Vixen VSD100 f3.8 5-element Apochromat

[Adam J]

1 hour ago, Rodd said:

And I said dont add a reducer thinking you will reduce your imaging time 2-3x.  2.8x is 2-3x in my book. 

Yes it is... but that's clearly not what I said. Read above, what I said was that a focal reducer of 0.8x will only provide a 1.56x increase in optical speed, 2.8x was with referance to the delta between F3 and F5. Hence in my book 1.56x is very different from 2-3x. 

1 hour ago, Rodd said:

Its just not so.  First of all, I went from F5 to F3.  For each stop its 2x faster.  if your telling me its 1.8...no difference,  I just don't see it. 

Well F-stop is a non linear scale used on camera lens. So to be clear moving from F5 to F3 doesn't equal 2 f-stops at all. 

The scale is as follows:

f1.4, f2, f2.8, f4, f5.6, f8, f11, f16

Hence:

(2/1.4)^2 = 2 = one f-stop 

(16/11)^2 = 2 = one f-stop

(2.8/1,4)^2 = 4 = Two f-stops

(8/4)^2 =4 = Two f-stops

1 hour ago, Rodd said:

And an Epsilon 180 IS 4x faster than an FSQ 106--no question--regardless of the FSQs focal ratio Why....BECAUSE OF THE APERTURE.

Sorry but I cant agree with that, its just not the case. If your using the same camera between those two scopes optical speed is determined by f-ratio. 

1 hour ago, Rodd said:

The proof is I still need 15-20 hours of integration regardless of whether I use a reducer or not.

Ok here is the thing...what if I told you that I would not necessarily expect you to be able to tell the difference in SNR between those two images by visual inspection alone?

Lets say that you take a 20hour image without a reducer and a 20 hour image with a reducer. We have shown above that the expected difference in optical speed is about a factor of 1.56x 

So for those two images to be of identical SNR your would expect to need 13 hours of exposure with the reducer and 20 hours without it. 

However the thing is that the way you perceive noise is non-linear and so its a well known rule of thumb that to make a significant impact on image quality you need to double the total exposure time. 

Hence to make your reduced image significantly better than your non reduced image you will need to expose for 26 hours not 20 hours. 

Looking at that the other way around if you want your non-reduced 20 hour image to be significantly better than the reduced 20 hour image then you will need to exposure for 40 hours. So 6 hours more vs 20 hours more integration with a reducer vs without before your see significant differences. 

But if both are 20 hours you are at the edge of being able to perceive the difference via simple visual inspection. Simple changes in imaging conditions and variable processing could easily obscure the SNR improvement. All things being equal the visual difference will be slight but probably better measured by software. 

So as I said above, everything is as expected and the main issue is that you have over inflated expectations of what the reducer will give you leading to disappointment. 

 

I think I am going to leave it here Rodd as if that has not convinced you nothing will. 

 

Best wishes, 

 

Adam 

 

 

 

Edited September 20, 2019 by Adam J

[Rodd]

14 minutes ago, Adam J said:

Yes it is... but that's clearly not what I said. Read above, what I said was that a focal reducer of 0.8x will only provide a 1.56x increase in optical speed, 2.8x was with referance to the delta between F3 and F5. Hence in my book 1.56x is very different from 2-3x. 

Well F-stop is a non linear scale used on camera lens. So to be clear moving from F5 to F3 doesn't equal 2 f-stops at all. 

The scale is as follows:

f1.4, f2, f2.8, f4, f5.6, f8, f11, f16

Hence:

(2/1.4)^2 = 2 = one f-stop 

(16/11)^2 = 2 = one f-stop

(2.8/1,4)^2 = 4 = Two f-stops

(8/4)^2 =4 = Two f-stops

Sorry but I cant agree with that, its just not the case. If your using the same camera between those two scopes optical speed is determined by f-ratio. 

Ok here is the thing...what if I told you that I would not necessarily expect you to be able to tell the difference in SNR between those two images by visual inspection alone?

Lets say that you take a 20hour image without a reducer and a 20 hour image with a reducer. We have shown above that the expected difference in optical speed is about a factor of 1.56x 

So for those two images to be of identical SNR your would expect to need 13 hours of exposure with the reducer and 20 hours without it. 

However the thing is that the way you perceive noise is non-linear and so its a well known rule of thumb that to make a significant impact on image quality you need to double the total exposure time. 

Hence to make your reduced image significantly better than your non reduced image you will need to expose for 26 hours not 20 hours. 

Looking at that the other way around if you want your non-reduced 20 hour image to be significantly better than the reduced 20 hour image then you will need to exposure for 40 hours. So 6 hours more vs 20 hours more integration with a reducer vs without before your see significant differences. 

But if both are 20 hours you are at the edge of being able to perceive the difference via simple visual inspection. Simple changes in imaging conditions and variable processing could easily obscure the SNR improvement. All things being equal the visual difference will be slight but probably better measured by software. 

So as I said above, everything is as expected and the main issue is that you have over inflated expectations of what the reducer will give you leading to disappointment. 

 

I think I am going to leave it here Rodd as if that has not convinced you nothing will. 

 

Best wishes, 

 

Adam 

 

 

 

There is theory and there is reality.  The two do not match.  Bottom line--if you want to take deep images in 3-4 hours--do not go the FSQ 106 and .6x reducer route.  It will not work.  You don't believe me--try it.

Rodd

[Adam J]

Just now, Rodd said:

There is theory and there is reality.  The two do not match.  Bottom line--if you want to take deep images in 3-4 hours--do not go the FSQ 106 and .6x reducer route.  It will not work.  You don't believe me--try it.

Rodd

I don't have that kind of money. 

[Rodd]

5 minutes ago, Adam J said:

I don't have that kind of money. 

OK--why is an Epsilon so much faster than the FSQ 106 at F3?   Because of aperture.  That is what I am saying.  I am not talking about reducers.  My main point was aperture effects speed more than focal ratio.  But together they effect speed the most.  That is my point.  Reducing a 4" scope to F1 will not be as fast as using a 180mm epsilon at F2.8....because of aperture.   You have taken the point into all this reducer math....thats not my point.

[Adam J]

2 minutes ago, Rodd said:

OK--why is an Epsilon so much faster than the FSQ 106 at F3?   Because of aperture.  That is what I am saying.  I am not talking about reducers.  My main point was aperture effects speed more than focal ratio.  But together they effect speed the most.  That is my point.  Reducing a 4" scope to F1 will not be as fast as using a 180mm epsilon at F2.8....because of aperture.   You have taken the point into all this reducer math....thats not my point.

I understand but there is a difference between pure optical speed and perceived image quality. Sometimes longer focal length alone will result in an increase in perceived image quality as SNR is only one factor, once you reach a certain level of SNR in an image then its no longer the deciding factor. 

Perhaps posting some example images would help demonstrate your point. 

Adam 

[Adam J]

10 minutes ago, Rodd said:

OK--why is an Epsilon so much faster than the FSQ 106 at F3?   

Btw whats the focal length of both those scopes?

 

[Rodd]

8 minutes ago, Adam J said:

I understand but there is a difference between pure optical speed and perceived image quality. Sometimes longer focal length alone will result in an increase in perceived image quality as SNR is only one factor, once you reach a certain level of SNR in an image then its no longer the deciding factor. 

Perhaps posting some example images would help demonstrate your point. 

Adam 

That wouldn't be of much help.  All I know is I bought the FSQ 106 and .6x reducer thinking that shooting at F3 would be similar to using an Epsilon.  I saw great images captured in 2-4 hours of data.  It does not matter if I use my TOA 130 at F7.7, my TOA 130 at F5.38, or my FSQ 106 at F3--I need the same amount of exposure time to make an image that I consider complete, an image with acceptable noise levels and structural detail.  I was taking 20 hour images and I wanted to be able to take images in 1 night instead of 4-5.   It just does not work that way.  I don't care what the theory says.  The primary benefit of  reducing a scope with a reducer is it increases the FOV.  I have not experienced any time savings at all.

Rodd

[Rodd]

14 minutes ago, Adam J said:

Btw whats the focal length of both those scopes?

 

F2.8 and F3--The Epsilon might be 3.2 depending on if its the 130 or 180.  Not sure what the focal length is of the Epsilon  around 500 maybe? The FSQ is F3 (318mm FL)

Edited September 20, 2019 by Rodd

[Adam J]

5 minutes ago, Rodd said:

F2.8 and F3--The Epsilon might be 3.2 depending on if its the 130 or 180.  Not sure what the focal length is of the Epsilon  around 500 maybe? The FSQ is F3 (318mm FL)

430mm for the 130 so F3.3. 

500 for the 180mm at F2.8. 

So several factors may be at play here. Thats a significant difference in focal length between the Epsilons and the FSQ.

What camera are you using? Is it the same on both?

Adam

 

[Rodd]

6 minutes ago, Adam J said:

430mm for the 130 so F3.3. 

500 for the 180mm at F2.8. 

So several factors may be at play here. Thats a significant difference in focal length between the Epsilons and the FSQ.

What camera are you using? Is it the same on both?

Adam

 

It is not possible to dissect this the way you want--I do not have an Epsilon.  Of course the focal length is longer, because the aperture is greater.  Its aperture that is the key--not focal length.   The only way to add more photons is to increase aperture.  Sure, you can spread the photons you have out differently among the sensor you are using, which may result in an apparent difference in speed--a wider FOV may seem to come to image a bit quicker than a restricted FOV with same aperture (Thats all a reducer does is spread out the same number of photons in a different manner on the sensor).  It may increase speed a bit--but not significantly, which is what I was after.    Or--you can change pixel size--thatt might do the same--but a person should not fall into the trap of thinking they will shave 40-50% off their exposure times by switching cameras.    I am tired of my images taking weeks to complete (due to clouds/Moon).  I want to finish them in one 6 hour period.   I posed this question to the forum and we discussed it ad nauseum.  I asked what would be better for this purpose, the RH 200 or the FSQ at F3.   It was a wash.   Now I know that the RH200 would be much faster due to the additional aperture (Don't like RH200 stars though so I am glad I did not go that route).  And I don't like diffraction spikes so I am glad I did not go the Epsilon route.  I am also not partial to hypserstar--so in the end, I guess the answer is a Rifast scope.  Bottom line, from a practical field tested perspective, if you want to significantly reduce the amount of time needed to complete an image without sacrificing quality (by significant I mean 2-3x faster), you simply have got to use more aperture--a reducer will not do it.  I have tested this on 4 scopes--all of which I can reduce by varying amounts.   regardless of which scope and which reducer---my images tend to be 15-25 hours (controlled mostly by the weather--if I have 12 hours and it will be cloudy for weeks--I will call it finished).

Rodd

[Adam J]

1 hour ago, Rodd said:

It is not possible to dissect this the way you want--I do not have an Epsilon.  Of course the focal length is longer, because the aperture is greater.  Its aperture that is the key--not focal length.   The only way to add more photons is to increase aperture.  Sure, you can spread the photons you have out differently among the sensor you are using, which may result in an apparent difference in speed--a wider FOV may seem to come to image a bit quicker than a restricted FOV with same aperture (Thats all a reducer does is spread out the same number of photons in a different manner on the sensor).  It may increase speed a bit--but not significantly, which is what I was after.    Or--you can change pixel size--thatt might do the same--but a person should not fall into the trap of thinking they will shave 40-50% off their exposure times by switching cameras.    I am tired of my images taking weeks to complete (due to clouds/Moon).  I want to finish them in one 6 hour period.   I posed this question to the forum and we discussed it ad nauseum.  I asked what would be better for this purpose, the RH 200 or the FSQ at F3.   It was a wash.   Now I know that the RH200 would be much faster due to the additional aperture (Don't like RH200 stars though so I am glad I did not go that route).  And I don't like diffraction spikes so I am glad I did not go the Epsilon route.  I am also not partial to hypserstar--so in the end, I guess the answer is a Rifast scope.  Bottom line, from a practical field tested perspective, if you want to significantly reduce the amount of time needed to complete an image without sacrificing quality (by significant I mean 2-3x faster), you simply have got to use more aperture--a reducer will not do it.  I have tested this on 4 scopes--all of which I can reduce by varying amounts.   regardless of which scope and which reducer---my images tend to be 15-25 hours (controlled mostly by the weather--if I have 12 hours and it will be cloudy for weeks--I will call it finished).

Rodd

The difference is most likely coming in the sampling and hence the scale of the noise presented in the final image. That is not normally a factor because at a given scale the noise scale will remain fixed, the problem comes when you try to present an image from a longer focal length at the same size as a shorter focal length image. The over sampled image will always look better even if both images have identical SNR. As an example if you take two images of M42 and you present them at full screen one long focal length the other short focal length cropped then the noise can be mathematically identical, however the more course structure of the noise in the critically sampled image will appear more noisy to the eye even if the resolved detail is identical, this would in turn prompt you to go for a longer integration. Hence why you should always image at a FOV that is appropriate to the target, essentially cropping is a bad for perceived SNR independent of image scale.

For this reason seeking a fast scope over and above a scope of an appropriate focal length is counter productive. So you should only have purchased the FSQ if you wanted to image wide field targets with it and present them uncropped and at the original image scale.

Adam

Edited September 20, 2019 by Adam J

[Rodd]

9 minutes ago, Adam J said:

The difference is most likely coming in the sampling and hence the scale of the noise presented in the final image. That is not normally a factor because at a given scale the noise scale will remain fixed, the problem comes when you try to present an image from a longer focal length at the same size as a shorter focal length image. The over sampled image will always look better even if both images have identical SNR. As an example if you take two images of M42 and you present them at full screen one long focal length the other short focal length cropped then the noise can be mathematically identical, however the more course structure of the noise in the critically sampled image will appear more noisy to the eye even if the resolved detail is identical, this would in turn prompt you to go for a longer integration. Hence why you should always image at a FOV that is appropriate to the target, essentially cropping is a bad for perceived SNR independent of image scale.

For this reason seeking a fast scope over and above a scope of an appropriate focal length is counter productive. So you should only have purchased the FSQ if you wanted to image wide field targets with it and present them uncropped and at the original image scale.

Adam

I am not referring to cropping.  Throughout the discussion I am keeping the images at what they are--so the image shot with the TOA 130 at F7.7 of M33 the galaxy fills the screen and for the FSQ at F3 the galaxy is small in the screen.  Or nebulae--or anything else.  I never use a reducer to expand the FOV then crop out the target later thinking it will reduce imaging time (that's a myth).  Sometimes I will crop out interesting features and portray them as images in their own right because with small pixel and the FSQ at F3 the resolution is still pretty good (2.46).  But when I say the time it takes to collect the data for the image I mean the entire FOV.  A good example is NGC 7000.  With the FSQ 106 at F3 I collected 10 hours of data while with the TOA at F7.7 I need about the same to get the same quality.  Now, it so happens that I really needed more than 10 hours so I am collecting more with the TOA 130 at F7.7 because my goal is not a comparison--but to render as good an image as I can.  Yes the  FOV of the FSQ at F3 is huge compared to the TOA at F7.7--but each image is what it is.  The important thing in my opinion, is I failed in my attempt to purchase a fast system.  I base that on my dissatisfaction with that element of it.  I love the setup for other reasons--but speed is not one of them.  If someone asked me what scope should they get so they could finish an image in a single day, I would not advise them get the FSQ and .6x reducer, because focal length on its own is meaningless.  F3 is only faster than F7 if it has more aperture.  You may define speed differently....I define speed as the number of hours my rig is imaging in the middle of winter when it is 10 degrees out side and I have to check focus at 4:00am. 

Case in point of aperture controlling exposure time--the sub lengths of the Liverpool telescope are usually 20-30 sec--and you get a few of them when you download to process the data.  The focal length of that system is over 5,000 and it still is blazing fast beuase of the 42" of aperture.

Rodd

[Adam J]

11 minutes ago, Rodd said:

But when I say the time it takes to collect the data for the image I mean the entire FOV.  A good example is NGC 7000.  With the FSQ 106 at F3 I collected 10 hours of data while with the TOA at F7.7 I need about the same to get the same quality. 

Rodd

Yes but your not resolving detail at the pixel level with the TOA at F7.7 hence you can use more noise reduction in processing without losing detail if your critically sampled with the FSQ your certainly over sampled with the TOA, in effect your lacking efficiency because although both images are good quality how many pannels do you need with the TOA?

Edited September 20, 2019 by Adam J

[Rodd]

36 minutes ago, Adam J said:

The difference is most likely coming in the sampling and hence the scale of the noise presented in the final image. That is not normally a factor because at a given scale the noise scale will remain fixed, the problem comes when you try to present an image from a longer focal length at the same size as a shorter focal length image. The over sampled image will always look better even if both images have identical SNR. As an example if you take two images of M42 and you present them at full screen one long focal length the other short focal length cropped then the noise can be mathematically identical, however the more course structure of the noise in the critically sampled image will appear more noisy to the eye even if the resolved detail is identical, this would in turn prompt you to go for a longer integration. Hence why you should always image at a FOV that is appropriate to the target, essentially cropping is a bad for perceived SNR independent of image scale.

For this reason seeking a fast scope over and above a scope of an appropriate focal length is counter productive. So you should only have purchased the FSQ if you wanted to image wide field targets with it and present them uncropped and at the original image scale.

Adam

OK--you want an example.  here are two images of the same target I shot--one at F4.3 (np101is 8x reducer) and one at F7.7 (TOA 130 native).  They have 23 and 22 hours respectively.  The widefield image needs all of the 23 hours and then some IMO.  Where is the time saving?

F7,7

 

F4.3

 

23 hour22hours

[Rodd]

9 minutes ago, Adam J said:

how many pannels do you need with the TOA?

Irrelevant IMO--at issue is how long it takes to create the image...for what it is.  The images can not be compared in any way except the time it takes to shoot them--which is the same.

Rodd

[Adam J]

16 minutes ago, Rodd said:

Irrelevant IMO--at issue is how long it takes to create the image...for what it is.  The images can not be compared in any way except the time it takes to shoot them--which is the same.

Rodd

Sorry cant agree with that Rodd they are not directly comparable because they are not images of the same thing. One is only a small part of a larger object the other is the entire object. Not to mention the fact that if you only take a image of the bright center area of course it looks better without the dimmer outer stuff

 

32 minutes ago, Rodd said:

Case in point of aperture controlling exposure time--the sub lengths of the Liverpool telescope are usually 20-30 sec--and you get a few of them when you download to process the data.  The focal length of that system is over 5,000 and it still is blazing fast because of the 42" of aperture.

 

If i had a penny for every time that example is used, that telescope has a liquid nitrogen cooled sensor with 90%QE and pixels the size of dinner plates. It performs exactly how you would expect according to the maths, nothing magical about it. Less read noise than a 8300 and about 100 times the pixel area or more. I can dig the spec out and show you that a 30second exposure is about equivalent to half an hour with a amateur camera on the same scope. I calculated that for Olly once in a different thread. 

Edited September 20, 2019 by Adam J

[Rodd]

6 minutes ago, Adam J said:

Sorry cant agree with that Rodd they are not directly comparable because they are not images of the same thing. One is only a small part of a larger object the other is the entire object. Not to mention the fact that if you only take a image of the bright center area of course it looks better without the dimmer outer stuff

As I said before--the images can not be compared to one another AT ALL--only the exposure times are being compared.  Image A for what it is compared to image B for what it is WITH RESPECT TO TEH TIME IT TOOK TO MAKE IT (EXPOSURE TIME).   I find they are about the same---unless you increase the aperture.

Rodd

[Adam J]

3 minutes ago, Rodd said:

As I said before--the images can not be compared to one another AT ALL--only the exposure times are being compared.  Image A for what it is compared to image B for what it is WITH RESPECT TO TEH TIME IT TOOK TO MAKE IT (EXPOSURE TIME).   I find they are about the same---unless you increase the aperture.

Rodd

yes but there is fainter nebulocity in one image than the other due to the fov. While the center so bright its not SNR limited and focal length plays a bigger part in image quality. The real test would be on a dim object like the squid.

Edited September 20, 2019 by Adam J

[Rodd]

12 minutes ago, Adam J said:

Sorry cant agree with that Rodd they are not directly comparable because they are not images of the same thing. One is only a small part of a larger object the other is the entire object. Not to mention the fact that if you only take a image of the bright center area of course it looks better without the dimmer outer stuff

 

If i had a penny for every time that example is used, that telescope has a liquid nitrogen cooled sensor with 90%QE and pixels the size of dinner plates. It performs exactly how you would expect according to the maths, nothing magical about it.

The aperture is much more critical to the exposure time than the nitrogen cooling and camera sensor.

[Adam J]

Just now, Rodd said:

The aperture is much more critical to the exposure time than the nitrogen cooling and camera sensor.

Sorry but its not, put a ASI1600mm pro on the liverpool telescope and you will need to expose for a very long time to reach the same SNR.

[Rodd]

6 minutes ago, Adam J said:

yes but there is fainter nebulocity in one image than the other due to the fov. While the center so bright its not SNR limited and focal length plays a bigger part in image quality. The real test would be on a dim object like the squid.

The real test is any target you want to shoot.  If I spend $10,000 to create an image in 1 night and it still takes me 6 nights over a month--its a failure (with respect to time). 

[Rodd]

5 minutes ago, Adam J said:

Sorry but its not, put a ASI1600mm pro on the liverpool telescope and you will need to expose for a very long time to reach the same SNR.

You don't need the same SNR--you just need to create an image faster than the FSQ 106 at F3 to prove my point.

[henry b]

I also have come  to the same conclusion as Rodd,  I must add it has been an expensive conclusion.  Henry B.

[Adam J]

6 minutes ago, Rodd said:

The real test is any target you want to shoot.  If I spend $10,000 to create an image in 1 night and it still takes me 6 nights over a month--its a failure (with respect to time). 

I do understand the way you feel mate. But put it this way if you wanted to image objects at 300mm focal length you could be using a ZS61 at F6 and that is going to be much slower. It really doesn't work when you compare the scope to a scope of a different focal length. Its Better to think of it as paying for a 100mm lens that operates at 300mm focal length than thinking in terms of f-ratio. your competitor is not a 130mm f7.7 its a 60mm F6......

 

Edited September 20, 2019 by Adam J

[Rodd]

10 minutes ago, Adam J said:

I do understand the way you feel mate. But put it this way if you wanted to image objects at 300mm focal length you could be using a ZS61 at F6 and that is going to be much slower. It really doesn't work when you compare the scope to a scope of a different focal length. Its Better to think of it as paying for a 100mm lens that operates at 300mm focal length than thinking in terms of f-ratio. your competitor is not a 130mm f7.7 its a 60mm F6......

 

The reason it will be slower is there is only 61mm of aperture instead of 106mm--that's the only reason