Driftwood Quantum X Settings, Series 2: Sedna A, B, C

proaudio4

SEDNA A AQ1: 24L (woops, meant to shoot 24H > see my next message... ) LENS: Panasonic 20mm 1.7 OIS: N/A F-stop: 51.7 ISO: 160 WB: Tungsten Profile: STD -2,-2,-2,-2 SD: SANDISK SDXC 64GB 95MB/s Streamparser screen capture: Senda A AQ1 ISO160 SS50 PAN 20MM 1.7 LENS CADENCE ISSUE.png

This shot was 2 stops under exposed, but still looks amazingly clean. Even more odd, I did not realize there was a cadence issue until I checked today.

Also, all other videos shot today looks great, but shows the cadence issue. All of them maxed out around 760KB, but fell in the cadence to around 200-300KB.

SEDNA A AQ1: Again, I accidently shot this in 24L, but it's good to know regardless. Also, don't bother tuning 24L if it takes away from 24H. This is IMO. LENS: Panasonic 14-140MM framed at 70mm on Strays Extreme chart. OIS: OFF (Still shot on tripod) F-stop: 5.8 ISO: 2500 WB: 4500 Profile: STD -2,-2,-2,-2 SD: SANDISK SDXC 64GB 95MB/s Streamparser screen capture: Senda A AQ1 Strays Extreme Chart ISO3200 SS1600.png

Ran 15:30 no errors. Spanned without issue. Interesting the i-frames are only 420KB average under this condition. But, more interesting is that ISO2500 certainly looks like it has much lower noise. Did not see a cadence issue under this particular test. Apparently this chart under these conditions did not demand as much bitrate which may have avoided the cadence issue.

BTW, using Streamparser 2.5 Is anyone else seeing this with Sedna A AQ1?

Nick, any thoughts on this? Man, I hope you figure this out since this looks great.

klem007

first small test with sedna C had some weird pixels around sharp edges in this last shot don't realy know why, elsewhere the footage holds very well in heavy grading; love that, the noise is so fine and devided under the 3 channels with nothing sticking out too much. I like c so far, i'm gonna try b next =) camera blocked in etc mode after 7 seconds =)

proaudio4

@spacewig>"What does he do with the grass he collects?"

Smokes it? ;)

islanders66

@driftwood, To get 720p to work I lowered the 60fps Frame Limit to 130000 so the 45mbs card doesn't write error, but the red flowers look weird like pixels sparkling or like little ants crawling around. I tried lowering the Frame Limit to 120000 but didn't help. The 24p looks great. Maybe someone else can test 720p outside with a 30 or 45mbs card.

proaudio4

Hmm. Maybe I was smoking that shit.... I just realized I was shooting 24L ! WTF, that explains the lower bitrate. Well, if anything.. This is good info regarding 24L, which I never use.

Also, I just looked at 24H, it's looks good with no cadence issue(of fucking course! lol!) I'll test for spanning now.

thepalalias

I mainly use 24L when I suddenly realize I need a much longer recording time and there`s no time update firmware. :)

proaudio4

@thepalalias Damn.... bummer man. I've been there before. 24L certainly shows the cadence issue. Oddly, the shit still looks good! lol

driftwood

@proaudio4 Of course it does!!!! None of my newer Quantum X settings are set up for L. You would have to pull frame limits right down and Im not prepared to foresake 24p at H for the inferior L. I am only after filmic settings. There's plenty of other settings that I do, and others do, if you want to save space, long durations etc... where L does not work. There are NO cadence issues. just that some of the i framess will be bigger and then flatten to a wrestle with the L setting bitrate/frame limit.

driftwood

@klem007 Nothing wrong with Sedna C in your video. The sharpe edges? Well, youve just found the problems in Panny's stock deblocking loop, this is being analysed for the up-and-coming new settings 'VY Canis Majoris' which will be a BIG analysis of deblocking in INTRA settings. The setting will employ Sedna matrices with new deblocking variables - hopefully better than Panny's default values.

Also, personally, I wouldnt choose C for that kind of shot :-) (use A)

driftwood

@islanders66 youre describing mosquito noise. 720p60 needs more analysis. 720p50 is easier to manage. I'll run some checks over 720 this weekend. When it works - it looks terrific.

proaudio4

Thanks @driftwood. 24L certainly was not my intention ;) I had a few to many last night and thought I was in 24H (This explains last night why I was writing Senda, instead of Sedna! LOL!)

I just checked spanning (24H) with SANDISK SDXC 64GB 95MB/s and all is well. This is your best work to date.

Thanks again! and looking forward to 'VY Canis Majoris' !

Zaven13

@driftwood. I am asking this question because you have tested the hell out of these settings and you probably have the answer on top of your head. How would you rate Orion v4b AQ3 against Sedna A AQ1? Thanks in advance.

driftwood

Both good, both different looks. Sedna explores measured frequency luminance and chrominance and is finely adapted Q in all freq range over the flatly 'try as high Q as you possibly can' approach with AQ set so high. Indeed Sedna is using very low QP throughout the freq range and if you set it to AQ3 or 4 it'll probably stop recording. Think of Sedna as simply, very tuned Quantisation.

jfro

Guess I'm going daffy. Just spent 10 minutes on personal-view and can't find the folder with patches. Where the heck is the Sedna patch?

driftwood

ALWAYS at the top page of this thread.

islanders66

@driftwood, thanks, I also tried B settings and got the card write error even after lowering the 60fps frame limit to 1300000.

It did look good when it worked. Kind of frustrating because 24 p looks so good but I'll just go back to Orion for now as a lot of my shots aren't static and I need 720 for my upcoming project.

killagram

@jfro check the 1st page of this topic.

thepalalias

@jfro More specifically, two of the zip files at the bottom of the first post. Some the earlier patches have links in the description but these do not. :)

jfro

Thanks all. Got it and camera is loading as we speak. Nice not to have to load the audio patch myself. Gonna load up the steadicam and take a trip outside. Driftwood, thanks again for your tireless work.

rajamalik

@driftwood it seems senda a b c helps in noise reduction,...iso 2000 looks like iso 1600 of orion v4b,.. senda c is more in reducing noise i believe,..i love the sharpness of senda A,...how can i get it in orion v4b the official intra matrix???

royfel

Broadcast promo (30sec) all driftwood, dancer part is Sedna c soft a20

Ntsc HBR Pana/ leica 45

cosimo_bullo

Nick, you've outdone yourself... Sedna A (AQ1) with panny 20mm stnd -2-2-2-2 is exquisite. The sharpness is there, but there's a smoothness I've never seen before, an organicness. Well done, well done, well done.

cosimo_bullo

@royfel - Just lovely...

rajamalik

@driftwood *** NEWS *** A new matrix and deblocking 'VY Canis Majoris' is being worked on to analyse the various stages of deblocking in INTRA settings. Issued 16th March 2012

How can we apply this new deblocking to orion v4b to get less noise in higher iso???? senda iso 2000 looks like orion v4b's 1600 iso,... a jump in noise,...looks good,.. i would love to add such in orion v4b,.. will u guide me in this,..??

Oedipax

@driftwood "Also, personally, I wouldnt choose C for that kind of shot :-) (use A)"

Referring to this video - I see mostly wide, landscape style shots... I thought C was the strongest for that (most detail, etc)? Do I have this backwards and A is preferable for detailed static wides?

I shot a bunch of locked off anamorphic shots on Sedna C today, still transcoding in 5DtoRGB but my initial impressions from MPlayer X are it looks very good. I'm just wondering if A would be better. There's still enough daylight here on the east coast US to do another round I suppose.

Howdy, Stranger!

Categories

Tags in Topic

Top Posters