AC97, is it ok, or does it suck?!

[Amused]

In all the boxes I've built for family and those I've built to be used in my GF's office, all have used Intel mobos or Asus mobos with Intel chipsets. Most of those have had onboard sound (to save a few bucks) and were perfectly fine for the environment they were being used in and the speakers they used with them. [edit]To the best of my recollection, all had AC97 onboard sound.

To be honest, if someone is going to be using cheap two piece speakers sets, I can't see any reason in adding the cost of a sound card when onboard audio is free and has never sounded that bad on the Intel/Intel based boards I've used.

I wouldn't use it for someone that does a lot of multimedia work or gaming, but for the other 80-90% of folks out there, I think it's fine.

 

[PH0ENIX]

Is it just me, or have a lot of people missed the thread title?

---
AC97, is it ok, or does it suck?!

NOT

Onboard audio, is it ok, or does it all suck?!
---

So If I read correctly, anything that doesn't use the AC97 CODEC, doesn't apply to this question.

So C-Media, and Intel's solution (whatever it is they're using) are somewhat irrelevant, yes?

I dont think anyone is trying to say that onboard audio is shonky and therefore useless in every case - just that the AC97 CODEC is very, very poor.

Oh, and winXP is supposed to support ViA's implementation of AC97 natively.
3 ViA boards later, still only the SiS worked out-of-the-box.

It seems like XP tries to assign the driver for the audio controller to the resources for the game port, and vice versa.
the only driver that worked for the Via AC97 was the 'Avance AC97'.
I've tried a few different releases of the 4-in-1, and every one either didn't support XP, supported XP but didn't include the audio driver for XP, only 9x/nt/2k, or contained the driver but it didn't function (it exhibited the same resource issues as the driver bundled with XP).

Like I keep saying - for a cheap audio solution it's great, you dont have to purchase additional hardware blah blah and if you're only using desktop speakers you probably won't notice, but if you enjoy listening to music, I recommend at least testing an alternate card.

Ph0.

 

[unclebud]

that ac97 on that msi board of mine sounds bad even in my hedpones -- the sounds lag, ie: demo four sounds like a dubbed action film
plus mouseover in menu sounds like bongz bongz instead of ding ding

 

[BFG10K]

Generally speaking I avoid all onboard components especially sound controllers and video controllers.

 

[bozo1]

This thread is totally riduculous. Read the damn thread title - it wasn't "Is Audigy better than my built in sound?" The simple question was "Is it ok, or does it suck" and that is totally subjective depending on what the user needs/wants. All integrated audio works. It may not work as well as you want, but it works. Does it suck? Of course not, how can something suck when it does exactly what it was designed to do?

Anyway the links to stupid articles pushing one persons viewpoint is pointless.

Comments like: "Eactly. If you have an actual AC'97 codec you will have the results listed no matter what board you have." is of course false.

How can you compare AC'97 to anything else? AC'97 is simply a standard. Each implementation of it will perform differently. This is like saying a 7200 RPM SCSI-2 hard drive will perform as well as a 15000RPM Ultra160 SCSI drive because both are SCSI. That is simply untrue. Do you realize that SB Live, SB Audigy, SC Turtle Beach, Hercules GTXP are ALL AC'97 compliant?

Coming in and crapping on someone elses thread with comments that don't pertain to their question, linking to articles whose technical accuracy is suspect at best, or making totally unfounded statements is just simply rude and is not thought of well here.

 

[spanky]

it's ok.

 

[bozo1]

Originally posted by: spankyOO7
it's ok.

ok

 

[PH0ENIX]

Sure you can - we're talking about the sound quality of the examples of the AC97 specification that we've seen so far.

or at least, _I_ am.

And it's not like saying "7200 RPM SCSI-2 hard drive will perform as well as a 15000RPM Ultra160 SCSI drive because both are SCSI"

It's more like saying "A voodoo 5 5500 will outperform a Geforce2 MX 400 regardless of brand"
(this is an EXAMPLE not a factual statement so shush if ur thinking of arguing)

Different brands and board designs of GF2s perform differently. But you can make a blanket statement because the V5 is much faster (EXAMPLE!).

Same applies - not all AC97 may be as bad as the ones described by myself and others, but...
As a general rule, if a majority of components built to a certain standard do not meet expectations, do you continue to support the standard?

Im not going to side with anyone here, i'm just going to say the same thing; AC97 quality is not what it should be.

And going by Bozo1;
The AC97 standard must have been designed to provide hi-noise, low bitrate (well thats the best way to describe the sound it puts out) audio, while still recieving minor, normally un-noticeable electrical interference - with incredible clarity

Again - if it's your own system, try another card; if you dont notice great, if you do, great.

But is it WORTH trying an alternate card?

That all depends. If you listen to rap, then it's not like the quality of your music is ever going to improve now is it?

Long Live Slayer, down with AC97, thankyou - drive thru.

 

[Amused]

Originally posted by: PH0ENIX
Is it just me, or have a lot of people missed the thread title?

---
AC97, is it ok, or does it suck?!

NOT

Onboard audio, is it ok, or does it all suck?!
---

So If I read correctly, anything that doesn't use the AC97 CODEC, doesn't apply to this question.

So C-Media, and Intel's solution (whatever it is they're using) are somewhat irrelevant, yes?

AFAIK Intel uses AC97. Or at least all the Intel and Intel based mobos I've used lately have had AC97. Maybe I should have made that clear in my post.

 

[gf4200isdabest]

I'll settle this guys!

AC97 is okay but it does kinda suck...

 

[merlocka]

MSI K7T pro2-RU (AC'97) -1.1dB

Can you explain how this measurement was made? Since your results are in dB I was assuming you were inputting a reference tone and measuring noise on the tone but you also mention that there is no signal present... so what is this -1.1 vaule referenced to? dB is unitless.

Also, you didn't mention which codec the MSI board uses. There are several different ones. This article here shows noise levels significantly different than you measured on the MSI board, yet they are both AC'97 solutions. So all AC'97 codecs are not the same?

so for the purposes of this discussion, AC]97 codec in particular, they're all the same.

I disagree, and I guess that's my main point. There are a ton of different companies which sell codecs for AC'97 applications. Analog Devices AD1885, VIA VT1611A, and Avance Logic ALC650 are just a few.

so, 7757524, my question is...

Do you understand the AC'97 specification? There is a nice explanation from the link above,

AC'97 is a concept which is not connected with a motherboard. It is a specification from Intel for an audio codec architecture (AC - Audio Codec) and digital interface AC-link, which connects this codec with Digital Controller. The latter can be realized in different ways: on a program level (only a bus controller is integrated into a chipset), on a soft hardware level (HSP, DSP), on a hardware one (DSP). In this case a controller or even the whole sound card are said to be implemented in the AC'97-standard (AC'97 compliant). The most of modern sound chips correspond to it: EMU10K1, CS46x0, FM801AU, YMF7x4, SSA7785, ES1970, AU88x0 etc.

Hmmm, that's pretty close to what I was thinking... except for the fact that even the EMU10K1 complies to the AC'97 specification. Wouldn't the EMU10K1 be the hardware in the Soundblaster Live which you included in your "review"?

The MSI KT7 Pro2 might have crappy audio, and many implimentations of AC'97 might be poor, but there's nothing wrong with AC'97. Of course anything integrated will, in general, be cheaper and "worse" than discrete AIC's. The reason for integrating it is to lower costs for OEM's.

And I'm pretty sure there is not much wrong with this El Cheapo FIC 810 mobo that I have sitting here with the AD1885. It sounds just as good as my SB live. I can assure you there is less than -11dB of noise being generated and there is no crosstalk being generated "between the codec and the connector on the back" lol.

 

[paralazarguer]

The tests were done using Algotest. The negative values represent noise levels in the line. -10dB represents more noise than -50db. Keep in mind that 6dB is roughly 4 times louder than 2 db and works on a similar scale as Ceclus. So, negative values are not as large as they seem. It's true that AC'97 is a specification rather than a chip. However, some boards identify their audio chips only as AC'97 advanced audio. This is what's written on the chip also. I can only assume that this is a codec of it's own similar to the C-media chip also measured. When I refer to AC'97 I'm reffering to this chip rather than the specification. When I say that all ac'97 will bench the same, I'm reffering to this unknown codec which goes by same name. Other integrated solutions such as the codecs you listed which go by the AC'97 spec will bench differently.

 

[paralazarguer]

between the codec and the connector on the back

I couldn't think of a better way to say it wasn't internal to the codec but the shielding in the connectors and EMI.

 

[Telinar]

Got amplified speakers that were made by someone more reputable than 'wang chung corp'?

Is that supposed to imply inferior quality? Oh c'mon. You can do better.

Anyway, after plugging in my nice speakers, the C-media on my Soyo mobo sounds decent. Sure I can tell the difference between this and my computer with the SB Live!, but the lesser quality isn't a huge sacrifice to save the $40 I would have spent buying a sound card.

 

[PH0ENIX]

Of course that's what it implies, Wang Chung was what they called that white kid who wanted to be asian in 'not another teen movie'.

Henceforth, tryhard, henceforth, not the real deal, and therefore - inferior quality.

And by some bizarre co-incidence you're right, I could do better - but for this thread im surprised i've made this much of an effort

 

[merlocka]

The negative values represent noise levels in the line. -10dB represents more noise than -50db. So, negative values are not as large as they seem.[/i]

dB is a logrithmic ratio between two numbers, but your data is not telling what the noise is relative to. You state that the MSI K7T pro2-RU (AC'97) produces -1.1dB of internal noise. If this is relative to the output, then you are saying that the device outputs nearly equal levels of singal and of noise (-1.1dB means that the noise would be about 90% the level of the signal). This would obviously be quite audible, disturbing, and it would start raining cats and dogs.

Keep in mind that 6dB is roughly 4 times louder than 2 db and works on a similar scale as Ceclus.

OK, beginning to doubt you a bit here... with power, 6dB represents a ratio of 4 but with voltage signals 6dB is a ratio of 2.
So in this case (I assume you not measuring noise power), 6dB would be about twice as large as 0dB. Tis irrevalent though, becuase you haven't explained how the measurment was made. And what the hell are you talking about wrt Celcius? dB's are a logrithmic system for measuring ratios. It's used because in certain systems the ratios would be meaningless if left in a linear scale. Celcius? Eh?

FYI -
The P4S533 uses a C-Media CM18738 codec, while the MSI K7T pro2-RU uses the VIA VT1611A codec. You might want to add that to your article.

I couldn't think of a better way to say it wasn't internal to the codec but the shielding in the connectors and EMI.

EMI - Electro Magnetic Interference, won't cause crosstalk in an audio path or connector on a motheboard. The audio won't "radiate" from one pcb trace to the other. Crosstalk occurs as a non-linearity of the mixing block in the codec. Sure, one billionth of a percent of the signal somehow magically jumps from one connector to the other but in reality all of the crosstalk is generated internally. Audio doesn't fly well.

 

[holdencommodore]

Just thought I should throw this in. It talks about Analog Devices new 1980 SoundMAX codec. It doesn't give much information, but gives a idea of where they will be heading with onboard audio.

Cheers

 

[Telinar]

Ok. I was mistaken. It just sounded to me there was something else being implied too. Anyway, my bad.

 

[Kingofcomputer]

take a look cmedia's website
http://www.cmedia.com.tw/

they call the CMI8738 as PCI 5.1ch Audio Chip,
CMI9739 as AC'97 6ch CODEC.

CMI8738 has way more pins than CMI9739.

 

[VirtualLarry]

Originally posted by: merlocka
In this case he wans't simply disagreeing, he stated that I didn't explain why...I did and he obviously didn't understand the explanation. Then I pointed him to it.

LOL, is it so obvious that I'm a dumbass? I need to learn to hide it better.

Some of the statements that you have made rergarding onboard sound using the AC97 specification have been...

No. AC'97 is horrible. Read this article which specifically lists AC'97 and how it compares to the live, audigy and Mia

The linked article does indeed compare the MIA, Audigey, and Live. Data is provided for DC offset and internal noise as well as other useful info. There is no specific data for the AC97 is there? Perhaps I am totally oblivious, can you cut and paste something from that link other than If we were comparing the Live to AC?97 integrated audio, we would have been doing back flips over the superior quality of this card. which provides specific data on a board using AC97 which backs up your later claim of sound will be muddled and squashed because of the 50+dB DC offset

I don't doubt that some boards using poor implimentations of AC97 could have this offset, which would certainly provide bad dynamic range, but this article you have linked to doesn't verify that. In addition, are you willing to claim that no implimentation of the AC97 specification could be absent of such significant DC offset?

I'm not doubting your audio 1337ness, you seem quite confident of it. I'm just curious why you keep linking to that article which does not specifically address a AC97 implimentation as you state.

For the record, the Aureal Vortex2 SQ2500, has an "AC'97-compatible CODEC" (interface), but IMHO it sounds better than an SBLive. I helped set my friend up with one for his home theater setup, and it sounded incredible, I didn't notice any significant reduction in dynamic range, etc.

The problem is with cheap DAC/ADC or CODEC implementations, not with the AC'97 standard itself, per se.

I do think that for the most part, integrated audio (AC'97 or no, it really doesn't matter) generally sucks as a rule. Generally integrated audio is based on price, not performance, and on the motherboard, there is little room for good isolation for analog signals.

 

[VirtualLarry]

Anyway, a blanket statement like "all c-media audio sucks" is obviously false. In fact, the CMI8738 has nothing to do with sound quality. Oh it's the sound chip, of course, but it's the DAC, ADC which is attached to it that matters. In the case of AC'97 audio you're dealing with a hardware CODEC which is a bit different. Since AC'97 IS itself a CODEC that would imply that all boards using AC'97 will suffer from exactly the same test results. Exactly the same codec: exactly the same results. (remember the codec is a combined DAC and ADC on one chip) The statement that all AC'97 "sucks" is correct

That's where you are unfortunately wrong. AC'97, as I understand it, is a hardware-level standard interface for CODECs. It is NOT a CODEC in and of itself. Rather, various companies make AC'97-compatible CODECs, that can be integrated into products. Much like the various chips used on ethernet cards, with seperate MAC and PHY chips, although that is going away, as most ethernet cards have fully-integrated chipsets now for cost reasons.

 

[VirtualLarry]

Originally posted by: bozo1
This thread is totally riduculous. Read the damn thread title - it wasn't "Is Audigy better than my built in sound?" The simple question was "Is it ok, or does it suck" and that is totally subjective depending on what the user needs/wants. All integrated audio works. It may not work as well as you want, but it works. Does it suck? Of course not, how can something suck when it does exactly what it was designed to do?

Anyway the links to stupid articles pushing one persons viewpoint is pointless.

Comments like: "Eactly. If you have an actual AC'97 codec you will have the results listed no matter what board you have." is of course false.

How can you compare AC'97 to anything else? AC'97 is simply a standard. Each implementation of it will perform differently. This is like saying a 7200 RPM SCSI-2 hard drive will perform as well as a 15000RPM Ultra160 SCSI drive because both are SCSI. That is simply untrue. Do you realize that SB Live, SB Audigy, SC Turtle Beach, Hercules GTXP are ALL AC'97 compliant?

Coming in and crapping on someone elses thread with comments that don't pertain to their question, linking to articles whose technical accuracy is suspect at best, or making totally unfounded statements is just simply rude and is not thought of well here.

Thank you! You're the first person with any sort of apparent technical clue to post in this thread, besides NFS4. Apologies for not reading your post, before I posted mine saying essentially the same thing, but not as well.

 

[absolutiza]

it is ok... IMHO

 

[mithrandir2001]

Everyday at work, I listen to my MPCs and OGGs through the on-board AC97 chip in my work PC. Compared to the TB Santa Cruz in my home rig, the AC97 is markedly inferior in quality. The biggest difference is the lack of dynamic range...music sounds flat and generally lifeless. The AC97 also is weak in the extreme bottom and upper frequencies. It's like listening through a Bose loudspeaker: no highs, no lows and it doesn't sound good loud. Mind you, I connect my Sennheiser HD-600 headphones directly to the headphone jack on each PC so perhaps my work PC cannot drive these 300 ohm headphones properly. However, the Santa Cruz can make them sing.

I'd add a sound card to my work PC, but last week I threw in a Matrox Millenium G200 (want my 1600x1200!) to replace the crappy on-board Intel video. Tired of upgrading the stupid work machine.

 

[snow patrol]

Originally posted by: Hardware
I am sure your ?5 speaker sounds great with a SB Platinum!

omg hardware! You're alive!

Come here and give me a hug