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 - Luther

Pages: 1 ... 11 12 [13]
Share Your Videos / Re: 3.5K Raw Testing - 5D III
« on: April 21, 2019, 04:21:21 PM »
Very nice dynamic range!

but this one beats lanczos3,bicubic,bell,triangle or mitchell when it comes to interpolating 75degree staircase artifacts.

I think ffmpeg has Spline64 on it. Might be easier to implement and has a good quality...

Hardware and Accessories / Re: EOS M lens suggestions
« on: April 21, 2019, 03:44:30 PM »
The helios + the crop will become a 350mm lens, which is impossible to shoot handheld, not to mentikn the extra glass makes the setup bulky and ineffective.

See the Viltrox speedbooster (works on EOS M).

Share Your Videos / Re: EOS M RAW (Super 8 film Look)
« on: April 17, 2019, 05:44:46 AM »
Nice! I liked you use 15fps.

Very very good film look simulation! Congrats. Only aspect that I would consider susceptible of improvement is the white balance which is maybe too white at times... a little bit of magenta or yellow cast could be used in some shots

Agreed. How are you converting the MLV? I think you might get a better result using MLVApp, configured to output LogC. Then convert with FilmConvert with the profile called "Arri / Alexa / LogCfilm" (note: not the normal LogC. This one uses a "film matrix" to convert the colors). The company Cinegrain also offers some very neat conversions and textures.
Check these curves:

vid.stab seems to work very nicely @masc ! I was watching the demos and... damn.

For the past weeks I've intensively gone through all threads regarding dual iso and ettr and I still haven't been able to know exactly a step-by step approach in how to use both for video in camera and when to use or not to use the histogram for this purpose.
I've understood all parts, but still how do they work together 'in practice' (on camera) is something that I have been struggling with. Anyone mind sharing through an example with concrete values?  :)

Here's how I use it:
- Enable Raw Histogram (don't enable yet DualISO)
- Set your FPS and Shutter, don't change after that
- Use ISO and Aperture to bring the exposure to the maximum, without clipping.*
- Look your histogram, the right side should not be crushed
- Enable DualISO. Set the recovery ISO to +2 EV.
- That's it.

After you process the file, it will be overexposed, so you should bring it back to normal exposure (also using histogram to guide you).

* Whenever possible and appropriate, I personally set the aperture to the "sweet spot" (see the MTF chart for your lens, it's normally between f/2.8 and f/5.6). Then I set the lowest ISO that crushes the right of the histogram just a little bit and, finally, bring back with a Variable ND filter. This way you have the best precision and highest SNR and DR.

Those prices are way above my budget.  :D

Indeed. Scary prices.

Say I want to send a film to a film festival that demands a DCP or Apple prores file. Am I better off buying DCI-p3 or rec2020 monitor if I am color grading in 14-bit cinemaDNG and in which working space should I set my NLE in AE or davinci, P3? rec709 or rec2020?

I'm not an expert, so don't quote me on anything I'm writing... that said, it depends on where you'll screen it. If you are going to be compliant to the UHD standard, your master format should be Rec.2020. I would recommend you do that, as most of the streaming services now support UHD standard and if you need to convert to Rec.709 8-bit it's very easy (the other way around is not).
Buy a monitor that has a big P3 coverage. As I said, as far as I know, there's no monitor today that has full Rec.2020 coverage and P3 is not so far from it.
About the 'working space', you should set ACES and then set the ODT to what you want (Rec.2020). Netflix has a tutorial here and more color grading resources here. I suggest you use ACEScct, as it simulates log.

ps: if you have $900 bucks, this one seems to fit what you need -

But with natural outdoor light, is there really a difference visible

Maybe not on Canon cameras, because it doesn't have precise speed. But I for one can tell the difference in motion blur when I see other value than 180 degree. I think it's more of a convention and people have adapted to the "cinematic feel" of 24fps/180degree. Not because it is intrinsically "cinematic", but because industry always used this settings. When shutter speed changes people notice something is different (see Salience bias and Latent inhibition).

If I try on 600D can't set shutter to 1/45 only 1/42 or 1/50, but by using 1/42 and in fine tuning -3.22 i got 1/47,908
and with 1/50 and Fine tuning +0,85 i got 1/47,995

Just for reporting my results

More precise version (got these timer numbers using g3gg0's EOSTimerGen):
- Enable FPS override
- Select 24.000, optimized for "Exact FPS"
- Change FPS timer A to 600 and FPS timer B to 2000
- Fine-tuning to +0.17ms

This gives 1/47.999 shutter speed on 600D.

What display is the best to color grade 14-bit uncompressed cinemaDNG, DCI-p3 or rec2020.
Is DCI-p3 enough to display all the bits or should I choose rec2020?

The best ones are probably Dolby PRM-4220 or Sony BVM-HX310. They are not affordable, though. I heard BenQ does decent monitors with affordable prices.
As far as I know, no monitor today has 100% coverage of Rec.2020.

If you're looking for precision, make sure you calibrate your monitor. See ColorMunki or similar devices. Also, most colorists recommend you use a wall painting of 18% gray and room artificial lighting using 5500K lamp with high CRI (>90 - see this link). This is to avoid some vision adaptations that can "distort" the perception (read Purkinje effect and Kruithof curve).

Thanks a lot. I will see if I can find a way to convert that to a matrix.

@Ilia3101 I'm not sure what I'm doing, but might be something: got a ICC profile from Rawtherapee and then, using the tool iccgamut from Argyllcms converted to gamut values. Result here:

edit: the "code" tag doesn't work because the file is too big, so I had to upload it.

ps: If you contact Beep6581 on github he could have the correct matrix.

I would add the AP1 gamut too if it weren't so difficult to find the matrix online. Maybe you know where it is ;)

Not sure if this will help (I'm not a coder), but I've found this CTL that can transform from AP0 to ACEScct. About CTL here.
Also, these are the primaries (it seems... found on github and on the document S-2016-001)

Code: [Select]
const Chromaticities AP1 = // Working space and rendering primaries for ACES 1.0
  { 0.713,    0.293},
  { 0.165,    0.830},
  { 0.128,    0.044},
  { 0.32168,  0.33767}

And the white point:

Code: [Select]
  { 0.32168,  0.33767}

You may be pleased in a few days ;) ;) ;) ;)

 :o Can't wait for it.

Code: [Select]
    { /* GAMUT_Rec2020 */
         1.72466, -0.36222, -0.25442,
        -0.66941,  1.62275,  0.01240,
         0.01826, -0.04444,  0.94329
    { /* GAMUT_ACES_AP0 */
        1.0498110175, 0.0000000000, -0.0000974845,
        -0.4959030231, 1.3733130458, 0.0982400361,
        0.0000000000, 0.0000000000, 0.9912520182

Hope I'm not spoiling the surprise. But, according to Walter Arrighetti, AP0 is mostly used for archival of footage, while AP1 is the one used for color grading (normally ACEScct).
Looking foward to version 1.7!

General Chat / Re: Guessing 1st April for 2019 - Share your thoughts
« on: April 02, 2019, 10:53:01 AM »
I've been using ML for years now. This forum and ML is what made me go to university study photography. Many of you have helped me through those years, in special a1ex and g3gg0, but also many other devs and users.
@a1ex, just wanted to say ML isn't for nothing. I truly think you have made a revolution with your development on ML. You gave people a tool that can record in a quality that once was only available in high-priced cameras. You gave people without money the possibility to compete in quality with high-end cinema, the possibility to express more freely. This isn't "nothing". I'm grateful to have this forum and ML.

As for the money, I agree with @Ilia3101 , a Patreon page could make it easier for people to donate.
Other idea I just had was to make a short film, just like Blender Foundation does. Could be a community effort and all the money gained through it can go to ML development.

I will discuss this topic in the forum in the right time and will let you know all the details, if someone want to help it's will be cool too.

About MLVApp, maybe a government funding? Natron got funded by INRIA and it seems to have worked well for them. I don't think this is an option for ML (because of copyright stuff), but MLVApp would have no problem trying.

Raw Video / Re: Colourful noise on raw dng files
« on: March 28, 2019, 10:45:40 PM »
Try MLVApp. If it stills there, see the FPN ("Pattern Noise") and Veritical Stripes options.
If it solves the problem, then the issue is with the software you're using. Else, if it does not solve the problem, try a clean install of ML, no modules loaded other than MLV, and record at low iso (100). If you still see the problem, it might be damage, but only devs can say that...

Very good.

Pretty new to sound recording/editing so there might be a few things here and there...

I'm new too to sound, but over these last few months I've been learning. I highly recommend the software from iZotope. In special RX7 and Nectar 3.

I want to see this kind of microcontrast for the surreal image, I like it :D

I see. An artistic choice, then.

The most shoots, much more than 1/96. I applied Pixel Motion blur in after effects at 90 degree, on almost all shoot.

I knew something was different. Have you tried Twixtor? It does a pretty good job for me.
Anyway, nice work. Hoping to see more from you.

Cool! The sky replacement is really good. You even applied a lens blur in it.
Two faulty points for me: there's too much microcontrast (like when use the option "clarity" in ACR) and I have the perception the slowmo have a "wrong" motion blur (your shutter speed was more than 1/96?).

Blowing up 4K when i export. (Becouse the youtube destroy all of the details, but in 4k is acceptable)

Not according to this:

Hm... none of them is usable in MLVApp without reinventing the world.


The processing would take hours without CUDA support. Yeah, not really feasible. But neural networks open a new world of possibilities for image processing. First time I read this stuff I couldn't believe it was true. Down below there's some impressive debluring algorithms [1][2][3][4], moire reduction [5] and super-resolution [6][7].

If someone here is looking to upgrade the demosaicing in MLVApp, I think it would be best to look what the high-end industry is using right now.
I'm not an expert, but most people recording with Panavision DXL2/mini or ALEXA LF seem to use Colorfront Transkoder or Codex Production Suite. ARRI recommends their own algorithm, called "ADA-5", that can be tested using their freeware software (sample footage from ALEXA here). This ADA-5 is now a standard called "RDD 31:2014" (couldn't access the paper myself).


General Chat / Re: Guessing 1st April for 2019 - Share your thoughts
« on: March 25, 2019, 11:26:13 PM »
4- Better adtg_gui which can work beside other modules like crop_rec.

ADTG as default ISO and gets integrated with Dual_ISO and HDR. That would be nice.
Most likely to happen:

6- None all of these, please try again next year or go to do these f***ing things yourself.

Share Your Videos / Re: 5D MKIII + MLRAW in Iceland
« on: March 25, 2019, 11:22:19 PM »
Very nice! You even got the aurora. I like the film grain you applied too.
Iceland is beautiful. The music on your video is good, but I think Sigur Ros or Ólafur Arnalds would be more appropriate :P

Right now you can simulate something very close to this effect: set contrast to 100 and clarity to -100. This should give what you are showing - just with boxblur instead of gaussian blur. ;)

Tested it. Works very closely to what I was describing. Thanks!

Luther I am a proud ML Canon 5D Mark II User, I use MLVFS to convert MLV files to Temp DNG, Use AfterEffect to edit DNG in Adobe Camera Raw, then AfterEffect Render DNG to ProRes 4444XQ. I would like to know what do you think is the best log profile to use and What is your Workflow.

@dariSSight I use Log-C from MLVApp. My workflow using MLV is: always get the best image first on camera (I normally ETTR to +1.5 stop) > Import on MLVApp > Exposure Correction (watch the histogram!) > WB Correction (using a 18% gray card you can just use the picker) > Change to Log-C > Export to ProRes 444.
Import on Premiere. Color grading using Lumetri curves and Filmconvert or Logarist (to simply transform to Rec.709). Denoise with NeatVideo and Sharpen with Samurai (from Digital Anarchy).

This is not the best workflow, though. For big productions the best would be to convert MLV to CinemaDNG, then use Resolve configure to use ACEScct.
If MLVApp adds ACES support and output to 10-bit Rec.2020 ProRes, I would easily just swith to do everything directly on MLVApp and use Premiere just to finish the video.

Did some hdr blending(48fps HDR on my eosm). Hdr iso settings can be found under Crop mode sub menu since HDR mode is broken on this model otherwise when it comes to raw recording. Export render time is about the same as any other file thanks to averaging filter instead of enfusing, aligning etc.

That's very interesting @Danne. Would be possible to blend the raw data and export to MLV? So you could import and have the full preview. Probably not, right, because the merge uses ffmpeg tblend?
Another question: the averaging filter will work if you use a slower shutter speed (more than 1/96, for 48fps) or will generate ghosts?

Thanks for all the efforts in this project. MLVApp is amazing. I use it in every production to convert MLV to ProRes in Log-C.

A higher color space is the only thing holding me from using MLVApp exclusively to process MLV files.
It would be great if you actually achieve that. Is there any specific problem you can't solve?
I started reading some stuff and found this camera spectral sensitivity for 5DMKII. Maybe this could be used to create an IDT? It would not be precise for other cameras than 5DMKII, but could be just precise enough. The ODT's can be found here.

I was thinking something like this:

High bit-depth output is important too, as on some companies (like Netflix) it's mandatory now to send HDR (meaning 10-bit Rec.2020).

Do you think it would be possible to implement a filter like Black ProMist? The idea is something like: gaussian blur on shadows and then change the blending mode, where gaussian intensity controls the amount of diffusion and opacity controls the overall intensity. Of course, it wouldn't be the same as the filter itself, but it would be close enough to "hide" crispy edges that digital cameras produces. I simulate that on photoshop this way:

Pages: 1 ... 11 12 [13]