Christopher, you’re going to be at RMAF with this possible mod? Let me know. While I think we’re generally responding with caution, I for one would be interested to hear a well thought out modification… and to finally meet in person!
Christopher, I do like the idea of a Hardware Upgrade sanctioned by PS Audio. Obviously the DS was designed with a certain price point in mind. But what if all replaceable parts, caps, iecs, transformers, fuses were upgraded by substituting the exact same values as the DS,
The DS might indeed be brought to its full potential.
Here is some ideas for you guys. For one a simple firmware rolling from the screen. The chip can hold many firmwares. So if the chip ws left in . A few firmwares could be selected from screen. Also a stripped FPGA for us who want ultimate performance .
al
I thought that at least there were enough spare resources on the FPGA now…(?)
adriaan said I also like the idea of a stripped down firmware FPGA version to optimize the sound quality. Perhaps the volume control coding and support for MP3 could be disabled.@Ted: is this a good idea?
There is no support for MP3.
Dennis
adriaan said I also like the idea of a stripped down firmware FPGA version to optimize the sound quality. Perhaps the volume control coding and support for MP3 could be disabled.@Ted: is this a good idea?
As other’s have said there is no support for MP3 - the least used feature is the deemphasys filter, but it’s just a few strategic adds in the right places so it doesn’t use a lot of resources.
There’s no fat left that I can find - I’m not saying there’s nothing left that could be optimized, but there’s nothing obvious left to play with. (For example there’s no debugging support left - I tossed the debugging helpers a few versions ago.)
FPGAs aren’t like the CPUs that we are familiar with - in a standard CPU using or not using an optional feature causes a change in control flow and also a change in power usage, and potentially could enable other work to get done instead. The variations in power consumption, heat generated (or even how many cores are powered these days) cause noise to be injected into the system.
In an FPGA (at least one without a traditional CPU configured inside it) optional features are running all of the time - using one or not doesn’t change the power used or allow other things to work faster… To a first approximation using an optional feature or not doesn’t change the power usage or noise generated at all. (More recent FPGAs can be configured so that the parts of the design that would have their answers dropped on the floor can get their clocks disabled to save energy, but I turn off such FPGA features to avoid introducing variability.)
So there’s no change in how many bits are flying around (or power used, or noise generated) when the volume is at 100 vs any other setting. I.e. the noise generated in the FPGA is independent of the volume setting.
I’m not saying that nothing would change with no volume control, but the noise in the FPGA would probably be just about the same.
For those interested here’s essentially how an FPGA works: each logic cell is a small lookup table followed by a flip-flop. The lookup table can be used to generate any possible outcome for the input variables. So for an FPGA where each cell has four inputs the lookup tables have 16 possible output bits indexed by the four inputs. Then the flip-flip is used to keep it’s old answer or to use the result of the table lookup. There is no flow control, every lookup table always is looking up, every flip-flop is always getting clocked. All arms of an “if” or a “case” statement are computed in parallel and a multiplexor is used to select the arm wanted. You can’t use some resources in one arm of an “if” differently than you do in the other arm so there’s much less variability than a more traditional processor. FPGA programming consists of filling in the lookup tables. (Each cell can also be used a little differently but the same ideas apply.)
Ted thanks for your answer. If ever in NYC look me up imown you a good meal of your choice and a drink.
Regadring the firmware rolling is there any hope for that. As with each release there is several losers how about them getting posted.
So there is always someone who like s something diffent then others right. So the frimware rolling as a the new tube rolling concept is really a good idea. Many months back I asked Paul to voice for head phones as well as speakers in fact I said back then his actual answer was its for speakers not headphones . I laughed to myself. Now headphones are used to voice as well as speakers ,To me headphones are perhaps more popular then speakers now as the cost for a top system is sooo much less than a full speaker rig. Now this leaves me to want more firmwares to roll . Why cause some music or headphones just need the dac to sound different. No one is going to change amps but a dac that can roll like tubes can be rolled in some DACS by the way is a very good idea. So ask me questions if need be , but from a user point of view do I as i ask and sell more product . The why Paul or whom ever feels it’s a speaker only world is truly living in the past .
The sites with the most traffic are headphones and iem,s. Not speakers . Lastly I truly do like the ps audio brand and it’s sound and weather my opinion matters or not is fine but still post on other sites and read the reactions. Your frimware rolling is a fantastic idea but how you go about it is wrong in my world. Having ten people tell the rest of us how it should sound is at the very least a dictator ship. The dac will remain it’s house sound anyway and get much more posts about the rolling. So if possible figure out my concept many want it and a such you shloud sell more product. . Just a thought . And yes or no please answer the question i pose please.
Al
Great idea AL,
PS Audio could allow the use of the SD card slot at the back to hold up to say 5 or 10 firmware versions that could be rolled using the remote control once the card is in place. This would be a powerful feature that would attract serious audiophiles who like to be on the cutting edge.
You could roll for HP type presentation, for preferred PCM playback, for preferred DSD playback and for stripped down “core mode” type of playback… for example. Going a step further, PS could even allow the customization of firmware version name so that version 0.61 could be replaced by PCM OPT to make it even more intuitive, for example. These could be unique selling points, at least for the moment. Carpe diem, PSA!
Wow, that’s the exact opposite of the direction I’d like to see the product go. I like to have an excellent source to build a system around and tweak it to get it as wonderful sounding as I can. As soon as the source starts changing it’s back to futzing and tweaking again. Earlier in my life that would have been just fine, now that I’m working about sixty hours a week week after week I just want to sit down and escape into my music. This new update wasn’t “free” for me (I couldn’t get it going on my own and had to buy a card from PS Audio for thirty bucks that I hope arrives soon) and I"m sure it will mean more adjusting and assessing. . . . The damned thing sounds amazing right now that it’s broken in and here we go again! I hope it’s a long time til the next update! I don’t like the “computer audio” aspects of this DAC and I don’t want to change firmware and hope that there’s a good chunk of time between this one and the next one.
Although I know this isn’t the most popular viewpoint here. . . maybe I’m not alone in some aspects of this.
Well, yes - the user friendliness aspect must not be forgotten.
Then probably the resistance also weakens.
lonson said Wow, that's the exact opposite of the direction I'd like to see the product go. I like to have an excellent source to build a system around and tweak it to get it as wonderful sounding as I can. As soon as the source starts changing it's back to futzing and tweaking again. Earlier in my life that would have been just fine, now that I'm working about sixty hours a week week after week I just want to sit down and escape into my music. This new update wasn't "free" for me (I couldn't get it going on my own and had to buy a card from PS Audio for thirty bucks that I hope arrives soon) and I"m sure it will mean more adjusting and assessing. . . . The damned thing sounds amazing right now that it's broken in and here we go again! I hope it's a long time til the next update! I don't like the "computer audio" aspects of this DAC and I don't want to change firmware and hope that there's a good chunk of time between this one and the next one.Although I know this isn’t the most popular viewpoint here. . . maybe I’m not alone in some aspects of this.
No one is forcing you to roll FW and heck, you did NOT have to update the FW this time if you were already happy with the sound. For those who like it as it is, they are TOTALLY unaffected by what AL suggested…you can let sleeping dogs lie. However, why begrudge others the opportuinty to switch between FW versions easiliy? You said it was OK when you were younger??? Well young people also own this Dac, so its not OK for them now? LoL
I know you cant be that mean-spirited.
For those who want to roll firmwares why not just keep a library of firmwares each on their own SD card and drop whatever one in the slot you want and reboot the machine?
At $2.50 a pop for a 2GB SD card seems like the easiest option.
I’m on the fence about this. While I do tweak (not as much as I used to) and the thought of flexibility sounds nice, I think it would be a rather large unplanned burden on the part of Ted and the rest of the PS Audio crew (I suspect this eats up a surprising number of people at PSA) in having to design the software variations necessary to do these things, in a way that can be easily supported and maintained. 1.2.1 went through dozens (if not close to a hundred) versions before even the Beta team got to test it, and once we got it there were around five different variations for us to sort through. That was for a planned improvement. I suspect these “tweak variations” would not need to be so complicated, but also not so simple that Ted could change a gate and then send it out the door. Even if the number of variations went down to about 10 (internal to PSA) for each tweak type you’re talking about potentially a lot of work for a lot of people. Take Paul’s Audio Decoding Engine view of the DS - treat the FPGA change as deciding how to vary the physical design of a discrete DAC (component selection etc). Now each variation has to be built (compiler options etc) and voice tested. The only real difference here is you have a team of people doing it in software rather than hardware. I actually think this would be easier to do in hardware; that paradigm has a much broader base of experience and knowledge history than what we’re doing in the DS.
Wisnon gets the whole point here .
Tony you are not understanding this correctly
you already have the firmwares to,release no work for ted.
With each new firmware there is sub firmwares so 121- A and so on. The one before this is 119 and its 119 -A and so on.
Lastly just a simple loader from the home screen to select one a many on the sd card is that is needed .
You,want buz talk firmware rolling see how many hits is generated. The concept is for the young unlike most here . Even though at 57 I say im old I am not like most here who change one thing and post about for weeks .
Move on chage is always good and anyone using headphones want to roll .
Honestly this is not just for me I am really showing how to generate buzz for your product. Two channel audio is a fraction of the headphone market wake up,ps audio and smell the firmware brewing.
Land pleae do not take my posting about this as a rant it’s not I am a passionate person. It’s hiw I do all,I do.
Read my caps thread ,many cause this is computor audio and most all who use a DAC are clueless to the hiw or why it sounds.
And as I understand the keep it dumb thing is fine it is not for all of us.
Speaking of mods, it would be interesting to listen to a DS with Jensen transformers 
The IEC socket and the internal power cables can be quite easily DIY-replaced.
wisnon saidHey please read my post again. Nowhere did I begrudge anyone, nor was there any mean spirit in my post, yours seems much more mean-spirited, LOL. I just stated my opinion that this is the opposite direction that I want to go. If PS Audio is going to release new firmware every month I'll get off the merry-go-round, that was not the type of product I spent my hard-earned money on or wanted to own. And yes I don't have to buy and install the firmware, and yes I get that others like this sort of thing. I AM letting sleeping dogs lie. All I did was state MY opinion that I'm not interested in this just as you are stating YOUR opinion that you are. What is so upsetting to you about that? Sheesh.No one is forcing you to roll FW and heck, you did NOT have to update the FW this time if you were already happy with the sound. For those who like it as it is, they are TOTALLY unaffected by what AL suggested…you can let sleeping dogs lie. However, why begrudge others the opportuinty to switch between FW versions easiliy? You said it was OK when you were younger??? Well young people also own this Dac, so its not OK for them now? LoL
I know you cant be that mean-spirited.
Finally, “earlier in my life” had nothing to do with age, but with time management. I’m helping my two aged parents suffering from dementia and physical ailments stay in their home which takes up sixty hours or more of my week. I have very little time to spend with my system and I no longer wish to spend fractions of that tweaking and futzing.
alrainbow said Ted thanks for your answer. If ever in NYC look me up imown you a good meal of your choice and a drink.I'd love to.
alrainbow said Many months back I asked Paul to voice for head phones as well as speakers in fact I said back then his actual answer was its for speakers not headphones . I laughed to myself. Now headphones are used to voice as well as speakers ,To me headphones are perhaps more popular then speakers now as the cost for a top system is sooo much less than a full speaker rig. Now this leaves me to want more firmwares to roll . Why cause some music or headphones just need the dac to sound different. No one is going to change amps but a dac that can roll like tubes can be rolled in some DACS by the way is a very good idea. So ask me questions if need be , but from a user point of view do I as i ask and sell more product . The why Paul or whom ever feels it's a speaker only world is truly living in the past .Arnie uses headphones for his listening when we're searching for the best versions. I'm not convinced that there are versions that are optimized for speakers vs headphones (or PCM vs DSD etc.) The only variable we are playing with is jitter. We try to minimize jitter to get the best version we can. With jitter a little higher we get a little more bite, a little more energy and perhaps a little more perceived detail. These are the things people who are used to bad digital may like. But just upping jitter is no guarantee that the DAC would sound "PCM". It is just a guarantee that no-one would listen again.
alrainbow said Your frimware rolling is a fantastic idea but how you go about it is wrong in my world. Having ten people tell the rest of us how it should sound is at the very least a dictator ship. The dac will remain it's house sound anyway...There are definitely differences between different FPGA compiles of the identical code (caused by the random number generator during the compiles) and it is the case that some may like one version over another, but from release to release there's no guarantee that we'd be able to identify a good headphone version vs a good speaker version (what ever that might be) or a good PCM vs a good DSD. The best we could do is a set of 5 (or whatever) not bad sounding releases with a recommendation for the default for those that don't want to play.
We don’t try 20 compiles of the FPGA code because we want to. We do it because we didn’t want to have the DAC cost significantly more if we attempted to fix the problem in hardware. The bad builds sound like crap and don’t maintain the house sound - some are flat and boring, some no one wants to listen to for very long.
But a few philosophical and practical points:
Allowing people to pick their favorites will encourage people to not update to newer versions with bug fixes because we can’t find what they like in the next version of the code. This is a headache and will force us to start testing each new set of PIC code with all previous releases of the FPGA code. It’s already enough work to test each new release - that would be a nightmare.
I used to think the best way to build a user interface was to give the user control over as many things as possible and let them pick the versions they liked best. That’s a cop out. The best user interface design is the one that doesn’t make the customer want choices… Notice that the DS doesn’t have filter selections? There’s no locking time vs. sound quality setting. There’s no measurement vs. sound quality setting. There’s no setting for where to roll off the ultrasonic noise… There’s no expected sample rate setting. There’s no choice for how much to upsample or to not upsample at all…
We got here by me choosing the best technical job I could at each juncture and indeed the whole DAC is from a dictatorship: me. I’ve done what I think was best at virtually every point. I didn’t have the budget to test even a small fraction of the design choices I made. What you have is what I got by picking the sound I like (and by the way, the sound that many PCM fans say they don’t like.) I’d veto any firmware choice that varied too much from the sound I’m striving for. We wouldn’t be here if I allowed people to vote on any significant number of the features.
Anyway perhaps this will help people to understand where I’m coming from and why I don’t think giving people a large choice of versions is a good idea.
Is it fair to say that your dictatorship-approach also have resulted in hundreds (if not thousand) of hours done at your own expense?
Especially if this include the prototype?
And if those hours had be payed for we are talking some big numbers here?
The reason for asking is that as a developer I know what it takes to succeed and what have to be sacrificed along the road…
But when you win, the feeling is great 
So we get a great new firmware, that apparently has universal praise, and we’re looking for:
-
a way to make the firmware updates easier
-
being able to switch between multiple firmwares
-
tuning for headphones
-
a way to put less stress on the fpga by bypassing volume, etc
??
Maybe we should enjoy the music a bit before looking to change things.
Ted Smith saidWe got here by me choosing the best technical job I could at each juncture and indeed the whole DAC is from a dictatorship: me.
Another reason why I like you. LOL