Disclaimer and Warning: This document has been determined to be exempts from ITAR regulations due to the following - All information contained below is either in the Public Domain or can be determined from Common Scientific Principles.
Undestanding System Gain and using this to compare scopes to determine suitability for ultra-dark conditions.
One of the most commonly heard metrics for measuring Night Vision is the gain - or how much it amplifies light - yet most of the time you only hear about the tube gain. While useful, if you do a search on gain, you'll also discover something called System gain - this is a relative measure of how much the entire Night Observation Device amplifies light, including loss through the lenses.
A good example of a definition of System Gain can be obtained from the TNVC website *1, and is as follows;
<div class="ubbcode-block"><div class="ubbcode-header">Quote:</div><div class="ubbcode-body">"Gain
Also called brightness gain or luminance gain. This is the number of times a night vision device amplifies light input. It is usually measured as tube gain and system gain. Tube gain is measured as the light output (in fL) divided by the light input (in fc). This figure is usually expressed in values of tens of thousands. If tube gain is pushed too high, the tube will be “noisier” and the signal-to-noise ration many go down. U.S. military Gen 3 image tubes operate at gains of between 20,000 and 45,000. On the other hand, system gain is measured as the light output (fL) divided by the light input (also fL) and is what the user actually sees. System gain is usually seen in the thousands. U.S. military systems operate at 2,000 to 3,000. In any night vision system, the tube gain is reduced by the system’s lenses and is affected by the quality of the optics or any filters. Therefore, system gain is a more important measurement to the user."</div></div>
from: "http://www.tnvc.com/faq/"
Yet despite almost every site saying how important it is, very few sites actually mention what the system gain is of different NODs.
As a rule of thumb, system gain used to be around 1/10th of the tube gain, but this doesn't really take other factors, such as lens choice, into consideration. To be militarily useful, it is generally considered that a NOD should have a system gain value of around 2000x to 3000x minimum. *1
So what should you expect? And how do you calculate the system gain of systems that do not have their gain listed?
Well, here's some well known "Typical" system gain levels. *2
AN/PVS-5 - 800
AN/AVS-6 Omni I&II - 2000
AN/AVS-6 Omni III - 2000
AN/AVS-6 Omni IV - 3000
AN/AVS-6 Omni VI - 5000
AN/PVS-14 Omni VI - 7000 (*3)
So although it's common to hear about system resolution. It's not all that common to see examples.
Additionally, because tube resolution and MTF drop off significantly as light levels deteriorate past the point where the tube can amplify the image to full brightness *2, the system gain is a very important component of understanding how well a particular night vision device will perform under extremely dark conditions.
For example, at moonlight levels, the PVS-5 has typical system resolution of .72cy/mr while at starlight levels, the PVS-5 has typical system resolution of around .4cy/mr - nearly half of what it can provide under starlight, while even Gen3 tubes of Omni I era don't fare much better *2.
Yet looking through the examples and taking lenses into consideration gives an interesting comparison.
Based on F-numbers, the F-numbers of those NODs are;
AN/PVS-5 - F1.1
AN/AVS-6 Omni I&II - F1.2
AN/AVS-6 Omni III - F1.2
AN/AVS-6 Omni IV - F1.2
AN/AVS-6 Omni VI - F1.2
AN/PVS-14 Omni VI - F1.2
Additionally, there is not a great deal of different in the ocular lens system of most common NODs either.
This information, when coupled with the system gain, leads to an obvious conclusion. If lenses have not significantly changed, the the most common factor in determining the system gain is the tube. As can be seen from the above tables, the gains in System gain since Omni I have been considerable.
Using this information, it is therefore possible to approximate the system gain of any system if you know the era of the tube and the F-number of the objective lens.
To do this, you need to also understand what the F-number of a lens is. Simply explained, it's a ratio of the lens diameter to it's focal length. As a result, a lens that was 26mm in diameter and 26mm focal length would have an F-number of 1.0 while a lens that was 12mm in diameter with a 26mm focal length would have an F-number of 2.0.
Now you remember basic area calculations don't you? You have to square the diameter of a lens to understand it's area and it's the same for F-numbers. To get it's relative area, just square it -
So a F-number of 1.2 would leave a relative area of 1.44
Let's compare that to say, a 3x adapter for a PVS-14, which has a F-number of 1.5. Squared, this has a relative area of 2.25. This means the ratio of F1.2 to F1.5 is 0.64.
As a result, you can assume that when you put a 3x adapter on a PVS-14, you reduce the maximum system gain by a factor of 0.64, or in other words, the maximum system gain is now around 4480, or 4500 if you like round numbers.
Likewise, if a riflescope has a F1.7 lens such as a D760 and a MX10160A tube of Omni VI era or later ( Omni VI exceeds all current Milspecs, including Omni VIII ) then you can likewise work out it's performance relative to a PVS-14.
Eg, The relative area factor is 0.49 ( 0.5 ) so the maximum system gain is still around 2500 - within acceptable Military levels.
Now is this all? Not quite - it's actually T-numbers that matter, not F-numbers. T-numbers actually measure how much light gets through a lens and take into consideration internal obstructions within the lens and other factors, but because the maximum a T-number can be is the F-number in a perfect lens, the T-number will never be higher and usually follows fairly closely to the F-number. They are also harder to find out for lenses, so for all practical purposes, F-numbers are good enough for our rough calculations.
Putting this information to use is just as easy as the above calculations. Recently I did a quick analysis of two scopes - the AN/PVS-27 and the CNVD-LR - both exceptional and well respected clip-on NODs.
The AN/PVS-27 has a F-number of 1.0 (*6)- one of the fastest of any NOD out of the factory. The CNVD-LR has a F-number of 1.85 (*7) - one of the slower NODs out of the factory. For this reason, they make for good calculations and examples of the importance of system resolution.
Using the above "rule of thumb" we can make an approximation that the max system gain of the PVS-27 is around 10,000 while the max system gain of the CNVD-LR, assuming the same high-quality MX-11769 of Omni VI or later design, is around 2900 - still at military levels, but barely scraping past Gen III Omni I and Omni II levels.
The final piece of the puzzle to understanding this subject is how this affects the final image quality and what significance does it have to someone using both systems. Well, there's a specification that's important to remember here: Screen Brightness. On tubes it's always specified as a maximum. That's because a tube will automatically reduce it's gain when there's enough light that it doesn't need to amplify it a lot, or when amplifying to the maximum level would exceed the maximum screen brightness.
The important thing to remember then is that the system gain only becomes important when it's at maximum, because the screen automatic brightness control will reduce the tube gain when it's not dark enough.
This means that two devices like the AN/PVS-27 and the CNVD-LR with the same tube will actually perform at the same level under high brightness conditions - and in fact, due to the better system resolution specification, the CNVD-LR actually outperforms the AN/PVS-27 under such conditions. by about 20% (*8)
A good example of a high-light comparison is this image: Resolution aside, similar performance gain-wise is present in both Gen2 PVS-5 images and the Gen3 MX10160A Omni V image.
However, as it get's darker, usually anything below starlight- such as when you're in a barren landscape or when there's dark clouds overhead, things will change. At the point at which the AN/PVS-27 runs out of system gain to maintain screen brightness, the difference will be that the resolution of the CNVD-LR is about half of that of the AN/PVS-27 and th AN/PVS-27 will be somewhere between three and four times brighter than the CNVD-LR.
A good example of this is a standard "starlight" image from both a Gen II PVS-5 and a GenIII MX10160A Omni V system.
PVS-5 - system gain approx 1000x
Gen III - system gain approx 3000x
Because of this, partly due to the reduced resolution and partly because of the lower brightness, you can also assume that the range of the PVS-27 will be more than double that of the CNVD-LR.
Something to think about when comparing scopes visually - no matter who or where you get it from.
So in evaluating the two, how do you know if you're at a reasonable darkness level? Especially as humans can't really tell the difference after about 1/4 moon? Simply use your best judgment based on what you can see and do your test on a starlight only night with heavy cloud cover. If the moon tables (*7) show any moon in the sky or within 1 hour of the horizon, it's going to be bright. If you can see the clouds as bright against the dark sky, it's too bright ( reflected light from nearby ground-mounted lights ). If neither of these factors are there and the clouds look like inky blots against the sky, conditions are perfect for your test. Just make sure equivalent tubes are in both.
Then after evaluating the difference between scopes, consider how dark it's likely to be when you're using your NOD and weigh that into the factor.
This is a test you can apply to any NOD comparison, BTW, not just those two - Just do the maths, work out the difference in system gain and take it from there -
And if you have any questions, please ask. That's what a forum is for
Regards
David
References:
*1: Many sites list this, eg;
http://www.morovision.com/terminology.htm
and
http://www.tnvc.com/faq/
eg: "U.S. military systems operate at 2,000 to 3,000"
*2 Past, Present and Future Night Vision Goggles in US Army Aviation
http://ftp.rta.nato.int/public//PubFullText/RTO/MP/RTO-MP-019///$MP-019-32.PDF
*3 - PVS-14 system gain typically ranges from 0.25 to 7000 depending on gain setting on a modern tube.
*4 - Approximate F-numbers for well known NV - Information provided by Litton ( Now L3 ) and ITT in public documentation - ITAR free.
*5 Information taken from datasheets for CNVD-LR -
http://nightvision-thermal.com/pdf/nitehog_cnvd-lr.pdf
*6 Information taken from data published on the AN/PVS-27 MUNS -
http://www.ownthenight.com/catalog/i177.html
*7 Moon phases can be obtained online: eg,
http://www.morovision.com/moonphases.htm
*8 - As per this thread where I previously calculated the difference between these scopes.
http://www.snipershide.com/forum/ubbthreads.php?ubb=showflat&Number=3118915#Post3118915
Additional-
High light comparison - PVS-5 and Gen III Omni V
http://aunv.blackice.com.au/userfiles/david-small-223.jpg
Low light comparison - PVS-5 and Gen III Omni V
Gen II PVS-5
http://aunv.blackice.com.au/userfiles/david-IMG_0474-800.jpg
Approximate System Gain: 1000x
Gen III Omni V.
http://aunv.blackice.com.au/userfiles/david-IMG_0406-800.jpg
Approximate System Gain: 3000x
Undestanding System Gain and using this to compare scopes to determine suitability for ultra-dark conditions.
One of the most commonly heard metrics for measuring Night Vision is the gain - or how much it amplifies light - yet most of the time you only hear about the tube gain. While useful, if you do a search on gain, you'll also discover something called System gain - this is a relative measure of how much the entire Night Observation Device amplifies light, including loss through the lenses.
A good example of a definition of System Gain can be obtained from the TNVC website *1, and is as follows;
<div class="ubbcode-block"><div class="ubbcode-header">Quote:</div><div class="ubbcode-body">"Gain
Also called brightness gain or luminance gain. This is the number of times a night vision device amplifies light input. It is usually measured as tube gain and system gain. Tube gain is measured as the light output (in fL) divided by the light input (in fc). This figure is usually expressed in values of tens of thousands. If tube gain is pushed too high, the tube will be “noisier” and the signal-to-noise ration many go down. U.S. military Gen 3 image tubes operate at gains of between 20,000 and 45,000. On the other hand, system gain is measured as the light output (fL) divided by the light input (also fL) and is what the user actually sees. System gain is usually seen in the thousands. U.S. military systems operate at 2,000 to 3,000. In any night vision system, the tube gain is reduced by the system’s lenses and is affected by the quality of the optics or any filters. Therefore, system gain is a more important measurement to the user."</div></div>
from: "http://www.tnvc.com/faq/"
Yet despite almost every site saying how important it is, very few sites actually mention what the system gain is of different NODs.
As a rule of thumb, system gain used to be around 1/10th of the tube gain, but this doesn't really take other factors, such as lens choice, into consideration. To be militarily useful, it is generally considered that a NOD should have a system gain value of around 2000x to 3000x minimum. *1
So what should you expect? And how do you calculate the system gain of systems that do not have their gain listed?
Well, here's some well known "Typical" system gain levels. *2
AN/PVS-5 - 800
AN/AVS-6 Omni I&II - 2000
AN/AVS-6 Omni III - 2000
AN/AVS-6 Omni IV - 3000
AN/AVS-6 Omni VI - 5000
AN/PVS-14 Omni VI - 7000 (*3)
So although it's common to hear about system resolution. It's not all that common to see examples.
Additionally, because tube resolution and MTF drop off significantly as light levels deteriorate past the point where the tube can amplify the image to full brightness *2, the system gain is a very important component of understanding how well a particular night vision device will perform under extremely dark conditions.
For example, at moonlight levels, the PVS-5 has typical system resolution of .72cy/mr while at starlight levels, the PVS-5 has typical system resolution of around .4cy/mr - nearly half of what it can provide under starlight, while even Gen3 tubes of Omni I era don't fare much better *2.
Yet looking through the examples and taking lenses into consideration gives an interesting comparison.
Based on F-numbers, the F-numbers of those NODs are;
AN/PVS-5 - F1.1
AN/AVS-6 Omni I&II - F1.2
AN/AVS-6 Omni III - F1.2
AN/AVS-6 Omni IV - F1.2
AN/AVS-6 Omni VI - F1.2
AN/PVS-14 Omni VI - F1.2
Additionally, there is not a great deal of different in the ocular lens system of most common NODs either.
This information, when coupled with the system gain, leads to an obvious conclusion. If lenses have not significantly changed, the the most common factor in determining the system gain is the tube. As can be seen from the above tables, the gains in System gain since Omni I have been considerable.
Using this information, it is therefore possible to approximate the system gain of any system if you know the era of the tube and the F-number of the objective lens.
To do this, you need to also understand what the F-number of a lens is. Simply explained, it's a ratio of the lens diameter to it's focal length. As a result, a lens that was 26mm in diameter and 26mm focal length would have an F-number of 1.0 while a lens that was 12mm in diameter with a 26mm focal length would have an F-number of 2.0.
Now you remember basic area calculations don't you? You have to square the diameter of a lens to understand it's area and it's the same for F-numbers. To get it's relative area, just square it -
So a F-number of 1.2 would leave a relative area of 1.44
Let's compare that to say, a 3x adapter for a PVS-14, which has a F-number of 1.5. Squared, this has a relative area of 2.25. This means the ratio of F1.2 to F1.5 is 0.64.
As a result, you can assume that when you put a 3x adapter on a PVS-14, you reduce the maximum system gain by a factor of 0.64, or in other words, the maximum system gain is now around 4480, or 4500 if you like round numbers.
Likewise, if a riflescope has a F1.7 lens such as a D760 and a MX10160A tube of Omni VI era or later ( Omni VI exceeds all current Milspecs, including Omni VIII ) then you can likewise work out it's performance relative to a PVS-14.
Eg, The relative area factor is 0.49 ( 0.5 ) so the maximum system gain is still around 2500 - within acceptable Military levels.
Now is this all? Not quite - it's actually T-numbers that matter, not F-numbers. T-numbers actually measure how much light gets through a lens and take into consideration internal obstructions within the lens and other factors, but because the maximum a T-number can be is the F-number in a perfect lens, the T-number will never be higher and usually follows fairly closely to the F-number. They are also harder to find out for lenses, so for all practical purposes, F-numbers are good enough for our rough calculations.
Putting this information to use is just as easy as the above calculations. Recently I did a quick analysis of two scopes - the AN/PVS-27 and the CNVD-LR - both exceptional and well respected clip-on NODs.
The AN/PVS-27 has a F-number of 1.0 (*6)- one of the fastest of any NOD out of the factory. The CNVD-LR has a F-number of 1.85 (*7) - one of the slower NODs out of the factory. For this reason, they make for good calculations and examples of the importance of system resolution.
Using the above "rule of thumb" we can make an approximation that the max system gain of the PVS-27 is around 10,000 while the max system gain of the CNVD-LR, assuming the same high-quality MX-11769 of Omni VI or later design, is around 2900 - still at military levels, but barely scraping past Gen III Omni I and Omni II levels.
The final piece of the puzzle to understanding this subject is how this affects the final image quality and what significance does it have to someone using both systems. Well, there's a specification that's important to remember here: Screen Brightness. On tubes it's always specified as a maximum. That's because a tube will automatically reduce it's gain when there's enough light that it doesn't need to amplify it a lot, or when amplifying to the maximum level would exceed the maximum screen brightness.
The important thing to remember then is that the system gain only becomes important when it's at maximum, because the screen automatic brightness control will reduce the tube gain when it's not dark enough.
This means that two devices like the AN/PVS-27 and the CNVD-LR with the same tube will actually perform at the same level under high brightness conditions - and in fact, due to the better system resolution specification, the CNVD-LR actually outperforms the AN/PVS-27 under such conditions. by about 20% (*8)
A good example of a high-light comparison is this image: Resolution aside, similar performance gain-wise is present in both Gen2 PVS-5 images and the Gen3 MX10160A Omni V image.
However, as it get's darker, usually anything below starlight- such as when you're in a barren landscape or when there's dark clouds overhead, things will change. At the point at which the AN/PVS-27 runs out of system gain to maintain screen brightness, the difference will be that the resolution of the CNVD-LR is about half of that of the AN/PVS-27 and th AN/PVS-27 will be somewhere between three and four times brighter than the CNVD-LR.
A good example of this is a standard "starlight" image from both a Gen II PVS-5 and a GenIII MX10160A Omni V system.
PVS-5 - system gain approx 1000x
Gen III - system gain approx 3000x
Because of this, partly due to the reduced resolution and partly because of the lower brightness, you can also assume that the range of the PVS-27 will be more than double that of the CNVD-LR.
Something to think about when comparing scopes visually - no matter who or where you get it from.
So in evaluating the two, how do you know if you're at a reasonable darkness level? Especially as humans can't really tell the difference after about 1/4 moon? Simply use your best judgment based on what you can see and do your test on a starlight only night with heavy cloud cover. If the moon tables (*7) show any moon in the sky or within 1 hour of the horizon, it's going to be bright. If you can see the clouds as bright against the dark sky, it's too bright ( reflected light from nearby ground-mounted lights ). If neither of these factors are there and the clouds look like inky blots against the sky, conditions are perfect for your test. Just make sure equivalent tubes are in both.
Then after evaluating the difference between scopes, consider how dark it's likely to be when you're using your NOD and weigh that into the factor.
This is a test you can apply to any NOD comparison, BTW, not just those two - Just do the maths, work out the difference in system gain and take it from there -
And if you have any questions, please ask. That's what a forum is for
Regards
David
References:
*1: Many sites list this, eg;
http://www.morovision.com/terminology.htm
and
http://www.tnvc.com/faq/
eg: "U.S. military systems operate at 2,000 to 3,000"
*2 Past, Present and Future Night Vision Goggles in US Army Aviation
http://ftp.rta.nato.int/public//PubFullText/RTO/MP/RTO-MP-019///$MP-019-32.PDF
*3 - PVS-14 system gain typically ranges from 0.25 to 7000 depending on gain setting on a modern tube.
*4 - Approximate F-numbers for well known NV - Information provided by Litton ( Now L3 ) and ITT in public documentation - ITAR free.
*5 Information taken from datasheets for CNVD-LR -
http://nightvision-thermal.com/pdf/nitehog_cnvd-lr.pdf
*6 Information taken from data published on the AN/PVS-27 MUNS -
http://www.ownthenight.com/catalog/i177.html
*7 Moon phases can be obtained online: eg,
http://www.morovision.com/moonphases.htm
*8 - As per this thread where I previously calculated the difference between these scopes.
http://www.snipershide.com/forum/ubbthreads.php?ubb=showflat&Number=3118915#Post3118915
Additional-
High light comparison - PVS-5 and Gen III Omni V
http://aunv.blackice.com.au/userfiles/david-small-223.jpg
Low light comparison - PVS-5 and Gen III Omni V
Gen II PVS-5
http://aunv.blackice.com.au/userfiles/david-IMG_0474-800.jpg
Approximate System Gain: 1000x
Gen III Omni V.
http://aunv.blackice.com.au/userfiles/david-IMG_0406-800.jpg
Approximate System Gain: 3000x
