#apertus IRC Channel Logs

off to bed now ... have a good one everyone!

good evening, meeting in 5 minutes

Today we'll do a special edition of our IRC meeting ... first because it will be the last before GSoC 2018 ends and then we want to recapture what was done during GSoC and figure out what we can change in the future

meeting time!

I've opened all the final GSoC project proposals you (the students) submitted as well as the task descriptions to each of the selected tasks

okay! :)

nmdis199-: please go ahead

So, most of my tasks were finished last week, I talked to alex about it and started documenting and creating github.io repo

All tools which were promised in timelines are ready and working on beta, the test result images are uploaded in doc, I have added link to source code and documentation here -> nmdis1999.github.io

That's all

lets see if anyone else is arround?

nmdis199-: thanks, so looking at the submitted proposal, I see ...

ArunM: ?

rahul__: ?

nmdis199-: what of those parts were implemented?

I made tool histogram.c instead (recommended by you) instead which was addition, I did suggested the image cropping algo in first month if I remember correctly.

ymc98: ?

I've talked about adding some better features in waveform and vectorscope tools, will talk to alex again soon

hello everyone

sebastian checked them :)

:)

I wanted to ask alex, that he worked on similar kind of tools in magic latern, if I remember our discussion, I'd like to read his codes and work and improve the current tool

I only see some plain histograms (i.e. no RGB) and some scatter dots labeled as "waveform" and "vectorscope", i.e. not exactly what I'd see on a camera screen

nmdis199-: sebastian checked them <- lets more call it, "I provided some feedback and improvement suggestions"

nmdis199-: IIRC, you are planning to spend some extra time on those tools in the forseeable future, right?

okay, so supragyaraj, you want to tell us something about what you did during GSoC and what the new features are now?

So, during GSoC, there were 2-3 different things we achieved

First, we created the way with which we will encapsulate all the meta and video frames and discussed the pathway for that

then, we created an emulation of the recording end and tested it's speed etc

after this, some time went into configuring the mlv structures to make it more suitable for beta

Finally, we shifted to the conversion from MLV to CDNG... tinkered around with MLVFS and MLV_DUMP

we added a linearization table in order to affect the CDNG according to the theoretical PLR block that we proposed

this is the final work product

and documented those....

MLV was still the most straightforward solution we can get

basically on 2 domains:

2. change current mlvfs/mlv_dump code in order to make PLR changes and add beta's color matrix among other things

I plan to... yes..

great to hear! :D

also a great thing!

rahul__: I think you're up next

yes please rahul__ go ahead

so, I had finished my IP design as well as successful simulations in the first two months. the last month was trying to place the IP in the beta pipeline, then the last week I visited the office in Vienna to expedite the work. after discussions with Bertl, I placed the IP in the output pipeline, (just after the frame reader) but after some experiments it shows that the pipeline was not able to handle delays (if the IP is

but I am not sure wether this would be possible by today

Hello everyone. I made a blog post describing the project, my progress and what's left to be completed. Here's a link : https://ymc98472005768.wordpress.com. Made few changes in the design for transceivers for MachXO2. With these changes incorporated, there isn't any requirement of having a 10 bit dummy data between packets for the serialization and deserialization. Also, the documentation to my code can be found here. https://docs.google.com/document/d/1r7

I will also be submiting

the presentattion as well as report will be on my github link soon today

Sorry. It's an accident

rahul__: checking with the original proposal, it looks like you are in Week 10 of your plans ...

yes, week 10 to 12 was to implement in the beta...but its not working properly as you know,,,, that was the main reason i came personally to vienna

thanks rahul__

rahul__: integrating new code into existing designs is always tricky

okay, do you plan to continue working on this after GSoC?

glad to hear!

also as we had discussed

about that: I installed a HDMI to USB capture card in the remote beta setup today

we still need to get it to work under linux though

Sorry for the interruption before. As i mentioned previously I made a blog post describing the project, my progress and what's left to be completed. Here's a link : https://ymc98472005768.wordpress.com. Made few changes in the design of transceivers for MachXO2. With these changes incorporated, there isn't any requirement of having a 10 bit data gap between packets for the serialization and deserialization. Also, the documentation to my code can be found he

so, what is working as of today?

SERDES transceivers for zynq and machxo2

okay, that's week 3 in your final proposal

Not yet. Working on it.

That's a yes

good to hear!

so, unless I miscounted, we had four reports, right?

Tofulynx is missing and ArunM

thought he would return today, as he said 10th to 13th

give me a sec

but overall we have a very good progress, which still needs evaluations, but this relies on the architecture etc.

Claudio has implemented and improved plain debayer processors, also ones in openmp and also fixed parallelization problems there

no email from google so far, probably it's org related

nmdis199-: supragyaraj: rahul__: ymc98: what could be improved for next years GSoC?

sure, that's why I ask :)

I think I loved the challenge part (before applying), adding familiarity with linux in project/task description could be helpful. I really liked the weekly meetings, so that should definitely stay :)

Overall I really enjoyed working here, I learned a lot. This being my first project had a really positive impact on me :)

As my experience, I really liked the way mentors responded and helped all along the way, The only issue that I faced (may not be the same for the rest) is the lack of detailed documentation of the code as most of the productive time went into that, so if the organisation could support the code base with proper documentation, then that would be really helpful for both the next lot of students as well the organisation, and as I

Well, i found this with magic lantern too....

code was not in one place... maybe I am wrong too...

but could not find a central place to find my references

Bertl: Like RexOrCine :)

yes, the main problem is that most folks who code/build do not want to spend time on documentation and those who like documenting often do not have the insight required ... it needs a lot of effort to make it happen :)

that said, documentation is very important and I'm happy to hear that you want to help out there

anything else you'd like to see improved in the future?

"practically", not talking about utopian world

yes, surely I think we GSOC students can contribute to this organisation in a big way like that after the GSOC ends....

by preparing a well documented code base

which kind of code docs would be suitable?

@students: okay then, thanks for the feedback and hope to see you all around in the future :)

:)

*doxygen adds

Btw, where can we get hold of AXIOM merchs? those tees look awesome ;)

ah good idea

only a few very small or very large ones left

very small would work for me :P

great :D

I think we will manage to send you some other stuff as well, like stickers and such ...

will the t-shirt print layouts be open-sourced? ;)

you will get it :)

Bertl: you said what we just wanted to say.... :)

I will get in touch with you when the time comes about tshirt sizes, shipping addresses, etc.

very happy to hear

Thank you, se6astian!

indeed, thanks to everyone and keep up the good work!

with an open source projects its always a coming and going and its hard to find long term contributors, having you around made everyone else more active and dedicated as well :D

we had them before GSoC, so I don't see a problem having them in the future as well

ok meeting concluded for today, you are very welcome to stick around (short term and long term) and feel free to ponder your mentors about tasks now after gsoc

Gn