CCE Crash = Restart?

[guboogi]

i'm using the following CCE guide from Doom9:

Getting The Most out of CCE

it says to first run a one-pass vbr first and then when that's done "while CCE is still open" to setup the multipass vbr. i did the one-pass and then as i was setting up the multipass CCE crashed. do i have to do the one-pass over again? does the one-pass and then multipass have to be right after each other? or can the program close and open again and then setup the multipass and encode.

i tried starting up cce again, loading the movie file and then setup the multipass but when i tried to encode it stalled and became unresponsive (it wasn't doing anything.) did i set it up wrong or is it because the program closed between the one-pass and multipass vbr's? thanks.

[adam]

No you don't have to do it all with CCE open the whole time. The first pass is simply used to create the VAF. This is necessary for multipass encoding because it contains all of the information gathered about the source during the scanning. After each pass the vaf file is updated with more accurate information about your source so that the encoder can better allocate bitrate.

Once you have the vaf you can close cce, reopen it again and begin the multipass encoding process or enter the advanced tab and custom tweak the bitrate allocation.

I'm not sure what is causing cce to freeze like that but it has nothing to do with you closing and reopening cce. The guide is just written like that for simplicity's sake I think.

[spooki]

maybe a silly question but why don't you just do a multipass vbr in the first place, i just select multipass vbr as the first step, put in the required average bitrate to get the required file size, change some of the other options as in the doom9 guide and then hit encode and leave running.

The first pass of the VBR creates the.VAF file then the remaining passes create the .mpv file, a 3 pass VBR produces great results.

[adam]

Cause he is using the "getting more out of cce" guide. Its intention is to get adequate results in a much smaller filesize for dvd authoring purposes. Also, despite requiring more manual work, it makes the additional passes unneccessary so it speeds up the encoding process. Personally I don't see the point of this method but others swear by it.

[guboogi]

hrm, i just followed the guide so i'm not sure why i didn't just do multi-pass off the bat. does doing it the way of the "getting more out of cce" guide really result in a smaller file size? can you explain why? i'm just curious.

[adam]

Does this guide result in smaller filesizes? Personally I do not think that it does. The guide basically allows you to do semi-manual bitrate distribution. The encoder scans the file once and gives you its rough estimation of where bitrate should go. From there you simply go through the movie and raise the bitrate to just prevent image distortion in all parts of your movie. The idea is that if you were to instead use the typical multpass vbr encoding method that you would more than likely choose a higher avg bitrate than you would need, and thus just be wasting space on your dvd-r\w. This guide allows you to use just enough bitrate for acceptable quality and no more. By acceptable I mean that there is no considerable image distortion in any scene.

Now this makes sense, but I think it could all be avoided with a little common sense. First of all the idea behind this method is to fit more than one movie onto a dvd-r, or maybe a movie plus alot of addional footage such as extras. Now using this method you would encode file 1 and get acceptable quality in a relatively small filesize, then file 2 and so on until you ran out of room on your disk. But with minimal experience in encoding you can quickly learn what a reasonable bitrate is going to be needed for a given source.

So it comes down to a simple pre-encoding decision, for instance, do I want to fit two movies on this disk or three? From there you can use traditional multipass encoding methods and just let the encoder work with the maximum amount of bitrate allowable as dictated by the size of your media. If you go with two movies, for example, you may be wasting bitrate, in the sense that you "could" have gotten away with less, but that wasted bitrate is not enough to be used for any additional footage so who cares?

I know that alot of work has gone into this method and it has evolved greatly due to collaborative help. This method may provide more of a quality boost than I suggested, I haven't really had a chance to fully test it yet.

Try it if you ever get your problem fixed but you should also just try regular multipass encoding, as I think its more practical. See which method you prefer.

[buckonine]

Adam - the single pass is not so much to create the smallest file size possible, but for me it is to create an initial VAF. at that point i do not use the lowest bitrate i can and keep it "in the green" i watch the output size and use the highest botrate i can and still fit the audio and extras on the disc. when doing this it is rare if ever that i have to go and change local botrate values. most of the time when you look through the movie in Q-scale any grey i see is already at a very high bitrate. and even with the grey in these areas i have never noticed and bad places when viewing the movie. the Q-scale is a CCE algorithim or something, it's just their own way to tell quality more than anything else. but it is a good guideline to use when in the advanced mode of VBR. while many people may do the single pass, then be excited about the small file size i believe most people see this as a dropping of point, and do not just accept the given file but "tweak" it to the point they really need. using this i have never made a disc that had a final size of under 4.35GB and most are within a MB or two of maxing out the disc. not sure if this was a direct rebuttal to your post, or if i just needed to ramble today =)

[adam]

buckonine the point of the first pass in any multipass encoding process is to essentially create the equivalent of CCE's VAF file. Basically this is just the intial information that the encoder gathers about the source, which it then adjusts with each additional pass so that it ultimately has the most accurate information before finally distributing the bits.

Unlike most encoders, CCE just lets you adjust local bitrate distribution manually, but of course only once there is something to use for analysis, hence the need for the VAF file.

If you look at my post you will see that I am simply outlining the purpose of the "Getting the Most Out of CCE" guide. This guide is used, or at least was initially used, exactly as I described it, to achieve an acceptable level of quality in a smaller filesize so that more content can be fit on large capacity media. I don't use the first pass in CCE in this way either. I simply use it so that CCE may perform multipass vbr encoding automatically. CCE allows you to completely tweak almot every facet of its bitrate distribution. Use that capability however you choose, I simply described one possible method.

The Q level is not something specific to CCE, and actually its not a direct measure of quality. Just like with any other encoder the Q level represents the encoded file's quantization.

I think you are misunderstanding the method described in that guide, however. The process does not stop after the first pass, hence the term "first pass." Just like the method you described, the 1st pass is basically just used to create the vaf file so that further tweaking can occur, and so that the encoder has a base set of values to start with. The "Getting the Most out of CCE" process is still multipass vbr its just a combination of automatic and manual vbr. The filesize can still be predicted so that the encoded file completely fits the dvd-r.

You said you have no problem fitting your movie + extra's onto one dvd-r/w, as well you shouldn't. This guide aims to fit 2 or maybe even three movies + extra's on 1 dvdr/w. That's the difference.