Using the PC for other things at the same : impact on quality?

[mrwul]

Up front, I know, I know, this is most likely a dumbest question ever, but eh ..

If I am using the PC for other things whilst DVD Rebuilder is doing its job, will that have any impact on the quality of the output?

I guess not, maybe it'll take somewhat longer before DVD Rebuilder finishes its job, but I cannot imagine why it might have an adverse effect.


Thanks.
--

[Atak_Snajpera]

If I am using the PC for other things whilst DVD Rebuilder is doing its job, will that have any impact on the quality of the output?
No .

[Ghitulescu]

However, if you drive your CPU and/or storage system to their limits for longer periods of time, they will start develop errors. Will these be visible or have an impact? Hard to say. Since the good old 386/16, no CPU ever has been error free. I quit programming some 20 years ago, back then there were "stress tools" to test your application under scarce resources. Today it's much simpler, get another 4GB of RAM, get a bigger HDD, get a faster CPU. To cite William Gates from memory, 640kB should suffice for everyone.

There are error correction mechanisms, however, try not to push the system too far and not for longer durations.

[setarip_old]

If I am using the PC for other things whilst DVD Rebuilder is doing its job, will that have any impact on the quality of the output? As was succinctly answered by "Atak_Snajpera", NO, multitasking will not impact the quality of the output of DVD Rebuilder...

[hello_hello]

However, if you drive your CPU and/or storage system to their limits for longer periods of time, they will start develop errors.

What sort of errors?

[Ghitulescu]

The type of errors one gets when he overclocks his CPU/PC/RAM/graphic card/etc.
Like a false bit in 10 Million. There errors are usually corrected, unless that bit occurs in the checking process, which is even less probable.
Server RAM always had ECC (as another defense against such errors), normal PCs don't.
But the bottom line is, NO, even if one such error will occur, it's hard to be seen on the final video. However, with the current PC configurations (consumer type), probably 1 file in less than 100000 of identical tries will be different than the rest of 99999.

[hello_hello]

The type of errors one gets when he overclocks his CPU/PC/RAM/graphic card/etc.
Like a false bit in 10 Million.

So if you overclock your CPU and it's stable are you saying it'll produce more errors than if it's not overclocked?
What's a false bit?

There errors are usually corrected, unless that bit occurs in the checking process, which is even less probable.

How are they corrected?
So when you said "if you drive your CPU and/or storage system to their limits for longer periods of time, they will start develop errors", are you now saying it doesn't matter, or are you making it up as you go?

Server RAM always had ECC (as another defense against such errors), normal PCs don't.

How does this show non-ECC RAM will start to develop more errors if you're working it harder?

But the bottom line is, NO, even if one such error will occur, it's hard to be seen on the final video. However, with the current PC configurations (consumer type), probably 1 file in less than 100000 of identical tries will be different than the rest of 99999.

So if you drive your CPU and/or storage system to their limits for longer periods of time, will they will start develop errors or not?

[Ghitulescu]

The answer is yes, driving a system (whatever it may be, CPU, RAM, HDD, even the data busses) to the limits, for longer periods of time, will increase the probability of errors according to an exponential curve.

For beginners like you I would suggest reading easier articles, like "when computers go wrong" (but not the anecdotal stories circulating on the net), then go deep down to systems' architecture and so on. Wiki is also a good starting point, it is however not authoritative. http://www.carnegiemellontoday.com/article.asp?aid=573

[mrwul]

Well, the idea was something like running e.g. file copying and Internet Explorer at the same time as DVD-Rebuiler.

In the past I had some (2-3 only) instances when burning a DVD and doing other things the same time.
When playing the DVD it appeared to be unreadable. So my guess was/is that maybe because of the extra simultaenous tasks
somewhere the burning went wrong from some reason or another.

I never got as far as to find out why the DVD burning went wrong: maybe it was just bad luck, or maybe it was
just a coincidence, or a glitch in the program. Either way, ever since, I hardly touch my pc when burning a DVD.

Though, technically, I cannot think of any reason why burning a DVD (considering the buffer size and so on) would
be impacted by something else, such as browsing.

I have DVD-Rebuilder run on a separate pc.

Then again, last time I had to check a few things, whilst DVD Rebuilder was running, which was the reason
for this question.

Anyway, thanks for the feedback!
=

[Ghitulescu]

Well, the idea was something like running e.g. file copying and Internet Explorer at the same time as DVD-Rebuiler.
The answer is No, unless you manage to get an unresponsive script that raises the CPU load to 100% for minutes and minutes. This will increase the chances to get an error (thus a quality issue if that pixel will be white instead of black). It's like in the plane, flying through clouds will increase the chances of a crash. How often did crash the planes flying through clouds?

Possibility and probability are two separate things :)
In the past I had some (2-3 only) instances when burning a DVD and doing other things the same time.
When playing the DVD it appeared to be unreadable. So my guess was/is that maybe because of the extra simultaneous tasks somewhere the burning went wrong from some reason or another.

I never got as far as to find out why the DVD burning went wrong: maybe it was just bad luck, or maybe it was
just a coincidence, or a glitch in the program. Either way, ever since, I hardly touch my pc when burning a DVD.

Though, technically, I cannot think of any reason why burning a DVD (considering the buffer size and so on) would
be impacted by something else, such as browsing.

That was the second issue I've touched - system / data bus agglomeration.

Even on systems with buffer-underrun protections (BURN-proof), the audio CDs will have an increased BLER which may be audible, so don't do "multitasking" when burning.[/QUOTE]

[hello_hello]

Well, the idea was something like running e.g. file copying and Internet Explorer at the same time as DVD-Rebuiler.

You'll find not only do most people here run an encode while doing other things, there are some who run more than one encode at a time. Myself included sometimes.
Because Xvid doesn't fully utilise a quad core CPU very well (usually only running it at around 50%) I'll sometimes run two encodes simultaneously, and often while watching another video or surfing the net at the same time etc.
My CPU is moderately overclocked, and I've suffered no ill effects from pushing it hard. What's the point of having a fast CPU and being afraid to use it to it's full potential?

In the past I had some (2-3 only) instances when burning a DVD and doing other things the same time.
When playing the DVD it appeared to be unreadable. So my guess was/is that maybe because of the extra simultaenous tasks
somewhere the burning went wrong from some reason or another.

I also used to do nothing else while burning a DVD, but these days I'm not scared any more. In fact I've got 3 DVD burners in my PC and I regularly burn three different DVDs simultaneously (running hard drives as a RAID-0 volume they can easily keep up). Sometimes I burn a DVD or two while also converting a video.
These days burners have pretty good buffer under-run protection so even if the burner isn't being fed the data as fast as it's burning it shouldn't be a problem. They can just stop burning and wait. Back in the old days if the drive wasn't being fed data fast enough, it could effect the burn quality.
I know my burning this way has no adverse effect on quality as I always run a quality check on the discs after I've burned them. If you're not burning audio CDs, then any improbable advice as to the effect buffer under run protection may have on their quality is irrelevant. You're burning DVDs.

[hello_hello]

The answer is yes, driving a system (whatever it may be, CPU, RAM, HDD, even the data busses) to the limits, for longer periods of time, will increase the probability of errors according to an exponential curve.

Speaking of ignoring the anecdotal stories circulating on the net.....
Your link didn't seem to have any relationship to the topic of whether pushing a CPU hard will increase the likelihood of errors, and your reply seems to indicate you're just continuing to make stuff up.

[Ghitulescu]

It's not my intention to cultivate you on my time and on other people's threads. Take it or leave it. Please provide counterarguments, not questions, and this in your own thread. As I said, you're supposed to read technical articles not internet things, if you can afford subscriptions to serious publications on this matter, or other subjects of interest for you.

[hello_hello]

I'll leave it.
I've encoded enough video to know there's no quality loss when multitasking while encoding.
Atak_Snajpera, who's the author of an encoding program, says multi-tasking won't effect quality. setarip_old also agrees.

Then there's you slowly backpeddling....
-"if you drive your CPU and/or storage system to their limits for longer periods of time, they will start develop errors."
-"driving a system (whatever it may be, CPU, RAM, HDD, even the data busses) to the limits, for longer periods of time, will increase the probability of errors"
-"Possibility and probability are two separate things"

Sounds like another of your nonsense claims, this time potentially causing the poor OP, who mightn't know not to take you seriously, to be unnecessarily paranoid about multi-tasking while encoding in case it effects the quality of the encode.

Why would I provide counter arguments?? I've not claimed pushing a PC hard for a long period of time won't cause errors.... I don't actually know. I do have enough common sense though to know it sounds unlikely if the PC is stable, and I do know to question your misinformation when it sounds illogical, and I do expect you'd have facts at your fingertips to support your claims instead of silly generalisations about false bits and wrongly colored pixels.... if you indeed know what you're talking about.
Once again though, those expectations have only led to disappointment, and I'm forced to assume you're again posting nonsense.

[Groucho2004]

I can't say I agree fully with anything that has been said in this thread. I have been building and using PC's for 21 years and I have seen pretty odd things happening during that time.

Atak_Snajpera pretty much said what I think but my answer would rather be "Unlikely".

Recently I ran an encode with DGDecodeNV and x264 and at the same time used Firefox 4 (which supports now hardware acceleration through the graphics card). My encoding process crashed shortly after. I'm pretty sure that the early implementation in Firefox is to blame (the hardware acceleration can be turned off in FF, by the way).

As soon as applications access stuff below the application layer (i.e. drivers), shit happens unless the programmer knows what he is doing.

My point is - The encoding process could just as well not have crashed but rather continued while Firefox was screwing around with CUDA. And who knows what impact this could have had on the quality of the encode.

Lastly, if one operates components within their specs it should not matter if they operate at their limits or not. If the component develops errors it is faulty.

[hello_hello]

Well, I'll agree.... I've not only seen odd things happen with PC, I've seen impossible things happen too.

So yeah, when software has got to play nicely and it doesn't, almost anything's possible. Mind you I think a problematic application causing a crash is one thing, but for it to effect the quality of an encode without crashing is probably something else entirely.

I'd definitely agree running components hard shouldn't cause them to develop errors if you run them within their specified limits. Many of us would be running our PCs outside of their "official" specifications and still not have problems. I know I've been running this PC moderately over-clocked since the day I built it, and I run the CPU flat out four hours on end. The only effect I can see it's had on my encodes is to shorten the amount of time they take to complete.

[GaPony]

If I am using the PC for other things whilst DVD Rebuilder is doing its job, will that have any impact on the quality of the output?

I guess not, maybe it'll take somewhat longer before DVD Rebuilder finishes its job, but I cannot imagine why it might have an adverse effect.
--

You actually answered your own question quite accurately. :)

[jdobbs]

The only possibility of conflict where might be if the other programs running were changings settings that are being used by BD-RB (like Win 7 DirectShow's preferred codec settings or selected codecs).