Is IIQ 16bit pointless at higher ISO?

[narikin]

I wondered if there is any point in using the higher bit option of Phase's IIQ-L 16bit, if you are going much above base ISO 50/100?
Has someone smarter than me done tests?!

-I can see very fine benefits of using it at 100asa, and maybe at 200asa it's still marginally worth it, but 400 and up, surely the noise and color degradation renders the finer details potentially in full 16bit, rather moot?

I'm happy with the IIQ-L file normally: smaller, quick shooting, and more shots per CF card.

[scyth]

but 400 and up, surely the noise

if you do not saturate your sensels (exposure = aperture/t-stop + exposure time) then the higher gain (might be a result of some higher nominal ISO dialed in, unless for example camera is doing ISO-by-tag - in this case there is no analog gain pre ADC and no digital multiplication of DNs post ADC ) in many sensors is actually better for S/N in deep shadows, for some sensors - quite irrelevant /ISO-less sensors or where ISO-by-tag is used/ and for some when you use some "intermediate" nominal ISO values it may be worse... so you want to test how ISO/gain is actually implemented in a particular P1 back/sensor combo before saying "surely"...

[ErikKaffehr]

Hi,

Probably so but may be not always.

Increasing ISO is essentially the same as underexposure. Full exposure at base ISO fills up the photodiode fully. If you reduce exposure with one EV it will be half full. Reducing well utilisation to half means you loose an EV in DR.

DR on the IQ3100 is around 15 stops. So, if you underexpose one stop you will have 14 EV of DR that fits exactly in a 14 bits wide data path.

There is an exception, though. Reducing exposure will protect specular highlights, unless raw data is manipulated when increasing ISO.

Best regards
Erik

I wondered if there is any point in using the higher bit option of Phase's IIQ-L 16bit, if you are going much above base ISO 50/100?
Has someone smarter than me done tests?!

-I can see very fine benefits of using it at 100asa, and maybe at 200asa it's still marginally worth it, but 400 and up, surely the noise and color degradation renders the finer details potentially in full 16bit, rather moot?

I'm happy with the IIQ-L file normally: smaller, quick shooting, and more shots per CF card.

[Bart_van_der_Wolf]

I wondered if there is any point in using the higher bit option of Phase's IIQ-L 16bit, if you are going much above base ISO 50/100?
Has someone smarter than me done tests?!

-I can see very fine benefits of using it at 100asa, and maybe at 200asa it's still marginally worth it, but 400 and up, surely the noise and color degradation renders the finer details potentially in full 16bit, rather moot?

Hi,

Good question. I have not done specific tests, but I do assume that above ISO 400 there will be insufficient effective DR to warrant 16-bit encoding, 14-bit should suffice. Whether ISO 200 or ISO 400 is the turning point, would take more analysis/testing to determine.

Cheers,
Bart

[Wayne Fox]

I'm happy with the IIQ-L file normally: smaller, quick shooting, and more shots per CF card.

I haven't tested it, just wondering how much smaller and quicker it is.

I would also wonder if it is enough to warrant considering, if it might be worthwhile to also use when the scene dynamic range isn't really pushing the extremes.  Would shooting in 16bit offer anything if there are only 8 or 9 stops of dynamic range in the scene?

I do know a scene I shot in Colorado where the colors were pretty much unbelievable, and the camera actually seemed to nail what I remember seeing (even though they were so extreme they were unbelievable on the screen and I had to tone them down).  My buddy shooting beside me on the a7r2 didn't get the same results, and can't quite get the tones out of the aspen trees that I did.  That may be more to do with C1 (my file) and Lr (his file).

[Steve Hendrix]

Ok, I may be wading in here where I normally don't want to, but help me understand why the question of the relevance of 16 bit capture is answered almost exclusively in terms of dynamic range?

You know, I've always been about the image. The image on your screen, the image you print on paper or whatever. But, always about the image. What does it mean in terms of end results, etc.

And for someone who leads toward the non-scientific approach, please try to keep this in laymans terms.


Thanks,
Steve Hendrix

[ErikKaffehr]

Hi Steve,

The reason is that DR is the amount of usable information of the image. It is defined as maximum usable signal (normally called full well capacity) divided by signal where Signal/Noise Ratio is 1.

One bit is needed for stop of DR. So if DR is 15 EV there is 15 bits of information, the rest is noise. With 13EV there is 13 bits of information and so on.

It is quite possible to transfer say 15 stops of DR trough say 12 bit of data using a tone curve, but using more bits than what corresponds to DR is just waste of band width. This may be over-simplified, but it is a pretty good rule of thumb.

A good article by Jack Hogan is here: http://www.strollswithmydog.com/how-many-bits-to-fully-encode-my-image/

Emil Martinec has a classic piece on the issue here: https://theory.uchicago.edu/~ejm/pix/20d/tests/noise/noise-p3.html

Getting back to the original posting. Assume that you expose fully to the right. That means that your relevant highlights go up to Full Well Capacity (FWC). You utilise the sensor fully. If DR is 15 EV it means that noise floor is 32768 times below FWC, so you need 15 bits of linear data to represent the usable output from the sensor.

Now, cut exposure by half, we no longer utilise the full FWC, just half of it. But, noise is still the same, at least regarding readout noise. So we now only have 16384 bits worth of data, needing 14 bits. We can multiply the signal with 4, but that would multiply the noise by four, too.

Astronomers may utilise deeper data (below SNR = 1).

A small note, if you pick up spec sheets from Kodak or Dalsa sensors they will mention readout noise in electron charges and saturation level also in electron charges. The saturation level is FWC. They may also give the dynamic range of the sensor, say 71 dB. You can convert dB to stops by dividing with 6 (6.021 to be more exact). Using either calculation you will end up with the figures DxO measures at base ISO in what they call screen mode and that figure normally fits vendor data very closely.

Best regards
Erik

Ok, I may be wading in here where I normally don't want to, but help me understand why the question of the relevance of 16 bit capture is answered almost exclusively in terms of dynamic range?

You know, I've always been about the image. The image on your screen, the image you print on paper or whatever. But, always about the image. What does it mean in terms of end results, etc.

And for someone who leads toward the non-scientific approach, please try to keep this in laymans terms.


Thanks,
Steve Hendrix

[ErikKaffehr]

I haven't tested it, just wondering how much smaller and quicker it is.

I would also wonder if it is enough to warrant considering, if it might be worthwhile to also use when the scene dynamic range isn't really pushing the extremes.  Would shooting in 16bit offer anything if there are only 8 or 9 stops of dynamic range in the scene?

I do know I of scene in Colorado where the colors were pretty much unbelievable,and the camera actually seemed to nail what I remember seeing (even though they were so extreme they were unbelievable on the screen and I had to tone them down).  My buddy shooting beside me on the a7r2 didn't get the same results, and can't quite get the tones out of the aspen trees that I did.  That may be more to do with C1 (my file) and Lr (his file).

[Wayne Fox]

Now, cut exposure by half, we no longer utilise the full FWC, just half of it. But, noise is still the same, at least regarding readout noise. So we now only have 16384 bits worth of data, needing 14 bits. We can multiply the signal with 4, but that would multiply the noise by four, too.

What if you still shoot ETTR, so you are close to FWC, but meaning there aren't any tones on the left side of the histogram. So dynamic range is half of the sensors capability.

Does still shooting in 16 bit offer anything, or would 14bit offer the same quality of raw capture?

(I'm with Steve on this, wondering if 16bit still offers the best file quality and best chance at the highest possible rendering in the raw processor even in cases like this)

[ErikKaffehr]

Hi Wayne,

No, I would not say data path width would matter with 8-9 stops of DR.

The greatest differentiator may the colour profiles used in the converter. Lightroom has several profiles. I always use home made profiles.


What it may be worth I enclose two screen dumps. I happened to shoot two very similar images with my P45+ and my Sony A7rII.

The first one shows the Sony A7rII image processed with different colour profiles, DCampProf (what I use), Camera Landscape and Adobe standard. I also include the P45+ mage processed in Lghtroom with a DCamProf profile.

The second one I also processed the image with Capture One, I have very little experience with that tool, I have done some local adjustment on the sky and bright side of the image, similar to Lightroom.

After taking that image, I set up the back for horisontal shooting and it came loose from the camera (altough it felt solidly attached) and fell four fee on pretty solid ground. It seems to survived with the sensor still in alignment. An ouch experience!

Best regards
Erik

I haven't tested it, just wondering how much smaller and quicker it is.

I would also wonder if it is enough to warrant considering, if it might be worthwhile to also use when the scene dynamic range isn't really pushing the extremes.  Would shooting in 16bit offer anything if there are only 8 or 9 stops of dynamic range in the scene?

I do know I of scene in Colorado where the colors were pretty much unbelievable,and the camera actually seemed to nail what I remember seeing (even though they were so extreme they were unbelievable on the screen and I had to tone them down).  My buddy shooting beside me on the a7r2 didn't get the same results, and can't quite get the tones out of the aspen trees that I did.  That may be more to do with C1 (my file) and Lr (his file).

[landscapephoto]

Ok, I may be wading in here where I normally don't want to, but help me understand why the question of the relevance of 16 bit capture is answered almost exclusively in terms of dynamic range?

You know, I've always been about the image. The image on your screen, the image you print on paper or whatever. But, always about the image. What does it mean in terms of end results, etc.

And for someone who leads toward the non-scientific approach, please try to keep this in laymans terms.

In laymans terms, higher dynamic range means that there are more levels to be encoded. For that you need more bits. Therefore dynamic range and bit depth are directly linked.

What the question means for the image is that, with higher iso, the image will be the same when using 16 or 14 bits. The answer, which I don't know, depends on the internal design of the camera. Specifically about the presence or absence of analog amplification before the ADC.

[ErikKaffehr]

The Sony sensor used in the IQ3-100MP has almost certainly no variable analog amplification before ADC.

Some Sony sensors use a trick invented by Aptina that reduces FWC at a certain ISO setting, we cannot say before we have seen photon transfer curves.

Best regards
Erik

In laymans terms, higher dynamic range means that there are more levels to be encoded. For that you need more bits. Therefore dynamic range and bit depth are directly linked.

What the question means for the image is that, with higher iso, the image will be the same when using 16 or 14 bits. The answer, which I don't know, depends on the internal design of the camera. Specifically about the presence or absence of analog amplification before the ADC.

[ErikKaffehr]

Hi Wayne,

In that case I would guess that the shadows are dominated by shot noise and neither DR or bit depth is really relevant.

So what happens when you rise ISO. Is the Raw data just multiplicated by 2 for each EV? Or is the raw data just tagged with the ISO value? What does the histogram show? Does the raw converter change noise reduction based on ISO?

The physics are well known, in camera processing is a great unknown…

Best regards
Erik

What if you still shoot ETTR, so you are close to FWC, but meaning there aren't any tones on the left side of the histogram. So dynamic range is half of the sensors capability.

Does still shooting in 16 bit offer anything, or would 14bit offer the same quality of raw capture?

(I'm with Steve on this, wondering if 16bit still offers the best file quality and best chance at the highest possible rendering in the raw processor even in cases like this)

[Bart_van_der_Wolf]

What if you still shoot ETTR, so you are close to FWC, but meaning there aren't any tones on the left side of the histogram. So dynamic range is half of the sensors capability.

Does still shooting in 16 bit offer anything, or would 14bit offer the same quality of raw capture?

Hi Wayne,

Increasing the ISO is typically done to compensate for underexposure. Underexposure means reducing DR. When the DR is low enough to not carry more than 14-bits of information, then there is little benefit in encoding with higher precision.

'Accuracy' and 'Precision' are two concepts that have a distinctly different mathematical meaning. As we underexpose, the average recorded signal will get less Accurate. There will be more noise fluctuation recorded instead of the actual signal level, and it's caused by Readnoise (affecting the shadows more), but there will also be less absolute shotnoise (affecting the midtones and highlights, and also less signal), but we can still encode that inaccurate noisy signal with very high Precision (e.g. 16-bit).

That's where Signal to Noise ratios are a useful concept, they tell us about the (average) Accuracy, regardless of the Precision (14 or 16-bits) with which we encode that noisy signal. More exposure usually means higher S/N ratio (because shot-noise increases less than the signal level increases). So then underexposure will reduce Accuracy to the point that it gets so inaccurate that there is little benefit encoding it with high precision.

As said, it would take more in-depth testing and analysis to mathematically determine the exact tipping point, but I expect it to be somewhere around ISO 400.

Cheers,
Bart

[Wayne Fox]

Increasing the ISO is typically done to compensate for underexposure. Underexposure means reducing DR. When the DR is low enough to not carry more than 14-bits of information, then there is little benefit in encoding with higher precision.

What if I'm not increasing ISO ... a low dynamic range scene which I would "overexpose" to get ETTR, so I end up with a histogram empty of any data on the left 1/4-1/3. Perhaps even if it isn't exposed to the right (example attached), there still isn't any data on the left and in this case on the right side of the histogram.

Does the 16bit file offer anything at all in flexibility in post processing?  Does the process of making a 16bit tiff from a 14bit raw produce an identical result or is there perhaps some small  gains to be had from making a 16bit tiff from a 16bit raw? (from a practical standpoint not necessarily from a bit by bit standpoint).

My unscientific "logic" based on this discussion tells me there is no point in shooting 16bit on scenes with lower dynamic range, or probably when pushing the ISO past 400 up since all of the available dynamic range can be encoded precisely/accurately enough with 14bits.

[Steve Hendrix]

In laymans terms, higher dynamic range means that there are more levels to be encoded. For that you need more bits. Therefore dynamic range and bit depth are directly linked.

What the question means for the image is that, with higher iso, the image will be the same when using 16 or 14 bits. The answer, which I don't know, depends on the internal design of the camera. Specifically about the presence or absence of analog amplification before the ADC.

Ok, thank you and everyone for pitching in on that question. But I am aware of the relationship between bit depth and dynamic range. My question was specifically why is Dynamic Range exclusively being discussed with regard to bit depth? Because for years, I've heard from clients who shoot 35mm DSLR and medium format and complain about color with 35mm DSLR. About the "global color response". Some have attributed this to CMOS vs CCD. I have not felt this to be the case. Instead, I attribute this to 14 bit vs 16 bit (if not completely, at least to a significant degree).

When you shoot IIQL 16 bit, IIQL, and IIQS, there are color differences. In the few tests I've done, it is very subtle, but there is a difference. Whenever I do this type of test, I feel like I can see it, and it feels to me like the 14 bit captures are splashing the color round while the 16 bit capture is differentiating more effectively and accurately. Today I shot a quick few shots at ISO 1600. There are color differences, it is very subtle, but I can see it. Inevitably, these differences can be hidden or magnified as one shoots many different scenes. Screenshots are below.

Keep in mind, for me, it's a moot point. I don't have any need to try and optimize storage or increase capture rate by shooting 14 bit files. I want the full 16 bits, no matter how subtle a difference there is.


Steve Hendrix/CI

[Wayne Fox]

Sort of my thinking as well, maybe even having to do with the question I asked earlier about more accuracy because the file isn’t being expanded to 16bit from 14bits.

I did notice 14bit still works in Lr, for some that might be useful.  I’m not particularly worried about file sizes (which aren’t remarkably smaller in 14bit) or capture speed either, but I do not like how i’m forced to use C1’s cataloging tools in my workflow, and can’t opt to use Lr for my DAM and preliminary image adjustments so I can move to C1 for the images I want to work on.  Additionally some images I’m more comfortable with LR’s local adjustment tools. 

Not sure that will make me switch, because I’m more afraid of leaving at 14bit and end up shooting a sunset where I need the dynamic range.

[Paul2660]

Sort of my thinking as well, maybe even having to do with the question I asked earlier about more accuracy because the file isn’t being expanded to 16bit from 14bits.

I did notice 14bit still works in Lr, for some that might be useful.  I’m not particularly worried about file sizes (which aren’t remarkably smaller in 14bit) or capture speed either, but I do not like how i’m forced to use C1’s cataloging tools in my workflow, and can’t opt to use Lr for my DAM and preliminary image adjustments so I can move to C1 for the images I want to work on.  Additionally some images I’m more comfortable with LR’s local adjustment tools. 

Not sure that will make me switch, because I’m more afraid of leaving at 14bit and end up shooting a sunset where I need the dynamic range.

Hi Wayne

From reading this can LR open a 14 bit IQ 100 file?  I have given up on LR Working with the 16 raw.

Thanks
Paul C

[Wayne Fox]

Hi Wayne

From reading this can LR open a 14 bit IQ 100 file?  I have given up on LR Working with the 16 raw.

Thanks
Paul C

Lr sees the files on the card and shows the image unlike the 16bit files.  I assume it will import them, though I didn't actually follow through.  but then I opened them in ACR and was able to manipulate them fine.

[Paul2660]

Thanks Wayne

I never thought to try the 14 bit files in LR. This is good to know as it looks like Adobe may never get support for the 16 bit versions.

Paul C