Driftwood Cluster X series 3:│moon T7│ЅріzZ T6│Nebula T7│Drewnet T9│

Jim_Simon

@GravitateMediaGroup My testing has been with the original T7 from the other thread. I have not tried 7b or 7c yet.

What frame rate are you seeing reported?

GravitateMediaGroup

T7c recorded at 23.9 for me

Jim_Simon

I wonder if Nick changed it in 7c.

Jim_Simon

So I made the Blu-ray...

This is hands down the best, most film-like image I've ever seen from my camera. And I actually think the 24 fps has something to do with it.

Watching film source, which is shot at exactly 24 fps, there is a smoothness to the motion that seems missing from video source, particularly with camera motion. If you watch a 23.976 video, there always seems to be a sort of shimmer in the details on pans and other camera movements, softening those details until the camera comes to rest, where they sharpen up again. Faster camera movement makes the shimmer more pronounced, but even with quite slow camera movement, it's still there to a degree.

That high frequency shimmer is largely missing from film content, even with faster camera motion, and I've often wondered if the difference is that film uses 24 fps, and video uses 23.976 fps. Looking at this 24 fps video, I'm fairly well convinced that is exactly the issue.

Once again, Nick, you are a god!

driftwood

Guys, if you look at the original moon T7 I also gave you 30 FPS and 60 FPS and of course 24FPS, moon T7b/c goes back to 29.97, 59.94 as well as 23.976.

So do you guys want both as trial tests?

producer

I can test the T7b and T7c too, but can't find them, where are their links?

driftwood

@producer @all. Hang on, I'm going to now include both versions in the link on page 1 in one zip file. One is labelled moon T7 24fps etc (setd) the other moon T7 23.976 (sete). ok.

driftwood

Cluster X DrewNET T9 1080/24 screengrab example and elecard quantisation evaluation - amazing at this detail. Third pic: motion vectors going to a static at 720p50 DREWnet. Note i frames go huge! in Cluster X as the image comes to a standstill. With movement b frames become larger and i/p frames naturally compensate in size to the bitrate.

bkmcwd

Hi Nick,

I will do tuning Spizz tomorrow. :-)

driftwood

Excellent. Thankyou Katsu @bkmcwd

Jim_Simon

@driftwood

Would it be possible to get 24, 29.97 and 59.94 in one patch? Or are they tied together somehow so that if we get the normal 29.97, we also have to take 23.976?

Because I'm loving this exact 24 thing. The motion just looks sooooo much better. But the issue is that Blu-ray and broadcast can handle 24 exact, whereas they both need the oddball 29.97 and 59.94 for the video look, they don't do 30 and 60 exact. (Meaning putting them on disk will invoke a small 'conversion' of sorts.)

driftwood

@Jim_Simon Yep just goto ptools, open the sete/24p moon T7 variation setting, then under the section titled 'GOP Related' in 'Patches for Testers' uncheck the GOPx2 Time for each of the ones you want to switch back to default - ie uncheck 720p60GOP x2 Time and 1080i60GOP x2 Time. Of course, don't uncheck the 1080p24GOPx2 Time. Then youre done. Save patch with shift to a new letter, then save firmware to sd card and flash your newly created firmware.

cjdincer

ı really like the idea of having 24 fps insted of 23.976. I can not do the tests cause my 2011 MBP's discrete radeon graphics card died on me last night and All I see the apple during the boot up than just gray screen. it is gonna take sometime to get it fixed ı guess. now I am left with an İpad. at least ı can follow the posts.keep testing guys.hope to cacth you up soon.

BurnetRhoades

@Jim_Simon you're not going to love the 24P thing once you have to sync up to non-throw-away audio recorded on an external device. There's a reason 23.976 exists. It's called post production.

matt_gh2

@BurnetRhoades Not too familiar with this, but are you saying if I shoot 24P and have audio recorded on separate recorder and I use a clapper board, that the syncing will be impossible. ..or is just the sync drifts out of sync after awhile? So I guess 23.976 makes syncing with clapper board accurate?

cjdincer

@burnetrhoades are you saying that the audio recorders record in 23.976 fps. never heard of it before.

BurnetRhoades

You will get sync drift in your audio if you shoot at 24fps. ALL "24P" cameras shooting at 23.98fps is not an arbitrary fact and it has nothing to do with "video look" like dear Jim is going on about.

24fps, don't do it, or, go with God. It's not my headache.

matt_gh2

Figured your were referring to drift. Thanks for heads up

driftwood

Guys, for non-narrative based stuff the 24p setting could prove useful so Im leaving it as a alternative.

flaschus

lol... just make sure u pack the ol' sound devices so u can get the 48.048 k audio to match up..

no probs in post

driftwood

Bring the revox.

BTW: Right, final DREWnet T9 being uploaded .

act

Hi! I have not been updated since Cluster v.6_6 Nebular. Actually, I tried about 3 newer settings, but can't find any most stable settings without bugs for long recordings at all modes.

I'd like know if there are newer settings with same stability, no freezes, spanning at all modes and about 44Mb/s @24L and 88Mb/s @24H. (I use Sandisk 64GB 45MB/s and 95MB/s) You know, we can't risk with client's "one of the lifetime" event like wedding, so, no way for using trial settings. But I don't stack with this old settings. I've seen that only one new settings Red Shift "on paper" looks like Cluster v.6_6 Nebular. But almost nobody post here clips about Red Shift, is it better then Cluster v.6_6 Nebular? What new settings have same stability like Cluster v.6_6 Nebular and worst to update from it?

driftwood

@act Give Cluster X DREWnet a try - most recordings around the 78-80Mbps for 24p H. Is that too much? Nebula will be coming soon but there doesn't seem much interest from the majority in a mid GOP like Nebula's 6 allround.

These Trial settings as you can see over the months from users work very well - and though they are experimental to a degree - they are kicked into shape over time until rock stable. At higher bitrates we can't guarantee spanning on all modes (can anyone?!). But rest assured with the card you have, moon T5, T7, Drewnet T8 and other Driftwood settings on 24p span reliably.

As you can see I'm mainly researching best quality at each GOP. Redshift is good low bitrate setting - try it too. Apart from that there's plenty of low bitrate settings around outside the House of Driftwood.

maddog15

@act

Did you try Slipstream #1B? Worked great for me. My camera functioned like it was using stock firmware - but shoots at 3 times the bit rate of the highest stock setting on 24H. (72Mb/s I believe.) No freezes, playback in camera, spanning etc. etc. ALSO... worked fine on my SanDisk 30Mb/s 32GB card! Mr. Driftwood has taken care of just about everyone's needs. In my humble opinion.

You'll find it here...then go to page 1

http://www.personal-view.com/talks/discussion/6429/driftwood-clusterx-series-2moonЅrіzznebuladrewnetslipstream-redshift/p1

driftwood

My FINAL 'Driftwood' tested version of DREWnet T9 (classic Long GOP) is now up on page 1. Its a truly brilliant setting.

A new version of Nebula 6 GOP will be up on page 1 shortly utilising the findings of my latest research.

Howdy, Stranger!

Categories

Tags in Topic

Top Posters