NPLB Project - Pixelmetric search for Voorwerpjes
-
Experimental search method I conceived to search for objects of interest for Galaxy Zoo and other Zooniverse projects, next to the traditional methods such as querying photometric catalogs.
More info, progress and updates will be posted here.
Posted
-
by Budgieye moderator
I saw someone demonstrate Aperture Photometry Tool if that helps
http://www.aperturephotometry.org/
There was also an easier tool, I forget the name.
Posted
-
by Ghost_Sheep_SWR in response to Budgieye's comment.
What?
EDIT; oh no I didn't mean existing tools, although I will be using APT in the future for another little project. Here I will describe the progress and results of a custom search method I conceived and am slowly creating to find objects of choice in a different way from any other methods.
Posted
-
NPLB stands for No Pixel Left Behind, because I like cheesy titles.
Most object searches depend on catalog searches, whether it’s photometric, spectral features or any other measured value. Mostly photometric searches anyway whereby the search is focused on one or more catalogued magnitudes / ratios of magnitudes. An exception is the shape-based g-lens algorithm searches.
But when we’re judging possible Voorwerp candidates for example it’s not about magnitudes, it’s color. A unique color for Voorwerpjes that changes with redshift and sometimes a bit of Milky Way dust reddening. Which ofcourse is a combination of magnitudes, but that’s not the point.
So why not actually base a dedicated search on color? That’s what I’m building right now, and AFAIK this has not been attempted before, which is exciting because the possibilities and effectiveness are unknown so far.
Pros and cons will follow, but one you can see for yourself; look at a random SDSS area and look at photometric objects. See how much area is missed, and not in a catalog?
Posted
-
NPLB Project Color
For NPLB to successfully work certain requirements have to be met regarding color; objects of interest need to have a unique color and the images used in the search must have a large color range for unique object colors.
Object colors
At the moment I’m convinced that there are object classes with unique distinct colors in different survey viewers. For example Voorwerpjes display a color range from deep blue to magenta to pinkish / purple in SDSS which is rarely seen in other types of objects, and consistently so. The same holds for other types as far as I know.
Hanny’s Voorwerp as imaged in SDSS
Color range
The color range of used images must be large enough to search for distinct colors. SDSS uses JPG (or JPEG) format to display RGB color images created from three seperate bands (gri) in their online navigation tool, as do many other surveys.
Standard JPG images are build up from colored pixels created from a red, green and blue channel with 256 different shades of grey per channel. This gives ~ 16.7 million unique possible colors for every single pixel in an image.I believe this is more than enough color range for NPLB to work. Perhaps even too much because if a single value in any of the three bands is off by 1, for instance red channel has value 134 instead of 135, this might mean missing an actual targeted object entirely 😄
Ofcourse this can be accustomed for by adding a proper ‘spread’ of RGB values around an actual sampled pixel color.
Posted
-
by JeanTate
Interesting project! 😃
Are you considering "photometric objects" only? If so, then catalog searches would seem to be both feasible and fruitful; for example, a search of (one of) the WISE catalogs, for objects with unusual (IR) colors.
FWIW, I used colors (among other things) to identify possible Huds (humongous distant spirals).
SDSS images are particularly colorful because of the Luptonizing, the clever transforms which emphasize otherwise very subtle color differences; maybe worth looking into?
Posted
-
by Ghost_Sheep_SWR in response to JeanTate's comment.
Thanks! 😄
Are you considering "photometric objects" only? If so, then catalog
searches would seem to be both feasible and fruitful; for example, a
search of (one of) the WISE catalogs, for objects with unusual (IR)
colors.Most definitely not, on the contrary! So I'm not arguing against photometric catalog searches. A couple of volunteers at the BYW:P9 project devised a smart query to search for Brown Dwarfs and outdid previous searches by scientists by far. They found dozens to hundreds of BD's, and there is no argiung that these kind of smart queries are ultimate in terms of effectiveness / efficiency. Well ok technically BD candidates but exactly the right colors + movement is nearly certified BD's.
What I want to do is what we're always doing in here at GZ; we judge by color, not combinations of magnitudes in the three different bands. So I will search on color not magnitudes. Seems to add up to the same thing but not quite so in my opinion.
A strong argument for trying this search experimental search method that is considers everything in the image, not just only photometric objects. For instance look at a random area in SDSS, cut away everything inside the photometric object circles and see what is left. A conservative guess from me is that at least 75% of the entire image will still be intact! And contains a lot of objects / information still; objects just too faint to become a photometric object, diffuse areas and faint clouds, outskirst of galaxies etc.
FWIW, I used colors (among other things) to identify possible Huds
(humongous distant spirals).How did you define the color? Combinations of magnitudes to use in a photometric catalog query?
SDSS images are particularly colorful because of the Luptonizing, the
clever transforms which emphasize otherwise very subtle color
differences; maybe worth looking into?Might be interesting, but not important for NPLB project. In fact it doesn't matter which transformations are done previous to the JPG RGB image, only consistency is important. The same transformations should be used throughout the whole survey so the same 3-color images result from it and I'm good. It doesn't matter which bands are used, which colors are assigned to which bands or anything else. If two Voorwerpjes at same ~z are greyish yellow then a third unknown Voorwerp at same ~z will look greyish yellow too, see what I mean?
It is a bit like I'm searching for certain kinds of photometric objects, except that every single pixel in the entire survey is a photometric object.
Posted
-
by NGC3314 scientist
One note of caution - the photometric object circles do not represent the area considered to be the objects by SDSS. The ones shown, for example, in the Navigator pages, just show the center locations. For galaxies, the light is traced out to a large radius (it has to work out the Petrosian radius, for example). In recent Navigator versions (DR12, for example) the "SDSS Outlines" option gives a better view of pixels that went into galaxy measurements. You might get interesting additional large things from smoothing the SDSS data (although that is almost surely better done in the FITS files than using the JPEGs just because of compression losses).
All sorts of weird things can happen when an object does not have a shape or intensity profile at all like a de Vaucouleurs or exponential profile - this is why I gave up on a pure catalog-color search for nearby Voorwerpjes doing much more good. And I was never all that interested in downloading the whole SDSS (which you may be...). I did find out that one has to be ready for a significant range in color based on the ratios of strong emission lines and (as you mentioned) redshift. Strongly star-forming galaxies would be a likely contaminating (or interesting) product, likely to mostly have higher surface brightness.
I can't resist: SDSS color rendition is exactly a mapping of magnitude differences, except very close to the background level where it might be as long as the SDSS sinh magnitude and Lupton color scaling thresholds match. (I'm not sure whether they do).
Posted
-
by JeanTate in response to Ghost_Sheep_SWR's comment.
Well, NGC3314 beat me to it, and said it far better than I ever could.
So just a couple of extras:
- I think it will be very challenging to identify real emission (i.e. that which is not noise) away from the boundaries of photometic SDSS objects; yes, sometimes SDSS misses very real objects (I even started a thread on just that), and sometimes there's a diffuse background SDSS doesn't recognize, but in my experience the former is rare and that latter hardly ever interesting
- how did I define color? The same way astronomers do: (magnitude1)-(magnitude2). For my Huds search, I found there is an oft-used relationship between color and redshift, for ellipticals, so I searched for high-z extended (to exclude QSOs) objects whose colors were quite different from that
- over in RGZ, there's a thread whose aims are somewhat similar to yours (I think): Green galaxies associated with RGZ sources. This has turned up quite a few very interesting objects, some of which may be voorwerpjies ...
Posted
-
by Ghost_Sheep_SWR in response to NGC3314's comment.
One note of caution - the photometric object circles do not represent
the area considered to be the objects by SDSS. The ones shown, for
example, in the Navigator pages, just show the center locations. For
galaxies, the light is traced out to a large radius (it has to work
out the Petrosian radius, for example). In recent Navigator versions
(DR12, for example) the "SDSS Outlines" option gives a better view of
pixels that went into galaxy measurements. You might get interesting
additional large things from smoothing the SDSS data (although that is
almost surely better done in the FITS files than using the JPEGs just
because of compression losses).TBH in the back of my mind I was a bit afraid this might be the case. But that would also mean photometric object values are, uhm, made 'from a lot of different colored pixels' I guess. well the other way around ofcourse but a large area goes into one value is what I mean.
And I was never all that interested in downloading the whole SDSS
(which you may be...).Yes. And if I get some actual results from this it's re-usable in the future. But not just Voorwerpjes or SDSS. Still assessing other object types to include, and with good results can expand to DECaLS / DES etc.
I did find out that one has to be ready for a significant range in
color based on the ratios of strong emission lines and (as you
mentioned) redshift.Yes I expect this. But I don't have to take into account emission line ratios / filters / redshifts. After sampling the target JPG cutouts for RGB pixel values to use for searching I will create one or two RGB values vs redshift plots so it should be fairly easy to interpolate and even extrapolate missing z RGB values to add to the pool of pixel values used in the search.
Strongly star-forming galaxies would be a likely contaminating (or
interesting) product, likely to mostly have higher surface brightness.Perhaps, I'm not sure I'll just have to run the experiment to find out. From what I've seen so far those are either light-blue through white colored?
I can't resist: SDSS color rendition is exactly a mapping of magnitude
differences, except very close to the background level where it might
be as long as the SDSS sinh magnitude and Lupton color scaling
thresholds match. (I'm not sure whether they do).This is like trying to read Chinese characters to me. But if I understand correctly it's something along the lines of 'RBG pixel values = magnitude ratios from the g r i bands'?
Thank you for the feedback and note of caution 😃
Posted
-
by Ghost_Sheep_SWR in response to JeanTate's comment.
I think it will be very challenging to identify real emission (i.e.
that which is not noise) away from the boundaries of photometic SDSS
objects; yes, sometimes SDSS misses very real objects (I even started
a thread on just that), and sometimes there's a diffuse background
SDSS doesn't recognize, but in my experience the former is rare and
that latter hardly ever interestingAs they say, the proof of the pudding is in the eating 😃 . But I'm not too afraid of a few random noise pixels matching the RGB values I search on; a quick visual check of the area will reveal those to be just that.
over in RGZ, there's a thread whose aims are somewhat similar to yours
(I think): Green galaxies associated with RGZ sources. This has turned
up quite a few very interesting objects, some of which may be
voorwerpjies ...Looks like a very interesting thread, although it's grown quite large in 3 years 😕 Perhaps someone might be interested in going through it looking specifically for Voorwerp candidates and posting them here.
Thanks for the feedback!
Posted
-
by vrooje admin, scientist
This is an interesting road to go down, but I suspect a potentially bumpy one, as Bill mentions. I agree it's pixel colors that matter, not so much integrated colors. Even in pixels with highly ionized gas you'll still potentially have contamination from the rest of the galaxy (or the galaxy behind it), so one way to clean that up might be to subtract off the smooth galaxy model, which would leave behind the things that deviate from that, like the bar, spiral arms, and ionized clouds of gas. A few test runs on known voorwerpjes might be able to reveal whether the considerable extra work that entails is worth it in terms of cleaning up the color ranges...
I can't resist: SDSS color rendition is exactly a mapping of magnitude differences, except very close to the background level where it might be as long as the SDSS sinh magnitude and Lupton color scaling thresholds match. (I'm not sure whether they do).
This is like trying to read Chinese characters to me. But if I understand correctly it's something along the lines of 'RBG pixel values = magnitude ratios from the g r i bands'?
Yes, though flux ratios (magnitude differences), and that might break where the detected flux is close to the sky background.
One thing's for sure: you'll find all the blue and green satellite trails! 😄
Posted
-
by Ghost_Sheep_SWR in response to vrooje's comment.
On the first part:
That might be true, but I’m setting this up as dumb as possible. No transformations, no creating JPG’s, no substractions, I would’t even know how to perform these transformations... Just static JPG’s as they are.
To clarify, this is the whole process in a nutshell:
- Extract pixel RGB values from target objects from survey JPG’s
- Use Photoshop to select all these pixels from large survey JPG’s
- Find coordinates of those pixels and asses if it might be a target object candidate
As simple as possible
On second part:
That might be the case if Hanny’s Voorwerp is pure blue, not if close but not exactly pure blue. Unsure atm if this is the case. For green I don’t think so because there won’t be pure green pixels in the pixel sample ( for Voorwerpjes not at least)
Thanks for the feedback!
Posted
-
by JeanTate in response to Ghost_Sheep_SWR's comment.
Looks like a very interesting thread, although it's grown quite large in 3 years 😕 Perhaps someone might be interested in going through it looking specifically for Voorwerp candidates and posting them here.
The >500 objects with the hashtag #green are a very mixed bunch. The most boring ones - which you'll likely encounter in your NPLB project - are artifacts, such as a "green fringe" which is mostly likely due to misalignment, differential refraction, etc. Others are QSOs with strong (broad) emission features which just happen to end up, redshifted, in the G of the JPGs (r-band in SDSS).
Some are certainly EELRs (extended emission line regions), and some may also be Voorwerpjies (are all EELRs Voorwerpjies? Are all Voorwerjies EELRs?). For example, SDSS J130854.52+562155.6:
Is the extended green emission an EELR? A voorwerpjie? (In this case I think we know the answer, because more research has been done on it; if all you have is an SDSS JPG, how could you tell?)
For the #green objects where the green is due to [OIII] 5007, there should be corresponding blue ones (lower redshift) and red ones (higher redshift). As compact objects, the former have been sought, and found; they're "blueberries" (lower mass counterparts to Green Peas, if I recall correctly). The latter? I'm not sure, but I've come across a few, so they certainly exist. As extended objects, some get called Green Beans; perhaps Hanny's Voorwerp is a low-z example? I think one of the Green Bean papers (perhaps the first?) reported one or two possible "Red Beans" (i.e. high-z Green Beans).
Then there are truly weird things, like the green blob near the zsp 4.623 quasar SDSS J123237.49+520343.8 (which I posted both here and in RGZ Talk; the green blob is a separate PO, SDSS J123237.28+520345.1):
A chance alignment of a green, foreground galaxy with a very distant QSO? A LAE (Lyman alpha emitter) right next to the QSO? An intense, very distant Voorwerpjie? An artifact??
Posted
-
by Ghost_Sheep_SWR in response to JeanTate's comment.
At the moment I’m still thinking about target object types. For the Voorwerpjes I’ll probably take just those that are in the papers, less or equal to ~25 in total. No green stuff in there AFAIK. Also considered green peas but I’ve learned a catalog query turned up 40k+ candidates so dismissed those. That just way more than I’m willing process, not to mention it’s very likely they will contain ‘pure’ green so will turn up every asteroid, satellite trail and cosmic ray hit in SDSS.
Targets I’m nearly certain to include in the experiment;
- Voorwerpjes; obviously 😄
- Quasars with ears; dunno, odd color and curious what more will pop up
- Brown Dwarfs; mainly on the lower end of sequence, T5-T8 and Y’s. I don’t expect too much of it in SDSS but still have to try for the BYW project
But I will definitely consider suggestions. Although they have to meet certain requirements; they have to be rare, have type specific color in SDSS and there has to be a minimum number of them present in SDSS to sample RGB values (hmm minimally around 5 known ones in SDSS or somewhere around that)
Posted
-
NPLB To-Do-List
PREP PHASE
- Acquire engine for customized survey JPG download
- customize engine for SDSS
- determine target object types. As of yet Voorwerpjes, Quasars with ears and Brown Dwarfs. Sugggestions welcome
- sample SDSS JPG RGB Pixel values from known targets
- Multiply sample pixels by minimally changing RGB values of sampled pixel
- Create SDSS Voorwerp RGB vs redshift plots. Check if similar is possible for BD's> many types + distances might get an erratic plot
- Interpolate and extrapolate sample RGB Pixel values from plot(s)
- Download SDSS
- Acquire Photoshop
MAIN PHASE
- Use Photoshop to extract all sample RGB pixels from SDSS survey
- Get frustrated
- Retrace coordinates of selected pixels
- Get elated
FINAL PHASE
- Assessment of candidates / target objects
- File remainder into the proper channels
Posted
-
Hmm it does seem that SDSS gri is the worst survey to search for late T / Y brown dwarfs, think I have to abandon those too / save for later for other surveys.
Perhaps found one Y dwarf visible in SDSS but after seeing absolutely nothing for T8.5-T9.0-T9.5 BD's in SDSS I now strongly suspect it might be noise after all.
DECaLS is be a bit better (grz) but looks like red-only objects in the JPGs, so still nearly useless for that kind of search.
Posted
-
by Budgieye moderator
Yes, SDSS doesn't "see" very far into the infrared. I looked for a brown dwarf in backyard worlds in SDSS.
There is a disk there, but it is very, very, very............ very, faint.
Posted
-
by Ghost_Sheep_SWR in response to Budgieye's comment.
Yeah I did check that one, but I'm fairly certain that isn't the Y-dwarf in SDSS;
- After checking a bunch of late T's and seeing absolutely nothing in SDSS I don't think it's even possible in SDSS gri bands. (maybe in z or y?)
- Even if visible in SDSS it would be very faint dark red-only, this object has dark greenish-blue tones
- Your object hasn't moved between SDSS-DECaLS, while the Y-Dwarf is a mover
- Your object is probably too far away though to be the BD, below I've centered your object and the position of the Y-Dwarf in WISE
You can check the position of the Y-Dwarf yourself in the DECaLS viewer, flipping between the two unWISE layers (using spacebar after consecutively selecting two different layers) you can even spot some movement.
http://legacysurvey.org/viewer?ra=126.2792&dec=28.1052&zoom=15&layer=sdss2
It might be one of the red objects below in DECaLS DR5 images, but not even sure about that since those objects don't seem to be in the direct path of motion.
Posted
-
Just messing around with the RGB plots, these are from the 4 'Quasar with ears' section. First one is RGB values vs redshift, second one is total RGB values vs redshift. Should be interesting to see how much the different RGB values influence the total RGB values vs redshift plot for the Voorwerpjes. I reckon the Voorwerpjes plot could also be very usefull as a tool to better judge Voorwerp candidates (quantification vs eye-balling) and guesstimate redshift when assuming a candidate is a Voorwerp.
Ofcourse the number of objects is too small (n=4) here for the Quasars with ears and I very much doubt we're dealing with the exact same object types. But it gives a good example of what I meant with plotting the RGB values.
Posted
-
by JeanTate in response to Ghost_Sheep_SWR's comment.
Hmm it does seem that SDSS gri is the worst survey to search for late T / Y brown dwarfs,
Given their blackbody temperature equivalent, and the SDSS filter transmission curves, why would you expect any brown dwarf to be visible in g, r, or i (unless it is a binary with a hotter star)? Even the z-band struggles, doesn't it?
Posted
-
by Ghost_Sheep_SWR in response to JeanTate's comment.
More hoping than expecting it to try to maximize output; the whole proces takes a fair amount of work and time and the extra effort of piggybacking secondary targets would hardly be felt overall.
Posted
-
Some very crude preliminary graphs I just made that I'd like to share, which I'm still mulling over to fully understand 😃.
First one is the lowest sampled RGB pixel values vs redshift, second the highest and third is a comparison of the total RGB values of lowest and highest. Note that RGB values not represent brightness but just colours. Also had to multiply redshift z value with 100 because excel is a bit sucky.
Some notes to the first graph; biggest blue spike at ~2.5 is Mkn 266 which by my count is unaffected by both continuum emission or MW Dust, so a 'low-z pure Voorwerp' that spikes in the blue. The biggest red spike is UGC 11185 which I have noted as affected by both continuum and 50% MW Dust which makes all RGB values go up but also redder, the size of the big spike makes me think there's also dust reddening from within the galaxy itself? The red spike left of that one is from the only unpublished one I found that's in the sample, which is also the only one that gets the full brunt of MW Dust
I've been warned the values might vary greatly, and these graphs really aren't usuable yet. But already fun to look at in combination with other information.
So the main things contributing to the colours (as mentioned before in this thread) are the Voorwerpjes itself (magnitude), redshift, continuum emission and Milky Way Dust reddening. Ofcourse I've also collected data on those factors, and its interesting to see how these things clearly contribute to the colour mix.
For now i think it should be feasible to create 'cleaner' plots that separate the different effects on the Voorwerpjes colour; a 'pure' unaffected Voorwerp plot, a continuum emission affected one and a MW dust affected Voorwerp colour plot. Should create a good spread of RGB values that represent the different states unknown Voorwerpjes might be in.
Some other thoughts;
- One oddball is The Teacup all the way at the end. First I thought it is redder because of redshifting into the SDSS i band, and I don't have it noted as being (highly) affected by MW Dust. But comparing with the other Voorwerpjes it seems it's affected by both continuum emission and dust. Looking at the different shape and colour of The Teacup compared with other Voorwerpjes; is the loop an object that also contains a lot of stars and dust? Can anyone confirm?
Also I think there are some interesting areas to look for Voorwerpjes;
-
Heavily MW Dust reddenend Voorwerpjes: aside from the one unpublished one there isn't one Voorwerp in the sample that is strongly affected by MW Dust (in / near the MW plane). Assuming a random distribution of Voorwerpjes across the sky this seems a bit unlikely?
-
Higher redshift Voorwerpjes; a pity there is such a big redshift gap between the last two Voorwerpjes, and no other known ones around redshift z 0.8-1.0. There simply should be many more assuming a random distribution and a much much larger area (space inside a growing sphere) at higher higher redshift
-
The Southern Hemisphere obviously. Not SDSS but DES should be good.
Posted
-
by ElisabethB moderator
Hi GSS
I'm not commenting on this thread (which is way over my head anyway ! ). But you might consider moving this over to the new platform, as I don't know how much longer this version of Talk will remain read/write. This version will be kept as an archive but it will no longer be possible to post in it. But as I said, I have no idea when that will happen.
Cheers E.Posted
-
by Ghost_Sheep_SWR in response to ElisabethB's comment.
If it’s ok with everyone I’m fine adding to this thread until that is no longer possible. If not let me know!
PS. I very much doubt this is over your head, it’s only been sampling pixels from SDSS JPG images with Microsoft Paint so far 😄
Posted
-
OH NO!
Imaging Extended Emission-Line Regions of Obscured AGN with the Subaru Hyper Suprime-Cam Survey
Ai-Lei Sun, Jenny E. Greene, Nadia L. Zakamska, Andy Goulding, Michael A. Strauss, Song Huang, Sean Johnson, Toshihiro Kawaguchi, Yoshiki Matsuoka, Alisabeth A. Marsteller, Tohru Nagao, Yoshiki Toba
https://arxiv.org/abs/1803.07241
Can never do projects in a relaxing pace, always run or get behind it seems.
Posted
-
So, now, some good news and some bad news, and some bad news with some good news 😛
Good news: I've managed to squeeze out a more useful graph of 'Pure' Voorwerpjes (haven't had the time for the messier ones, no not 'Messier catalog' ones, messier (sp?)) ones. The RGB values here are the average value of all sampled values per Voorwerpje, so that takes care of any anomalous pixels a bit. To my surprise had to ditch an additional two out of my chosen set of 'pure' ones because they spiked relatively to the others. For now I assume they too suffer from continuum emission contamination o_O .
EDIT: This part was based on a faulty graph so isn't relevant anymore:Bad news: it's clear that the RGB values steadily decline towards zero / noise level, RGB values can be anything from 0 to 255 and in SDSS images values around '10-15' for every band is not a good thing.... I suspect average SDSS noise is somewhere between RGB values of 1-10, so following the trend downwards pure Voorwerpjes are probably indiscernible beyond z=0.055, or at least before z=0.06 which isn't all that deep. I think the big blue spike of the first one is because below a certain z threshold it gets into the SDSS sweetspot (close enough for SDSS to pick it up really well)??
EDIT: graph accidentally took the wrong values, will update it.
I was planning to test it on 1 or more Voorwerp candidates from the Voorwerp thread to see how it holds / how the candidates turn out in terms of Voorwerp signature and estimated pixelmetric redshift but I really want to check out the paper from above post.
Bad news: as far as I can see there are really 10 new Voorwerpjes in that paper, so 10 less to discover....
Good news: 10 new voorwerpjes might mean 10 new objects to sample from!
PS; if anyone wants to try to use the graph for themselves; just download the SDSS JPG image of the candidate, open with Windows paint and then use the 'Color Picker' tool to sample the colour, go to 'Edit colors' to see what RGB values that colour has and then draw the 3 horizontal lines starting at the Y-axis pixel values. Then draw a straight line downwards to the x-axis where it hits the correct colour line. If the graph and the candidate are good the three lines should hit the x-axis at around the same point, and so also give an indication of the redshift, which you can match with the redshift of the supposed host galaxy. HAVE FUN! 😃
PS2: my earlier comment about the Teacup redshifting into i-filter is rubbish. First the OIII emission will ofcourse get into the r (green) filter, which gets us (surprise!): Green Peas and Green Beans...duh
Posted
-
So that was quite the Excellent misstake, apologies for that.
Here is the correct graph, based on the average sampled RGB pixel values from a chosen set of 'pure' Voorwerpjes eg. supposedly not affected by regular starlight / dust from the host galaxy or Milky Way dust.
The blue peak at z ~0.05 (5.0 in graph) is actually Hanny's Voorwerp, seems a bit of an outlier to the colour trend although the sample is still pretty small (n=7).
Which gets us the (linear) trendline graph for these RGB line values:
Seems like a pretty steady RGB ratio vs redshift, and as is obvious extends much farther out than z 0.06 than I thought previously, at least for red + blue. Which makes sense since the Voorwerpjes follow a general colour trend from dark blue to magenta along redshift. That is until the OIII emission is redshifted into the SDSS r band which is mapped to green.
Assuming a linear trend I've manually extended the trendlines until they hit pixel value 10, after which they become meaningless since they descend into SDSS noise level.
I've tried testing the trendline graph on LEDA 4678815 but failed miserably. It is tainted a lot more with at least MW dust than I was hoping for, and perhaps also dust from the galaxy itself.
Quite hard to even find a good 'pure' voorwerp candidate that (by eye) isn't affected by continuum emmision / tidal debris / host galaxy dust or MW dust, don't know why that is.
Here you can see it's location relative to the Milky Way plane & dust: http://legacysurvey.org/viewer?ra=51.3055&dec=40.7017&zoom=4&layer=sfd
Posted
-
I've managed to split the messy ones in a red-dominated and blue-dominated group. But it isn't really possible to extract usefull trendlines from them as there are just too many variables in play that affect the colour and too little subjects in the samples. One major contributor seems to be distance to the nucleus; the closer to the nucleus the higher the RGB values get. But then there is also continuum emission (and other emission lines perhaps), host galaxy dust, MW dust, redshift, energies involved etc...
Remarkable that the Voorwerpjes all started with Hanny's Voorwerp, which is a kind of a special case even within the group of Voorwerpjes.
Also notable is that to me the Teacup looks like an oddball Voorwerp with a very different colour, much more whiter / towards pink in comparison with the blue and magenta of other Voorwerpjes. But within the red-dominated group it isn't really remarkable in colour space.
Below is the comparison with Mkn 1498 and the Teacup in SDSS images and in the average RGB colour values of those two. Mkn 1498 is the closest Voorwerp to the Teacup in redshift and RGB values. In the SDSS image the Teacup looks to me as being a very different colour, in the average RGB comparison not really so much. It's a bit darker mainly because of the redshift difference but that's about it.
Odd isn't it? Must have something to do with how are eyes / brains perceive colours? Or the colours surrounding the Voorwerp influencing perception like in optical illusions?
Posted
-
NPLB To-Do-List
PREP PHASE
- Acquire engine for customized survey JPG download
- determine target object types: Main target is Voorwerpjes, secundary Quasars with ears
- sample SDSS JPG RGB Pixel values from known targets
- Multiply sample pixels by minimally changing RGB values of sampled pixel Not quite, seems smarter to fill up the 'missing pixels' between close RGB pixel values
- Create SDSS Voorwerp RGB vs redshift plots As of yet not quite as useful as I presumed for using for estimating redshifts of candidate Voorwerpjes. But useful for extracting pixel values
- Interpolate and extrapolate sample RGB Pixel values from plot(s) Used average sampled pixel values for interpolation of RGB pixel values. Also two trendline plots for inter- and extrapolation of pixel values
- customize engine for SDSS
- Download SDSS
- Acquire Photoshop
MAIN PHASE
- Use Photoshop to extract all sample RGB pixels from SDSS survey
- Retrace coordinates of selected pixels
FINAL PHASE
- Assessment of candidates / target objects; extract more RGB pixel values from candidates to get an average RGB value, estimate continuum emission / host galaxy dust / MW dust contribution to pixel values, estimate redshift of candidates if possible, acquire redshifts of likely host galaxies.
Posted
-
Have been thinking a bit about the broader application possibilities of using pixels / pixelmetric redshift.
In response to a Voorwerp tag by CecilaB in this thread I tried to fit the average RGB values of the magenta part on my ‘Blue trendline chart’. The galaxy has a redshift z=0.080585.
The messy Blue chart is built up from only 6 data points (confirmed Voorwerpjes) vs redshift, which each consist of the average RGB value of about 6 sampled pixels. From that chart 3 trendlines are calculated, one for each colour band. To get a more realistic green trendline I had to manually alter the first data point.
If everything is perfect the three lines should hit the same redshift value with the goal of quantifying judgement of candidates and at the same time getting an estimate on it’s redshift, which should ~coincide with the host galaxy redshift.
Resulting graph from test:
It is obvious that it is not yet usable to better judge a Voorwerp or establishing redshift. But keeping in mind the trendline input is only 6 data points and the actual redshift of the galaxy is z=0.080585 this seems pretty impressive already.
Just wondering about the applicability on using it on regular galaxies to get a pixelmetric redshift estimate next to photometric redshift (PhotoZ).
Posted
-
Tiny update.
Since this is all unchartered territory it seems every single thing I need for the complete process poses a problem that needs to be solved.
Right now took some hurdles in (trial for now) Photoshop, and still 2-3 hurdles to take. So getting pretty close now 😃
Posted
-
First test run results (+ first hits)
Ok so with only a tiny fraction of my selected pixel values I'm already getting A LOT of hits. So I'm probably still doing something wrong with the pixel selection, eg Photoshop takes also values between selected pixel values instead of only pixel values.... or something like that.
A lot of artifacts from brights stars, some magenta at galaxies but I think I'm also getting EVERY SINGLE SMALL RED/MAGENTA POINT SOURCE. Ehm are they just red dwarf stars or also [OIII] red/purple peas?? Will post pics later.
So instead of getting almost no hits I'm getting way too many, which means I have to get really really strict selection criteria. Something like must be multiple pixel hits, not a point source or smallish high-z galaxy, must be next to a galaxy etc. etc. Also means it will be a much more enormous task than I envisioned, needing to eye-ball all the hits and stuff.
Here is a small selection of some of the more or less promising looking hits, but I still have to really improve and streamline the whole process and apply selection criteria for candidates (Or post all the somewhat more promising looking hits) since the goal is to create an efficient + effective search method, not check every object in SDSS.
last one two stars with standard 1-sided colour glow?Posted
-
Can’t we make 4-valued or 5-valued pixel maps? RGB pixels are ofcourse tailored to our physiology, so for our eyes to perceive colour there is nothing more to add once the red, green and blue values are determined.
This is a very human-biased system, for a computer it doesn’t matter if a point of information has 3 values or 4 or 10.... ofcourse the display of a 3> values pixel would still be just RGB to suit our eyes&brains. But even there might be a possibility to display more values using transparancy on multiple pixel layers?
For example let’s say we have an RGB pixel made from the SDSS gri layers. Then the u’ and z’ layers don’t contribute to the information packed in the pixel. Just go from (random values) RGB(110,30,56) to ugriz(40,110,30,56,54). The 8-bit RGB system has ~16.5 million unique pixels. Adding one value would mean ~4.2 billion unique pixels, five values a trillion unique pixels(!).
So instead of thinking of pixels as a square colour thing but as an n-value informational point using the existing single band SDSS ugriz images we could already generate pixel maps (coordinates+5 values) with a trillion unique ‘colours’.... hmm
Posted
-
Right now the NPLB Voorwerpje search system is up and running! 😄
(abandoned the 'Quasars with ears' for now)
Many hits with most being just ordinary (foreground star) artifacts which I have to weed out. When in doubt I'll simply add the coordinates to the other ones I will be posting. There also seem to be some regular spiral starforming stuff, but since they are real hits I will not omit them from the hit lists. And by the looks of it I'm even catching some purple peas (or red dwarf stars?)!
Here are the very first pixel hits, literally just got them by running the sequences! All coordinates and images are centered on the actual pixel hit location.
Also keep in mind that the posted cutout images usually are a much darker than what I'm seeing in Photoshop, so a dark magenta blob or protrusion might be clearly visible for me while very hard to see here in the cutout images.
0.0469, 0.0003
0.0330, -0.0381
359.9001, -0.0602 Hmm this one disappeared already in DECaLS (/SDSS)??
0.0469, 0.0003
0.0330, -0.0381
359.9001, -0.0602
Posted
-
Yeah this will probably become a more 'skim&run' type of search because I'm getting a tad too much pixel hits. Just speedsearch for the very best looking candidates and keep those coordinates, forget the rest.
Here is the first batch of hits, which is probably overkill to register them all.
EDIT: added the DECaLS images
359.9330, -1.2747
0.0531, -1.4611
359.9655, -1.4321
0.0475, -1.5140 Faint ring in DECaLS
0.0992, -1.8162 pixel hit on opposite sides of the bright object
0.0673, -0.2598
0.0456, -0.4557
359.9022, -0.4871
0.0248, -0.5603
0.0823, -0.9281
0.0598, -0.9905
359.9046, -0.9910
359.9044, -1.0580
0.1069, -1.1735
0.0963, -1.1781
359.9231, -1.1688
359.9878, -1.2436
359.9330, -1.2747
359.9849, -1.2815
0.0663, -1.3522
0.0531, -1.4611
359.9655, -1.4321
0.0475, -1.5140
0.0253, -1.5136
359.9007, -1.5326
359.9809, -1.6201
0.1232, -1.6510
0.0992, -1.8162
0.0932, -1.8191
0.0648, -10.0698
0.0386, -10.0685
0.0097, -10.0704
0.1180, -10.2042
359.9618, -10.2198
359.9092, -10.2744
0.1058, -10.4037
0.1096, -10.4260
359.9432, -10.4379
359.9098, -10.4719
359.9521, -10.5349
0.0282, -10.5727
0.0648, -10.6372
0.0332, -10.6796
359.9996, -10.6492
0.0792, -10.8839
359.9778, -10.9015
0.1218, -2.0724
0.0394, -2.0581
0.1047, -2.1238
0.0390, -2.3333
0.0763, -2.3836
0.0011, -2.3592
0.0874, -2.4403
359.9274, -2.8027
0.0283, -2.8697
0.0555, -3.1241
359.9035, -3.1370
359.9815, -3.2636Posted
-
There is also quite a bunch of 'grey peas' in there; a bit dull grey with a hint of pink. Is this a known type, or perhaps just regular high-z starforming galaxies? (not depicted in below images btw)
http://legacysurvey.org/viewer?ra=0.0013&dec=-7.3995&zoom=16&layer=decals-dr5
http://legacysurvey.org/viewer?ra=359.9245&dec=2.6367&zoom=16&layer=decals-dr5
http://legacysurvey.org/viewer?ra=359.9330&dec=3.1792&zoom=16&layer=decals-dr5
http://legacysurvey.org/viewer?ra=359.9361&dec=3.8609&zoom=16&layer=decals-dr5
http://legacysurvey.org/viewer?ra=0.1107&dec=6.2077&zoom=16&layer=sdss2
http://legacysurvey.org/viewer?ra=0.0878&dec=6.6413&zoom=16&layer=sdss2
ra dec 359.9880, -3.7190 0.0688, -4.3049 0.0926, -4.4980 0.0195, -4.5307 359.9582, -4.6877 359.9617, -4.6905 359.9610, -4.7952 359.8976, -4.7806 0.0856, -4.9788 0.0489, -5.1545 359.9985, -5.1108 0.0183, -5.2054 0.0231, -5.2121 359.9250, -5.3122 359.9841, -5.4021 0.0297, -5.4609 359.9990, -5.4571 359.9614, -5.4653 359.9057, -5.5740 0.0235, -5.7116 0.0674, -5.8428 0.0508, -5.9167 359.9905, -5.9702 0.0439, -6.1017 359.9080, -6.2184 0.0131, -6.2092 0.0159, -6.2416 0.0000, -6.5541 0.0967, -6.9283 0.0619, -6.9827 0.0120, -7.2287 0.0014, -7.3996 0.1172, -7.5178 0.0803, -7.7241 0.1167, -7.7996 0.0089, -7.8646 0.0304, -7.9053 0.0447, -8.1078 0.0442, -8.1089 0.0830, -8.1824 359.9403, -8.2122 0.0797, -8.2675 0.0365, -8.3369 359.9462, -8.3617 359.9012, -8.4341 0.0950, -8.3839 0.1071, -8.4897 0.0150, -8.5476 359.9576, -8.7234 359.9903, -8.7613 359.9590, -8.8312 0.0198, -8.9426 0.0179, -8.9438 0.0565, -9.0569 359.9075, -9.3381 359.9937, -9.6928 0.0110, -9.6547 0.1251, -9.8242 0.0881, -9.8614 359.9104, -9.8602 359.9102, -9.8650 359.9225, -9.8645 359.9399, -9.9283 0.0907, 0.2351 359.9421, 0.1381 0.0895, 0.5180 359.9558, 0.4149 359.9613, 0.6579 359.9540, 0.6109 0.1093, 0.8947 359.9195, 0.7876 0.0547, 1.1237 0.0541, 1.1173 0.0484, 1.1165 359.9953, 1.0610 359.9411, 1.1727 359.9439, 1.3459 0.0922, 1.2952 359.9485, 1.2075 0.0296, 1.6102 0.0139, 1.7031 0.1157, 2.0402 0.0791, 2.0323 0.0839, 1.8727 359.9538, 10.4175 0.0864, 10.5694 359.9184, 10.6732 359.9316, 11.0095 0.0542, 2.1380 359.9607, 2.4928 0.1109, 2.4095 0.1209, 2.3435 359.9247, 2.6365 0.0794, 2.9340 359.9028, 2.9089 359.9134, 3.1368 359.9128, 3.1370 0.1022, 3.2560 359.9329, 3.1793 359.9565, 3.7539 0.0664, 3.6213 0.0657, 4.0093 0.0328, 3.8588 359.9361, 3.8612 0.0956, 4.2156 0.1134, 4.1714 0.0645, 4.1774 0.0215, 4.0923 359.9100, 4.4587 359.9095, 4.3962 359.9169, 4.5233 0.1184, 4.9022 0.0759, 4.8996 359.9407, 4.9080 359.9458, 4.8777 359.9706, 4.7004 0.0087, 5.0438 359.9158, 5.0234 0.0422, 4.9299 359.9680, 4.9370 0.1048, 5.2687 359.9687, 5.4953 0.0049, 5.4445 0.0019, 5.4409 0.0007, 5.4419 0.0009, 5.4423 0.0803, 5.4577 359.9004, 5.7544 359.9357, 5.7176 0.0806, 5.5866 359.9715, 6.0205 0.0406, 5.9969 0.0860, 5.9096 0.1106, 6.2080 0.0520, 6.1583 359.9860, 6.1018 0.0017, 6.0666 359.9339, 6.3908 359.9132, 6.3518 0.0822, 6.3337 0.0879, 6.6412 359.9225, 6.6316 0.0137, 6.8396 359.8970, 7.2915 0.1115, 7.5504 0.1152, 7.4399 0.0549, 7.7504 359.9013, 7.5933 359.9555, 7.5785 359.9023, 7.7843 359.9098, 8.1683 0.0789, 8.2589 0.0172, 8.2756 359.9621, 8.2927 359.9824, 8.5598 0.0055, 8.7687 0.0243, 9.0655 0.1112, 9.0753 0.1098, 9.0041 0.1098, 9.0041 0.0912, 9.4995 359.9127, 9.3923 0.0142, 9.3537 0.0986, 9.3742
BTW does anyone know how to neatly stack lists of coordinates here in TALK so they can be copy-pasted in the the SDSS Image List Tool? EDIT: nvm I found it.
Posted
-
I've managed to get one 'grey pea' with a BOSS spectrum, which says it's a QSO / Quasar at z = 0.7800
SDSS J235936.09+054515.9 1237678778478100658
Is it possible I'm also getting a lot of hits on greyish quasars without a spectrum?
etc. etc. Many more in above coordinates list, can be paste here: https://skyserver.sdss.org/dr12/en/tools/chart/list.aspx
Posted
-
A little sneak peek of the second SDSS field I'm just starting at.
A small part of 10x10 annotated SDSS images 2048*2048 pixels each in the background, overlaid with SDSS voorwerp pixel hits ready for inspection 😃
Posted
-
Voorwerp candidates
EDIT: reprocessed SDSS gri image
359.9329, 21.8659 Also seems to be a magenta area on the opposite side mirrored on the nucleus.
0.0575, 14.0888 clearly some sort of stream there, regular tidal stream?
359.9039, 14.8072 NGC 7800 SIMBAD many pixel hits all around, also in the magenta-ish outskirts. Definitely [OIII] in there, spikes in the spectrum.
Thought I went mad when seeing a second galaxy in SIMBAD Aladin light view 😄
Posted
-
Can be pasted in the SDSS IMAGE LIST TOOL
ra dec 359.9329, 21.8659 0.0575, 14.0888 359.9039, 14.8072 0.1198, -10.9740 0.0366, -10.9966 359.9832, -11.0085 359.9810, -11.1231 0.0803, -11.1607 359.9748, -19.0176 359.9604, -19.3732 359.9545, -19.4804 0.0617, -19.7490 0.0298, -19.8112 359.9456, -19.8294 0.1288, -19.7991 0.0830, -20.1979 0.0130, -20.5591 0.0060, -20.6274 359.9967, -20.6438 359.9079, -20.6963 0.0404, -20.6540 359.9943, -20.9928 359.9358, -21.0599 0.1056, 11.2420 0.0338, 11.2310 359.9171, 11.1446 359.9997, 11.1230 0.0100, 11.4962 0.0902, 11.4074 0.0818, 11.3021 359.9160, 11.6055 359.9300, 11.5669 0.0341, 11.6123 0.0665, 11.6094 0.0196, 11.8549 359.9967, 11.9701 0.0672, 12.2155 0.0092, 12.5947 0.0532, 12.7399 0.0980, 12.8537 0.1140, 12.8698 0.0349, 13.2650 0.1257, 13.4788 0.0245, 13.7996 0.0189, 14.1244 359.8988, 14.1469 0.0390, 14.0196 0.0177, 14.4387 0.0251, 14.4659 359.9622, 14.7104 0.0406, 14.6562 0.1151, 14.9607 0.0288, 14.8598 0.0590, 15.1714 0.1165, 15.3759 359.9561, 15.3402 0.0986, 15.5906 359.9267, 15.5578 0.0123, 15.8799 0.0296, 15.8383 0.0009, 15.7111 0.1108, 15.9401 0.1206, 15.9263 359.9570, 16.2291 0.0151, 16.2630 0.0813, 16.5418 0.0553, 16.3681 359.9176, 16.8480 0.0285, 16.9777
Posted
-
Candidate, 2 hits, small blue / magenta area, big cloud towards 1:00 o'clock, other cloud-like structures around.
0.0841, 26.0007 + 0.0831, 26.0006
Posted
-
Candidate + hitlist
Reprocessed SDSS gri image
359.9459, 18.7217
Can be pasted in the SDSS image list tool
ra dec 0.0831, 26.0006 0.0841, 26.0007 0.1077, 17.3952 0.0065, 17.3337 0.0777, 17.6119 0.0133, 17.6269 359.9636, 17.5877 0.0462, 17.5449 0.0732, 17.4880 0.0754, 17.4882 0.1105, 17.8735 0.0681, 17.7745 359.9768, 17.9463 359.9459, 18.7217 0.1209, 18.8742 0.0302, 18.8549 359.9953, 19.2032 0.1204, 19.1092 359.9785, 19.3943 359.9807, 19.3966 0.0073, 19.2486 359.8990, 19.2576 0.0500, 19.6230 359.9584, 19.6029 0.0715, 19.5844 359.9947, 19.5603 359.9901, 19.8233 0.0936, 19.7554 359.9020, 19.6720 359.9033, 19.6730 359.9867, 20.0511 359.9026, 19.9074 0.0860, 20.2381 359.9669, 20.2166 359.9022, 20.2067 359.9119, 20.1613 0.0591, 20.1865 359.9473, 20.1208 359.9079, 20.1183 0.1172, 20.5211 0.1206, 20.3201 359.9047, 20.7034 359.9433, 20.7000 0.0971, 20.9124 359.8941, 20.8905 0.0243, 20.8841 0.1170, 20.7934 0.0916, 20.7985 0.0927, 20.8008 0.0755, 20.7883 359.9447, 21.1704 0.0103, 21.2843 0.0457, 21.2149
Posted
-
Candidates + hitlist
0.0106, 23.1141
0.0474, 26.6447
0.0318, 28.8782
359.9682, 30.1994
ra dec 359.9914, 21.4747 0.0789, 22.3016 0.0992, 22.2795 0.1309, 22.2201 359.9425, 22.1872 0.1134, 22.4919 359.9246, 22.3016 0.0168, 22.8876 359.9716, 22.8239 0.0384, 22.8139 359.8945, 22.8281 0.1281, 23.1345 0.0800, 23.1327 0.0106, 23.1141 359.9682, 23.0843 359.8967, 23.0778 359.9082, 23.0047 0.0170, 23.2778 0.0886, 23.2167 359.9393, 23.5104 359.8993, 23.8152 0.1294, 23.7175 0.0757, 24.0037 0.0879, 23.8408 359.9705, 23.8643 359.9676, 23.8548 0.1143, 24.2288 0.1144, 24.2292 0.0442, 24.2533 359.9496, 24.2841 0.0320, 24.4988 0.0205, 24.4646 0.0078, 24.4417 359.9593, 24.3065 359.9942, 24.5328 359.9440, 25.1163 0.0600, 25.0900 0.0953, 25.2811 0.0610, 25.1725 0.0621, 25.1729 359.9231, 25.4756 0.0309, 25.4030 359.9321, 25.7313 359.9273, 25.6675 0.0573, 26.0281 0.0791, 26.4151 0.0792, 26.4155 359.9523, 26.4201 359.8904, 26.3453 0.0153, 26.2806 359.9594, 26.2705 0.0474, 26.6447 0.0681, 26.6256 0.0119, 26.6202 359.9002, 26.8005 0.0801, 27.1404 359.9960, 26.9332 359.9238, 26.9822 359.9392, 27.2542 0.0771, 27.3999 0.0751, 27.6735 0.0806, 27.6023 0.0122, 27.9848 359.9479, 27.9192 359.9096, 28.1015 0.0271, 28.4259 0.0286, 28.4185 0.1262, 28.6067 0.0318, 28.8782 359.8803, 28.8870 359.9823, 28.8204 0.0442, 28.7392 0.0991, 29.0221 359.9652, 28.9454 0.0102, 29.3333 0.0760, 29.5385 359.9083, 29.5101 359.9395, 29.4298 0.0552, 29.6786 359.9741, 29.6794 359.8818, 29.6800 359.9465, 29.9735 0.0337, 29.8703 359.9682, 30.1994 0.0881, 30.0951
Posted
-
Candidate + hitlist
In the DECaLS SDSS image cutout a tidal stream arcing back can be clearly seen as well as a secondary magenta blob. The upper triangle is positioned on the pixel hit, the lower one on the secondary magenta object / Voorwerp candidate.
359.8807, 32.3300
ra dec 359.9547, 30.3108 0.1025, 30.5595 0.0263, 30.7346 0.1260, 31.0489 0.0716, 31.0282 0.1193, 30.9069 0.0447, 31.2953 0.0981, 31.2813 0.1131, 31.1118 359.9392, 31.4652 359.9128, 31.4479 0.0719, 31.3834 0.0874, 31.3163 0.0543, 31.7310 0.0374, 31.7251 0.0020, 31.7321 0.0020, 31.7329 0.0309, 31.6758 0.0712, 31.9614 359.9439, 31.9211 359.9833, 31.7670 359.9019, 32.1850 0.1376, 32.0902 0.0898, 32.0626 359.9302, 32.0732 359.9181, 32.3377 359.8807, 32.3300 359.8919, 32.2556 359.9379, 32.5997 0.1138, 32.8550 0.0093, 32.7110 359.9671, 33.0637 0.0873, 32.9145 359.9846, 32.9047 0.0062, 33.1639 0.0084, 33.1641 359.9145, 33.5226 359.8810, 33.4739 359.9223, 33.4477 0.0229, 33.4226 359.9105, 33.2967 0.1052, 33.8467 0.0743, 33.8019 0.0403, 33.7445 359.8831, 34.1685 359.9081, 34.1073 0.1119, 34.0721 359.9422, 33.9701 359.9238, 34.3502 359.9531, 34.2288 359.9555, 34.2301 0.0932, 34.6183 0.0965, 34.6167 0.0075, 34.5555 0.0245, 34.5243 0.1159, 34.4898 359.9152, 34.4533 0.0015, 34.6207 0.0865, 35.2178 0.1407, 35.1623 0.0387, 35.1521 359.8707, 35.0749 0.0074, 35.5015 359.9192, 35.3741 359.8901, 35.5198 0.0722, 35.8386 0.0041, 35.8214 0.0227, 35.7299 0.0854, 36.1527 0.0657, 36.1121 0.1348, 36.0544 359.9368, 36.0451 359.8968, 36.0207 359.9272, 36.5702 0.0218, 36.7105 359.9117, 36.6341 0.1038, 36.6558
Posted
-
I wonder why I wonder why, why am I getting so much hits in a rectangular shape... hmmm?
http://legacysurvey.org/viewer?ra=0.3584&dec=6.3160&zoom=11&layer=sdss2
Posted
-
Candidates + hitlist
0.2525, -0.9418
Strange object with protrusion in SDSS, DECaLS and Panstarrs
0.2818, -3.0339
Tidal bridge between two galaxies or Voorwerp in between?
0.3154, -3.4279
ra dec 0.2434, -0.0632 0.1598, -0.1163 0.2656, -0.3958 0.2245, -0.3907 0.2269, -0.4144 0.1567, -0.7298 0.2525, -0.9418 0.1927, -1.0683 0.1573, -1.2021 0.2008, -1.3255 0.1949, -1.4219 0.2956, -1.6974 0.1598, -1.8998 0.3197, -10.0404 0.3176, -10.1813 0.2284, -10.2546 0.1235, -10.3537 0.1873, -10.4037 0.1871, -10.4023 0.1621, -10.4572 0.2602, -10.5423 0.2602, -10.5562 0.2698, -10.7748 0.3435, -10.8311 0.2577, -10.8916 0.1783, -1.9320 0.1432, -1.9670 0.2668, -1.9714 0.1944, -2.3120 0.1395, -2.3927 0.1354, -2.3946 0.1725, -2.4984 0.1565, -2.5100 0.2500, -2.6031 0.2696, -2.6781 0.2818, -3.0339 0.2485, -3.0485 0.2414, -3.2030 0.1656, -3.2472 0.2106, -3.3111 0.3154, -3.4279 0.1458, -3.4007 0.2145, -3.4292
Posted
-
Candidate + hitlist
Big odd-coloured object visible in DECaLS (other side too?)
0.2364, -6.8899
ra dec 0.2276, -3.4540 0.1814, -3.4761 0.2578, -3.5584 0.2640, -3.5868 0.2072, -3.5857 0.2111, -3.6139 0.2008, -3.6509 0.2974, -3.6604 0.2728, -3.7024 0.2367, -3.7117 0.2946, -3.7327 0.2386, -4.0689 0.3024, -4.2289 0.2378, -4.3285 0.2002, -4.4147 0.3259, -4.4287 0.2223, -4.4609 0.1218, -4.4587 0.2330, -4.5526 0.1344, -4.8344 0.1193, -4.9687 0.2507, -5.0256 0.1870, -5.0658 0.2347, -5.1645 0.2812, -5.1818 0.1797, -5.3632 0.1666, -5.3637 0.2022, -5.5820 0.2819, -5.6258 0.3402, -5.6347 0.2277, -5.7411 0.2216, -5.7357 0.3336, -5.9168 0.2702, -5.9323 0.1359, -6.0960 0.1805, -6.5197 0.1988, -6.8090 0.2364, -6.8899 0.3286, -7.1589 0.2155, -7.2670 0.1891, -7.3905 0.1756, -7.4058 0.1461, -7.4528 0.2505, -7.4715 0.2030, -7.4761 0.2291, -7.5242 0.1891, -8.0568 0.2408, -7.8500 0.2373, -7.8873 0.3324, -8.0166
Posted
-
Finally got a spectrum hit on one of the little critters, a broadline QSO. Also on an irregular galaxy which turned out to be a starburst galaxy (0.2083, 1.2306) so that's maybe not so good for the Voorwerp search but it is good for confirming [OIII] hits 😃
I think when I'm finished with the first degree of SDSS I'll check them all for spectra so I can round 'em up in (candidate) groups.
Here is the quasar + spectrum pixel hit object at Ra, dec 0.1613, 1.2406, redshift z=1.8352;
SDSS J000038.65+011426.3 SDSS ObjID 1237678617417875953
SIMBAD SDSS J000038.65+011426.3 -- Quasar
I think this one counts as a candidate purple pea;
0.1881, -8.7235
On another note, anyone know what this pixel hit object is? Looks extended and very disturbed. SDSS, DECaLS SDSS, DECaLS red and Panstarrs gri images. Starburst dwarf galaxy?
0.2304, -9.6453
Right, onwards with the hunt!
Candidate and hitlist
0.1192, -9.6077
Can be pasted in the SDSS IMAGE LIST TOOL
ra dec 0.2784, -8.1078 0.2132, -8.1209 0.1400, -8.3361 0.1593, -8.3648 0.1194, -8.4998 0.2039, -8.5330 0.3113, -8.5897 0.3314, -8.6034 0.2265, -8.6363 0.1905, -8.6450 0.1881, -8.7235 0.1180, -8.8654 0.2769, -9.0558 0.1238, -9.1594 0.1775, -9.2128 0.2621, -9.2456 0.2251, -9.2642 0.3018, -9.3287 0.3027, -9.3289 0.3045, -9.3404 0.3298, -9.3575 0.3219, -9.3691 0.2012, -9.3822 0.1843, -9.4356 0.1840, -9.5776 0.1214, -9.5171 0.1721, -9.5465 0.3001, -9.5922 0.1192, -9.6077 0.2028, -9.6336 0.2304, -9.6453 0.1894, -9.7366 0.2584, -9.7948 0.3416, -9.7941 0.1976, -9.8286 0.2123, -9.8234 0.2727, -9.8920 0.2188, -10.0267 0.1774, 0.1631 0.1510, 0.1363 0.2517, 0.3534 0.2266, 0.9120 0.1634, 0.8364 0.1961, 1.0877 0.1297, 1.0295 0.2044, 1.0290 0.1779, 0.9610 0.2872, 1.3826 0.1613, 1.2406 0.2083, 1.2306 0.1893, 1.5008 0.2477, 1.6503 0.1502, 2.0578 0.1298, 1.8875 0.2618, 10.1974 0.2434, 10.1904 0.2865, 10.0986 0.1405, 10.0044
Posted
-
Got myself a hot subdwarf with 3 pixel hits no less at 0.2778, 11.0099 😄 don't know if this means I might be finding unknown hot subdwarfs candidates further on but 3 hits is quite a lot for a single object.
Candidates + hitlist
0.2143, 10.7666
0.1468, 4.1592
ra dec 0.1935, 10.2512 0.3401, 10.2073 0.1290, 10.5554 0.2716, 10.8037 0.2143, 10.7666 0.1553, 10.6986 0.3426, 10.6866 0.2778, 11.0099 0.2788, 10.9795 0.3013, 10.9114 0.1592, 2.2821 0.1636, 2.2749 0.1768, 2.2704 0.1408, 2.0686 0.1694, 2.0600 0.2574, 2.3595 0.2375, 2.5817 0.2820, 2.7672 0.1493, 2.7381 0.1960, 3.3389 0.3435, 3.3226 0.2901, 3.1679 0.1311, 3.8043 0.1864, 3.7911 0.3115, 3.7781 0.2068, 3.6955 0.2052, 3.6967 0.1954, 3.9923 0.1518, 3.9804 0.1545, 3.9810 0.1468, 4.1592 0.3058, 4.1454 0.3240, 4.1242 0.2151, 4.0481 0.1804, 4.0486 0.1927, 4.4679 0.3384, 4.3816 0.3175, 4.3765 0.2876, 4.3811 0.2386, 4.3571 0.2319, 4.3566 0.3279, 4.3237 0.1415, 4.2705 0.3041, 4.8524 0.1523, 4.7982 0.2078, 5.1047 0.2556, 5.0836 0.1778, 4.9819 0.1455, 5.3050 0.3430, 5.2044 0.2669, 5.1410 0.3338, 5.4961
Posted
-
WOOHOO got my second spectrum quasar hit, and also one I'm proposing as candidate / target because it looks like it's got some serious action going on at 10 o'clock.
Candidates + hitlist
Quasar at z=3.307 Ra,dec 0.1415, 5.9649 The red images are DECaLS z band and DECaLS Residuals, quite some heavy stuff going on if it's that visible in this redshift range (although seeing isn't lineair within our weird universe, I know I know...)
0.1316, 6.8266
0.3175, 9.5813
ra dec 0.1523, 5.7008 0.1415, 5.9649 0.2264, 6.1985 0.1951, 6.0326 0.1824, 6.3516 0.2989, 6.5734 0.3365, 6.8567 0.2069, 6.8324 0.1316, 6.8266 0.1329, 6.8267 0.2298, 6.8080 0.1799, 6.7768 0.1839, 7.0875 0.2642, 7.0202 0.3139, 6.9961 0.2273, 7.3426 0.2091, 7.2861 0.2333, 7.5629 0.2329, 7.7114 0.2109, 8.1920 0.1602, 8.0372 0.1365, 8.4274 0.2690, 8.2780 0.2715, 8.2787 0.1663, 8.6121 0.2516, 8.5687 0.1732, 8.4658 0.1400, 8.8492 0.3378, 8.7482 0.1910, 8.7544 0.1989, 8.7418 0.1422, 8.7405 0.1436, 9.0249 0.1665, 9.0207 0.2500, 8.9784 0.2589, 8.9446 0.1888, 8.8925 0.1940, 9.1415 0.3257, 9.4779 0.3175, 9.5813 0.1619, 9.8806 0.2261, 9.8729 0.1763, 9.8632 0.2384, 9.7999
Posted
-
ra, dec
0.2226, -10.9710
0.2285, -11.0141
0.1928, -11.1033
0.2403, -11.1599
0.2796, -18.8931
0.2000, -18.9199
0.1176, -18.9679
0.2003, -18.9590
0.2000, -18.9618
0.2021, -18.9615
0.2307, -18.9551
0.2300, -18.9623
0.2309, -18.9530
0.2319, -18.9608
0.3182, -19.0504
0.1743, -19.0875
0.1773, -19.0865
0.2061, -19.3583
0.2753, -19.3392
0.2728, -19.3375
0.3355, -19.4026
0.2424, -19.4392
0.3486, -19.7701
0.3143, -19.7507
0.2065, -19.8603
0.1942, -19.8811
0.2546, -19.8876
0.2569, -19.8887
0.2340, -19.9182
0.1904, -19.9338
0.3176, -19.9393
0.3331, -19.9358
0.1334, -20.1614
0.1590, -20.2458
0.1621, -20.2475
0.1628, -20.2475
0.1659, -20.2473
0.2389, -20.2727
0.3387, -20.2875
0.3187, -20.2950
0.1654, -20.6025
0.2896, -20.6398Posted
-
Candidates + hitlist
3 hits, possibly a MW star but their seems to be stuff around in PanSTARRS too
0.3353, -21.0777
Seems like a high z starforming galaxy. Pixel hit on top but also has an extended structure at the bottom.
0.1415, 13.8560
ra, dec 0.2279, -20.6488 0.2504, -20.6730 0.2284, -20.6782 0.3087, -20.6860 0.3404, -20.7058 0.3397, -20.7049 0.2642, -20.7042 0.1302, -20.7291 0.1685, -21.0242 0.3331, -21.0785 0.3368, -21.0770 0.3372, -21.0816 0.2704, -21.0985 0.2348, -21.0901 0.1861, -21.0917 0.3182, -21.1654 0.1363, 11.2717 0.2556, 11.1767 0.1745, 11.1465 0.3017, 11.4073 0.2529, 11.4089 0.2923, 11.3808 0.2964, 11.3808 0.1750, 11.3698 0.3435, 11.7363 0.2932, 11.7185 0.1400, 11.6587 0.3221, 11.5342 0.1530, 11.5712 0.3221, 11.9327 0.3091, 11.8656 0.1899, 11.7960 0.2743, 12.1259 0.1651, 12.0904 0.2771, 12.3997 0.1330, 12.3055 0.2347, 12.8129 0.1206, 12.7946 0.1657, 12.7505 0.3068, 12.7305 0.2345, 12.6783 0.1283, 12.6869 0.1278, 12.6707 0.3041, 12.9855 0.1990, 12.9376 0.3207, 12.9302 0.3320, 12.9043 0.2496, 12.8911 0.2561, 12.8546 0.1855, 13.2573 0.2882, 13.1805 0.2980, 13.1605 0.2190, 13.1572 0.2658, 13.3070 0.1219, 13.2923 0.2900, 13.6268 0.2375, 13.6317 0.2353, 13.6340 0.1415, 13.8560 0.1859, 13.8783 0.2827, 13.8589 0.2761, 14.0444 0.2029, 14.3737 0.2279, 14.3644 0.2758, 14.3618 0.1589, 14.4643 0.1557, 14.4340 0.2832, 14.7843 0.1847, 14.7409 0.2408, 14.7342
Posted
-
Candidate + hitlist
Hit near the top of the galaxy, big cloud above that area at 12:30
0.2338, 15.6700
ra, dec 0.3262, 14.9893 0.2727, 14.9635 0.2206, 15.1907 0.1399, 15.1842 0.1890, 15.1271 0.1331, 15.0897 0.1609, 15.0695 0.2209, 15.4544 0.3376, 15.4251 0.1380, 15.3742 0.2481, 15.3646 0.1925, 15.3118 0.2338, 15.6700 0.2623, 15.6426 0.2373, 15.6061 0.3081, 15.4932 0.2231, 15.9057 0.2561, 15.8967 0.2454, 16.1196 0.1201, 15.9264 0.2893, 16.3509 0.2783, 16.3568 0.2490, 16.3474 0.1497, 16.3249 0.1755, 16.3170 0.2177, 16.2906 0.1291, 16.1790 0.2916, 16.5340 0.2939, 16.5353 0.2005, 16.5195 0.1396, 16.5130 0.1173, 16.4831 0.2847, 16.3803 0.3260, 16.7155 0.1966, 16.6747 0.1434, 16.6855 0.1567, 16.6525 0.1534, 16.6350 0.2623, 16.6346 0.1922, 17.0036 0.1800, 16.9601 0.1399, 16.8943 0.1655, 16.8586 0.2100, 16.8123
Posted
-
Pitched the JPG pixel-based search method at SDSS / Karen Masters but immediately hit a brick wall. Photometric database queries already do what you want to do, you need to use FITS files instead of JPG’s due to many reasons etc.
I wonder, am I that bad in explaining things or is this project / method really that sucky compaired to existing methods?
In any case more reason to carry on and find out, time will tell. Anyone fancy candidates so far or meh?
Posted
-
by CeciliaB
I've checked quite a few of the coordinates. The trouble is, the procedure is a bit time consuming, especially the emptying of the Image List box, before you paste the coordinates. Also, the targets are often very distant which makes them difficult to assess.
But do carry on, one never knows what may pop up one day : )
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Ah yes I see. With candidates I mean only the ones that I explicitly post as candidates with images. All the other coordinate hitlists are real pixel hits but I don’t consider them to be actual Voorwerp candidates. I just post them for completeness or in case anyone is interested in checking to see what I’m getting (like you 😃 )
Thanks!
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Was wondering, how do you empty the Image list box?
I click in it, use Cntrl+a to select all and then hit Delete, about 1sec work for me. (or hold-click and swipe over all prefilled coordinates to select all)
Posted
-
by CeciliaB
I click in the Image List box and then press the Delete button and hold it in, until all the digits which keep coming up for a long time have disappeared. That takes quite a few seconds.
I didn't think one could get rid of digits which hadn't yet appeared, by using your method. lol, I have no webb skills 😄
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Ah I see that takes time yes.
I usually just copy ra+dec, click in ra box, ctrl+a ctrl+v, delete dec coords, click dec box, ctrl+a ctrl+v delete ra coords.
But point taken will post links with future candidates.
Posted
-
by CeciliaB
That method didn't work for me, but your description of how to empty the Image List box is a good tip. I'll stick to that procedure.
And a link is always helpful : )
Posted
-
Candidate + hitlist
Promising candidate, looks like a merger possibly triple (quadruple?), with at least 1 AGN involved. I've marked the first hit, a second magenta area and another bluish area at 9:00 from the main merging galaxies with the SDSS triangle markers.
0.2223, 18.5087
Likely merging pair 2MASX J00005427+1830215 in SIMBAD
ra, dec 0.2803, 17.2331 0.2236, 17.1925 0.1280, 17.2008 0.2353, 17.0680 0.1395, 17.0378 0.2759, 17.4510 0.2359, 17.4268 0.1848, 17.4421 0.1151, 17.2513 0.1166, 17.2528 0.1239, 17.6706 0.1420, 17.5008 0.3131, 17.4939 0.1144, 17.8967 0.2207, 17.8044 0.3126, 17.7842 0.1461, 18.0501 0.1469, 17.9860 0.1847, 17.9185 0.1860, 18.3056 0.1205, 18.1251 0.1380, 18.1325 0.1401, 18.1328 0.2223, 18.5087 0.2672, 18.4264 0.1382, 18.4200 0.2076, 18.3907 0.2151, 18.3948
Posted
-
by CeciliaB
Wow! This must be your most promising candidate so far. The scenario with the blue and magenta coloured merging galaxy, which has an AGN, together with the strong tidal tails, one leading up to an interacting galaxy, ought to be a perfect condition for a voorwerp to arise.
The colour of the object at 9 is right for a voorwerp too, I think. Without doubt a great find!
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Thanks, does look good. But confidence dwindling a bit because ive also hit a few starforming patches before. (And didn’t get a hit on the blue patch...)
Posted
-
Faint candidate + hitlist
Very faint but looks like an reddish arc-shaped structure at 9 - 11 o'clock from the pixel hit
0.3293, 18.8133
SDSS J000118.74+184857.4 - 1237679502171504804
http://legacysurvey.org/viewer?ra=0.3281&dec=18.8160&zoom=16&layer=sdss2
ra, dec 0.2535, 18.7799 0.1399, 18.7640 0.1407, 18.7635 0.2718, 18.7162 0.2723, 18.7156 0.2727, 18.7148 0.3068, 18.9596 0.1677, 18.9487 0.3036, 18.8782 0.2949, 18.8724 0.3293, 18.8133 0.2937, 18.8053 0.2537, 18.8212 0.1287, 18.8106 0.1981, 19.1187 0.2798, 19.0791 0.1831, 19.0561 0.2357, 19.0417 0.3415, 19.0174
Posted
-
by CeciliaB
The arc almost looks as if it is part of a tilting ring.
Posted
-
Yeah, if it isn’t just random noise does look like it could be a tilted ring with another arc more to the right side. Excellent observation! 😃
Posted
-
Posted
-
Candidate + hitlist
Looks like some sort of streamer / outflow towards 6 o'clock, particularly visible in Panstarrs r band with adjusted autoscale
Ra,dec 0.2414, 20.1658
http://skyserver.sdss.org/dr12/en/tools/explore/Summary.aspx?id=1237680246814474485
http://legacysurvey.org/viewer?ra=0.2415&dec=20.1657&zoom=16&layer=sdss2
ra,dec 0.2344, 19.4171 0.2324, 19.3910 0.2676, 19.2309 0.1325, 19.5462 0.2385, 19.5888 0.2542, 19.5418 0.2492, 19.8315 0.3224, 19.8154 0.2344, 19.7729 0.1395, 19.7574 0.1380, 19.7585 0.2373, 19.7481 0.2664, 19.7426 0.1772, 19.6770 0.2094, 19.7127 0.2444, 20.1003 0.1583, 20.0589 0.1614, 20.0585 0.1582, 20.0631 0.1562, 20.0635 0.1789, 20.0302 0.1766, 20.0279 0.1751, 20.0256 0.1684, 20.0244 0.1889, 19.9383 0.2073, 19.9565 0.2920, 19.9673 0.3121, 19.9285 0.2332, 19.9085 0.2881, 20.2855 0.2414, 20.1658 0.1751, 20.1091 0.1195, 20.5202 0.1787, 20.5133 0.1753, 20.4323 0.3325, 20.3379 0.2232, 20.3591
Posted
-
Candidates + hitlist
Ra,DEC 0.2319, 20.5775
(not an SDSS photometric object)
http://legacysurvey.org/viewer?ra=0.2319&dec=20.5775&zoom=16&layer=sdss2
Ra,DEC 0.2953, 21.1765
(not an SDSS photometric object)
http://legacysurvey.org/viewer?ra=0.2952&dec=21.1765&zoom=16&layer=sdss2
0.3151, 20.6894 0.2319, 20.5775 0.2261, 20.5630 0.3415, 20.7832 0.3404, 21.1802 0.2953, 21.1765 0.1667, 21.0996 0.3427, 21.0697 0.1542, 21.3317 0.1768, 21.2287 0.3364, 21.5435 0.2971, 21.7400
Posted
-
Hitlist
ra,dec 0.1969, 22.0128 0.3276, 21.9501 0.2679, 21.9347 0.1483, 21.9358 0.1208, 22.2793 0.2071, 22.2423 0.2794, 22.2271 0.2472, 22.1628 0.2548, 22.4664 0.3376, 22.4510 0.1684, 22.3190 0.2845, 22.7350 0.2963, 22.6661 0.2460, 22.8618 0.1525, 23.1544 0.3262, 23.0853 0.2480, 23.0408 0.1445, 23.0247 0.2594, 23.2102 0.2292, 23.1929 0.3040, 23.5551 0.2580, 23.5335 0.2578, 23.4967 0.1880, 23.4409
Posted
-
Candidates + hitlist
Looks like two (SDSS greenish) streams on opposite sides, also visible as 'blobs' in DECaLS and PS1 at 8:00 and 2:00 o'clock
Ra,dec 0.3273, 23.8629
http://legacysurvey.org/viewer?ra=0.3273&dec=23.8627&zoom=16&layer=sdss2
Ra,dec 0.1661, 23.8590
http://legacysurvey.org/viewer?ra=0.1662&dec=23.8585&zoom=16&layer=sdss2
Can be pasted in the SDSS DR12 Image List Tool
ra,dec 0.2952, 23.8347 0.1068, 23.8075 0.1296, 23.6631 0.3273, 23.8629 0.1661, 23.8590
Posted
-
Candidate + hitlist
At first look the magenta feature at the bottom looks like an reflection artifact. But upon closer inspection it seem to be too strainght for an artifact, and it can't be found below other bright objects nearby. From PS1 it is clear there are definitely two objects above the feature, but no relation between them or the line feature can be deduced, neither if the two objects are two ordinary MW stars. Still really wonder if the magenta line feature really is an artifact, PS1 / DECaLS not of any help here.
Opinion anyone?
Pixel hit coordinates
Ra,dec 0.3297, 24.4427
Ra,dec 0.3283, 24.4419
http://legacysurvey.org/viewer?ra=0.3278&dec=24.4396&zoom=16&layer=sdss2
Hitlist
ra,dec 0.3478, 24.2804 0.2654, 24.1819 0.3215, 24.4928 0.1607, 24.4674 0.3283, 24.4419 0.3297, 24.4427 0.2803, 24.3775 0.2751, 24.2805 0.1952, 24.2850 0.1305, 24.2931 0.3202, 24.7079 0.1739, 24.7131 0.3039, 24.5155 0.2205, 24.5352 0.3437, 24.9084 0.1133, 24.9191 0.1263, 24.8343 0.1961, 24.7500
Posted
-
by CeciliaB
I'm sorry, but I can only see one straight line below the two (what I think must be) MW stars, and that looks like one of four diffraction spikes. Is that the magenta line feature you mean?
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Refraction spikes always radiate outwards from the center of a bright object, not perpendicular to the object. Even if it was an optical artifact it would have to be an arc or circle with the bright object as center. Also other bright objects would have to show the same artifacts.
If they are MW stars or not I don’t know, how to tell the difference for sure?
This is the magenta line I’m talking about. I think it will be good for me to check the FITS files later.
http://legacysurvey.org/viewer?ra=0.3273&dec=24.4396&zoom=16&layer=sdss2
Thanks for taking a look 😃
Posted
-
by CeciliaB in response to Ghost_Sheep_SWR's comment.
Ah, of course I saw that line straight away in the DECaLS images, but I didn't think that that was the line you meant, since I couldn't see it in the SDSS images.
To me the line does look like some sort of artifact, being too straight and sharply defined to be anything natural. Isn't there a similar, weaker one visible below the first one in the zoomed out image?
With such big diffraction spikes my guess is that the two objects must be MW stars.
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Yes it’s straightness does imply artifact, but I’ll not be foiled by that! Everything else points at not an artifact so far..
The DECalS image is ofcourse SDSS image in DECalS viewer so that data is hidden in SDSS.
Not sure which weaker one you mean, can’t seem to spot t so far.
Diffraction spikes to me mean nothing else as something being bright, although they do appear as MW stars.... Perhaps GAIA DR2 might reveal more, or checking DSS images for movement.
Posted
-
by CeciliaB in response to Ghost_Sheep_SWR's comment.
Perhaps I'm mistaken, but I think I can see a second dimmer and more uneven magenta line in the same direction as the first line, about 2 cm from the bottom of your zoomed in image.
Sorry, I didn't notice that the DECaLS image was a SDSS image in DECaLS viewer.
We'll see what your follow-up may lead to... 😃
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Ok thanks, appreciate your feedback 😃
I’ll check it out later, cant say I’m seeing it now but you might be right!
Posted
-
Yeah must be an artifact (or something transient, maybe in Earth's atmosphere?). Can only find it in SDSS g band, and luckily there are two SDSS coverage images; one that contains the streak and one that doesn't. So I guess you're right about it being too straight to be real 😃
Posted
-
Hitlist
ra,dec 0.1848, 25.0689 0.1478, 25.0576 0.2865, 25.3392 0.1146, 25.3474 0.3340, 25.2927 0.2049, 25.2025 0.3478, 25.5369 0.3508, 25.5389 0.3549, 25.5220 0.2189, 25.5210 0.1165, 25.4787 0.1948, 25.7929 0.2855, 25.7483 0.3410, 25.6882 0.3355, 25.6510 0.3392, 25.6026 0.2183, 25.9124 0.1726, 25.8413 0.3460, 25.8823 0.3490, 25.8803 0.3525, 25.8823
Posted
-
by CeciliaB
Better luck next time 😃
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
No luck! I've got meself a METHOD HAHAHA
Perhaps not a 'real' candidate but this one looks just odd:
http://legacysurvey.org/viewer?ra=0.1744&dec=26.1710&zoom=16&layer=sdss2
Candidates + hitlist
Seems like a decent outflow / loop at 3 o'clock
0.2296, 26.1005
http://legacysurvey.org/viewer?ra=0.2302&dec=26.1004&zoom=16&layer=sdss2
Sinlge hit in between Mrk 1137 and MCG+04-01-009
0.1326, 26.3163
http://legacysurvey.org/viewer?ra=0.1329&dec=26.3160&zoom=15&layer=sdss2
ra,dec 0.1743, 26.1710 0.2446, 26.1614 0.2043, 26.1077 0.2296, 26.1005 0.1683, 26.3735 0.1688, 26.3682 0.2922, 26.3646 0.2855, 26.3317 0.1326, 26.3163
Posted
-
by CeciliaB in response to Ghost_Sheep_SWR's comment.
I'm sure you'll have a find sooner or later 😃 Remember the old saying: If at first you don't succeed, try, try again.
The second candidate is interesting with that mysterious dark magenta cloud! It looks as if it's in front of the bright round object. But if that is a star, the cloud would have to be behind it of course. One is easily deceived when it comes to deciding which object is in front of the other.
Posted
-
Today a very magenta object hit, a Voorwerp candidate without a pixel hit + hitlist
Not a candidate but interesting colour, very likely a good purple pea candidate. Discussion please in this thread thanks
ra,dec 0.3158, 27.0070
SDSS / DECaLS SDSS & PS1 gri images
Candidate without a pixel hit, stumbled upon it while examining a hit above it, Frustrating that I can't seem to find it in DECaLS or PS1 images....
ra,dec 0.3560, 27.3080
Nearest SDSS photometric object is SDSS J000124.77+271830.0
http://legacysurvey.org/viewer?ra=0.3560&dec=27.3080&zoom=16&layer=sdss2
Hitlist
ra,dec 0.1115, 26.8308 0.1248, 26.8039 0.1236, 26.8035 0.1207, 26.7610 0.2717, 26.7461 0.2864, 27.1405 0.2843, 27.1392 0.2241, 27.1255 0.3004, 27.0962 0.3259, 27.0978 0.3159, 27.0383 0.3158, 27.0070 0.2833, 26.9508
Posted
-
Candidates + hitlist
0.2440, 28.6212
http://legacysurvey.org/viewer?ra=0.2440&dec=28.6211&zoom=16&layer=sdss2
0.2162, 28.6110
Nearest SDSS photometric object SDSS J000052.00+283643.5
http://legacysurvey.org/viewer?ra=0.2162&dec=28.6110&zoom=16&layer=sdss2
ra,dec 0.2420, 27.2313 0.2429, 27.2326 0.2123, 27.2094 0.2675, 27.5159 0.3011, 27.5067 0.3037, 27.5058 0.3546, 27.3920 0.2632, 27.7457 0.1515, 27.7092 0.3371, 27.7102 0.1800, 27.9784 0.2072, 27.8929 0.2039, 28.1888 0.1502, 28.1665 0.1707, 28.0652 0.1166, 28.0589 0.1290, 28.0356 0.1289, 28.0366 0.3073, 28.0340 0.2705, 28.4120 0.1185, 28.3576 0.2052, 28.2778 0.1819, 28.6687 0.2440, 28.6212 0.2162, 28.6110 0.1919, 28.4922 0.3307, 28.5277
Posted
-
by CeciliaB
Funny that - in the first candidate - there seems to be such a blue outflow at three o'clock from the biggest galaxy in DECaLS SDSS images, whereas in SDSS the outflow towards six o'clock is more visible.
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Yeah definitely, DECaLS viewer renders them much brighter, and also posting original SDSS images here seems to put an extra layer of ‘darkening’ over them. A bit frustrating because what I see doesn’t display here so well :\
That’s why I also post ‘enhanced SDSS images’, just did so for the last two candidates where they are much better visible. But it’s an extra step in the process and you caught me in between 😄
FYI first upload image here while on PC, save it on my smartphone, edit on smartphone, upload to imgur and then add it here in the existing post.
Posted
-
by CeciliaB
Many steps, besides checking all the hits! But it must be exciting to see what pops up 😃
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Yes and no.
Getting waaay to many hits mainly from bright stars. Voorwerp colours seem to be exactly the same range as all the magenta star / bright object artifacts so it becomes a bit of a drag too. At the start I imagined I would a have covered a couple of degrees by now. But still wrestling with the first half degree lol.
On the upside for most of the recent candidates I would have never found the ‘hidden’ magenta blobs buried in the data, so yeah thats fun to make em pop out by editing an image 😃
Posted
-
by CeciliaB
I hear that it's time-consuming, but you don't have to rush, do you? To do a little now and again would be enough, wouldn't it, and that would help to make the work even more enjoyable, I think 😃
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Oh yeah definitely! From the start however I was very eager to ‘clear’ SDSS and then move on to the DES field. Ah well AFAIK there isn’t a known Voorwerpje to use as pixel sampler in the Southern hemisphere yet... or is there Bill Keel?
Thanks for your support! 😃 😄
Posted
-
Hit on the top-side of a Broadline QSO. Not sure if this should be tagged as candidate, but it does seem to have some notable features such as protrusions, faint clouds closeby and a red splotch on top in PS1 y band.
Ra,dec 0.2405, 29.7105
http://legacysurvey.org/viewer?ra=0.2408&dec=29.7102&zoom=16&layer=sdss2
last two PS1 y/i/g and PS1 gri.
Also notable is that I got 5 hits in a circular pattern around this source. That might sound like a way artifact pixels would appear in my search method, but the opposite is true. Almost all artifact colour pixels around bright sources are either just one pixel on one side (right-side usually magenta-ish), or two pixels on opposite sides of the source. How this one registered is quite rare. But still big chance this is just a very bright MW star with artificial magenta colours around it, can't really find anything useful on it so far.
Opinions?
http://legacysurvey.org/viewer?ra=0.2088&dec=28.9135&zoom=16&layer=sdss2
Hitlist,can be pasted in the SDSS image list tool.
ra,dec 0.3236, 28.8957 0.3145, 28.8947 0.3205, 28.8334 0.2449, 28.7769 0.1902, 28.7727 0.1898, 28.7733 0.2068, 28.7601 0.1328, 28.6852 0.3024, 29.1111 0.2892, 29.0628 0.2872, 29.0595 0.2846, 29.0596 0.1567, 29.0670 0.2239, 28.9792 0.3460, 28.9952 0.2332, 28.9045 0.2077, 28.9121 0.2097, 28.9121 0.2108, 28.9130 0.2114, 28.9140 0.2088, 28.9154 0.1637, 29.2853 0.2909, 29.3284 0.1276, 29.2696 0.1464, 29.2432 0.1550, 29.2367 0.2873, 29.2695 0.3028, 29.1957 0.2891, 29.1808 0.2891, 29.1819 0.2607, 29.1483 0.2216, 29.1780 0.3195, 29.5580 0.1705, 29.5381 0.1531, 29.4319 0.1745, 29.3786 0.1430, 29.7482 0.2405, 29.7105 0.2731, 29.6736 0.3593, 29.9774 0.1862, 29.9855 0.1675, 29.9656 0.1866, 29.9032 0.1100, 30.1690 0.1805, 30.1613 0.1368, 30.1435 0.2740, 30.1170 0.2014, 30.0855 0.1662, 30.0317 0.1737, 30.0163
Posted
-
by CeciliaB
Interesting with that protrusion from the QSO, even if it may not be a candidate.
The second hit, unfortunately, seems to show one of those magenta bright object artifacts you keep getting. As you would have seen, other stars in DECaLS SDSS images have the same colouring. One wonders why this particular star was picked out by your method?
Posted
-
by Ghost_Sheep_SWR in response to CeciliaB's comment.
Not sure if it can be considered a candidate, perhaps a bit too faint to be sure. At least the spectrum supports high energy activity.
On the second hit; I picked it out because of the many pixel hits in circular shape, but I do get a lot of hits on the other MW stars magenta artifacts. Just getting better at ignoring them, now I’m getting much quicker in dismissing them. At first I was afraid of perhaps dismissing one that’s actually a Voorwerpje, now I just think if the hit is inside the bright / magenta part it doesn’t matter because it wouldn’t be a Voorwerp anyway; too close to the object / nucleus 😃
Posted
-
ra,dec
0.1034, 30.3614
0.2546, 30.5474
0.3000, 30.5235
0.1926, 30.4970
0.2298, 30.4758
0.2097, 30.4512
0.1687, 30.4504
0.1048, 30.4400
0.1957, 30.7855
0.2168, 30.6643
0.1526, 31.0831
0.1498, 31.0865
0.1110, 31.0938
0.2487, 30.9997
0.2456, 30.9961
0.1865, 30.9742
0.3112, 30.8807
0.1031, 31.2689
0.2397, 31.2836
0.2427, 31.2867
0.2771, 31.2641
0.2496, 31.2055
0.1656, 31.1857
0.1467, 31.1469
0.1471, 31.1475Posted
-
project temporarily suspended due to another faster search method in DES and SDSS / DeCALS.
But the pixel hunt will continue! 😃
Posted