dcsang
Canadian & Not A Dentist
Let me preface this post with the following: I do not own nor have I seen an M8. I am going by reliable information that I have from a trusted source; one whom I have known for a number of years and one who has seen and has images from a production body.
Now, what I have heard is that the M8's images; at least at 320 ISO (as I was told) had excessive amounts of moire. There was "purple fringing" everywhere as well.
I was surprised to say the least after reading all these threads here claiming the quality of the M8 being up to snuff.
I asked what lens was used (since the Chromatic Aberrations are usually a product of the lens to the best of my knowledge) and was told that the shots were take with the brand new Leica 28mm. Again, a bit shocked, as this is a brand new production lens released along with the M8.
Do I believe this? Well, I want to see the images first and hopefully I'll be able to do so at some point in the coming week.
The story is that a certain "local to me" online luminary has a production M8 body and the 28mm lens tesing out before producing a review. This particular person dropped into a favourite haunt of mine and one thing led to another and these comments (re: purple fringing and moire) were made by my source and NOT the online luminary. One thing my source said to me was, when I asked what the images reminded him of, "they look like they came from a Kodak 14N . . . " That was what really disappointed me.
That being said, as I have stated, I have not seen the images so take what I have typed for what it may or may not be worth. I would be extremely disappointed if the M8 exhibited such behaviour as Leica is banking a lot on this imho.
Dave
Now, what I have heard is that the M8's images; at least at 320 ISO (as I was told) had excessive amounts of moire. There was "purple fringing" everywhere as well.
I was surprised to say the least after reading all these threads here claiming the quality of the M8 being up to snuff.
I asked what lens was used (since the Chromatic Aberrations are usually a product of the lens to the best of my knowledge) and was told that the shots were take with the brand new Leica 28mm. Again, a bit shocked, as this is a brand new production lens released along with the M8.
Do I believe this? Well, I want to see the images first and hopefully I'll be able to do so at some point in the coming week.
The story is that a certain "local to me" online luminary has a production M8 body and the 28mm lens tesing out before producing a review. This particular person dropped into a favourite haunt of mine and one thing led to another and these comments (re: purple fringing and moire) were made by my source and NOT the online luminary. One thing my source said to me was, when I asked what the images reminded him of, "they look like they came from a Kodak 14N . . . " That was what really disappointed me.
That being said, as I have stated, I have not seen the images so take what I have typed for what it may or may not be worth. I would be extremely disappointed if the M8 exhibited such behaviour as Leica is banking a lot on this imho.
Dave
Gid
Well-known
I hope you are wearing your asbestos suit 
The fire department are on standby
BTW the Kodak 14N was rated very highly up to ISO100. DR and colour rendition were regarded as very good indeed, so if Leica have got as good as the Kodak's image quality and improved the noise, that can't be bad.
The fire department are on standby BTW the Kodak 14N was rated very highly up to ISO100. DR and colour rendition were regarded as very good indeed, so if Leica have got as good as the Kodak's image quality and improved the noise, that can't be bad.
Olsen
Well-known
- I don't believe this. But. What we shall be prepared for is that Canon's EOS 400D delivers files that are just as good. Even better in some respect, as low noice on high ISO.
icebear
Veteran
Close up
Close up
Hi Dave,
I’ve seen an approx. 16x20’’ printout made on an Epson 2400 from a picture taken with a M8 [firmware vers. 1.06] with the latest 50 Summilux. The guy who took the picture and printing was in a hurry getting it done and not an expert in the digital workflow according to himself and obviously no one can claim to be an expert with the M8 yet. And this picture looked pretty darn good to me. Looking very closely you could detect slight moire in red areas of the US flag. I have no doubt that this could still be optimized using different settings. I have no info on the ISO rating the picture was taken at. Let’s wait and see for some more official work shot with the M8. I would not give too much about obviously negative comments at this stage.
Close up
Hi Dave,
I’ve seen an approx. 16x20’’ printout made on an Epson 2400 from a picture taken with a M8 [firmware vers. 1.06] with the latest 50 Summilux. The guy who took the picture and printing was in a hurry getting it done and not an expert in the digital workflow according to himself and obviously no one can claim to be an expert with the M8 yet. And this picture looked pretty darn good to me. Looking very closely you could detect slight moire in red areas of the US flag. I have no doubt that this could still be optimized using different settings. I have no info on the ISO rating the picture was taken at. Let’s wait and see for some more official work shot with the M8. I would not give too much about obviously negative comments at this stage.
jaapv
RFF Sponsoring Member.
I always have problems with " an unnamed friend who is supposed to be an expert says..." type of reviews. I tend to believe guys like Sean Reid, Phil Askey and others...
boilerdoc2
Well-known
Say hi to Michael R for me!
Steve
Steve
thurows
Established
Morié becomes more apparant when you shoot something with a small distinct pattern like a herringbone suit. There use to be a bunch of filters around in the late 90s for use on cameras that didn't have an AA filter. Maybe they'll pop their heads up now that the M8 is out.
dcsang
Canadian & Not A Dentist
boilerdoc2 said:
Shhhhh...
According to Jaap ... this guy is unnamed
Dave
jaapv
RFF Sponsoring Member.
No asbestos suit needed as far as I'm concerned If it is M.R., I'm sure he'll write so in his review....
To come back to the original post: We did see purple fringing on one shot on this forum already. It was hardly surprising: The lens was a Noctilux wide open, the situation was OOF 100% contrast. No surprise to see a purple fringe then. The Noctilux was uncoded, and superb lens though it is, no one would call the old dear apochromatic...
As for Moire: there is a tradeoff between detail rendering and anti-aliasing. The balance struck by Leica seems to give extremely good pictorial rendering. However, I have one DNG file where I can see aliasing in the beard of the model. I think it was taken using beta firmware and I opened it in Lightroom. It may well be that the combination of the final firmware and Capture One will surpress the moire a bit more. Having said that, I really had to look for the aliasing.
If it is M.R., I'm sure he'll write so in his review....To come back to the original post: We did see purple fringing on one shot on this forum already. It was hardly surprising: The lens was a Noctilux wide open, the situation was OOF 100% contrast. No surprise to see a purple fringe then. The Noctilux was uncoded, and superb lens though it is, no one would call the old dear apochromatic...
As for Moire: there is a tradeoff between detail rendering and anti-aliasing. The balance struck by Leica seems to give extremely good pictorial rendering. However, I have one DNG file where I can see aliasing in the beard of the model. I think it was taken using beta firmware and I opened it in Lightroom. It may well be that the combination of the final firmware and Capture One will surpress the moire a bit more. Having said that, I really had to look for the aliasing.
Last edited:
John Camp
Well-known
Outback Photo has posted an M8 downloadable 10mp RAW file of one of their common test subjects here:
http://www.outbackphoto.com/reviews/...8_review.html#
If you greatly enlarge the shot, you will find purple fringing on brilliantly lit metallic surfaces (look at the front wheels on the car on the right, or the gas cap on the car on the left and some other possible artifacts on the right side letters of the sign on the building (on the B.) I had mess around with the file for...uh, let me think here...about 14 seconds in Lightroom to get rid of it, at any size that I would actually print.
As for moire, I can't find any in this photo, but since there's no physical filter in front of the sensor, I would be surprised if there isn't any in photos with more severe tight patterns (like tweed).
This is a good shot to play with -- brightly lit outside, with a variety of forms and colors. But since this is the 39th shot taken with the camera, I'm not fully confident that the shooters are yet completely in control of the camera (although Uwe has a lot of experience will all kinds of cameras, and his reviews and commentaries are excellent.) I'm also not fully confident that this is the latest firmware, since the file name on the thing is 0610R8, unless that R8 is a complete coincidence.
JC
http://www.outbackphoto.com/reviews/...8_review.html#
If you greatly enlarge the shot, you will find purple fringing on brilliantly lit metallic surfaces (look at the front wheels on the car on the right, or the gas cap on the car on the left and some other possible artifacts on the right side letters of the sign on the building (on the B.) I had mess around with the file for...uh, let me think here...about 14 seconds in Lightroom to get rid of it, at any size that I would actually print.
As for moire, I can't find any in this photo, but since there's no physical filter in front of the sensor, I would be surprised if there isn't any in photos with more severe tight patterns (like tweed).
This is a good shot to play with -- brightly lit outside, with a variety of forms and colors. But since this is the 39th shot taken with the camera, I'm not fully confident that the shooters are yet completely in control of the camera (although Uwe has a lot of experience will all kinds of cameras, and his reviews and commentaries are excellent.) I'm also not fully confident that this is the latest firmware, since the file name on the thing is 0610R8, unless that R8 is a complete coincidence.
JC
ywenz
Veteran
From the DNG sample, it is obvious that the M8 has some serious purple fringing issues. The lack of the AA filter means very sharp details, as evident, but also jagged lines that are not perfrectly horizontal or vertical(See the chrome trimming on the Accord) You can compare 100% crops from my 5 year old D30, my cheap Canon A70 P&S and the more modern 20D:
M8:
The Canon DSLRs shows NO purple fringe in areas of high contrast (blown out whites and dark backgrounds) The D30's rendering of the highlights are spectacular! What a feat for circa 2000 technology!
D30:
20D:
The little P&S shows purple fringe, which is acceptable for what it is.
A70:
I'm guessing these issues with the M8 are sensor related, but hopefully, they can be resolved with new hacks at the firmware.
M8:
The Canon DSLRs shows NO purple fringe in areas of high contrast (blown out whites and dark backgrounds) The D30's rendering of the highlights are spectacular! What a feat for circa 2000 technology!
D30:
20D:
The little P&S shows purple fringe, which is acceptable for what it is.
A70:
I'm guessing these issues with the M8 are sensor related, but hopefully, they can be resolved with new hacks at the firmware.
Last edited:
Bob Ross
Well-known
Hi John,John Camp said:
While "tweed" gets mentioned for moire potential, I think we will see it first on bridge railings, fences and architecture details. If you are aware that there is potential for moire in your shot, you can often alter the angle slightly, to avoid it. I supposed that could be called side stepping the issue
......or maybe, side step bracketing Bob
M
Magnus
Guest
hey... it' a digital camera ..
purple haze, orange fringes red whatevers are part of the game ...
purple haze, orange fringes red whatevers are part of the game ...
Gabriel M.A.
My Red Dot Glows For You
Hear hear.jaapv said:
I know somebody who talked to the sister of a coworker of the in-law of the receptionist of the uncle who has personally shaken the hand of the grandfather of the driver of Charles de Gaulle's hairdresser. She said color film was all the rage.
Very reliable source.
Olsen
Well-known
Moire is a hell of a problem on 1Ds II too. If you shoot cloth patterns of certain types the groom looks like a vocalist of a dance band. I am sure it is not solved on the M8 either. Hasselblad claims you have to have 39 million pixels to get rid of it. Moire is a curse of the digital age.
I am looking forward to see how hardware and software fixes the hard vignetting, AC and worse - the soft corners, that might be expected of a digital camera with such high angle light rays hitting the sensor. Vignetting can be easily fixed. A bit more difficult is AC. Soft corners; impossible. Resolution that is not there can not be added.
That said, I think Leica did a tremendous good job on the digital R. It is one of the very best pro D-SLRs that deserves success.
I am looking forward to see how hardware and software fixes the hard vignetting, AC and worse - the soft corners, that might be expected of a digital camera with such high angle light rays hitting the sensor. Vignetting can be easily fixed. A bit more difficult is AC. Soft corners; impossible. Resolution that is not there can not be added.
That said, I think Leica did a tremendous good job on the digital R. It is one of the very best pro D-SLRs that deserves success.
M
Magnus
Guest
Olsen,
thank god digital photography has supplied us with the means to increase saturation right up untill the amounts that please us ...
thank god digital photography has supplied us with the means to increase saturation right up untill the amounts that please us ...
M
Magnus
Guest
Originally Posted by jaapv
I always have problems with " an unnamed friend who is supposed to be an expert says..." type of reviews. I tend to believe guys like Sean Reid, Phil Askey and others...
yeah me too ... the word of Rumsfeld, GWBush and the other "established" are water on my tongue for me ...
I mean, really how could they be wrong, millions of people listen to them... and ahhh, I also believe everything I read in the newspapers ....
I always have problems with " an unnamed friend who is supposed to be an expert says..." type of reviews. I tend to believe guys like Sean Reid, Phil Askey and others...
yeah me too ... the word of Rumsfeld, GWBush and the other "established" are water on my tongue for me ...
I mean, really how could they be wrong, millions of people listen to them... and ahhh, I also believe everything I read in the newspapers ....
Bob Ross
Well-known
Hi Ywenz,ywenz said:
I think your "serious" adjective is probably an exaggeration in keeping with your usual style
Real purple fringing has been traced to Lateral chromatic aberration in lenses and is defined: "Variation of image size according to color. Color fringes are created around images of white objects near the limit of the lens field. An off-axis fault unaffected by stopping down."
Chromatic aberration is different and caused by different colors focusing at different planes in non-Apo lenses. It is reduced by stopping down. I suppose the question to be asked, since Leica optics are very well corrected, are the off-set micro lenses contributing in some way.
The lack of an AA filter simply lets the sensor define the aberrations better along with all the fine detail, for which everybody seriouly hungers. This sort of thing shows up in film, too, especially if you look as close as you do. The jagged edges are probably from the processing routine, rather than the Bayer nature of the sensor.
Keep up the good work. We need a serious peeper around here
Bob
JohnL
Very confused
I have three potential concerns about the M8, two of which are raised in this thread:
1) Purple fringing: I also saw the Noctilux shot and it looked to me (hard to judge from the screen) like it was a hard edge of a blown highlight (white cup) against a dark background. This *could* be caused by CA but I rather suspect it might just be pixel overflow (aka "bloom"). We need to see tests of situations like this with exposure adjusted to avoid the blown highlights. I'm sure we will before long.
2) Moiré: I have the impression this will be an issue but rarely, and that when it occurs we'll be able to handle it in PP.
3) The alleged fact that I've seen mentioned both here and on LUF that the DNG files only contain 8 bits of data per pixel (despite the fact that Leica specs mention 16 bits quite specifically). Apparently Leica somehow is getting more than 8 bits of data per color channel from that, perhaps through keeping all the information (not rounding the results) from the demozaicing algorithm. In one sense this seems to go against the basics of information theory, but if it works in practise then great! There is additional information in the algorithm itself, of course, but whether there is enough to avoid posterization in PP only time will tell.
... these are the issues that I'm trying to find out more about.
1) Purple fringing: I also saw the Noctilux shot and it looked to me (hard to judge from the screen) like it was a hard edge of a blown highlight (white cup) against a dark background. This *could* be caused by CA but I rather suspect it might just be pixel overflow (aka "bloom"). We need to see tests of situations like this with exposure adjusted to avoid the blown highlights. I'm sure we will before long.
2) Moiré: I have the impression this will be an issue but rarely, and that when it occurs we'll be able to handle it in PP.
3) The alleged fact that I've seen mentioned both here and on LUF that the DNG files only contain 8 bits of data per pixel (despite the fact that Leica specs mention 16 bits quite specifically). Apparently Leica somehow is getting more than 8 bits of data per color channel from that, perhaps through keeping all the information (not rounding the results) from the demozaicing algorithm. In one sense this seems to go against the basics of information theory, but if it works in practise then great! There is additional information in the algorithm itself, of course, but whether there is enough to avoid posterization in PP only time will tell.
... these are the issues that I'm trying to find out more about.
Last edited:
jaapv
RFF Sponsoring Member.
John Camp said:
This is the thing that stuns me each time I play around with these M8 files - the speed and ease with which they post-process.
Starting with scanned film -I lay no claim to great expertise, but I think I'm no beginner either- let's just say life is too short... Before they are halfway decent I need about every trick in the book.
Then my Canod 10D, which taught me the basics:Convert from Raw after slamming all the automatic adjustments to zero, do levels and curves, contrast,finetune colour balance, complicated sharpening, 10 mins per file, then Digilux2, which taught me the ins and outs of noise reduction, but was better:5 minutes. And now the M8: Open in lightroom, a few clickes, export to PS, crop, a bit of Focal Blade, done in 60 seconds, perfect file ready for print. I love it!
Last edited: