@feha I´d love to watch the original file... looks good on youtube.. did you have any issue with 25p?
Waiting for the 720 Intra for super slow motions... ;-P I already have a headache to understand the best and/or the differences of settings for the HBR 25p mode.. Now the news with the matrixes... I´m very gratefoul for your work and effort, but for people that are not so long in this forum like me is very much stuff!!
@driftwood Nick, I did some testing with Orion 4b and Sandisk & 64GB 95mb/s & voigt 25mm found the same results as @Mr_Moore
Profile: Nostalgic -2 0 -2 -2
24H 1/250 ISO 1600 — spanned
24H Ext Tele 1/250 ISO 1600 — spanned
Vari 80% 1/250 ISO 1600 — recording stopped after about 8 seconds. No lockup. Plays back in camera
Vari 80% Ext Tele 1/250 ISO 1600 — recording stopped after 7 seconds. No lockup. Plays back in camera
HBR (NTSC) 1/250 ISO 1600 — “Motion recording was canceled . . .” after approximately 40 seconds. Lockup. MTS file 0 bytes.
HBR (NTSC) Ext Tele 1/250 ISO 1600 — “Motion recording was canceled . . .” after approximately 40 seconds. Lockup. MTS file 0 bytes.
720p (NTSC) 1/250 ISO 1600 — spanned (video stream bit rate 53.3 mbps according to MediaInfo)
720p (NTSC) Ext Tele 1/250 ISO 1600 — “Motion recording was canceled . . .” after a few seconds. Lockup. MTS file of 0 bytes.
Not sure how these modes are supposed to work or if 24H was the main focus. Just giving you feedback.
@driftwood, I have been using Quantum X Orion CBR all week to do B roll on a doc and I love it, so does the editor I have to say. I have noticed some banding in the red channel but it's only visible on high iso or badly exposed shots. I have tried out Pictoris tonight cause I would love to shoot some interviews with it. I really like the new matrix over the Pictoris orion matrix. How much does it differ from the orion matrix, can you see problems intercutting the two?
Nick. My quest has been to have a great HBR NTSC picture, today I chose x v5cVBR. Did a test, no crashes on my air conditioner fins of death. Since it is wrapped I ran it through 5dRGB and Cineform> Cineform rendered it better IMO especially the motion. Now another problem: banding please take a look at this clip
any suggestions? ISO 320 (yup I went to 400 first) NOS-2 everything. Sorry I imagine the Driftwood test kitchen is a little cluttered lately.@sohus Hey! You’re much more daring than I am with your shutters speeds. I suppose that explains shorter recording times you got the HBR mode.
@driftwood Nick, in case you overlooked my report, here’s the direct link so you don’t have to hunt for it: http://www.personal-view.com/talks/discussion/comment/44878#Comment_44878
@everyone I believe shooting static stress charts for the 720p mode is not as useful, since it’s not an all-intra mode. I guess an animated sequence with two different charts would be better, unless you’re willing to babysit the camera and wiggle it around all the while.
@izash i filmed with the lumix 14mm and canon 24-70 with kippon adapter. Smooth Profile all -2. Im thinking of putting sharpness up to 0 though. Then edited in CS5.5. Exported as prores 422. Recut in FCP, then graded with Color.
@Fulgencio I used on HBR 25p mode, no problems, if i have more time will upload mts files, but i do not have a pro account at vimeo ... There was no any correction, used: Fim Mode Vibrant: 0, -2, +2, -2 Exposure mode Manual: 1.4 50 iso 160 - 320 ... Focus Manual
@Mr_Moore I like overall stability so that's why I try these speeds. I also use a motion death chart that really tortures settings. Still, older settings used to work without problems.
I just tested Pictorus Orion matrix CBR on Transcend 24H not spanning HBR not spanning 720p speed error after approximately 1 minute Also 1/250 ISO1600 This is with Voigt, Panasonic lenses will be worser (sharper!)
I hoped that this setting would be easier on the card but it is not problem solver for me.
Some more testing on Pictorus Orion Matrix CBR:
720p60 fails on Sandisk 95mb/s and on Transcend 32GB Class 10 720p50 works on both
24H and HBR are not spanning on Transcend, they stop a 7:00 and 7:30 minutes.
Some more work to do @driftwood :)
@sohus @driftwood "720p60 fails on Sandisk 95mb/s and on Transcend 32GB Class 10 720p50 works on both"?
I dont get it. 720p60 works perfectly with my 8gb 30mb/s Sandisk Extreme HD Video class 10. :/ I have been planning to buy 95mb/s 64GB Sandisk SDXC card, but as far I see, apparently its not guaranteed to work (with Orion), is it?
What kind of death charts are you using? Id like to try one too.
@sohus Thanks for more test results!
Speaking of test. Only now, having conducted just a few (and not very exhaustive) test, do I realize how much work goes into developing and testing new settings.
I have been planning to buy 95mb/s 64GB Sandisk SDXC card, but as far I see, apparently it’s not guaranteed to work (with Orion), is it?
Apparently not, however frustrating it is for us owners of 95mb/s 64GB cards.
I probably wouldn't use HBR at all if you could stream 720/60 out of HDMI without it switching to 1080/30i. The 720/60 looks better and is plenty for a 1080/60i broadcast conversion.
Oh well, can't have it all! Or, well, you can, just spend more money. xD
@blackspot I took all 4 death charts that are commonly used here, but them in a Motion/AE project and put some wiggles/scales on them and loop that animation. It is very reliable for me as settings that survive it never fail me in the field because it has resolution, sharpness and heavy motion (stops a few times but always moving).
The 1/250 and ISO1600 test I do is extreme torture, I am sure you can manage in the field with ISO 1/50 and ISO800 if you need to but I prefer a setting with some overhead so I am (almost) never surprised in the field. There are easy tricks to 'tame' a patch, like lowering the AQ from 4 to 3 and lowering the frame limit. I think these trade-offs can be worth the extra room they give you in terms of reliability and using cheap cards.
i.e. Orion 4b doesn't last 50 seconds on Transcend Class 10 but when I lower the frame limit a bit and set the AQ from 4 to 3, I get 4gb files with 700k iFrames (instead of 780k) which for me is worth the trade-off. That's what I am doing right now, testing @driftwood's great settings and see what I have got.
Orion 4b works fine for me on the Sandisk 64gb 95mb/s and I can shoot 40 seconds on my Transcend and the 24L setting lasts 4GB on it or forever with the 95mb/s. So that can also be a strategy. Most clips I shoot are not longer than 40 seconds, unless they are interviews so I can shoot them on 24L anyway.
The important thing is that at a certain point you pick a setting, know what it can and can't and live with the limitations. Otherwise you will be testing forever. Especially with @driftwood 's insane working speed. He dropped like 4 new settings on us today and now we have choices of CBR/VBR and different matrixes. Overwhelming.
I am especially looking forward to @stray 's analysis of the matrixes in Pictorus. I think I stick with Orion matrix, I really like it and you can use sharpness if you need it (with non-Panasonic lenses).
Sorry for the blunt question but - out of the quantum X:es, which you reckon would make most sense for me to use for a music video shoot we're doing this weekend? It will be shot indoors, controlled light (some flourescent). I want to shoot it as flat as possible, long dof. Got a 64gb 95/s sandisk and some 16gb 30/s and spanning is not a priority.
I'm asking since I have not had time to do any extensive testing.. way too much work atm. Did a BW shoot today with Orion v4, high iso (6400), looking forward to editing that / checking out how it compares to earlier similar attempts.. I'm set on using Orion, but maybe someone with the experience of the latest patches can tell me if that is a good idea or if I should go with other settings.
Here is a quick test of Orion V4. Backyard on a monopod with the SLR Magic 12mm. It was very close to the web and hard to keep from touching the lens.
I will have a film shooting at this weekend. Which patch would you guys suggest for the most filmic style. I've heard from many people that the quantum x v2 patch should be the filmicst of all patches.
I have filmed some szenes for this movie with quantum 9b before these other patches came out.
What would you mention?
thx for the help
Spotted the identical files (they are also here: http://www.personal-view.com/talks/discussion/comment/45001#Comment_45001 ) but that's because I was curious as too how much different you CBR x VBR approaches would be, and surprise! They're almost identical except one uses AutoQuantizer and the other uses Quantizer. Is that all there really is to it? :)
@Driftwood. Just combined your Orion v4b 24P settings with your Pictoris CBR Orion 1080i60 and 720p settings. Did some quick tests of the 1080i/720p/HBR using Sandisk Extreme 32GB 30/mbs and Panasonic 14-140mm AF, OIS. I also changed 720p AQ from 4 to 2 to reduce the chance of write errors.
1080i60 FSH/FH fails after 6 seconds.
720p60 SH/H work until I zoom toward 140mm on a more detail subject and as soon as the bit rate goes over 60MB, it fails on both settings. Video quality looked very good though.
HBR worked without a problem however, too noisy for my taste.
@driftwood: I'll be testing your 720/60p settings on GH2 with Sandisk 32gb 30MB/s cards this weekend, along with 1080/30p on my GF2. Will report back on my results.
i did test the 720p pal mode, in the motocross clip.
It was shot on Quantum X 'SpanMyBitchUp' v2b
i did not find any problems, all went fine even high iso , high shutter and tele2 mode on...
I've got some v9b hack footage being cut into a show, sitting in the DI theater coloring it, parked in front of an 18 foot screen and piping multi formats out through a 2K projector.
Shaking my damned head at how well this stuff is grading, and how bad it stomps on 5D footage in the same context...
@driftwood : I'm testing your "Quantum X v5b 'ß Pictoris' VBR / hybrid GH2 meets AVC-Intra Matrix" and I'm finding a bit of a big quantization range:
Static SCENE (low light) at ISO800 (file attached)
Same scene same aperture at ISO3200 shows even bigger variations:
Frame--Type QP---Min---Max--Range---DC Skipped QST-High
00000 (IDR) 20 20 20 1 8 0
00001 ( I ) 20 18 28 11 7 0
00002 ( I ) 20 18 28 11 7 0
00003 ( I ) 20 18 28 11 7 0
00004 ( I ) 20 18 28 11 7 0
00005 ( I ) 20 18 28 11 7 0
00006 ( I ) 20 18 28 11 7 0
Is that OK? Should I load one of them in StreamEye and see which blocks are being skinned?
(EDIT) I did, and it doesn't look that good, notice how at about 4/5 of the image the encoder kind of like enters into "Quantizer Fallback" (sorry, I can't find a better expression).
Perhaps the issue here is that it's already limited at ISO800 as one might guess from the attached XLSX. :(
Needs a biggest base quantizer, perhaps?
It looks like you're new here. If you want to get involved, click one of these buttons!