Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

EBI vs SNR and thoughts on some specs

  • Please log in to reply
29 replies to this topic

#1 cmormando

cmormando

    Sputnik

  • -----
  • topic starter
  • Posts: 47
  • Joined: 14 Sep 2023
  • Loc: South Central PA

Posted 23 January 2024 - 08:20 PM

Equivalent background illumination or signal to noise ratio, what do you guys think is more important? 

 

AND...

 

How would you guys feel about a tube with the following specifications? For a GEN III thin film green phosphor 

 

Photocathode sensitivity: 2280

Halo: 0.8 

EBI: 0.25

SNR: 27 

Gain: 66000

Resolution: 72 


Edited by cmormando, 23 January 2024 - 08:37 PM.


#2 chemisted

chemisted

    Apollo

  • *****
  • Posts: 1,116
  • Joined: 24 Feb 2012

Posted 23 January 2024 - 08:54 PM

Equivalent background illumination or signal to noise ratio, what do you guys think is more important? 

 

AND...

 

How would you guys feel about a tube with the following specifications? For a GEN III thin film green phosphor 

 

Photocathode sensitivity: 2280

Halo: 0.8 

EBI: 0.25

SNR: 27 

Gain: 66000

Resolution: 72 

If it is an Elbit tube, it will be an excellent performer. 



#3 Souldrop

Souldrop

    Vostok 1

  • *****
  • Posts: 128
  • Joined: 16 Feb 2020
  • Loc: Southern Middle TN, USA

Posted 23 January 2024 - 09:38 PM

People get too caught up in chasing specs imo. You could do better and you could also do worse. As long as you are happy with the spots/cosmetics of the tube I personally think thats a solid performer and gets you solidly in the game.

I do suggest doing a price check in your market to ensure you’re paying a fair price.

Edited by Souldrop, 23 January 2024 - 09:39 PM.


#4 cmormando

cmormando

    Sputnik

  • -----
  • topic starter
  • Posts: 47
  • Joined: 14 Sep 2023
  • Loc: South Central PA

Posted 23 January 2024 - 10:03 PM

People get too caught up in chasing specs imo. You could do better and you could also do worse. As long as you are happy with the spots/cosmetics of the tube I personally think thats a solid performer and gets you solidly in the game.

I do suggest doing a price check in your market to ensure you’re paying a fair price.

Cosmetics are excellent. I’m not chasing specs, but I understand what you’re saying. It’s a tube I already have (non astronomy related purchase and paid 2600$), so I’m curious to know if there are better tubes out there and where this one stands. I have nothing to compare it to. 



#5 Souldrop

Souldrop

    Vostok 1

  • *****
  • Posts: 128
  • Joined: 16 Feb 2020
  • Loc: Southern Middle TN, USA

Posted 24 January 2024 - 01:24 AM

Cosmetics are excellent. I’m not chasing specs, but I understand what you’re saying. It’s a tube I already have (non astronomy related purchase and paid 2600$), so I’m curious to know if there are better tubes out there and where this one stands. I have nothing to compare it to.


Ah gotcha. Your tube is likely comparable to most mil spec gen 3 green tubes being produced (US). Maybe slightly low SNR, but your EBI is good. Starting a few years ago L3 (and to some extent elbit) really started putting out some impressive tubes.

If looking to upgrade it shouldn’t be too hard to find tubes that have 36+ SNR and EBI less than 0.7. My personal preference is to sacrifice EBI for SNR as it feels more “glasslike” with higher SNR. (I should make the disclaimer astronomy usage is not my primary reason for owning NV so it’s easier to justify sacrificing some EBI for SNR)

#6 chemisted

chemisted

    Apollo

  • *****
  • Posts: 1,116
  • Joined: 24 Feb 2012

Posted 24 January 2024 - 06:48 AM

Cosmetics are excellent. I’m not chasing specs, but I understand what you’re saying. It’s a tube I already have (non astronomy related purchase and paid 2600$), so I’m curious to know if there are better tubes out there and where this one stands. I have nothing to compare it to. 

There is an important fact that people on this forum seem to have missed so I will highlight it again.  Photocathode sensitivity of Elbit tubes is expressed in units of microamps per Watt while other manufacturers use units of microamps per lumen.  I'm pretty sure you have an Elbit tube since you quote EBI to two decimal places.  That makes it a near clone of the green phosphor tube that GeezerGazer used in his head-to-head comparison of a top end L3 tube and an F9800 tube which showed they both performed the same.  Read the last two posts of the following thread including the link to the 2020 comparison:  https://www.cloudyni...fication/page-2

 

On your data sheet you will have a value for Radiant @ 830nm.  I consider this a better determinant of tube response and have provided an analysis in the following thread:  https://www.cloudyni...-radiant-830nm/

 

So, here's the summary.  The tube you own matches or exceeds the key specs of the green phosphor tube that Ray showed equaled a top end L3 tube.  In the following brackets are your values compared to the green phosphor tube that Ray used [SNR: 27 vs 28.5; PS: 2280 vs 2274; Gain: 66K vs 58K; Halo 0.8 vs 1.1 and EBI: 0.25 vs 0.93]. You have a truly excellent tube.  Congratulations and get out there and explore the night sky with it.  


Edited by chemisted, 24 January 2024 - 06:52 AM.

  • cmormando likes this

#7 sixela

sixela

    James Webb Space Telescope

  • *****
  • Posts: 17,658
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 24 January 2024 - 08:12 AM

Screenshot from 2024-01-24 14-19-20.png

 

doesn't seem to be a typo (since the energy and the luminous flux of the tungsten lamp are inherently tied --that's why you use a standard source-- the units in which to express the luminous flux is something you can pick). And yes, that means that Elbit could actually be better than it seems when you look at just the number.

 

I haven't found what the "W" refers to. It certainly ain't the W of a tungsten bulb source used to produce X lumen (these will produce roughly 15 lm per Watt).

 

Perhaps Elbit doesn't want you to compare the numbers with L3. Why these two companies don't both specify radiant sensitivity at 830 or 880 nm is also anyone's guess.


Edited by sixela, 24 January 2024 - 09:08 AM.


#8 chemisted

chemisted

    Apollo

  • *****
  • Posts: 1,116
  • Joined: 24 Feb 2012

Posted 24 January 2024 - 10:30 AM

 

Perhaps Elbit doesn't want you to compare the numbers with L3. Why these two companies don't both specify radiant sensitivity at 830 or 880 nm is also anyone's guess.

Actually, it looks to be the other way around.  By not providing Radiant @ 830nm numbers, L3 doesn't want you to compare their tubes to Elbit.  

 

If you divide the 830 nm value by 2 you can get a pretty good approximation of the 880 nm value.  For example, I have both numbers for my F9800YG tube.  They are 289 mA/W and 148 mA/W for 830 and 880 values respectively.



#9 cmormando

cmormando

    Sputnik

  • -----
  • topic starter
  • Posts: 47
  • Joined: 14 Sep 2023
  • Loc: South Central PA

Posted 24 January 2024 - 11:03 AM

There is an important fact that people on this forum seem to have missed so I will highlight it again.  Photocathode sensitivity of Elbit tubes is expressed in units of microamps per Watt while other manufacturers use units of microamps per lumen.  I'm pretty sure you have an Elbit tube since you quote EBI to two decimal places.  That makes it a near clone of the green phosphor tube that GeezerGazer used in his head-to-head comparison of a top end L3 tube and an F9800 tube which showed they both performed the same.  Read the last two posts of the following thread including the link to the 2020 comparison:  https://www.cloudyni...fication/page-2

 

On your data sheet you will have a value for Radiant @ 830nm.  I consider this a better determinant of tube response and have provided an analysis in the following thread:  https://www.cloudyni...-radiant-830nm/

 

So, here's the summary.  The tube you own matches or exceeds the key specs of the green phosphor tube that Ray showed equaled a top end L3 tube.  In the following brackets are your values compared to the green phosphor tube that Ray used [SNR: 27 vs 28.5; PS: 2280 vs 2274; Gain: 66K vs 58K; Halo 0.8 vs 1.1 and EBI: 0.25 vs 0.93]. You have a truly excellent tube.  Congratulations and get out there and explore the night sky with it.  

Indeed, it is an elbit. Thank you for this detailed and helpful response! 



#10 Souldrop

Souldrop

    Vostok 1

  • *****
  • Posts: 128
  • Joined: 16 Feb 2020
  • Loc: Southern Middle TN, USA

Posted 24 January 2024 - 11:10 AM

To OP -  I agree with Ed that I would try to enjoy the tube you have unless you just find something lacking. Though I do stand by what I would look for if wanting to upgrade. Unless you're looking at a major improvement to SNR (~10 SNR) I think any upgrades will feel lackluster for most given the price difference (unless you find a good deal somewhere). 

 

There is an important fact that people on this forum seem to have missed so I will highlight it again.  Photocathode sensitivity of Elbit tubes is expressed in units of microamps per Watt while other manufacturers use units of microamps per lumen.  I'm pretty sure you have an Elbit tube since you quote EBI to two decimal places.  That makes it a near clone of the green phosphor tube that GeezerGazer used in his head-to-head comparison of a top end L3 tube and an F9800 tube which showed they both performed the same.  Read the last two posts of the following thread including the link to the 2020 comparison:  https://www.cloudyni...fication/page-2

 

On your data sheet you will have a value for Radiant @ 830nm.  I consider this a better determinant of tube response and have provided an analysis in the following thread:  https://www.cloudyni...-radiant-830nm/

 

So, here's the summary.  The tube you own matches or exceeds the key specs of the green phosphor tube that Ray showed equaled a top end L3 tube.  In the following brackets are your values compared to the green phosphor tube that Ray used [SNR: 27 vs 28.5; PS: 2280 vs 2274; Gain: 66K vs 58K; Halo 0.8 vs 1.1 and EBI: 0.25 vs 0.93]. You have a truly excellent tube.  Congratulations and get out there and explore the night sky with it.  

 

I have a few NVDs on hand with spec sheets and am slowly working on trying to write up/capture some comparisons for potential buyers (or other interested parties) for both terrestrial use and astronomy use. PCS is actually one I am struggling to capture well; I personally feel SNR and EBI are the easier specs to talk about as they both are easier to simply and qualitatively describe what difference in specs look like visually. PCS...that one is more difficult and in my experience is difficult to call it a smoking gun for tube performance. For example 2 of the L3 unfilmed tubes have the following specs:

 

Tube 1: 
PCS: 2159

EBI: 0.9

SNR: 35.9

Halo: 0.7

 

Tube 2:

PCS: 2717

EBI: 1.5

SNR: 36.7
Halo: 0.7
 

They seem more alike than different to my eyes.


  • cmormando likes this

#11 sixela

sixela

    James Webb Space Telescope

  • *****
  • Posts: 17,658
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 24 January 2024 - 04:02 PM

 

 

If you divide the 830 nm value by 2 you can get a pretty good approximation of the 880 nm value. 

That is not necessarily valid across all GaAs photocathode types. They don't all have the same spectral sensitivity curve.



#12 WheezyGod

WheezyGod

    Apollo

  • -----
  • Posts: 1,352
  • Joined: 07 Jun 2021
  • Loc: Massachusetts, USA

Posted 25 January 2024 - 07:30 PM

To OP - I agree with Ed that I would try to enjoy the tube you have unless you just find something lacking. Though I do stand by what I would look for if wanting to upgrade. Unless you're looking at a major improvement to SNR (~10 SNR) I think any upgrades will feel lackluster for most given the price difference (unless you find a good deal somewhere).


I have a few NVDs on hand with spec sheets and am slowly working on trying to write up/capture some comparisons for potential buyers (or other interested parties) for both terrestrial use and astronomy use. PCS is actually one I am struggling to capture well; I personally feel SNR and EBI are the easier specs to talk about as they both are easier to simply and qualitatively describe what difference in specs look like visually. PCS...that one is more difficult and in my experience is difficult to call it a smoking gun for tube performance. For example 2 of the L3 unfilmed tubes have the following specs:

Tube 1:
PCS: 2159
EBI: 0.9
SNR: 35.9
Halo: 0.7

Tube 2:
PCS: 2717
EBI: 1.5
SNR: 36.7
Halo: 0.7

They seem more alike than different to my eyes.


PCS should produce a brighter output at a given level of gain. The EBI is likely offsetting the improvements I would expect you might see when viewing nebula. Have you tried comparing the two when looking at globular clusters?
  • Souldrop likes this

#13 sixela

sixela

    James Webb Space Telescope

  • *****
  • Posts: 17,658
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 26 January 2024 - 03:59 AM

No, a higher PCS does not yield a brighter image at the same gain.

Rather, it’s one of the factors that determines the gain (and SNR, since the MCP needs less electron multiplication to deliver the same total gain if the PCS is high). Look at the units for gain; light at both sides, not current at one side and light at the other.

A high PCS also means a high QE and that also lowers the shot noise from the input at low illumination levels, but at these levels the EBI also becomes relevant.

Edited by sixela, 26 January 2024 - 04:06 AM.

  • cnoct and Souldrop like this

#14 Souldrop

Souldrop

    Vostok 1

  • *****
  • Posts: 128
  • Joined: 16 Feb 2020
  • Loc: Southern Middle TN, USA

Posted 26 January 2024 - 02:48 PM

Wheezy I’ve tried on a couple different targets and uses. My notes mainly mentioned the difference in EBI being the more likely discriminator between the two performance wise. The higher pcs tube is slightly “calmer” than the slight difference in snr would imply, but I wrote it off as some of the noise being lost in the EBI haze.

I will be revisiting astronomy comparisons in the coming months and will try to make more detailed/critical notes.

#15 Mauro Da Lio

Mauro Da Lio

    Viking 1

  • -----
  • Posts: 838
  • Joined: 12 Sep 2004

Posted 27 January 2024 - 05:26 AM

As long as one does not know the spectral responses, that improvement might happen in the IR, where a lot of airglow is present (in this case it would be a worsening) or it might not affect the H-alpha line (in which case it would be a waste of money).



#16 Mauro Da Lio

Mauro Da Lio

    Viking 1

  • -----
  • Posts: 838
  • Joined: 12 Sep 2004

Posted 28 January 2024 - 08:05 AM

The question of what are the concrete improvements on astronomical applications and what are the costs is well posed and I would say they are within the broad objectives of this topic.

 

Specs, and in particular SNR are gross figures that do not necessarily translate into user experience under the sky. 

 

For h-alpha what matters is the improvement on that particular spectral line and for broadband objects under dark sky, what matters is the response once the IR airglow is cut out.

 

We do not know how SNR changes with passband filters. If one device obtains higher specs thanks to, say, improved sensitivity in 1000-1200 nm, that may be an advantage for terrestrial an military applications, but it is a disadvantage for astronomy because those wavelengths are plagued by airglow. 

 

That is why Alexis compared the old and new device in B, OIII and H-alpha in post 21, concluding that the spectral sensitivity look to be similar.

 

However, granted that the spectral sensitivity is the same in the two devices, it still remains to clarify what is the visual experience improvement in H-alpha and, say, galaxies, and how much that costs.

 

We all emphasize characteristics like EBI and SNR, but to be honest it is not clear what the neat visual effect is. In fact this thread asks whether it is better a lower EBI or a higher SNR.


Edited by Mauro Da Lio, 28 January 2024 - 08:06 AM.


#17 sixela

sixela

    James Webb Space Telescope

  • *****
  • Posts: 17,658
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 January 2024 - 03:00 PM

The visual experience in H-alpha seems to be finer detail (due to the better SNR and resolution) and a blacker background for the really faint detail (due to the better EBI), even though that has no effect on brighter portions of the objects (and when you use less aggressive filters or are in a more polluted site, the EBI difference is also less relevant, although I'm surprised that with 3nm filters you can still see the EBI difference in winter Bortle 8 skies. But then, even if both EBIs are low, there is a factor of three difference...).

 

On galaxies, the better EBI is completely irrelevant, but the better SNR does help to see more detail, both finer detail in the bright high contrast regions and larger detail with less contrast. It's all very subtle, I'd say, but it's there.

 

That said, there seems to be a better MTF in the new tube that is not necessarily a result of the better SNR (given the way it's measured, it won't measure MTF at large spatial frequencies).

 

The new tube also has a different phosphor, so perhaps the phosphor screen yields a better image. That doesn't seem to be something that you can attribute to the different price bins of the OVNI-Ms, that's probably an evolution in technology in the underlying device.

 

One thing is sure: even though the improvement is quite subtle (so if you want the most "bang for the buck" then a device with SNR of 30-32 might be better for you, and 64 lp/mm is quite enough, so a "2000 FOM" device might be just fine as long as the high FOM isn't because of an irrelevant higher resolution in lp/mm), but if I'd have to make a "Sixela GUTTS" score I probably would not take the log of the SNR numbers. 37.6 is quite obviously better than 30, much better than the roughly 1dB difference would make you believe once you express it in dB.

 

What comparing different technologies has shown me, though (in my case, recently, Green phsophor Photonis, green phosphor Elbit and white phosphor OVNI-M by Harder) is that each brand of device has a type of noise and image quality that is just different and not expressed in the milspec numbers. I could simply not quantify it. As one experiment that Mauro made me perform shows, even putting a blue filter behind the white phosphor NVD changes the subjective "feel" of the noise.


Edited by sixela, 28 January 2024 - 05:58 PM.

  • Souldrop likes this

#18 chemisted

chemisted

    Apollo

  • *****
  • Posts: 1,116
  • Joined: 24 Feb 2012

Posted 28 January 2024 - 03:27 PM

 

 

. . . but if I'd have to make a "Sixela GUTTS" score I probably would not take the log of the SNR numbers. 37.6 is quite obviously better than 30, much better than the roughly 1dB difference would make you believe once you express it in dB.

Since you referred to my algorithm I will respond.  You are clearly being disingenuous as GUTTS is composed of SNR, Photocathode Sensitivity and Gain (with a hat tip to EBI).  Your new tube has vast improvements in the first two of these: 37.6 vs 30 and 2370 vs ~2000.  You haven't posted the gain of the new tube but it seems likely it is also larger as your first tube was ~58K.  Quite honestly, if you do not take the log of SNR your calculated improvement would far exceed the improvement in visual details that you have so nicely summarized.


Edited by chemisted, 28 January 2024 - 03:29 PM.


#19 sixela

sixela

    James Webb Space Telescope

  • *****
  • Posts: 17,658
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 28 January 2024 - 04:47 PM

I don't believe that the difference in PCS is what I'm seeing.

 

Personally I would place a higher emphasis on SNR and a lower one on PCS (since it's not a system metric but a photocathode metric, a high one should be reflected in higher gain and SNR which are already in the formula).

 

And no, I don't think it's disingenuous of me to say that I PERSONALLY would do things differently (I didn't name it "Sixela GUTTS" for nothing); you're perfectly entitled to stick to your guns, but I'd place a higher emphasis on SNR in a combined scoring metric, that's all.

 

The improvement in gain, BTW, on the new device is minimal. But then I rarely observe at full gain anyway (although I do tend to use slightly higher gain with the new device).

 

You're right that the improvement is probably not 30 vs 37. But it's nog log(30) vs. log(37) either, it seems to be somewhere in between. Unless, of course, a log is indeed what you need for the direct influence of better SNR but there's another intangible that improves the views and correlates with SNR.

 

The SNR used in milspec rating, BTW, is a rather crude metric, meant to be measured easily, but perhaps not the ideal one for our uses. Not that I have anything better or that we stand any chance of getting something more descriptive from manufacturers.


Edited by sixela, 28 January 2024 - 05:56 PM.


#20 DeepSky Di

DeepSky Di

    Skylab

  • *****
  • Moderators
  • Posts: 4,473
  • Joined: 15 Aug 2020

Posted 28 January 2024 - 05:54 PM

Everyone - please stay on topic and answer the OP's question. Any comments on other topics should be posted on those other topics, not this one. Clear skies!



#21 a__l

a__l

    Fly Me to the Moon

  • -----
  • Posts: 5,024
  • Joined: 24 Nov 2007

Posted 28 January 2024 - 09:12 PM

 

 

That is why Alexis compared the old and new device in B, OIII and H-alpha in post 21, concluding that the spectral sensitivity look to be similar.

 

 

I don't see post 21?


Edited by a__l, 28 January 2024 - 09:56 PM.


#22 a__l

a__l

    Fly Me to the Moon

  • -----
  • Posts: 5,024
  • Joined: 24 Nov 2007

Posted 28 January 2024 - 09:25 PM

The question of what are the concrete improvements on astronomical applications and what are the costs is well posed and I would say they are within the broad objectives of this topic.

 

Specs, and in particular SNR are gross figures that do not necessarily translate into user experience under the sky. 

 

 

I agree! That's why I raised this question. Unfortunately, I don't see him either.


Edited by a__l, 28 January 2024 - 09:57 PM.


#23 a__l

a__l

    Fly Me to the Moon

  • -----
  • Posts: 5,024
  • Joined: 24 Nov 2007

Posted 28 January 2024 - 09:32 PM

"The new tube"

 

sixela, in post #17 everything is blurry. I would like specifics in the dark sky. Therefore, I have to repeat the question about this, if you have a good observant report for 20", I will definitely compare it with my tube on my 24" telescope.

The best comparison is glass - new NV. Without narrowband filters, which are more or less understandable.


Edited by a__l, 29 January 2024 - 12:33 AM.


#24 sixela

sixela

    James Webb Space Telescope

  • *****
  • Posts: 17,658
  • Joined: 23 Dec 2004
  • Loc: Boechout, Belgium

Posted 29 January 2024 - 03:18 AM

I wasn’t under the impression you could order me around to do tests for you.

Of course in due course there will be observing reports; I tend to always observe objects both with glass and NVD (at least galaxies, and it’s Galaxy Season!)

“Specifics in dark sky” are rather harder to get close to Full Moon ;-).

Edited by sixela, 29 January 2024 - 03:18 AM.


#25 Mauro Da Lio

Mauro Da Lio

    Viking 1

  • -----
  • Posts: 838
  • Joined: 12 Sep 2004

Posted 29 January 2024 - 08:45 AM

 

What comparing different technologies has shown me, though (in my case, recently, Green phsophor Photonis, green phosphor Elbit and white phosphor OVNI-M by Harder) is that each brand of device has a type of noise and image quality that is just different and not expressed in the milspec numbers. I could simply not quantify it. As one experiment that Mauro made me perform shows, even putting a blue filter behind the white phosphor NVD changes the subjective "feel" of the noise.

I found that rods have longer latency. They integrate light over about 100 ms. That lets them "average" the noise. https://en.wikipedia.org/wiki/Rod_cell

 

When using a blue filter after the NVD one probates shifts the importance of the rods vs cones in the mesonic range.

 

Concentrino SNR, I said elsewhere that the phosphors used in the military NVDs are faster than the optimal latency for astronomy. Phosphor with longer integration time would smooth the noise and obtain higher SNR at the price of slower response time (which does not matter in astronomy). 

 

You said the phosphor in the new device is different... maybe...




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics