What's new
Medium Format Forum

Register a free account now!

If you are registered, you get access to the members only section, can participate in the buy & sell second hand forum and last but not least you can reserve your preferred username before someone else takes it.

H3D

My "motive" for wanting to discuss the 16 bit issue had nothing to do with the nefarious claims Q.G. has so delusionally stated. I hope this is quite clear.

The reason this interested me, is, being a professional in the digital imaging industry, and having designed with a number of CCD imaging sensors, I have yet to see a commercial imaging sensor that can give 16 good bits, and this claim did not make any sense to me. I wanted to either 1) find out if there was such a sensor, or 2) point out the potential of this being misinformation about the capabilities of the sensor.

I believe "The 16 Bit Mystery" has now been solved. If anyone gets any feedback/information from Hasselblad as to why they claim 16 bits, I'd appreciate hearing about it.

Regards,

Austin
 
Hi Franc,

> Although you cannot distinguish that many different colors you > can still see a difference.

I disagree, and have seen no evidence that shows that. In fact, there is a lot of evidence that shows that you can't see any difference, except in very limited circumstances (see below). There is also no commercial digital camera that can actually capture 16 bits, so where would you get this image data from?

As I've said, you do need to process the lower bit depth data in a higher bit depth realm, or you get combing...which you can see once you re-setpoint the data...it shows up typically as posterization. THAT is probably what you are seeing, but it is not because the capture was done in 16 bits, it is because the color manipulation was done in 16 bits. They are not the same.

> To your point about usable bits, I would tend to disagree, with 16 bit depth > of color all 16 bit are relevent and in a properly exposed photograph you have > every pixel recording relevent data.

Clearly not true. The 39M imaging sensor used in the Hasselblad back is only 12 bits (confirmed by Kodak BTW). If you know of a CCD or CMOS commercially available sensor that can actually give 16 bits, I'd love to see the spec sheet.

> The camera manufacturers have made it easy > to make sure all your data is relevent and the device we have is the clipping > indicator. If 2 or more bits of the 8,12 or 16 bits at the bright end of the > scale are the same, the clipping indicator will show you on your preview screen.

I haven't seen a "clipping indicator", but a histogram is simply showing that your image data values are within the range the camera can capture, no matter how many bits the sensor and/or A/D have. It's no indication of how many bits your camera is/actually uses though.

> But my basic point is this, If you properly expose your image you will get > TRUE 16 bit depth and you will see the difference in the image.

But your point is incorrect. The facts are in the imaging sensor's data sheets, if you know how to read them. Here is what to look for. The data sheet will either say that the sensor has a dynamic range of 65536:1 or a dynamic range of 96.32dB.

Dan Margulis did quite a bit of work some years ago, and showed that 8 bits/color could not be distinguished from higher color bit depth (except in very very limited circumstances). And, having performed the same experiments over many years of digital imaging, this is correct in my experience.

And in many others as well:

http://www.digitaloutput.net/content/ContentCT.asp?P=350

This is good reading for someone wanting to see facts about the 8 bit vs 16 bit argument.

Regards,

Austin
 
Many thanks Franc - a terrific explanation - I have learnt a lot! And now Jurgen has added motivation to polish the pixels on his new Hassy back and blow us away with his wonderful images!
happy.gif


So, I suppose (correct me if I am wrong please) this bit depth stuff and number of colours is similar to computer screens and video cards as well as scanners. Maybe it is the low bit depth stuff that makes some digital images look "plastic" to me - the short colour range / tonality abruptly changes to the next colour / shade in the image (I hope I make myself clear here). Likewise the super poster I saw on the weekend made from a Hasselblad H3D was unquestionably "real" and the scene so lifelike I thought I was there.

Touché Colin! And Gilbert said: "Better yet, your M7 and a couple of rolls of film and nice sunrise stroll." So Sooooooo true. Since being a passionate photographer, I have learnt to treasure beautiful light!

Wow Marc the images are fantastic in every way. Those tulips are sensational.

Thanks Marc for your comments on 16bit printing - all you said makes much sense. And I agree about taking advice from well proven sources. When you sent me back those sensational scans of my trannies and the s&le prints done at 360dpi, I made a mental note that there must be a real reason to print at that resolution and in future I will request that when I have quality large prints done.

Marc said: "Maybe we should have a "credentials" thread and get it over with once and for all : -) " Not a bad idea...... oooops, but I don't think I have any!!
happy.gif
happy.gif
.

But on a more serious note, actually our bio section should enable some credentials info, that way when we see a post that is important to us we can consider that in light of the person's credentials / background - sort of put the comments in perspective. Of course it's not a string of qualifications and letters after the name so much as hands on actual experience I suppose.
 
Franc, how do you think analogue slide film compares in "bit depth" - does it produce millions or trillions of colours? Or, is it impossible to determine? If it is up there in the trillions of colours, I suppose scanning may limit the ultimate bit depth of that digital file?

Marc, from your experience using the 39mp back versus slide film, do you "see" a difference in the level of colours - does one look to have greater bit depth to the other?

And this makes me wonder (sorry for the "curve ball"): but, does the lens glass and its coatings affect the "bit depth" or number of colours available to be received on the sensor / film?
 
Marc

I envy you for the wonderful tulips . Out of two reasons .
a) The tulips we have in our garden this year are not very nice . I just don't know why .
b) The wonderful colours in your shots . I have not got the colour process with my CFV under full control yet .
This year , HASSELBLAD Germany offers the "HASSELBLAD DIGITAL SCHOOL" again , a professional course in cooperation with ADOBE and EPSON .
The topics are very extensive and will give you training for the complete workflow , from shooting to printout . The working horses are H3D39 , 503CWD , FLEXTIGHT X5 and EPSON K3 LF printers . Handling , optimizing and archiving your images using PSCS3 (DNG and HASSELBLAD 3F) formats is an other topic . Training is done in small groups in the studio and using MAC computers .
I have booked me in for the 13th of june in Munich . No booze with good bavarian beer
z04_bier01.gif
that day .

Regards Jürgen
 
G'Day all:

Jurgen - mate - I'm sure you meant to write ... "No booze with good bavarian beer that day ... until AFTER the training.." . Do you know, by the way, are they having the HOLGA digital seminar as well...

Simon - Galbo - great question for Franc about the film. I've wondered that myself. It's organic, and so I guess, repeat guess, the potential information from film is to some degree still limited by current scanner technology - sensors again?? No doubt Marc's drum is way ahead of my flat bed in being able to extract (create?) digital information. But then it all comes back down to how you get to see (or try to see) the end result - on a monitor, on paper, etc. Of all the folks on this forum, there may not be two who have the same visual acuity, and probably none with matching interpretive 'brainware'. Not that it matters, of course. At the end of the day, we like the things we like, and we are moved by the things that move us. Horses for courses.

Cheers,

Colin

PS This forum is much nicer when the passion is kept at moderate levels, don't you think.
wink.gif
 
Colin

My friend
The training starts at 10:00 and will definately last till 18:00 . And after that , I have to travel 230 km . In my wonderful bavarian station car , known as BMW . But then . . . . . I can assure you , with all the hopefully good HASSELBLAD DIGITAL information in my head , I will have a good bavarian beer .
z04_bier01.gif
cheers .
Hopefully the colour management info will still be there next day . :)
 
Juergen,

Hefeweizen I hope? Helles is my preference :cool:

Oh, by the way: the Ektachrome shots from the Potsdammer Platz nightshots in Berlin came out very nicely. The Portra look also OK, just need to do some scanning before I can post things to the forum. I'll probably never learn to translate colour negative directly in real colors in my head ;)

Wilko
 
If the passion is for photography, count me in : -)
Austin may well be completely correct in his technical explanation of bit depth. To me it just means true accuracy in reporting specs by the digital back makers ... may not be truly accurate.

The difference between 8 bit and 16 bit, and not being able to distinguish the difference does not match with my personal experience, nor that of any PhotoShop expert, or master printers' advice. Not that we all couldn't be wrong.

However, if that's true, and I am processing 16 bit files for no good reason, that is a different story. 16 bit files from a Hasselblad H3D/39 take up a huge amount of memory. I have a job on my desktop that has resulted in 35 gigs of Tiff images. Combined with other current jobs it has reduced my computer to the IQ of an idiot.
 
Hi Marc,

> Austin may well be completely correct in his technical explanation of > bit depth.

What I was trying to do was understand the claim that the Hasselblad back was 16 bits. I believe it isn't, and have substantiated that. I hadn't thought an explanation of bit depth was necessary since we have such an advanced group of users here. But, simply put bit depth (in this context) is the maximum number of discernable tones that can be represented per color (like Franc said). Really, that definition isn't arguable. I don't believe you're questioning that this is a correct definition of bit depth, but that you are questioning the discussion surrounding 8 bit vs 16 bit. At least that's my impression. Please correct me if I'm mistaken.

> The difference between 8 bit and 16 bit, and not being able to > distinguish the difference does not match with my personal experience, > nor that of any PhotoShop expert, or master printers' advice.

Actually, the reference I cited, both Dan Margulis and the author of this article:

http://www.digitaloutput.net/content/ContentCT.asp?P=350

are well known PS experts, who say there is no difference. It is interesting that no one was able to pony up the goods to Dan's challenge.

But, if you have other resources that show otherwise, I'd be more than glad to review them. This "16 bit myth" has been around for a while (as well as many other myths surrounding digital imaging), and it's been discussed over and over. Perhaps I should send it in to the TV show Myth Busters ;-)

What I always suggest is for people to experiment for themselves. But, the problem is most people don't know how to conduct a "scientific" experiment correctly, and come up with inaccurate results. Somewhere I saw a good outline for an experiment of this nature. I'll see if I can find it.

Regards,

Austin
 
Hi Jurgen,

> ...with all the hopefully good HASSELBLAD DIGITAL information > in my head, I will have a good bavarian beer.

This will most certainly reduce your bit depth ;-)

Regards,

Austin
 
Austin

As long as my bit depth does not fall under 12 bits , it will be allright .
happy.gif


I am very much looking forward to attend that HASSELBLAD DIGITAL seminar , because I still do not achieve the bright colours Marc does .
Hopefully it will bring me the knowledge , I am looking for .

Jürgen
 
" ... but that you are questioning the discussion surrounding 8 bit vs 16 bit. At least that's my impression."

That's correct. I understand bit depth as much as I need to, so it's a question of whether I'm wasting time and space saving 16 bit files... which is actually a different conversation.

However, the discussion began concerning the capture device and we've digressed to bit depth during post workflow.

In your linked web article, he says:

Quality Image Capture Technology is Critical
What I have commented on so far is related to working in Photoshop. This bit-depth issue is sometimes confused with the technical details of a good scanner or digital camera that can deliver high quality images. It is well known that to obtain high quality images, a scanner or digital camera usually requires higher bit-depth options. Meaning that you often need more than 8-bits per pixel to be captured during the scan or capture phase. This input scanning or digital camera strategy allows the resulting file to achieve good shadow detail and maintain color saturation. It is very typical for an input device to capture over 8-bits per pixel and is a positive attribute to an imaging system. This can include high-bit scans s&led down to the best 8-bits or high-bit scans that are imported directly into Photoshop.

Now, as far as working in PS, I wonder why they include 16 bit, and now even 32 bit? ... and since the linked article was published, PS has included more tools in the 16 bit palette. It seems foolish if it has no visible effect ... which I'm not sure is true, but am open minded about it since these file sizes are 2X those of a 8 bit file.

I "think" I see a difference, which is why I use 16 bit workflow, as well as saving them as 16bit Tiffs ... and his may be because I really push the limits of digital capture, quite frequently shooting high ISO in low light, and doing a some severe manipulations like perspective corrections and removing distortions as well as using the Liquify tool to reshape things.

Interesting.
 
Hi Marc,

> Now, as far as working in PS, I wonder why they include 16 bit, and > now even 32 bit? ... and since the linked article was published, PS > has included more tools in the 16 bit palette.

Because more people are working in 16 bit modes, since storage is so cheap these days.

> It seems foolish if > it has no visible effect

Whether it does or doesn't, doesn't matter. It's a matter of perception. They want to sell software/upgrades, and if they don't give people features they want (or think they want), sales will decrease.

> ... which I'm not sure is true,

That's the problem...you can never prove it is true, just that it isn't true, by showing one image that makes a difference. Personally, I think there are circumstances that it will make "a" (unquantified) difference, like skies, but for %99 of most people's work, it won't make any noticeable difference. But, it's the *chance* that it *may* make a difference (even though you can't see it ;-) that keeps most people who use 16 bits using 16 bits.

Regards,

Austin
 
Austin:

I skimmed the article and noticed that it was written 4 years ago. Has current software improved to the point of making a measurable difference in processing 16 bit information.

Regards:

Gilbert
 
Hi Gilbert,

> I skimmed the article and noticed that it was written 4 years ago. Has > current software improved to the point of making a measurable > difference in processing 16 bit information.

That's a very good question. The software has improved you are correct, but I do not believe these improvements will give any different results, with respect to the 8 vs 16 bit debate. What has happened, is since more people are working in 16 bit mode (for what ever reason), the programs have accommodated that by having all the tools also work in 16 bit mode.

I'll say for clarity, there is nothing at all wrong with working in 16 bit mode. The issue is that you may not need to, but there is no harm in doing so. And, I will add, that when I talk about this, I am talking about color only.

B&W is a different story. You do in fact need to work in 16 bit mode for B&W, or you will most certainly see visible differences. With color, you get 24 bits (3 colors * 8 bits/color), with B&W, you have only one color channel...so 8 bits is 8 bits. You'll get combing which will typically lead to posterization when doing anything but the most subtle tonal manipulations in 8 bit mode.

Regards,

Austin
 
Austin

I have attended a color mangement workshop just 2 weeks ago and we could use an APPLE screen as well as an EIZO screen . Both of the same size .
There was a true visible difference , looking at the same image on both screens .
The instructor ( not an EIZO man ) told us , that the EIZO is the only screen on the market , which can produce 16 bit colour depths images . The APPLE screen can't . If any printer on the market can reproduce this was not said and terefore I do not know .
But the given ex&le shows me , that 16 bit depth can't just be a marketing argument , it makes sense and is visible .

Jürgen
 
You must log in or register to reply here.
Back
Top