Preparing 720p File for Tmpgenc Authoring Works 5 (Smart Render)

[LIBERT4D]

I got it working... just not sure about how to determine the bitrate settings. I'm good to go and compliant, otherwise. Any thoughts on that would be appreciated.

I like their products for sure, just have to know when to "not" use them I think. Or, in this case, do the transcoding in their dedicated program rather in the authoring one. I consider myself to be pretty tech savvy but unfortunately with so many variables I'm just really better off sticking to a GUI like this. I tried, though! For weeks... haha

[poisondeathray]

Did TMPGEnc end up using pulldown or real duplicates ?




RE: bitrate:

Filesize = Bitrate * Running Time

You generally plan the project ahead of time. The project determines what bitrate you use. Use a BD bitrate calculator , since BD medium is fixed capacity like a DVD 5/9 . e.g. how much are you trying to fit on a disc ? how many titles ? any other features ? multiple audio tracks ? etc...

If you don't really know or care, then do some tests on representative samples. Use a particular set settings and encode serial tests adjusting the bitrate , examine them, adjust the settings and go again

[LIBERT4D]

Well.. since I'm prioritizing quality over anything, that math may be true but not applicable to me.

Was just wondering if I was missing something obvious in what i was doing but I guess trial and error with testing outputs is the way? Seems like there ought to be some sort of math involved with finding the sweet spot that would reproduce the original with very little quality loss. Figuring it based on project size is completely the wrong way to go as I'm not looking to compress anything at all; I'm trying to preserve the original footage in the best way possible.

[poisondeathray]

What concerned you about mastering works output ? You said some issue about the quality ? What specifically? What settings did you use ?

The "best" possible re-encoding method is to use --pulldown double . That's clearly the higher quality, better way to go. It's a shame if tmpgenc doesn't accept it. Twice as many frames has a cost . Sure , b-frames don't "cost" very much, especially when you have true duplicates , but this stop motion content already has strings of 3 duplicates, which means 6 when you hardcode dupes to 59.94. Blu-ray only supports 3 consec. b-frames, that means you start using more costly P frames to store the residuals for the duplicates.

Was just wondering if I was missing something obvious in what i was doing but I guess trial and error with testing outputs is the way? Seems like there ought to be some sort of math involved with finding the sweet spot that would reproduce the original with very little quality loss. Figuring it based on project size is completely the wrong way to go as I'm not looking to compress anything at all; I'm trying to preserve the original footage in the best way possible.

There isn't a "magic" formula, because content is decoded, uncompressed before it is input into x264 to be re-encoded. It's purely dependent on the source characteristics

If you're looking for "archival" quality and truly looking to "preserve the original footage in the best way possible", then how about forgetting about blu-ray and using a data disc or hdd ? There will always be some loss when you re-encode with lossy codec . If you still want "regular" blu-ray , and only concerned about quality, then use the highest bitrate you can to fit the disc. Or chose some value in between if you want a tradeoff. The "best" way is to do some test encodes, look at the output

Alternatively, you can do "capped CRF" encoding . Basically this is a 1 pass CRF encode , which is an approximation of "quality" , but with VBV restrictions to make it BD compliant. The lower the CRF value, the higher the quality. e.g you might use a range between 14-20 . Test one at 14, one at 20. At CRF 0, it is "mathematically lossless" , but will be many times larger than the original, and not compliant with BD. Settle on a value that you think is "visually lossless" .

[turk690]

Did TMPGEnc end up using pulldown or real duplicates ?

I'm curious about this, too. If TAW can do it, how come TVM can't? Is it a subtle way to make one buy the other?

[LIBERT4D]

Ill check out the CRF setting, I think I recall seeing that. Wasn't sure if i should go with VBR, CBR or that.

I'll have to look at the output more closely when I've rested better... Insomnia for the win (loss). Could have just been my vision that was blurry. Probably not a good time to be doing video encoding...but time to kill is time to kill!

I'll keep all your suggestions in mind and see what I can come up with. Thank you!

[LIBERT4D]

Wellll...now I am having audio sync issues. Hell of a thing to test and fail on due to how time consuming it is to convert a 45 min file or greater. I even ran the files through videoredo first to try to catch any timing problems but still ran into them anyway.

[poisondeathray]

Describe the sync issues. Constant delay? (e.g. 200ms always) , Progressively worsening (e.g. in sync at beginning, gradually worse at the end ? ) , or is it only after authoring ?

Was this mastering works ? What does mediainfo (view=>text) say about the source file? about the output file ?

Do some tests on faster encoding presets until you get the process down

[LIBERT4D]

In one particular file it became out of sync in the middle of the file somewhere; it was about a 100ms offset. Now.. an offset across the entire file wouldn't work, as I tried that and the dialogue at the beginning of the clip was way off.

I get the same result in every program I try. Authoring works, mastering works, premiere, encore. So there's a problem with the original file. When I play the original file, though, it's in sync.

What info in particular might you be looking for with mediainfo?

[smrpix]

Wellll...now I am having audio sync issues. Hell of a thing to test and fail on due to how time consuming it is to convert a 45 min file or greater. I even ran the files through videoredo first to try to catch any timing problems but still ran into them anyway.

This is why I initially suggested you let TAW do the encoding. When all is said and done, you haven't saved any time by not using the slower, fully compliant internal encoder (same encoding engine as VMW, btw.) And if you believe you can now magically take your new formula and it will work on all files, while I genuinely hope for your sake you're correct, my own experience says otherwise.

[poisondeathray]

In one particular file it became out of sync in the middle of the file somewhere; it was about a 100ms offset. Now.. an offset across the entire file wouldn't work, as I tried that and the dialogue at the beginning of the clip was way off.

I get the same result in every program I try. Authoring works, mastering works, premiere, encore. So there's a problem with the original file. When I play the original file, though, it's in sync.

Is it just that one clip ?

Or all of them, with the same type of sync issues ? or different type of sync issues ?

So Authoring works, mastering works, premiere, encore are having problems , have you tried anything else ?

What did you use to play the original file ?

[poisondeathray]

Were any of the files in sync? Did you test them after encoding, or only after authoring ?

Was the sample you uploaded "Video+Hot+Enough+for+Ya++wwe.mp4" earlier in sync after your processing? or was that one ok ?

Because I just tested it and that seems to be ok both after encoding & authoring, with either method


Post the mediainfo report, because I'm looking at several things, including the A, V durations, to see the framerate and if there is a frame doubling flag, and to see what settings are passed when you use TMPGEnc's x264.

So if you're inserting duplicates, how are doing that ?

If TMPGEnc is compatible with double pulldown, it really is better for quality; but also faster since twice as many frames will obviously take longer to encode