I doubt you screwed up, just that, for whatever reason aacgain doesn't like not having the /k missing, that is ALL we have changed.
Will give it a quick go..
+ Reply to Thread
Results 271 to 297 of 297
.. am assuming the latest executables work for you? i.e. you can succesfully run the program without the /k option?
.. the idea I had was a bit silly but to "replace" the /k with another switch, in this case /a just in case it was something to do with the number of command line switches being sent, I looked at the code before to see if these are referenced elsewhere or as variables/parameters bit couldn't see, but thought it worth trying.
Sadly, it didn't work
Okay I hadn't ran it myself at the time without the /k option...however I now have....exact same issue you are experiencing...and the answer to the problem is if there is no /k then there must be a /c
With /k gone & /c added things work peachy.
OK great, will give that a go in the morning
Very nice to see the improvements in the program where I think the last version to download?
Success. All files were processed and all files when checked had a gain of ~89db.
I loaded one of the MP3 from the 'fixed' files into Audacity and you can see some clipping but typically this is only bang, crash, wallop gunfire type segments and playing it, it does not sound distorted.
The file in question, when processed with 'prevent clipping' enabled took an 81.5 audio track and REDUCED it to 77.0 so dialogue sections were very quite.
I then decided to run one of the Audacity in-built filter "Compression". I'm no expert but having used this in the past it has the effect of making quite sections a little louder and, if left at default settings, will reduce the volume of any clipping audio. What it actually does is a bit more involved but that's the effect you get (when run on music it has the effect of making the audio sound more dynamic).
It worked a treat, here is a screenshot of a small section that I processed. Original Audio on top, Processed below.
I know one of the issue when 'normalising' video is the conflict between loud action sequences and quiter dialogue, maybe this is the answer?
Just to thank you once again for the time and effort you've put into this so far and taking the time to explain it.
I'm sure Nologic will be along soon but the 1.40 version posted on 21st Jan works fine if you are happy to have the output audio processed as to not allow clipping. This would probably be fine for most people but I had a requirement which meant I needed higher gain so hence we investigated having an option to remove this check and most of the last dozen or so posts have been feedback and bug hunting on that.
The 1.40 install has a separate program Video Gain MP3.exe which allows for the creation of AVIs (in most cases). Again, this may not be relevant to many but I needed it as I had a requirement for the file to play on an AVI only playing DVD player.
Yeah as Malcolm stated the latest can be found here.
I'm starting on the new major version update now...the newer GUI is going to take a fair amount of new code to have everything behave correctly...everything else shouldnt' be that big of a deal. Heh probably already jinks myself.
Just my two-cents worth: If you make one bit louder and not another bit, then you are doing compression. Normalize means changing the level of everything exactly the same amount.
Hi everyone. First post. Hoping someone can help.
I'm trying to use the latest GUI but all I get with any settings is an error stating "Failed to Create Finalized Video For".
I'm trying to level a lot of AVI files. I've tried doing it 1 at a time or all at the same time but either way produces that error. I've tried compatibility mode and running as admin, still same error. Also tried both .exes included in the most recent version (v1.40). Is there a command line I need to use with the Video gain.exe? Or am I just being plain stupid?
If you want screenshots of the errors or a log etc. just let me know.
Thanks in advance
"Failed to Create Finalized Video For" is actually slightly generic...which is my fault. This message is generated when ether the original merge of data streams fails...or when moving the merged streams to the final destination.
I've had this happen to myself several times...and most the time it's not clear what went wrong.
First thing to check is the most silly....which is that the user account that the script is running from can create files and folders in the destination location. So while in that account simply right click on the background of the folder and select New\?? where ?? is take your pick...folder, bmp, txt...anything. Once created simply delete...as you now know you have write permissions.
After that you can see about updating FFMpeg:
Grab a shared build rather than static...extract and overwrite whats required in Bin\FFMpeg\.
Then you can see about updating MKVToolNix:
Heh be sure to grab the Windows build. Do the same as above overwriting the proper files in Bin\MKVToolNix\
Those are the only two things that have been updated since my last release...which may or may not improve the situation.
Past that...upload a sample file somewhere and PM the link and I'll take a look and see if I can find out whats going wrong...and if there is anything I can do about it.
Keep in mind I'm just tossing on a pretty face onto some awesome command line tools.
Just wondering, is the volume levelling option still configurable? I'm used to using the command line for the old one without the GUI but got tired of having to move 700+ files around all the time.
If possible, could you give me a quick guide on what settings I should use for optimal quality and level volume? Don't want to spend loads of time processing videos just to find that some have become too quiet to hear or too bad quality to see.
The 64bit version of FFMpeg is likely more stable than the 32bit one...however since all the original app's are 32bit, I'd suggest for the time being not using the Direct Replace option...as things might get nuked somehow if things go wrong when piping 64bit to 32bit. I just recently moved over to Win7 64bit...so maybe sometime in the future I'll really investigate the stability of mixing 64 & 32bit app's. For now be careful.
Gain at present isn't available through the GUI, this is something that you have to edit the source code to do. A couple posts back discuss this very thing...at present it defaults to 89db provided there is no clipping...if clipping is present then it'll lower it to a point where there is no longer any clipping present.
Honestly I feel this is a good default and should be left alone...as one can always turn up the volume on their speakers...and clipping is certainly unwanted by any sane person.
There will without a doubt be video's that are to quiet...how often largely depends on original source material...if it started off excellent then it shouldn't happen to often...but it will happen.
Ideally the next version should cut this down more...but it will still happen...at this point its unavoidable.
The defaults that the app opens with are what I consider decent if you are starting out with good source material. Any time you encode things get worse...so if you have the room to spare...I wouldn't encode the video stream...but if you don't the defaults are a good starting point. With Video the lower value equals higher quality...but also increased file size compared to using higher values.
Now Audio encoding will only happen if the audio stream isn't already ether MP3 or AAC...at which case it'll encode to AAC. At present AAC is done by Apples encoder, which is far better than the Nero one I used prior, and produces better audio than the original Lame MP3 encoding done by the original script...however newer Lame builds do nearly as well over all as Apples AAC encodes. Now with Audio the higher the value the better the quality is.
Really this stuff is in the ReadMe.txt
I would generally just use the defaults, with the one covet being that if you have cartoon animations that you intend to encode that you move them to a different folder and run the app with the Animation tuning option selected for those, as it'll help prevent jittering from appearing on solid lines, which is very distracting. Also with Animation you may want to try the x264 10bit encode as it'll generally get a smaller file size, for that type of material.
Last edited by Nologic; 12th Mar 2013 at 04:48.
Ok, I'll bare that in mind if anything goes wrong.
89db is just right from me really so I doubt I'll need adjustment, just wondered if the facility was available. Only issue with turning the volume up on my speaker setup is it has 2 settings: A) Can't hear it very well or B) Deafening (Probably should lower my wattage really... 2000w through a 5.1 system is overkill but oh well)
I did have to change the "force MKV" option. I left them all un-ticked as my media player (XBMC4XBox) doesn't support .MKV.
Audio stream is probably MP3 for all videos as they're all standard .avi.
I'm terrible for ignoring ReadMe files. I'll have to have a read through on this one.
Line jitter I have had some major issues with and was never aware of what is was called or why it was happening so you may have just solved that for me haha! File size isn't really an issue, I can always get a 1TB IDE or larger SATA and convert it to IDE for use in the XBox. PC storage is practically unlimited, I must be nearing 4TB with the amount of 2.5" HDDs I have just lying around.
Thanks again Nologic and also great work so far
Notes: Version 1.40 serious bug. Regardless of the frame rate is changed to 50, at 1.37 this was not. And bitrate changes.
Suggestions: Add the ability to measure the volume in video files. That is, just do batch analysis without normalization.
Well, add. And focuses on version 1.37 - it is the most successful so far, although some bugs are there (I can list if you're interested).
Yeah give me a list of what you've found, and I'll see what I can do.
Really like to know what makes 1.37 work better for you than 1.40
Well, maybe you already know them:
- When using the "Direct Replace" may damage the files of (or the video will not be seen, or the sound properly normalized)
- If you do not have enough disk space for processing files, you also may be damaged
- With the normalization of more than 19 files, some are not processed (issue any appropriate error)
- Variable bitrate is rounded to the constant (that is, there were VBR 284, CBR 320 or 256 was)
The latest version has improved much more still gives error: Failed to Create Video Stream 01 For: in some files in the first attempt over the second time converts without problems, how to fix this problem?
@ Art Blos -
It shouldn't be doing anything to the bitrate of MP3's...as the code just copies it...I might could see it getting out of sync and such...but it shouldn't be doing anything to the bitrate...at least my code isn't instructing it to do anything.
@ Belini -
Yeah I've seen that time and time again and haven't a clue as to why that happens. I noticed while working with MaxxUK that some of the video's have compressed video streams...so I kind of wonder if maybe that might contribute to the issue...but it's all guess work. So I might start using MPEG4Modifier to unpack video streams.
I haven't gotten around to doing the big update I've been talking about...however I'm looking into going back to how SoopaFresh originally did things...in some respects. Mostly this means re-encoding the audio...I believe the volume would likely be more predictable & would not require a Replay Gain aware media player. However the down side is of course that when ever you encode something there is going to be damage done...to the stream.
Anyways the little work that I have done is swapping out Unlocker for FileASSASSIN as I feel it works better, and switching from AutoIt's process killing abilities to that of PStools...again because I feel it works better. Out side of that I haven't really done jack.
Course both of those are mostly because AACGain sometimes ether locks up or locks files in the file system...which makes a mess of things.
Is more complex than I thought then, as it happens to fail the first time and work in another attempt could put in code to convert all files when it gave some error try again.
I could probably set it to try once again...I don't want to introduce a possible infinite loop...I rather have the script trying to get as much done as possible rather than sitting beating it's head against a wall forever.
I'll see if I can't get a release out tomorrow.
Thanks for listening, I'll be waiting for a new version to test.
@NoLogic & others:
Is there an official download webpage for VideoGain?
Is there a tutorial/explanation for VideoGain?
I need some input for example:
1.) Does a higher Audio quality number means better quality?
2.) If I NOT tick MKV, MP4, H.264 checkboxes is then video part untouched and only audio volume adjusted?
I want only to increase/decrease MP3 audio tracks inside an *.avi video