We are in a preproduction stage for a GH2 Hack multicamera set up, (7) for a classical chamber music concert. The concert is about 90 minutes. We are trying to obtain at least 95 minutes continue recording with the GH2 hack later version. The recordings have a random stop- interruptions before the card even reach 11, 30, 44 minutes.
Setting: Movie Mode/1080/24/ 44MB/s. We have reformatted the card, used a different ones.The cards are Sandisk 32MB 95/s. Any help would be appreciated
Agreed, but exposure aside..
Proaudio4, thanks for sharing the link. To be fair, the location matches the shots but not the time, and to be a little more close to comparison, in day light a 10 to 15 minute apart can make a difference in contrast ratio. But it is still helpful to see them.. Thanks once again.
It's not mine, but I saw this a while back.
DrDave, thanks for the extra info. It is very appreciated.
@willianeleman Thanks for the info. I have had good luck then on strings. This is what I would say about the dynamic range--with the Canon to me the shadows look like shadows, the look really alive and detailed, and, well, shadowy. But I never find in one complex scene that either cam is better. The GH2 is super detailed. The Canon will not go that extra layer of sharpness. I find in video of very old instruments that the Canon color seems to bring out something in the color and grain of the old wood. So it looks more rich and antique. But that could be a grading issue. In the studio, with controlled lighting and a big monitor I almost always prefer the GH2. From a bit farther away I appreciate the Canon camcorderiness.
mmband, Thank for your reply and suggestions. You have an interesting point of view on diversity - working with different cameras: The diversity of the products doesn't come from the technology I use, but from the way I tell the story, and for this to happen I always try to make sure that technology doesn't get in the way of the story.
I have a concern about working with diverse cameras within a project. I want to avoid disaster in postproduction. If I were to experiment with diversity, it would not be within a project but by project: one project would be GH2, another one D5, Alexa, etc.
For more information about working with multicamera, I would highly recommend the book by Mitch Jacobson, "Mastering MultiCamera Techniques" From Preproduction to Editing and Deliverable.
Have a great day
Мost importantly all the camera set to the same maximum settings especially the white balance,and never use the automatic white balance setting.Аbout the contrast and color no problem using Final Cut Pro X 10.3 and Motiоn 5.02 Рeview these programs and you will find a super quick editing of your video using a variety of video cameras. I also believe that by working with one kind of camera you lose the diversity of your products
JDN, Thank for your suggestions.
We have Clipwarp and 5D to RGB. As for sound: we are using the in house multi channel sound engineering as a double system. The mix is going to be done in post and from there synchronized to the video. The audio from the cameras is going to be used only for sync references. Since we are synchronizing with PluralEye, we do not have the need for TC. Timecode is only add for the delivery formats
@willianaleman if you use clipwrap to remove the ACVHD wrapper it will also merge your spanned files for your so you're not dealing with a bunch of 6-8 minute clips in FCP. We do offline/online for max quality, treating the unwrapped .movs as the masters, adding reel and TC with QTchange, importing only as proxies and then bringing in at 1:1 once cut is locked.
But I'm anal.
Out of curiousity, what is your sound setup? Are you micing the room as backup in case something goes wrong off the board? Or is there a board?
DrDaveCamera, about the workflow:
About the XA10/Vixia HF G10.
1 - For the footage I have seen, there is too much block side effect apparently caused by the compression level and/or small sensor. 2 -The quality of DOF in the G10 is too blurry to be used as a master shot. Probably it would be good enough for close up. I have seen the same stairstepping in the G10 in string instruments, specially in long shots. One workaround would be not to pan nor tilt hard on the string instruments to diminish this side effect, common to most of the pro video entry level. But this is when technology gets in the middle of the creative process values.
For your posting, in your opinion, does the G10 support a higher dynamic range than the GH2 hacked version?
Balazer, even when we already did the crossgrade from FCP 7 to AVID Media Composer, we are not ready to edit in AVID yet. Then, we are going to be editing in FCP 7.
@willianaleman Look forward to seeing your video! Would love to see a screen grab ofthe harp strings if you have one. I compared quite a few cams, including the hacked GH2 at high bitrate, and although I prefer my GH2 for most things, for musical instruments, cellos, violins, harps, double basses, pianos, and so on, the parabolic curves of the surfaces as well as long thin strings produced too much stairstepping for me. The other issue was concert lighting, which produces a huge dynamic range. I simply could not stuff all that range into the GH2. That why I mixed the two together, sharpness on the closeups with DOF from the GH2. But grading was a terrible pain to match the cams so maybe you are better off. Interested in what you did not like about the Canon XA10/Vixia HF G10, and I hope you detail your workflow from the GH2s so I can pick up a few tricks :) PS those screen grabs look great!
Sony Vegas Pro has a nice multi-camera editing mode. I'm sure other products do as well. It's a huge time saver.
DancingCamera, I would like to hear, if possible, about your editing workflow with this type of formats and multi-camera models.
Have a great day.
DancingCamera, thanks for the references. Those shots look amazing. How do you compare the sharpness between the AF100 and GH2? I haven't seen yet convinced footage online of the AF100 that would motivate us to take it as part of our considerations. Looking forward to see the edited version of these shots.
I really appreciate your reply.
@willianaleman Here are a few screen grabs from a shoot we did last night. Ungraded shots directly from VLC. We needed a 1.5hrs on each side of intermission, so we went with @Driftwood's "spanmybitchup." Shot 24L at 80% I think it holds up pretty well considering it was 1h41min per Sandisk 64gb 95mb/s on the "L" setting (and actually would have shot quite a bit longer than listed). The full shoot was done with several other cameras (AF100s, HPX170), but figured to try the GH2 for close follow-shots, and we were all surprised (again) at how well this patch held up.
Thanks again to Nick and as always, Vitaliy.
DrDave, thanks for your reply. To make postproduction easy, the first step we are taking care in preproduction is not to mix camera models or formats. And the quality of the footage obtained from the Canon G10 are under our expectations, specially the compression level. We are studying the cinematic limitations of the camera in preproduction. We have three more weeks before the shooting.
Thanks once again for your reply.
I recommend adding to the GH2s at least two, and preferably four, Canon Vixia G10 or better for this kind of shoot. The GH2 will give you great DOF for closeups, but the G10 is far better for typical musical instrument issues. Better color, far better dynamic range, and the huge problem of stair stepping that occurs on curved surfaces and long strings. You will also get the benefit of instant object tracking, with or without exposure, and a host of other handy features. The Canon sensor is excellent at resolving moving strings as well, which is one of the hardest things to video. The excellent sharpness of the GH2 will complement the Canons both on closeups and distant stage shots. Look forward to seeing the results.
duarttix,
Yes, this is what I meant.
duartix,
The ISO is 640.
Cake v1.1 did span, but it doesn't interrupt the recording process as the other patches we used, which was our biggest frustration during the tests.
We just did another test with a SanDisk Extreme, 30MB/s class 10. It gave us 1 hour 42 minutes.
Thanks for asking.
@willianaleman, thanks for your feedback. I'm pleased that you found success.
Using Cake, did you encounter any recording failures or failures to span the 4-GB limit? Were you satisfied with the video quality?
@duartix, I believe that he meant that Cake spanned, and that the others did not.
Cheers @willianaleman .
I'm glad we could all be of service, I'm just a bit lost on your post, did Cake 1.1 span or not?
What ISO are you using?
The conclusion: Thanks to all who replied and those whose suggestions we have followed in one way or another. To make a long story short, here is our experience. Without the information we have found here we wouldn't have been able to find success in our quest with the GH2 hack version.
Gaol: to make the GH2 hack version record continuously for the longest period of time possible. Our goal was 95 minutes. In multiple repeated test, at the moment, we have an average of 2 hours with a Sandisk Extreme Pro 32 GB/95MB/s, Class 10.
Working patch: balazer's Cake V.1.1 (GH2 constant quality variable bit rate encoder settings) It seems to be that the constant quality "only" setting is not able to record for a long time recording period. It produces an unstable spanning, causing rounding interruption around the 4GB, FAT32 limit storage-time recording of this format. Thanks to duartix for suggesting this patch and explaining the reason why to use it, and of course, to balazer for the work.
Unsuccessfully we tried more than 5 patches versions, followed by tweaking almost all the parameters, from Rec Quality, Bitrate FPS using Ptools3/Patch GH_V10. bin. We worked in Mac and PC. The application that works fluently in Mac for us is Crossover Game in OS 10.5.8
We didn't know what wasn't working in our case: the card speed, card brand, the patch, camera model, etc. By process of elimination, we reset the GH2 to factory preset. Surprisingly, we have almost 2 hours of continuous recording, but with its limit of 24MB/s. Then, we knew that the variable in the test were the patches, and not any of the other factors listed above. At that moment we gave a try to duartix's suggestion about the quality variable bit rate encoder settings patch by balazar.
Here is the direct link to balazer's Cake patch. http://personal-view.com/talks/discussion/2123/gh2-constant-quality-variable-bit-rate-encoder-settings/p1
End of the story. We are open to questions.
Have a great day.
Now from the technical and cinematic point of view, we are going to start testing the other known rumored limitations of the DSLR camera.
we can get long time recording with new driftwoods... 64gb 95m/s sandisk with quantum v9b http://web.forret.com/tools/filesize.asp?speed=154&unit=Mbps&dur=3200 with quantum 100 http://web.forret.com/tools/filesize.asp?speed=100&unit=Mbps&dur=5000
@JDN, you can read what I said about ISO noise levels in these topics:
http://personal-view.com/talks/discussion/1280/gh2-iso-noise-bug-at-320-640-and-1250
It looks like you're new here. If you want to get involved, click one of these buttons!