Leica M Monochrom DNG raw question

[koferk]

Advertisement (gone after registration)

hi, i hope Im in the right place for this question.

 

say i have a Leica M Monochrom DNG raw file, then i convert it to a TIF file using photoshop.

 

will there be any difference between the original DNG and the converted TIF (besides file size) ??

 

since there is no demosaicing, i cant see any difference. is that true ?

 

 

thanks in advance,

koferk.

[mydarkroom]

I find the TIFF files not so great to recover shadows. Other adjustments also didn't look quite the same.

 

And you are right, the TIFF file is 2-3x larger than the DNG.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[koferk]

Thanks for the quick reply!

 

Did you check it or you just assuming ? can you explain it ?

I can't see why/how the monochrome DNG and converted TIFF are any difference..

 

I have an idea -

Maybe the DNG to TIFF is doing some sort of down-sampaling from the true sensor resolution ?

[jaapv]

Yes - there is a difference - the DNG file contains the raw data and will not render an image you can see (except for the embedded small jpg). It must be converted to for instance a tiff file or similar in order to be visible in the first place.

[koferk]

I know, this all true for RGB raw sensor.

I found this for example -

http://www.bobatkins.com/photography/digital/raw-flowchart2.gif

but this refers to the an RGB raw file.

 

i still cant see why it would matter if you play with the white balance on the DNG file or on a TIFF file (ONLY with the Leica M Monochrom raw files).

 

if someone can explain this to me I'll be super glad

 

Thanks again

[koferk]

Interesting, I also found this -

Leica M-Monochrom Hands-on Preview: Digital Photography Review

"There are drawbacks, of course - the 'headroom' found in Raw files comes mainly from the fact that bright regions have usually only over-exposed one of the three color channels, with usable data still available for the other two channels. With a true mono sensor, any overexposure is absolute - once the channel has clipped to white, there's no chance of recovery. Equally, anyone who has got used to producing mono images by converting color images, with all the selective color mixing that brings, will have to get used to pulling the correct color filter out of their camera bag at the point of capture."

 

this means that its not better at recovering shadows as user mydarkroom wrote above. so I'm still not sure if there is a difference between the original monochrome DNG and a converted TIFF.

[250swb]

Advertisement (gone after registration)

I find the TIFF files not so great to recover shadows..

 

As Jaap has said, a dng file is in the ether, it doesn't exist, it is a range of variables waiting to be 'set'. On the principle of 'rubbish in, rubbish out' if you haven't got the shadows sorted out during the processing of the dng file then the 'real' file, the TIFF, won't have the shadows either.

 

The 16 bit RGB Tiff file is the most accurate way to represent your MM .dng images. It also has the most flexibility for manipulation in post processing, but it can't invent what isn't there, so shadows and highlights should be set correctly in ACR to begin with.

 

Steve

[koferk]

i understand what you are saying.

again, this all true for rgb sensor data - where you have a mosaic bayer filter and so you can get more detail by using other channels in acr.

 

but in this specific case, at least by my understanding, the raw dng might be used as a tiff container, but there is no difference. they are both contain only one channel of data. that is - one number for each pixel and NOT RGB matrix calculation per pixel. say 0 = black and 100 = white 50=grey.

 

they are both the same, except file size and genuininty - the raw file is an evidence it came from the camera if that worth anything..

 

still waiting for a good explanation of this..

thanks for the replies

[koferk]

im referring the quality available, to the flexibility.

 

ok, one is a ready to print/view file (tif) and one is not (dng mono).

 

but this is not my question.

im asking if there is any difference in quality, flexibility, adjustments, and so on..

[jaapv]

There is indeed no Bayer interpolation on the raw data of the Monochrome, for the simple reason that there is no Bayer filter, but that does not change the fact that DNG is not an image file format, but a file containing data to be converted to a file that can be read as an image. As Steve says, 16 bit TIFF is a format that will give you the full content of the DNG.

As for adjusting etc, you cannot manipulate a DNG file until it has been converted into TIFF.

There is a difference between Lightroom and Photoshop here, btw. Photoshop will convert the raw data in ACR (Adobe Camera Raw) and export them to edit them, whereas Lightroom will simulate all your edits in preview JPGs, store them in a sidecar file and apply them on exporting the file using the same underlying software as Photoshop.

Edited May 5, 2013 by jaapv

[250swb]

they are both the same, except file size and genuininty - the raw file is an evidence it came from the camera if that worth anything..

 

 

A TIFF file is evidence it came from the camera, there is no higher level of purity in a .dng file which is just a collection of data waiting for you to show a preference for how it should look.

 

Any default 'look' of the image as it comes out of the camera in the software generated thumbnail is made by a technician who wrote the software, it is not a perfect representation of reality, or a demonstration of how special the sensor is, or how the individual photographer see's. This is why Leica bundle Silver Efex with the camera, and this is why you need to make your .dng file into an TIFF RGB file, because Silver Efex does not work in .dng or Greyscale.

 

In addition if you sent an image off for publication they would want it in TIFF (or JPEG)format and an RGB colour space whether or not all the channels were variations of grey. For instance, if you sent a publisher or other third party like an online printing company a .dng file they would open it in whatever software they were using. So it could very well look totally different to them than it would to you on your computer, given .dng is only a collection of data waiting to be organised. If the image is in TIFF format, in the required colour space (even for monochrome), what you see and send them is exactly what they see on their screen (assuming everybody is working with calibrated monitors). If instead of posting an image in LUF's Photo Forum as a JPEG, and posted a link to your .dng file, and people found the time to open it, many people will see something different to you if they use a different .dng processor. So there is no purity or value in showing it came from the camera in a .dng file, no genuinity, but instead it is verging on chaos.

 

So to reiterate, you adjust the initial interpretation of tones and the black and white points when converting your .dng file, but you work in post processing and treat as the final image the TIFF file. If you are from a darkroom background think of it like this, the .dng file is your negative (the clue is in the name), you make a print and decide on the paper grade and do all your dodging and burning to make the perfect image, this is your TIFF. Then when you want to distribute this 'master image' you make a copy print at a smaller size, this is the JPEG you post on camera forums.

 

Steve

Edited May 5, 2013 by 250swb

[koferk]

I know all that.

I will try to rephrase my question -

 

lets say I have a DNG-mono file and the converted tiff file.

 

now, i want to play with the white balance and export in tif.

 

which will be more flexible ? i think they both have the same flexibility. same dynamic range. same abilty to recover shadows and highlights.

 

is that true? if not, why not?

[pop]

The information contained in both files is the same. Any differences in processing would be due to the software being used. Hence, if you're intending to use Photoshop, it's a Photoshop question.

 

Monochrom files do not, BTW, have a white balance, as the white balance is a property of color images.

 

For those interested in such things: DNG is a kind of TIFF.

[koferk]

so in terms of shadows and highlight reviving they are the same as well, right ?

 

everything is eventually software independent, but if the software is smart, it shouldn't matter one bit because it is the same data, its just the defaults knobs that might change, you can always change them back

[250swb]

If you have processed the .dng file with the maximum amount of detail in the highlights and the shadows (no clipping) the resulting TIFF will be the same, it is a lossless format, so can be edited and resaved without losing any information, unlike a JPEG.

 

Steve

[sandymc]

In truth there is very little difference between a Monochrom DNG and a monochrome TIFF. However, the DNG does have a potential advantage, which that it keeps data that is below black level, "blacker than black". In a TIFF, this information is cut off. A raw developer may be able to use that information to do a better job of shadow recovery.

 

Sandy

[250swb]

Is there something in the water?

 

If it were possible to have anything 'blacker than black' it would be wasted information, just as any clipped blacks or whites are wasted information. A TIFF file does not 'cut off' any of the information that is saved.

 

Steve

[sandymc]

Is there something in the water?

 

If it were possible to have anything 'blacker than black' it would be wasted information, just as any clipped blacks or whites are wasted information. A TIFF file does not 'cut off' any of the information that is saved.

 

Steve

 

Water? Not water.....

 

But no, not necessarily wasted information. Look at it like this: Where to place the black level is a decision about where the noise floor is. But noise level is actually somewhat malleable depending on noise processing in the raw converter.

 

Sandy

[mydarkroom]

Interesting, I also found this -

Leica M-Monochrom Hands-on Preview: Digital Photography Review

"There are drawbacks, of course - the 'headroom' found in Raw files comes mainly from the fact that bright regions have usually only over-exposed one of the three color channels, with usable data still available for the other two channels. With a true mono sensor, any overexposure is absolute - once the channel has clipped to white, there's no chance of recovery. Equally, anyone who has got used to producing mono images by converting color images, with all the selective color mixing that brings, will have to get used to pulling the correct color filter out of their camera bag at the point of capture."

 

this means that its not better at recovering shadows as user mydarkroom wrote above. so I'm still not sure if there is a difference between the original monochrome DNG and a converted TIFF.

 

The info in the article is correct but it refers to the highlights. Indeed the bayer sensor allows you to recover a bit more on that side. I was talking about the shadows. The MM captures amazing details on the shadows that can be pushed in post processing (sometimes noise is introduced if you push a lot).

 

Back to your question: I have no scientific explanation and have not done extensive testing but I find the DNG and TIFF files different to manipulate. The DNG allows me to pull more details from the shadows without introducing artifact. My comparison is done the following way: Adobe Camera RAW, RAW Photo Processor 64, or Noise Ninja is used to to generate the TIFF file. Then that TIFF file is imported into Lightroom and compared to the original DNG form the camera. If this is not the comparison you wanted, please disregard all my comments.. I misunderstood your question. Ideally you want to download a sample MM DNG file and do the comparison yourself. This will allow you to compare the flexibility of each file according to your needs and usual photo manipulation.

 

Cheers

 

Raf

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[250swb]

Water? Not water.....

 

But no, not necessarily wasted information. Look at it like this: Where to place the black level is a decision about where the noise floor is. But noise level is actually somewhat malleable depending on noise processing in the raw converter.

 

Sandy

 

Thank you, I see what you mean now Sandy .

 

But in this instance the Monochrom only uses .dng or JPEG file formats, unlike some cameras that can internally generate TIFF files. So if we assume JPEG is out of the equation the user still has to take the image through a RAW/dng processor like ACR to get to a usable image, so the actual floor of the black output (and taking noise into account) will be seen and can be adjusted via the histogram, unless simply pressing 'Auto' doesn't do it all for you anyway.

 

By using Adobe's .dng converter we can of course turn the .dng straight into a TIFF without any of the normal tonal controls available in ACR. And this was indeed a workaround used by some photographers when Aperture didn't support the M (or does it still not?). But this would be the only time that a TIFF file would be compromised and where an out of camera JPEG might be the better option.

 

Steve