Messages

[like this]

So... did you figure out how to program the image processing chip?

Good a1ex if you do not want comments n00bs should close the forum, because I believe that's what it is for. Otherwise, transfer all of the ML codes for a non index page on onion and live in peace, just posting the change log and nightly.

I'm sorry, I know how to be stressful to manage such a large project like this, I know you must be bursting with so many useless questions like this, but please keep temperance.

I repeat, I'am not a developer, I'am a user with questions and with motivation to help with this.

Guys, it is clear that I am not talking about complex processing RAW matter!
The processes that speak of NR are simple processes applied to JPEG, for example, or H.264 (what I think to be more difficult in real time). I'm not saying to create a "non-local algorithm" within the scripts ML! No, I'm not troll guys. I'm talking about simple NR, the same by canon but improve, just it.

So it must be true.

Irony detected. 
I'm sorry, I have no expertise.

Not even a1ex, I'm trying to get somewhere.
If you could interpret, higher clocks can process more complex code. If the noise reduction given by Canon may change soon, with higher processing powers that can be improved.
DIGIC processing noise reduction, no? Wiki says that.

I'm not trying to troll anything, just acquire  this knowledge about ML. 

Also, do not think so unlikely that it can be improved when says about the AF.

edit: I'm not looking for any label or status here, I'm looking for quality. You as the owner of an open source project should know better identify identification true trolls.

[Auto focus]

Could overclock be benefited with this hardware? I understand it's a bit dangerous, considering that there users of countries with extremely high temperatures, which could overheat the system, but perhaps this speed benefit functionality somehow?
I do not see how, but I'm just throwing things in the air ...

Maybe there was a gain on features that are automated, such as Auto focus and auto exposure? And the noise reduction on jpeg and h.264?


See on wiki:

DIGIC 4

Canon claims improvements such as:

    Much faster image processing when compared to previous processors
    Improved noise reduction in high-ISO images
    Improved performance while handling larger 14-bit RAW images
    Live Face Detection AF during Live View
    H.264 1080p encoding

Hi everybody!
I just thinking other day: in some part of ML codes the DIGIC is overclocked, or somewhat, or not? I don't know how this will help, but maybe the encode of h.264 more fast, more bitrate can be pushed?

Psycheodelic! I like it, very good.
The landscapes shots are just so amazing! The b&w [yosemite?] are fabulous. 

Go to "syncronize" option, on upper left.  HERE.

Print your settings on Cineform, please. The camera not produce interlaced, so, it is a software configuration and not your hardware.
Maybe your sequence on fcp is in interlaced? Try to change it, if is. Maybe your config. on cineform produce this...

[HERE]

HERE

Great swordinhand!

[bennyray1]

bennyray1 and D6-2035. Thank you so much. You do not know how much your comments motivate me. Thanks a lot guys!

[technique]

After seeing this video, I am going to have to rethink my attitude towards artifacts!

Ah, yes, the art of spoiling our youtube videos. I intentionally threw a dirty grain of Cinegrain, but it seems that the motion estimation youtube does not like to keep them! In vimeo only give to upload at 720p (I'am not PRO), and I did not miss it.

But the content is way too "family" for the style.

Yep, I tried not to do anything of the viral type, this is exactly what you said: New Video family. I had some recordings here and did not want to lose them in HD, so I decided to post it here and here.

I'd be interested in hearing some details about your software and workflow.

So, I used AE and PP CC. I did slow with Twixtor PRO 6 without any vector, so some bizarre interpolation artifacts on water. As the shots were 720p with 60fps and I wanted to make a 1080p video, I interpoled them with VideoEnhancer.
Before working on anything, I transcode to Cineform, incidentally.
After that I did the composition in Premiere, played the audio in audition CC (only found in 44.1hz and 320kbps music, so it is a drug), and then I colored the video with After Effects using OSIRIS from VisionColor (incidentally, used the Visiontech 2 for shots), also used the FilmConvert PRO, with appropriate profiles, and also the Colorista II, RedGiant, for push some tones... is that. I render to Cineform 444, and then did the encode for web with MeGUI (x264 10bit, CRF 12 and --fgo 15 ; audio with qaac on HE-AAC).




Thanks for the comment and especially critical. Not only live praise and believe that we will grow much criticism.
Just to portray me, I made this video without much technical involvement. I believe there comes a certain time the technique takes care of us, and it hides all the essence for which we are working, it is making something beautiful and sensitive. I do not know if I got it, probably not, but I tried briefly.

[What your thoughts on this?]

This modification is really necessary? I mean, people have come to conclusion that ISO's greater than 1600 are pushed digitally by the camera, and a1ex has said he prefers to push this information in post-production than letting the camera do.
What your thoughts on this? I think we could save time about it and turn in the study of ADTG, which seems the most promising. It is just a warning in the official announcement... 

Summer video, for fun guys! 

Yes, comfirmed for my camera.
ISO 6400 different locations
ISO 1600 same locations.

I didn´t try ISO 3200 yesterday.

Edgar

Maybe because they are not real ISO's, but pushed digitally?

[Astra Image]

Astra Image

Maybe your compilation of dcRAW is 64bit and you run o OS 32bit? Anyway, try VM with Win8 and you will get the response...

aldy, several factors may have influenced so that this problem, is hard to say exactly what it is.
First you are using an outdated OS. Have you tested directly at the command line without this GUI? What is the file you are trying? Perhaps a conversion to dng?
Try something like this, if you do not give, upload the file you're trying to modify, and I maybe try to analyze and see if it's something physical on file, if not, maybe is your OS... try VirtualMachine with Win8.

Congrats escho, great.
Maybe an implementation of these code in MLV2DNG? Just one un-official mod, for now?

[Here]

Here

[not turn off the camera after recording a sequence]

COCO, which its formatting? Directly from the camera?
You may have some scruffy block ... try to use a tool for low-level formatting, and then use the SDFormatter. If not, try the Paragon alignment.
Probably the files were corrupted ... the most you'll be able to do is extract some DNG's... an important tip is to not turn off the camera after recording a sequence, because it may still be processing what will corrupt the entire sequence. The ideial is to do nothing, and wait maybe 5 seconds only.
ps: try exFAT, too.

[samples]

FPN samples. Simple silent pic from 600D. Appear more pattern on deep shadows. On crop mode, this not appear [or less].

Are you sure a1ex? I tested at least 4 footage with the 600D and the first coordinate badpix appear... 

edit: see the first coord. on this archives https://app.box.com/s/t1voim3ew086ko0aze81

0k.
On here:

Code Select Expand

X: 47,26
Y: 3,26
And [maybe] here:

Code Select Expand

X: 2,70
Y: 0,37

There's bad pix not removed.

I not test very much, but it's very effective! But, I think this is slow yet... on my garbage HD, this run approx. 89% more slow...

I think I've compiled the wrong branch, can you download it again?

Work very good now.
Any idea for calculate the line and column of the bad pix? There's some badpix than escho modifications not interpolate... I can indicate it on some file, but I think better indicate the line/column.