#apertus IRC Channel Logs

yup

good morning

Hi, a reply to a comment made yesterday --quote 20:56 aombk: this whole dslr blog/news sites thing is very disappointing. after having contacted many of them i am starting to believe most of them host ONLY paid articles/advertisements. 116 --end quote. To be honest, I think quite a few blogs have reported on it and you can't really blame them for not writing articles on Apertus daily as there is not anything really new to report

statistically campaigns gain momentum at the end of their funding period

also we will announce some more big news today that hopefully get us some attention again

Is an option for the supporters to make a deposit of 2000 Eur? Maybe you don't get all the 100000 Eur for development but you want lose all the rest...

daFred: you mean asking people for a downpayment for the 1900/2300€ alrady now?

a heatsink alone is no solution, you need to not just collect the heat but also move it away somehow

Besides the its nice to have a strong metallic body - the lens can be heavy there can be follow focus

se6astian: yes, maybe a option for the last two days. Better then a stone on the head :-) ((austrian english...))

The only reason i was not talking of the metal body with no fan earlier was bacause the crowdfunding campaign seemed and still seems a much bigger concern

anton7: where would the connectors for the modules be in your concept?

but the Beta has the extensions module at the front sides not the back

Sebastian, didn't think hard of connectors... Maybe vertical on right and left so as not to interfere with vertical air flow?

A sealed camera would have its advantages

we wil hold another survey about the enclosure and see what people prefer

Fair enough

http://www.ioindustries.com/provideo/products/4ksdi_overview.html Same sensor. Seems to do well without a fan.

That's what IO told me at IBC. They're also somewhere mentioned in a Cmosis brochure. But yes, different processing. I do think it is essential to keep the sensor dirt free. Having it different would be a major design flaw.

Both Arri (Alexa) and Sony (fs-7) are working with heatpipes. Cion seems to be pretty unprotected, though.

do not confuse 4K and UHD

Bertl: DP is not TMDS and DP is better and cheaper than HDMI

danieel: thanks for clarifying

Speaking on the cooling again - if it turns out that a metal camera body is not able to dissipate the heat from Zynq then it doesn't seem likely a small fan on the top would do the job either. Then ribs on the back side and a large external fan still look like a better choice.

we are always open to suggestions, especially if they come with good arguments, like for example test results

the tube in front of the sensor is a big heat sink. We also can add some ribs at the outside of the tube...

but note that the sensor does heat up on the back more than on the front/side

this is one concept we already considered, we will see what is doable there and what not

so it would really help to find alternative socket solutions for the CMV12k pingrid

EUR 90 I mean..

Well once the socket has been soldered to the board perhaps one could take a dremel or wire cutters to the socket to remove the side walls and allow the metal plate to go under?

Andon wasn't very cooperative in regard of changing anything, well, I'm sure they would for the right price, but do you want to pay the same for sensor and socket? :)

I had a look at the so-called peel-away sockets, which might be suitable for the purpose

even after contacting them several times and through different channels

i had no problem with andon, they were quite surprised that I know of a product which is "future" and the sockets were instantly ready (for KAC)

the campaign seems to have had a push forward for some reason

I guess it is the ML team-up

anton7: not a stupid idea

uncooled sensor without any thermal management gets to at most 45deg (internal reading).. so you should rather cool/separate the fpga, than care about a sensor (big package with relatively low power)

It would be interestng to test if the most important is having the sensor as cool as possible, or have the same temperature all the time (which is a bit different), for correct calibration for instance.

both is relevant, higher temperature on the sensor means exponentially more thermal noise

it is not about predictability, rather to say - stability

nobody said anything about predicting noise

I meant what I wrote, it gives a predictable noise pattern

so compensation and algorithms have to adapt to the changes

do you know at what temperature the fpga runs curently in alpha, and at what max temp it can run?

for FPGA tempertures on the alpha you have to ask se6astian

Thanks. 85° is pretty high, so I guess the sensor will need more cooling than the fpga.

gotta go

big update ahead tonight :)

https://en.wikipedia.org/wiki/Thermoelectric_cooling ?

it will definitely be the preferred solution for the AXIOM Gamma

Re attaching the Peltier - have you seen what that guy did with Canon d450? He attached the copper plate to the sensor let the plate go outside and attached the Peltier between the copper plate and an external heatsink

The sensor went subzeri

Well I don't know how much energy that takes of course

Could be an upgrade pack for Beta - so long as you design it in such a way that it is possible to add later

I have the feeling that Peltier cooling will consume too much power on battery. That's just a feeling :-)

even if zynq is well cooled and does not heat the sensor up and if the emphasis is on stability... and the battery is biggish... and there is a way to switch it off..

Another idea: record sensor temp alongside raw, apply different noise compensation depending on temp in post.

Sure, shall we start a new page?

Updated: http://wiki.apertus.org/index.php?title=Cooling and http://wiki.apertus.org/index.php?title=Noise_Reduction

if there is a temperature difference, yes, i.e. it is the reverse operation

Panasonic PGS looks very interesting

max 9A @ 15V :-/

but it doesn't mean that you have to use it at the highest current

that is 15 Watt, multiply it with the efficiency and you get the potential heat transport

Need to compare that with how much CMV12000 and/or Zynq is drawing, right?

Actually CMV12000 spreadsheet lists power consumption of only 2W

sexy?

you find my back sexy? :)

any native speakers here?

like the last update

you can have a look please

great

I'm doing a draft post for that now

wunderful!

how do I comment/chat in googledocs?

it looks odd having the prices to three decimal places

the campaign looks like it might slowly but steady take off.

very likely yes

as we can already output 108ßp60 on the Alpha

aombk: the cmv2000/4000 can be considered a crappy and small sensor :)

it wont get much cheaper than that!

the only thing that might be confusing about the new release is the first image is a bit ambiguous about the frame rates

the cameras capability at launch will probably be like 1FPS

then it would be worth saying that!

if you want to get highest framerate when the device is shipped, just let us know and we'll defer shipping till we 'finished' the software

Bertl: I'm just concerned there might be some vocal people who complain that the camera that just arrived in the post can only do 1fps

I'm pretty sure we will get those

but as I said, communication is key, we will ask folks and those who do not want the hardware early, will get the hardware when the software is 'useable'

have 'earliest adopter' group that know in advance the limitations before

I will change the first image

of course 'usable' will mean many things to many people :P

it is not that we fund the camera, we are funding the development, so all the numbers are basically guestimates

se6astian: the Super35 specs also list 4096x3072 which won't be an output option at launch

explaining that takes a lot of text/space

that space we do not have in the image

hmm

to me it is clear that the FPS is sensor max

why would the max Frame Rate refer to the camera when everything else refers to the sensor

IMHO it is perfectly fine, if you really want to emphasize on it

then it should be clear for everybody that the values refer to the sensor

added a sentence above the specs

there will always be people interested in the project who are wildly optimistic and only briefly glance at the details

perfect

are the monochrome sensors the same as the normal versions but with the RGB colour filters removed?

can users modify the colour sensors to make them monochrome?

i dont know for sure but i believe this will not be possible

aombk: ah ok :)

I'd be terrified of performing surgery on a £1,000 sensor though

no chance, those are micro filters directly attached to the microlenses

i am never terrified about doing these kind of things

will there be an integrated infra-red filter on the cameras?

that would be a nice option to have

olpf?

Bertl: cool :) I'd love to do some infra-red photography or video

oh,i see now olpf and ir filters are no that expensive

please reconsider OLPF. Image really looks amateuristic without, or at least think about making it swappable, like in some Red cameras.

this company does addons for DSLR users: http://store.mosaicengineering.com/

anton3, not really

there are pros and cons for the OLPF, so we won't force it on the end user

I tried some blur on the test footage, but the girl's cheek still has some awkward green and red dots in the highlight. You want people to look natural right? image here: http://www.blog.kurtaugustyns.com/wp-content/uploads/2014/09/AXIOM-Alpha-Sample-Footage-Selection-Ungraded-0014007b.jpg

also the color moiree in the test footage at the horse race. quite annoying to me.

and it's not just a case of finding a OLPF that's the correct size

it needs to be carefully matched to the size/density of the sensor photosites

anyway i personally consider olpf a small detail

if we have enough money to test filters, we will probably do it at some point

I'm off for a nap now ... bbl

do you people know anything similar to adobe onlocation?

new crowdfunding update posted

will be interesting to see what the night will bring...

Really love the addition of a 16mm sensor. I was just about to ask about an option for the Truesense chip found in the Ikonoskop and Digital Bolex. I love some depth of field. Is the CMV2000 the same family as the CMV12000. It would be nice to have matching colors for people using both 16mm chip and super35. Very nice update!

ikonoskop/DB use a CCD sensor

plus very limited with FPS

some people still claim CCD is superior to CMOS - just like some people say that about analog film (where I tend to agree more)

And some years have passed too.

yes there are stats

so I will be able to tell tomorrow

well every donation counts no matter how small it is!

nevermind.. webpage says 5 months. dunno how I missed that

soon

time for bed

good night

goodnight!

AndroUser2: what do you mean?

I see a page that is almost empty

Thx