Filter issues, E. Puts

[Peter Branch]

Advertisement (gone after registration)

LFI explained it. Two factors were involved:

 

1.Nobody expected it to be such an issue. After all, D70 type sensors exhibit much the same type of spectral response, as do some others, but there was scarcely negative feedback on those.

2. Leica was tweaking WB right up to the time of release (hoping to get it down to an acceptable level?), so testers were told to disregard colour balance issues.

 

In retrospect not very smart, but understandable and certainly no hidden agenda.

 

So now we are invited to accept that LFI is independent and impartial!

 

There can be little doubt that Leica were trying to “optimise” the colour response prior to release and were plagued by the IR sensitivity, the evidence is everywhere. All photography is a matter of compromise and subjective judgement – but – we should not be expected to accept basic faults that have been known about, investigated and warned against over decades. Just go and look at the literature published by Kodak, Fuji, Agfa and others over the last four decades. Regardless of the WB that Leica eventually came up with for JPEG files the reviewers are known to have used DNG files. The IR linked problems were spotted by numerous users within hours of the cameras becoming available.

 

It is not a conspiracy by Leica and I don’t think my words could be taken to imply that it is. In the strict legal sense I doubt it is a conspiracy amongst the reviewers but they are behaving in a collective manner which defies logic. If they were told to ignore any colour response issues why have they not come out and said so themselves? And why are they now downplaying the issue to a degree that makes them look foolish?

 

I had a long discussion with a senior Leica manager back in May/June 2006 about when the digital M would be available and what was the state of development. In the course of that discussion it became evident that sample cameras were already in the hands of selected people outside the company. I said I hoped they were giving the cameras a “good going over” – he immediately responded by saying that the problem with “testers” was that they were so experienced that they did not make the mistakes that ordinary users make and further more if there were some anomaly they would be the first to find a workaround – how prophetic!

[Guest guy_mancuso]

Since it is in the literture than I can confirm that there is a ON/IR filter in the menu

[chetccox]

Guy -

 

You pretty much already did, and I figured you knew. I read between the lines several days ago on one of your posts. Thanks for the info. Now, if they would just release the upgrade!

[Peter Branch]

What it actually says is 'higher than 1.10'.

 

My apologies - you are correct. I had read it as 1.10 or higher but that is not correct. It does say "...higher than 1.10......"

 

Could this be the fabled version 1.13 that has allegedly been spied on reps cameras?

[Guest guy_mancuso]

Hopefully soon Chet, I would imagine there still tweaking it some and working on fixes. But so far it looks good folks is all i can really say. Maybe there throwing in some of our feature sets, not sure but would be nice . I personally think firmware will be upgraded for awhile with more new things coming that we want. once the fixes are in they can concentrate on the improvement stuff.

 

BTW just so folks know Chris Tribbles survey was forwarded to them and they responded immediately and thanked him for all the info and mentioned it is very useful to them. So we owe Chris a big thank you for putting out the effort there on the survey. I think info like that is vital to a company to improve there products. I scan this forum everyday looking for stuff like that so I can forward it on. There probably pretty tired of me but I feel if it helps the end user and leica than it is worth the effort. So do keep the issues out there, nothing worse if we miss something and it starts showing up

[jaapv]

So now we are invited to accept that LFI is independent and impartial!

 

There can be little doubt that Leica were trying to “optimise” the colour response prior to release and were plagued by the IR sensitivity, the evidence is everywhere. All photography is a matter of compromise and subjective judgement – but – we should not be expected to accept basic faults that have been known about, investigated and warned against over decades. Just go and look at the literature published by Kodak, Fuji, Agfa and others over the last four decades. Regardless of the WB that Leica eventually came up with for JPEG files the reviewers are known to have used DNG files. The IR linked problems were spotted by numerous users within hours of the cameras becoming available.

 

It is not a conspiracy by Leica and I don’t think my words could be taken to imply that it is. In the strict legal sense I doubt it is a conspiracy amongst the reviewers but they are behaving in a collective manner which defies logic. And why are they now downplaying the issue to a degree that makes them look foolish?

 

I had a long discussion with a senior Leica manager back in May/June 2006 about when the digital M would be available and what was the state of development. In the course of that discussion it became evident that sample cameras were already in the hands of selected people outside the company. I said I hoped they were giving the cameras a “good going over” – he immediately responded by saying that the problem with “testers” was that they were so experienced that they did not make the mistakes that ordinary users make and further more if there were some anomaly they would be the first to find a workaround – how prophetic!

 

Where in my post did I say LFI was objective?

No- we don't think LFI is independent and impartial. It is a way for Leica to say, leak and explain things that they can't or won't say as a company. It is a pretty good read, and I must say, it reads more objectively than some of the "free" press I have seen. Let's call it "embedded".

And yes- it was well known in the beginning of 2006 that the camera was being beta-tested. One did not need to speak to "senior managers" for that. Simply reading LFI -or even the internet forums sufficed.

 

If they were told to ignore any colour response issues why have they not come out and said so themselves?

 

You should inform yourself better - that is exactly what they have been saying, like for instance, LFI.

 

Or M.R.

As I see it, it is inconceivable that Leica's engineers and its alpha and beta testers did not see and were not aware of this problem. Early external testers (like me) missed it at first. One has to shoot certain types of subjects under certain light for it to be obvious. And for the type of shooting that most people do with an M Leica – documentary – whether someone's jacket or hat was a slightly different shade than it really was isn't readily apparent after the fact. Also, shooting a standard target like a Macbeth chart doesn't display the problem.

 

Sean has a similar text on his site,

 

and so on...

[skunkworks]

Advertisement (gone after registration)

Erwin may know something about optics but some of his testing methods and writings on the subject of digital have been hilarious they are so clueless.

I've read several posts that indicate that Mr. Puts doesn't know about digital photography but I haven't read anything he's written that's "clueless". Can any of Puts' detractors give examples? I'm not trying to start an argument but I enjoy his writing, and I consider him an authority; if I have the wrong opinion of the extent of his knowledge I'd like to understand why.

 

Even Erwin's lens reviews I found less then informative -as opposed to say Sean's lens reviews that always contain extensive real world examples which provide some useful information about whether a lens is likely to be to your taste or not.

(I have an optics fetish and an analytic mind, so I thoroughly enjoy Puts' reviews as a geeky indulgence.)

[lct]

What it actually says is 'higher than 1.10'.

True in German, French, English and Spanish but not in Italian if i'm not wrong:

"Assicurarsi che la propria LEICA M8 disponga di una versione del firmware 1.10 o superiore"

[chetccox]

Your translation is correct. So, 1 out of 5.

[Guest tummydoc]

Since it is in the literture than I can confirm that there is a ON/IR filter in the menu

 

If I'm using one of my older uncoded lenses will I be able to tell the M8 I have an IR filter on it, or only with coded lenses?

[jaapv]

It will only make sense with coded lenses, as the camera uses the information to correct - and that is lens specific. Without "knowing"at least the focal length, the medicine is worse than the affliction.

[Guest guy_mancuso]

You can turn the filter on , no problem there . The only issue will be if it is 35mm and wider it will NOT correct the cyan cast UNLESS there is a menu option to do so but if you have a 50mm and above and have the IR filter on than you will get the color presets from the camera so your good as gold there. There is another feature but i can't talk about that.

 

Let's see if I can talk about this a little. A couple things that are good about 1.10 or higher

 

First is color with the IR filters on and selected than you get the benefit of the color presets that are in the firmware, so any lens with a filter on can take advantage of that.

 

Coding really is only needed for 35mm and wider and the firmware will correct the cyan cast if the camera KNOWS what lens is one there, right now it is coding that will trigger that.

 

Now if Leica decides to implement a optional menu item for the user to select the focal length than that would correct the cyan cast. This is a feature set that the users here have asked for to be implanted into the firmware. This will at least get you very close to a Leica coded lens maybe not exactly but close. This is a guess at the moment but seems logical a 25 Zeiss maybe a little different than a leica 24mm

 

I am intentional leaving out a certain section here but it will be a benefit to users.

 

Honestly not sure how much i can talk about things so I am probably on the edge here and I am sorry I can't talk about it more. But it will be out soon and it should be very good for the end user

[Guest guy_mancuso]

If you happened to miss the thread on what features the LUF members asked for in order that was sent to Leica to hopefully implement in future firmware . Here it is

 

1. Lens selection option "in menu" Generic 15, 21,24,28,35 otherwise with coding it is automatic regardless of setting

2. Optional 16 Bit setting like the DMR or better option

3. EV/ISO option with controls of either arrows or wheel , protect or set buttons. Whatever is fast and confirmed. Idea flash in VF what was set when confirmed

4. B&W optional b/w review, without having to write JPEG

5 Warning of non coded lens

6. Full name of lens in EXIF data sample 50mm summilux 1.4 as it is read by the coding

7. 15mm F4, 24mm 1.4 or F 2 , 28 1.4 ( some really fast wide angles )and than a line of lenses like the 28 2.8 that will be less costs like a 75 2.8 , 90 2.8 or think about 1.33 image circle digital specific lens

8 SDHC compatable or new standard This is not a leica call per say but industry standard

9 Travel charger smaller

[Jamie Roberts]

BTW--and for what it's worth--anything Leica can do in firmware you can do in Photoshop (or in some future RAW converter that corrects lens vignetting).

 

So I wouldn't worry too much about cyan vignetting on non-coded lenses. It *is* a pain, but there will be a number of good (and pretty easy) ways to correct in post. And FWIW, I'd rather deal with a vignette than the distortion in wides I get with other systems...

[rosuna]

IR light focus on a different plane than visible light, isn't?

Does IR contamination affects sharpness?

[jaapv]

Yes

[Jamie Roberts]

IR light focus on a different plane than visible light, isn't?

Does IR contamination affects sharpness?

 

In theory, yes.

 

In practice, I've never seen an unsharp unfiltered M8 print that couldn't be explained by (my) user error, unless the presence of IR is so strong the colours mush together (as in Sean's Christmas tree / light shot).

[Guest tummydoc]

It will only make sense with coded lenses, as the camera uses the information to correct - and that is lens specific. Without "knowing"at least the focal length, the medicine is worse than the affliction.

 

Agree except if the firmware has different sets of colour algorithms for with and without the filter. In that case shooting through an IR filter that the M8 doesn't know is there, might muck up the colour balance. Sure it would be a rightable wrong with a custom profile, but if one were shooting with an assortment of coded and uncoded lenses it could get to be quite a challenge parcelling them out to batches in order to run them through different conversion profiles.

 

The reason I brought it up is that my first and second M8s, before they died, were running 1.09 and with an IR filter in place the colours were better (IMHO) than what I'm getting now with my latest M8 running 1.092...whilst on the other hand 1.092 is much better compared to 1.09 when not using the filter. Thus I wondered whether 1.10 might have the 1.092 colourset when running with the IR-filter disabled, and revert to the 1.09 colourset with the filter enabled.

 

Thanks Guy for offering up as much as you can. Sounds very promising. I appreciate that the M8 may rightly be more empowered and made convenient through the use of lens coding, however as a collector and knowing as Leica surely must that for many people the plethora of older lenses is a selling point, I'm hopeful they will do their best to retain the utmost usability of those older optics.

[rosuna]

Another point.

Erwin Puts compares two cases: with filter and without filter.

However, the two real alternatives for Leica were: with filter on the lens and with filter on the sensor.

IR filters have effects on the final image, in addition to the IR cutting, and this is true for filters on lenses and for filters on sensors.

Leica's solution takes off the IR light before it comes in the system, not after (the case of sensor's IR filters).

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Jamie Roberts]

Another point.

Erwin Puts compares two cases: with filter and without filter.

However, the two real alternatives for Leica were: with filter on the lens and with filter on the sensor.

IR filters have effects on the final image, in adition ti the IR cutting, and this is true for filters on lenses and for filters on sensors.

Leica's solution takes off the IR light before it comes in the system, not after (the case of sensor's IR filters).

 

Not quite. There *is* an IR filter on the M8, IIRC. It's just not strong enough to filter ALL the IR late in the higher red spectrum. Hence the need for a filter.

 

If they had gone for a stronger (thicker) filter, the sharpness we all love in the M8 would have been compromised.

 

I've seen all kinds of differences now with filter or without, but sharpness just isn't one of them (caveats on heat lamps, christmas tree closeups, and other torture tests)