dethis2

Joined on Oct 3, 2011

Comments

Total: 4, showing: 1 – 4
In reply to:

Stanislav Evstatiev: Just HAD to comment on this!
1. Compare? No? What's the ...
2. Low light (noise) performance - Just sliding a tiny bit down and seeing the text will give you a whole new level on how noise and contrast are a whole different level on the medium format.
3. Tiny vs FF in macro should give ADEQUATE understanding of FF vs MEDIUM? What is wrong with this word actually... Yep, they might be not sufficently bright lenses in the medium format to match 1.2 on FF, but how many 1.2 FF with more than a finger tip center sharpness can You list?
4. Portrait lens obviously start at 50mm for the most and goes to 100 mm for some pretty obvious perspective distortion reasons. Yes I've shot portraits with 8mm and it works - but these are not my USUAL ones :-) I've thought You shot the same shot along with the Fuji GFX 50S with Fujifilm GF 63mm f/2.8 R WR Lens on it? No? You didn't? Then what is the point of showing some sunset light unrelated portrait shot with one of the three best of FF world cameras?

"Others, .. Which is the a7R II and which is the GFX in the rollover ?"

It's easy even if I had not seen the samples before .. the left frame is wider .. the left is Sony A7rII.

Crossformat comparisons would be better if you had used "fair_framing" i.e subjects covering equal proportion of the frame on both 3/2 & 4/3 . With the current framing (equal height) you give a small advantage (area, resolution and noise) to the taller 4/3 ..

For this you need to rearrange the reference frame to sqrt(2) / 1.0 aspect ratio** and frame so that the 3/2 format covers the full height, the 4/3 format covers the full width. Equal "Dead space" area will exist at the sides (3/2) or top/bottom (4/3). Which makes res/noise comparisons fair .. also easier to find proper scale factors.

** (h * 3/2*h)=(w * 3/4*w) gives w = h * sqrt(2) .. happens to be exactly the ratio of most paper sizes (A, B)

Link | Posted on Mar 24, 2017 at 13:42 UTC
In reply to:

The Squire: Doesnt seem to specify a minimum bit rate?

You can have 10bit but at the same bit rate as a 'normal' 8bit video file, but you'll loose most of the benefit.

See the criticism of the new Sony FS5's 1080p 10bit mode - The normal bit rate cant cope with 10bit depth, resulting in compression artifacts.

UHD, 10bit, 4:2:2 at 200Mbps would be a good step up from the codec we see in most of the enthusiast cameras today.

Encoding at 10bit in fact saves bandwidth for the same quality ..

http://x264.nl/x264/10bit_02-ateme-why_does_10bit_save_bandwidth.pdf

http://ateme.com/en/technology/white-papers/Using-10-bit-AVC-H-264-Encoding

Link | Posted on Jan 7, 2016 at 04:04 UTC
In reply to:

dethis2: RAW conversion with Rawtherapee ..
Looks like LX100's sensor is very much alike GH4's regarding color filters and raw white level.
You need to add the following paragraph in RT's camconst.json file

{ // Quality X, preliminary support
"make_model": "Panasonic DMC-LX100",
"dcraw_matrix": [ 7122,-2108,-512,-3155,11201,2231,-541,1423,5045 ],
"ranges": {
"black": 16, // 16 is BL offstet calculated from exif data.
"white": [
{ "iso": 100, "levels": 2700 },
{ "iso": 125, "levels": 3100 }, // gh4 data
{ "iso": [ 160, 200, 250, 320, 400, 500, 640, 800, 1000, 1250, 1600, 2000, 2500, 3200, 4000, 5000, 6400, 12800, 25600 ], "levels": 4080 }
]
}
},

jpeg
https://drive.google.com/file/d/0B0NqktTgc54sa1JCWGhWNXM1dWc/view?usp=sharing
Settings
https://drive.google.com/file/d/0B0NqktTgc54sYmxmU2ZKbEh1Um8/view?usp=sharing

Maybe GX7 matches better .. I didn't try it
Do you find the colors in my conversion wrong ?.

Link | Posted on Oct 7, 2014 at 03:11 UTC

RAW conversion with Rawtherapee ..
Looks like LX100's sensor is very much alike GH4's regarding color filters and raw white level.
You need to add the following paragraph in RT's camconst.json file

{ // Quality X, preliminary support
"make_model": "Panasonic DMC-LX100",
"dcraw_matrix": [ 7122,-2108,-512,-3155,11201,2231,-541,1423,5045 ],
"ranges": {
"black": 16, // 16 is BL offstet calculated from exif data.
"white": [
{ "iso": 100, "levels": 2700 },
{ "iso": 125, "levels": 3100 }, // gh4 data
{ "iso": [ 160, 200, 250, 320, 400, 500, 640, 800, 1000, 1250, 1600, 2000, 2500, 3200, 4000, 5000, 6400, 12800, 25600 ], "levels": 4080 }
]
}
},

jpeg
https://drive.google.com/file/d/0B0NqktTgc54sa1JCWGhWNXM1dWc/view?usp=sharing
Settings
https://drive.google.com/file/d/0B0NqktTgc54sYmxmU2ZKbEh1Um8/view?usp=sharing

Link | Posted on Oct 7, 2014 at 02:51 UTC as 55th comment | 2 replies
Total: 4, showing: 1 – 4