Maybe I shouldn't have....

[georg]

Advertisement (gone after registration)

The sensors are nearly identical, the M8-sensor is even a little bit more advanced (see technical data from Kodak).

The Firmware can cause subtle differences in controling the hardware (power/heat management...) but was also optimized in the M8 to enhance speed.

 

The D/A-converters are most likely the same, both are as much 16Bit-cameras as others, using 14Bit-converters and putting final data in a 16Bit (DMR) or 8bit (M8) log-file (which has to be converted into a 16 not 8bit Tiff).

 

When we do a comparison we need to use the same lens: the 90AA (only modern lens available for both systems)

And we have to carefully adjust speed (160ASA M8 = 100 or 200ASA DMR or something completely different?) and profiles, otherwise we don't compare cameras, we compare systems/workflows!

[isaac]

Hallo George!

The D/A-converters are most likely the same, both are as much 16Bit-cameras as others, using 14Bit-converters and putting final data in a 16Bit (DMR) or 8bit (M8) log-file (which has to be converted into a 16 not 8bit Tiff).

Very interesting! And how can you explain the fact, that in a DMR DNG ("Leistungsbeweis - Bali Tanzgruppe_de.DNG") all 16 bits are valid?

[luigi bertolotti]

Sure-

First M8, then DMR

 

Uhm... if I had to say something ... "DMR better in highlights, M8 better in shadows"... did you save the DMR as a 16 bit TIFF ? The above pics make me think that (seeing the foliage in the enlarged crop, forefront) there is really something that could be credited to this.... but I admit that confrontation is subtle...

[luigi bertolotti]

Hallo George!

 

Very interesting! And how can you explain the fact, that in a DMR DNG ("Leistungsbeweis - Bali Tanzgruppe_de.DNG") all 16 bits are valid?

 

Is what George said... DMR's DNG is a 16 bit - depth file.

[mgcd]

In the context of another thread, I shot the DMR and the M8 side by side. No contest, the DMR won hands down...

 

No kidding! Of course the DMR wins hands down. Just to be absolutely clear, the DMR isn't 14 bits reprocessed into 16 bits, it is native 16 bit A/D. There is simply no comparison and this is the main reason why I did not buy the M8 (although I wanted to...).

 

http://www.l-camera-forum.com/leica-forum/digital-forum/36403-your-favourite-dmr-pictures.html

 

Cheers,

[Overgaard]

I don't think one should underestimate the software made by Imacon. In my experience, they deal with light and colors in a very "Leica manner," meaning accuracy in tones, colors, light rays, etc.

Different than many other who concentrate on detail sharpness, often digitally enhanced. Which is why - I learned - that when you look at details enlarged, Imacon scanners, etc may look less detalied (less digital sharpness). But when you view at a distance, you notice that the images touched by Imacon has clarity, accuracy in colors and a 3D appearance. And thus look as they contain more detail.

 

And after all, pictures are viewed as pictures, not as blown up details.

 

Besides that, I see a great difference in raw converters. I use FlexColor now for DMR as that is the "native" Imacon software and make the best results.

 

That said, the M8 is not bad. Digital Leica's has a great future with new hardware, firmware upgrades and new raw software. Look at the last five years, then think five or ten years ahead.

[sean_reid]

Advertisement (gone after registration)

Hi Jaap,

 

I'm not sure if you've done this already but for this kind of comparison it is helpful to do focus bracketing (10 steps each way) with both cameras so as to rule out slight focus differences as a confounding variable. With camera's capable of these quality levels that can end up being important.

 

Cheers,

 

Sean

[earleygallery]

Andreas Kaufmann: 'You can be assured that to us it´s very clear what's going to happen with regard to the topic SLR. We´re developing a solution which we will present to the public.... This fall in Cologne you will be able to get a clue'... the new camera 'shall more than satisfy the needs of our customers'.

 

Sorry LCT but I meant that we won't see a new DSLR launched and therefore no new lenses either, unless we're in for a surprise of course!

[jaapv]

Picked the best ones in a series of four, Sean I still see differences, so your idea of a series of ten is probably better. Lack of time, though

[jaapv]

No kidding! Of course the DMR wins hands down. Just to be absolutely clear, the DMR isn't 14 bits reprocessed into 16 bits, it is native 16 bit A/D. There is simply no comparison and this is the main reason why I did not buy the M8 (although I wanted to...).

 

http://www.l-camera-forum.com/leica-forum/digital-forum/36403-your-favourite-dmr-pictures.html

 

Cheers,

 

That does not sound very convincing. We are talking about the difference between excellent and excellenter here. It is a bit like saying "I did not get a BMW 7 because the Merc S-class is more silent.. What camera did you buy then?

[rick_dykstra]

... but I meant that we won't see a new DSLR launched and therefore no new lenses either, unless we're in for a surprise of course!

 

Hi James. I hope we're in for a suprise too. May I ask, what is the basis for your thoughts that we wont see a new DSLR (or announcement of impending release)?

 

My kids sometimes prepare themselves for disappointment by taking an artificially pessimistic outlook, just in case. Have we been bad?

[earleygallery]

From what Dr Kaufmann has said in recent interviews, and some of our members were told at the recent Leica day, from Stefan Daniel. Of course it could be a decoy but probably not.

 

http://www.l-camera-forum.com/leica-forum/customer-forum/56309-infos-forum-meeting-2008-a.html

[lct]

According to Kaufmann's and Daniel's statements, we will have a 'clue' of the forthcoming DSLR at Photokina but no new R lenses there. The clue could be a mock up or anything else but, at least, we should know if our lenses will become obsolete as suggested by the rumors of a new mount widely spread but this otherwise excellent forum.

[georg]

I still haven't seen a valid 1:1 comparison! In fact, the pictures posted here don't look to well at all and I don't see where the DMR wins hands down!?

 

The discussion 16 vs 14bit is quite old. In fact, nobody seemed to know but as some specialists pointed out that even the MFDB have 14Bit-converters (claimed as 16bit-systems), because these are the highest quality converters available and the dynamic range of today's sensors is simple not high enough to judge more than that.

But I've spoken with a leica-engineer and he told me that the only difference between M8/DMR is the fact that they use 16bit linear recording for the DMR and 8bit log recording for the M8 - no different conversion technology at all, that's why early prototypes of the M8 created 20MB-16Bit-linear-files just like the DMR. They decided to go with 8bit-log because it's more efficient (more colour resolution where needed instead of the same resolution over the entire spectrum) and even in 1:1 comparisons they couldn't tell the difference - and that's my experience since 2 years M8, too.

10bit log-files are the industrial standard in film-industry, too and they scan film-stock with 15 apertures dynamic range! The Arriscan (500.000$ scanner) for example takes two pictures of the negative to capture the whole dynamic range, creating a true 16bit-lin-file and than it's usually converted to 10bit-log!

 

The profiles for the DMR were better, creating better colours "out of the box" while the M8-profiles had to be changed (because if IR-filters) and still causing some minor trouble (red skin tones in C1). But that's workflow, that's software and not the camera itself!

[Jamie Roberts]

Jaap--in the examples you're showing the exposure // contrast // red saturation is so different that it's kind of an impossible comparison as posted....

 

Whatever the cause--because I don't know how you processed the files--it's the colour and contrast difference that's being shown here. I don't see any difference in tonality from the samples that couldn't be fixed in a RAW tweak or a profile because the light isn't particularly tricky here...

 

It is true, though, that "out of the box" the DMR gives better base colour than the M8, especially with FlexColor (as Georg said above).

 

Where I see the biggest differences myself is usually in the upper quarter-tone (midtone to highlights) in harsh light or backlight (at low ISOs), and in skintones with the Imacon profile in C1. The DMR just rocks in those conditions, and is usually a wee bit better there overall and with highlight transitions. But again, both M8 and DMR look like they belong together when processed in a similar manner and printed, IMO. There's a much bigger gap between either and, say, the 5d (I process all three together all the time, and I've given up trying to get the 5d files to match the M8 and DMR until I tweak each one for print, and then I can at least get the colour in the same ballpark).

[jaapv]

For comparison I developed the both at the default settings in ACR. Of course I can get them close in PS I was hoping some others would chip in with comparing shots.....

[hoppyman]

Different sensors, different firmware, 14 bit v 12 bit.

Steve, I understood that that both sensors are 16 bit. The M8 A/D convertor does apparently discard the lowest 2 bits due to noise contamination. That doesn't make it 14 bit capture though! I don't know about the DMR. Is the file size difference between the two 10MP sensors not due to the compression method used in the M8? Not meaning to start a tech argument.

I have seen Raw files from the DMR and they are certainly superb.

[mhoersch]

I don't quite see what you want to prove with shots at different ISO settings, different lens designs with different focal lengths and, obviously, different exposures - except that they look different. What is this supposed to say about file quality of DMR versus M8?

 

Sorry, but I don't get your point.

[jaapv]

Just a discussion point -that is what forumns are for - nothing scientific, I make no claim.Btw: How are you going to get the same Iso settings between the M8 and the DMR?

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[mitchell]

Jaap,

 

Thanks for this comparison. What I see in it is that they are remarkably close.

 

I have to agree with Jamie that the differences seem to come mostly from exposure differences. In both cases the M8 looks slightly overexposed compared to the DMR. This seems especially apparent in the second shot. Exposure differences may be unavoidable given their different base ISOs.

 

Best,

 

Mitchell