#apertus IRC Channel Logs

morning folks!

Good Afternoon!

I won't be able to make it for the weekly meeting today. I came down with fever and going to bed early. Nothing much to report for this week. Have been working on MachXO2 serdes.

understood, all the best!

Ok! :)

Helllo nmdis199-!

evening!

ymc98_2 told us earlier to be sick and not able to attend today

nmdis199-: you are first, please go ahead

Last week was pretty good, after talking to sebastian on irc I did wrote a code for calculating luminance and it showed result however, since I didn’t know what data was being provided by dev/mem I couldn’t have verified it.

hello!

Now, this was the point where my code on running on beta showed Segmentation fault.

Therefore, mid-week Bertl asked me to write modified version of mimg (which is in processing tools) and explained me what it does and how.

it is also uploaded on the beta

This can be useful for future purposes as well as on my waveform tool for writing an image to the memory which is a much better way.

that's all :)

load_raw only deals with raw image, because it wasn't wise to use register

bascially the counterpart to cmv_snap3

ah, so you fill the framebuffer instead of reading it out

and the idea is to write the histogram data into the framebuffer so cmv_snap can read it and save it into an image correct?

the main idea is to be able to run the various tools (waveform, histogram, etc) on 'known' good test data (raw images)

great, many thanks for reporting!

otherwise TofuLynx is up next

Ok :)

yes, sure go on :)

https://docs.google.com/document/d/18o9ajnAF1_DEy2CfYi_jYxDmurW-KEtojqL3_c3SPBU/edit?usp=sharing

So far, the only thing I have to finish is the SHOODAK2 documentation. The documentation helped me notice some areas of the code that I can optimize

By the way, I didn't do so much last week because I was occupied with my last university exam

may I suggest to add some example debayered images in the documentation

Thanks for the suggestion!

:)

https://www.irccloud.com/pastebin/907fvQuD/

I have made the required ‘focus peaking kernel’ as reported in the last meeting,

This time I ran functional and timing tests so as to ensure the proper functioning of the module.

In the coming week I will be done with my university exams and will try to incorporate my design into the existing Pipeline.

ya the link was deleted

rahul_ do you have a way to simulate the focus peaking kernel output with a provided input image?

I dont have a way to simulate.

simulation would be possible but rather tricky

yes

I upgraded the remote beta webcam

as the design is for the output pipeline, it would also be possible to test it with uploaded raw images on the Beta

ok, once I finish the incorporation of my design in the pipeline, I will start testing it on beta

I think so

it would be possible to store the peaking output into memory as well

(but we'll see what we can do once we get there :)

sure

supragya is up next

so, the last week, I added linearization table into already existing MLVFS

however, from mlv, dngs are coming out fine with linearization table effecting the colors

then, I will look into new blocks in mlv that needs to be added for our use and implement it

(I updated the status to the mentors using the email... last friday if I am not wrong)

any images I can share via social media about this?

also from any other students: rahul_ TofuLynx nmdis199- ?

so I am not sure if that serves the purpose

that would be great!

Okay, then will do upload a blog (preferably on medium) :)

supragya, you did made a video right? Is it working the Qt one ?

these (above) are the images...

the linearization table was synthetic... but modifies MLVFS for us...

nmdis199-: not sure what you are asking me...

The video for building up OpenCine (the one which I tried making on my ubutnu :P )

its in here now: https://www.apertus.org/opencine

TofuLynx: yes please

Do you have any preferred image?

the axiom labs one would be great

se6astian ,I'll email the blog you then, I don't have much images right now (because we haven't plotted waveform tool yet) so will that do?

the image situation should soon change when you can write and read images to/from memory

yes it should be :) will do

sure!

so that is progressing

we secretely updated our brochure: https://apertus.org/AXIOM_Beta_Brochure_02.04.pdf

work on finishing the compact enclosure is also slowly progressing

finally we did some extensive testing on the laser cut enclosure which will also be quite helpful for the future Compact version of the Beta

the data should be available today or tomorrow

thats it from my side!

Great :)

feel free to use the channel for general discussion/chatting again

bye!

rahul_: one way to test your blocks easily would be to wrap them in a simple HLS design

reading the data from DDR and writing the resulting image back into DDR

TofuLynx: https://www.apertus.org/logo

thank you!

it's titillium?

I will be away for a moment

Portugal is playing now

off for a nap ... bbl

haven't written this here yet: since i have some space left on the edge of the sdi module, i'll add a legacy synchronization input to the board; apparently that's something that is still used in production today

oh and fun fact: the version in the bigger package of the clocking chip i chose for the sdi board is also used in the new readout system for the atlas experiment at cern