is the alc888 a good dac ?
if not what should i upgrade it with ?
>inb4 $500 hardware
be resonable.
why there are /hpg/ threads but no /dac/ threads ?
>>54378284
>is the alc888 a good dac ?
I thought it was a codec, not a dac.
>>54378284
>but no /dac/ threads ?
Because it's (mostly) a placebo unless you're going for multiple channels with a goal of studio quality. There's also no way of saying which on board DACs perform the best as this varies wildly with how the mobo manufacturer uses them. Using NVidia graphics cards is also bad for your built in DAC sound quality as they pull more power through the mobo than using an IGP or an AMD card.
All in all, the minor difference in sound quality isn't worth the hassle. If you insist on having a separate DAC then just get a USB one.
>>54378363
interesting, thanks for the tip
bump.
>>54378284
Because many DACs are audibly transparent. Anything past 4-bit delta-sigma is good enough desu.
>>54378834
what bout SNR ?
It's decent and depends on MB construction and internals to conquer noises when moving mouse etc.
See also - https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=auto&tl=en&u=http://pclab.pl/art17707-12.html&usg=ALkJrhh5-Wo79z1wyYeH8O1r8O0v-2hDuA
>>54378284
/hpg/ discusses DACs as they are a necessity with computer audio. Today /hpg/ is infested by a few placebophiles so I suggest you stay away and lurk more. A dedicated DAC thread would be extremely boring as audio DA-convertes are already so good that there's little to no reason to get something fancier than your on-board in most cases(at least as far as audio quality is concerned) and people here probably don't know a thing about DACs in other circuits. It would yield no discussion other than audiophool nonsense.
As for the ALC888 codec, you can probably find a nice spec sheet for it. Problem with on board solutions is mostly when you apply an impedance varying load to the output such as headphones for it to drive. Codec might be really good at driving most headphones but motherboard manufacturers commonly like to put resistors into the output as a poor overload protection which results into limited output power and high output impedance which can have large effect on how the headphones sound. This stuff is never specified by the manufacturer either. Some might have audible problems like hiss and EMI.
Ask yourself this: Why do you need a new DAC and for what purpose? Usually when people talk casually about DACs it's often a lot more than that(preamp, headphone amplifier, digital processor etc).
>>54379299
thanks for clearing that up. i personally have no problem wirh my alc888, even tho i'm using 250ohm headphones, the sound is loud and clear, and as far as i'm concerned there are no background noise at all.
i was just curious about other audio solutions
>>54378284
ALC888 is a codec dummy.
>>54379299
I have a modded Audigy 2 ZS and I bet it's better then most of the onboard shit.
>>54380212
wtf you're talking about ?, they're virtually identical
Bought a $25 USB DAC and it definitely sounds better than my alc888. Still not as good as my iPad mini 4 tho.
>>54380750
what DAC did you get ?
>>54380648
But you see the Audigy 2 ZS is slightly better.
I replaced the opamps and capacitors on mine too.
>>54380188
Not OP, but I never understood this. Is it a hardware codec? What exactly does the hardware do if the codec is simply software?
>>54380856
1db difference isn't noticeable, hell its barely measurable !.
you have no idea what you're talking about
>>54380797
http://www.amazon.com/Signstek-Coaxial-Converter-Decoder-Analogue/dp/B00FEDHHKE
>>54380648
wht program is that?
>>54380941
rightmark audio analizer
>>54380916
>realtek
>no 96khz
>>54380930
it uses a PCM2704 which is according to the datasheet a 16bit dac, i doubt it sounds better, but it does have a good SNR according to the datasheet.
also obsolete.
>>54381031
the alc888 can go up to 192khz on 24bit
>>54381031
>96 kHz
Pls go away.
>>54381031
For DERPBASS IS NOT NEEDED, GOY
>>54381080
>>54378834
How many bits does it take for digital volume control?
>>54381439
1 bit gives you about 6 dB range. So 16 bit gives you a dynamic range of 96 dB. Pain sets in at 125 dB.
>>54381545
So if I have a 16 bit DAC and set my volume half way, I loose half my dynamic range?
>>54381545
thats not how it works.
>>54381608
That is just for the information you can save. If you set your DAC to a lower volume, your range just shifts (until noise needs to be considered). So when you set the volume lower or higher, you just shift the range. And if I'm not mistaken, this is linear in the dB scale. So if a 100 dB signal is changed to a 90 dB signal, you also change a 20 dB signal to a 10 dB one.
>>54378284
It's a budget solution found on lower end motherboards. Any decent higher end board has at least ALC1150 these days.
>>54381545
fef you have no idea what you're talking about
>>54381656
I'd love to learn how it really works. This supports what I read somewhere else with the 6 dB per bit:
https://en.wikipedia.org/wiki/Audio_bit_depth#Dynamic_range
>>54381608
not exactly, i'll give you an example.
lets say you have an 8 mega pixel picture, and your screen is 1280x720.
you can still see the picture however allot of pixels will be merged.
>>54381796
see >>54381801
the picture example is well suite for it, i think.
>>54381801
But a screen is a digital way of output. A DAC gives you analogue output.
Best codec is ALC898
>>54381796
>16-bit int
>96 dB
>in any shape or form divisible
According to you every bit in a uint32_t gives me 134217728 different steps. Congrats at failing basic math.
For reference:
2^16 = 65.536 steps
96 dB / 65.536 = 0.00146 dB (beyond placebo)
>>54381888
i wonder if i can desolder the codec on my motherboard and solder a better one on it.
it would be technically possible since most dac use the i2s bus, but i think it would be a giant pain in the ass.
>>54381844
everything you see or hear is anolog, humans don't work in binary
>>54381681
If you have 12 bits for the volume, you only have 4 for everything else? Is that why when you turn the music up, it sounds better? Wouldn't a better solution be to place the volume control in the current stage after the accs, rather than in the voltage stage? You wouldn't loose any detail, right?
>>54381942
How do you get the 134217728 different steps? Also, you don't account for the fact that 10dB difference at a low volume/signal are a much smaller step than at a high signal.
>>54381942
Some DACs only have 90db, while most have about 70db of control range.
>>54382006
>the 134217728 different steps
Was meant as 134217728 steps per bit as per his formula (maximum/"bits"). Which is ofc bullshit as every extra bit more in a int gives you twice as many values (or range if you keep the delta even).
>the fact that 10dB difference at a low volume/signal are a much smaller step than at a high signal
Again wrong. Decibel describes a ratio and not absolutes. A 20 dB difference will always be a factor 100 difference in power and a factor 10 in amplitudes, no matter if 0-20 or 80-100.
>A 20 dB difference will always be a factor 100 difference in power and a factor 10 in amplitudes, no matter if 0-20 or 80-100.
>>54382211
That's exactly what I mean. The factor is the same, the difference in amplitude (sound or voltage) isn't. Log scales, motherfuckers.
