My current desktop - Phenom X3 720 undervolted to 0.75V @ 800MHz and 1.2V @ 2.8GHz, HD5670
Idling - 85W
Linpack - 150W
Furmark + CPU forced 800MHz - 140W
Linpack + Furmark - 190W
Linpack w/o undervolting - 173W
Linpack @ 800MHz w/o undervolting - 107W
Linpack @ 800MHz w/undervolting - 99W
TV rig - Athlon64 3400+ undervolted to 1.1V @ 1.9GHz, 9300GS (a.k.a 8400GS)
Idling - 65W
Prime95 - 80W
Previous previous previous rig - Pentium 4 1.8A Ghz, GeForce2 MX400, factory tuned
Idling - 55W
Load - 70~80W
W A R N I N G !
W A R N I N G !
This page is full of non-facts and bullsh!t, (just like the internet and especially forums and other blogs), please do not believe entirely without exercising your intellect. Any resemblance to real things in reality is purely coincidental. You are free to interpret/misinterpret the content however you like, most likely for entertainment, but in no case is the text written on this blog the absolute truth. The blog owner and Blogger are not responsible for any misunderstanding of ASCII characters as facts. *cough* As I was saying, you are free to interpret however you like. *cough*
Wednesday, December 29, 2010
Monday, December 13, 2010
Where you place your volume control - it matters
Self-loopback of recording interface (TASCAM US-144) -

Except the SNR of TASCAM US-144 is actually around 102-103dB; it is bottlenecked by its analogue out. Strange, I know.
High distortion values, probably from the mic preamp. Fortunately we're only looking at noise today.
Now, an explanation of all the candidates -
- DAC out (no volume control available), software volume control used, input of TASCAM US-144 @ max
- Preamp front headphone connector out, software volume control used, input of TASCAM US-144 @ max
- Preamp at volume required for 1dB, software volume control @ max. TASCAM input @ max
- Preamp @ maximum volume, software @ maximum volume, TASCAM input adjusted to get 1dB

Detailed results here:
Where you put your volume control - RMAA results
As you can see, even though my DAC and it's preamp stage is capable of over 100dB SNR (and possibly bottlenecked by the recording interface), if the input to the preamp is low volume noise gets amplified along with the tiny signal and kills SNR.
The second and third tests are to compare between using software to reduce to volume entering the preamp and using the potentiometer to achieve the same thing. As you can see, there is not much difference. The bottleneck is not the quantization noise that occurs at low digital volumes, but the huge 50Hz hum from the transformer plus other noises.
The first test is to test the typical scenario where one chooses to control volume at the digital source. For this, I put WDM volume @ 25/100 and master @ 6/100 in order to give -1dB with the TASCAM @ max volume. If those scales are linear, I would've lost around 6-7 bits of resolution and added some more noise, but the leftover signal can still theoretically do better than 81dB, which is full of mains hum.
So conclusion - do not output volumes too low from your source, or risk having noises from inside the source dominate.
Unfortunately, sometimes the default volumes of DACs are too loud, I have to put my Windows' volume @ 25/100 and 8-16/100 most of the time, even with my amp @ 9 o'clock.
But does that really matter?
Maybe, not really.
Sneak peek:

Except the SNR of TASCAM US-144 is actually around 102-103dB; it is bottlenecked by its analogue out. Strange, I know.
High distortion values, probably from the mic preamp. Fortunately we're only looking at noise today.
Now, an explanation of all the candidates -
- DAC out (no volume control available), software volume control used, input of TASCAM US-144 @ max
- Preamp front headphone connector out, software volume control used, input of TASCAM US-144 @ max
- Preamp at volume required for 1dB, software volume control @ max. TASCAM input @ max
- Preamp @ maximum volume, software @ maximum volume, TASCAM input adjusted to get 1dB

Detailed results here:
Where you put your volume control - RMAA results
As you can see, even though my DAC and it's preamp stage is capable of over 100dB SNR (and possibly bottlenecked by the recording interface), if the input to the preamp is low volume noise gets amplified along with the tiny signal and kills SNR.
The second and third tests are to compare between using software to reduce to volume entering the preamp and using the potentiometer to achieve the same thing. As you can see, there is not much difference. The bottleneck is not the quantization noise that occurs at low digital volumes, but the huge 50Hz hum from the transformer plus other noises.
The first test is to test the typical scenario where one chooses to control volume at the digital source. For this, I put WDM volume @ 25/100 and master @ 6/100 in order to give -1dB with the TASCAM @ max volume. If those scales are linear, I would've lost around 6-7 bits of resolution and added some more noise, but the leftover signal can still theoretically do better than 81dB, which is full of mains hum.
So conclusion - do not output volumes too low from your source, or risk having noises from inside the source dominate.
Unfortunately, sometimes the default volumes of DACs are too loud, I have to put my Windows' volume @ 25/100 and 8-16/100 most of the time, even with my amp @ 9 o'clock.
But does that really matter?
Maybe, not really.
Sneak peek:
Wednesday, December 1, 2010
End of blogging spree
If you noticed, I've been posting regularly, every two days.
I actually tried to make it regular for this month. Had 2 to 4 articles of buffer at any one time.
Well, that's the end of this. Time to study for exams.
This post was completed on 22/11/10
I actually tried to make it regular for this month. Had 2 to 4 articles of buffer at any one time.
Well, that's the end of this. Time to study for exams.
This post was completed on 22/11/10
Monday, November 29, 2010
Games and 2GB of RAM
We've entered the age of 64-bit computing. Well, at least most of the new computers have an OS and a processor that are capable of 64-bit computing, but nobody's using it.
And with 64-bit address space, we finally break through the memory limit to use 4GB (otherwise capped @ 3GB by Windows XP, for a somewhat related, somewhat unrelated reason) and 8GB and beyond.
But the performance reviews of games still show that 2GB is the sweet spot for price-vs-performance - games are still happy with 2GB of system ram and increasing further yields much less benefit.
Thinking about the 32-bit Win XP, a reason flashed across my mind.
Windows XP allows each process to use a maximum of 2GB of virtual address space.
And most games are written to be compatible with XP since a lot of people still use it.
So...... 2GB limit it is.
The good part is this shows that current-day programmers can still make efficient programs, if they want to, or have to now.
The bad part is this shows that games are super bloated.
And with 64-bit address space, we finally break through the memory limit to use 4GB (otherwise capped @ 3GB by Windows XP, for a somewhat related, somewhat unrelated reason) and 8GB and beyond.
But the performance reviews of games still show that 2GB is the sweet spot for price-vs-performance - games are still happy with 2GB of system ram and increasing further yields much less benefit.
Thinking about the 32-bit Win XP, a reason flashed across my mind.
Windows XP allows each process to use a maximum of 2GB of virtual address space.
And most games are written to be compatible with XP since a lot of people still use it.
So...... 2GB limit it is.
The good part is this shows that current-day programmers can still make efficient programs, if they want to, or have to now.
The bad part is this shows that games are super bloated.
Saturday, November 27, 2010
Random thought: Data Volumetric Flow Rate
100Mbps - that's pretty easy to hit nowadays.
If each bit is a physical 1 or 0 that is 1cm tall 1cm wide 1cm long,
That's 100 million of them coming out of the cable every second.
Or 100 thousand litres every second.
Or enough to fill your room in less than a second.
How did they use to be able to store bits on paper and punch cards?
Information explosion, it's scary.
If each bit is a physical 1 or 0 that is 1cm tall 1cm wide 1cm long,
That's 100 million of them coming out of the cable every second.
Or 100 thousand litres every second.
Or enough to fill your room in less than a second.
How did they use to be able to store bits on paper and punch cards?
Information explosion, it's scary.
Thursday, November 25, 2010
New marker for measuring show pace?
There are two types of shows
- the type that you can watch while eating, and
- the type that can't
because things happen too fast you will miss something when eating. Yes I do enjoy my food.
- the type that you can watch while eating, and
- the type that can't
because things happen too fast you will miss something when eating. Yes I do enjoy my food.
Labels:
Anime
Subscribe to:
Posts (Atom)