Your preferred image resolution for Astrobin AstroBin Platform open discussions community forum · Dan Brown · ... · 16 · 281 · 2

I have, till recently, uploaded images to AB that are about 15 megapixels. At this size the image supports the click zoom feature, a very convenient viewing experience. I feel that viewing images that don't support the click zoom feature usually take too much time to load and when loaded the viewing experience can be awkward, so typically I don't download full res versions. 
Last week I noticed that a lot of the IOTD winning images are much larger than my usual 15 megapixels. I decided to upload my most recent image at its full 61 megapixels. It definitely shows more resolution doing so but I don't know if anyone will bother to download it. 
What is your preference?
Dan

Hey Dan, I always prefer native resolution.  You can still have a high resolution image and maintain click to zoom if you upload a JPG.  JPG even on the highest quality setting yields an image that is a quarter of the size with minimal loss. 

That being said I don't pay much attention to the click to zoom as I feel it ruins the viewing experience for me. Just personal preference.  I want to see the whole image on my screen, and fill it.    If I'm dissecting the image quality then the click/zoom is nice.    

In no way at all do I mind downloading the full resolution image.  Even the largest ones download in less than a second and I'm on starlink.  I honestly pay little attention to the warning when I upload.

Chris White- Overcast Observatory:
My preference is to view an image with the ability to click-zoom.  I also find that some full resolution images are less interesting when zoomed in too far.... but the value in this is that you can really see minute details and max out what the telescope is capable of....

So with my own images, I'm posting both versions.  A downsampled version that is half the native resolution, and set this as my final image so that its easier for people to navigate to a higher res version and then I post a full resolution version as a revision (not final) that people can click on and download to pixel peep should they decide to do so. 

https://www.astrobin.com/5uu47w/C/

The click-zoom final version is 19mb, while the full res version is 65.

Good idea, I like it.
Thanks,
Dan

Chris White- Overcast Observatory:
My preference is to view an image with the ability to click-zoom.  I also find that some full resolution images are less interesting when zoomed in too far.... but the value in this is that you can really see minute details and max out what the telescope is capable of....

So with my own images, I'm posting both versions.  A downsampled version that is half the native resolution, and set this as my final image so that its easier for people to navigate to a higher res version and then I post a full resolution version as a revision (not final) that people can click on and download to pixel peep should they decide to do so. 

https://www.astrobin.com/5uu47w/C/

The click-zoom final version is 19mb, while the full res version is 65.

I too have been doing this with my own recent images.  My last image of M31 is jpg and while a 3 panel drizzled image, still can be viewed with click and drag.  However the jpg format is still a lossy compression and I can see issues.  Also, AstroBin appears to carry out its own compression with our downloads and these images have additional artifacts than the original jpg images had.  However they are still very nice.  In the example of mine above, I see diffraction ring-like artifacts around many of the star-like objects in the dark background.  To me these are not an issue for the full effect of what I wanted to convey for the full frame image.  However, this post was also meant for people to pixel peep, because as a drizzled image and a three panel mosaic, there is necessarily much more detail to be seen at quite high magnifications.  So I also included a full scale png image.  Its over 98MB, but I have no issues looking at the image.  In any case, this allows for some few who wish to roam around the field and see the many features of this large close galaxy and some of the smaller galaxies, star associations, reflection nebulae, globular clusters, red, blue and yellow supergiant stars and emission nebulae within a natural RGB  field.   For those unaware PNG is a lossless compression and I recommend using it.  At least for galaxy images because the vast dark smooth backgrounds allows for very significant compression.  In the full size image of the png file, I see none of the artifacts that I see in the jpg.  It looks very close to my original working (XISF) image.  This for a 20k+ X 12k+ image.  For those who choose to navigate to that revision only!

Seems to me that file size is irrelevant (except for the AB limit for click zoom). The relevant issue is to provide native image scale or perhaps drizzled image scale for a drizzled image. I say this because one might choose to crop an image to draw the viewer's attention to the subject matter. For example, an imager may target  an interesting but small galaxy. As such, the sensor field of view may be relatively large and not present the subject matter optimally or artistically. In such case the cropped image  file size is smaller than the full FOV image, but still has the same image scale.

If I am faced with such a situation, I like to present the crop as the final image - at native image scale but also present the full FOV in a revision - hopefully still at native image scale.  I say 'hopefully' because I may wish to stay under the click zoom limit for this view. It seems to me that it is inconsiderate or confusing to the viewer to present additionally a second full FOV image that exceeds the click zoom limit which may have only a marginally better image scale.

Jim Thommes:
It seems to me that it is inconsiderate or confusing to the viewer to present additionally a second full FOV image that exceeds the click zoom limit which may have only a marginally better image scale.

How is it inconsiderate or confusing if the description of the revision clearly states the purpose of the revision?  I see it as an additional choice given to those who might be interested, for whatever reason they choose.  Unless there is some cost or risk to the viewer?  I think that the image offered above by @Chris White- Overcast Observatory is a perfect example where both the final image and the full scale image being offered is of great benefit to the user.  Very much appreciated by me and I assume enough to others.

File size can be relevant if comparing two images using different compression methods, or degrees of compression used within a single method, such as jpg.  Given a large native resolution file that is on the edge of meeting the click/zoom feature, meeting your standards of image quality may not be met if you choose to meet the click/zoom number.  This seems to be irrelevant with any non-mosaic image, with current sensor/file sizes for single frame images with the allowances here on Astrobin.  Yet it still warrants checking your images each time you upload them to ensure that each level of scale presentation is acceptable to your standard.

The FAQ recommends uploading JPEG, I do believe Salvatore has replied with this answer in the past as well, unless something has now changed that I am unaware of (source).

JPEG is a recommendation.  The thread you link too does not state that he expects that to be exclusive.  The thread includes someone worrying about the use of png files.  Seems like that person is unaware that png can be used as a compression method to reduce file size as well as jpg.  The difference is png claims that compression is lossless.  My experience with png is that it compressed files clearly look better, even better than my jpegs compressed with maximum retention of original qualities.  Both levels, colors, and artifacts caused by compression are worse with jpeg compression at max quality vs png with high compression.  But that is highly image dependent.  That said, even though is see the differences, for the examples I provided above, I still used the jpg compression for my final image, the mouse-hover image, etc.  Only for the high quality revision did I provide a png compressed image and I clearly stated my reasons for doing so.  No AstroBin rules broken.  AstroBin is always within its rights to choose not to allow png files, but as long as they do, I think I may take advantage of that.  My working in PI with 32 bit images, etc. only to have them cut to 8 bit is something that may push me one way or the other depending on my standards.

I should add that there is a cost to using png.  They seem to be larger as a file than jpg.  So that is a cost.  When choosing a fairly high (but still lossless) level when compressing png files, the algorithms take a lot longer to achieve the compression with png files.  This is significant.  Impatient people should use jpg.

Timothy Martin:

Alan Brunelle:
Also, best check that those files are faithfully reproduced after you upload to AstroBin.

I am soooo not in control of that. In fact, I just uploaded a jpeg that was 132MB, 18844x12420, and the click-to-enlarge message says it's 17MB and enables the click-to-enlarge. So clearly something substantial is going on behind the scenes on AB. All I can do is, as I said, upload the best image I possibly can. Then it's really out of my hands. Pretty much everything is that way--Facebook, APOD, APOD GraG, Astrophotography Prize. Heck, even getting something printed requires some externally imposed restriction.

*Oh, I hear you!   That's why I resort to trial and error.  Some of my posts start with revisions well down the alphabet. I think I'm getting better though.   Just a week ago I finally figured out a way to delete pages stuck in my staging area. Too embarrassed to ask.

Hi @Jan Erik Vallestad ,

I do trust that you see no differences as you state.  A lot of these issues are very image specific.  And that is why have repeatedly stated that we make these choices when they become important to us.  To show you a very simple example of what I am seeing, I give you the examples below:

JPG compression, at the very highest quality compression: 


Same field with png, high compression:


Note: These two images are screen clips directly from the images as presented on the AstroBin site at the highest scale. Therefore these are both at full image scale dimensionally.

In the case of these two images, the jpg image is from my top image seen first in my post and it works with the mouse-hover feature and click-zoom.  Because I spent a lot of time finding a software to do lettering and annotation, learning to do so to my standards precisely for the clarity and benefit of the user of my image, I felt that the jpg was fine for a high level presentation and even "OK" for a click-zoom, but that I wanted to offer the user a high quality image in case they wanted to use the best image as a tool to tease out some of these GC from their starred images, which are often tightly close to stars, etc.  

I have other examples where these types of artifacts impact actual nebula features.  I may post examples, however, with the holidays here and my family visiting, let alone that my wife and I are moving in January (my work computer is unavailable), I risk having my family kicking me out of the house if I spend too much time on this stuff right now.

Note in Salvatore's comments that he cannot make any claims as to how any software implements its algorithm to generate jpgs, and leaves it precisely to the user to work out what is best for them.  I use three software packages to generate my jpegs.  PI, (always set to 100% quality), Irfanview (which offers a lot more options to reduce file size, bit depth, scale, etc. than PI.)  This I used often to reduce the image scale of my files.  I often find that full scale is unnecessary for many of my images, nor did I want to release these full-scale if I find the full screen full frame image is all that I want to convey.  Finally, I started using GIMP to do my annotation and it also saves as jpeg. None are transparent as to what they actually do.  However understand that the compression is only one side to this equation.  How Astrobin treats the data and uncompresses and displays the image is another side and, again, I would stress that the user is responsible to find what suits them.

I actually save my images as XISF for fidelity and compatibility to quickly revise in PI, and jpeg and now PNG or whatever are in the posts I present publicly.

I also understand that presenting data at the highest level of perfection is not necessary nor even desirable to people who post here.  Consider:
Some might want the posted images to not look perfect.  Reasons: a. Concerns about theft of images, of which there are examples here on AB.  b. the compression defects at the highest scale can act a bit like watermarks on the image.

Me, I am less concerned about theft.  Yes, I would certainly feel violated if I saw one of my images used commercially.  However, I welcome anyone to download my images for personal use, such as on their computer desktop, or even if they want to print, all without asking.  Actually, I would be honored.  And holding my best quality on my hard drive until I am buried seems a waste to me.  As if I am hording something of value to use in my afterlife...

Best regards, CS and happy holidays!
Alan