Jump to content

M8 Review at B&H


Bill W

Recommended Posts

  • Replies 74
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

It's a crude fix for slow internal processing. It's also 1. not as intelligent as Nikon's compressed NEF (that yields higher effective bit depth and can be switched off in higher end models too) and 2. not comparable to Canon CRW/CR2 that uses lossless compression. Trying to explain that as some especially clever and new way to store images is simply ridiculous.

I'm glad that one more person who really understands the implications speaks out ... I have to thank another person who initially called me a troll, it was exactly him who drove me to spend hours (if not days) on those specs, and source codes - now when you know that I'm not a computer person. LOL

Link to post
Share on other sites

I think trying to compare across brands (as I think Alan was suggesting above) would just bring more confusion. At that point we would be looking at more than just the M8 storage algorithm. We'd see the results of sensor size, lens design, sample variation etc. Nothing wrong with that if someone just wants to say "Mine is better than yours," but it gets us into another sphere.

 

Where we started, I think, is what we still want to know: Everyone agrees that the M8 turns out good files with its unique DNG. We want to know if an ordinary, uncompressed DNG would do better. The only way to know that is if Leica were to decide to make that option available. Since they haven't done so, they clearly feel it isn't necessary; and therefore they probably won't give us the choice, no matter how many of us would like to see it offered. (I don't think we want another menu choice if it doesn't make a real distinction.)

 

In other words, it really doesn't matter if the 5D outperforms the M8 in some situations, or if the M8 tops the 5D in others. We already know that the range of applicaton of a rangefinder isn't as broad as that of a reflex, and that a reflex isn't as suited to some jobs as a rangefinder.

 

We also know that many reviewers rank the M8's image quality as good as or better than other offerings.

 

The only thing we don't know is whether the M8 is limited by its compression scheme.

 

But keep in mind, we like Leica lenses; and all lens designs are compromises. We might want to admit that we like the M8's images as they are, and that it may be worth keeping this compression to avoid other compromises.

 

--HC

Link to post
Share on other sites

Guest guy_mancuso

Advertisement (gone after registration)

Simon hope you have some money ready.:D

 

I'll take a 90mm macro and adapter that will be fine.

 

 

Leica did not see any visual difference between there prototype M8 16 bit to final 8 bit scheme.:D

Link to post
Share on other sites

But you've missed the point, Guy ... as "enjott" has pointed out, it's about the headroom for adjustment, not direct visual difference.

 

What about this, you go suggest Leica provide several 16-bit enabled prototypes to folks who are willing to "fight" for my prize, I'm sure we can find folks who are willing to host and join such a game ... geez, there're people starting a openraw website for absolutely nothing :D ... then I'll sponsor a M8. LOL

Link to post
Share on other sites

Guest guy_mancuso

I do agree there is head room in 16 bit and i think it would be a nice idea to see it as a option for M8 users. I would certainly use it, I do think though we maybe leaving something out regardless of 16 or 8 bit and that is the color gamut of the M8 and from some charts i seen earlier it is a huge color space that there using , so there certainly is a lot play in the files today. But I'm all for it and you guys have my vote and support for 16 bit as a option. i will let the color gods fight for that M8 .:) I will take myself out of the running, they will bury me with there math and science. ROTFLMAO

Link to post
Share on other sites

Amazing how this subject comes up again and again, along with lots of speculation presented as fact ("no camera can output true 16 bit color" - plain wrong).

 

The M8 look up table essentially causes a gamma encoding of the image. I modified dcraw so it will output the image without the LUT applied, and such a conversion looks indeed similar to jpegs straight from the camera.

 

Because most bits are used to represent the brightest tones, most of what you lose is in the "artificial film shoulder" that, when working in 12+ bits, allows us to get back details from what appears as perfect white when represented by 8 bits per color. What's probably also affected is the ability to completely change white balance, though I haven't tested that.

This is never visible without heavy image manipulation so it's not the end of the world like that reviewer seemed to think, but there's not the slightest doubt it's unworthy of a camera of that price. It's a crude fix for slow internal processing. It's also 1. not as intelligent as Nikon's compressed NEF (that yields higher effective bit depth and can be switched off in higher end models too) and 2. not comparable to Canon CRW/CR2 that uses lossless compression. Trying to explain that as some especially clever and new way to store images is simply ridiculous.

No one ever claimed the difference between 8 bits and higher color depth is directly visible (although there are worst case scenarios where this is indeed possible), the important property is headroom for editing. These are really only "half raws".

 

Writing images in the 12-bits JPEG mode that no other camera maker has yet used at least would have been something new...

 

Brilliant compression solution or crude fix... The experts have lost the laymen:( I'll go back to taking and printing my photographs and leave the bits,bayers and pixels for those that like to consume them....:rolleyes:

Link to post
Share on other sites

This is often repeated in online forums, and it's grossly wrong. Just about everyone can make out a doubling in luminance (hence the choice of gamma 2 for NTSC and its prevalence in RGB), so the number of bits required is log2 of the luminance range of the output device. A flat screen with a 2000:1 contrast (luminance range) for instance will require 11 bits for smooth gradation. (snip)

Thanks for this correction. Clearly better screens with more dynamic range need more steps (=bits) to maintain smooth gradation. Of course, this just increases my concern about the shortage of bits in the DNG files, especially if we plan to work on them in PP. This is my main worry, because I regard the final product to be the print, not the view on the screen.

Link to post
Share on other sites

We want to know if an ordinary, uncompressed DNG would do better. The only way to know that is if Leica were to decide to make that option available.

No, we already know this. First, it's obvious from theory alone. Second, you can just use any camera with "real" raw capture and experiment. Third, the reviews at dpreview actually tell you how much dynamic range you gain through raw when they review SLRs. The review for the 5D demonstrates this nicely.

 

Here's a quick and not very good example, but I took this just a few hours ago.. part of a panorama shot:

http://leitseite.net/photo/tests/350d_clipping.jpg

 

Left is the jpg from the camera, right an unadjusted raw conversion with EV -2 exposure compensation. The example isn't that good because most of the clipping occured abruptly and the raw image is clipped too.

But there are a few parts where the jpeg is white and the raw clearly shows detail. For example at the copper roofs to the left of the clock tower. Or the corner of the roof at the lower right of the image.

 

The additional dynamic range surely isn't huge, but still an advantage. And these are only 12-bit images from an ordinary Canon 350D.

 

Since the M8 could already write full raw data with earlier firmware revisions, they should simply bring it back and leave it to the user if to use compression or not. For me, the slow buffer writes wouldn't be a problem at all, especially not with such a camera.

The digital Leica products seem to lack an implementation of lossless compression in hardware, unlike the special processors certain Japanese companies use. The DMG didn't compress its raw files either. An embedded general purpose CPU is very likely too slow to do this quickly enough.

 

Something is very odd about the camera being able to write 8 bits per pixel quickly, but not 14 even in twice the time (hope I got this right, how many shots per second can the M8 take in DNG?). And no, today's embedded CPUs in such devices are not 8 bits as suggested in another thread. They're 16 at the very least, more likely 32 and can handle a 16 bit value without a huge number of clock cycles. 8 bits is nowadays what your washing machine or toaster might use.

Link to post
Share on other sites

Guest sirvine

You're comparing JPGs to RAW from a 350D to prove that the M8 compression scheme is losing highlights? I really have to question the fundamental logic there.

Link to post
Share on other sites

You're comparing JPGs to RAW from a 350D to prove that the M8 compression scheme is losing highlights? I really have to question the fundamental logic there.

 

 

You just have to shoot the same image with an M8 and let's say a Canon DSLR using raw. Then you could see if one does better than the other in this regard.

Link to post
Share on other sites

You're comparing JPGs to RAW from a 350D to prove that the M8 compression scheme is losing highlights? I really have to question the fundamental logic there.

There is no fundamental difference in this aspect between a Kodak KAF100xx sensor and a Canon 704W. Or between any digital still camera sensors for that matter, from a 10-bits P&S to a large format module for aerial reconnaissance.

Just like 8 bits in GIMP stay 8 bits in Photoshop or Bibble Pro. The math always stays the same: 8 bits gamma encoded vs. 14 bits linear, the latter of which means half of the recordable values get wasted on the upper end of the dynamic range [edit: this is the reason for the "expose to the right" recommendation.]

 

The difference should be stronger with the M8's sensor. It's 14 instead of 12 bits and the photosites are larger (more "useful" bits).

 

[edit: of course I meant DMR a few posts above, not DMG..]

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...