Lowest Bit Rate without DVD Choking

[monteie]

I have been playing with DVD's and SVCD's for some time now. I have recently started moving away from my original prcess to create them and trying different thnings. Currently I am trying to see how MVR (Manual Variable Rates) in TMPGenc work. I was wondering for the Minimum BitRate value, what should it be? The default seems to be 300 but can you go lower. Is it advisable to do so?
If it's any help at all I am using a Pioneer 5-disc player.

Thanks Guys...

[poopyhead]

lowest bitrate doesn't really matter...i believe i have mine set at 0

basically with VBR, it uses the range between the min and max values you have set....but that doesn't mean it will use the min or max value...usually, it's something in between

[monteie]

I am wondering what the point is then? I have been in the past using Constant Quality to encode and it has worked very well. I just thought that using Variable rates would save on space without sacraficing quality. I guess that was a mistaken assumption. Can you tell me what the difference between Constant Quailty and Constant Bitrate is?

Thanks for the Help

[Mirror_Image]

I did a google search for:

CQ constant quality

and this was the top result, very informative site:

http://www.tecoltd.com/enctest/concepts.htm

this was the third link on the search result page:

http://tangentsoft.net/video/mpeg/enc-modes.html


http://www.google.com/search?sourceid=navclient&querytime=C2W&q=CQ+constant+quality

[Greg12]

with DVD, VBR is best, and the more passes the better

[poopyhead]

VBR kicks the crap outta CBR because VBR allocates bandwidth better......

for CBR, the slow scenes get as much bandwidth as fast, motion scenes....obviously, there's something illogical about that...wouldn't it be much better if some of the bandwidth was diverted from the slow scenes to the fast, motion scenes? i mean, the slow scenes don't need that much bandwidth to begin with, so taking some away won't make a difference...however, fast, motion scenes need more bandwidth to be crisp and clear

[monteie]

Thanks for the posted links. I understand what you are saying about CBR using the same bitrate stream for fast and slow portions of the movie. It is more bandwidth. I would like to use a VBR but there are two problems. 1) I won't sacrafice quality to save on CD's. What is the point of that @ today's prices? 2) A multiple pass seems to be the way to go but I am assuming that it would added several hours to the process. I would rather use the extra CD than wait the extra hours to just get the same result in the end. Acutally I am not convinced that a 2 pass would be better than CQ as far as the end result. Just out of curiousity... Do you guys now what the difference in encoding times would be (ball park guess) betwen a single pass VBR and a two pass VBR?
I have a PIII 700, w 128 RAM if it helps any..

Thanks again guys...

[cornsoft2002]

Yes,

so except for the audio ripping and encoding to mp3

it is fair to say that n-passes take n times the single pass!

I always use 4 pass VBR and for a average movie length of say 100 min
it takes 10 or 11 hours on my 1.33 Gh Athlon (but thats including the audio encoding, subtitle ripping, muxing and creating vcd imager cue/bin files

btw, You won't see difference in more than 4 passes

gr.

Bart.

[monteie]

Do you find from you 4 passes a great savings in file size? I would expect that would be your reason for doing it.

[cornsoft2002]

No,

it is not for saving space it's for better spreading the available bandwith of data to either simple parts and complex part of the movie

If you use CBR all scenes will have the same bitrate
With VBR complex scenes may use a higher (so better) bitrate
The passes are for calculating the best encoding.

Bart.

[monteie]

I am wondering if I would gain from using that technique..... I currently encode using CQ with a bitrate of 2520 and a CQ value of 90. Do you think that I would be able to get better quality than that from performing a 4 pass VBR? Would my Pioneer support a higher bitrate? I wonder... The quality of the tests that I do are really good as far as I am concerned.
As well...... I was reading your previous post about your encode times. I also read in the thread that each pass takes up the same time as a normal single pass. If this is true and you are encoding 100 min @ 10 to 11 hours that would mean that you would encode @ my method or at a single pass VBR at 2.5 hours.... Is what I have stated correct... ? If so I think that it is time for me to upgrade the processor

[mickboss]

Best advise:

- If you plan to fit no more than 40 minutes of video / CD:
Use CQ_VBR at 70, 1-pass in TMPGEnc

- If you plan to fit +40 video / CD:
Use Cinemacraft with 3/4/5 passes VBR (40-50 minutes / 50-60 minutes / +60 minutes)

Cinemacraft is about 3-4 times faster than TMPGEnc.


I take about 6 hours to encode a 2-hour film with ripping, subtitles, sound, 4-pass:

Athlon XP 1700+ @ 1800+
256 Mb SDRAM PC133
2 7200 rpm MAXTOR HDDs
Asus DVD-ROM 16x ATA100
Windows XP
...

Almost all standalone DVD players alow (at least) up to 3000 Kb/sec
BUT remember: 3000 kb/sec = Video + Audio + 50 Kb/sec for stuff...
So: Video = 3000 - (112/128 kb audio) - 50 = (+-) 2830 Kb/sec

Conclusion: You can safely use video bitrate up to 2830 Kb/sec

My Nintaus alow up to 3500 Kb/sec (video+audio)
My Mustek 560 alow up to aprox 7000 Kb/sec (video+audio)
8)

[monteie]

I took a read around the net about the CinemaCraft product and it looks very impressive. I believe that the SP v2.0 is what I need. Can you tell me where I can get it other than the Company Page? Do they have a free version that you would know of?

Thanks