mr.duck said:Metronome has the added advantage of upsampling to 192 KHz (maybe the dac such as Opus or Buffalo perform better at 192 KHz?). But big disadvantage it cannot resample to 44.1 KHz. The twisted pear modular design is great... but adding up all the bits gets a wee bit expensive. So many options, so little time!
😀
I can't speak as to which method sounds "better" because that is subjective. What I can say is that the ASRC should be technically better at producing a low jitter I2S source. I personally would (and do) use a metronome to reduce jitter from a PC source.
You actually can resample to 44.1khz with metronome, but you would need to use a different XO, or disable the on-board XO and supply and external master clock. I am not exactly sure why you would want to, but you could. 🙂
Cheers!
Russ
Russ White said:
Ah excelent! What frequency XO is required for 44.1 KHz?
As for why I would want to... to compare 192 vs 44.1 sample rates with redbook material of course 🙂 To upsample or not to upsample.
Also gives the opportunity to compare the on board crystek to an external clock such as tentlabs, hagclock, DIY clock, etc (if I ever feel the need to try that).
ti SRC4192 and EIAJ data (from CXD2500)
Hi,
I'd like to know, can I use the SRC4192 for convert/upsampling the sony CXD2500 EIAJ output's data to create 64fs I2S signal for Sabre8 (Buffalo).
Many thanks in advance!
Hi,
I'd like to know, can I use the SRC4192 for convert/upsampling the sony CXD2500 EIAJ output's data to create 64fs I2S signal for Sabre8 (Buffalo).
Many thanks in advance!
Hi Guys,
As a newbie I apologise in advance if I ask stupid questions.
Russ, I know you may already have mentioned this but this thread is already 66 pages and it's taking me awhile to get through it all...
just wondering what program you are using to do the PCB artwork and whether you would be happy to share the artwork with people who want to modify it.
Storm
As a newbie I apologise in advance if I ask stupid questions.
Russ, I know you may already have mentioned this but this thread is already 66 pages and it's taking me awhile to get through it all...
just wondering what program you are using to do the PCB artwork and whether you would be happy to share the artwork with people who want to modify it.
Storm
Re: ti SRC4192 and EIAJ data (from CXD2500)
I don't see anything in the SRC4192 datasheet about supporting that format, so my guess is it won't work. 🙂
Cheers!
Russ
cartman said:
I don't see anything in the SRC4192 datasheet about supporting that format, so my guess is it won't work. 🙂
Cheers!
Russ
s1b2storm said:
Hi Storm,
I use Diptrace CAD software. I moved to it from Eagle and I never looked back. Its very good software. 🙂
I don't generally share my PCB designs as it takes a lot of work to design them. 🙂
Cheers!
Russ
Russ,
I appreciate the amount of effort you've put into this. I actually bought a couple of your blank boards.
There was also some talk in one of the early posts about getting the opus board measured. Did you ever manage to get it onto an AP2 or similar?
As the components I source are likely to be different to those on your ready built units it'd be interesting to compare measurements between stock, my build and the Wolfson eval board.
Storm
I appreciate the amount of effort you've put into this. I actually bought a couple of your blank boards.
There was also some talk in one of the early posts about getting the opus board measured. Did you ever manage to get it onto an AP2 or similar?
As the components I source are likely to be different to those on your ready built units it'd be interesting to compare measurements between stock, my build and the Wolfson eval board.
Storm
Opus is high bang/buck!
I have had Opus boards waiting for an ambitious multi-channel DAC/pre project for computer plus SACD, but I had difficulty getting started. All the current news and anticipation surrounding the second-generation ESS Sabre chip have spurred me into action to just make my on-hand Opus boards play. ...hoping, I suppose, to better withstand the inevitable lust and anticipation during development until the new chips and products are both available and dialed-in.
I decided to use the Opus boards only to upgrade my old computer sound board (ESI Wami-Rack 192X). After some micro-soldering into vias spaced 2mm apart (using a stereo-microscope
) I now have 3 stereo pairs handled by 3 Opus boards. WOW! The Wolfson 8741 chip is no slouch! Of course, Russ has been saying that all along, but it's true... Outputting directly into TP amps with nothing else added, IMHO, the naked Opus is VERY high bang/buck!
Nice work Brian and Russ! ...take your time with that crossover and control software for the new ESS Sabre chip! 😀 While some believe that "the good is the enemy of the best", I think they are worried about marketing wine. For me, good is good!
The performance, robust applicability, and price of a naked Opus are impressive.
Cheers,
Frank in Mpls.
I have had Opus boards waiting for an ambitious multi-channel DAC/pre project for computer plus SACD, but I had difficulty getting started. All the current news and anticipation surrounding the second-generation ESS Sabre chip have spurred me into action to just make my on-hand Opus boards play. ...hoping, I suppose, to better withstand the inevitable lust and anticipation during development until the new chips and products are both available and dialed-in.
I decided to use the Opus boards only to upgrade my old computer sound board (ESI Wami-Rack 192X). After some micro-soldering into vias spaced 2mm apart (using a stereo-microscope
) I now have 3 stereo pairs handled by 3 Opus boards. WOW! The Wolfson 8741 chip is no slouch! Of course, Russ has been saying that all along, but it's true... Outputting directly into TP amps with nothing else added, IMHO, the naked Opus is VERY high bang/buck! Nice work Brian and Russ! ...take your time with that crossover and control software for the new ESS Sabre chip! 😀 While some believe that "the good is the enemy of the best", I think they are worried about marketing wine. For me, good is good!
The performance, robust applicability, and price of a naked Opus are impressive.
Cheers,
Frank in Mpls.
Q: for Brian, Russ, or ?
Hi Guys!
I have been enjoying the Opus boards very much and have a basic question:
I am running the 8741 version in hardware mode. For some recordings I would like to be able to switch from filter #2 to filter #3 using an external switch. ...in other words, change the state of FSEL from "+" to "0". Is this possible using the I/O header? [I presume that '+' on the Opus state switch = 'Z' on the WM8741 data sheet...] Thanks!
Rationale:
I like my music near 'performance levels'.
Unfortunately, my listening room has marginal acoustics (square and not very absorptive) so my powerful midrange speakers can easily produce harshness when recordings include significant overtones. With the Opus, the recordings that tend to be strident (and high SPL in general) are more enjoyable using filter #3 (hardware mode: OSR = '+', FSEL = '0'). However, sweeter, better (and quieter) recordings are much more life-like using filter #2 (OSR and FSEL = '+'). For example, using filter #2 saxophones 
seem absolutely live and fine acoustic instruments 
have much more genuine timbre. It would be sweet to toggle the DAC filter depending on the recording quality and/or desired volume. I would normally glue egg cartons to the ceiling and walls but for reasons I don't understand my wife objects to that solution! 
Frank
Hi Guys!
I have been enjoying the Opus boards very much and have a basic question:
I am running the 8741 version in hardware mode. For some recordings I would like to be able to switch from filter #2 to filter #3 using an external switch. ...in other words, change the state of FSEL from "+" to "0". Is this possible using the I/O header? [I presume that '+' on the Opus state switch = 'Z' on the WM8741 data sheet...] Thanks!
Rationale:
I like my music near 'performance levels'.
Unfortunately, my listening room has marginal acoustics (square and not very absorptive) so my powerful midrange speakers can easily produce harshness when recordings include significant overtones. With the Opus, the recordings that tend to be strident (and high SPL in general) are more enjoyable using filter #3 (hardware mode: OSR = '+', FSEL = '0'). However, sweeter, better (and quieter) recordings are much more life-like using filter #2 (OSR and FSEL = '+'). For example, using filter #2 saxophones seem absolutely live and fine acoustic instruments have much more genuine timbre. It would be sweet to toggle the DAC filter depending on the recording quality and/or desired volume. I would normally glue egg cartons to the ceiling and walls but for reasons I don't understand my wife objects to that solution! Frank
Hi Frank,
Yes that is precisely what that header is for. 🙂
Just keep in mind you will need to Cycle the power to the DAC for the change to take effect.
You will need to pull the pins high are low using an external switch or some other mechanism.
I would suggest a pull-up resistor and a switch which closes to GND.
You will then leave the on-board switch open for that position.
Cheers!
Russ
Yes that is precisely what that header is for. 🙂
Just keep in mind you will need to Cycle the power to the DAC for the change to take effect.
You will need to pull the pins high are low using an external switch or some other mechanism.
I would suggest a pull-up resistor and a switch which closes to GND.
You will then leave the on-board switch open for that position.
Cheers!
Russ
Russ White said:
Russ, you're so generous with your time! I apologize... , I am very naive in this arena plus I suspect a typo above. To confirm, let me rewind and play this back: Is it the case that a pin = "Z" when it is power cycled as neither high nor low? Then, I see that the pins are shorted to ground at dip switch position '1'. Is that situation considered 'high'? ...meaning that dip switch position '0' is low? (
...or vise versa? 
) So, if the dip switch is on '+' and I then add a resistor between FSEL header and ground with an external switch (and power cycle, of course), that pin state becomes '0'? ...and that is the same as cycling with the dip switch set to '0'!!! Whew! Can do!!! 😀 What value resistor?Cheers!
Frank
PS. application in computer sound board shown in attachment - boards are stuffed into a 1U rack box...
Attachments
I think it means
a- leave the dip switch in "0" or unconnected
b- Connect a wire to pin header and wire to high (VDD) through a resistor (lets guess 1K)
c- connect a switch to the pin that when closed, connects to gnd.
I think I'm going to enable the different H/W filters too.
(Very happy to see some people still experimenting with the OPUS)
a- leave the dip switch in "0" or unconnected
b- Connect a wire to pin header and wire to high (VDD) through a resistor (lets guess 1K)
c- connect a switch to the pin that when closed, connects to gnd.
I think I'm going to enable the different H/W filters too.
(Very happy to see some people still experimenting with the OPUS)
francolargo said:
Russ, you're so generous with your time! I apologize... , I am very naive in this arena plus I suspect a typo above. To confirm, let me rewind and play this back: Is it the case that a pin = "Z" when it is power cycled as neither high nor low? Then, I see that the pins are shorted to ground at dip switch position '1'. Is that situation considered 'high'? ...meaning that dip switch position '0' is low? (
Cheers!
Frank
PS. application in computer sound board shown in attachment - boards are stuffed into a 1U rack box...
Hi Frank,
My mistake. I forgot that the pins are 3 state. Sorry for the confusion.
You would essentially need to mimic what the on-board switch does.
Basically the "-" (It looks like a 1) pulls the pin to GND.
"O(as in Oscar)" leaves it open.
"+" pulls it to VDD (3.3V).
What I would do is use a 3 state switch. One end would be high(VDD), middle open, other end low(GND).
You don't need a resistor at all in this case, but I would probably use a 1K resistor on the high side out of an abundance of caution, in case you accidentally have a short through the switch .
Cheers!
Russ
so many questions...
Thanks for the contribution glt and thanks Russ for clearing up how to proceed!
I'm still curious about something...
Page 40 in the WM8741 product review concerning pin 25: 0 = low input rate, Z = medium input rate, 1 = high input rate. The filter properties, as shown on pages 49-51, have very different values depending on the input rate. To what input rate do the charts refer? Is it the actual I2S input rate or merely the setting that one selects using pin 25? With pin 25 set for "z" or medium (88.2/96 kHz), I can play 44.1 kHz material just fine so is the only consequence of the setting in the filter set that is subsequently available using pin 4? If I have pin 4 set to "z" (for filter 2) and am playing 44.1 material even while pin 25 is set for 88.2/96, then what filter am I getting? Filter 5 or filter 2??? If the board actually plays, does anybody know of any consequences of the input rate (pin 25) setting other than the filter sets available in hardware mode?
Ulitmately, I know what sounds preferable in my system but I am interested in understanding the chip, the filters and their differences.
All the best,
Frank in Mpls.
Thanks for the contribution glt and thanks Russ for clearing up how to proceed!
I'm still curious about something...
Page 40 in the WM8741 product review concerning pin 25: 0 = low input rate, Z = medium input rate, 1 = high input rate. The filter properties, as shown on pages 49-51, have very different values depending on the input rate. To what input rate do the charts refer? Is it the actual I2S input rate or merely the setting that one selects using pin 25? With pin 25 set for "z" or medium (88.2/96 kHz), I can play 44.1 kHz material just fine so is the only consequence of the setting in the filter set that is subsequently available using pin 4? If I have pin 4 set to "z" (for filter 2) and am playing 44.1 material even while pin 25 is set for 88.2/96, then what filter am I getting? Filter 5 or filter 2??? If the board actually plays, does anybody know of any consequences of the input rate (pin 25) setting other than the filter sets available in hardware mode?
Ulitmately, I know what sounds preferable in my system but I am interested in understanding the chip, the filters and their differences.
All the best,
Frank in Mpls.
Re: so many questions...
Frank,
That setting is critical. The DAC will not function if it is not set correctly. As per the datasheet the setting needs to match the incoming sample rate of the PCM data. If it does not you will get noise or nothing. I can't remember which... 🙂
The filters are different for each sample rate. Basically internally different FIR coefficients are used depending on the indicated sample rate.
Cheers!
Russ
francolargo said:
Frank,
That setting is critical. The DAC will not function if it is not set correctly. As per the datasheet the setting needs to match the incoming sample rate of the PCM data. If it does not you will get noise or nothing. I can't remember which... 🙂
The filters are different for each sample rate. Basically internally different FIR coefficients are used depending on the indicated sample rate.
Cheers!
Russ
Hey Russ,
Thankfully, that Beethoven now wafting up the stairs doesn't sound too much like noise! 🙂 I don't want to distract you from the other cool projects on your plate!
You said elsewhere in this thread that you preferred filter #2, which according to the data sheet is only available in hardware at 88.2kHz. That's what I thought the TP Opus manual 2.0 pdf suggested and what I was listening to... even using I2S @ 44.1. I need to relate the Opus dip switch settings to the "0", "Z", and "1" states in the Wolfson data sheet. I now gather 1=+=high, Z=O=Open, and zero=minus=low.
When I open up my chassis to tap into the I/O headers I'll double-check my favorite settings and re-evaluate all the functioning states in this simple implementation. I'd like to try upsampling in software but thus far, no results.
to be continued...
Frank
Thankfully, that Beethoven now wafting up the stairs doesn't sound too much like noise! 🙂 I don't want to distract you from the other cool projects on your plate!
You said elsewhere in this thread that you preferred filter #2, which according to the data sheet is only available in hardware at 88.2kHz. That's what I thought the TP Opus manual 2.0 pdf suggested and what I was listening to... even using I2S @ 44.1. I need to relate the Opus dip switch settings to the "0", "Z", and "1" states in the Wolfson data sheet. I now gather 1=+=high, Z=O=Open, and zero=minus=low.
When I open up my chassis to tap into the I/O headers I'll double-check my favorite settings and re-evaluate all the functioning states in this simple implementation. I'd like to try upsampling in software but thus far, no results.
to be continued...
Frank
WM8742 is out!!
Hi guys!
26th Jan 2009 Wolfson announced WM8742
http://www.wolfsonmicro.com/whatsnew/press/press/PI320
pin compatible with 8741/8740
available for sampling now.
it seems they improved filtering.. who is going to be the first to evaluate it (and possibly compare it to older versions)???😎
btw thanks to Russ and Brian.
I'm listening now to opus8740/metronome via usb (foobar+asio4all) with great pleasure.. BTW better than my previous dual mono 1798 COD/IVY combo.
ciao
Vale
Hi guys!
26th Jan 2009 Wolfson announced WM8742
http://www.wolfsonmicro.com/whatsnew/press/press/PI320
pin compatible with 8741/8740
available for sampling now.
it seems they improved filtering.. who is going to be the first to evaluate it (and possibly compare it to older versions)???😎
btw thanks to Russ and Brian.
I'm listening now to opus8740/metronome via usb (foobar+asio4all) with great pleasure.. BTW better than my previous dual mono 1798 COD/IVY combo.
ciao
Vale
follow-up: 8741 hardware mode
OK, I opened up the chassis and tinkered some more and I have two items to report.
1. Those Wolfson 8741 chips just want to play! I had mistakenly configured them in serial control mode, but they were still working well. I re-configured them to hardware mode and learned that they will play 44.1kHz I2S regardless of the oversampling rate setting. The SQ is indeed inferior if the OSR setting does not match the source. Interestingly, whatever the defaults are for 2 wire serial mode, that sound was almost as good as hardware mode with the best filter.
2. My favorite filter:
Source material is ALAC files in foobar 9.4.X with DSP crossover, room correction, and kernel streaming. (ASIO is incompatible with some DSP plugins) The Opus DACs are directly connected (no caps) to TXO balanced amps driving home-brew JBL Pro speakers.
The filter choices in hardware at 44.1 kHz are numbers 1, 4 and 5. In my system, staging and detail were best with filter number 1. I hear this particularly in choral music and well-recorded live performances. Well-recorded percussion is strikingly realistic. Also, I agree with Russ that the clipping control really helps with some recordings so I have that enabled as well. One of my 'evaluation' tracks is Diana Krall singing 'the look of love', because there are some clean but very 'hot' vocal highlights. The clipping filter helps control them without ruining the brightness.
FYI, my switch settings for 44.1 material are as follows:
....l..O..+
1 |......X|
2 |X......|
3 |...X...|
4 |X......|
5 |......X|
6 |X......|
7 |X......|
8 |......X|
9 |X......|
When I get around to testing 88.2 or other higher sample rate I2S, I'll report back with a comparison and my favorite filter for that material.
Best,
Frank in Mpls.
francolargo said:
OK, I opened up the chassis and tinkered some more and I have two items to report.
1. Those Wolfson 8741 chips just want to play! I had mistakenly configured them in serial control mode, but they were still working well. I re-configured them to hardware mode and learned that they will play 44.1kHz I2S regardless of the oversampling rate setting. The SQ is indeed inferior if the OSR setting does not match the source. Interestingly, whatever the defaults are for 2 wire serial mode, that sound was almost as good as hardware mode with the best filter.
2. My favorite filter:
Source material is ALAC files in foobar 9.4.X with DSP crossover, room correction, and kernel streaming. (ASIO is incompatible with some DSP plugins) The Opus DACs are directly connected (no caps) to TXO balanced amps driving home-brew JBL Pro speakers.
The filter choices in hardware at 44.1 kHz are numbers 1, 4 and 5. In my system, staging and detail were best with filter number 1. I hear this particularly in choral music and well-recorded live performances. Well-recorded percussion is strikingly realistic. Also, I agree with Russ that the clipping control really helps with some recordings so I have that enabled as well. One of my 'evaluation' tracks is Diana Krall singing 'the look of love', because there are some clean but very 'hot' vocal highlights. The clipping filter helps control them without ruining the brightness.
FYI, my switch settings for 44.1 material are as follows:
....l..O..+
1 |......X|
2 |X......|
3 |...X...|
4 |X......|
5 |......X|
6 |X......|
7 |X......|
8 |......X|
9 |X......|
When I get around to testing 88.2 or other higher sample rate I2S, I'll report back with a comparison and my favorite filter for that material.
Best,
Frank in Mpls.
Opus testing: software oversampling
To complete the Opus testing at oversampling frequencies I had to remove the software DRC from the signal path (for compatibility). I'm not sure of the reason, but my woofers are playing flatter now than before the Opus upgrade. I have no idea whether the WM8741 is responsible, but now bass sounds better un-filtered even though one cancellation effect is still evident in the scans.
I tested 88.2 and 176.4kHz up-sampled in software from 44.1 music. I felt that regardless of the filter, 88.2 was not as clean as 44.1 running filter #1. Best of all, however, was 176.4kHz where both filters #2 and #3 sounded quite good. Filter #2 is less 'damped' and offers noticably more brightness, space and timbre. Filter #3 produces less detail and poorer staging, but I never heard even the slightest sibillance - even from over-saturated jazz flute. Filter #2 is my fave. For difficult recordings (read: nasty - like grand opera recorded in some opera houses) I will switch to filter #3. The oversampling clearly reduces the 'fatigue' factor and that translates to better dynamics and a more powerful presentation. [The bass from the WM8741 is ear-opening, and in my system is being achieved using simple TXO monoblock amps
driving JBL 2235H's.]
Again, considering what one would spend configuring the 'trendiest' stereo DACs for 6 channels plus I/V, the Opus seems like an unbeatable bargain. ...Perfect for those like me whose DIY objectives include 'frugal-audio'! Again, nice work and thanks to Brian and Russ!!!
Frank in Mpls.
P.S. Here are my favorite oversampled dip-switch settings:
....l..O..+
1 |......X|
2 |X......|
3 |...X...|
4 |X......|
5 |......X|
6 |X......|
7 |...X...|
8 |......X|
9 |......X|
To complete the Opus testing at oversampling frequencies I had to remove the software DRC from the signal path (for compatibility). I'm not sure of the reason, but my woofers are playing flatter now than before the Opus upgrade. I have no idea whether the WM8741 is responsible, but now bass sounds better un-filtered even though one cancellation effect is still evident in the scans.
I tested 88.2 and 176.4kHz up-sampled in software from 44.1 music. I felt that regardless of the filter, 88.2 was not as clean as 44.1 running filter #1. Best of all, however, was 176.4kHz where both filters #2 and #3 sounded quite good. Filter #2 is less 'damped' and offers noticably more brightness, space and timbre. Filter #3 produces less detail and poorer staging, but I never heard even the slightest sibillance - even from over-saturated jazz flute. Filter #2 is my fave. For difficult recordings (read: nasty - like grand opera recorded in some opera houses) I will switch to filter #3. The oversampling clearly reduces the 'fatigue' factor and that translates to better dynamics and a more powerful presentation. [The bass from the WM8741 is ear-opening, and in my system is being achieved using simple TXO monoblock amps
driving JBL 2235H's.]Again, considering what one would spend configuring the 'trendiest' stereo DACs for 6 channels plus I/V, the Opus seems like an unbeatable bargain. ...Perfect for those like me whose DIY objectives include 'frugal-audio'! Again, nice work and thanks to Brian and Russ!!!
Frank in Mpls.
P.S. Here are my favorite oversampled dip-switch settings:
....l..O..+
1 |......X|
2 |X......|
3 |...X...|
4 |X......|
5 |......X|
6 |X......|
7 |...X...|
8 |......X|
9 |......X|
Hi Frank,
Its very cool to see you getting so much use out of the Opus. 🙂
Bravo!
It seems you have also found the same nice synergy I have between the TXO (and to an even greater extent Sympatico) and balanced DACs. 🙂
Running your gear fully balanced results in better dynamic range, which is why I think I will probably never have a single ended power amp again. 🙂
Cheers!
Russ
Its very cool to see you getting so much use out of the Opus. 🙂
Bravo!
It seems you have also found the same nice synergy I have between the TXO (and to an even greater extent Sympatico) and balanced DACs. 🙂
Running your gear fully balanced results in better dynamic range, which is why I think I will probably never have a single ended power amp again. 🙂
Cheers!
Russ