothetical proposals.

 

In my last reply to you, I neglected to reply to your final comment/question:

 

"Effective resolution is the power to resolve fine detail and it is measured in lines or line pairs per image height (or mm), not pixels. That’s where different cameras and lenses differ, even when their megapixel count is the same. If you do not care for effective resolution, what do you care for?"

 

My obvious answer to that question is the matted and framed images that hang on walls. None of them are of charts with line pairs as the subject!

 

I am sure many here understood what you said at least as well as I have, go easy on the ones who don't agree with your conclusions. Now I will click the ignore button; it's been nice communicating with you.

Link to post
Share on other sites

....If you can't see a difference in the print, why even discuss this?

 

I am sorry, but I can't understand, why you believe that.

10.3 MP Colour have the same amount of pixels than 10.3 MP Monochrome, yes, but we agree the size of the file does not show it's quality, right?

To describe any colours which is not exactly blue, green or red, the colour version has to use more than one pixel to do this, yes?

At a monochrome sensor each pixel has just to describe the intensity of the light which is falling on him, right?

That means in my eyes, the possibility to describe details must be better with a monochrome sensor and of course you will see that in a print.

Link to post
Share on other sites

I am sorry, but I can't understand, why you believe that.

10.3 MP Colour have the same amount of pixels than 10.3 MP Monochrome, yes, but we agree the size of the file does not show it's quality, right?

To describe any colours which is not exactly blue, green or red, the colour version has to use more than one pixel to do this, yes?

At a monochrome sensor each pixel has just to describe the intensity of the light which is falling on him, right?

That means in my eyes, the possibility to describe details must be better with a monochrome sensor and of course you will see that in a print.

 

Eckart,

 

We seem to be talking about various aspects of an image file and not always discussing the same one. Please bear with this rather lengthy reply with information you probably know already because it may help clarify my position. Feel free to disagree with any points I make; there will be no ill feelings or assumption of limited comprehension on your part.

 

All sensors are monochrome devices with a finite number of photosites, which in the case of the M8, is about 10.3 million. Calling it 15mp or 20mp does not make it so. The M8’s sensor only reacts to the intensity and amount of light that falls on it within a given timeframe. The 10.3mp are real pixels, what is interpolated is their color. Regarding file size – yes, we agree up to a point. Obviously if you take a 10.3mp image file and interpolate it up to 20mp, it will increase in file size but not in quality. Pixel size is more important than pixel count. I do not entirely agree with your conclusion, and I will try to explain myself better starting with an example and propose another conclusion.

 

Some camera manufacturers play games with resolution numbers by claiming more “effective” resolution than the number of photosites on the sensor – Fuji and Sigma come to mind. Fuji claimed due to the unique shape of their photosites (hexagonal as opposed to square or Nikon’s rectangular), and diagonal arrangement, their cameras doubled the “pixels.” They sold a lot of “12mp” cameras with 6mp sensors. I tested them, and while they were among the best 6mp dSLRs at the time (exceptional color!), they were still 6mp despite Fuji’s in-camera conversion up to 12mp. Sigma makes similar claims for their Foveon sensor, but resolution is still judged by the number of addressable points on a sensor (i.e. photosites), anything else is simply interpolation.

 

All digital cameras capture exposures as analog voltage signals sent through an A/D converter to become a digital Raw file. The Raw file data contains the amount of light that struck each photosite, identifies the Bayer filter color covering each photosite, and includes the metadata recorded by the camera such as dark frame subtraction, date, time, lens, optical correction, exposure, etc. At this point there is no color image and the Raw file is just a data file with a linear gamma of 1.0. Until it is processed through a Raw file converter (which besides demosaicing, applies a gamma correction curve), it looks extremely dark like an underexposed image. We are never meant to see that image because it would really hurt camera sales.

 

Your Raw converter software affects every aspect of your output image file including apparent detail, color, and tonal range at the top of the list. Demosaicing can be accomplished with as few as four photosites to determine the values of the two missing colors in any given pixel, but you would not be very pleased with the results. With better sampling of more surrounding photosites, the results improve dramatically. Every Raw converter has proprietary algorithms for assigning color to individual pixels. I use most of the pro-level Raw converters and have observed very noticeable differences between them on the same files. Even newer versions of the same coverters can show surprising improvements on the same files.

 

Another major influence on the final output image is the earlier mentioned A/D converter, which takes the analog output from the sensor and converts it to the essential digital bits, but that is a complicated subject to cover in this thread. The point I am making is there are far more things having greater impact on image quality than the sensor’s Bayer array. The worst image destroyer is the AA filter, which the M8 fortunately lacks. I do not believe the Raw files as delivered out of an M8 and converted by a good Raw converter will lose any noticeable detail in a B&W conversion because of the Bayer array. I also don’t believe an M8 stripped of its Bayer array will produce a noticeable increase in “perceived” resolution.

 

When advocates of a B&W only M8 sensor can show me comparisons from such a sensor with the current sensor having B&W conversions made of test images with identical subjects from the best Raw converters, I am willing to learn and adapt my thinking. So far, all I have read was speculation and theory. My personal experience contradicts the speculation.

 

To sum it up, a pixel is the smallest element of your digital image. The smallest subject element you can capture will be the size of your pixel. This is how resolution is commonly defined. Anything smaller than the photosite will only produce an average of the smaller subject element and its surroundings within that pixel. One can shoot line pair charts trying to prove a point, but in the end, a line chart is not what you or buyers of your prints will want to display on a wall, is it? There many ways to get that image on a wall, but as I stated, unless you interpolate the image up or down, a 10.3mp image printed at 300ppi will still print a 3916 x 2634 pixel image that is 13 inches by 8.78 inches on a piece of paper, canvas, whatever...

Link to post
Share on other sites

Advertisement (gone after registration)

Sorry - I'm an empiricist. I've posted an image showing that the M8 (with bayer filter pattern) can resolve fine details down to one pixel wide.

 

In the face of empirical evidence, words and theories count for very little.

 

"The great tragedy of Science - The slaying of a beautiful hypothesis by an ugly fact." - T. H. Huxley

 

This thread can grow for the next 100 years - but until someone posts an IMAGE SHOWING how and that an un-bayer-filtered M8 sensor will resolve details SMALLER than one pixel, it's just hot air.

Link to post
Share on other sites

I dont think that im interested of making m8 sensor to monochrome..

 

just make one NEW sensor with more photosites crammed in.. ,

 

it is maybe less complicated to make a monochrome sensor producing more photosites in the same area, allowing higher tolerances compared with color ones. Use software tweak for dynamic range. Throw out RGB :)

Link to post
Share on other sites

I've posted an image showing that the M8 (with bayer filter pattern) can resolve fine details down to one pixel wide.
Actually one line pair across three pixels, the typical resolution of a Bayer sensor. Nothing special I’m afraid.
Link to post
Share on other sites

Back in 2003, at the LHSA meeting in St. Louis, Stephan Daniel asked for a show of hands: How many people would buy an M8 with a B&W-only sensor if Leica produced it? He mentioned several things that Michael has talked about, including the greater sensitivity and greater effective resolution.

 

Many of us raised our hands. However, the LHSA is a self-selecting group of people, not typical of photographers in general, and probably only partially typical of Leica M photographers. My guess is that Leica looked at the extra cost of producing such a camera and decided that it wouldn't sell enough to be worthwhile.

 

The consolation prize is that the M8 is a wonderful B&W camera anyway. Use the channel mixer, and you can select your tonal response with far more variation than a few colored filters give you with B&W film. In poor light, you can select the color mix that gives you the best rendition or the least noise, or balance the two.

 

Of course, a "Bayerless" monochrome sensor would give less noise at higher ISOs. This would be the prime advantage, as the standard M8's resolution is already pretty spectacular due to the lack of and AA filter.

 

But I have to ask myself, would I buy two M8s just to get the monochome sensor? Probably not. I couldn't afford it new, and I suspect an monchrome M8 would not lose its value on the used market as fast as the standard one.

 

--Peter

Link to post
Share on other sites

Very interesting, Peter,

but 2003 was 3 years before the release of the M8, don't you think, he was asking in this times not for a additional option but more for a only-option?

It was a major thing to bring the M and digital together, as I understand are some problems not this serious with a monochrome sensor, for example the flat light at the corners.

Maybe that's why he was asking.

But now, the M8 is out and have been sold not bad. The M9 seems to rise at the horizon.

To bring out a monochrome M should be a lot easier and cheaper now, than in 2003.

In this thread and in the German one, there have been quite a lot users who found the idea fascinating and could imagine to buy such a specialist.

A out coming M9 will sell good, but I am afraid, not as good as the M8, price wise and because of already having a M8. Maybe it would be not such a bad idea to start this balloon as well, when it is not too expensive to realise, because this would be a perfect second body.

Link to post
Share on other sites

Last week, I wrote an email to Stefan Daniel from Leica about this theme.

Today I received the answer.

 

Sehr geehrter Herr Friz,

vielen dank für Ihre Anfrage nach einer Schwarz-Weiss M Kamera. Wir hatten derartige Konzepte am Anfang der M8 Entwicklung untersucht, ein Qualitäts- und Empfindlichkeitsvorteil wäre deutlich zu erwarten, leider haben die wirtschaftlichen Randbedingungen (zu erwartende Stückzahl und Entwicklungskosten für den Sensor) eine Realisierung nicht zugelassen. Wir haben auf das Thema jedoch weiter ein Auge.

Besten Dank für Ihren Hinweis.

 

 

He said (No professional translation;)),

1. they thought about that in the beginning of the M8 development, but decide in this time against.

2. there would be obvious an advance of quality and sensibility

3. they have and will have an eye on this idea

Link to post
Share on other sites

  • 2 months later...
  • 4 weeks later...

One of the things I love about the Epson rD1 is the ability to shoot in raw but have it set to b&w. The screen shows the image in B&W and usind the Epson software to download the image it comes up in B&W. I wish at the very least I could do that with my m8.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...