View Profile
View Forum Posts
Private Message
Hello. I use Sony Vegas 12.0. An annoying problem happens while I edit videos. When I work with a considerable amount of parts of videos, they start to glitch in some frames. It's hard to explain, so I will give you an example of what happened many times. I am trying to edit two videos.
I'm splitting each video in many parts in order to merge certain scenes of them in just one video (at the same time, I am inserting a song in the background). The more I work with those videos, more glitches I get. I've been extremely careful cropping the exact frames to make a good video, however, when I re-watch them, I see that at some points, for less than 1 second, the video shows an other scene which was not there. It's not an editing mistake, it just shows up. Sometimes the scene is from an other part of the same cropped video or it is a scene of a part I've already removed from my project.
It's disapointing, because I am trying to do a clean editing and I take so much time to do that. The temporary solution I got is rendering each part I've already "finished". So I import this part back into my project in order to merge them with the other parts I am finishing and rendering. However, it's not an acceptable solution, since that way I can't change effects and other things I've already inserted in that rendered video when I re-watch the whole work to see if it's good.
Since it is harming the quality of my editings, I have some questions for you guys:
1. Is there a name for the problem I've just described?
2: What could be causing that problem? Is it really a glitch of Sony Vegas? Is it a matter of hardware? Is it a matter of codecs?
3: Please, how could I fix it?
4: Seriously, I just want to do good works. So I ask you for recommendations of another programs (or procedures) which can help me to do good editings without that kind of problem. Even if you don't know how to answer any of the other questions, I would like you to recommend a better program than Sony Vegas if you can.
+ Reply to Thread
Results 1 to 20 of 20
-
-
Sorry you're having this edit and combine problem, but what does "videos" mean? The term covers a lot of territory. Are they similar video types? Are they already encoded to lossy formats? Are they the same frame size and frame rate? Why are you cropping? Etc., etc., etc. More info, please.
- My sister Ann's brother -
Sounds like you're trying to make the system work too hard.
What type of footage are you using? I would bet it's some sort of h.264 long-gop video, possibly variable frame rate as from a phone. And if your system specs are correct, i5 with 4GB RAM, you're likely underpowered.
There are two basic workflows to try.
First is to work in smaller chunks. Which is basically what you're doing -- rendering sections as you go. You should be able to simply render in place without having to reimport and replace the material.
The next step would be to go back and make tweaks as needed, so you never need the entire timeline unrendered at once. You can also literally break your larger show into smaller timelines, render each sub-section on its own so all you have to make adjustments to are the transitions between sections when you rejoin.
The second would be a proxy workflow. Make lower resolution i-frame versions of your footage, edit and then relink to the full resolution material before output.
And of course you can mix and match techniques as well.
Vegas is an excellent NLE, but any excellent NLE, even with high end system specs, will choke on effects-heavy, long-gop, vbr, long duration timelines.Last edited by smrpix; 27th Aug 2015 at 07:09.
-
another option would be to convert all your videos to a lossless yuv format before editing. the files sizes would be increased greatly but they would all edit much easier than what i suspect is long gop format mp4 you are using now. there are lots of codecs to choose from but some of the popular free ones are UT, lagarith, and huffYUV.
--
"a lot of people are better dead" - prisoner KSC2-303 -
Difficult to say exactly what to do. But the owner's PC specs say a single 500GB hard drive, which would go pretty slow with lossless and the drive would soon fill up. We don't know what formats the owner is using and I'm afraid he doesn't know how to tell us. So smrpix has the best advice: work in smaller segments, then join at the end. That's a lot of lossy re-encoding, but without more information that's the best we can offer.
Last edited by LMotlow; 27th Aug 2015 at 10:59.
- My sister Ann's brother -
Thanks for the answers. I'm newbie, but I've been working with Sony Vegas for some time.
Info about the first video: AVC .avi, 23.976 fps, 720:544 (4:3), 920 kbps, Time: 23min38s
Info about the second video: MPEG-4 Visual (XviD) .avi, 29.970 fps, 1280:720 (16:9), 9175 kbps, Time: 1min38s
Information from MediaInfo. The most problematic video is the first one.
I'm cropping because I just need about 1 min from the first video. I'm trying to combine scenes from both videos to make one single video. The resulting video will have about 1min30s (probably less than that). The only effects I'm doing is re-sizing the second video to 720:544 and inserting a flash trasition for trasitions between certain scenes. So I suppose it's not a complex editing...
I'm using a HDD, but in both my internal and external HDs, I don't have much space so it's not an option to convert them to lossless formats... I don't mind converting to any other lossy formats, I just don't want resulting low-quality videos. By the way, if I don't recall bad, I've already tried to convert the first AVC .avi to XviD .avi when I worked with it last year. The same problem happened.
Since I've got that described problem for some other times, I would like at least to know what's really causing it...
Do you need any other information?Last edited by Clever Sleazoid; 27th Aug 2015 at 21:05.
-
That's not much from mediaInfo, which has a lot more info that you've posted (did you bother to notice the little message near the bottom of MediaInfo's abbreviated front window?). But it's enough to let us know that it's neither the first video that's the problem, nor the second video that's the problem. The problem is joining two very dissimilar types of video into the same clip.Originally Posted by Clever Sleazoid
Seriously, If you "don't mind other lossy formats", you'd best resign yourself to throwing concerns about quality out the window. Lossy formats were not designed for what you're trying to do, and Vegas won't make them look any better.Originally Posted by Clever Sleazoid
Yes. Open MediaInfo with those videos again , then on MediaInfo's top menu bar click on "View", and from that submenu choose the "Text" view. When you see the text page for each video, notice that you can select all of the text with your mouse, copy it to the clipboard, and paste that text report for each video right here in the forum as plain text. If that seems to complicated, you can also export mediaInfo's detailed "Views" as a plain text file that you can also post here.Originally Posted by Clever Sleazoid
From your brief description we don't know if you're dealing with interlaced video, but likely video 1 is progressive low-bitrate, low-quality material encoded with huge GOPs and is from a movie. It is totally incompatible with the second video, whether you resize the second video or not. Vegas users will be able to advise you on how to rework in lossy formats to make those videos more alike. The simple answer is, even if you could talk Vegas into making those two segments more alike, you won't have much to show for all the work you're making Vegas do.- My sister Ann's brother -
Most likely it has to do with temporal compression as mentioned earlier (long GOP). What you describe is fairly common - It especially affects AVI container with videos that use b-frame compression. AVI container does not handle temporal compression very well. In fact, if you just remux it to an MP4 (not re-encode, just remux) , often the problem goes away. When you converted it last year, you probably used the same temporal compression which causes the problem in the first placeOriginally Posted by Clever Sleazoid
You can use lossy compression, but it needs to be I-frame only (intra compression). Every frame is encoded separately (complete) and does not depend on information or references from other frames - so you never get mixed up frames even with AVI container. This also means you need higher bitrates and larger filesizes for a certain level of quality. You decide how much quality you're willing to sacrifice -
No, it's worse than that. Anyone notice that the O.P. is starting with 720x544 (just slightly more narrow than 4:3) and 1280x720 (16:9), and resizing the 16:9 video down to 4:3 to match the much shorter low-quality clip?
- My sister Ann's brother -
These might be issues to consider - It's not clear how he's organized the project - but that has nothing to do with the problem he asked about (frame mixups, glitching)Originally Posted by LMotlow
-
Yep. One might consider that. Maybe. Sorry to divert the thread with a quality issue, but the O.P. brought it up first. I'll remember to ignore that in this thread.
Last edited by LMotlow; 27th Aug 2015 at 22:46.
- My sister Ann's brother -
Here you are
General
Complete name : I:\Video 1.avi
Format : AVI
Format/Info : Audio Video Interleave
File size : 224 MiB
Duration : 23mn 37s
Overall bit rate : 1 323 Kbps
Writing application : VirtualDubMod 1.5.4.1 (build 2066/release)
Writing library : VirtualDubMod build 2066/release
Video
ID : 0
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L5.1
Format settings, CABAC : Yes
Format settings, ReFrames : 16 frames
Codec ID : h264
Duration : 23mn 37s
Bit rate : 920 Kbps
Width : 720 pixels
Height : 544 pixels
Display aspect ratio : 4:3
Frame rate : 23.976 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.098
Stream size : 155 MiB (70%)
Writing library : x264 core 55 svn-662
Encoding settings : cabac=1 / ref=16 / deblock=1:2:2 / analyse=0x3:0x133 / me=umh / subme=6 / brdo=1 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=2 / 8x8dct=1 / cqm=0 / deadzone=21,11 / chroma_qp_offset=0 / threads=1 / nr=0 / decimate=1 / mbaff=0 / bframes=2 / b_pyramid=1 / b_adapt=1 / b_bias=0 / direct=2 / wpredb=1 / bime=1 / keyint=300 / keyint_min=25 / scenecut=30 / rc=2pass / bitrate=920 / ratetol=1.0 / rceq='blurCplx^(1-qComp)' / qcomp=0.60 / qpmin=10 / qpmax=51 / qpstep=4 / cplxblur=20.0 / qblur=0.5 / ip_ratio=1.20 / pb_ratio=1.30
Audio #1
ID : 1
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 3
Mode : Joint stereo
Mode extension : MS Stereo
Codec ID : 55
Codec ID/Hint : MP3
Duration : 23mn 37s
Bit rate mode : Constant
Bit rate : 128 Kbps
Channel(s) : 2 channels
Sampling rate : 48.0 KHz
Compression mode : Lossy
Stream size : 21.6 MiB (10%)
Alignment : Split accross interleaves
Interleave, duration : 42 ms (1.00 video frame)
Interleave, preload duration : 500 ms
Writing library : LAME3.97
Audio #2
ID : 2
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 3
Mode : Joint stereo
Mode extension : MS Stereo
Codec ID : 55
Codec ID/Hint : MP3
Duration : 23mn 37s
Bit rate mode : Constant
Bit rate : 128 Kbps
Channel(s) : 2 channels
Sampling rate : 48.0 KHz
Compression mode : Lossy
Stream size : 21.6 MiB (10%)
Alignment : Aligned on interleaves
Interleave, duration : 42 ms (1.00 video frame)
Interleave, preload duration : 504 ms
Audio #3
ID : 3
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 3
Mode : Joint stereo
Mode extension : MS Stereo
Codec ID : 55
Codec ID/Hint : MP3
Duration : 23mn 37s
Bit rate mode : Constant
Bit rate : 128 Kbps
Channel(s) : 2 channels
Sampling rate : 48.0 KHz
Compression mode : Lossy
Stream size : 21.6 MiB (10%)
Alignment : Split accross interleaves
Interleave, duration : 42 ms (1.00 video frame)
Interleave, preload duration : 500 ms
Writing library : LAME3.97I must admit, most of those info sound like an unknown language for me. Some things you guys are saying sound like that too. Specifically, I have no idea of what long GOP or i-frames/b-frames mean. But you're helping a lot, I really thank you.General
Complete name : J:\Video 2.avi
Format : AVI
Format/Info : Audio Video Interleave
File size : 109 MiB
Duration : 1mn 38s
Overall bit rate : 9 316 Kbps
Writing application : Lavf55.34.101
Video
ID : 0
Format : MPEG-4 Visual
Format profile : Simple@L1
Format settings, BVOP : No
Format settings, QPel : No
Format settings, GMC : No warppoints
Format settings, Matrix : Default (H.263)
Codec ID : XVID
Codec ID/Hint : XviD
Duration : 1mn 38s
Bit rate : 9 175 Kbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16:9
Frame rate : 29.970 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.332
Stream size : 108 MiB (98%)
Writing library : Lavc55.52.102
Audio
ID : 1
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 3
Mode : Joint stereo
Mode extension : MS Stereo
Codec ID : 55
Codec ID/Hint : MP3
Duration : 1mn 38s
Bit rate mode : Constant
Bit rate : 128 Kbps
Channel(s) : 2 channels
Sampling rate : 44.1 KHz
Compression mode : Lossy
Stream size : 1.50 MiB (1%)
Alignment : Aligned on interleaves
Interleave, duration : 26 ms (0.78 video frame)
Interleave, preload duration : 26 ms
Writing library : LAME3.99.5
I'm glad you guys got right what the problem looks like, so I think I don't need to upload the video to show what's happening.
So if I got it right, just by using so different types of videos, I'm kind of overloading Sony Vegas, right?
Well, maybe for you a 720:544 video could not sound so good in quality, but it's really pretty good, so it is even very beautiful on my 32'' LCD TV. Well, you have to take into consideration it's a cartoon as well. The second video is not a cartoon. I'm re-sizing it to 720x544, so I forgot to say it gets just part of the screen (it was intended, by the way, since I don't want to show the whole screen, anyways; I'm using the option "Event Pan/Crop"). About the matter of quality, I just wouldn't like to see it to become like that: https://www.youtube.com/watch?v=YROHSAyeO78
The first video is so much better than that. I would like the resulting video to stay at least so beautiful as the 1st video was in the beginning. It doesn't matter if it's not HD. I would not like an incredibly large file, anyway, so I thought it was better re-sizing the second one to 720x544 than doing the opposite (which wouldn't get the first video any better).
So let me see, again, if I got it right: do I just need to remux both videos to a mp4 container? Or will it be necessary to convert them to I-frame only (whatever it means)? If the second one is the solution, how do I do that?
Plus: what should I have done before trying to use Vegas to merge so different videos in only one video?
Thank you, you guys are being very helpful. I hope I'll be able to improve my editings, since the problem seems to be much simpler than I thought.Last edited by Clever Sleazoid; 28th Aug 2015 at 20:53.
-
You can think of long GOP as frames not being complete by themselves - they rely on information from adjacent frames. They have P (predicted) and B (bi predicted) frames. Linear playback is usually not a problem, but frames can get mixed up when you do non linear seeking (e.g. editing, scrubbing around on the timeline). It's especially bad with AVI and b-frames, because AVI container wasn't designed with temporal compression in mind. It uses 1 frame in, 1 frame out. Not only can you get misorded frames when using AVI - you can get black or green frames and sync issues - those sorts of problems are common in vegas and other NLE's with these types of files. I-frame only encoding (intra encoding) just means each frame is independently encoded and complete, without relying on information from adjacent frames . I-frame is less efficient because it doesn't "borrow" from similarities from adjacent frames. That's why the filesizes will be larger for similar quality
Remuxing probably won't work on that one, because it uses 16 reference frames, which is the max for AVC. It's basically very compressed and not meant to be edited so vegas might still have problems with it. But you can try remuxing it if you want (no quality loss, no big filesize change, very fast). You might use mymp4boxgui, or yamb beta 2 .
The safest and most stable thing to do is re-encode it to an I-frame intermediate. You should use a lossless codec as best practices, but if you don't have the HDD space then something lossy but I-frame will work. Cineform works very well in vegas, but even at the lowest quality setting will be a few times larger than your sources. That would be my first choice for lossy/ near-lossless I-frame codec. I've seen sometimes xvid or AVC using x264 have other additional, different problems (not related to your specific issue or I-frame vs. long GOP) in vegas and NLE's. So I would avoid using those and stick with formats known to be very stable in vegas. You could use virtualdub to encode using cineform, but you need to install the free go pro cineform studio first. If you still want to use xvid, you could use vdub too and encode with "maximum I-frame interval" set to 1 (default is 300) - that will give you I-frame only encode. If you are using x264vfw, in the extra commandline box enter --keint 1
The aspect ratio topic is important , but I would start another thread for that if you still need help. The reason is someone searching for help on the original topic (glitches, misordered frames, etc..) will not want to sort through a big discussion on aspect ratios and setting up project settings before finding the relevant info -
IIWY, I'd:
1. Get a larger (fast) HDD (preferably internal, but if it's fast enough, external ought to work also). Seriously, spend the money, sorry.
2. Convert to lossless or nearly-lossless aka slightly-lossy I-frame format. The 1st clip is using >120:1 compression ratio, the 2nd is >36:1 (not surprising: AVC is usually much more efficient than Xvid). If you go lossless, it is ~2.5:1 (so a growth in filesize of ~48x and ~14.5x respectively) or go the slightly lossy route of ~8 or 10:1 (for a growth of ~12-15x and ~3.5-4.5x respectively). They'll be way bigger, but still actually easier to edit. Beware going totally uncompressed, though, because then you run the risk of taxing the HDD/buss subsystem speed limits.
3. You also need to address the issue of mixing of framerates. Is it possible to use IVTC on the 29.97, or (alternately and less optimally) TC on the 23.976 footage? They should both match, and it's MUCH better to get it fixed before doing the edit rather than have Vegas make some non-optimal decision on export. Kill 2 birds with one stone and do the framerate conversion when you do the frametype conversion.
4. Keep it in that intermediate format throughout until all the edit/composite/titling/FX work is done (and you have an "edit master" clip). Then SEPARATELY do a single, final conversion to the format you need for output.
5. Last, but certainly NOT least (and maybe it shouldn't even be last): do some more learning about modern digital video. Compression/codecs & "efficiency"; compression ratios & bitrates - both CBR & VBR & CQ/CRF; GOPs & Frametypes - both I, P and B; Framerates & TC/IVTC - both CFR & VFR; and Aspect Ratios - both Pixel/Sample and Display/Frame. Understanding of these are ALL important fundamentals in technically getting and maintaining good images these days.
Scott -
Thanks for all the explanation. I'll try that. Since I don't have much space in my HD, I'll just render the part I want to work with and use an I-frame encoding, so I won't have to work with a gigantic file.Originally Posted by poisondeathray
Hello, thanks for the answer.Originally Posted by Cornucopia
1. I don't have any money to spend now. Buuuuuuuut what speed is "fast enough"?
2. As I said to poisondeathray, I'll do that, but I'll work with just one part of the video.
3. No idea about what you mean with that, bro...
4. Yeah, I'll do it. When you say "separately" you mean, when I'll have finished the editing on Sony Vegas, right?
5. Do you have any stuff to recommend for newbies?
Thank you, guys, you have been so helpful. I'll tell you if I get any problem doing the stuff you said. Maybe I'll make another thread, depending on the case -
1. Fast enough = figure out what bitrate your files need for the quality you're going for, then adjust for multiple tracks, overhead & inefficiency, simultaneous I + O. My guess is you'd want something capable of sustaining read & write speeds at least 4x the bitrate.
2. OK
3. What I mean: unless you want VFR (variable framerate) video - IMO a thoroughly BAD idea all around - you have to MATCH the framerates somehow. Either making the 23.976->29.97 or making the 29.97->23.976. IIWY, I'd apply pulldown to make the 23.976 -> 29.97, as most NTSC users are already accustomed to it and the reverse would be quite lossy & noticeable.
4. Yes, that's what I mean.
5. Plenty of stuff (too much to go into now, though). Let us know as your project(s) and needs progress, and we'll try to address them.
Scott -
Well, thank you for all the help. I've managed to do the editing. Since it can help someone someday, I will tell you what I did.
Well, I couldn't convert anything to CineForm. For real, I don't really know if I could, since all the files I converted to CineForm couldn't be read. VirtualDub couldn't do anything with the first video (although it could convert the second one, it couldn't even manage to trim the first one at all).
Then, that's what I did: I used Sony Vegas to trim the first video, so I could work with the only part I needed. I rendered it as mjpeg, since it's I-frame. It was the only codec I could use, since all the other ones got some kind of problem. That process also converted the video to 29.970 fps 720x480, so it was useful. I got a 339 mb file for 6min10s. Too big, but accetable for my HD. Audio got a bit unsynchronized, but it was not a big issue, since I would edit it as well.
Even converting it to mjpeg, some frames were still messed up. So I got some wrong frames at the wrong parts. However, I think it must be normal, since the first file had a really long GOP, right? I could remove those wrong frames, and they didn't come back, as it was happening in the beginning, when a lot of new wrong frames popped out of nowhere. So yeah, I could work well with the file, as I wanted. There was no need to convert the second video, I could work well with it. Maybe because it was short.
The funny part is that when I finished my editing, after having all that work, I got a video with a really bad audio. I rendered it as Lame MP3 128kbits 44100 Hz. When I played the video, audio seemed to change, sometimes it sounded too close or too far. I couldn't get what was causing that, but rendering the audio as the same format, just changing it to 48000 Hz, the audio became good. -
Sounds like lots of corruption there, due IMO mainly by using the wrong tools on iffy material.
When you start with material with issues, you cannot use your standard tools - you must process & upconvert to standard formats. THEN edit. And when doing so, since your material had issues already, you shouldn't expect anything less than remaining working with (ONLY) digital intermediate formats until final conversion after editing, compositing & processing AND RENDERING is done. This means BIG HDDs (hence the recommendation).
Not many tools work properly with VFR. About the only ones I would trust are AVISynth and ffmpeg, and a few single/special-purpose tools.
Similarly, if you want to do optimal framerate conversion, I'd primarily go with AVISynth.
Scott -
Not a good way of doing it - you "main" video will have duplicates introduced by vegas every 5th frame when you converted it to 29.97. Jerky playback. You should have matched the project settings .The secondary video is 29.97, but if it's from the same or similar series, it's probably telecined (3:2 pulldown) - you should inverse telecine back to the original 23.976
Vegas had the initial issues decoding in the first place, so I would have avoided that when converting to mjpeg => that's probably why they were decoded out of sequence before you even began to edit. But you did notice that the frames "stayed in place" afterwards - that's because mjpeg is I-frame
Maybe some things for you to think about for future projects, unless you want to redo this one properly -
I'll try to use other way next time. I used Vegas just because all the programs I've tried to use to convert it were not working, especially VirtualDub. Vegas was the only one which didn't convert the first video to a black video.
Quote
Similar Threads
-
Effect FX Editing in Sony Vegas
By imrealdumb in forum EditingReplies: 14Last Post: 14th Sep 2013, 19:36 -
Sony Vegas Pro 8.0 won't render videos?
By pixelhurricane in forum EditingReplies: 19Last Post: 25th Jan 2013, 15:47 -
Sony Vegas 11 GPU rendered videos look terrible, help!
By Mitalis in forum EditingReplies: 1Last Post: 28th Sep 2012, 23:34 -
Best settings when editing videos with Vegas 11 Pro to publish on Youtube
By siopilos in forum EditingReplies: 9Last Post: 13th May 2012, 10:57 -
Sony Vegas Editing help
By blecap in forum EditingReplies: 2Last Post: 30th Mar 2011, 02:08