Driftwood Quantum X Settings, Series 4: Cluster v2, Mysteron, Sedna, Orion...etc...

No_SuRReNDeR

-

shian

Dude, chill. I know. That's why I said it was accurate.

thepalalias

This weekend I did some testing by the observatory - shooting the same shot with various ISOs in a situation where light could not be compensated by aperture, etc. I hope to post some shots today or tomorrow that give a sense of the underexposure noise vs high ISO noise for people that have not tested themselves. But my memory is that I stopped at ISO 640. I did the testing with either Sedna or Mysteron and cannot wait to look at my notes and find out which one. :)

No_SuRReNDeR

-

yeehaanow

@jvalal Quadrocopter.us It's the cinestar 6 kit with a gimbal of my own design. I plan on doing some back to back flights with Sedna vs mysteron soon. Trying to figure out a good 60p patch to pair with them too.

thepalalias

Here is an update from @Driftwood (Nick).

"Cluster v2 is almost finished, Canis is still undergoing tests... Ive had a busy few weeks and will continue to be filming projects until end of April so I have little time to finish my testing until last week of April. Plu wiki and so much else! "

He asked me to pass that long to everyone, so there you go.

So no pestering for new settings until the end of April, got it? ;)

Seriously, the settings Nick has made so far are already really, really great. The other day I was filming Muay Thai and I shot some of the footage with Mysteron in 50 or 60P on a GH2 and some of it with the stock settings on a GH1 in 60P. What a difference.

luxis

@thepalalias yeah i totally agree for the mysteron...i went back to it from sednaAQ1_B...the image is soo freaking nice is scary.and the bit rates are sometimes half of sedna which has better grain, etc but the jump in size and and all the issues i had with it stopping on the 720 50p are just not worth it for me at this moment...when i have all the 95/65 san disks i need i might reconsider but the end of April(Cluste v2, Canis;) will be sooner than that for sure...

@awasome,thanks for the flight...wait, do i feel a little envy for the hexacopter?... ; )

jvalal

@yeehaanow the footage is smooth. I thought it was a dolly at first. A bit pricey for my wallet right now, but something I'll keep my eye on as prices drop.

adamquesada

So it looks like most prefer Mysteron.

peternap

@adamquesada Overall, yes. I'm still on the fence about 720 60 since I shoot that the most. There just isn't a lot of gain over Sanity.

rajamalik

@adamquesada most prefer sedna c and some prefer sedna b,.. i prefer sedna A,.. but all of them prefer the Aq1 versions of sedna,... ;)

rajamalik

@driftwood is it true until april last we wont be getting canis majoris,.,. i thought u had set this between april 14 & april 20??????????????

blackspot

@driftwood Hi! Is it possible to decrease Mysteron Burst-set. bitrate a lil bit to get record times like 30sec? or does it "screw things up" someway?

modernhuman

Some thoughts about Picture Profiles:

I've been testing Vibrant (-2,-2,-1,-2) and Smooth (-2, -2, 0, -2), with Sedna A AQ1 (24H). I've pretty much come to the conclusion that Vibrant has more accurate (Canon like) colors, but Smooth might have a tad (1/4 stop) more dynamic range. But Smooth does weird stuff with greens and reds, it's also a lot more desaturated, and in general the color rendition is strange.

Based on tests Vibrant appears to hold up better in grading too. Using this logic: 1. Underexposing when capturing, to preserve highlights 2. More extreme grade to push up low level details (from underexposed image) 3. Colors desaturate when pushed 4. Vibrant has more color to begin with, so it holds the color better

If you do the same thing with smooth you lose more color and have to add it back in, this seems backward to me. Always cut or take away, before adding back in.

Anybody have any thoughts on this or perhaps have discovered similar and/or different results?

conscius

new shortfilm I filmed for a friend. Sedna A Q20 (L settings)

Later this week I'm going to shoot my own shortfilm. I'm going to use Sedna A Q20 24p H.

dbp

@modernhuman I've been sold on Vibrant as my goto setting for a while. It seems to be the cleanest, and to my eyes, it does offer the best color. The biggest difference is in the greens...Standard and Smooth, even with additional saturation to match Vibrant, have weird looking greens to my eyes. Especially smooth.

I'm with you, too. I'd much rather take away saturation then add it back in on an 8 bit 4:2:0 codec, regardless of the bit rate.

The only time I'd ever shoot smooth is if I was 95% sure that I wanted a flatter, desaturated look as my final look. Whenever I'm not sure, and I want room to mess around, I'll go with Vibrant. It may not have the highest dynamic range, and you may be stuck with the contrast, but you have to make compromises somewhere. I'd follow the whole "shoot flat, grade to whatever" workflow if I was shooting Red Raw files, or even AVC-Intra 100 10 bit 4:2:2, which grades far better than any GH2 hack.

duartix

My humble opinion about ISO, fundamented on a few tests I've done while trying to access the relation ISO/bitrate spending: avoiding the ISO bug, ISO160 is definitely the cleanest and ISO320 is cleaner than ISO640 so I believe base is ISO160 because noise & bitrate rise normally and linearly.

@Mr_Moore said:

"We don’t know for sure, but I believe the reasoning is as follows. The base ISO for the GH2 in stills mode is 160. Assuming that one video pixel is interpolated by using four stills pixels, 160 × 4 = 640."

@duartix thinks:

I've seen that logic repeated, but not knowing whether the sum of the pixels is averaged in the end, can defeat it instantly.

RRRR

regarding grading e.t.c. Gh2 footage holds up well enough to be pushed pretty far. The only difference when grading and choosing color profiles is if you want to push or pull color. Vibrant is no more natural looking than the others, but you might think so if you have shot a lot with Canon. Unfortunately there is no truely flat profile - such flatness has to be achieved by other means: lighting, filters e.t.c.

IMO the old underexpose trick does not work well with the gh2. Instead I prefer to expose as far to the right as possible on the histogram, meaning that I do not have to lift areas that are likely to have noise in them.

But if you have a gameplan for a certain look and know you can achieve it with good results, then that is a good gameplan and one you shouldn't doubt too much.

modernhuman

Thank you @dbp, seems that we have come to a similar conclusion. Although, with numerous tests, I don't think smooth has, in the real world, anymore useable dynamic range. But it does have less contrast, in the low mids. So that could equate to less noise, if you're not having to push that range up in the grade. Either way I think I'm liking vibrant more and more.

@RRRR I definitely agree that neither profile, nor camera, have perfect colors, but Vibrant has far more color information, and the greens and reds have a more natural appearance. As for underexposing or shifting the histogram left, instead of right, I will have to do more tests. Ideally, the lighting is controlled and the exposure is correct, but when shooting in a forest or some difficult location, underexposure seems a bit more favorable. Definitely, overexposure (and pull back in the grade) is a terrible idea. I guess I'm just saying err on the side of underexposure, when in doubt.

obscura

@rajamalik (@adamquesada) see, link below, there is no banding problems with it, it's "Sedna Q20 ver.A", but there are videos with for example Sedna's incl. AQ1 ver.A, raja I know you are still fan of it, ver.B C etc. and even Q20 ver.A with (incl. blue sky) banding problems. But no problems with this one Q20 ver.A, see also comments there, vimeo.com/39188868

@all Isn't it strange? (some with problems and some not, they are all shooting respectively with same settings/mayne even setup)

obscura

@conscius nice waiting for the shortfilm with Sedna q20 ver.A settings

Zaven13

@yeehaanow. If you are interested in a good 60p to combine with Driftwood's 24p settings, I am using a hybrid setting that uses Sedna A Q20 for 24H, L and Sanity 4 for FSH, 720p60 and HBR. However, I am using Sedna matrix for Sanity instead of the stock matrix. I am very happy with this combination setting. I am getting anywhere from 35 - 50 mb for 60p, FSH and HBR with great quality. BTW, Sedna A Q20 in 24L with averaging 75mb produces great image quality and it spans. All of these modes work with Sandisk Extreme HD 32GB 30mb/s card including EXTC mode.

shian

I'll be covering the different modes and their effect on dynamic range in the CGT Film School series (I've already covered "smooth").

And those shooting Vibrant clearly aren't using ColorGHear, or otherwise they'd know better to squeeze as much detail as you can get into the image at the extreme ends of the dynamic range first, and then let more powerful processors and algorithms handle contrast and saturation in post.

luxis

what so wrong with standard then? 4x-2

YOSS

STRESS TEST MYSTERON + SANDISK 30MB/S 32GB

@jfro , this is continuing what we talked a few days ago.

-This is only a stress test. Don't look at the composition, focus or motion blur... it's only about bitrate & writing speed errors on the mentioned card-.

Everything was shoot on 14-42 kit lens at 14mm & f8. shutter around 100, iso 320.

-1st shot (clip 41), 24H, smooth -2-2 0-2 / walking--> Got a stopped with a speed error at 1'30" aprox.

-2nd shot (clip 43), 24L, smooth -2-2 0-2 / only moving the camera around--> Stopped MAnually at 4'30" aprox.

-3rd shot (clip 44), 24H, standard 0 0 0-2/walking --> Got a stopped with a speed error at 2' aprox.

-4th shot (clip 45), SH 60p, smooth -2-2 0-2 /walking--> Got a stopped with a speed error at 3'30" aprox.

-5th shot (clip 46), 24H 300%, smooth -2-2 0-2 /walking--> Stopped MAnually at 10'30" aprox clip length 3'20".

-6th shot (clip 47), 24H, smooth -2-2 0-2 /moving the camera/ the subject was less busy that shot 1. Did that in order to try improving recording time--> Got a stopped with a speed error at 3' aprox.

-7th shot (clip 48), 24H, smooth -2-2 0-2 /walking / Another variation, walking. This time got stopped with a speed error at 4'10" aprox. File size 3,99GB. Span error (the only one I had in this test) All other files were smaller.

80% doesn't work. EX tele yes. HBR 30p yes. -didn't test on that a lot. just tried to record.

CONCLUSSIONS:

Mysteron is great, at least for my needs.

You can get wide shots (usually for establishing purposes) 1'30" should be enough.
If you want to shoot a time-lapse using video, go for the 24H 300% or the 24L settings. Or better, get an intervalo-meter (20 USD on ebay) and do it properly with manual lenses & stills (JPEG/RAW).
720 60p works great! 3' should be enough for any slow motion!
If you are following an action, it should be a similar situation from my test "7th". I got about
4 minutes 3.99GB. The limit is the span.

Thanks VK & Nick Driftwood for making this cheap plastic box a real video camera! -with intra-frame Super Hi bit-rates!!!!!

-It's much more than I expected for the price!

note: with the same card, I couldn't get 60p with sedna.

-Hope it's clear and will help.

Howdy, Stranger!

Categories

Tags in Topic

Top Posters