Some news about the M8

[Riley]

Advertisement (gone after registration)

akk !

ofcourse i should have asked

 

why is it not an option to coat the sensor glass (the sensor has a thin glass coating) with an appropriate IR coating ?

[pascal_meheut]

You would have the answers by simply reading the beginning of this thread...

[xrogers]

Hi Folks, I have an M8 on order, but i am a little worried because i have decided to go for CZ lenses insted of Leica due to cost, and would be interested to know how this would affect what i can do with the camera. I understand i can get a filter to sort out the blacks, but what about the other problems with none coded lenses as the CZ lenses would be.

Many Thanks Northern Qtr

 

Here's my current mental summary of Sean Reid's, Micheal Reichmann's, and Guy Mancuso's experiments.

 

You don't get lens data in your photo's information without coding. This is not important.

 

Apparently there is some small amount of vignetting correction applied for wide angle coded lenses. This is nice, but isn't very important---the camera works very well with a wide variety of uncoded lenses. Don't worry about it.

 

For the IR shift issues, you have the following options with uncoded lenses:

 

1. Forget about the filters, go with the camera as is. This is a real option for many people. If you get off color in a photo where it matters, fix it in photoshop. The fixing may be trivial, or it may be a lot of work, depending on the image. Even people with coded lenses can and will take this option.

 

2. Put filters on your lenses, and use them as is without any coding. For lenses 28mm and higer, this appears to work just fine (Sean has been using a 50 Nokton with a filter on the M8 with fine results).

 

For very wide lenses, you get a cyan color shift in the corners (this has been verified with a Zeiss 21mm, whether or not a 24mm shifts color in the corners was unknown last I saw). For these wider lenses, photoshop would likely provide an answer again (some sort of graduated mask in a color correction layer), but that hasn't been tested.

 

If very accurate color is critical to your work, use filters on lenses 28mm and higher. Don't use wider lenses, or be prepared for some photoshopping when using wider lenses.

 

3. Aftermarket coding of lenses or lens adapters. Someone will be doing this, and nobody knows how well it will work at this point.

[Riley]

You would have the answers by simply reading the beginning of this thread...

 

well maybe i read the wrong part

 

"the reason the filter cannot be in front of the sensor is that an interferential filter cannot be coated to avoid reflection"

 

but they are operating on the premis that a 'filter' cannot be installed in front of the sensor

im not suggesting that

 

im suggesting coating the sensor glass directly

not an option for current sensors, its a factory job

 

Riley

[pascal_meheut]

Because according to what they say, an IR coating only would reflect too much light. Another possible reason is that it would induce the cyan shift in the corners too and be much more difficult to fix by firmware.

 

We do not know for sure, we are just guessing based on the few informations we have?

[Guest guy_mancuso]

This is my opinion how this all will work or my best theory of it.

 

If it stands the fix for Leica is using the IR blocking filters. Than you should use one for each lens. From 28mm up there will be no cyan corner cast but from 28mm to wider you will have cyan corners( still waiting on the 24mm ) Now from my prspective a coded lens will have the information to reduce or eliminate that color cast cyan in the corners with the 28mm down. This can be done in two ways one with the coding as we mentioned but also in a menu selection you can pick your focal length and the firmware applies the right correction for that lens. This will be able to let you use 3rd party lense that can't be coded , this is a nice option that we hope to have. So this is still in the air as to what Leica will do and handle this.

 

Now we need C1 to come out with a new profile for the use of the IR filters as it stands today from my testing you get a overall color shift that a new profile needs to be in place that corrects for the use of the IR fileter , the one that is in place is for without. This is not that hard and if anybody is good at this stuff than Phase One is, C1 has been around a long time and has dealt with magenta casts on MF backs.

 

Now in my testing with the IF filter the green blobs and blue refections are gone , that is a non issue now . Als the streaking which it really is and not banding has to do with a overzealous frequency on the sensor that maybe a hardware fix that needs to be fixed by repiar or replace the camer, not sure what Leica's plans are on that but as it stands today through my tsting with the IR blocking filter it does reduce the effect becuase some of that streaking from light sources is IR light it is not the cause nor the cure but it does help it a little to reduce some of it.

 

 

These are what i have figure out so far along with Sean and other board members trying to work the issues and get us out ther shooting. The issue is today that whatever leica will do is speculation and heresay or closely worded not confirmed and until they confirm what we are talking about what they will do than we are in holding pattern. Honestly there maybe a way to not use the filters at all but personally i like the horse talking in my ear before I believe anything, we simply need more facts.

 

 

I should ad there are a couple threads were information or samples to view will give you a better understanding. Mine are the Cut filters and the ISO 1250 threads along with the DMR Vs M8 thread which is a intersting look at the differences between the two. I consider the DMR the gold standard and want my M8 looking just like it. I don't think anyone would argue that point given the image quality of the DMR , so my wish is Leica and Phase is to use the DMR as the standard to adjust too. Also plase look at Seans threads on the board his wedding one is also a good example of the IR filters. Plus a few other members also ran some tests . We all hate testing but it needs to be done. Also Marc has a few threads that are very good. So look around gain the info need than make decisions based on gut feelings and the research that you gain.

[hrosell]

Advertisement (gone after registration)

This is my opinion how this all will work or my best theory of it.

 

If it stands the fix for Leica is using the IR blocking filters. Than you should use one for each lens. From 28mm up there will be no cyan corner cast but from 28mm to wider you will have cyan corners( still waiting on the 24mm ) Now from my prspective a coded lens will have the information to reduce or eliminate that color cast cyan in the corners with the 28mm down. This can be done in two ways one with the coding as we mentioned but also in a menu selection you can pick your focal length and the firmware applies the right correction for that lens. This will be able to let you use 3rd party lense that can't be coded , this is a nice option that we hope to have. So this is still in the air as to what Leica will do and handle this.

 

Now we need C1 to come out with a new profile for the use of the IR filters as it stands today from my testing you get a overall color shift that a new profile needs to be in place that corrects for the use of the IR fileter , the one that is in place is for without. This is not that hard and if anybody is good at this stuff than Phase One is, C1 has been around a long time and has dealt with magenta casts on MF backs.

 

I have the expectation that when I receive my M8 the streaking issue will not be there. I can live with the cuting filters provided we can obtain them for free or at a really downgraded cost. However, it's much more difficult to deal with the coding of the lenses. I have six Leica M glasses and I live in Argentina, so the matter of cost is truly significant. The filters woud have to arrive here via courier (more costs involved) and customs would charge 55% on item + freight. Sending/retrieving lenses for coding plus the coding itself would become prohibitive.

So, couldn't we insist before Leica that they include the manual coding in the menu? and of course the reshaping of the C1 profiles.

 

Cheers

[Guest guy_mancuso]

Just need to code 28mm down . I would code the 28mm . I have a some cyan bleed here. Now a profile will correct this also

 

Watch Ray 's shirt BTW a fine member here on our forum. He has a BLE shirt on. We still need a profile and colors are bouncing but the was the 28 mm F2 cron and with and without filter. You can tell immediately which one has the filter.

 

Sean may have more info on the 28mm since he is working those lenses

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Guest guy_mancuso]

Now look into the garage , the exposures vary slightly but are you noticing more DR with the filter

[harmsr]

Guy loaned me a 486 52mm cut filter with step-up rings from 46 and 49 so that I could use it on all of my lenses.

 

My lenses are 28 Cron ASPH (coded), 35 Lux ASPH (coded), 50 Lux ASPH (coded), 75 Cron ASPH, Tri-Elmar 28-35-50 second gen.

 

The IR block filter completely eliminated the magentas. Prior to this the magenta cast showed up in all kinds of things. It was not strictly on blacks. Guy took some photos of my car (gray) and myself (blue shirt) that were very magenta without the filter, but on color with the filter.

 

The IR block filter really does need a new color profile as it does affect the general color. It seems to remove a little red from everything. I'm not technical enough on light transmission or filters to know if this is true or if the combination of the filter and the sensor coating is starting to remove a little of the visible red spectrum also.

 

I reshot some night scenes with cars and neon in them last night which had the "green blobs" previously which did not have the "blobs" when I used the filter, however some minor banding could still be seen.

 

Now it is time to play in C1 & Lightroom to see how these shots actually come out when post processed.

 

Ray

[Guest guy_mancuso]

Thanks Ray for jumping in there , sorry I had to post you pic to illustrate this , hope you did not mind.

 

Right now Pase One needs to be working on this profile and do a update.

 

My wish list C1 profile and batteries for God sake

[Northern Qtr]

Thanks xrogers for your help, I have defo decided to go with CZ lenses as i see no point in spending twice the amount on leica glass when the CZ lenses are just as good. I suppose i will just have to wait and see what the outcome to the problems are along with the rest of the people on this forum. By the way i have waited so long for this camera from Leica, i did try the Epson RD1 but i had to return 3 bodys due to the frame lines which were out of alignment and hot pixels. Once again many thanks

[grober]

 

Just published part 4 of the review series which concentrates on this topic. Am currently working on a series of tests as well:

 

 

Cheers,

 

Sean

 

Sean,

 

I'm aching to read your latest on your site but can't remember my password. Please answer the e-mails I have sent to you on your site.

 

Thanks!

 

-g

[marknorton]

I'm really struck by the change in colour between the two pictures - it may be just the profile, but the vegetation looks much better in the filtered shot. Guy, straight out of the camera, can you say which is more accurate?

[sdai]

I'm wondering too, Mark ... the vegetation does seem to be greener but so is the sky, and roof???

[Samir Jahjah]

Here's my current mental summary of Sean Reid's, Micheal Reichmann's, and Guy Mancuso's experiments.

 

You don't get lens data in your photo's information without coding. This is not important.

 

Apparently there is some small amount of vignetting correction applied for wide angle coded lenses. This is nice, but isn't very important---the camera works very well with a wide variety of uncoded lenses. Don't worry about it.

 

For the IR shift issues, you have the following options with uncoded lenses:

 

1. Forget about the filters, go with the camera as is. This is a real option for many people. If you get off color in a photo where it matters, fix it in photoshop. The fixing may be trivial, or it may be a lot of work, depending on the image. Even people with coded lenses can and will take this option.

 

2. Put filters on your lenses, and use them as is without any coding. For lenses 28mm and higer, this appears to work just fine (Sean has been using a 50 Nokton with a filter on the M8 with fine results).

 

For very wide lenses, you get a cyan color shift in the corners (this has been verified with a Zeiss 21mm, whether or not a 24mm shifts color in the corners was unknown last I saw). For these wider lenses, photoshop would likely provide an answer again (some sort of graduated mask in a color correction layer), but that hasn't been tested.

 

If very accurate color is critical to your work, use filters on lenses 28mm and higher. Don't use wider lenses, or be prepared for some photoshopping when using wider lenses.

 

3. Aftermarket coding of lenses or lens adapters. Someone will be doing this, and nobody knows how well it will work at this point.

 

 

And for each of these options, you will need a specific color profile !

We all try to smooth or digital workflow, now Leica solution is complicating everybody's life with their filter "solution".

[hrosell]

Just need to code 28mm down . I would code the 28mm . I have a some cyan bleed here. Now a profile will correct this also

 

Guy,

Sorry, you are right and you stated this before. However I do have both a 2.8/24 and a 2.8/28, both non-Asph. My point about cost still subsists: two lens back and forth from this area plus the coding mean about 10% of the M8 value, at least!, so a manual alternative in the menu would be very welcome!

 

And I adhere to your plea for extra batteries (I had already asked for an extra one but it seems it may take some time o get it) and the C1 new profiles.

 

Regards

[wparsonsgisnet]

Now look into the garage , the exposures vary slightly but are you noticing more DR with the filter

 

Guy, It appears, however, that the darkness around the front door is darker without the filter. So, the white wall in the garage must be reflecting IR and the front door, not.

[Guest guy_mancuso]

Most like Bill , even in the shade IR is present and the garage is just a white blanket of walls

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[harmsr]

Mark,

 

Neither one is really accurate as it is out of the camera (these photos). We had to add a little magenta back into it in C1 to get it close.

 

The tree in the second shot gives the appearance of how the color looks when using a polarizer in this light, while the rest is too cyan across the whole frame.

 

The first shot is closer on the trees, but off on the rest (too magenta).

 

It really seems as if the IR filter MUST be used to stabilize the light (remove the IR) and then you can adjust in C1. You can get very close in C1 with color adjustments or adjusting the custom calibration sliders in LightRoom, once you use the IR filter.

 

It appears as if the 486 filter / sensor cover combo may remove a little too much. (In other words it actually dips into the magenta/red of the visible spectrum.) However, this is now across the entire frame and is correctable.

 

As Guy mentioned, with the filter, DR and sharpness appear to increase.

 

As a side note, when using the IR filter the AWB works MUCH better. It still needs some firmware tweeking, but is by no means as inconsistent as it is without the filter.

 

Ray