Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Comparing MPC and OGG at same bitrate (Read 7799 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Comparing MPC and OGG at same bitrate

Ok I'm a little new to these codecs but i've been experimenting alot with vorbis and mpc.
My question for everyone is which is better...:confused:

MPC using the standard settings
or
Ogg Vorbis using a -q 5.25

When i mean better i'm talking pure sound quality.

I get around the same files sizes (approx 170 kbps) and my ears can't tell the difference right now but I'm in the process of building speakers and pricing  recievers so I'll probably be able to tell the difference down the road.

Please try to be impartial to your favorite codec, but i don't need an explanation as to why you think what you do, and try to base your answer specifically on sound quality not any other basis.

Thanks alot!

Comparing MPC and OGG at same bitrate

Reply #1 –
If you're talking pure quality and nothing else, MPC is currently still the way to go. I base this on my experience with listening to both codecs, and the fact that most people I think capable of doing decent listening tests agree.

Quote
but I'm in the process of building speakers and pricing recievers so I'll probably be able to tell the difference down the road. 


But this is a misconception. Hearing artifacts has very little to do with your equipment, but with training and physical abilities. If you don't hear difference on your current equipment, you likely aren't going to hear anything wrong on the new one either.

(BTW. Please don't split up discussions like this over two forums; that just makes things confusing. I deleted the identical thread in the Ogg forum)

--
GCP

Comparing MPC and OGG at same bitrate

Reply #2 –
Quote
Originally posted by Garf
Hearing artifacts has very little to do with your equipment, but with training and physical abilities. If you don't hear difference on your current equipment, you likely aren't going to hear anything wrong on the new one either.


I couldn't agree with this statement more.  In the very vast majority of cases, hearing an artifact is only a matter of training and physical capability.  Equipment very rarely has anything to do with this.

I speak this point of view from personal experience.  For the last few months I've often been confined to performing listening tests in much less than ideal situations, specifically on a laptop with some not so great Sony headphones, yet I still am able to hear all the artifacts I heard before this, and I'm usually still able to hear most of the new ones which other people point out in different clips.

Comparing MPC and OGG at same bitrate

Reply #3 –
:confused:

I've experienced the contrary. I've made encodes that sounded fine on my $75 PC speakers but clearly artifacted on my $300 Sennheisers. I've downloaded MP3 albums on various file-sharing services, burned them to CD and played them in my car and often didn't find much wrong with the sound. I've played that same CD on my home system and found it sounded clearly worse from artifacting. Of course equipment matters. I remember when I bought my HD-600s, I sat down and listened to most of my albums all over again because there was so much information I never heard before. It wasn't a matter of learning to hear better, but I had superlative equipment that revealed things that my old cheap $25 headphones couldn't.

Comparing MPC and OGG at same bitrate

Reply #4 –
Sorry about posting 2 threads i was just trying to get the biggest responce.
So can i assume from your lack of a vorbis/mpc responce that you Dibrom & mithrandir agree that mpc would be a better choice?

Comparing MPC and OGG at same bitrate

Reply #5 –
Quote
So can i assume from your lack of a vorbis/mpc responce that you Dibrom & mithrandir agree that mpc would be a better choice?

It might be that, or it might be the fact that almost idential questions to this have been asked on this board several times before. Take the time to read what's already on the board -- you might find some interesting discussions (and even a bit of flaming) in there.

Comparing MPC and OGG at same bitrate

Reply #6 –
Quote
Originally posted by mithrandir
:confused:

I've experienced the contrary. I've made encodes that sounded fine on my  PC speakers but clearly artifacted on my 0 Sennheisers. I've downloaded MP3 albums on various file-sharing services, burned them to CD and played them in my car and often didn't find much wrong with the sound. I've played that same CD on my home system and found it sounded clearly worse from artifacting. Of course equipment matters. I remember when I bought my HD-600s, I sat down and listened to most of my albums all over again because there was so much information I never heard before. It wasn't a matter of learning to hear better, but I had superlative equipment that revealed things that my old cheap  headphones couldn't.


Yes, of course switching from very bad to very good equipment will reveal a few things extra, but it's not the most important thing.

--
GCP

Comparing MPC and OGG at same bitrate

Reply #7 –
Quote
Originally posted by Garf
If you're talking pure quality and nothing else, MPC is currently still the way to go. I base this on my experience with listening to both codecs, and the fact that most people I think capable of doing decent listening tests agree.


It might be so for some artifacts like pre-echo and such, but how do you evaluate how well a codec encode high frequencies if your speaker doesn't even produce remotely-decent high frequencies at all? And how about sound staging and resolution?

Quote
Originally posted by Dibrom
I speak this point of view from personal experience.  For the last few months I've often been confined to performing listening tests in much less than ideal situations, specifically on a laptop with some not so great Sony headphones, yet I still am able to hear all the artifacts I heard before this, and I'm usually still able to hear most of the new ones which other people point out in different clips.


Is it possible that it's because you've been so well trained (or gifted) in this regard? For someone who is just starting to learn, wouldn't it be rather difficult under the same circumstances? (When I listen to MP3 on my laptop with cheap headphones, I hear a lot of noises, and I have a hard time distinguishing machine noises from encoding artifacts. )
tw101

Comparing MPC and OGG at same bitrate

Reply #8 –
Quote
Originally posted by tw101

It might be so for some artifacts like pre-echo and such, but how do you evaluate how well a codec encode high frequencies if your speaker doesn't even produce remotely-decent high frequencies at all? And how about sound staging and resolution? 


I'm not sure what your problem is here. If MPC would have problems with that, we'd have heard them.

--
GCP

Comparing MPC and OGG at same bitrate

Reply #9 –
Quote
Originally posted by Garf


I'm not sure what your problem is here. If MPC would have problems with that, we'd have heard them.

-- 
GCP


Arrrggg!! It's my fault. I quoted the wrong paragraph when I wrote my questions. (It's almost 7 in the morning, and I haven't slept yet.) Sorry about that.

What I really wanted to quote was this paragraph from you:

Quote
But this is a misconception. Hearing artifacts has very little to do with your equipment, but with training and physical abilities. If you don't hear difference on your current equipment, you likely aren't going to hear anything wrong on the new one either.


I may be rambling, but I'll try to explain a little.

Many users have the same idea as the orginal poster or me--i.e., "I may not be able to hear the difference now, but I may if I get better equipment".

Then came your (and Dibrom's) statement which seems to suggest that if I can't distinguish a, say, 128 k MP3 from the original CD on my crappy little PC speakers, I'm not going to tell a difference even if I play them on high-end equipment. (That's how I infer from your explanation, sorry if I got it wrong.)

This is counterintuitive to me, hence my questions in the post you quoted: are all artifacts equally easy to hear, regardless of equipment? My intuition is some artifacts might have less to do with equipment, while others have a lot more to do with it.

Put it another way: if 128 k MP3 lacks, e.g., some bass, but my $20 PC speakers aren't giving me any bass anyway, I certainly won't hear the difference between the MP3 and the original CD for this particular artifact. (Am I right in calling this an artifact. I seem to remember people say any difference is an artifact.) But I should be able to hear the diffenrence given better equipment, shoudn't I?

Now I'm really going to bed.
tw101

Comparing MPC and OGG at same bitrate

Reply #10 –
Quote
Originally posted by tw101
This is counterintuitive to me, hence my questions in the post you quoted: are all artifacts equally easy to hear, regardless of equipment? My intuition is some artifacts might have less to do with equipment, while others have a lot more to do with it.

Put it another way: if 128 k MP3 lacks, e.g., some bass, but my  PC speakers aren't giving me any bass anyway, I certainly won't hear the difference between the MP3 and the original CD for this particular artifact. (Am I right in calling this an artifact. I seem to remember people say any difference is an artifact.) But I should be able to hear the diffenrence given better equipment, shoudn't I?


The answer relates to the nature of psychoacoustic artifacts, vs the standard types of differences you would hear in various sound systems.

Psychoacoustic artifacts are very specific, and they are probably not what you would expect.  For example, "lack of bass" is an example of one of the many misconceptions people seem to have with the nature of these artifacts.  It's something that doesn't really happen in psychoacoustic codecs.  You might get distorted bass, but you aren't going to get a lack of it like you would on an inadequate system.  Most other psychoacoustic artifacts are like this as well.  Very rarely do you get a subtle change in the "feel" of the sound like you would with different systems, it's usually some sort of distortion, or something which appears to be added to the sound.  Pre-echo is one example of this.  Dropouts are another.  Noise pumping, knocking, fluttering, etc... these are the type of differences usually heard.

One possible exception to this would probably be the lowpass (or highpass) some coders use, but that's not really a psychoacoustic based artifact (unless the lowpass is adaptive and driven by the psymodel maybe).

At any rate, it's a little more complex than how I described it, but it basically boils down to the fact that the differences (flaws) that these coders show when they are not transparent, are very different from that of conventional audio.  The two are almost completely unrelated, and that's why equipment very rarely has anything to do with being able to hear them.

Comparing MPC and OGG at same bitrate

Reply #11 –
Quote
Originally posted by tw101
Is it possible that it's because you've been so well trained (or gifted) in this regard? For someone who is just starting to learn, wouldn't it be rather difficult under the same circumstances? (When I listen to MP3 on my laptop with cheap headphones, I hear a lot of noises, and I have a hard time distinguishing machine noises from encoding artifacts. )


For some artifacts which are completely noise based, like noise pumping or very very quiet pre-echo, it might be easier to learn on a cleaner system.  I don't think it will make a really huge difference though if you just attempt to train yourself to hear what constitutes an artifact, and what is just your machine.

Have you taken a look at ff123's page?  You might want to look into that if you'd like to know what many of the typical artifacts sound like.  Of course, many of the examples he gives are near worst case scenarios (to emphasize the effect and maximize audibility), but the information there is very applicable elsewhere.

Comparing MPC and OGG at same bitrate

Reply #12 –
Quote
Originally posted by Dibrom


Psychoacoustic artifacts are very specific, and they are probably not what you would expect.  For example, "lack of bass" is an example of one of the many misconceptions people seem to have with the nature of these artifacts.  It's something that doesn't really happen in psychoacoustic codecs.  You might get distorted bass, but you aren't going to get a lack of it like you would on an inadequate system. 


Thanks for the explanation. But if my speakers aren't giving me bass at all, even distorted bass won't be easy to hear, isn't it?

I guess I'm like many laymen who know very little about how psychoacoustic works, and like to blame equipment for everything, much like tennis players tend to stare at their rackets when they lose a point.

Could you (or others) please help clarify one more thing which has been bugging ME for some time. In the process of trying to convert some my self-appointed audiophile friends from CBR 320 to aps, I got the following argument the most:

The lower the bitrate, the lower the sound "resolution", and it won't sound good on good equipment, even if you can't hear any difference on PC speakers. It's like reducing CD's sample rate from 44.1 k to 22.05 k or less. The resolution is just not enough.

I've tried to counter the argument without much success.  In addition to the fact that I know too little to argue effectively, I guess the lack of persuasiveness has to do, at least in part, with my lack of self-confidence -- i.e., I suspect what they say is right!

To a layman, I guess it's only natural to think that when you throw out more than 80% of the data, the "resolution" has to suffer.

From your explanation, and from what I read in the archive and various websites, I got the idea that psychoacoustic audio compression doesn't work that way. But could you be more specific about how to counter that argument?

TIA.
tw101

Comparing MPC and OGG at same bitrate

Reply #13 –
Quote
Originally posted by Dibrom


Have you taken a look at ff123's page?  You might want to look into that if you'd like to know what many of the typical artifacts sound like.

Yes, I started "training" myself with ff123's samples the day I registered as a member here. But it's going very slowly, for I found it not very easy even for those "obvious" samples. (That's why I'm blaming my equipment.  ) But I'll keep trying. Thanks for the advice.
tw101

Comparing MPC and OGG at same bitrate

Reply #14 –
Quote
Originally posted by tw101

The lower the bitrate, the lower the sound "resolution", and it won't sound good on good equipment


This is wrong in 2 ways:

a) There is a minimal resolution where the sound will be transparent. Adding more above that is just a waste no matter what the equipment.

b) It's not only about allocating bits, but also _where_ and _how_ to allocate them. The Dibrom presets are much smarter in this than the normal settings, and hence will sound better even at a lower bitrate.

It's possible for an mp3 to allocate more than 320kbps at times, due to a feature called the bit reservoir. Thanks to this, the smarter presets can allocate more than 320kbps when they think it's necessary. CBR320 Can do it too, but it won't be as smart.

--
GCP

Comparing MPC and OGG at same bitrate

Reply #15 –
here are your basic options:

quality = mpc
compatability = ogg

you pick

Comparing MPC and OGG at same bitrate

Reply #16 –
Wrongo.

It should be:

Quality at >160kbps = MPC
Quality at <160kbps = Vorbis

Comparing MPC and OGG at same bitrate

Reply #17 –
Quote
Originally posted by tangent
Wrongo.

It should be:

Quality at >160kbps = MPC
Quality at <160kbps = Vorbis


well, it may just be me, but I don't see mpc being in any portable devices or such related, any time soon if at all.

Comparing MPC and OGG at same bitrate

Reply #18 –
Quote
Originally posted by Garf


This is wrong in 2 ways:

a) There is a minimal resolution where the sound will be transparent. Adding more above that is just a waste no matter what the equipment.


Thanks, but could you give me an idea what the minimal resolution for transparency is?

I think most people (me included) conceive this sound "resolution" as something similar to picture resolution. Hence the question: since the lowest resolution necessary for a picture to look good depends on how good the display equipment (monitor or printer) is, isn't it the same for sound? Or am I getting all wrong?

Quote
b) It's not only about allocating bits, but also _where_ and _how_ to allocate them. The Dibrom presets are much smarter in this than the normal settings, and hence will sound better even at a lower bitrate.


I understand that. But that's not my question. My question was: "is it true that better equipment won't allow me to discover more deficiency in an mp3 (or mpc or ogg).

Or let me put it this way: assuming for a particular clip, an MP3 created with LAME aps sound transparent to me on my crappy little PC speakers, will it still hold up when I burn it to audio CD and listen to it in a decent home audio set?

I think this is one of the most troubling problem for many users. Many times I've seen people say "aps sounds fine to me but I'll encode in apx or api anyway, just to be safe." I think these people have the same idea about resolution as I do.
tw101

Comparing MPC and OGG at same bitrate

Reply #19 –
Mathematically, "transparency" would be defined as minimum bitrate required to represent output signal with exact amount of noise as estimated by psychoacoustic model.

Using some "perfect" psychoacustic model and perfect algorithm for noise allocation, this range should be in range of 110-130 kbps/s for average music.

Now, you must add following things to that value (110-130 kbps)

- Imperfection of the noise allocation algorithm (every codes has it!)  + 10-50 kbps

- Imperfection of the psychoacoustic model +5-20 kbps

- Your personal difference from hearing model  -10 - +30 kbps


So, this value is not equal between codecs, between implementation of codecs, and between people - usually, for standard-based codecs, like MPEG, some average value is estimated and I can give you that results

MP2 - 256 kbps
MP3 - 160-192 kbps
AAC - 128-132 kbps

(These results are based on highest quality implementations, not every implementation i.e. brand is this good)

But these values are purely based on averaging, and might differ between listeners and listening conditions, for example MP3 fails on some files even on 320 kbps - i.e. some people could hear artifacts...

I am sure that people responsible for MPC and Ogg development could give some approximated results, but again - this might be very different for your ears and taste


-- Ivan

Comparing MPC and OGG at same bitrate

Reply #20 –
Quote
Originally posted by meff
well, it may just be me, but I don't see mpc being in any portable devices or such related, any time soon if at all.

If you need portable device compatiblity, then use MP3 or WMA8. This discussion is about quality. Even though Ogg and MPC have little or no hardware support, you can't overlook the fact that both are the best lossy encoders for anything above, say, 96kbps. Some of us will not forfeit quality for compatibilty with portable devices.

Comparing MPC and OGG at same bitrate

Reply #21 –
Quote
Originally posted by mithrandir

If you need portable device compatiblity, then use MP3 or WMA8. This discussion is about quality. Even though Ogg and MPC have little or no hardware support, you can't overlook the fact that both are the best lossy encoders for anything above, say, 96kbps. Some of us will not forfeit quality for compatibilty with portable devices.

Erm.. Ogg Vorbis is also the best below 96kbps

Comparing MPC and OGG at same bitrate

Reply #22 –
iirc, mpc preforms pretty bad at low bitrates.

as far as this thread being about quality, one still has to take in consideration about compatability, and its future.

i'm not trying to talk biased, but mpc seems to be knit to a niche market, with a sketchy future. ogg has events lined up, and upon 1.0-final there will be many advancements in alot of areas in compatability.

as far as i am concerned, wma and mp3 are shit.

if your not going to sacrafice quality for compatability, why not use a loseless codec?

im sorry.. it just seems clear to me
ogg for compression and future compatability
loseless for archival
most people use mpc at xtreme/insane anyways so its kind of pointless.

i can see mpc fading out sometime in the future..

but thats just me

Comparing MPC and OGG at same bitrate

Reply #23 –
Quote
Originally posted by Ivan Dimkovic

some average value is estimated and I can give you that results

MP2 - 256 kbps
MP3 - 160-192 kbps
AAC - 128-132 kbps

(These results are based on highest quality implementations, not every implementation i.e. brand is this good)


I suppose those values are valid for AACdemo2.2
What would be AACenc 2.15's values?

Comparing MPC and OGG at same bitrate

Reply #24 –
Everybody has different encoding needs and there currently isn't a single format that is able to satisfy everyone. However, I believe that the number one concern of the hydrogenaudio forums is quality. If it wasn't, then we'd just stick with MP3 and close shop.

MPC may offer lessons on how to achieve startling transparency at a moderate bitrate. Ogg may offer lessons on how to use open source to increase the chance of an encoder's long-term survival. It's true that MPC is not well-suited for low bitrate encodings and at higher bitrates, Ogg has a greater occurance of artifacts than MPC. They are rather complimentary in this regard.

I don't worry about compatibility or the future, frankly, with MPC. I basically encode the audio CDs that I own so that I can listen to them on my PC with the convenience of a media player like WinAMP (particularly convenient for my work PC), rather than having to shuffle physical media in and out of my stereo. On the go, I just play my original CDs in my portable CD player...I've been doing that for over a decade. As for the future, as long as the MPC decoder is open source (which it is), then I don't worry about my files. Plus, the MPCs I have are from audio CDs I physically have, so I really don't have anything to lose.

As for lossless, there's still no comparison. 200-250kbps (for lossy) and 850-950kbps (for lossless) are in just totally different ballparks. Sure, going lossless would mean keeping it perfect, but this is at the expense of quadruple the disk space. A 60GB hard drive could probably hold all of my 600-700 CDs...because this would be at 225kbps, not at 900kbps. 240GB worth of disk storage - as required for 900kbps lossless - is sufficiently more expensive. Given how transparent an encoder like MPC is, it's just not worth it since I have the original audio CDs on hand should I ever want to "start over".

Frankly, Ogg isn't "there yet". At least with MPC, my encodes sound stable/mature, i.e. hardly any room for improvement. Say I went with Ogg now, to buy into the heralded longevity and compatibility. Well, then I'd have files that weren't exactly perfect (i.e. less perfect than MPC) and would probably want to reencode later as Ogg matures. What am I losing, then, if I go with MPC? I get the quality now. If you are going to reencode in the future anyway, then stick with what's best today. But then again, I don't think I'll have to reencode my MPCs because their quality is already so high.

I could be the only person in the world using MPC for all I care. As long as I can encode and decode, then I am satisfied. Back in the DOS days, I recall shunning PKZIP for ARJ, since the latter was superior in terms of compression and flexibility. I can still decompress all those ARJ files today (not that they contain anything useful anymore  ).

But again, everybody has different needs so people need to go their own route. We seem to have a tendency for imposing a "mandate" on what encoder to use, i.e. if somebody else uses a "competing" encoder, then they must be ignorant and misguided. That's for the people who still encode 128kbps Xing MP3s. :eek: