12- vs 14-bit RAW - worth it?

Started Sep 17, 2013 | Discussions
eaa
eaa
Senior MemberPosts: 1,839Gear list
Like?
12- vs 14-bit RAW - worth it?
Sep 17, 2013

I know thwere are several dual-system users in here, some of which use cameras w/ both 12- and 14-bit RAW. What's your take on that? Any visible difference/advantages in the resulting images?

I suppose Adobe RGB is a must to even see it, as the sRGB gamut may be to narrow to utilize the depth of 14-bit anyway?

I see the E-M1 now has 14-bit (as the first Oly body ever), so I wonder if this is a real benefit, or if it's more a gimmick that'll just slow the processing pipeline down?

Sorry for even mentioning the E-M1 in here (wrong forum and all), but as a long term 4/3 user, 1022 is my "home" forum, and I'm a bit curious if this feature will improve anything over my E-5.

Matrix from Ming Thein's E-M1 review:
http://blog.mingthein.com/2013/09/11/the-2013-olympus-om-d-e-m1-review-2/

-- hide signature --
 eaa's gear list:eaa's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital 35mm 1:3.5 Macro Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital 14-54mm 1:2.8-3.5 +6 more
Olympus E-5 Olympus E-M1
If you believe there are incorrect tags, please send us this post using our feedback form.
dingenus
Senior MemberPosts: 1,700
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to eaa, Sep 17, 2013

14bit, are you sure? I could not find that 14 bit in the official Olympus documentation.

Reply   Reply with quote   Complain
OlyChamp
Senior MemberPosts: 4,340Gear list
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to dingenus, Sep 17, 2013

Ming's comparison chart says 14.

-- hide signature --

"I take photographs, if you want them back see my lawyer!!"
E500, 14-45mm, 40-150mm, 70-300mm, E620, 14-42mm, 40-150mm.

 OlyChamp's gear list:OlyChamp's gear list
Olympus E-500 Olympus E-620 Olympus Zuiko Digital ED 14-42mm 1:3.5-5.6 Olympus Zuiko Digital ED 70-300mm 1:4.0-5.6 Olympus Zuiko Digital ED 40-150mm 1:4.0-5.6 +2 more
Reply   Reply with quote   Complain
rovingtim
Veteran MemberPosts: 8,006
Like?
What really bugs me
In reply to eaa, Sep 17, 2013

EDIT: to answer your question, from what I have seen, it is extremely rare to be able to see the different between 12 bit and 14 bit ... especially as jpgs (the most common display file type) are 8 bit.

It really bugs me that when some people claim a weight advantage, they double the 4/3rds focal length 12-35 = 24 - 70 and then IGNORE the aperture f2.8 = f2.8.

If we're talking exposure, fine. F2.8 = f2.8.

However, the same way 12mm in 4/3rds gives you 24mm in FF, f1.4 gives you the same DoF as f2.8 in FF. To pretend otherwise is deceitful, ESPECIALLY when comparing WEIGHT.

Reply   Reply with quote   Complain
eaa
eaa
Senior MemberPosts: 1,839Gear list
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to dingenus, Sep 17, 2013

dingenus wrote:

14bit, are you sure? I could not find that 14 bit in the official Olympus documentation.

The tech-spec from DPR doesn't mention it, but Ming Thein's matrix clearly states so.
Dunno where he has his data from, as I've not seen this mentioned in any other (p)reviews either, so far...

-- hide signature --
 eaa's gear list:eaa's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital 35mm 1:3.5 Macro Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital 14-54mm 1:2.8-3.5 +6 more
Reply   Reply with quote   Complain
Kevin Sutton
Contributing MemberPosts: 678Gear list
Like?
Groan.........
In reply to rovingtim, Sep 17, 2013

rovingtim wrote:

It really bugs me that when some people claim a weight advantage, they double the 4/3rds focal length 12-35 = 24 - 70 and then IGNORE the aperture f2.8 = f2.8.

If we're talking exposure, fine. F2.8 = f2.8.

However, the same way 12mm in 4/3rds gives you 24mm in FF, f1.4 gives you the same DoF as f2.8 in FF. To pretend otherwise is deceitful, ESPECIALLY when comparing WEIGHT.

Double groan............................................. 

 Kevin Sutton's gear list:Kevin Sutton's gear list
Fujifilm X10 Olympus E-M1 Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital ED 7-14mm 1:4.0 Olympus Zuiko Digital 1.4x Teleconverter EC-14 +14 more
Reply   Reply with quote   Complain
djbrom
Senior MemberPosts: 1,233Gear list
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to eaa, Sep 17, 2013

We don't really know yet. Ming is apparently trying to confirm if the raw files are 12 or 14bit. File size would indicate they are still 12bit but they are slightly larger in size than the e-m5 which makes sense seeing the pixel count is a little larger than the em5.

I guess we will see in time

 djbrom's gear list:djbrom's gear list
Olympus OM-D E-M5 Olympus E-M1 Olympus Zuiko Digital ED 50mm 1:2.0 Macro Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Panasonic Lumix G 20mm F1.7 ASPH +9 more
Reply   Reply with quote   Complain
eaa
eaa
Senior MemberPosts: 1,839Gear list
Like?
Re: What really bugs me
In reply to rovingtim, Sep 17, 2013

rovingtim wrote:

EDIT: to answer your question, from what I have seen, it is extremely rare to be able to see the different between 12 bit and 14 bit ... especially as jpgs (the most common display file type) are 8 bit.

Yes, shooting sRGB and JPEG would probably nullify any effects and advantages from 14-bit, as that depth is probably applied to RAW only. But from AdobeRGB RAW > 16-bit TIFF, maybe the color bit-depth will increase, and thus better sustain color transitions and tonality. So if JPEG should not benefit from it, maybe AdobeRGB TIFFs for print will? I mean, what's the point otrherwise?

It really bugs me that when some people claim a weight advantage, they double the 4/3rds focal length 12-35 = 24 - 70 and then IGNORE the aperture f2.8 = f2.8.

If we're talking exposure, fine. F2.8 = f2.8.

However, the same way 12mm in 4/3rds gives you 24mm in FF, f1.4 gives you the same DoF as f2.8 in FF. To pretend otherwise is deceitful, ESPECIALLY when comparing WEIGHT.

Quite OT, but for what it's worth, the smaller (m)4/3 lenses do have a weight advantage, as they yield the same reach and exposure as an FF lens having twice the FL and the same brightness. But DOF is 2 stops deeper, as an intrinsic part of the (m)4/3 system. But as DOF alone isn't the only criteria for comparison, one cannot just compare a 4/3 12-35 lens w/ a two stops slower FF 24-70. Once you go to wider apertures than the FF lens has, m4/3 has an advantage (higher shutter speeds, no need to bump ISO, better IBIS). That said, this is debated to death elsewhere, so lets stay on topic, and drop the equivalence turmoil here...

-- hide signature --
 eaa's gear list:eaa's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital 35mm 1:3.5 Macro Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital 14-54mm 1:2.8-3.5 +6 more
Reply   Reply with quote   Complain
MatijaK
Regular MemberPosts: 237
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to eaa, Sep 17, 2013

1 bit = 1 stop

12 bits = 12 stops

14 bits = 14 stops

... if you want to store information in a 1:1 ratio. You can squeeze a lot more theoretical bits of DR into 12 bits, but you're going to have to use lossy encoding; think mp3 or jpg files compared to wav and tiff.

As for the gamut, it's a physical measure, not a logical one. Take a rubber band, mark some notches into it and pretend it's sRGB. Then stretch the band and now you have aRGB, but the number of notches remains the same. (That is the reason why wide gamut monitors are a horrible atrocity and why after several years of marketing BS they have all but vanished from new monitor models.)

Reply   Reply with quote   Complain
Sergey_Green
Veteran MemberPosts: 6,782Gear list
Like?
Don't think Adobe even sees it ..
In reply to eaa, Sep 17, 2013

eaa wrote:

I know thwere are several dual-system users in here, some of which use cameras w/ both 12- and 14-bit RAW. What's your take on that? Any visible difference/advantages in the resulting images?

I suppose Adobe RGB is a must to even see it, as the sRGB gamut may be to narrow to utilize the depth of 14-bit anyway?

I see the E-M1 now has 14-bit (as the first Oly body ever), so I wonder if this is a real benefit, or if it's more a gimmick that'll just slow the processing pipeline down?

There were several threads on this subject on Nikon forum after D300 came out. The consensus was that 14-bit did not really add much, if at all.

http://www.dpreview.com/forums/thread/2294948

Here is a good link on the subject,

Are 14-bit Raw Images Really Any Better Than 12-bit Raw?

On the other hand I never really cared for space, and always used 14-bit. The camera was fast enough.

Sorry for even mentioning the E-M1 in here (wrong forum and all), but as a long term 4/3 user, 1022 is my "home" forum, and I'm a bit curious if this feature will improve anything over my E-5.

Neither 12-35 nor 17-55 are equivalent to 34-70/2.8 on a full frame camera. It would need to be constant (and excellent from the start) 12-35/1.4, and made for mFT format specifically. Olympus 12-35/2.8 is as 24-70/5.6 on FF, would be rather inexpensive and uninspiring lens at best.

What, 12-35/2.8 starts from € 967,-- ? - Someone must be kidding  !

-- hide signature --

- sergey

 Sergey_Green's gear list:Sergey_Green's gear list
Samsung Galaxy Note 10.1
Reply   Reply with quote   Complain
eaa
eaa
Senior MemberPosts: 1,839Gear list
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to MatijaK, Sep 17, 2013

MatijaK wrote:

1 bit = 1 stop

12 bits = 12 stops

14 bits = 14 stops

Hmm - (maybe i don't grasp this) but how come then, that many cameras have a deeper DR than their bit-depth should allow for? I.e the E-P5 has a DR of 12.4 EV (provided it is 12-bit), and the 14-bit Nkon D800 has a DR of 14.4 EV?

... if you want to store information in a 1:1 ratio. You can squeeze a lot more theoretical bits of DR into 12 bits, but you're going to have to use lossy encoding; think mp3 or jpg files compared to wav and tiff.

If I understand you correctly, this is the answer to my question above?
Lossy encoding by menas of data compression?
That would reduce tonality, wouldn't it (even w/ a deep DR)?

As for the gamut, it's a physical measure, not a logical one. Take a rubber band, mark some notches into it and pretend it's sRGB. Then stretch the band and now you have aRGB, but the number of notches remains the same. (That is the reason why wide gamut monitors are a horrible atrocity and why after several years of marketing BS they have all but vanished from new monitor models.)

I guess 12- or 14-bit processing also affects color tonality, not just DR?
In that case, as a wider gamut represenst a bigger part of the complete physical spectrum, will not the representation of that part benefit from the "color resolution" within 14-bit processing?

-- hide signature --
 eaa's gear list:eaa's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital 35mm 1:3.5 Macro Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital 14-54mm 1:2.8-3.5 +6 more
Reply   Reply with quote   Complain
Messier Object
Senior MemberPosts: 2,880Gear list
Like?
Re: What really bugs me
In reply to rovingtim, Sep 17, 2013

rovingtim wrote:

It really bugs me that when some people claim a weight advantage, they double the 4/3rds focal length 12-35 = 24 - 70 and then IGNORE the aperture f2.8 = f2.8.

If we're talking exposure, fine. F2.8 = f2.8.

However, the same way 12mm in 4/3rds gives you 24mm in FF, f1.4 gives you the same DoF as f2.8 in FF. To pretend otherwise is deceitful, ESPECIALLY when comparing WEIGHT.

Bugs me that this keeps getting raised.  I could argue that at 300mm it's a real issue, but then Bob and GB and Sergey will just drop in for a visit and wind this thread up to 149 with the same stuff as before.

The bottom line is that 600mm FF lenses are monsters

Peter

 Messier Object's gear list:Messier Object's gear list
Olympus E-300 Olympus E-330 Olympus E-30 Olympus E-5 Canon EOS 5D Mark III +23 more
Reply   Reply with quote   Complain
rovingtim
Veteran MemberPosts: 8,006
Like?
Re: What really bugs me
In reply to eaa, Sep 17, 2013

eaa wrote:

rovingtim wrote:

EDIT: to answer your question, from what I have seen, it is extremely rare to be able to see the different between 12 bit and 14 bit ... especially as jpgs (the most common display file type) are 8 bit.

Yes, shooting sRGB and JPEG would probably nullify any effects and advantages from 14-bit, as that depth is probably applied to RAW only. But from AdobeRGB RAW > 16-bit TIFF, maybe the color bit-depth will increase, and thus better sustain color transitions and tonality. So if JPEG should not benefit from it, maybe AdobeRGB TIFFs for print will? I mean, what's the point otrherwise?

Ever see Spinal Tap?

http://www.youtube.com/watch?v=NrVCjnRdB_k

It really bugs me that when some people claim a weight advantage, they double the 4/3rds focal length 12-35 = 24 - 70 and then IGNORE the aperture f2.8 = f2.8.

If we're talking exposure, fine. F2.8 = f2.8.

However, the same way 12mm in 4/3rds gives you 24mm in FF, f1.4 gives you the same DoF as f2.8 in FF. To pretend otherwise is deceitful, ESPECIALLY when comparing WEIGHT.

Quite OT, but for what it's worth, the smaller (m)4/3 lenses do have a weight advantage, as they yield the same reach and exposure as an FF lens having twice the FL and the same brightness. But DOF is 2 stops deeper, as an intrinsic part of the (m)4/3 system. But as DOF alone isn't the only criteria for comparison, one cannot just compare a 4/3 12-35 lens w/ a two stops slower FF 24-70. Once you go to wider apertures than the FF lens has, m4/3 has an advantage (higher shutter speeds, no need to bump ISO, better IBIS). That said, this is debated to death elsewhere, so lets stay on topic, and drop the equivalence turmoil here...

Not disputing the details you bring up, but he is specifically comparing weight. If you compare weight, it is disingenuous to do what he is doing. m4/3rds should still come out ahead because of the smaller sensor, but the advantage is no where near as great as it is implied here.

Reply   Reply with quote   Complain
MatijaK
Regular MemberPosts: 237
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to eaa, Sep 17, 2013

eaa wrote:

... if you want to store information in a 1:1 ratio. You can squeeze a lot more theoretical bits of DR into 12 bits, but you're going to have to use lossy encoding; think mp3 or jpg files compared to wav and tiff.

If I understand you correctly, this is the answer to my question above?
Lossy encoding by menas of data compression?
That would reduce tonality, wouldn't it (even w/ a deep DR)?

Yes, but as with mp3 and jpg, you don't care about it. 14 stops/bits of DR can fit into 12 bits easily, without you noticing much (if anything at all).

As for the gamut, it's a physical measure, not a logical one. Take a rubber band, mark some notches into it and pretend it's sRGB. Then stretch the band and now you have aRGB, but the number of notches remains the same. (That is the reason why wide gamut monitors are a horrible atrocity and why after several years of marketing BS they have all but vanished from new monitor models.)

I guess 12- or 14-bit processing also affects color tonality, not just DR?
In that case, as a wider gamut represenst a bigger part of the complete physical spectrum, will not the representation of that part benefit from the "color resolution" within 14-bit processing?

Here's the thing: you're doing that on a monitor with 8 bits per channel (which isn't going to change any time soon), not 12 or 14. The monitor is limited in terms of what it can display. Just like sRGB monitors can't display truly intense shades, the aRGB monitor can't display fine shades without dithering and faking them because its gamut is stretched, and it actually loses dynamic range because not everything is in the extreme - a large part of the colours it can display needs to be discarded.

Let's use the rubber band example once again. Take two rubber bands this time and mark six notches, making the total 8 (start + six + end). Leave one band still (that's sRGB) and stretch the other one (aRGB). The notches don't overlap, obviously.

Pretend the bands are the red channel, you're shooting a rose in your back yard, and it's partially cloudy; in other words, the maximum red isn't that intense.

Your sRGB rose band will cover, say, 3/4 of "maximum red" and use 6 notches (incl. the start), which means you'll get 2^6 = 64 shades of red.

Your aRGB band is stretched to, say, double the length of the sRGB band, which means that the red of that rose is now covering 3/8 of the aRGB band's length, so you get only 2^3 = 8 shades of red. The rest is discarded. If you don't discard it, your rose photo will look like it's been taken at mid-summer high noon. Dithering ahoy! You now have 8 shades in aRGB to represent what 64 shades did in sRGB.

The same applies in reverse; your sRGB won't capture the True Glory of the Rose at Mid-Summer High Noon when you actually shoot it that way and will look dull compared to the aRGB version, but - so will your prints. Even sRGB has much a greater gamut range than all printing processes.

In conclusion, aRGB would only be useful if monitors accepted a lot more bits than they do, if graphics cards used a lot more bits than they do, and if operating systems knew how to work with more bits. However, none of that is true, so aRGB can, should and must be considered harmful, because it's detrimental to image quality.

Reply   Reply with quote   Complain
rovingtim
Veteran MemberPosts: 8,006
Like?
here's a video 12 bit v 14 bit
In reply to rovingtim, Sep 17, 2013
Reply   Reply with quote   Complain
Chris Mak
Senior MemberPosts: 1,177
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to MatijaK, Sep 17, 2013

MatijaK wrote:

1 bit = 1 stop

12 bits = 12 stops

14 bits = 14 stops

... if you want to store information in a 1:1 ratio. You can squeeze a lot more theoretical bits of DR into 12 bits, but you're going to have to use lossy encoding; think mp3 or jpg files compared to wav and tiff.

As for the gamut, it's a physical measure, not a logical one. Take a rubber band, mark some notches into it and pretend it's sRGB. Then stretch the band and now you have aRGB, but the number of notches remains the same. (That is the reason why wide gamut monitors are a horrible atrocity and why after several years of marketing BS they have all but vanished from new monitor models.)

A wide gamut monitor allows for exactly what it says: a wider gamut, meaning a higher degree of variation in tonality and color. But in itself a wide gamut monitor like the Eizo coloredge that I use, is not going to give you better looking images. For a start, wide gamut is much more prone to all sorts of issues than a proper sRGB monitor. I'm on a Mac, and there are many threads on fora about the inability of the Mac OS to deal properly with wide gamut screens. For a start, getting a wide gamut monitor means a compulsory step into hardware calibrating, sRGB simulation on a wide screen isn't really wonderful and also makes the benefit of a wide gamut screen non existent. With a wide gamut screen that is hardware calibrated, you get the option to work in a wide gamut color space, mostly approaching aRGB. Whether this has any advantages, is really up to you and how you view/edit your images. Same goes for 12 versus 14 bit. 14 bit allows for higher DR and more latitude in (RAW) post processing, but you are not going to see much of that, if any, when you are working/viewing on a sRGB monitor, certainly when it isn't even calibrated. I can tell you that I see the differences. I went from Oly 12-bit E5 raw to now the Pentax K5IIs 14 bit raw. That's not really fair comparison, but the difference on a wide gamut screen color-wise and tonality wise is véry substantial. When editing for print as 16-bit Photoshop PSD files in a wide color space, there is almost unlimited latitude for raising shadows or midtones using luminance masks, without colors or tonality being affected. A wide gamut screen shows a lot of what you are doing. But when I save an image for the web in resized sRGB jpeg, and view it on a sRGB monitor, will it at all look like a 14-bit wide gamut edited image? No, obviously. Even the editing latitude won't necessarily show, the lighting might have been different, good as well be straight out of camera 8bit sRGB.

Much the same applies for 10-bit video card color: the Eizo coloredge alows for 10-bit color (billon possible colors instead of 8-bit millions of colors) through the display-port connection, and there are endless discussions whether you see any difference. Apple even decided not to support 10-bit video in Mac Os, because of the possible issues (and probably because they are too busy getting everyone on Ipad/phone/pod). Still, you can get 10-bit enabled (Windows) workstations for 3D editing/CAD.

That Olympus should support 14-bit Raw in it's pro-model, is only obvious. What benefit you are going to get of of that, depends...

Chris

Reply   Reply with quote   Complain
eaa
eaa
Senior MemberPosts: 1,839Gear list
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to MatijaK, Sep 17, 2013

MatijaK wrote:

In conclusion, aRGB would only be useful if monitors accepted a lot more bits than they do, if graphics cards used a lot more bits than they do, and if operating systems knew how to work with more bits. However, none of that is true, so aRGB can, should and must be considered harmful, because it's detrimental to image quality.

Thx for this down-to-earth example, it makes it a bit (pun intended ) easier to comprehend.
IOW, using aRGB is mostly a waist, for both screen and print use?

Because any theoretical gain in colors are either lost, stretched, interpolated, compressed, faked and what not, due to the sub-par/sub-bits equipment used to show them, be ithey printers or screens?

In that case, to maintain a predictable presentation throughout any workflow, sRGB ought to be the gamut of preference?

What then justifies the very existence of aRGB?
And/or even 14-bit in-cam processing?

-- hide signature --
 eaa's gear list:eaa's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital 35mm 1:3.5 Macro Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital 14-54mm 1:2.8-3.5 +6 more
Reply   Reply with quote   Complain
Chris Mak
Senior MemberPosts: 1,177
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to MatijaK, Sep 17, 2013

MatijaK wrote:

eaa wrote:

... if you want to store information in a 1:1 ratio. You can squeeze a lot more theoretical bits of DR into 12 bits, but you're going to have to use lossy encoding; think mp3 or jpg files compared to wav and tiff.

If I understand you correctly, this is the answer to my question above?
Lossy encoding by menas of data compression?
That would reduce tonality, wouldn't it (even w/ a deep DR)?

Yes, but as with mp3 and jpg, you don't care about it. 14 stops/bits of DR can fit into 12 bits easily, without you noticing much (if anything at all).

As for the gamut, it's a physical measure, not a logical one. Take a rubber band, mark some notches into it and pretend it's sRGB. Then stretch the band and now you have aRGB, but the number of notches remains the same. (That is the reason why wide gamut monitors are a horrible atrocity and why after several years of marketing BS they have all but vanished from new monitor models.)

I guess 12- or 14-bit processing also affects color tonality, not just DR?
In that case, as a wider gamut represenst a bigger part of the complete physical spectrum, will not the representation of that part benefit from the "color resolution" within 14-bit processing?

Here's the thing: you're doing that on a monitor with 8 bits per channel (which isn't going to change any time soon), not 12 or 14. The monitor is limited in terms of what it can display. Just like sRGB monitors can't display truly intense shades, the aRGB monitor can't display fine shades without dithering and faking them because its gamut is stretched, and it actually loses dynamic range because not everything is in the extreme - a large part of the colours it can display needs to be discarded.

Let's use the rubber band example once again. Take two rubber bands this time and mark six notches, making the total 8 (start + six + end). Leave one band still (that's sRGB) and stretch the other one (aRGB). The notches don't overlap, obviously.

Pretend the bands are the red channel, you're shooting a rose in your back yard, and it's partially cloudy; in other words, the maximum red isn't that intense.

Your sRGB rose band will cover, say, 3/4 of "maximum red" and use 6 notches (incl. the start), which means you'll get 2^6 = 64 shades of red.

Your aRGB band is stretched to, say, double the length of the sRGB band, which means that the red of that rose is now covering 3/8 of the aRGB band's length, so you get only 2^3 = 8 shades of red. The rest is discarded. If you don't discard it, your rose photo will look like it's been taken at mid-summer high noon. Dithering ahoy! You now have 8 shades in aRGB to represent what 64 shades did in sRGB.

The same applies in reverse; your sRGB won't capture the True Glory of the Rose at Mid-Summer High Noon when you actually shoot it that way and will look dull compared to the aRGB version, but - so will your prints. Even sRGB has much a greater gamut range than all printing processes.

In conclusion, aRGB would only be useful if monitors accepted a lot more bits than they do, if graphics cards used a lot more bits than they do, and if operating systems knew how to work with more bits. However, none of that is true, so aRGB can, should and must be considered harmful, because it's detrimental to image quality.

Don't forget about hardware calibration on good wide gamut screen monitors, which can be 10/12 or even 14-bit. You are not going to calibrate a wide gamut screen through your 8-bit video card...

Reply   Reply with quote   Complain
Chris Mak
Senior MemberPosts: 1,177
Like?
Re: 12- vs 14-bit RAW - worth it?
In reply to eaa, Sep 17, 2013

eaa wrote:

MatijaK wrote:

In conclusion, aRGB would only be useful if monitors accepted a lot more bits than they do, if graphics cards used a lot more bits than they do, and if operating systems knew how to work with more bits. However, none of that is true, so aRGB can, should and must be considered harmful, because it's detrimental to image quality.

Thx for this down-to-earth example, it makes it a bit (pun intended ) easier to comprehend.
IOW, using aRGB is mostly a waist, for both screen and print use?

Because any theoretical gain in colors are either lost, stretched, interpolated, compressed, faked and what not, due to the sub-par/sub-bits equipment used to show them, be ithey printers or screens?

In that case, to maintain a predictable presentation throughout any workflow, sRGB ought to be the gamut of preference?

What then justifies the very existence of aRGB?
And/or even 14-bit in-cam processing?

God forbid, that won't get us forward. sRGB is horrible on wide gamut screen in certain tones/colors, most noticeably intense yellow. The point is: there will always have to be standards that allow for various needs. If one only needs sRGB, fine, but the development into larger gamut/higher bit software and hardware is essential. Hollywood now edits on 10-bit monitor output, perhaps that will help it gain acceptance

Reply   Reply with quote   Complain
eaa
eaa
Senior MemberPosts: 1,839Gear list
Like?
Re: here's a video 12 bit v 14 bit
In reply to rovingtim, Sep 17, 2013

rovingtim wrote:

and here's a guy who thinks he's found an advantage:

http://www.earthboundlight.com/phototips/nikon-d300-d3-14-bit-versus-12-bit.html

Thx for that one!
It is truly informative, even downright educational.
I guess the 10 FPS in the E-M1 is dependent on more than just SAF then, as to maintain speed and deep buffers, 12-bit processing may be a prerequisite...

-- hide signature --
 eaa's gear list:eaa's gear list
Olympus E-5 Olympus E-M1 Olympus Zuiko Digital 35mm 1:3.5 Macro Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD Olympus Zuiko Digital 14-54mm 1:2.8-3.5 +6 more
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads