Jump to content
IGNORED

DIY Project High Performance Audio PC with high quality wiring


Recommended Posts

37 minutes ago, StreamFidelity said:

... "Unfortunately" the direct CPU power supply from a Keces P8 with 12V is still better than the complete power supply from the Taiko ATX. That can change with a better LPS.

Despite the quality of the Taiko module, the direct power supply by the Keces would be superior (12v CPU) ?
Interesting but it minimizes the contribution of the Taiko.

Link to post
Share on other sites

In my experience the ATX supply is less important for SQ than the EPS 12V supply to the CPU, perhaps trying the HDPLex on the ATX and the Taikoo on the 12V could be a useful test to undertake?

 

 

ISP, cat8.1, Zylix switch, Finisar <1321>Solarflare 8522, AMD 9, Aorus X570 Elite, Corsair AX1200i, 8Gb Apacer Ram, Pink Faun I2S OCXO on LPS, home grown RJ45 I2S cable, Metrum Amethyst (modded to accept I2S), Klangfilm 204a, Klangfilm Trionor (3Tesla 835, JBL2402)

Link to post
Share on other sites

I anticipate that a good old linear front will be needed to make that Taiko DC DC shine. I passed, I could not justify it to myself shelling out this sort of money unseen/unheard...

ISP, cat8.1, Zylix switch, Finisar <1321>Solarflare 8522, AMD 9, Aorus X570 Elite, Corsair AX1200i, 8Gb Apacer Ram, Pink Faun I2S OCXO on LPS, home grown RJ45 I2S cable, Metrum Amethyst (modded to accept I2S), Klangfilm 204a, Klangfilm Trionor (3Tesla 835, JBL2402)

Link to post
Share on other sites
15 minutes ago, MarcelNL said:

I anticipate that a good old linear front will be needed to make that Taiko DC DC shine. I passed, I could not justify it to myself shelling out this sort of money unseen/unheard...

 

Nenon has made it clear the winning formula is an unregulated linear to be fair.  I will in time also build that.

Link to post
Share on other sites
2 hours ago, guiltyboxswapper said:

Nenon has made it clear the winning formula is an unregulated linear to be fair.

 

 

Either that is the solution or instead of a good mid-range LPS like Keces P8, high-end power supplies like Paul Hynes SR7 or Sean Jacobs DC4 are used instead. I hope someone try it because I don't have this power supplys.

 

To try to give a rating (more * are better):

***** 2 Keces P8 + Taiko 500W ATX
**** 1 Keces P8 + Taiko 500W ATX
* 2 Keces P8 + HDPLEX 400 ATX

 

I feel the difference between Taiko Audio and HDPLEX is very big. 👍

 

 

Grigg Audio Solutions Owner

StreamFidelitys Setup

Sonus Faber Amati Futura | T + A M10 | T + A SDV 3100 HV | fis Audio PC | HFX RipNAS Solid V4 | GigaWatt PC2-EVO + | Solidsteel HJ-3 / HY-A

Link to post
Share on other sites

@StreamFidelity just to be clear, are you feeidng the USB or I2S output separately or through the MB?

ISP, cat8.1, Zylix switch, Finisar <1321>Solarflare 8522, AMD 9, Aorus X570 Elite, Corsair AX1200i, 8Gb Apacer Ram, Pink Faun I2S OCXO on LPS, home grown RJ45 I2S cable, Metrum Amethyst (modded to accept I2S), Klangfilm 204a, Klangfilm Trionor (3Tesla 835, JBL2402)

Link to post
Share on other sites
53 minutes ago, StreamFidelity said:

Either that is the solution or instead of a good mid-range LPS like Keces P8, high-end power supplies like Paul Hynes SR7 or Sean Jacobs DC4 are used instead. I hope someone try it because I don't have this power supplys.

 

 

Nenon has clearly posted that any regulated solution fed into Taiko's DC DC (that he has tried to date) isn't as effective, so much so he dropped the development of his own highly regulated PSU collaboration with SJ.  And he has access to the DC4 stuff, so that speaks volumes. 

 

We'll have to try it for ourselves ofcourse to confirm.  

 

Link to post
Share on other sites
20 minutes ago, guiltyboxswapper said:

We'll have to try it for ourselves

He also recommended ca. 30V as an optimal voltage to charge the Taiko ATX board.

 

Unfortunately my arrived with some issues and I'm waiting for Taiko answer.

Link to post
Share on other sites
1 hour ago, MarcelNL said:

just to be clear, are you feeidng the USB or I2S output separately or through the MB?

 

That's a really good point. Neither one nor the other in my setup. I am using NAA with the HQPlayer.

The specialty is the Solarflare XtremeScale X2522 NIC, which is powered directly from the motherboard. Could that explain why Taiko ATX works so well in my system?

Grigg Audio Solutions Owner

StreamFidelitys Setup

Sonus Faber Amati Futura | T + A M10 | T + A SDV 3100 HV | fis Audio PC | HFX RipNAS Solid V4 | GigaWatt PC2-EVO + | Solidsteel HJ-3 / HY-A

Link to post
Share on other sites
16 minutes ago, StreamFidelity said:

The specialty is the Solarflare XtremeScale X2522 NIC, which is powered directly from the motherboard. Could that explain why Taiko ATX works so well in my system?

 

My gut feeling is that the entire motherboard, including the Solarflare, is benefiting from the Taiko. With good power every small little things eventually adds up to provide a better sound. 

Link to post
Share on other sites
7 hours ago, StreamFidelity said:

Summary

 

While the replacement of the HDPLEX 400W DC-ATX converter by the Taiko Audio 500W DC DC-ATX converter was immediately audible, I had to prick up my ears when changing the CPU power supply. "Unfortunately" the direct CPU power supply from a Keces P8 with 12V is still better than the complete power supply from the Taiko ATX. That can change with a better LPS.

 

Interesting. Still lot of unknowns with the Taiko but based on your experience I have a feeling that feed supply is equally important to bring out the best of this DC-ATX.

Link to post
Share on other sites
1 hour ago, StreamFidelity said:

 

That's a really good point. Neither one nor the other in my setup. I am using NAA with the HQPlayer.

The specialty is the Solarflare XtremeScale X2522 NIC, which is powered directly from the motherboard. Could that explain why Taiko ATX works so well in my system?

but, but....you surely use either USB or I2S or SPDIF or something to feed your DAC the digital signal coming out of your MB?

 

ISP, cat8.1, Zylix switch, Finisar <1321>Solarflare 8522, AMD 9, Aorus X570 Elite, Corsair AX1200i, 8Gb Apacer Ram, Pink Faun I2S OCXO on LPS, home grown RJ45 I2S cable, Metrum Amethyst (modded to accept I2S), Klangfilm 204a, Klangfilm Trionor (3Tesla 835, JBL2402)

Link to post
Share on other sites
12 minutes ago, Nenon said:

What feeds the Taiko ATX is crucial. It makes a huge difference. Definitely worth trying the higher voltage unregulated LPS. 

Hi Nenon - I understand you voiced the combination of Taiko ATX and unregulated power supply. 1) Which audio PC did you use for the voicing? Is it Taiko or something else? Also, 2) what is your “religion of sound” or voicing “philosophy”? I am trying to understand better how those will sound together. 

Link to post
Share on other sites

Is anyone running the HDPlex 500W ATX? Did a search but seems to be very little feedback on the sound quality. It seems like a nice unit that eliminates need for DC conversion

Link to post
Share on other sites
  • 4 weeks later...

FYI - here's an official SFP28 from Solarflare and its model SOLR-SFM25G-SR-LL could be found under the "Supported Transceiver" section

 

https://www.hamburgnet.com/products/ethernet/solarflare

 

25GB ETHERNET LOW LATENCY SR SFP28 OPTICAL TRANSCEIVER

https://www.directdial.com/us/item/solarflare-sfp28-module/solr-sfm25g-sr-ll

https://www.shopblt.com/item/xilinx-solr-sfm25g-sr-ll-25gb-ethernet-low-latency/y510_solrsfm25gsrll.html

https://www.shopsaitech.com/ProductDetail/SOLARFLARE-25GB-ETHERNET-LOW-L/57548022/true/SOLR-SFM25G-SR-LL

 

Therefore -LL should actually stand for Low Latency and just wondering if there were anything special about that particular model? Solarflare was taken over by Xilinx while the archived version (dated November 2019) of the original site didn't seem to have SOLR-SFM25G-SR-LL (UPC # = 816191010991) listed anywhere

 

https://web.archive.org/web/20191128023407/https://www.solarflare.com/xtremescale-nics/accessories

https://web.archive.org/web/20191129075357/https://www.solarflare.com/xtremescale-nics/accessories

 

And then there's a similar one from Axiom SOLR-SFM25G-SR-LL-AX (UPC # = 840177844270) while that costs about the same

 

https://www.axiomupgrades.com/productdetail/SOLR-SFM25G-SR-LL-AX/

https://www.directdial.com/us/item/25gbase-sr-sfp28-transceiver/solr-sfm25g-sr-ll-ax

https://www.shopblt.com/item/axiom-solr-sfm25g-sr-ll-ax-25gbase-sfp28-transceiver-solarflare/axiom_solrsfm25gsrllax.html

 

$730 for a pair of SOLR-SFM25G-SR-LL, obviously they better be able to make a significant difference for audio purposes. Otherwise, why would anyone bother to pay THAT much while we could get these affordable cards from this seller with 100% positive feedback?

 

https://www.ebay.com/itm/173382619434

https://www.ebay.com/itm/174601092676

https://www.ebay.com/itm/174713063688

 

Those cards are indeed coming from China, though.

Link to post
Share on other sites
1 hour ago, seeteeyou said:

FYI - here's an official SFP28 from Solarflare and its model SOLR-SFM25G-SR-LL could be found under the "Supported Transceiver" section

 

Thank you very much for this information. I will watch the market. Maybe there are cheaper used ones.

Grigg Audio Solutions Owner

StreamFidelitys Setup

Sonus Faber Amati Futura | T + A M10 | T + A SDV 3100 HV | fis Audio PC | HFX RipNAS Solid V4 | GigaWatt PC2-EVO + | Solidsteel HJ-3 / HY-A

Link to post
Share on other sites

But in the end, has the quest for the lowest latency really been evaluated against a quality ethernet card (with a good buffer). These cards consume more power and generate more heat, so is it worth the risk?
How to evaluate the latency with the final streamer? 

Link to post
Share on other sites
  • 3 weeks later...

Tips for refining music files on the hard drive with resampling

 

introduction to the topic

 

Anyone who saves their music files on the hard drive has a fairly large playing field in front of them. For example, it's about the sampling rate (CD format 44.1kHz or HiRes from 88.2kHz) and also the bit depth (CD 16Bit or HD 20-32Bit). While the sampling rate expands the frequency response, the bit depth is about the dynamics.

 

Sampling rates:

 

PCM 44.1: 2 Hz -20 kHz
PCM 48: 2 Hz -22 kHz
PCM 96: 2 Hz -40 kHz
PCM 192: 2 Hz -80 kHz
PCM 384: 2 Hz -100 kHz
PCM 768: 2 Hz -120 kHz

 

Bit depth:

 

The word length / word width in bits (bit depth) encodes the volume of the analog signal (loud / quiet). 1 bit corresponds to 6 dB level difference. 1 bit therefore always includes a volume range (dynamics) of 6 dB. The 16-bit CD enables 65,536 (216) different signal levels to be recorded. A 16-bit x 6 dB CD has a dynamic range of 96 dB. HD (High Definition) with 24 bit x 6 dB theoretically achieves 144 dB dynamic range. The technical standard of good DACs usually limits the resolution to 20 bits = 120 dB.

 

FLAC vs. WAV

 

Another question deals with the audio format, for example FLAC (Free Lossless Audio Codec), a codec for lossless audio data compression. Anyone who rips their CDs will have already dealt with the WAV format. It is an uncompressed format, so the files are larger than FLAC. I prefer WAV, because the unpacking step as with FLAC is not necessary.

 

Application: Low Frequency Interpolation

 

This is where a free application for Windows comes into play: LowFreqResample65536x16x8. This application increases the sample points (taps) and the bit depth through interpolation. While the bit depth is always converted to 32 bits, the taps can be freely selected by specifying the sample rates. 65536x16x8 means 8 million taps (8,388,608). 

 

It was developed by a Japanese man who is also known for MinorityClean and the Bug Head Player. Here is the download page (in Japanese): http://www.mics.ne.jp/~cdorya/. Scroll all the way down and download the zipped file. Then unzip the file and select the Low frequency Resampler folder. The application can be run immediately without installation. What exactly the program calculates is not clear to me. Does the developer think low frequency resampling only calculates the deep bass? My ears tell me otherwise.

 

I tried it with 256 samples = 32,768 tabs (256x16x8). A maximum of 4 threads can be activated. The music file is simply dragged and dropped into the window. Strangely enough, when you click Start, you have to enter the number displayed. Only then does it start.

 

41298902yz.png

 

Even an 8 core CPU like Intel i9-9900K has a lot to do.

 

41298903uj.png


As a result, a subfolder is created that conveniently contains the sampling rate and of course the new file. The original file is therefore not touched!

 

41298906cm.png

Comparison original
Original size 44.95 MB
Length 4 minutes 27 seconds
Channels 2 (stereo)
Sample rate 44.1 KHz;
Sample size 16 bit
Bit rate 1.411 kbps

 

Compare 256 samples
Size 89.89 MB (0% Compressed)
Original size 89.89 MB
Length 4 minutes 27 seconds
Channels 2 (stereo)
Sample rate 44.1 KHz;
Sample size 32 bit
Bit rate 2.822 kbps (DVD)

 

The file size and the bit rate have thus been doubled. I repeated it with 512 samples = 65,536 tabs (512x16x8) and it took a few minutes. This program is not for weak computers. It is played back quite normally, here with Roon, for example.

 

41298907ey.png


And how does it sound?

 

My experiments and audio samples with a sound engineer brought the breakthrough from 1,048,576 samples = 134,217,728 tabs (1,048,576x16x8)! I'm not sure I got the formula right. But if so, I'm at 134 million Tabs. the conversion of just one music file takes hours. I was rewarded with finer resolution and improved timing.

 

Summary

 

The files on the hard disk can be further refined by interpolation. In addition to increasing to 32 bits, the taps can be interpolated by selecting the sample rate. Files in FLAC and WAV format can be processed. FLAC files are automatically converted to the uncompressed WAV format. The program Low Frequency Interpolation shows some parallel to PGGB. However, PGGB has more functions and must be installed first. Low frequency interpolation, on the other hand, is slim and something for experimentation at your own risk.

Grigg Audio Solutions Owner

StreamFidelitys Setup

Sonus Faber Amati Futura | T + A M10 | T + A SDV 3100 HV | fis Audio PC | HFX RipNAS Solid V4 | GigaWatt PC2-EVO + | Solidsteel HJ-3 / HY-A

Link to post
Share on other sites
29 minutes ago, StreamFidelity said:

Application: Low Frequency Interpolation

 

What exactly the program calculates is not clear to me. Does the developer think low frequency resampling only calculates the deep bass? My ears tell me otherwise.

 

A little more info here: https://www.audio “science” review/forum/index.php?threads/low-frequency-interpolation-another-bs-software-from-the-author-of-minority-clean.23781/

My digital system is wired ethernet (RedNet/Dante) but strictly offline. I don't download or stream music from the internet.

System here

Link to post
Share on other sites
49 minutes ago, Iving said:

 

I did not manage to deposit the correct link. Presumably our forum has an aversion to audio “science” review. 😂

 

I know KSTR from German-speaking forums. He writes well-founded. I am always grateful when I understand the technical construct. But at the same time humble, because not everything can always be explained with today's technology. My ears always win.

 

The named sample jumps are wanted by the developer (it was written somewhere in the developer blog).

 

Besides some measurement results, did KSTR "hear" the sampled music? There is nothing about that in the report.

 

I also use MinorityClean as a CPU jitter cleaner because it "sounds" better. Everyone as he likes. 😉

Grigg Audio Solutions Owner

StreamFidelitys Setup

Sonus Faber Amati Futura | T + A M10 | T + A SDV 3100 HV | fis Audio PC | HFX RipNAS Solid V4 | GigaWatt PC2-EVO + | Solidsteel HJ-3 / HY-A

Link to post
Share on other sites
5 minutes ago, StreamFidelity said:

 

I did not manage to deposit the correct link. Presumably our forum has an aversion to audio “science” review. 😂

 

I know KSTR from German-speaking forums. He writes well-founded. I am always grateful when I understand the technical construct. But at the same time humble, because not everything can always be explained with today's technology. My ears always win.

 

The named sample jumps are wanted by the developer (it was written somewhere in the developer blog).

 

Besides some measurement results, did KSTR "hear" the sampled music? There is nothing about that in the report.

 

I also use MinorityClean as a CPU jitter cleaner because it "sounds" better. Everyone as he likes. 😉

Yes there's a history. Chris had his reasons for preventing direct ASR links.

 

If I can make the time I might give LF Int. a personal evaluation. I just ordered a larger Optane AIC and will have to reinstall Windows. I just read about 11 Superlite. I have a lot to keep me busy. I don't know how you and Nenon keep up the frontier pushing so consistently!

 

btw I found LFI in the bughead package which is the penultimate not the last zip in your link.

My digital system is wired ethernet (RedNet/Dante) but strictly offline. I don't download or stream music from the internet.

System here

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×
×
  • Create New...