Closed jovo closed 8 years ago
@Aeusman Let's look at this after you are done with finals.
@kunallillaney good call! want to put a milestone with a date on this? up to you guys when, maybe last monday of may?
I don't want to commit to a milestone for @Aeusman. Since he might have other things in the pipeline. I assigned it to him for now so we don't drop the ball over summer.
cool!!! can you send a ndviz link? @alexbaden
On Mon, Jul 25, 2016 at 12:35 PM, Alex Eusman notifications@github.com wrote:
http://openconnecto.me/ocp/ca/allen_gold_16_012/image/xy/0/0,1000/0,1000/46/window/0,100/
Ingested
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/neurodata/ndstore/issues/259#issuecomment-235008028, or mute the thread https://github.com/notifications/unsubscribe-auth/AACjcvTFA__AMzYII4Nj6bf0mQSMvJg6ks5qZOXHgaJpZM4IBicj .
the glass is all full: half water, half air. neurodata.io, jovo calendar https://calendar.google.com/calendar/embed?src=joshuav%40gmail.com&ctz=America/New_York
Traveling today. viz.neurodata / token / channel /
Will need to be windowed in the console first or you'll cache black tiles.
Alex
On Mon, Jul 25, 2016 at 13:21 joshua vogelstein notifications@github.com wrote:
cool!!! can you send a ndviz link? @alexbaden
On Mon, Jul 25, 2016 at 12:35 PM, Alex Eusman notifications@github.com wrote:
http://openconnecto.me/ocp/ca/allen_gold_16_012/image/xy/0/0,1000/0,1000/46/window/0,100/
Ingested
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <https://github.com/neurodata/ndstore/issues/259#issuecomment-235008028 , or mute the thread < https://github.com/notifications/unsubscribe-auth/AACjcvTFA__AMzYII4Nj6bf0mQSMvJg6ks5qZOXHgaJpZM4IBicj
.
the glass is all full: half water, half air. neurodata.io, jovo calendar < https://calendar.google.com/calendar/embed?src=joshuav%40gmail.com&ctz=America/New_York
— You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub https://github.com/neurodata/ndstore/issues/259#issuecomment-235055404, or mute the thread https://github.com/notifications/unsubscribe-auth/AFkzUkF-OfYR9hGflaZlt-tVkDVR0E14ks5qZQvfgaJpZM4IBicj .
i tried using the token listed with channel "image". it looks all black.
i tried changing the max intensity using the scroll bars. but i didn't know the values it had, and it still looked black.
The first ~15 slices appear to be mostly black/contain no data (at least for this dataset)
Yup, it will be black until a window is set in ndstore and memcached is flushed. On Mon, Jul 25, 2016 at 13:26 joshua vogelstein notifications@github.com wrote:
i tried using the token listed with channel "image". it looks all black.
i tried changing the max intensity using the scroll bars. but i didn't know the values it had, and it still looked black.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/neurodata/ndstore/issues/259#issuecomment-235057500, or mute the thread https://github.com/notifications/unsubscribe-auth/AFkzUjZ9aghY-XKgcKC6JfS27GRnBhTkks5qZQ22gaJpZM4IBicj .
shouldn't the "max" slider do that dynamically?
Nope. The problem is the 16 bit data is rendered to 8 bits on the server. The window tells the server how to handle that resampling. The client can do some minimal tweaking of the 8bit data, but if your range is 0-1000 and you sample 0,2^16 into 0,2^8, all your data is lost as noise and the 8bit image ends up black.
We could look into rendering 16 bit data on the client to get around this problem, especially since datascope / MySQL isn't reliable enough to run big histogram jobs at the moment. Something to discuss Weds or next week.
Alex On Mon, Jul 25, 2016 at 13:36 joshua vogelstein notifications@github.com wrote:
shouldn't the "max" slider do that dynamically?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/neurodata/ndstore/issues/259#issuecomment-235059244, or mute the thread https://github.com/notifications/unsubscribe-auth/AFkzUtSw2ayUHJZqA3-wGIaBVgOMis0Rks5qZQ9TgaJpZM4IBicj .
ok. let's put an issue in ndviz to create a project for this with appropriate windowing?
I think my comment was misinterpreted, even when manually windowing the data by pulling pngs the first ~15 slices are blank. but at slice 46 we can see data so I would recommend testing based on that.
https://github.com/BigNeuron/Data/releases
@Aeusman @j6k4m8 anybody want to look at it?