Monochrom Sensor Corrosion... opinion needed

[MIkhail]

For me I shoot as if I am a large format shooter and try to make the best at image capture. Rescuing images with post processing is not the way to go, and the best images really only need tweaking in Lightroom. I never add sharpness. Think of how large format shooters make negatives for contact printing.

Cal

Agree with all of that.

 

[Calzone]

Important note: I think shooting the Monochrom (original) is likely the most unforgiving digital camera ever made, because when you blow the highlights there is no way to recover. Gone forever. In this regard is totally unforgiving, kinda like driving a Porshe too fast in a curve, but if you nail the exposure you will be highly rewarded.

The M-246 seems to have more/wider dynamic range and is more forgiving in regard to the highlights.

Anyways shooting the MM will make you a better shooter.

Cal

 

[MIkhail]

Important note: I think shooting the Monochrom (original) is likely the most unforgiving digital camera ever made, because when you blow the highlights there is no way to recover. Gone forever. In this regard is totally unforgiving, kinda like driving a Porshe too fast in a curve, but if you nail the exposure you will be highly rewarded.

The M-246 seems to have more/wider dynamic range and is more forgiving in regard to the highlights.

Anyways shooting the MM will make you a better shooter.

Cal

You got me thinking... Is the use of color filters with digital camera even warranted? Not just an MM but any camera. i know that you can get everything in post-process but I wonder if in post you are tweaking something at the expense of something else... Maybe this deserves a separate thread.

 

[willie_901]

Blown highlights happen when the sensor signal capacity is exceeded, or when the sensor is not clipped but instead ISO amplification exceeds the maximum signal capacity of the analog-digital converter. The former occurs at base ISO and the latter happens above base ISO. In extreme overexposure, both happen.

This is the case for all digital cameras (except a few Nikon DSLRs that don't use electronic amplification to increase ISO).

Clipping highlights at the sensor is an analog phenomenon. It is difficult to describe exactly when information is irreversibly lost. The opposite is true for the anaog-digital converter. Once the incoming DC signal level exceeds the ADC's design maximum all information is lost.

Another difference involves how the camera data stream designers define 0 EV for the raw data file. While the total usable EV range (dynamic range) can't be increased or decreased, shifting where 0 EV falls can make either highlights or shadows appear easier to recover.

I have seen similar differences (apparant ease of highlight recovery) between different brand of DSLRs and mirrorless cameras. The difference between the default renderings of Nikon raw and Fujifilm raw in LR are significant. For Nikon raw I always started with selectively pulling highlights while I always start selectively pushing shadows with Fujifilm raw. Neither increases the DR... they just make full use of the DR defined by the original shutter and aperture settings (exposure).

None of this means the data stream and, or meter response of the original MM doesn't behave differently than the M-246. It doesn't mean learning to expose the original MM properly won't make one better at exposure. It just means there are technical differences and you have to work differently with the two cameras.

 

[pechelman]

Use Heliopan 2X yellow filters marked "Digital" that have additional IR and UV filtering for lower noise, higher signal to noise ratio, and less clipping. Basically speciffically with the Heliopan filter it seems to hit the sweet spot of the sensor.
Cal

agree with everything you said in this post, and especially liked your Moby Dick histogram comment. Such a good visual.

Regarding the heliopan "digital" benefits, I did some testing, admittedly pretty crude, but did it in high sun with materials that I know are UV and IR reflective. I tested the "light yellow 2x digital" heliopan versus both 022 yellow & 040 orange and then again with all 3 filters stacked with a bw uv/ir cut stacked ontop. Also tested with no filter and just the uv/ir cut to get a baseline. To boil it down, I couldnt detect any significant improvements in the colored filters based on their uv/ir capabilities and felt that certainly at least 90-95% of the differences on my screen were due to the color of the filter itself. Maybe the colored filters are by their nature uv/ir filters as well, but the spectrographs Ive seen from B+W dont seem to indicate that. I mean to revisit this when summer comes back around and there's more UV/IR in the sky.

Having said all that, I've also found that the heliopan light yellow is a great match to the 246 sensor which is already pretty contrasty. I feel it adds just enough separation to the midtones to make them malleable in post (so things dont just look like a mess of gray) without lightening or darkening the highs or lows too much. Also has the benefit (or not) of having the lowest filter factor. The BW 022 is good too sometimes, certainly with a bit more risk in blowing some highlights more I think, and the 040 I feel is a special use only, and should be treated similar to a red filter on film. It can have a very strong impact.

You got me thinking... Is the use of color filters with digital camera even warranted? Not just an MM but any camera. i know that you can get everything in post-process but I wonder if in post you are tweaking something at the expense of something else... Maybe this deserves a separate thread.

Traditional colored filters really only have a serious use on black and white (digi or film) imo. Certainly there were color filters to adjust for light temperature on film, but that's no longer an issue with raw shooting. If you consider a 720nm IR filter a color filter, then that's about the only "color" filter I can think of to use on color sensors.

For the 246, I really think that leaving a filter on all the time is reasonable. Take it off at night for sure tho as you'll probably get some reflections from incident light sources you wish you didnt have. In the case of the heliopan, the use of a filter is of course at the expense of a few things, but with the lower filter factor, it's honestly minimized. Unless you want a higher filter factor to shoot close to max ap in bright sun. Tangible compromises are generally slower shutter/higher ISO, increased chance of flare\reflection, and potentially a change in contrast you didnt anticipate (e.g. shooting in the mountains and creating a haze). Tangible benefits are generally more mallebale files (imo), a little protection of the front lens element if you shoot in dirty environments (beaches, snowy mountains, deserts), and potentially a little extra sharpness by reducing\eliminating chromatic aberrations (lens only has to focus one color of light now).



Regarding tonality; Books, shows, studying the masters' photos, etc. Also, I would suggest silver efex, but only as an easy analysis tool. Open up a photo and use the histogram tool to see what zones your tones are measured at. Certainly zones arent the end all be all and this is probably a rough approximation/conversion for the digital world, but I've always felt it helps calibrate my eye to what I'm looking for. Used in combination with a REAL calibration tool I think there's a lot of value in this.
AND EDIT: And the other obvious thing here is that you also need to print your work. Sorry I did not mention that before.

 

[MIkhail]

Traditional colored filters really only have a serious use on black and white (digi or film) imo. Certainly there were color filters to adjust for light temperature on film, but that's no longer an issue with raw shooting. If you consider a 720nm IR filter a color filter, then that's about the only "color" filter I can think of to use on color sensors.

For the 246, I really think that leaving a filter on all the time is reasonable. Take it off at night for sure tho as you'll probably get some reflections from incident light sources you wish you didnt have. In the case of the heliopan, the use of a filter is of course at the expense of a few things, but with the lower filter factor, it's honestly minimized. Unless you want a higher filter factor to shoot close to max ap in bright sun. Tangible compromises are generally slower shutter/higher ISO, increased chance of flare\reflection, and potentially a change in contrast you didnt anticipate (e.g. shooting in the mountains and creating a haze). Tangible benefits are generally more mallebale files (imo), a little protection of the front lens element if you shoot in dirty environments (beaches, snowy mountains, deserts), and potentially a little extra sharpness by reducing\eliminating chromatic aberrations (lens only has to focus one color of light now).

That's understandable.
I was thinking of any ways to possibly decrease the contrast of digital color files intended for further conversion to black and while.
This is in order to avoid that terrible "b/w HDR effect" which I shown earlier in this thread... maybe.
Than again, it is probably not worth it, correct exposure and correct post-processing should take care of things.

 

[Calzone]

You got me thinking... Is the use of color filters with digital camera even warranted? Not just an MM but any camera. i know that you can get everything in post-process but I wonder if in post you are tweaking something at the expense of something else... Maybe this deserves a separate thread.

MIKhail,

I happen to like shooting SLRs because I think using a Polarizer to control contrast and saturation is best. Perhaps again this is borrowed from landscape photographer's and large format shooters.

I think with any digital manipulation rounding off and signal errors get accumulated and compounded the more one relies on digital manipulation. My approach to minimize post and to do as much as possible at image capture is maybe a more "organic" way that is more pure.

I know that Heliopan filters marked "Digital" have both IR and UV filters built in that go unmentioned and unnoticed. I use Heliopan Digital filters on all of my digital cameras because it cuts down on both clipping and noise.

My use of yellow filters on my Monochrom I attribute to Michael Reichman's reporting in a preliminary review that Leica engineers stated that for apromatic response on the Monochrom sensor a slight yellow filter should be used. This is how I discovered the Monochrom's "Sweet-Spot." Separately when comparing Heliopan filters against B&W filters I stumble upon and discovered that the Heliopan filters marked "Digital" removed the IR and UV signals that contributed to noise and clipping.

At a PhotoPlusExpo I questioned a Heliopan representative to get the definitive answer of what made Heliopan filters an enhancement.

Cal

 

[MIkhail]

I use Heliopan Digital filters on all of my digital cameras because it cuts down on both clipping and noise.

Cal

Would you possibly have any examples to illustrate this?

At a PhotoPlusExpo I questioned a Heliopan representitive to get the definitive answer of what made Heliopan filters an enhancement.

Cal

What did he say?

 

[Calzone]

Phil,

My simple testing drew the conclusion about Heliopan "Digital" filters by a simple controlled experiment with only one variable: comparing a Plain 2X yellow Heliopan filter against a Heliopan 2x yellow filter marked "Digital."

In this A-B test I used the histogram on the back of my camera and the clipping indicators set at 1%. The conditions were high contrast summer light where I knew I would have clipping. I was surprised on how different the histograms were. The clipping became less or non existing (repeated testing under different lighting).

Like Willie suggests clipping is the result of overwelming the sensor, but removing non visual signal that can tip the amount of light into being too much signal is kinda shown to either eliminate or curb clipping.

Interesting to hear about your experience with the M-246 with filters.

Cal

 

[Calzone]

Would you possibly have any examples to illustrate this?



What did he say?

M,

Years ago I had a theory about these "Digital" filters having both IR and IR filtering, and I posted about it here on RFF. Some people challenged my assumptions and pointed out that IR and UV are opposite ends of the light spectum.

The rep manning the booth gave me a pamphlete and we discovered together in writing that in fact "Digital" marked filters have additional UV and IR filtering built in.

When I bought more Heliopan filters I took note that this is also in the instruction booklet that comes inside the packaging.

The histograms don't lie. I did the direct comparision with older 2X yellow Heliopan filters against the newer versions marked "Digital." Single variable and using the histogram and clipping indicators to quantify the result.

Pretty much a straight forward "controled experiment."

Cal

 

[pechelman]

Thanks, Cal.
Wondering if there's any difference in the cover glass or IR sesnor filter strength on the M9M and M246's that make any screw on filter differences less. I sort of thought there wasnt, but who knows.
Certainly, I did see a clear difference in the histograms and what was in the image on my M240 but those same things were not present on the M246 for me.

Either way, for completeness, I meant to link back to the original post I made a while back about this;
http://www.rangefinderforum.com/forums/showpost.php?p=2654228&postcount=4731

 

[jsrockit]

Important note: I think shooting the Monochrom (original) is likely the most unforgiving digital camera ever made, because when you blow the highlights there is no way to recover. Gone forever. In this regard is totally unforgiving, kinda like driving a Porshe too fast in a curve, but if you nail the exposure you will be highly rewarded.

The M-246 seems to have more/wider dynamic range and is more forgiving in regard to the highlights.

Anyways shooting the MM will make you a better shooter.

Cal

I think this goes for any CCD sensor.

 

[Calzone]

I think this goes for any CCD sensor.

John,

Is it not more so with a monochrome sensor? My thinking is because with a color sensor one does have an opportunity to recover due to the Bayer Filter Array.

Cal

 

[Calzone]

Thanks, Cal.
Wondering if there's any difference in the cover glass or IR sensor filter strength on the M9M and M246's that make any screw on filter differences less. I sort of thought there wasn't, but who knows.
Certainly, I did see a clear difference in the histograms and what was in the image on my M240 but those same things were not present on the M246 for me.

Either way, for completeness, I meant to link back to the original post I made a while back about this;
http://www.rangefinderforum.com/forums/showpost.php?p=2654228&postcount=4731

Phil,

Thanks for the link.

You provide some useful data. I noticed the same results when comparing B&W and Heliopan grading. The 2X seems to be the best tailored for my Monochrom.

Perhaps Leica cleaned up the histogram by adding more or better UV and IR filtering to the sensor on the M-246. Clearly the M-246 is a more advanced camera. Maybe that is why you did not notice any difference?

It would be interesting to do the controlled experiment that I performed using just the histogram and clipping indicators with the only variable being a Heliopan filter marked "Digital" verses one not so marked.

All I can say with certainty than my Monochrom responds rather drastically, not only in the shape and level of the histogram, but also in clipping, due to less noise (I consider the non visual information of UV and IR noise). Really the change for me is signal to noise.

Because of a possible advantage I now favor Heliopan filters marked digital on all my digital cameras. There is a possibility the effect I clearly see on my Monochrom might not be relevant say on my SL or your M-246, but out of habit I go Heliopan "Digital" anyway.

Cal

 

[pechelman]

Would be happy to try a non-digital vs digital heliopan comparison one day. Maybe if I'm ever in NYC or if you are in Denver and we have a chance to meet? Unfortunately, I just have no need to buy another filter (of the same color!) and hate to buy something with the sole intent of returning it. Doesnt feel totally right to me.

The only last oddity here is that I would have expected the M240 and M246 to react the same or similar to IR contamination since I would presume they have identical sensors minus the CFA. It was clear as day on the M240 as has been documented everywhere on the web, but really negligible or not present on the 246.

I suppose an additional test I could do is with an old hoya rm72 filter and an IR flash trigger (aimed at the lens) I have laying around. Might be something fun to occupy my time between turkey comas this weekend 😉

 

[willie_901]

Filters can limit the total light energy reaching the sensor. Many do this with frequency selectively.

However this is just a screw-on exposure compensation tool. In terms of exceeding the full-well capacity of the sensor one could also use the exposure compensation dial or in manual exposure just compensate with the meter. The effect is a constant so nothing changes... i.e. the signal-to-noise ratio is not improved using a filter unless it somehow causes you to maximize exposure (maximize signal) when the shutter is open. When exposure is maximized a filter can not improve the SNR compared to using no filter when exposure is equally maximized.

In terms of information content (tonality in the case of the MM) selective frequency filters do alter the relative sensor pixel photon counts as Sal mentions. This is not a trivial effect and many find it useful. And, as Sal stated, this could also make it less likely to exceed the sensor sites' full-well capacity (just as the EC dial does). But these are two different effects.

When exposure is maximized, the relative SNR at different pixels may be different but the total SNR will always be maximized when exposure is maximized.

Obviously filters are pure analog devices. The term digital filter is an oxymoron. However the term digital filter is a convenient way to indicate the filter was designed to minimize light energy contributions outside of the visible spectrum and, or selectively affect light frequency transmission.

Very few still camera sensors have a significant response to UV light, so the main advantage is at the other end of the spectrum. This is only relevant to exposure. Relative differences in pixel photon counts with selective filters can produce desirable aesthetic results.

 

[Calzone]

Filters can limit the total light energy reaching the sensor. Many do this with frequency selectively.

However this is just a screw-on exposure compensation tool. In terms of exceeding the full-well capacity of the sensor one could also use the exposure compensation dial or in manual exposure just compensate with the meter. The effect is a constant so nothing changes... i.e. the signal-to-noise ratio is not improved using a filter unless it somehow causes you to maximize exposure (maximize signal) when the shutter is open. When exposure is maximized a filter can not improve the SNR compared to using no filter when exposure is equally maximized.

In terms of information content (tonality in the case of the MM) selective frequency filters do alter the relative sensor pixel photon counts as Sal mentions. This is not a trivial effect and many find it useful. And, as Sal stated, this could also make it less likely to exceed the sensor sites' full-well capacity (just as the EC dial does). But these are two different effects.

When exposure is maximized, the relative SNR at different pixels may be different but the total SNR will always be maximized when exposure is maximized.

Obviously filters are pure analog devices. The term digital filter is an oxymoron. However the term digital filter is a convenient way to indicate the filter was designed to minimize light energy contributions outside of the visible spectrum and, or selectively affect light frequency transmission.

Willie,

I am using the word "noise" to define unwanted signal.

The elimination of unwanted signal also allows for maximizing exposure and increasing the signal strength of the visual information. In my framing the SNR is higher because like you suggest exposure is able to be maximized, but only because signal that is non visual is removed vial use of a filter.

This is comparing with and without filter and also adjusting exposure like you suggest.

Cal

 

[Calzone]

Back in July I contacted Leica N.J. to get on a waiting list for sensor replacement. I was told it would be about 4 months wait until I would get contacted to send in my Monochrom.

I contacted them today because more than 4 months has transpired and got a shipping label. The expected turnaround currently is 8-10 weeks. I presume the repair time got extended from 6-8 weeks due to the holidays. It seems that this waiting list works well.

It was explained to me that the real bottleneck is the back-order of sensors that is currently 4-5 months. I was also told that my Monochrom will get a complete overhaul and come back as if like a new camera.

February 2015 my Monochrom will be 4 years old. I'm hoping to get it back by then. Also note that Leica does not want you to send the body with the battery.

Cal

 

[cz23]

That's interesting, Cal. I sent mine to NJ a month ago without getting on a wait list. Yesterday they assured me 10 weeks maximum turnaround. So I question the benefit of the wait list.

In any case, it sure seems like a long wait. And I'm not even sure mine is corroded.

John

 

[user237428934]

And I'm not even sure mine is corroded.

John

So why exactly did you send in your camera?