• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
ZEISS DTI thermal imaging cameras. For more discoveries at night, and during the day.

SV's didn't really blow the little BN's away! (1 Viewer)

If I'm understanding this correctly, without actually seeing the ISO tables, this is a disappointment to me. I think by far the best information about transmission are full spectral transmission curves. "Night" a "day" transmission are already dumbed down from that, but at least as long as I thought they represented two points along the transmission curve they were of some use in making a crude estimation of the value and shape of the curve between and a little to either side of the points. If, however, night and day are weighted values, so that each one is off the plotted curve by some unknown amount, then they become completely useless for estimating the shape of the curve or the true transmission values. Too bad.
 
Last edited:
Hi Ed,

you are right about this. According ISO 14490-5 - Test methods for transmittance - as the relative spectral luminosity factor for photopic (day) vision is used as specified in Table 2 of CIE Publ. 18.2:1983. For scotopic (night) vision it’s the same source, Table 3. The measurements and calculations should be done in steps of 5nm over the whole visible spectrum (370-780nm).

Steve

Thanks, Steve.

What is most misleading, unfortunately, is the meaning of "percent" when applied to the day/night values. It is the percentage of the max 'visible energy' not the max physical energy, but no one explains that as they should.

Something else has dawned on me that seems to be a shame. All of the information necessary to assess color fidelity numerically is available to the analyzer but it isn't reported. Here is what the survey authors say on the subject:

...Colour fidelity.
Perfect colour fidelity from binoculars is sure to be of high importance for demanding birdwatchers. To enable colours to be reproduced perfectly, the transmission spectrum of binoculars must ideally lie as closely as possible within the spectral range of 500-630 nm (blue green-yellow-orange-red). Binocular manufacturers may achieve this by applying various different coatings. The perceived colour fidelity of the different binoculars on test is provided in the table. You can also check this easily yourself by looking at an even white surface through the binoculars with one eye while looking at the same surface directly with the other eye. If both eyes are equally sensitive to colour, you will see any differences in colour directly.

The day-value applies to white light, i.e., the full input spectrum. The same computation, however, could be done for each of the color frequency ranges show below. This would result in seven additional numbers representing the brightness of each color band, which could be presented in either tabular or graphic form. This would certainly improve on the purely subjective methods these authors (and others) use.

Why haven't the test instrument designers included this feature (or similar ones)? Probably because there is not enough demand for it, and probably because the test instrument operator could produce such results by brute force by doing seven analyses, each with a different frequency range. It's up to the survey researchers to ask (and pay) for that. But why don't the manufacturers report it?

Ed
 

Attachments

  • Color spectrum frequency bands.jpg
    Color spectrum frequency bands.jpg
    68.4 KB · Views: 54
Last edited:
If I'm understanding this correctly, without actually seeing the ISO tables, this is a disappointment to me. I think by far the best data about transmission are the full spectral response curves. "Night" a "day" transmission are already dumbed down data, but at least as long as I thought they represented two points along the resonse curve they were of some use in making a crude estimation of the value and shape of the curve between and a little to either side of the points. If, however, night and day are weighted values, so that each one is off the response curve by some unknown amount, then they become completely useless for estimating the shape and value of the curve. Too bad.

Hi Henry,

Yes, the best data about instrument "transmission" are the spectral energy outputs. The best data about what the human perceives (i.e., sees), however, is determined by multiplying that output by the human spectral sensitivity function. Fortunately, that psychophysical function is well documented and standardized, so there is nothing dumbed down.

When you reflect on this more, my friend, you will find that it addresses a great deal about apparent brightness that you are very interested in.

Respectfully,
Ed
 
If I'm understanding this correctly, without actually seeing the ISO tables, this is a disappointment to me. I think by far the best information about transmission are full spectral transmission curves. "Night" a "day" transmission are already dumbed down from that, but at least as long as I thought they represented two points along the transmission curve they were of some use in making a crude estimation of the value and shape of the curve between and a little to either side of the points. If, however, night and day are weighted values, so that each one is completely off the plotted curve by some unknown amount, then they become completely useless for estimating the shape of the curve or the true transmission. Too bad.

Henry,

if the transmission is done according to the above mentioned ISO-standard you have always the raw data for each 5nm step. It's no problem to generate a curve from that. Tau night and tau day (don't have the greek letters here) aren't just the result of two measurements at 505 or 550 nm, but is a kind of mean value that takes into account all calculated values over the whole visible spectrum. The single values are calculated from the raw measured data by two different formulas (sorry, too complicated to write without formula editor), one for day and the other for night sensitivity given by the CIE. Modern manufactores of spectrometers do use a software that does all the calculations automatically based upon measured and saved values. Of course I have no idea how the values quoted in this thread did come into existence.

Steve
 
Hi Steve and Ed,

I'll just have to learn to live with this, but I'm not happy with the full spectrum transmission being reduced to two numbers by any formula. I understand that the original data still exist, but unless they're made available there is no way for those of us who haven't seen them to work back from the day and night figures, which is what I mean by dumbed down. Then there is the question, as Steve mentioned, of whether the ISO standard is always used. I was told by a source at Zeiss that, at least as of a few years ago, their day and night figures were simply 510 and 550 nm and I have no idea what the earlier tests I mentioned in this thread used as a standard.

I've also been perplexed lately by the rather different looking transmission curves being generated by different reviewing sites for the same binocular models as well as some odd anomalies in some of these curves. I always hoped that more real data would put to rest (at least for me) the wide range of subjective reports about "brightness" in different binoculars. But now, alas, it seems that even the measured data are not in agreement.

Henry
 
Last edited:
... I always hoped that more real data would put to rest (at least for me) the wide range of subjective reports about "brightness" in different binoculars. But now, alas, it seems that even the measured data are not in agreement.

Test variability is a chronic problem for a variety of reasons, all of which I'm sure you understand. However, getting past variability in subjective reporting first requires understanding how physical measurements (however variable they may be) relate to subjective responses. Unfortunately, I've done a rather poor job at explaining why transforming the raw transmission data to psychophysical estimates of those subjective responses is an essential requirement for progress.

Hang in there, Henry. :t:

Ed
 
Last edited:
Henry,

Here's a hypothetical question that may help. Assume you had two binoculars with spectral transmissions that differed in only one respect: the percentage numbers shown for frequencies are identical but appear in a completely different order for each binocular. The mean and variance of overall spectral transmittances therefore would be identical.

Would the overall "brightness" of the two binoculars be the same or different for an average human observer?

Ed
 
Last edited:
Henry,
True, two point on a curve is a better representation of that curve than weighted averages of the curve centered at those two points. But, two points on a curve is a totally SUCKY representation of the curve. The weighted averages about 510 and 555 nm, if indeed weighted by the shapes of scotopic and photopic sensitivity are, however, THE BEST representation of how bright a binocualr will look by night and day that I can think of. I see where you're coming from but I'd prefer the weighted average, if we could trust it. As you point out, we can't trust much of anything so the whole thing is king of moot huh. So, cheer up, things are no worse than ever man.
Ron
 
From sept, '10:

Dennis:
"2nd that except add the Nikon 8x30 EII to the mix. It is clearly better than the Leica 8x32 BN. You may prefer the EII to the SE because it has a wider FOV. The Leica 8x32 BN's are definitely not as good as the newer alpha binoculars especially the Zeiss 8x32 FL. Even the Zen Ray 8x42 ED's are better than the Leica's."


Me:
"Dennis, please don't try to convert this into another EII thread. I'm of the thinking that this is about more than the GREATEST VIEW, or THE BEST VIEW FOR THE BUCK, or PORRO PERFECTION, yada yada yada.

Look, we mostly all love the EIIs, OK? I do and recently bought a pair. Very nice alright and I concede, based strictly on the view, is generally better than my 8x32 Trins.

This is a thread about a superior birding binocular. Something greater than the sum of it's parts. Something that can't be reduced to a couple of catch phrases.

The Leicas have so many advantages that, to me, make them a superior birding tool. I started to list them but then thought "what's the point?" Not everyone places ultimate value on that last tiny bit of view, looking at stationary targets, in clean dry conditions, really... not everyone."
 
From sept, '10:

Dennis:
"2nd that except add the Nikon 8x30 EII to the mix. It is clearly better than the Leica 8x32 BN. You may prefer the EII to the SE because it has a wider FOV. The Leica 8x32 BN's are definitely not as good as the newer alpha binoculars especially the Zeiss 8x32 FL. Even the Zen Ray 8x42 ED's are better than the Leica's."


Me:
"Dennis, please don't try to convert this into another EII thread. I'm of the thinking that this is about more than the GREATEST VIEW, or THE BEST VIEW FOR THE BUCK, or PORRO PERFECTION, yada yada yada.

Look, we mostly all love the EIIs, OK? I do and recently bought a pair. Very nice alright and I concede, based strictly on the view, is generally better than my 8x32 Trins.

This is a thread about a superior birding binocular. Something greater than the sum of it's parts. Something that can't be reduced to a couple of catch phrases.

The Leicas have so many advantages that, to me, make them a superior birding tool. I started to list them but then thought "what's the point?" Not everyone places ultimate value on that last tiny bit of view, looking at stationary targets, in clean dry conditions, really... not everyone."

I have changed my opinion on that. I still think the Zeiss 8x32 FL's are a little better optically than the Leica BN's but I prefer the view of the BN's to the Zen Rays now even though they do have a little more CA than the Zen Rays. I definitely prefer the ergos and the build quality of the BN's and of course their size to the Zen Rays. I wouldn't pay twice as much for the FL's though because I don't think they are worth the difference. The BN's represent a sweet spot of great optics, size and compactness and build quality and durability for me. For $600.00 I think they are hard to beat. They are not the top optics out there but they are very good and very close to even my Swarovski SV for alot less money.
 
Henry, Ed and Steve (hinnark);

I had a look at the ISO document Steve referenced since I had it. I had used the portion relating to the transmission curves but had missed the part about day and night calculations, since I did not have the appropriate CIE documents concerning D65 and photopic, scotopic data.

I went this morning and pulled what I had and made a spreadsheet for the equations listed using data I had on hand, they seem to fit close enough that they may be the same data files.

I had some files with day, night figures that Ed (Elkcub) had sent me in the past and when I used them, my figures agreed with the figures on the data sheets Ed had supplied. I used them with some of my transmission curves and the results seem to match close to some of the test results recently posted.

Some of today’s results are:

Binocular Day Night
Swift 8.5x44 61.58% 57.11%
Trinovid 8x20 82.91% 79.46%
Ultravid 8x20 89.76% 86.44%
LXL 8x20 81.24%` 76.67%

It appears what is being done is that the D65 light value for a given wavelength is being multiplied times the optic transmission value then multiplied by the photopic (scotopic) value for the wavelength, then divided by the value of the D65 light time the photometric value of that wavelength, then integrated over the visible spectrum.

This would seem to be a reasonable value for converting a radiometric curve to a photometric value.

There are a couple of problems I see with this. First, they are using D65, or a CCT of 6500K, and I prefer 5000K for my optics since it fits my particular photopic curve better, and also my color preference. I did substitute Illuminant A for D65 and it only changed the day and night numbers by a few percentage points though.

Second, peoples individual photopic and scotopic curves deviate from the average by quiet a bit. One or two reviewers can be considerably off the norm and report what they see and be perfectly valid. The norm would not come into play until a significantly higher number of reviewers brought the average closer to the norm.

For some time I have been using the color-rendering index as a guide for color deviation. The Trinovid and Ultravid above, for example:

Ultravid, CCT 5124K and white point of .3414, .3445 Ra=97.5%, I pulled out the strong red, strong yellow, strong green and strong blue values. Note these are not spectral colors but the CIE composites. They are 95.7, 96.7, 95.4 and 91.

The same numbers for the Trinovid are CCT 5051K, white point= .3432, .3446 with the R9, R10, R11 and R12 values of 91.3, 95.9, 94.3 and 89.8.

The CRI seems to tell me more about what colors I will perceive visibly a lot more than any transmission figures.

As some of you are aware, I quit relying on transmission curves for any brightness estimate, relying more on invariant radiance as a function of small solid angle or f:/number instead.

I am going to revisit those experiments in light of this new data though, and see if it makes any appreciable difference.
 

Attachments

  • test photopic.jpg
    test photopic.jpg
    143.8 KB · Views: 69
  • test scotopic.jpg
    test scotopic.jpg
    148.9 KB · Views: 66
  • Swift 8,5x44 HHS links.pdf
    290.9 KB · Views: 50
... It appears what is being done is that the D65 light value for a given wavelength is being multiplied times the optic transmission value then multiplied by the photopic (scotopic) value for the wavelength, then divided by the value of the D65 light time the photometric value of that wavelength, then integrated over the visible spectrum.

Yeah, that confirms it. Thanks, Ron. :t:

... Second, peoples individual photopic and scotopic curves deviate from the average by quiet a bit. One or two reviewers can be considerably off the norm and report what they see and be perfectly valid. The norm would not come into play until a significantly higher number of reviewers brought the average closer to the norm.

Not sure I agree completely. An individual's deviation from the average would be very hard to measure without a well-equipped visual psychophysics lab. Nonetheless, even for individuals the day/night metric values should predict the order of perceived brightness very closely assuming the lighting conditions are comparable. Humans only make relative judgments about brightness anyway.

On post #82 I mentioned: "The day-value applies to white light, i.e., the full input spectrum. The same computation, however, could be done for each of the color frequency ranges show below. This would result in seven additional numbers representing the brightness of each color band, which could be presented in either tabular or graphic form. This would certainly improve on the purely subjective methods these authors (and others) use."

Would you mind integrating the Swift data over the various color bands to see what percentage differences there are among them? They might also be put on a bar chart to show color emphasis.

Ed
PS. I had you in mind when I mentioned someone else doing all the work. :-O
 
Last edited:
The weighted averages about 510 and 555 nm, if indeed weighted by the shapes of scotopic and photopic sensitivity are, however, THE BEST representation of how bright a binocualr will look by night and day that I can think of.
Ron


That´s the way I see it. But, men, we are now amongst what visual perception and perception of color are all about. To understand what's really going on here we have to go back to university for a couple of semesters to learn only the knowledge base and go on with some research because one thing is for sure: The knowledge about visual perception is still by no means at all complete. Would be a pleasure though, to hit the books with you guys again. :)

Steve
 
Ed,

I will have a look at this later. I don’t think it will help anything though. The linear spectral bands are just going to follow the photopic sensitivity curve and there is no way to integrate the photometric values (color mixing of RGB factors) in a simple way from spectral (radiometric) data.

Probably way over my head.
 
Ed,

I will have a look at this later. I don’t think it will help anything though. The linear spectral bands are just going to follow the photopic sensitivity curve and there is no way to integrate the photometric values (color mixing of RGB factors) in a simple way from spectral (radiometric) data.

Probably way over my head.

Not sure what you mean by linear spectral bands, but OK. The limits of integration (summation) for the calculation you mentioned earlier can be changed arbitrarily to produce sub-spectral values. In fact, the formula apparently assumes one value for each 5nm band and then adds them up.

Anyway, we can't get everything done in one day. Maybe someone will show up with a programmable spectrometer and make it easy. ;)

Thanks again,
Ed
 
Last edited:
Ed and RonE:

Interesting your very technical take and observations on this subject. For me I will take
the Dr. Gils van Ginkel, Europa test as a good guideline.

It seems this testing across a wide variety of optics, using precision measuring instruments,
would be one the best tests I have found. I find the graphs with full curves of the spectral
measurements are very well done.

Much of what I find about the discussion above is quite obscure, and for the average observer
has not much merit.

I would recommend taking it over to another thread. Not relevant here.

Jerry
 
Last edited:
Ed and RonE:
...

Much of what I find about the discussion above is quite obscure, and for the average observer
has not much merit.

I would recommend taking it over to another thread. Not relevant here.

Jerry

I was just going to comment very positively about the Europa tests, since they are the poster child of how to publish day/night values with some understanding of their importance. I must have offended you by saying that their explanation of the terms was not quite all it could be.

Speaking for myself about your rude statement, I have to say that you didn't originate this thread and really don't have much standing as to what is "obscure" or "relevant," — much less what has "merit." If you don't understand or like what's being said you can always participate elsewhere.

Ed
 
I was just going to comment very positively about the Europa tests, since they are the poster child of how to publish day/night values with some understanding of their importance. I must have offended you by saying that their explanation of the terms was not quite all it could be.

Speaking for myself about your rude statement, I have to say that you didn't originate this thread and really don't have much standing as to what is "obscure" or "relevant," — much less what has "merit." If you don't understand or like what's being said you can always participate elsewhere.

Ed

Ed:

Take a look at the start of this thread, the Swaro. SV. and the Leica BN.

The thread drifted into another whole discussion, optics brightness and evaluation.

The Europa tests offer something that most can read with some ease.

Forgive me, some of the posts earlier, could be used in another discussion, please
start one.

They are not relevant to "this" thread. I was not trying to be rude, but trying to get
the discussion to point.

Jerry
 
Much of what I find about the discussion above is quite obscure, and for the average observer
has not much merit.

I am probably not even "the average observer," not having owned or seen through most of the binoculars discussed on BF, but this thread is a good example of why I read this forum regularly and have learned a great deal here.

The OP made a very subjective claim that surprised many, pleased some, and left a few scratching their heads about how to attempt to quantify pleasing (if no longer state of the art) optical qualities.

While much of the above discussion is challenging and occasionally arcane, it is advancing our knowledge base, and the BN/SV comparison is but a provocative pretext.

Carry on. :t:

David
 
Jerry,

Sorry I'm afraid I agree with the others. I find the technical turn of the thread fascinating and positive.

"For me I will take the Dr. Gils van Ginkel, Europa test as a good guideline."

I'm going to take issue with you here. While I find the inclusion of the transmission data laudable, both originally and on rereading, I found it's factual errors annoying, the data presentation highly misleading and left me highly suspicious of the motivation behind the report.

"If you read through the Ginkel Europa tests, the chart of transmission for day and nite organizes these nicely for a wide range of binoculars. One that stands out is the Bushnell 8x43 Elite, which has a night trans. of 83, Day 88, the comment they make on this one it has a Pink-red cast." Most others have a more neutral trans. curve, as shown on the bar graphs."

Was that what you were supposed to think? Look carefully at the plots for the Elite and the SCL. There is a couple of blips in the Elite transmission curve which biases the result and doesn't, certainly can't, represent the night/day brightness performance. This is a powerful argument for the methods ED and Ron have discussed. In fact if you look carefully, the curve practically mirrors the SCL up to 600nm where Elites continue to climb the the SCLs go into decline. I've not tried any of the pairs here and couldn't say if the Elites are "pinkish" or the SCLs "bluish" (or maybe greenish?). Im sure this is very much a point of view! A bias that may well vary with the user.

"Most others have a more neutral trans. curve, as shown on the bar graphs."

Just feel I need to point out the deplorable practice of changing the Y axis dimensions. It is highly misleading. A short axis make the curves look flatter. Please take a closer look.

I really can't understand how you think this a good example of how to communicate test results. Please don't try to stifle a healthy discussion which might just lead to an more accurate binocular evaluation. Some of us who aspire to own an alpha pair one day value objective information.

David
 
Last edited:
Warning! This thread is more than 13 years ago old.
It's likely that no further discussion is required, in which case we recommend starting a new thread. If however you feel your response is required you can still do so.

Users who are viewing this thread

Back
Top