Unistellar eVscope - More Experiences (V. 1.3)

Introduction | The New App Version (Version 1.3.0) | The New Image Format and the New Image Processing | Small Bug List | Suggestions for Improvement | First Conclusions on Version 1.3 | Links || Appendix: Report on App Version 1.3 for Unistellar

On this page, I describe my experiences (no observations) with my electronic 4,5" Newton telescope Unistellar eVscope (I took part in a Kickstarter campaign in mid-November 2017; it arrived on January 27, 2020). This fifth "experiences" page covers first experiences with the new app version 1.3 (from May 20, 2021 on). Possibly, these are useful for others who want to acquire the eVscope as well...

Note: App version was released on September 2, 2021, and there were few opportunities for observations with app version 1.3; so this reports remains somewhat "rudimentary"...

Further Notes: See page Overview of the eVscope Pages for just that! For a detailed version history of the app, see page App Version History.

 

Introduction

Photos: My eVscope (End of January 2020)

Read First, Please!

My eVscope arrived at my home at the end of January 2020. During the first weeks, it had an app with a version number below 1.0. My experiences with this early app version are reported on page Unistellar eVscope - First Experiences. For links to further experiences, see page Overview of the eVscope Pages.

Note: Those who already start with a newer app version do not need to read the above-mentioned pages.

The following experiences are based on the app version 1.3 and refer initially only to changes compared with the previous state; there was no update for this version.

On the one hand, according to Unistellar, some bugs were fixed, on the other hand, some changes were made regarding the functionality and the user interface. The following additions are perhaps the most important ones:

A complete list of improvements can be found in the following. For a more detailed version history, see page App Version History.

 

The New App Version (Version 1.3.0)

The new app version (V 1.3.0) is available in Europe since May 21, 2021, and I downloaded it on the very same day. First of all, the app required me, as did the two previous versions, to update my eVscope, which succeeded right away. The new app version offers, according to the App Store, the following new functions (in my own words):

Unistellar published a short video about the new app version:

Status: No updates to version 1.3.

Comets of 2021 Added to the Data Base

The comet hunters among us will be happy about this! I have not checked this, but trust there fully on Unistellar...

Four Times Larger Observing Record

According to a video about the new app version, the photos are now scaled up to four times the original size (from 1.2 to 4.8 MB). I first thought that what already looked blurry in full size on my computer or iPad before, would now look even blurrier, because the camera does not get more pixels this way... I will explore this matter further down!

Note: On July 17, 2021, I noticed that upscaling only occurs in Enhanced Vision mode. Lunar and planetary images in Live View mode remain unchanged at 1280 x 960 pixels (rectangular format without overlay).

Unistellar Experience Improvements

Regarding this aspect, I was not able up to now to find out what this refers to...

Various Bug Fixes

As always, no information on this...

 

The New Image Format and the New Image Processing

See also my Report on App Version 1.3 for Unistellar from the beginning of September 2021!

So what does the statement "Four times larger observing record for even more stunning memories of your stargazing" mean in practice? To be honest, this was not easy for me to clarify. As already written, upscaling the sensor images does not add any new information, and the eVscope's photos, which appear blurry at full size anyway, become even blurrier this way. So much for the theory! But before I pass judgment, I first wanted to look at the new results and compare them with the previous ones. This is what I do on this page. Besides, I am not interested in the new size, so I compare the old results with the new ones after scaling the new ones back down to 1280 x 960 pixels. Then the photos are the same size again and thus directly comparable.

When comparing old and new photos or when looking directly at the new photos, I noticed the following things:

In the following, I will try to substantiate these observations with examples.

Colors and Color Fringing

The new app version 1.3 has a slightly different color reproduction than the older versions. Primarily, the colors look a bit more saturated, which is certainly a matter of taste. This increased colorfulness can also be seen in brighter stars, which now often have distinct color fringes, sometimes in violet, sometimes in blue, sometimes multicolored. The photos of the older versions had more of a kind of faint rainbow color fringing.

         

M 17 - Aug 23, 2020

 

M 17 - Jul 17, 2021 (new)

 

M 17 - Jul 17, 2021

   

M 17 - Aug 23, 2020, photo on top processed

 

M 17 - Aug 23, 2020 (old)

 

M 17 -Jul 17, 2021, photo on top processed

Photos: Demonstration of the stronger color saturation and the purple color fringes around bright stars due to the new processing

Contrast and background

The contrast of the photos seems to have increased, perhaps making the photos appear more brilliant. The background also appears darker compared to my older shots. However, the higher contrast is disadvantageous for bright cores of globular clusters (and probably also for galaxies), because the washing out of the core can hardly be controlled.

         

M 92 - Sep 15, 2020, manual exposure (old)

 

M 92 - Sep 15, 2020, processed

 

M 92 - Sep 15, 2020, processed and sharpened

         

M 92 - Jul 18, 2021, automatic exposure (new)

 

M 92 - Jul 18, 2021, processed

 

M 92 - Jul 18, 2021, processed and sharpened

Photos: The photos demonstrate the darker background and the higher contrast - the core of M 92 is more washed out with the new processing (automatic exposure)

Example DSO: Comparison of Nebulae

Large area (galactic) nebulae offer, in my opinion, a good opportunity to compare the old and the new image processing. For this, you should, however, look at the photos in the "sensor size" 1280 x 960 pixels and not just the smaller thumbnails (click the previews to see the larege versions). I also scaled the new photos back down to the sensor size so that old and new image processing can be better compared. Below I show photos of M 8 (Lagoon Nebula), M 16 (Eagle Nebula), M 17 (Omega/Swan Nebula), and M 20 (Trifid Nebula). While the old photos typically represent some of the best I was able to get out of the eVscope, the new photos were taken at half moon, and presumably the eVscope was also not properly collimated.

    

M 8 (Lagoon Nebula) - old processing

 

M 8 (Lagoon Nebula) - Version 1.3

 

M 16 (Eagle Nebula) - old processing

 

M 16 (Eagle Nebula) - Version 1.3

 

M 17 (Omega/Swan Nebula) - old processing

 

M 17 (Omega/Swan Nebula) - Version 1.3

 

M 20 (Trifid Nebula) - old processing

 

M 20 (Trifid Nebula) - Version 1.3

Apart from the problems of the new photos (noise, amplifier glow), the new photos look to me "a bit" more detailed and also not as "painted" as the old ones. Readers of these pages may disagree, but at least you can take a closer look at the comparison photos! After looking at these (and other) photos, I decided not to ask Unistellar for an image processing that "paints" less.

Note: Here, old and new photos have been brought to the same size, namely the sensor size, for a direct comparison. Of course, this is only one of many comparison possibilities. Further below, I describe my observations on the iPad, where the new size is exploited in the display. There you can see a clearer difference in favor of the new app version! But afterwards, I found out that the new versions look nearly identical on the iPad after being scaled down to the old format...

Example DSO: Comparing Globular Clusters (M 92, M 13)

Globular clusters consist of many fine stars, here the new image processing of the new app version has also brought improvements - except for globular clusters with very bright cores (M 15, M 92), which are washed out and the details destroyed. In the following, I present an example with M 92 (with ASI224 and PS72 for comparison):

         

M 92, eVscope-20200915-201415.jpg

 

M 92, eVscope-20210718-222729as.jpg

 

M 92, PS72*, ASI224, stack

   

M 92, eVscope-20200915-201415as.jpg

 

M 92, eVscope-20210718-222729as.jpg

 

M 92, PS72*, ASI224, stack processed

Old image processing

 

New image processing, orignal photo downscaled to 1280 x 960 pixels

 

For comparison: Photo with PS72 and ASI224 from a stack of single frames

*) My ED refractor PS 72/432 seems to have some color issues...

With M 13, I find differences in background darkness and sharpness:

   

M 13 - Sep 15, 2020

 

M 13 - Sep 15, 2020, processed

 

M 13 - Sep 15, 2020, processed and sharpened

   

M 13 - Jul 18, 2021

 

M 13 - Jul 18, 2021, processed

 

M 13 - Jul 18, 2021, processed and sharpened

The new photos of M 13 look a little "oversharpened" compared with the old ones (look at the large versions!).

"Graininess"

The first photos taken in Enhanced Vision mode looked very noisy with the old processing, since they consist of the overlay of just a few shots. The longer you stayed in Enhanced Vision mode, the less noisy and the more detailed the photos became (although there are testers according to whose opinion the photos did not improve after a certain duration in EV mode...).

Im folgenden drei aufeinanderfolgende Stadien eines Fotos von M 101 mit der alten Bildverarbeitung, darunter aufgehellte Versionen, um das Rauschen besser zeigen zu können (große Versionen aufrufen!):

Below, there are three successive stages of a photo of M 101 with the old image processing; below them are brightened-up versions to show the noise better (open the large versions!):

         

M 101 - 20210306-215743

 

M 101 - 20210306-215831

 

M 101 - 20210306-220137

   

M 101 - 20210306-215743, made brighter

 

M 101 - 20210306-215831, made brighter

 

M 101 - 20210306-220137, made brighter

With the new image processing this is also true, but the visual impression is a bit different: Initially, the photos now seem to have a certain "graininess", which decreases over time, but according to my impression does not vanish completely, especially if you edit the photos afterwards (by increasing the contrast, for example). I can only guess where this "graininess" comes from, and suspect that upscaling the photos is the cause.

Below, there are six successive stages of a photo of M 101 with the new image processing; below them are brightened-up versions to show the noise better (open the large versions!):

         

M 101 - 20210718-223222

 

M 101 - 210718-223246

 

M 101 - 20210718-223309

   

M 101 - 20210718-223222, made brighter

 

M 101 - 210718-223246, made brighter

 

M 101 - 20210718-223309, made brighter

         

M 101 - 20210718-223349

 

M 101 - 20210718-223429

 

M 101 - 20210718-223536

   

M 101 - 20210718-223349, made brighter

 

M 101 - 20210718-223429, made brighter

 

M 101 - 20210718-223536, made brighter

In the following, I present section with 1:1 pixeld of the bright versions:

         

M 101 - 20210306-215743 (old), made brighter, section

 

M 101 - 20210306-215831 (old), made brighter, section

 

M 101 - 20210306-220137 (old), made brighter, section

   

M 101 - 20210718-223222 (new), made brighter, section

 

M 101 - 210718-223246 (new), made brighter, section

 

M 101 - 20210718-223309 (new), made brighter, section

   

M 101 - 20210718-223349 (new), made brighter, section

 

M 101 - 20210718-223429 (new), made brighter, section

 

M 101 - 20210718-223536 (new), made brighter, section

Less Aggressive Image Processing ("Painting")?

Especially when looking at photos of large nebulae or galaxies, I was not the only one who noticed that the image processing of the eVscope is quite aggressive and "brushes away" details - the photos look a bit like watercolor paintings. This is also noticeable when comparing eVscope photos to photos taken with other astronomy cameras. The most direct comparison is with the ASI224MC camnera, which uses the same sensor as the eVscope, albeit in a different mode.

Since I am currently (spring/summer 2021) unable to observe my favorite object for such tests, the Orion Nebula M 42, I made my first comparisons between the old and the new image processing using M 51 (Whirlpool Galaxy). My first impression here is that the new image processing does not "paint" quite as much as the old one (but not everyone was able to confirm this). Below, I show examples that can be used to check this assumption. I also show M 51 photos taken with the ASI224MC for comparison.

         

M 51 - Mar 6, 2021, processed

 

M 51 - 30.5.2021, processed with Polarr

 

M 51, PS72*, ASI224, processed

Old image processing

 

New image processing, orignal photo downscaled to 1280 x 960 pixels

 

For comparison: Photo with PS72 and ASI224 from a stack of single frames

*) My ED refractor PS 72/432 seems to have some color issues... It also needs a UV/IR cut filter, which was missing.

Similar to globular clusters, the core of galaxies can be more washed out with the new image processing (not visible here). Please inspect the large versions (click on the previews) to see the differences in the image processing! The old processing looks a little like "phantasy"...

App Version 1.3 on the iPad

Initially, the Unistellar app only ran on iPhones (which is why I had to buy one, unfortunately...) and not on the iPad; since app version 1.0.5, it can also run on the iPad. However, I have to admit that I always found the eVscope images on the iPad very blurry and therefore preferred my iPhone for its sharper (and smaller) image. When I used my iPad to control the eVscope for the first time after a long break in late October 2021, the image no longer seemed that blurry to me, which I expected to be an effect of upscaling the image size. Of course, I don't have a comparison between the app versions anymore, so I can only describe my impression here....

So I thought! But I still have eVscope photos in Apple's Photos app stored, taken with different app versions! For comparison purposes, I created an album to where I moved some new and old photos from DSO. And when comparing the photos, there is a clear difference between version 1.3 photos and older versions! The old photos are scaled up in the iPad, which can also be seen, and look "coarser" than the new ones, which are probably even scaled down slightly. With this comparison, it is then also immediately noticeable that the new photos are less "brushed". However, the effect seems to be less due to the scaling than to the image processing alone. I scaled down photos in the new format to the old size and transferred them to the iPad. There was virtually no difference between these and the large-format images on the iPad.

The lesson I take from this is that a comparison is not always easy, and the results often depend very much on the conditions and sizes under which the comparison is made. The new format may only realize its full potential when the photos are displayed larger than the original 1280 x 960 pixel format. On the other hand, the downscaling example shows that the differences between old and new photos are essentially due to differences in image processing.

Conclusions

The new image processing actually seems to bring advantages in the presentation of the sky objects, such as more details and sharpness as well as a darker background. Increased contrast and saturation, however, are a matter of taste. Now, I only wish that the graininess of the photos and the amplifier glow can be reduced...

Note: I will probably create another page where objects can be compared with with respect to the old and to the new image processing. My impression is that the new image processing delivers much sharper images (after scaling down to the sensor size).

 

Small Bug List

I postpone my bug list to version 1.4, which also is said to have fixed a lot of bugs...

 

Suggestions for Improvement

Wish for Unistellar: Error message as to why the Enhanced Vision mode was aborted.

Possible causes could be:

And as always, I wish that important image data would be saved with the image!

 

First Conclusions on Version 1.3

App version 1.3 has brought some useful improvements, particularly for the look of the photos. Once again, a "thank you" to Unistellar for that! But the company also did not implement a number of features that I had hoped for and had asked Unistellar for several times already.

See also my Report on App Version 1.3 for Unistellar from the beginning of September 2021!

Note: I draw generals conclusion about the eVscope after owning it a little more than a year of ownership, based on app version 1.2, on page Unistellar eVscope - Second Conclusions (Version 1.2).

 

Links

 

Appendix: Report on App Version 1.3 for Unistellar

Finally, I would like to give some feedback on app version 1.3. It was already released at the end of May, 2021, but somehow I never managed to put my feedback together. And it is still not "final", after all...

I would like to restrain myself to giving feedback on the new image size and processing. When I first read the statement "Four times larger observing record for even more stunning memories of your stargazing", I was not quite sure what this means. But I soon found out that it means upscaling the images from the sensor size of 1280 x 960 to a "whopping" 2560 x 1920 (2 x in linear size).

My first thought was that this is not a good idea, since upscaling does not deliver any new information and makes the image softer, overall. This is similar to the digital zooms of digital cameras (they come in various "spices"...), and typically you should not go beyond a factor of 1.4x there.

A second finding was that the new images are rather grainy, which is new. The graininess is reduced over the observing time, but does never really disappear. It is rather evident if you brighten up the photos for demonstration purposes, and may nearly become invisible if the sky on the photo is dark. I do not know the cause for the graininess. Upscaling be one reason, a changed image processing the other - probably a mixture of both...

But as always, things are not that easy! I found the following changes:

iPad: I rarely used my iPad for controlling the eVscope, because I found the images, similar to my laptop, fuzzy and somewhat disappointing, whereas they looked much nicer on my iPhones, because they were a lot smaller on these.

Recently, I used my iPad again (because it has more "stamina"...) and was pleasantly surprised by the sharpness of the photos shown on the screen!

I no longer have access to older app versions, but in the Photos app, I still had some photos taken with older app versions of objects that I also had recently photographed. Well, that really is a difference and a huge improvement, which I had not expected!

I considered that upscaling for the old versions and downscaling for the new versions might cause the difference. But that is not the case! I transferred some downscaled photos (1280x960) taken with the new app version to my iPad, and they looked nearly identical to the large versions. So, it is the new image processing that makes the images look sharper. So, this was an interesting experience!

More Detail: I always disliked the "painted" look of the eVscope images, that is, the processing "brushed away" the details. This was especially apparent on photos of large nebulae, such as M 42, M 16, or M 17. This is also apparent, when I compare eVscope photos with my photos taken with the ASI224, which uses the same sensor (albeit in a different mode...). On a first brief look, the images now seem to look more similar, but I still have to check that thoroughly...

I had always planned to send you an e-mail and ask for less painting. This has never happened, but you seem to have guessed my wish - the new image processing "paints" definitely less and reveals more detail. Not every eVscope user may recognize this, but for me, it is evident. I found this already for M 16 and M 17 (and some more nebulae...), but will have to wait until winter for M 42...

So, this is also a nice surprise for me, which comes somewhat unexpected because of my preconception regarding the upscaling of images. Anyway, I scale my images down and back to 1280 x 960 for archiving purposes (and presenting them on my Website)...

So all I would wish is that the graininess of the photos (and the amp glow) can be reduced...

Wish List

Finally, I would like to return to my wish list, which is rather short this time.

First of all, I have to repeat my request for some sort of storage of the image data, be it EXIF data (preferred), an addition to the file name or some text file. The data may be stored in the data that go to SETI, but I do not have access to the SETI server (and would have to store the data there first, which would cause delays...).

Secondly, I would like to have an error message that explains why the Enhanced Vision mode has aborted. There may be many reasons for this (clouds, too many discarded frames, shocks, etc.). I had an e-mail exchange with another eVscope user who complained about lots of EV mode aborts and was unsure how he could improve on this...

So that's it for today! If you need images examples, please let me know! But I think that the Unistellar engineers know the photos that the eVscope produces well and do not need sample images...

 

An den Anfang   Homepage  

gerd (at) waloszek (dot) de

About me
made by walodesign on a mac!
17.10.2021