filmov
tv
Alexa 35 vs Venice 2: Codecs, Sensitivity and Noise Tests (4K)
Показать описание
Taken from the test sequence that I've shot for my upcoming online serial. DI files will be in UHD and main deliverables will be in HD, so I don't want to record beyond UHD resolution. ARRIRAW and HDE were out of my time and budget, so I've decided to test ProRes on Alexa 35 and X-OCN on Venice 2.
Please note that RAW codecs (ARRIRAW, X-OCN, etc.) will be noisier than video codecs (ProRes, DNxHR, etc.) because of the lack of internal processing. So ProRes on Venice 2 will be LESS NOISY than X-OCN and ARRIRAW on Alexa 35 will be NOISIER than ProRes.
A much more fair comparison would be ARRIRAW (or HDE) vs X-OCN XT or ProRes on both. Unfortunately I didn't have time to test them all. Maybe next time...
PS: Because of the YouTube's compression, I advise you to watch this video on 1440p and beyond, even if your screen doesn't support it.
🔗 Link to original file (27GB, UHD, DNxHR HQX, 10-bits)
----------------------------------------------------------------
CHAPTERS:
00:00 Intro
00:11 Arri Alexa 35
02:22 Alexa 35 - Noise Pattern, Noise Floor
02:35 Sony Venice 2 - 8K
04:17 Venice 2 - Noise Pattern, Noise Floor
04:32 Alexa 35 vs Venice 2
05:00 VS: Noise Pattern, Noise Floor
----------------------------------------------------------------
Arri Alexa 35 vs Sony Venice 2 (8K)
Codecs, Sensitivity and Noise Tests
Featuring
Azra "The Sleepless" Topuz
Crew
Ugurcan Taylan Adiyaman
Gokhan Can Guzel
Salih Bolme
Melihcan Guler
Rentals
SVC Camera
Flare Isik
----------------------------------------------------------------
Internal ND filters were used for 3200 ISO tests.
4x vs 8x multipliers were used for Noise Pattern tests because of different pixel counts.
(~8m pixels for Alexa 35 vs ~16m pixels for Venice 2)
All shots were matched with X-Rite ColorChecker Passport Video.
Davinci Resolve Noise Reduction Values:
Temporal NR: 2 frames, Better, Medium
Temporal Treshold: Luma & Chroma 9, Motion 3
Output Device Transform:
ACES 1.3 Rec709
----------------------------------------------------------------
Shot & Edited by
Engin "Teddy" Hayiroglu, ɢʏᴅ
----------------------------------------------------------------
SVC Studyolari
Istanbul - 2024
Please note that RAW codecs (ARRIRAW, X-OCN, etc.) will be noisier than video codecs (ProRes, DNxHR, etc.) because of the lack of internal processing. So ProRes on Venice 2 will be LESS NOISY than X-OCN and ARRIRAW on Alexa 35 will be NOISIER than ProRes.
A much more fair comparison would be ARRIRAW (or HDE) vs X-OCN XT or ProRes on both. Unfortunately I didn't have time to test them all. Maybe next time...
PS: Because of the YouTube's compression, I advise you to watch this video on 1440p and beyond, even if your screen doesn't support it.
🔗 Link to original file (27GB, UHD, DNxHR HQX, 10-bits)
----------------------------------------------------------------
CHAPTERS:
00:00 Intro
00:11 Arri Alexa 35
02:22 Alexa 35 - Noise Pattern, Noise Floor
02:35 Sony Venice 2 - 8K
04:17 Venice 2 - Noise Pattern, Noise Floor
04:32 Alexa 35 vs Venice 2
05:00 VS: Noise Pattern, Noise Floor
----------------------------------------------------------------
Arri Alexa 35 vs Sony Venice 2 (8K)
Codecs, Sensitivity and Noise Tests
Featuring
Azra "The Sleepless" Topuz
Crew
Ugurcan Taylan Adiyaman
Gokhan Can Guzel
Salih Bolme
Melihcan Guler
Rentals
SVC Camera
Flare Isik
----------------------------------------------------------------
Internal ND filters were used for 3200 ISO tests.
4x vs 8x multipliers were used for Noise Pattern tests because of different pixel counts.
(~8m pixels for Alexa 35 vs ~16m pixels for Venice 2)
All shots were matched with X-Rite ColorChecker Passport Video.
Davinci Resolve Noise Reduction Values:
Temporal NR: 2 frames, Better, Medium
Temporal Treshold: Luma & Chroma 9, Motion 3
Output Device Transform:
ACES 1.3 Rec709
----------------------------------------------------------------
Shot & Edited by
Engin "Teddy" Hayiroglu, ɢʏᴅ
----------------------------------------------------------------
SVC Studyolari
Istanbul - 2024