My evolution in video/photo (assist, idea, other)

02/24 edit:
Add new night shot picture

02/11 edit:
Reorganisation post
1 overall, important information
2 material, Software, other information
3 Picture and video actual progress
4 Light archive of 3 previous post

Add video for comment

_

Hi everyone :slight_smile:
I started a few days ago to focus on the realization of photo/video with objective to make , among other things, light rewiew.
I go little by little having everything to learn in the job.

Few word that I could develop later.
For video job:
-Once I solve the problem of focus which does a little what it wants, I am with 2 remaining problem.
The shaking of the rather unsightly video that I tried to remove by software, result inconclusive, I think I should invest in a stabilizer.

  • After make a stable video I have for project enhance video by artificial intelligence, good result just need massive “power” processing/computing power AKA Graphic card to resume :frowning:
    I’ve actually
    Ryzen 5 3600
    GTX 1660 Ti
    16 GO RAM
    I’m proudly has an average of 6 seconde/frame … for a first draft who goes into training another (without touching the colorimetry of the video :stuck_out_tongue: ) all in a fairly high noise.
    So it’s impossible and i’m in reflexion for the utility to invest in new configuration to give me the means to do something nice.

I am focused on moving video, static video is less problematic.

For the picture :person_facepalming:
AĂŻeAĂŻeAĂŻeouilleouille as have said on my country
A lot of information, setting to learn all at once and more than a camera with a really no translation that does not help me in setting and a MASSIVE issu with software when I can make a good photo.
The setting of software are most difficult compare to the camera :smiley:

Lot of work in perspective :innocent:

-

If you have any suggestion advice and other I take
Or just give your opinion on the general project it also suits me.
I take all.

Camera
Panasonic Lumix FZ-82

Software
Topaz labs AI ( video, Gigapixel, Denoise, Sharpen)
GIMP2
Darktable
VSDC Free editor (Pay version)

PC fo do that
I’ve actually
Ryzen 5 3600
GTX 1660 Ti
16 GO RAM.

I’m thing about pay for Adobe Premiere Pro seem be good to correct little issu on video (luminosity, colors…), but nor for the moment because of my PC level restriction, no interest to buy 20€/month in this condition

Picture advancement

02/24
New picture shot, i’ve work better my camera preset I’m thinks its a good step up result :slight_smile:
I’ve found the base issu for colorimetry change, AI software use in default special RGB output, so i check the various RGB output exit and find a nice for good result
No tripods but quality is nice
F: 2.8 SS: 10 ISO 800

Video
Two video avalaible for the moment focused on ES3

Isee two major issu on this video
-The shacking camera

  • Noise of my respiration, you can hear it sometimes like I’m been holding fart :weary: I try especially not to breathe too hard to reduce the noise to the maximum, hence this impression

For solve this I need to buy stabilizer, If you have suggestion
For sound I think I need buy additional microphone to connect to the extension port of the camera to capture the sound more in front of the camera and not behind, again if you have suggestion.

_

Second video, static video.

Niwalker ES3 Beamshot compare to various lights (See description, timeline, information). - YouTube
I’’m too fast in rotation, need be more slow.
400 iso, potentially too dark, for me no, depend our screen.
Not perfect pattern.

if you have anything else to say about it I take :slight_smile:

If you have any comment or suggestion you can :slight_smile:

Archive 02/01

Ok, after cry a lot I think i’m able to correct the auto correction of Darktable (But I still do not understand and especially how to remove the auto correction, finaly I had to change many color parameters and not only the white balance)
Look this interface, thanks the 400 page PDF for basique explain of this software

Result for tonight

compare to JPG

Mitigate but for first try it’s ok
We can see i’ve only focus on the blue bezel (to make real blue JPG is little exaggerated on his color), other color are very decrease :frowning: especially the red on the ES3 top head.
For the quality/resolution/detail improve by AI it’s nice keep in mind that my basic photo is not perfect I can improve it in several points for better difference JPG/ final work.

You can see the difference with RAW+ AI work on this point

Compare to JPG

I have enough for today and probably tomorrow :stuck_out_tongue:

Archive:
Which brings me to a first problem I submit to you, if anyone has an idea.
I’ve buy Niwalker ES3 create a topic to speak about this light, and in addition to my first impression I also wanted to share some photo


Here is my problem, the image on the right is my final rendering in GIMP2 and the colorimetry has changed without me understanding why, without understanding how to put it back original.
The image at left is the original photo in JPEG record at the same time as the RAW in the camera.
But for work with pretty nice quality on software I need to use the RAW.
On the center you see the RAW pics after AI software job in Darktable software obligation to work RAW picture in GIMP2 software for the last work on pics and conversion on PNG/JPG to share.
Darktable is required to load a RAW file into GIMP, but for unknow reason they load picture in GIMP with this colorimetry.
I don’t see when on Darktable or GIMP the option take decision to change colorimetry O_0
Seem be white balance change on the GIMP picture white are white.
If anyone knows how to fix this problem I take.
Quick edit, I’ve identify issu, It’s Darktable who apply auto white balance, don’t know how remove this auto change but i can change manuel each pics

You say you need to use RAW, but then you load it into GIMP. GIMP can’t edit RAW files. So, if you like the jpg the camera produced better than darktable’s, why not just use that one?

Darktable is a high-end, extremely configurable raw editor/converter. It might be the most flexible, but it’s not the easiest raw converter. The darktable devs, and lots of users are active at https://discuss.pixls.us/ It’s auto settings are sometimes not as good as the camera, so for “developing” raw images, it usually requires manual intervention. Almost all the open source raw editors are like that.

GIMP can use RAW pics in addition with Darktable, they use Darktable to load the raw pics on his interface to use it with its own functionality.

I must have misspoken, I only say for unknow reason Darktable apply is personal setting on the base picture to give a result that is not realistic as show in my first picture(head bezel are not purple).
Originaly I use it only to load RAW in GIMP for make an other base work to AI again in upscale.
In the immediate I just wanted to produce not ugly photo and not start directly by learning all the art of colorimetry of a photo.
It’s antinomic with my work method step by step.
Colorimetry it’s after produce reasonable picture.

I don’t see where I say JPG are better.
Why GIMP, only because I have the basics of its operation and its in adequation with my step plan.
But I agree only Darktable could do the job for as long as I have hours to spare

No, you’re misunderstanding. Darktable develops (or converts) the raw image to something GIMP can edit. GIMP cannot edit raw images.

That’s because you aren’t doing what you think you are doing. You’re not just using darktable to load a raw image into GIMP. You’re using darktable to develop the raw image and convert it into something gimp can edit. That includes all the same steps the camera does to convert the raw to a jpeg: denoise, white balance, etc. except the actual jpg compression. So, once you have it in GIMP, it isn’t a raw image any more.

Since you don’t say specifically why you “must” use a raw, I don’t know what you’re trying to accomplish, but if you really have to edit a raw image, then you need to do that in darktable (or rawtherappee, or other raw converters) that actually use the raw image. GIMP doesn’t.

Your main question was why the white balance in the darktable-developed image wasn’t as good as the jpg from the camera.

Basic it’s always better to use RAW and compress after you do what you want with the picture
I know it’s no RAW on GIMP it’s clearly says on Darktable “export to GIMP after close” So if GIMP can load RAW they don’t use Darktable and if they use Darktable for load RAW after Darkatable export it’s not a RAW.

Once this point is cleared up, My software to enhance auto with pretty good result is Few Topaz labs software.
AI processing is largely better with RAW compare to JPG.
After this process i would be able to make few small modification on GIMP
I’ve try solution provide by Topaz AI software after AI for other format compression
I’ve try RAW convert by Adobe DNG
I’ve try the way provide by GIMP to load RAW.
Adobe I don’t like Adobe and the result not good on GIMP, less “easy”
Topaz compression after AI not good on GIMP
Darktable good result after export in GIMP.

So I must use RAW for upgrade picture quality and for the moment I would use it on GIMP.
Trying to keep a little compressed image devoid of metadata.

I’ve see the possibility with Darktable (I’ve also seen is capability to change blue on purple without having a simple way to remove this auto correction) for upgrade a RAW picture manually, interresting for the futur, not for the moment.

I’ve also say the purple color on the ES3 head is the result of Darktable, ES3 bezel on reality are blue as on the JPG and the RAW before and after AI job.
Sorry i can table say more essentially because I can’t figure out how to make sure that Darktable doesn’t apply its auto correction and after its auto correction it indicates all the parameters changed to 0/ in their base value, while they have been modified .

I hope I’ve made that clear. :slight_smile:

@manithree
Thank’s for Darktable forum link, I’m going to look at it more in depth and probably make a post for understand some detail about RAW.
Actually I don’t understand why base RAW picture are so distant to reality in color fidelity, I’m ok it’s for do our personal modification, I don’t understant why RAW a so distant to reality in color O_0

Edit 03/14
I receiv gimbal Saturday :blush:
It’s DJI Ronin-SC

I was counting on it as early as Saturday, but I had a little too much hope that it would be just plug and play product.
Its always looks simpler in video, so I’ve pass my Saturday to setting it and Sunday for understand the app.
I think can try it Tuesday night.
It’s quite heavy and rather cumbersome but the result is here, i’m very surprised by the app who can offer a lot of setting choice :slight_smile:
I have a major deception to the stabilisation by software, my video is bad, for me, with it I can do pretty nice walking night and other dynamic plan.

03/16 edit:
It’s as incredible as it is heavy and I had 1.5 kilos on my arm :heart_eyes:
I’m very satisfied of this first try, I check my vids some minor issu but in fact the gimball give perfect stability.

It’s random test i’ve film the moon a full zoom (x60) stability is pretty nice this is filming at arm’s length

Quick edit and post on my chanell today or tomorrow
I’ve make aprox 4 minute for each light dynamic move with
Convoy M21B GT FC-40
Astrolux MF01 mini 4000k
Astrolux MF01S 4000k
Emisar D4SV2 XPL-HI 5000k frosty optic
Sofirn SP36 Samsung LED
Niwalker ES3 Samsung LED
Nightwatch NSX4E SFS80 LED

Edit: playlist ready:

I use Canon RAW files, RAW mode by nature has no in body conversion of any kind. Therefore, yes RAW files are not nearly as colorful or realistic as most JPG outputs. That is all controlled in post processing. There are several adjustments in body that can and will make a difference in the RAW output. Most notable would be your color saturation settings and curves. Also be mindful of using the same color space through all steps of conversions. If I were you, I would work on determining the best results from the JPG conversion processor on you camera and not attempt so many time consuming and complex steps in post. Another option is Photoshop, it actually can use profiles for your selected body to assist in the final post conversion. It also uses lens profiles to correct for inherent lens flaws upon import. Most of my RAW conversion work has been done in Photoshop over the years, I am not an expert but worked on the side under one. Another needed step is to calibrate your output devices, this will allow images to look the “same” across screen, paper and optionally camera screen. My newest bodies are a 5D ii and 7D and I do use a Lumix G5 for light weight trips. I do edit with decent monitors and usually only for personal output on web pages and such. I did have a few years where I did some studio product stills and processed them for paper, much of this work was done in film and medium format :). Had to color match for paper changes, ink batches and such when working in early digital.

I do not shoot video, so I really cannot comment on that process at all.

Darktable and other open source raw processors use the lensfun library to do the same thing. I don’t know how it compares to Photoshop’s profiles (I haven’t used Photoshop since the 1990’s), but it seems quite adequate.

Hi everyone, I’m resurrecting this topic after making a few changes to my equipment.
I’ve now:
Camera
Canon EOS R7
Sigma 20mm F/1.4 DG HSM Art.

Software
Topaz labs AI ( video AI , Photo AI, Gigapixel, Denoise, Sharpen)
GIMP2
Darktable
VSDC Free editor (Pay version)
DaVinci resolve free

PC fo do that
Ryzen 9 5900X
RTX 3070 Ti
64 Go RAM

Thank you mattlward I was discovering the joys of photo work and I was using software (a bit badly) which has evolved (it or my equipment I don’t know which has corrected what).
What annoyed me was that it randomly changed 1 colour (blue) and even afterwards it was hard to find a blue colour without altering the others…
This problem is now behind me :slight_smile:

So I’m back on this topic to continue sharing my progress and the associated chain.
I was looking at several camera models and I came across a good opportunity on a Canon EOS R7 with its adapter ring for a Sigma 20mm F/1.4 DG HSM Art.
I had seen other full frame sensors, but they were relatively expensive and I couldn’t find anything at the price I liked, so I fell back on this APS-C sensor, which I thought would be a nice way of continuing to explore the world of photography and video (and not just at night).

For the nightshots that I share here.
I finally have a colour setting that I’m happy with for video and photos.
Well, depending on the screen, there are differences in tint, but I’m quite happy with the overall level that comes out on the different screens at my disposal. I manage to bring out the various tints in a relatively realistic way.

As of today, I have 5 videos with the correct tint settings and only the last 2 have the maximum image quality.
Here they are from the most recent to the oldest.

CRI compare

X75 compare

X75 compare

CRI compare

CRI compare

I didn’t have time to really work on the photos, as I already wanted to make videos of satisfactory quality.
On all my videos I put the JPEGs at the end, I’m thinking of a solution to share them in a nicer way than a series of photos posted one after the other.
Some picture of my last video.

Acebeam E70 1* Getian FC 40 4500K 95 CRI

Emisar D4SV2 4* Nichia 519A Dedome 4000K 95 CR

Niwalker ES3 3 x 3 Samsung LH351D 4000k 90 CRI

Astrolux MF01 Mini 7* Luminus SST 20 4000K 95 CRI

Sofirn SP36S 4* Samsung LH351D 5000k 90 CRI

Example for Nichia
Original JPEG

Upscale/denoise/sharpen

Upscale/denoise/sharpen various setting change.

I haven’t yet given much thought to the issue of colour reality in photography, the results are very similar, JPEGs seem a little too saturated and RAW work tends to smooth out the tones that light can have.
For the time being, my problem is mainly one of quality, with low-light and inhomogeneous outdoor shots complicating my work, especially with a sensor that isn’t the best for this situation.

But I can already see that most of the comparisons will be made on video, and the photos will be more for beautiful images rather than realistic ones.
I’m not just planning to make comparisons, I also want to share certain night scenes by making beautiful images and videos.
Still, I’ve been happy for a few months now to be able to do things that I like and to finally be able to think about buying new lights. :smiley:
The equipment had a certain cost and I had to put aside certain pleasures, like buying lights. :slight_smile:

I should be getting the Nitecore MH25 pro with the new “made in Nitecore” LEDs pretty soon in the few videos I’ve seen it looks like it has a terrible green tint and I want to compare it to my other green tints and the colorimetric style of my videos.
But the beam seems pretty crazy, I love big hot spots and it looks big, round and relatively unique (without being a LEP).
Performance interests me too.
And I’m still a Nitecore boy. :stuck_out_tongue:

That’s all I could say for tonight, I’ll probably keep this topic updated more often.
I’m open to all your comments/remarks

Have a nice night :slight_smile: