Is DVDshrink slower on Athlons?

[Timster]

Is DVDShrink slower on athlons than pentiums? I've tried DVDShrink versions for the past few years and everytime I think the thing is slooooowww. Running XP Sevice pack 2 on an athlon 1200ghz. For comparison, Backing up my copy of "Saving Private Ryan" will take DVDShrink about 3 minutes to analyze and about 1hour 20 minutes to transcode. DVD2One alongsiide DVD43 on the other hand will transcode it in 30 minutes with no need to analyze. I personally think the quality is better on DVD2one. So I'm wondering why people praise this program other than the fact it's free? Or is that it, it's free? I'd rather pay $40 and back up my DVDs in 40-45 minutes. Less stress on my DVD player and processor which will in the long run probably save me the $40. Just curious.

[Capmaster]

Everyone has apps they like and apps they don't like so much. It's like taste in food or cars. Everyone has their favorites.

I happen to like a slick little app called Vobrator that I almost never hear mentioned here.

You like DVD2One, I like DVDShrink. Simple as that.

[US Guy]

Is DVDShrink slower on athlons than pentiums?

My question to you is what Pentium are you based on??

I have both AMD, TB900 and XP200; and a P4 3.0E.

Personally I still like the AMD better given the same PR; and for sure at same clock rate.

By the way, shrinking does not only rely on CPU alone. The ROM speed (locked and unlocked is also a factor) and system resources. Of course there are more...

[US Guy]

Everyone has apps they like and apps they don't like so much. It's like taste in food or cars. Everyone has their favorites.

I happen to like a slick little app called Vobrator that I almost never hear mentioned here.

You like DVD2One, I like DVDShrink. Simple as that.

You are not answering the question !!!

[wulf109]

It's possible that Pentium's are faster. If you look at CPU reviews on sites like Anandtech,Pentium's seem to score higher then AMD XP's on video conversion tasks. But I'm generalizing.

[johns0]

Everyone has apps they like and apps they don't like so much. It's like taste in food or cars. Everyone has their favorites.

I happen to like a slick little app called Vobrator that I almost never hear mentioned here.

You like DVD2One, I like DVDShrink. Simple as that.

You are not answering the question !!!

You are not answering the question either :P @Timster,Have you compared the result with a pentium 1.2 ghz?Some systems are faster than others even with the same hardware and o/s but with differences in how the systems are setup with various software and different drivers can influence the speeds.

[US Guy]

You are not answering the question either :P @Timster,Have you compared the result with a pentium 1.2 ghz?Some systems are faster than others even with the same hardware and o/s but with differences in how the systems are setup with various software and different drivers can influence the speeds.[/quote]

I am still waiting for him to answer my question.

Besides, what you wrote is a repeat of what I wrote. STFU if you do not know the inside story.

[US Guy]

Personally I still like the AMD better given the same PR; and for sure at same clock rate.

By the way, shrinking does not only rely on CPU alone. The ROM speed (locked and unlocked is also a factor) and system resources. Of course there are more...

@ johns0

Isn't this answering:
Given the same PR, better yet at same clock speed. I like the AMD better, sufficient enough to say NO AMD is not slower.

[Capmaster]

Everyone has apps they like and apps they don't like so much. It's like taste in food or cars. Everyone has their favorites.

I happen to like a slick little app called Vobrator that I almost never hear mentioned here.

You like DVD2One, I like DVDShrink. Simple as that.

You are not answering the question !!!

I was responding to this:

So I'm wondering why people praise this program other than the fact it's free? Or is that it, it's free? I'd rather pay $40 and back up my DVDs in 40-45 minutes. Less stress on my DVD player and processor which will in the long run probably save me the $40.

And I wasn't aware you were appointed "post quality judge"

[hetvarken]

Well, I own both an Athlon T'bird 1400 (FSB 266) desktop machine with SDRAM and a Pentium 4 2.8 GHz laptop (FSB 533) with DDRAM. The laptop is only marginally faster

I have to confess that this laptop is a Dell
So that might explain it....

[Timster]

I wasn't asking if a Athlon or Pentium was faster. Rather if DVDShrink was faster on a Pentium than an Athlon. Some programs are optimized for Penitums and some for athlons. I only have an athlon so I can't do any testing. I was always baffled by people stating how fast DVDShrink is when I find it painfully slow.

[budz]

I have to confess that this laptop is a Dell
So that might explain it....

What does Dell have anything to do with how fast DVD SHRINK will encode faster or not? Having a Dell laptop doesn't mean anything. I have a pent 3 933 mhz pc that probably can keep up with your laptop!

Now to the question at hand if AMD is slower. I can't fully answer that because I don't own a AMD pc. But in my experience 80 mins. to encode using DVD SHRINK isn't that slow. Have you ever ripped the movie with DVD DECRYPTER just to see how fast your DVD-ROM reads the dvd files? I would think that would make a difference too not just the speed of the CPU. Whjle you're letting DVD SHRINK do it's thing do you have any other applications running? surfing the net? Because multitasking while shrink is running you're using up CPU resources. I've noticed this so I don't run any other applications but I do surf the net. Anyone can correct me if I'm mistaken about this.

If you feel that using DVD2ONE is better for you then use that program. As Capmaster said everyone has their favorite applications to use. I personally like using DVD REBUILDER for the movies that I LOVE! It's a long process but the video quality is stunning that it blows any other program away in my opinion of course!

[Alan69]

IDVD2One alongsiide DVD43 on the other hand will transcode it in 30 minutes with no need to analyze. I personally think the quality is better on DVD2one.

Does it strike you how it may be a bit less productive to spend $40 on software to speed up your backups on a slow computer vs using that in what it cost to upgrade to something more modern so everything runs faster instead of only one program?

Shrink is relatively painless and operates quite well on a faster system, you're 2 or 3 times more sensitive to speed issues than most because of your slower CPU. Yes there are minor speed differences between CPUs, but they'll be majorly swamped by using a 1.2 GHz processor when 3's are available..

Shrink also has a deep analysis, which you probably don't use because it'd take an extra hour instead of the 20 minutes for most, so you aren't even judging by the quality they get. If you like DVD2One, then use it and don't worry about what other people say. It's hard for you to do an apples to apples comparison to what they say anyway with a 1/2 or 1/3 of their speed CPU, so your answers will end up different to their's anyway. Priorities shift when you're doing something 3 times slower..

But do note that it hardly makes sense to worry about time on a drive etc when drives are $9 on sale. With the fact that you are still in the hour or two range, and it is so easy to run these overnight, your priority should be on best quality even with a 1.2 GHz processor since it's only done once. But a ~$100 upgrade to a 2.4 or so wouldn't be a bad investment either.

DVD2One may be noticably better in some ways, with the $40 price meaning paid development. But with 2.4 GHz machines all over the place and DVDShrink being free and generally fast enough, most just don't care that much about any speed differential, and no one can blame them for that really. And with no analysis and at the same bitrate, I think you've just justified picking a favorite program from it operating faster on your system anyway. Any longer movie with changes in action level should show up the no analysis fairly readily. Assuming it's still adaptive on the fly, no pre-analysis shouldn't be that much worse, but you'll notice it here and there when bitrate is tight.

Alan

[Cobra]

I hear that Athlons are generally better in gaming, but Pentiums are quicker for video encoding. Personally, I run a heavily-overclocked Athlon XP2500.

As far as AMD vs Intel wars go, let's not go there. It's been a subject for many arguments and it has no definitive answer. It is personal opinion and taste. The same goes for your choice of DVD copying software - you get to the end with a backup of your DVD that you are happy with.

US Guy - please don't stir up trouble. You'll only get a yellow - not worth it, I'd say.

/ Moderator Cobra

[lordsmurf]

When it comes to transcoding with a compressed-domain transcoder, no. I don't see any real difference that's worth talking about. What few jobs I do with DVD Shrink, I all do on an AMD system. I've never noticed it to be any faster or slower than the P4 with this software. Not anything worth counting (few minutes, maybe?).

[jabloomf1230]

Most of the websites that do benchmarking, now include media encoding in their benchmarking suite for CPUs. For example, here's a test from www.anandtech.com that compares 3.2 GHz Pentium 4 CPUs with Athlon A64 3400+ & FX51 CPUs, Scroll down to see the media encoding graph:

http://www.anandtech.com/showdoc.aspx?i=1946&p=5.

The Pentium processors were about 20% faster in encoding files to MPEG 4, for what it's worth.

Here's another benchmark from www.ZDNet.co.uk that compares the Athlon A64 3800+ and the Pentium P4 3.6 GHz. Again, scroll down to see the media encoding graphs:

http://www.zdnet.co.uk/print?TYPE=story&AT=39164010-39024015t-30000019c

In this test, the results are mixed, depending on whether the specific encoding software supports Intel's hyperthreading technology or not. If HT is supported by the software (for example, in TMPEGEnc Express 3.0), then the Pentium has a big advantage in speed.

[andkiich]

I can Shrink a DVD-9 to DVD-5 in 50 minutes or less.

If it's taking your machienthat long, then it might be a configuration issue. The machine I use for shrinking is a dedicated DVD creating machine, so there is nothign extraneous on it.

[proxyx99]

It is not about Athlon or P4 power and speed. DVD2One uses a very efficient fast transcoding technology. I was never able to get more info on it (never really applied myself to it either) but nothing else can explain why my transcoding time for a particular material drops from 1.5 hrs or even longer with some other programs to just 15-20 min. with DVD2One.

This is not about the processor but rather a shortcut in the technology. DVD2One seems to be sharing this technology with at least 2 other programs that I'm aware of. Traditional transcoding or encoding (whatever it means) is a lengthy process. At first I just couldn't believe my eyes when DVD2One has finished in less then 20% of time needed by Pinnacle's Instant Copy with no noticeable sacrifice in quality.

DVD2One transcoder is not a typical transcoder/encoder as seen in TMPEG for example. From the bits and pieces that I've read on the net it is rather an optimized compressor using a constant compression ratio. It seems to be closer related to RAR/ZIP/ACE etc. tools then mpeg2 encoders. The difference is that it approaches mpeg2 file from the file compression angle rather then video content (fields and frames, color, detail etc.). I see it more as zipping the mpeg2 rather then rewriting it in a traditional sense. Of course it is not that simple as the mpeg2 bitrate is lowered according to the disc size/ selected content. It's mpeg2 analysis routine is also very different from DVDShrink.

I wish I was able to better lay out what sets these two tools apart and hope some other participants will shed more light onto this issue.
DVD2One has limited disc content editing options and is in that respect far behind DVDShrink. But the rest of it really shines. As a simple "few click", no frills, backup tool it has no competition in terms of speed. Moreover, its track record speaks for itself. Great quality and speed, how can you not like it?

[jimmalenko]

I hear that Athlons are generally better in gaming, but Pentiums are quicker for video encoding.

I have heard the same - something to do with Pentiums being better at repetitious floating point calculations or something. I'm sure those with more technical knowledge could confirm/deny this.

As far as AMD vs Intel wars go, let's not go there. It's been a subject for many arguments and it has no definitive answer. It is personal opinion and taste. The same goes for your choice of DVD copying software - you get to the end with a backup of your DVD that you are happy with.

Here here. The post before this strays way into the software debate IMO. For the purposes of this thread, we are talking about DVDShrink's performance specifically on Athlons as opposed in Intel chips.

[adam]

DVD2One is a compressed domain transcoder just like dvd shrink or instant copy. Its transcoding algorithm may or may not be more efficient then its competitors, I have no idea, but its fundamentally the same. All compressed domain transcoders do the same thing. They only partially decodes the data and requantize coefficients. The real reason that DVD2ONE is faster is because it only operates in 1 pass (maybe newest version supports more?). It simply transcodes the asset at the set percentage ratio. It does not analyze the motion vector data vs. quantize data amounts to determine where it would be best to transcode, like Instant Copy and DVD Shrink do. This does speed up the process, but it also degrades quality more and can lead to undesired results like having the first half of the movie compressed by 50% and the last half only comperssed by 10%, whereas DVD Shrink could achieve more even compression. Whether or not you notice any quality difference is subjective, but the more compession you use the more noticable this limitation will become.

Timster if you are letting DVD Shrink do the ripping then it is the reading of the data that is the bottleneck. On my burner it takes me 45 mins to rip to my hard drive. From there I can transcode in DVD Shrink in like 15-20 mins. But If I let DVD Shrink rip the DVD then the whole thing would take like 1.5 hrs. That's not DVD Shrink's fault. It can only transcode the data as fast as its being fed from the ROM.

Also, on a fast pc the hard drive itself can be the bottleneck. If reading and writing from the drive, you could easily double your transcoding time by letting your drive get too fragmented. By setting Shrink to read from one drive and write the another, my transcoding times are nearly cut in half. There may be some actual differences between transcoding speed in Shrink when using AMD vs Intel chips, but I'm willing to bet that the real difference in speed you are seeing is due to the configuration of the pc's, not the chips they are running on.

[proxyx99]

The original poster seems to be baffled with the speed difference between DVDShrik and DVD2One and assuming that:
-the technology in both is the same (which is not)
-it must be an AMD vs. Intel thing (or software specific CPU optimization).
Sice both asumptions seem to be a bit off (to be polite) one needs to look into the technology involved rather then the CPU issue. Ignoring this would make any argument a pure nonsense in this context. None of the posts seems to have explained what this huge difference in transcoding may be attributed to. If someone assumes that it is a CPU issue he/she is dead wrong.

@adam PC config has no noticable effect, transcoding iis quite slow in terms of HD speed capability and at most you can save ypourself 5-10% (if that) of the time by writing to another drive.

Also, DVD decrypter seems to be the fastest ripper. Calculating ripping into the total transcoding time may additionally skew the image.

Your assumption about uneven compression/bitrate looks like a pure speculation (unless you can back it up). Never heard of DVD2One or alike to vary the ratio that dramatically (from 10% to 50% ???). It sets the ratio at the beginning of the process and maintains it throughout. Frankly, I'm more impressed with the DVD2One then DVDShrink quality.

[adam]

-the technology in both is the same (which is not)

Yes it is, like I just explained. There are only so many different algorithms used for compressed domain transcoding, and they all operate by requantizing coefficients. That is about the only thing you can do without fully decompressing the program stream. Again, the speed increase with DVD2ONE is a result of its lack of pre-transcoding analysis to compare motion vecotor information vs. quantize information. Its a shortcut, its not a different "technology." Its like using one mpeg encoder to encode in 1-pass VBR and using another to encode in 2-pass VBR. Both encoders are doing fundamentally identical processes, one is just factoring twice as much data into its equation, and thus takes longer to perform its encoding algorithm.

@adam PC config has no noticable effect, transcoding iis quite slow in terms of HD speed capability and at most you can save ypourself 5-10% (if that) of the time by writing to another drive.

Even 5% is quite noticable, but I find that on a fast pc the difference is much more then that. If your hard drive is fragmented then the difference can easily be massive. This goes even for encoding where the CPU processing time is easily 500% longer.

Again, on a fast pc transcoding is not overly slow compared to HD speed. It doesn't take me much longer to copy a full DVD from one drive to another then it does to transcode it from one to the other.

[proxyx99]

There are only so many different algorithms used for compressed domain transcoding, and they all operate by requantizing coefficients. That is about the only thing you can do without fully decompressing the program stream. Again, the speed increase with DVD2ONE is a result of its lack of pre-transcoding analysis to compare motion vecotor information vs. quantize information. Its a shortcut, its not a different "technology."

I'm not arguing about the principles. I call it technology you call it methodology. Both mean the same in the context of the original question: is this a CPU (or maybe software) related?

One more thing, PC issues you named although valid were more of an problem in win95/98 times. Not entirely (to be fair to you) true today considering hardware we all use. Fragmentation you are talking about is a non issue on a ripped DVD files that are usually written contiguously due to their size. Check it out but mpeg/video or other large files are written in huge chunks, almost never chopped up. Fragmentation slowness would be more in a range of 30% decrease of max HD speed in worst case scenario.

[adam]

Your assumption about uneven compression/bitrate looks like a pure speculation (unless you can back it up). Never heard of DVD2One or alike to vary the ratio that dramatically (from 10% to 50% ???). It sets the ratio at the beginning of the process and maintains it throughout.

You are misinterpreting what I am saying. If you tell the transcoder to compress by X% then its going to compress the whole movie by that much. But NO compressed domain transcoder compresses each frame equally. Its impossible because there is no way to compress motion vector data without fully decompressing the stream. Thus all compressed domain transcoders operate by only compressing quanitize information. But each frame contains varying amounts of quanitze information. If one frame contains a great deal then you can compress it quite a bit with no noticable difference. But if it contains very little, then the slightest compression will result in noticable quality loss. Thus programs like DVD Shrink and Instant Copy do a preanalysis pass to determine, roughly, which sections of the movie contain the most quanitize information. Then they can make educated guesses as to where to apply more compression. This ensures more "even" compression in the sense that all frames are compressed by a ratio relative to how much quanitize data it contains. (if there is more to compress, then it gets compressed more.) This avoids the unfavorable result of having 25% left of your movie to compress, and that section only containing 5% of the total quantize information.

The author of DVD Shrink has discussed this function several times both on his forum and on doom9.net, and it has been quoted here at videohelp several times as well. It is a very common and accepted feature in compressed domain transcoders and as I said, earlier versions of DVD2ONE did not perform it. Maybe the latest version does, I have no idea but it is a fact that performing this extra analysis adds time to the transcode and has the potential to increase the resulting quality.

As for quality differences between the different transcoders, I never suggested any one was better then another. Whether your transcoder of choice performs this function or not, the resulting level of quality is still entirely subjective.

[proxyx99]

....Then they can make educated guesses as to where to apply more compression. This ensures more "even" compression in the sense that all frames are compressed by a ratio relative to how much quanitize data it contains. (if there is more to compress, then it gets compressed more.) This avoids the unfavorable result of having 25% left of your movie to compress, and that section only containing 5% of the total quantize information.

That is not what I've read. Instant Copy (DVDShrink as well) tries to save more space by applying more compression to less complicated images and less compression to more elaborate scenes (panning, fast moves). Logically this would result in less even compression curve then going by a preset percentage as DVD2One does... so I don't buy it. It somehow doesn't fit together. But that is a side issue and I do not intend to overanalyse it.
In principal we both agree pointing the original poster to a different aspect of the problem (rather then making it an AMD vs. Intel dabete). And that was my intention. Thanks for your input.

[adam]

One more thing, PC issues you named although valid were more of an problem in win95/98 times. Not entirely (to be fair to you) true today considering hardware we all use. Fragmentation you are talking about is a non issue on a ripped DVD files that are usually written contiguously due to their size. Check it out but mpeg/video or other large files are written in huge chunks, almost never chopped up. Fragmentation slowness would be more in a range of 30% decrease of max HD speed in worst case scenario.

Its exactly the opposite. Since the files are so large there is less chance of there being enough free sectors in "a row" to accommodate it. Thus the file gets broken up into many fragements. Rip a DVD to your hard drive and then run defrag. You will see that the vobs are among the most fragmented files on your hard drive. The larger the files you are working with, the more fragmentation comes into play. Besides, I don't care whether you are encoding mpeg or playing solitaire, a highly fragmented hard drive slows every process down regardless of your os or hardware.

[adam]

Compressed domain transcoders cannot look at the complexity of the image at all because they cannot decode it all like an encoder can. Things like "panning" and "fast moves" are stored as motion vector information which transcoders can't touch. All transcoders can do is requantize coefficients. The only question is how high do you change that 1 number for each frame.

But yes, we agree that the original poster is either comparing two very differently powered/configured machines, that just happen to have different chip manufacturers, or that there is some other variable or bottleneck causing the discrpancy, beyond the already mentioned speed differences between the two transcoders.

Ok, the FAQ on doom9 has been updated quite a bit on this matter. The additional processing done by Shrink and Instant Copy is described pretty specifically there. As to what degree the latest version of DVD2ONE performs similar functions, I don't know, but considering the speed difference my guess is that it does substantially less. Check out the second to last post in this thread in particular.

http://forum.doom9.org/showthread.php?s=&threadid=63587

[proxyx99]

Adam, write a large mpeg/vob to an empty HD and run a defrag. Even though you would see the whole thing as one chunk (just written to it) defrag will attempt to optimize it. That is the flaw of a defrag not a proof of file fragmentation.
As to slowness, it's simple. Just check your HD (or volumes within) with Sandra or similar. Compare speeds on dedicated volumes with mpegs vs other volumes containing regular "stuff". Meve files and test again. You'll notice a difference but not up to 500% as you seem to imply (this is "out of memeory" issue not fragmentation).
On a little populated (normal operating state 10% slowness) volume compared to say 90% data load your HD should be within 20% of speed difference (30% as compared to empty one max.). If you get more then it's your particular HD issue rather then a general rule.

PS. aside from PC config issues on 3 PC's I use at home (1 laptop) DVD2One performs like a champ as opposed to DVDShrink or other software so I fully understand the original poster cdilemma.


Btw. I read the post (your link) and they say:
"InstantCopy attempts to solve this problem by distributing compression evenly over all I,P and B pictures (I think), and then calculating the resulting error for each picture."

Personally I'm not comfortable with the "I think" portion that suggests speculation rather then the solid knowledge. Again, it goes against the logic which crumbles towards the end of this particular post (but I'm not even going to attempt to analyze that in depth, I'm not well equipped with this specific knowledge to raise doubts here). The point has been made and for the sake of this thread we have reached enough clarity for someone to understand the basics. If someone wishes to explore it further it'll be to their benefit.

[adam]

This is getting so off topic. I will just say that I totally disagree with your description of hard drive fragmentation and its effects, and just leave it at that. I'm sure you will agree that it can at least have some effect on transcoding speed, and thus if you are going to compare speeds between two different setups it is something that should be taken into account.

Personally I'm not comfortable with the "I think" portion that suggests speculation rather then the solid knowledge

Then read the direct quote stated by the Pinnacle employee where he says that IC decodes each frame twice to determine the ratio of compression applied relative to its neighbors, resulting in a 2-pass process. As for DVD Shrink, you've got the author himself describing his own 2 pass process. All I'm saying is that if DVD2ONE does not perform similar functions, which I believe to be the case, then that explains its speed advantages.

[proxyx99]

I concur. DVD2One has implemented a simpler, speedier approach and many users seem to like it, me included. I used Instant Copy for some time and have to say DVD2One was a shock to me when used for the first time. As I said it is hard not to like it. Fulcilives pointed me to it some time ago and it delivered on the promise. (btw. what happened to him? did he quit?). (hey, fragmentation component of this discussion was introduced by you so don't complain it gets off-topic ).
I just love its simplicity and have no intention to recreate the DVD to my liking, I'm just happy with the options it offers (I'm a simple guy in that sense no time for bullshit pseudo-editing).

To close the fragmentation impact: all I can say it depends on the user and PC maintenance habits. This is a non-issue for me. I researched it (at a major defrag maker S......c) and came to some conclusions that I shared here. That's all.