ONLY SETTINGS COMPARISON, NO SMALL TALK!

Kihlian

Hope this is welcome.

Since there is much interest about "the best settings" and a lot of good settings are coming out for different purposes, but the useful settings confrontation are rare and scattered over the topics. I propose to start this topic to collect this confrontation. Rules:

1) Only video or screen shot confrontation with different settings, but SAME: ISO, Picture profile, light condition, Focus, subject, registration mode.

2) No post of Video or screenshot of only 1 settings!

3) No comments, otherwise the discussion deviates on discussions about the weather.. (If VK welcome this topic hope it delete all non sense comments)

Possible exception:

1) The creator of the settings can also add comments to comparisons

Kihlian

So I start with a confrontation that I have made right now between: SEDNA AQ1A, GOP3ZILLA_2.11/modified, ORIONv3. The shots where -3 underexposed.

ORION=32822 Kbts/s average GOP3ZILLA=24978 Kbts/s SEDNA=56576 Kbts/s

EDIT: 720 50p

Stray

@towi Its only really fair to compare a GOP1 setting to a longer GOP by making sure that you have selected a B-Frame (or P-Frame) from the long GOP setting. Thats what I always did in the past, and the longer GOP setting would hold up well against the GOP1, it was better in most cases too. In fact when comparing a GOP12 setting I would compare the frame right in the middle of two I-Frames and it still maintained comparable quality. This was before the matrices which mitigate the smaller I-Frame sizes of GOP1, so I figure now the difference may be less obvious. Mind you, using the same matrix the quality may be better in the longer GOP. Still, this was a 720p test (which reaches maximum quality at much smaller I-Frames), in very poor light of a low contrast scene, the CBR-like setting is going to win that everytime.

towi

@Stray

okay, I get your point. Makes sense.

However you'll always see decreased IQ on P- and B-Frames, especially when you look at the noise pattern in the blacks (even in Ralph_B's Sanity patch ... which works surprisingly well). With motion the decreased quality of B-Frames is pretty much leveled out... IMHO (I've compared Sedna A, Sanity 4 and GOP3ZILLA - with motion they produce a very similar noise pattern).

Vitaliy_Kiselev

Guys, either stop talks or I close the topic.

thepalalias

Will be conducting tests on Sunday that meet all of the criteria @Kihlian specified, to post here and elsewhere.

If you can contribute your time or equipment please e-mail me (this in L.A. and Pasadena). Do not respond in this thread - it will clutter.

thepalalias

Sedna AQ1 A vs Mysteron

Shot with Rokkor 58mm f/1.4 at f/2.0, WB 4500K, ISO 1600, in 24H. -2 for everything (Smooth profile if I remember correctly), 24H.

Sedna had average datarates in the 126 to 130 mbps ergion and the Mysteron had 116 to 120 mpbs on average.

I picked a frame from each sequence that the most prominent searchlights in similar positions to make comparing easier. Shots were done in PAL mode and I also shot HBR, SH, FSH and EX versions of all 4 for both Mysteron and Sedna AQ1 A.

I have also included the same shots with extra sharpening, as well as the same shots with no sharpening but a contrast and gamma adjustment designed to emphasize noise.

rajamalik

@thepalalias will u try sedna aaq1 vs mysteron burst ???????? stability , detail and filmic test??

Kihlian

Test: Smooth all -2 with a Canon FD 50 mm at F2, ISO 160 all 50p. I took the frame number 12 so for every setting I think I grab the I frame. The white stripes are moved from a fan. Tried to build a high contrast scene with deep blacks and blow out whites. N.B. Unfortunately in the last shoot (cluster v7) the chair was moved from my niece so I put it back but not perfectly. Sedna Q20 A = 52208 KBit/s M6 Sports E = 38494 Kbit/s FlowMotion v 2.02 = 34428 Kbt/s ValkyrieApocalypseNow= 42293 Kbit/s ClusterV7´ApocalipseNow´Nebula6Gop= 59497 Kbt/s

thepalalias

The following was an attempt to test general detail in a partially underexposed, mostly low detail, high ISO environment. Feedback should be put into another thread or sent by message.

Lumix 20mm f/1.7, MF (identical distant focus point for all shots), 60P SH Mode
ISO 12,800, 1/125, Standard -2, -2, -2, -2

Lighting was unchanged. Camera moved slightly between takes (was resting on a surface, not on a tripod) but was essentially static during shooting.

Here is a B-frame comparison (frame # checked with StreamParser 2.7) of the 16 settings that shot b-frames, exported without further processing to lossless PNG from VirtualDub.

Frames are in ascending order of the size of the resulting PNG file, not the original compressed frame size. To see which setting was used, just hover over the picture.

Settings tested:

Cluster 6 Nebular
Cluster 6 DREWnet
CM Night
Cluster 7 AN Nebular
Quantum 9b Sedna Adapted
Sedna Q20 A
Valkyrie 44 TYPE-ZERO2 Candidate 1.2 'AN' w cbrandin...
Mysteron
Flowmotion 2.02
SMBU v2
No Adverse Affects
Crossfire Var 1 Skintone Soft
Crossfire Var 3 CM Night
Sanity 5
Crossfire Var 2 Green Screen

Howdy, Stranger!

Categories

Tags in Topic

Top Posters