I've got a quandry that I'd like to run by the forum.
I'm converting videos with audio (primarily 5.1 DTS and AC3) into WMV form. Now, my native encoder -- Microsoft Expression Encoder -- is incredibly finnicky when it comes to DirectShow filters, so a lot of my DTS/AC3 audio formats get a random error that's essentially unfixable when I try converting the audio via the VBR unconstrained route. I've searched for other filters (if anyone knows any non-DirectShow filters to try, please let me know) but no dice. I'm currently using LAVFilters with minor success.
My solution to that error, though, is to 1) convert the DTS/AC3 source to 640kbps AAC and then into VBR unconstrained or 2) convert DTS/AC3 directly via 1-pass CBR.
TL;DR version:
Will I be losing any quality going from DTS/AC3 -> 640kbps AAC -> VBR aside from the drop in bitrate?
Will I be losing any quality going from DTS/AC3 -> 1-pass CBR aside from the drop in bitrate?
Which method (between the above two) do you think would be the better option?
Thanks all.
Try StreamFab Downloader and download from Netflix, Amazon, Youtube! Or Try DVDFab and copy Blu-rays! or rip iTunes movies!
+ Reply to Thread
Results 1 to 30 of 56
Thread
-
-
VBR audio is EVIL. Unless you have a way to go back in time 10+ years, there is no compelling reason today to EVER use it.
Both of your conversions will lose quality. You don't say what format you are converting into, but unless you are converting to FLAC or WAV or an HD audio format like DTS-HD, your conversions will all be lossy. You don't say what your final bit rate is either. I'd suggest that your 2nd method of going directly to 1-pass CBR should be the best. -
Yikes, which is VBR audio evil? Would you say that CBR is the way to go, then?
I'm converting to WMA Professional (might be WMA10?). In the interest of space saving, I've been converting down to a final bitrate of 440. -
If bitrate of source audio is high, then convert AAC, yes, data loss, but cannot recognize your ears. If you want high quality 5+1 audio, than 576 kbps is enough.
-
The DTS bitrate are usually 1500kbps, with AC3 being dependent but usually lower quality (assumed).
576 output you say is good enough for high quality? -
Yes. 576 kbps 5+1 channel = 192 kbps stereo channel audio. If your ears not most most sensitive, you cannot recognize any different.
One question: If your disk space 25 GB, and you have one film, then do you want compress data? No. Because disk space enough. Like Bluray discs.
But, if you encode video, you want good "file size/quality" ratio.Last edited by Hikmet; 23rd May 2013 at 15:54.
-
Seriously --- the best solution starts with dropping Expression Encoder.
If you like the ASF container and the Windows Media formats that much, this is what you should do:
1) create an adequate Avisynth script for the source file;
2) use the Windows Media Encoder engine through Zambelli's wmcmd.vbs <-- clicky. -
I don't understand what your question was. It seemed like a statement. I don't have the money for a large RAID setup to store oodles of high quality, large file size videos, so I have to compress things. And I'm trying to find that ratio, and basically have one that makes me happy right now.
I expected this. Unfortunately, I have no idea how to use Avisynth. I'm not looking to switch encoders, either, because I really like what MEE offers aside from its frustrating refusal to function with DirectShow filters, but I'll give anything a look if you can direct me to something that would teach me how to use it (or teach me yourself). -
I said, bluray standart bitrate is high, but it's not necessary. Bitrate is high, because disk space is not problem. Why we want encode videos? Because we want good quality but small size videos and no restrictions. If we want edit a video, then need uncompressed or high bitrate videos. But we want watch a video, then no necessary uncompressed videos or ultra high bitrate.
Ultra high bitrate values is not necessarily. Calculate proper bitrate and encode videos. If know encode operations and configurations, then one person no catch on differences uncompressed and compressed videos.
My friend think like you. He has 20-25 GB file size videos and he think this is necessary. And 20-25 video full over 1 TB HDD. And I told him this is not necessary and I encode his some videos. He watch this videos on 100+ inch tv and his monitor, then he say, "You say true. This big file size is not necessary."
If we want do video archive, then file size is most important. And I don't want waste my disk space with unnecessary datas. -
So you're preaching to the converted, basically. Your friend is not like me, however. I'm well aware of the quality/size ratio and have done many tests on my own to discern what settings work best for me and my tastes.
I'm still waiting to hear from jman98 about why VBR is so evil. Unless anyone else would like to tackle that statement? -
-
-
I've used nothing but VBR MP3 and AAC over the years and never had a problem. DTS and AC3 and WAV are CBR (at least in my experience) so they don't count.
-
-
So new consensus is that VBR is not necessarily a bad thing?
I like to use VBR for encoding video because it gives me the bitrate when I need it (I go unconstrained, as well) and cuts it down when I don't. Let's me get more bang for the buck with space and, again, done tests myself and found I enjoy it and it suits my needs. -
Hell no.
VBR was necessary 10+ years ago when all people had to burn was CD-Rs and disk drives were expensive. Yes, the world is not going to implode if you use VBR. I find it absolutely fascinating that ONE guy posted what you wanted to hear and now all of a sudden you are completely and utterly vindicated. Every damn post we have here that says "Don't do X" has a guy like manono who pops up to say "I've never had a problem with it".
If you need a higher bit rate, then use a damn higher bit rate instead of using VBR for trickery to get it. Getting a stiffy over VBR is a bit like being super excited because the store next door offered you that HDTV for one dollar cheaper and that's just the biggest deal ever to you. The space savings are minuscule with VBR over CBR and the problems are greater for reasons stated, but if you just cannot live without VBR and somehow being told to consider CBR is a bruise to your ego, then use VBR. -
You dropped your "VBR is evil" line and ran, giving me no explanation. So I asked others to chime in. One person said it confused encoders, one said they've never had a problem, and the other said that programs -- not VBR -- were the issue. So that would be two people for VBR, seemingly, and one against. I'm bound to go with the majority on this. However, not wanting to come to a premature conclusion, I posited "So new consensus is that VBR is not necessarily a bad thing?" I had been operating under the assumption (and researching on my own) that VBR was now not a good option, and was looking for someone to confirm or deny after your denial-and-run. Like I said, I had a new majority saying that no, it was not a bad thing, so I asked a question to double-check that. Every post here was not aimed specifically at my question; most were telling me what I should change rather than telling me which method is better. And that's fine, I expected people to say "don't use X to convert," "don't use Y settings."
Now my knowledge of CBR and VBR is amateur. CBR is constant, so constant quality. VBR is variable, so it varies. Essentially, it means that VBR may give me a smaller file size OR will allow me to use a lower bitrate while still getting some of the positives of a higher bitrate when it is needed. Because I'm constrained when it comes to space -- as I've mentioned -- I can't just increase the bitrate to get better audio. If I had all the space I wanted, I would just NOT encode anything and keep everything at source. Unfortunately I don't have the money to do that, so I have to encode things down. The settings I've arrived at were the product of a lot of tests done by myself and I was happy with the results. Not considering changing those options, you told me to. I asked for a reason. I'm still trying to make up my mind, so my ego isn't bruised at all. You yourself still haven't provided me with your reasoning for why CBR is better. You're mainly pointing out why VBR isn't good, which is something I don't necessarily care to hear. If it's no good, tell me why and then explain which options are better and why. -
Stop trying to pit friend against friend. Use whatever you want. I just said I've never had any problems. I make it and mux. I don't edit or do anything difficult with it. I just make my own videos with it, and I upload videos to YouTube with it. No problems. I don't count anything downloaded that doesn't work right as a strike against VBR encoding as who knows what some idiot did to screw it up.
jman98 is well-known around here for his position and he's certainly not alone. You said you've done your own tests so use your own judgement.
Two for and one against does not a consensus make. Set up a poll if you like. -
You've been told VBR *can* be a problem with some programs. You yourself confirmed it in your first post. So don't do that.
I have to say, jman has a point. VBR does not confer a worthwhile advantage, IMO. And your workaround of encoding twice is bound to degrade the quality anyway. I'd just go with encoding once with CBR and be done with it.Last edited by fritzi93; 24th May 2013 at 22:17.
Pull! Bang! Darn! -
I don't get it. I've used VBR for audio more times than I could count. Many, many times for MP3 encoding and each and every time for AAC and it's never caused me a playback problem. Ever.
Calling VBR encoding for audio "trickery" makes no sense to me. Anyone here use CBR encoding for video?? I doubt it, so why is audio any different? I've read through the thread but not read any specific examples as to the problems VBR audio can cause or why it's evil.
Does anyone consider FLAC to be evil? I only ask because I was under the impression FLAC uses VBR encoding. -
Probably these are not very-good answers, but hey
1) lossy audio is less bitrate-hungry than lossy video;
2) video is not audio with one more "dimension"
Does anyone consider FLAC to be evil? I only ask because I was under the impression FLAC uses VBR encoding.
http://codecs.multimedia.cx/?p=313 -
Surely the principle's exactly the same though. More bits where they're needed, less where they're not.
I wonder if CBR MP3 could be considered "slightly evil", given it can use a bit reservoir to allow temporary changes to the bitrate, even when it's a CBR MP3.
http://en.wikipedia.org/wiki/MP3#VBR -
VBR audio is fine for distribution / final format for most purposes
But VBR audio can cause sync problems in many editors that need non linear access
(Even avisynth has EnsureVBRMP3Sync(), which is usefull for all types of VBR audio, not just MP3)
http://avisynth.org/mediawiki/EnsureVBRMP3Sync -
To clear up for the thread:
My issue was not using VBR audio, but using ANY 2-pass audio encoding option in MEE with DirectShow decoders. I just choose to use VBR encoding as my 2-pass method. The only option I have that doesn't give errors is 1-pass CBR.
Another thing to clear up:
Slightly fixed my issue. Initially, trying to run videos through to convert I noticed an alarmingly large amount of them failed with the 2-pass error. Before, I would only get a couple errors. For whatever reason, it dawned on me that the only thing I was changing was the versions of LAV Filters that I used. I decided to roll back to the oldest version that I could (0.46 was the earliest usable release for me) and that seems to have solved a lot of my problems. I've had one video still get the error so far -- out of testing only a few -- but hope to test a lot more in the next coming days. -
Well directshow is not necessarily frame accurate - it might give different results on the 1st pass vs. the 2nd pass. You might luck out, you might not. Thus directshow should be avoided at all costs (only used as a last resort) .
I would use another method like avisynth with an accurate, non directshow source filter + wmnicenc . I think older versions of expression were able to accept avs scripts as input, not sure about the current status -
Do you know any good non-DShow filters?
Also, I don't know if this is relevant, but I have no problem reading or opening the files -- it's the actual encoding where I get the error. And it's definitely a random error. It's my estimation from what I've seen thus far that Avisynth is an input method?
I have absolutely no idea how to use Avisynth or how to tie it in with other programs. I would like to go that route, but I can't find any good info anywhere to teach me. And I'm not good at coding at all in any form, so I need a guide that says "Do A, B, C, D and you get X" before I can even start to understand how it works. -
Can you be more specific about the errors ? What kind of errors ? What is the error message ?
Seemingly "Random" errors (maybe encoding aborts at different places) might also indictate hardware issues e.g. bad memory stick. Run some hardware tests e.g. memtest86+, check temps, run stability programs like prime95, linx, occt etc...
If you've mostly fixed it by using a different version of lav, and it only fails consistently on 1 file, then maybe that file is bad ? Then that suggests there was a problem in decoding with the other version of lav that you used
Are you encoding audio only (WMA) , or video and audio ? What is the source video format and container ? (you already mentioned audio was AC3 & DTS) -
The error, I believe, is "Error: bad audio profile selected." Which means you tried 2-pass audio, that ain't happenin'.
I can assure you I am long-tenured in experiencing and trying to diagnose this error. MS has confirmed it's DShow-related error that's pretty random when it pops up. Their only solution was "pay us and buy the pro and it won't happen -- maybe." Hardware is all stable. Just went through about a week or OC'ing and stress-testing the PC, so she's running smooth.
It doesn't fail consistently on one file. Well, it will always fail on one specific file, but the file isn't bad (I do believe). I can watch it in its native form and convert it using other applications. I've tried my darndest to find what separates an error file from a non-error file but to no avail.
I'm encoding audio and video together. Never had video issues before, just this 2-pass audio issue. -
One common reason for 2pass error but 1pass is ok is different framecounts or audio samples between the passes (directshow is a common culprit, being non frame accurate) . I would use ffmpegsource2 (ffms2) , as it indexes the audio (and video), and is frame accurate (both passes will be the same)
Hardware is all stable. Just went through about a week or OC'ing and stress-testing the PC, so she's running smooth.
It doesn't fail consistently on one file. Well, it will always fail on one specific file, but the file isn't bad (I do believe). I can watch it in its native form and convert it using other applications. I've tried my darndest to find what separates an error file from a non-error file but to no avail.
You can think of avisynth as a frameserver. The .avs file will "look" like video & audio to the other program. You just load the .avs (it's just a text file) in the other program
1) Install avisynth
2) Download ffmpegsource2 plugin, vanilla 32bit version, unzip the .dll and .avsi and put in the avisynth/plugins folder and they will autoload
Here is the main page
http://code.google.com/p/ffmpegsource/
Direct download page to version 2.17
http://code.google.com/p/ffmpegsource/downloads/detail?name=ffms-2.17.7z
3) Create a text file with notepad in the same directory as video . Write the following , change names and file extensions to match (e.g. "video.avi") , save it. Rename the .txt extension to .avs (e.g. myscript.avs)
Code:FFMpegSource2("video.ext", atrack=-1)
When you have more time you can read about the wonderful avisynth - it's very useful for many, many things
http://avisynth.org/mediawiki/Main_Page
http://avisynth.org/mediawiki/Main_Page#New_to_AviSynth_-_start_here -
I'm confident in the stability of my hardware.
I'm not sure how else I would go about error-checking in a file. Again, I can convert the same file using other encoders. Beyond that, I'm unaware.
I tried your Avisynth method and, using MEE, it returned the same "Error: Bad audio profile selected" message.
I've perused those Avisynth links, as well. As stated, I have no experience so I found them far too complicated. The first couple steps for okay, but after that it became way too dense way too quickly. I need someone to teach me how to use it to do exactly what I want, or else I'll never learn. That's just my learning style: I find out what I need to know to get what I want done, and then go outwards from there.
Similar Threads
-
MKV/DTS files to MKV/AC3 & AAC
By ktklein72 in forum AudioReplies: 0Last Post: 15th Feb 2013, 12:43 -
What's the difference betwen AAC, AC3, and DTS audio's for movies?
By Eagleman in forum AudioReplies: 3Last Post: 29th Aug 2011, 14:36 -
Loss of Video Quality During SWF to DVD Conversions
By rckowal in forum Video ConversionReplies: 4Last Post: 11th Aug 2010, 09:16 -
MeGUI audio encoding dts/ac3 to aac-mp4, keep getting errors
By runLoganrun in forum AudioReplies: 5Last Post: 29th Jul 2010, 18:15 -
Best method to decode DTS WITHOUT QUALITY LOSS?
By m1ndripper in forum AudioReplies: 12Last Post: 22nd Nov 2009, 17:07