Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - 2blackbar

Pages: [1] 2 3 ... 20
where in the source i can change quality of jpeg images exports to be higher? i can see some jpg2k code but cant find quality values

It is possible to use sinus in transfer function and create non linear curve but whats the exact formula for non linear film gamma ?

aaah... 10 years with magic lantern and im still a noob  :P

Danne i asked previously about it (fps override) but i think that in older builds from 2019-2020 i was able to use 16fps with shutter 32 in crop modes like 1:1 or 2.8k, i cant do it now, i can use 16fps and shutter 32 with uncropped 1080 tho.
Is it possible somehow to unlock lower shutter values ? Now lowest shutter in 1:1 crop 1920x1280 is 1/48, when i go lower then its like 4000 again.
Im using it for 8mm film emulation , 16fps and 180degree shutter which is about 32.
fps override code is a separate code you didnt worked on ? Maybe i could bring more shutter values in 1:1 or 2.8k crop modes on my own but i have no clue where to look.

I find it very useful that you opened transfer function for processing and i could create more looks more freely using luts that just werent working with previous locked algorithms.
Is saturation algorithm able to be opened like that ?
What i really want to do is to make it behave identical to HUE vs saturation or luminance vs saturation algos because their algorithms are much cleaner, i find saturation slider below vibrance to be very noisy when moving up, introducing a lot of artifacts while adding saturation especially in reds, while hue vs saturation have almost none even when maxed out.
Or can i somehow swap algorithms in the sourcecode ?
Sometimes i cant use luma vs sat or hue vs sat because theyre already taken and used to correct for something else.
here are maxed out Hue vs sat, saturation slider (one below vibrance), original image.
saturation slider artifacts appear very soon when i bump it up.

thanks! This is responsible for grain size ?
Code: [Select]
double weight = sumL / 1.5 / 65535.0;I see that in 1080 or 2800 pixels fooage grain size is different, relative to resolution, so id like to change the size for 2.8k

Markus is there some non complex way for me to change in the sourcecode when lut is applied in mlvapp pipeline ? I want it to be absolute last thing applied to footage.
I want it to be after chroma smoothing, after stabilisation and after adding toning, white balance change etc.
Most luts react differently when i give the footage blue tint and i use "toning" panel to do that but its affecting currently afterlut footage not pre-lut.
Or am i wrong and it is at the end ?

Do You guys know what kind of tonemapping its used here?
It looks so much better and brings up a lot of detail , from what i see at the end hes using 3 sliders, one for hightlights, one for mids and one for shadows to bring them back.
Its kinda like luma adjustment overlay for each of 3 sections thats blended on top of the footage.Luminance ciurve in mlvapp gives similar result but theres something more to it in their method.
would something like this be remotely possible in mlvapp ?

if youre on windows, compiling mlvapp isnt that hard , you open it in qtcreator

If i would want to swap "set" button funcionality to make it go into 1:1 1920x1280 mode  with good preview would simply swapping names around in ML source work ?  i just want that mode to be accessed quicker.
Or is there some other way?

I would not recommend 5DII raw if youre using mlvapp and you dont know color science cause it has some issues in there , all colors but green are washed out undersaturated and overall you need to work with it quite a bit to make it look close to real life colors.DNG files work fine in other programs tho but with mlv people mostly use mlvapp.
Go for 5DIII, maybe it doesnt have such issues.

i have that one, didnt testes yet like that tho

i also  add sharpness to 64 and upscale 1920 to 2500 then go h265, would  that make a difference?

i monitorr using phone and hdmi usb grabber but battery drains very fast , this ssetup is possible but its just better to use monitor with own battery.
unless you have dc adapter close.
youd need 120hz phone tonavoid heqdache in vr

Is it normal that it takes twice as long to exportvideo with grain enabled - 45mins vs wih disabled - 22 minutes?

you have a lot to catxh up to, read and watch zeek vids , easiest route so you wont be asking questions
- m has a lot of rcording modes., highest true res in 24fps is 2800*somethindont remember., in cinemascope crop ratio,about super 16mm frame size not aps-c.
- 5k is slightly better detail than normal 1768* somethin hd mode but isnt real 5k.there are modes with low fps and higher res but 2.8k is highest true res with 24fps
- 3k will drop frames in 24fps, camera card reader is not fast enough to catch up with raw stream.5k and 16 fps is probably actual 5k but higher than 16 wont record without dropping frames.
- m is a true  2.8 k raw camera.other higher res modes upscale in post unless you shoot with low fps.

yup Danne is responsible for M price bump  :P , its because of great magic lantern support for it, people are discovering it does almost 3k raw in crop.
I remember buying M for around 150 euro with kit lens in 2014 and i sold kit lens, i think like 2 years ago ive seen m going for 100 as well but now they want as much as for 5DII.

Camera-specific Development / Re: Canon EOS R / RP
« on: February 07, 2021, 04:32:58 PM »

crop_rec and derived builds / Re: Danne's crop_rec_4k experiments for EOS M
« on: February 06, 2021, 12:51:24 AM »
just get all folders besides dcim and throw thwm to new folder on card, then move autoexec and studf from ml zip file to card root.
you will regret not having version that wprks best once you collect about 10 builds , you will lose track, so i name the folder for eg working hdmi ,let it stay on card its a small size,  i juggled many build whwn testing my own builds and danne's , so naming folders andnkeeping on sd card helped

crop_rec and derived builds / Re: Danne's crop_rec_4k experiments for EOS M
« on: February 04, 2021, 04:19:36 PM »
Great to see new modes introduced by Danne in this camera , downloading !

Ah i knew somethng was off, do you know where i can look up the hue adjustments to apply in mpvapp?
are there more cams affected by this besides 5d2?

crop_rec and derived builds / Re: Danne's crop_rec_4k experiments for EOS M
« on: February 02, 2021, 12:31:08 PM »
 I can tinker when i get enviroment installed again, i hope not much was changed in how it was done, im currently investigating if higher or lower rolling shutter values give me that motion blur im after, cause i see that exact fps actually has bigger rolling shutter value than other fps override presets and  ill try if i can get the motion blur in 2.8k to look similar to non crop.
 In 1080mcm With 17.8(lowest rolling shutter settingi can get ) and 1/50 i get typical motion blur as if not using fps override, with 18.8 which has exact fps setting i get i think less motion blur but i can get shutter 1/45 to make it less stuttering than 1/50 while not looking like internaced cause of too much blur, and i got 25.5 rolling shutter to try it out , that gave me even less blur but i have to test that out by filming moving subjects.
 Its boring stuff for someone who doesnt care but im trying to find a way to get rid of micro jitters while maintaing nice motion blur, dont care if it says shutter 1/40, if it looks good for me then its good.
Thats why i filmed wits exact fps and 1/41 shutter a lot, i could go handheld and not have micro jitters like on 1/50 because motion blur actually looks different.
I know that progressive 24fps footage stutters with 1/50 shutter but not as much as DSLR's and theres a bit of info on it how rolling shutter affects that.But here i actually use higher rolling shutter values to drown micro stutter in motion blur cause that blur looks different, withhout fps override and shutter 1/45 footage already starts to look like interlaced with too much motion blur.

Basicaly , motion blur without fps override and 1/50 looks more like its current frame and previous frame but the moving part (thats supposed to be smudged out) is just 2 frames in one.It looks to stuttery for me, with fps override exact fps and 1/45 or 41 shutter i get better non stuttery motion blur.
That regular motion blur without any fps override turned on looks Kinda like this :
So im looking really for blur characteristics not lower or higher rolling shutter values just to get best math results, im just sharing the info in case i forget later.

Raw Video Postprocessing / Re: UGLY clipping samples
« on: February 02, 2021, 12:19:02 AM »
Yes i had that especially at night when filming traffic lights or cop lights, i dont have dngs now tho.

It was but Id like im using tonemapped  and camera matrix most of the time, colors dont look that good without it. But i tried it , changing different methods like srgb and not using camera matrix and it really doesnt fix the colours whatever i do in mlvapp, i have to use HSL curves to bring back reds and squash the orange.
Theres just no red in that 5D2 image from mlvapp, its like orange and its strange.IT affects the sking and its the reason i dont film with 5D2 that much.
I brought it up cause maybe there is something thats causing this difference especially if in lightrooom both look the same.

crop_rec and derived builds / Re: Danne's crop_rec_4k experiments for EOS M
« on: February 01, 2021, 11:51:03 PM »
I dont see these regs affecting rolling shutter value, but i see that rolling shutter lower than 23ms in crop modes will get me no preview/glitchy image.I have it in fps override menu at A ft-106 and B FT+5) but anything lower than 23 and theres no image, i dont know if it was like that long time ago but i remember using fps override in crop modes with success but on other hand iwanst able to record for so long so maybe its a trade off.

Pages: [1] 2 3 ... 20