I just received this neat little camera and I'm trying to figure out the particulars of the MTS files that it shoots. I'd like to edit the 3D in 3D Stereotoolbox but when I look at the files they are not separate files or side by side as I expected.

Anyone with 3D editing experience with these types of files who can educate me as to how to break these files into component parts and adjust convergence please chime in and help out. Are they muxed permanently or can they be adjusted?

My understanding is that only Sony Vegas Pro version 10d can edit these files.

Bruce, can you please shoot a very short clip and send it to me for analysis. I don't have one of these cameras yet but I assume they are using MVC, which we don't have a means to demux yet.

The JVC TD1/HMZ1 seems to have the option for MVC or AVC side by side.

I wonder if Cineform Neo3D could deal with the files? This could be worth an inquiry to Cineform, or perhaps post the question on their dvinfo forum. This may be an expensive solution unless you are aready into Cineform processing but their 3D tools work very well.

They normally start with the two separate tracks, convert them to Cineform AVI (or MOV) and then do the mux. I don't know if they can start with a file that is already muxed. I think they can handle MTS files okay but I don't know about muxed MTS.

The TD10 looks very interesting - anxious to hear more about the results.

I don't think any 3rd party software can deal with demux of MVC yet. I just sent David Newman an email and notified him of this thread so maybe he can shed some light on it.

When the consumer divisions of JVC and Sony put MVC into the cameras I don't think they intended anyone to actually edit it, just drop it on a blu-ray or PS3 for 3D playback. Now that the professional divisions have adapted the design it becomes the preferred way to go for full raster left and right.

There are several people that I know about who have this camera and are editing the MVC files using Sony Vegas Professional version D.

Peter Wimmer has a MVC to AVI prototype converter here: http://www.3dtv.at/Downloads/MVCConverter010_en.msi . It will produce two AVI streams from the MVC (MPEG stream or MP4 file) and it will not work directly on MAC, but it is a step to separate the two streams out of the single MVC stream. The AVI streams can be subsequently converted. This converter currently is at beta level. The converter only works under windows. Some additional information from Peter:

"The MVC to AVI Converter converts MPEG Transport Streams (*.mts, *.ssif) or MP4
files containing 3D video data encoded with the H.264 MVC codec to two
uncompressed AVI files, one AVI file for the left and one for the right view.

Click 'File|Add Files' to add an MVC file to the batch list, then click
'File|Start Conversion' to convert all files in the batch list to AVI format.
The AVI files are stored in the same folders as the MVC files. The 'Left' or
'Right' suffix is added to the AVI filenames.

The audio track is only decoded if the ffdshow Audio Decoder is installed."

Sorry guys I didn't mean to abandon this thread - just been caught up in massive production overload. So this turns out to be a bit of a curiosity. I'll post a clip in a day or so Tim and others can take a look at it - on the road right now.

I just received this neat little camera and I'm trying to figure out the particulars of the MTS files that it shoots. I'd like to edit the 3D in 3D Stereotoolbox but when I look at the files they are not separate files or side by side as I expected.

Anyone with 3D editing experience with these types of files who can educate me as to how to break these files into component parts and adjust convergence please chime in and help out. Are they muxed permanently or can they be adjusted?

Sorry guys I didn't mean to abandon this thread - just been caught up in massive production overload. So this turns out to be a bit of a curiosity. I'll post a clip in a day or so Tim and others can take a look at it - on the road right now.

How are you getting the video off of the camera? This camera and the ability to edit its video came up before, Alister Chapman posted about it on Jan 6th of this year and I think there maybe another thread as well. Anyway my question is whether you are recording to the "soldered" flash memory or using a removeable SD card for example. I was hoping that by using the removable media the file format would be more universally editable as compared to downloading from the internal storage.

The recording medium is irrelevant. I'm not sure if the Sony TD10 has an option to record a simple side-by-side in AVCHD, but if it doesn't then it seems only Sony Vegas is currently supporting ingest of MVC.
The JVC TD1(consumer) and HM-Z1U(pro version) give you the option of full raster MVC or side by side AVCHD.

Tim is correct that there are limited options for post production work with these new MVC formated 3D cameras. We are now in discussion with three vendors of H.264 decoders, all with upcoming MVC support. We currently (not shipped) have support for progressive MVC sources only, yet the first Sony cameras are interlaced. I understand that the Sony camera software does support splitting the MVC stream into two standard AVCHD streams, which can than be loaded in Neo (via FirstLight) for batch conversion in a CineForm 3D file -- still that is a kludge. We believe will have full MVC support for M2TS and MP4 wrapped media in the August-September timeframe based on the reports we get from the codec vendors. If a decoder component is available sooner, we make our upgrades very quickly, so fingers cross.

New link for MVC splitter under windows: 3dtv.at - MVC to AVI Converter Editions & Prices (http://www.3dtv.at/Products/MvcConverter/Editions_en.aspx)

We are using the Sony TD10 as a "D" camera on a 3D shoot now. I'll get Tim a clip to analyze and I'm going to give the software that Pavel suggests a try also.

Will post results as soon as I can.

Has anyone tried Stereosplicer?

StereoSplicer (http://web.me.com/ijunji/Challenge!_REAL_3D_and_Mac_E/StereoSplicer.html)

In working with the Fujifilm W3 3D camera I used this to separate the streams into individual AVI's and then converted to a format FCP could work with. The W3 had MPO and AVI 3D files. Might be worth messing around with.

Can anyone post a short clip from there Sony TD10 with native folder structure. Just has to be 5 or 10 seconds. Enough to try and over come the challenges.

My Sony TD10 has shipped from Sony and should be here any day now.

This is just like when HDV was first introduced.

We used the Sony TD10 as a "D" camera on the Indy 500 3D shoot, so I have some MTS files from it that I'm trying to "break open" to get discreet L and R files for NLE editing.

So far I've tried the following;
PMB software - comes with the camera and has a hard time recognizing footage even when it sees the camera plugged into the computer. Seemingly worthless, I am unable to determine what if anything this POS software does.

MVC to AVI converter - doesn't seem to know what to do about this footage. Says it's processing it, then gives error message: "Main Concept Adobe2 H264/avc video encoder is not compatible with the video format" I've installed FFSHOW etc, but I guess I still don't have the right codec installed to make this work.

Sony Content Management Utility - downloadable from the Sony site, it does pull in the footage, but doesn't seem to want to do much of anything with it.

Sony Vegas 10.0D - Finally I see two video tracks on a timeline with the audio in between. Not knowing much about how to use this program (FCP user that I am) I can't see how to export these tracks separately into L and R files as it appears I should be able to.

Anyone with some good knowledge of how to do this? It seems like it might be the final step towards generating file streams that can be converted to be used in another NLE like FCP.

Suggestions welcomed. I can post a short clip for analysis if someone instructs me the correct way to do so on the forum.

As far as the 3dtv.at S/W goes, people definitely split files with it from that camera. Perhaps email Peter Wimmer. He has been on this forum several times, but he is mostly on the yahoo group: 3dtv : 3d stereoscopic amateur videos (http://movies.groups.yahoo.com/group/3dtv/). You can ask there. There were some updates lately.

Someone posted a procedure for the file splitting for the Vegas 10.0d:

Sony HDR-TD10E
3D STEREOSCOPIC VIDEO SPLIT LEFT-RIGHT MVC FILES
1 - In Vegas 10.0d, configure the project as "LEFT ONLY" (in 3D mode)
2 - Take a file imported from the TD10E with Sony PMB
3 - Render .....
4 - Reconfigure the project as "RIGHT ONLY"
5 - Take the same file and replace it in its track
6 - Render as step 3 and save it also in the desired folder, renamed-R

I went from Vegas into Final Cut and Tim's filter many times. I normally render into ffdshow MJpeg at the desired quality, and move it into my Mac. It usually imports into Final Cut (AIC) quite easily, even though it is time consuming.

This is a hot topic on many forums. Unfortunately Apple is not very interested in BluRay or in 3D, and the people supporting these media and formats don't port their S/W onto Mac.

Thanks for all that info Pavel. I'll give your suggestions a try even though I'm not that familiar with the Vegas program.

i own a TD10E and found the camera to work very well.
The only problem is the way you work and what you want to get.
If you are looking for professional output, there is no solution for instance out of creating 3D MVC blu-ray.
Unfortunately the sony is 50/60i and this is a pain since 3D MVC Blu-ray is 24p only.
If you are an happy owner of a TD10E, you can shoot 50i, deinterlance to 25p and just use it as 24p.
for 60i, there is no solution yet since converting 60i to 24p will create lot of problem that could be irrelevant in 2D but that are really bad in 3D.
If you are just shooting for internal/personal stuff and accept to use some proprietary solution, then there is no problem.For example, you can split the MVC to left/right files, edit with your favorite application, then reassemble the left/right to side by side and burn it to a blu-ray or a usb key.

My TD10 has three options for frame rate; 24P, 60P and 60i at 1080x1920. I have only shot in the 24P mode but have not as yet gotten into a program that will tell me whether it is native 24P (23.98) like Quicktime Inspector. I figure it's 23.98 embedded in a 60i stream - AKA Sony PSF.

I have some files on a timeline in Vegas 10D, but don't know the program well enough yet to determine the frame rate from that program - but I'm working on it.

Bruce, you don't have to know and use much of the Vegas S/W. I know how to use it about intermediately. I thought it would be useful for it's auto-alignment feature, but that isn't perfect, and Vegas kept crashing on me, wasting lot of time and nerves, so I wouldn't recommend it over FC and Stereo3D Toolbox anyway. I would just bring the files in, place them on the timeline one-by-one, output L and R in Mjpeg format, transfer that output onto Mac (unless you are running the Vegas on Mac already), and finish the post with Tim's S/W and FC. You don't have to learn much of Vegas for that. I just don't like it for real editing that much. Sony should stick to H/W. Vegas is in beta stage of development in comparison to Final Cut. Just MHO.

Someone posted a procedure for the file splitting for the Vegas 10.0d:

Sony HDR-TD10E
3D STEREOSCOPIC VIDEO SPLIT LEFT-RIGHT MVC FILES
1 - In Vegas 10.0d, configure the project as "LEFT ONLY" (in 3D mode)
2 - Take a file imported from the TD10E with Sony PMB
3 - Render .....
4 - Reconfigure the project as "RIGHT ONLY"
5 - Take the same file and replace it in its track
6 - Render as step 3 and save it also in the desired folder, renamed-R.

A little extra help here if you don't mind;

1. Step by step on how to configure project as LEFT ONLY (in 3D mode)
2. Why imported from Sony PMB and not directly from AVCHD/STREAM/*.MTS?
3. What are the proper Export settings? I can only see Export to Tape, DV, or PSP ?

Thanks for chiming in.

Sorry Bruce, I don't have Vegas 10d any more. I let the demo version expire, because it was way too unstable. Maybe someone else,or on the Vegas forum can help with step-by-step.

I've successfully extracted a full frame L & R clip from Sony TD10 footage.

I used Sony Vegas 10D version NLE to extract the full resolution clips, and then exported to AVI, MXF, MOV, and MP4

1 - 13 second clip (X2 L+R) =
MP4 - 40MB (x2)
MXF - 87MB (x2)
AVI - 1.35GB (x2)
MOV- 2.70GB (x2)

The MXF which is Sony XDCAM @ 50MBs turned out to be the most efficient and data compact without any resolution sacrifice. AVI and MOV were uncompressed but obscenely large for that short of a clip. MP4 was a Sony XDCAM @ 35MBs which most closely replicated the original AVCHD data rate.

I used Final Cut Pro XDCAM Transfer software (Mac) from Sony to re-wrap the MXF file to MOV and placed both clips in the FCP bin. Dashwood plug in was applied to adjust convergence, and I watched the full screen composite on a Samsung 42" 3D HDTV.

Mission accomplished, TD10 footage can now fit into either an Avid or FCP workflow with this process, although it does span two different OS which makes it kind of kludgy. Hopefully a more direct method will arise soon to avoid this log jam of NLE's.

Now I'm trying to determine if there is a way to batch process in Vegas, but I'm suspecting that there isn't for this type of activity.

Anyone who have a TD10, can upload a very short native MVC file?
I work with Vegas for many years, and with 2 cameras sidebyside,
but i like to upgrade to TD10 or the new HXR-NX3D1.
I like to see the quality of the footage, the wide angle, barrel distorsion,
chromatic aberration, low light sensitivity, etc. before buying.
The 3cm interaxial is not too small? I hope it will be ok also for very near subjects,
because now i have 7cm minimum with my cams, and often this is just
too much for close subjects. And it is true that you can set the parallax
with a knob, but not with toe in the lens, instead the sensors are wider than
fullhd, so i can play with the extra pixels on the sides? That's what i have read.

Can somebody help me? Then i can help in using of Vegas.
regards

Here's a little clip from a recent U/W shoot. Sorry it's not much - actually the only small clip I have and it's a result of an accidental start and stop of the Record function. Hopefully it will help.

I'll try to shoot some better short clips soon.

http://beampath3d.com/pub/Bimini_25_MVC.zip

Thanks, even Vegas Moviestudio can read this file, and use as stereo3D source,
output to MVC, etc..
I only have to set the "'pair with next stream" option in the clip properties dialog.

Here are a few more short clips. At least these have some color range and some motion.


http://beampath3d.com/pub/Paramotor_16_MVC.zip
http://beampath3d.com/pub/Paramotor_17_MVC.zip
http://beampath3d.com/pub/Golf_MVC.zip

Thanks, looks very good to me!
Can this cam shoot real 24p, or just 50/60i?
Stereoscopic player cannot deinterlace the MVC file..

The 3cm base work well with close subjects!
The opposite situation: sunday i filmed an airplane in a show,
with 15cm base, but this was too small. I get little depth.
So now i have to make a 60cm rig :-)

Unfortunately the TD10 is limited to 1080 60i in 3D. This is a MAJOR limitation. I would think it's a deal killer for all but the home movie consumer. Even then, the user will need one of the 3DTV's that support 1080 60i frame-packed mode (e.g. Panasonic or Sony).

Stereoscopic player should de-interlace the MVCs properly - it just doesn't detect the need to do so automatically. Go to "Video Properties" under the "File" menu, select the "Settings" tab and click on the "Interlaced Video" option.

The narrow I.A. is better than the alternative (too wide) that most side-by-side cameras have suffered with since the dawn of time. It will help the consumer user from ripping viewer's eyes out with excessive parallax.

Yes, but 60i is perfect for 120Hz shutter systems.
Who wants to create 24p the european 50i version would be better,
beacuse it can be deinterlace to 25p, and slowdown to 24p.
Yes, Stereoscopic player has some deinterlace, but this simply duplicate one field,
so usless, compared to real Blend or Bob method. In the codec properties (ctrl-v)
i cannot set deinterlacing.

I think Stereoscopic Player's de-interlace is going to use the native de-interlace in your video card. In the case of modern NVIDIA cards, it's motion-adaptive and very good. On ATI cards, the default on-card de-interlace is selectable via the Catalyst control panel. Playback looks very good from Stereoscopic player on my 120Hz LCD page flipper w/ NVIDIA de-interlacing.

The big problem with 60i is that 1080 60i or 30P (the natural rate to de-interlace into) are NOT mandatory HDMI 1.4a 3D formats and therefore are not widely supported - so the "out of the camera" experience for some users will have to be anamorphically scaled and not full res. Also, 3D Blu-Ray is limited to 24P and there is no good way to get there from 1080 60i. You can always scale to 720 30P... but that sucks!

For deinterlacing: would be good, but i read in the text file:

* Deinterlacing support (drop-field deinterlacing). The internal deinter-
lacing is not used if a compatible video decoder is detected. In this
case, deinterlacing is automatically enabled in the decoder and the
internal deinterlacing is disabled.

Drop is bad. You are sure, that you see 60frames, and not 30p?

We're seeing 60i motion-adaptivly de-interlaced into 30P.

CoreAVC MVC decoder (that Peter Wimmer uses in Stereoscopic Player for MVC decoding) is almost certainly using hardware de-interlacing via DirectShow. It's using DXVA2 or CUDA hardware acceleration for decoding. If the DirectShow decoder can't do the de-interlacing, Peter falls back and discards one field for playback. That is certainly NOT happing on any of our page-flipping PC's here.

Are you seeing juddery playback of the MVC files from Stereoscopic Player on a page-flipping display? That is a tell-tail sign of dropping a field.

We've been trying to de-interlace, scale and modify the time-base of the native 1080 60i footage here. We've tried yadif, the motion-adaptive scaler/de-interlacer in Compressor and the optical-flow-analysis based Nuke/Kronos package. Neither offer ANY visual improvement in de-interlacing this footage in real-time with an NVIDIA card. We've tried:

- 1080 60i to 1080 60P
- 1080 60i to 1080 30P
- 1080 60i to 1080 24P (a crap shoot, I know)
- 1080 60i to 720 60P (had high hopes for this one, but no better image than 720 30P)
- 1080 60i to 720 30P

We were hoping to crack the 1080 60i fields in half (to 1920 X 540 60P), then scale to 720 X 1280 by borrowing spatial information from the original 1080 frame. There are apparently broadcast hardware processors (e.g. Teranex vc100) that can do this effectively. We have failed to replicate the process in software so far. All hopes of getting some production use out of the TD10 are pretty much dashed here. Here's hoping the HXR-NX3D1P ships on schedule.

Great, then it's time to upgrade my 7600GT :P

OK - have a decent path to 720 60P working. Used Peter Wimmer's MVC-to-AVI. Used AVISynth to crack 1080 60i into 540x1920 60P (split the fields). Then used Compressor's statistical prediction resize filter to go from 540x1920 to 720x1280. Added a pinch of gamma correction ('cause something along the way stepped on the gamma) and the results are passable 720 60P 3D.

I'm test encoding now to see how much difference additional temporal information makes using this conversion as apposed to using a native 720 60P 3D rig.

If anyone wants the simple AVISynth script and/or Compressor settings, let me know.

Ok, i tested with the latest player which claims:
"Added support for interlaced MVC content. "

I use a geforce 210 (dxva and cuda support) and also
an Ati 5450, but i cannot see any hardware deinterlacing.
Graphics drivers are updated.. So what's wrong?
I see the horizontal lines, or dropped resolution when check the
"interlaced video" option in video properties window.

But if you see 30P from 60i source then it's also not the best, because
the best is 60 individual frame from a 60i source, so very smooth
motion.

"have a decent path to 720 60P working."

Great, and this is also a 3D Bluray standard :-)

I'm a little stumped by the official Sony specs posted at this URL.
Sony : HXR-NX3D1E (HXRNX3D1E) : Technical Specifications : United Kingdom (http://www.sony.co.uk/biz/product/nxcamcorders/hxr-nx3d1e/technicalspecs)
________________________________________
Recording format
Video Format: 3D: 3D : MVC (1080/60i, 50i, 24p : original format), HD: HD : MPEG4-AVC/H.264 AVCHD format compatible (1080/60p,50p : original format), STD: MPEG-2 PS
Audio Format: 3D/HD: Linear PCM/Dolby Digital 2ch, 16bit, 48kHz, STD: Dolby Digital 2ch, 16bit, 48kHz
Recording frame rate
3D: 3D (28Mbps) 1920 x 1080/(60i,50i,24p)/16:9, HD: PS (28Mbps) 1920 x 1080/(60p,50p)/16:9, HD: FX (24Mbps) 1920 x 1080/(60i,50i,25p,24p)/16:9, HD: FH (17Mbps) 1920 x 1080/(60i,50i,25p,24p)/16:9, HD: HQ (9Mbps) 1440 x 1080/(60i,50i)/16:9, HD: LP (5Mbps) 1440 x 1080/(60i,50i)/16:9, STD: SD (9Mbps) 720 x 480/60i or 720 x 576/50i /16:9, 4:3
________________________________________

It sure looks like the camera records 3D at 24P, but the frame properties of recorded footage clearly indicate it is 29.97. Now I'm also looking at de-interlacing solutions for possible conversion to 23.98 but I'm not holding out much hope that it will look any different than Sony Z1U footage did when it would get converted to 23.98 - lots of strobing and other ugly artifacts.

...but i cannot see any hardware deinterlacing.
Graphics drivers are updated.. So what's wrong?

Not sure. Works great here on Geforce 540M and GTX 580. Can you see a difference when you enable the switch for de-interlacing in Stereoscopic Player? Are you viewing on a 120Hz LCD or over HDMI 1.4 to a 3DTV? If you play in 2D from Windows Media Player, does it de-interlace any differently than when played from Stereoscopic Player in 3D?

I'm a little stumped by the official Sony specs posted at this URL.
Sony : HXR-NX3D1E (HXRNX3D1E) : Technical Specifications : United Kingdom (http://www.sony.co.uk/biz/product/nxcamcorders/hxr-nx3d1e/technicalspecs)


Those specs are for the yet-to-be-released Sony HXR-NX3D1, we're discussing the TD10. They are likely the same basic camera with the exception that the NX3D1 supports 24P in 3D and has XLR audio inputs (and costs 2X as much as the TD10).

Those specs are for the yet-to-be-released Sony HXR-NX3D1, we're discussing the TD10. They are likely the same basic camera with the exception that the NX3D1 supports 24P in 3D and has XLR audio inputs (and costs 2X as much as the TD10).

Good point, hadn't made that distinction.

So, what would the best deinterlacing/24P conversion routines to run on TD10 footage to make it the most compatible with other 1080/24P footage like from a Panasonic 3DA1?

Compressor? other workflows? I've been reading your thread posts about converting to 720 but I'm not sure that is the way to go with a 1080 3D project. It sucks to find all this out after the fact, but I'll be looking more closely at the NX3D1 now.

So, what would the best deinterlacing/24P conversion routines.

There is no good path from the TD10s native 1080 60i to 1080 24P.... it's just not really practical. There are optical flow re-timers (Nuke Kronos) that may get you somewhere, depending on the footage composition - but - the results are generally not good.

Going to 720 60P, which is another mandatory HDMI 1.4a 3DTV standard, works pretty well. You can use AVISynth to de-interlace to 1080 60P, then scale to 720. Results are fairly good. For fast action 720 60P IS the way to go in 3D. You need as much temporal resolution as possible to capture action, especially when the motion is parallel to the sensor plane. In 3D, blur = BAD.

It is a real bummer that Sony didn't implement 1080 24P in this camera. Also a bummer than GoPro's are 1080 30P (but, will do 25P in PAL mode at least).

Anyone find a all mac solution for this yet?
Been looking everywhere...
Thanks

Taj, I don't think Sony has made Mac processing of these files a priority - yet. You need to use the Sony Content Management Utility (Win only) to process the M2TS files into discreet L / R files and then you can edit/process them on a Mac in FCP, Avid, AE, etc. I know that the next update of Neo 3D from Cineform will process and mux them directly, but that's an expensive software package at $1K USD. Right now only CMU and Sony Vegas 10D,E or 11 will edit them natively - Windows only.

The main problem with the CMU from Sony is that you need to plug in the camera while you install it, then it runs ok without the camera plugged in after that. Don't know why they did this.

So the best answer for exclusive Mac users is probably to use Boot Camp, install any version of Windows from XP to Win 7 (I would skip Vista unless it's free), then install CMU with a camera (owned or borrowed), then make the files accessible from the Mac boot side to edit them.

Hi! I'm new to 3D, the TD10, and this forum.

I cut on FCP, so I started looking for a solution that could allow me to edit there. I tried Wimmer's MVC to AVI Converter, but there was a big problem! The (beautifully compressed) .m2ts codec used in the TD10 mushroomed in data size, when converted to AVI. It was literally 100X more data; one 100MB file expanding into 2X5GB AVIs. So 10GB of footage would expand to 1TB! And that's all before transcoding the second time, from AVI to MOV.

That's a huge workflow bottleneck, in transcoding time, sheer data space, and edit throughput. How much easier it would be, just to cut natively in the .m2ts codec!

My first work-around was to pull selects from the footage, with a combination of Wimmer's Stereoscopic Viewer, and Sony's PMB (file management software that comes with the camera), which can trim the .m2ts files without changing their format. In theory, then, the selects would expand to smaller AVIs, although it would still take up a lot of space. Also, archiving footage gets a lot easier.

But once I started working in Vegas 11, which can edit the .m2ts files directly, I decided that was the only way to go. As an FCP user, it's a bit frustrating to learn Vegas' very different logic, but ultimately it's worth it, to be able to keep my file sizes small, and my workflow lean.

I have just shot an interview on the TD10, with the GoPro 3D rig as my b-camera. Since Vegas can cut the Cineform codec, I'm hoping I can sync up the two cameras there, without having to transcode anything. I'm not yet sure how best to finish the project (it will include some stereoscopic CGI), but I am definitely sold on Vegas (and hence, PC) as my near-term cutting solution. A learning curve in the short term seems like it will pay off well, over time, at least until FCP or Avid starts cutting .m2ts natively.

Matt, my experience with TD10/NX3D1 files mirrors yours, however, I found Sony Vegas to be a pale comparison to FCP / Cineform Neo3D workflow so I have only been using it to separate the L/R images and converting/muxing them with Cineform MOV to use in FCP. I can see why some would opt to stay in Vegas and edit 3D natively, however I had serious playback issues with M2TS files in Vegas and not in FCP, so I didn't bother learning a new timeline routine. Also the 3D adjustments palate was rudimentary at best with very few options for fine tuning or ghost busting etc.

Vegas does a great job of creating full resolution L/R images from the MVC files and that is good enough for me.

Another note, I stated in an earlier post on this thread that the Sony CMU & PMB programs would work without the camera being directly hooked up to the computer, but this appears to not be the case. I have found though, that if your MVC files are still on an SD card in the camera that it is a much easier process to render full L/R frames out from those programs directly from the camera as they can be queued in a batch for processing saving mucho time. Unfortunately this isn't always the case as data gets transferred and then SD cards get re-used, so I'm working out the possibility of re-recording the MVC / M2TS files back to SD cards to insert in the camera for processing. This is ass-backwards of course but for some reason Sony has not made it possible to utilize the separation of clips from existing files on a hard drive - go figure. The fact that the Sony MVC files Right image is not a full resolution but rather a Delta file embodying only the differences between it and the full resolution Left image makes using the Sony software all the more important for image fidelity.

Cineform tech support tells me that the next update of Neo 3D will include native MVC processing for files from JVC, Panasonic and Sony.

Cineform tech support tells me that the next update of Neo 3D will include native MVC processing for files from JVC, Panasonic and Sony.

Sounds great. Thanks for the heads up. Did they give you an ETA on the Neo3D update?

The usual: ADN

(any day now)

Matt, my experience with TD10/NX3D1 files mirrors yours, however, I found Sony Vegas to be a pale comparison to FCP / Cineform Neo3D workflow so I have only been using it to separate the L/R images and converting/muxing them with Cineform MOV to use in FCP. I can see why some would opt to stay in Vegas and edit 3D natively, however I had serious playback issues with M2TS files in Vegas and not in FCP, so I didn't bother learning a new timeline routine. Also the 3D adjustments palate was rudimentary at best with very few options for fine tuning or ghost busting etc.

Vegas does a great job of creating full resolution L/R images from the MVC files and that is good enough for me.

Another note, I stated in an earlier post on this thread that the Sony CMU & PMB programs would work without the camera being directly hooked up to the computer, but this appears to not be the case. I have found though, that if your MVC files are still on an SD card in the camera that it is a much easier process to render full L/R frames out from those programs directly from the camera as they can be queued in a batch for processing saving mucho time. Unfortunately this isn't always the case as data gets transferred and then SD cards get re-used, so I'm working out the possibility of re-recording the MVC / M2TS files back to SD cards to insert in the camera for processing. This is ass-backwards of course but for some reason Sony has not made it possible to utilize the separation of clips from existing files on a hard drive - go figure. The fact that the Sony MVC files Right image is not a full resolution but rather a Delta file embodying only the differences between it and the full resolution Left image makes using the Sony software all the more important for image fidelity.

Cineform tech support tells me that the next update of Neo 3D will include native MVC processing for files from JVC, Panasonic and Sony.
Thanks for sharing your experience, Bruce. I have not yet had the chance to work with the Cineform Neo on FCP, since I only have the GoPro software. My experience with Vegas, too, has been stuttery playback, at best (even sometimes of 2D!), as compared to Stereoscopic Viewer, which plays back the same files, fine, on the same system. I have been hoping that this was mostly a matter of a slow PC, and that adequate CPU/GPU/RAM would bring Vegas up to snuff, but that' s just theory, so far. I've been cutting professionally on FCP for a decade now, so I'd sure prefer to keep cutting there (I wonder how FCP X fares with .m2ts files?).

As for Vegas' convergence controls, I'm still pretty unsure about them. I first tried to clean up the ghosting using the "horizontal offset" slider, and then whoops, realized I was killing my depth. So far, I've had the best success using the "Auto Correct" function in the "Stereoscopic 3D Adjust" filter (although that also crashes Vegas, a lot). I've seen the Cineform Neo tutorial, and it sure seems more powerful. Good to know that it is.

Regarding copying .m2ts files onto an SD card, and re-inserting that in the TD10, I'd like to know if you have any success with that. The instruction manual with the TD10 suggests that it should be possible, but does not say how. I tried that procedure, in order to be able to show selects to others (and set the camera to view the SD card as its medium), but the camera did not see any of the files.

If indeed Cineform's Neo will (any day now) natively handle MVC, then I'd much prefer that, to stick with FCP. Either way, I'll need to buy a new computer to handle all the throughput, so the main question is to stick with Mac, or move to PC...

All good points Matt.

"I wonder how FCP X fares with .m2ts files?" It seems to handle them very well, however, sticking an MVC / M2TS file into it will only give you 2D. I am impressed with only one thing about FCP X which is it's very robust playback engine. It's really something, but I'm afraid it's the only thing I can recommend it for as it's other lack of funtions cancel everything out for me. I'm waiting to see if Apple makes this once again a professional program like FCP 7. They are losing huge market share every day they delay for sure.

"I have been hoping that this was mostly a matter of a slow PC, and that adequate CPU/GPU/RAM would bring Vegas up to snuff, but that' s just theory, so far" Not so much theory really. I rebuilt my computer to an 8-core i7 with 16GB RAM and still couldn't playback some files without stutter in Vegas. Same files play fine in PPro CS5 on Windows and PPro CS5 & FCP7 on Mac.

On controls like ghost busting and features like 3D text, Tim Dashwood's Stereo Toolbox 3D can't be beat. He has a much more robust toolset than either Neo3D or any NLE standard tools like Vegas or Avid MC 5. Avid MC6 looks pretty promising for 3D though, but expensive.

I'm still experimenting with repositioning the downloaded camera files back onto the SD card and using the camera with CMU or PMB software. I'll post an update when I get more info.

With my i7 2600K overclocked to 4.3 Ghz, the playback of Sony TD10 mvc files is fine. It has been fine for Vegas Pro 10e, it is fine for Vegas Pro 11 too. There is not a lot of playback reserves, but it works.

The Vegas convergene tool work fine - the adjustment of the horizontal difference between the two videoparts adjust in the end of the day the position where the convergence is zero, so what comes out or what stays behind the monitor. It is especially great if you use in the internal preview window a difference picture, where you see in a very quick way where the convergence is zero - and use a nvidia 3D vision with a 120 Hz monitor for the preview of the video (even that still works with a Quadro only). The only range for improvement would be to see how many pixel you shift the video parts (so to show the parallax in a direct way).

I never ever would like to edit MVC file directly in the camcorder or with an sdhc chip.Why not move that to the harddisc? Vegas generates the audio waveform files, and that is not fine in the camcorder.