The Fiio X5 Thread
Nov 6, 2013 at 4:11 AM Post #1,637 of 19,652
So was just looking at some specs, the SOC in the X5 which is the Ingenics 4760b 600mhz is kinda old scool in performance, plus it uses a 130nm production process, meaning the heat output and battery life will not be great, compared to Ingenics newer chips like the 4770 which are 1ghz and use 65nm process. So its a shame that the X5 cannot get an upgraded SOC at double the price of the X3. Especially since cheaper contenders like the DX50 are using very good SOCs like the RK2926 1ghz on the 55nm process.
 
Nov 6, 2013 at 4:21 AM Post #1,638 of 19,652
  So was just looking at some specs, the SOC in the X5 which is the Ingenics 4760b 600mhz is kinda old scool in performance, plus it uses a 130nm production process, meaning the heat output and battery life will not be great, compared to Ingenics newer chips like the 4770 which are 1ghz and use 65nm process. So its a shame that the X5 cannot get an upgraded SOC at double the price of the X3. Especially since cheaper contenders like the DX50 are using very good SOCs like the RK2926 1ghz on the 55nm process.

Trance, you are having fun aren't you? Since when the specs were so important in our decision? I look at the DX50 thread, every page I see people are complaining about problems, some decided to give up already, nothing seems to be working smooth generally. Some people are happy some people are angry. You can get a better chip like that but offer a problematic product, I rather to use good old one but at least it is stabile.
 
I don't know much about the technical things but I learned that implementation makes the final performance, not just a single chip or something like that.
 
Apart from that, I trust FiiO/James's decision on that chip. Everything was covered in X3 will also be covered. I do believe their optimization is good enough to get a smooth experience. Since the beginning James told us, we want a build a dap, which sounds good and also functions perfectly. Which is a quite rare thing in high resolution daps if you think about it.
 
I think we should have done with specs and trust experiences as well as our hearing.
 
That wasn't directly against you, I am just fed up with people on every thread, discussing numbers on the papers and deciding what is best what is worst. I think this is one of the hobbies that we should trust our experience instead of technical specs.
 
Nov 6, 2013 at 4:23 AM Post #1,639 of 19,652
  So was just looking at some specs, the SOC in the X5 which is the Ingenics 4760b 600mhz is kinda old scool in performance, plus it uses a 130nm production process, meaning the heat output and battery life will not be great, compared to Ingenics newer chips like the 4770 which are 1ghz and use 65nm process. So its a shame that the X5 cannot get an upgraded SOC at double the price of the X3. Especially since cheaper contenders like the DX50 are using very good SOCs like the RK2926 1ghz on the 55nm process.

 
 
I think fiio decided o go with this soc because it has more support compared to rockchip
 
Nov 6, 2013 at 4:31 AM Post #1,640 of 19,652
  Trance, you are having fun aren't you? Since when the specs were so important in our decision? I look at the DX50 thread, every page I see people are complaining about problems, some decided to give up already, nothing seems to be working smooth generally. Some people are happy some people are angry. You can get a better chip like that but offer a problematic product, I rather to use good old one but at least it is stabile.
 
I don't know much about the technical things but I learned that implementation makes the final performance, not just a single chip or something like that.
 
Apart from that, I trust FiiO/James's decision on that chip. Everything was covered in X3 will also be covered. I do believe their optimization is good enough to get a smooth experience. Since the beginning James told us, we want a build a dap, which sounds good and also functions perfectly. Which is a quite rare thing in high resolution daps if you think about it.
 
I think we should have done with specs and trust experiences as well as our hearing.
 
That wasn't directly against you, I am just fed up with people on every thread, discussing numbers on the papers and deciding what is best what is worst. I think this is one of the hobbies that we should trust our experience instead of technical specs.

 
I do agree with you that final implementation is much more important than the chips used. While discussing specs is kinda pointless, imo, I am more interested in final measurements, which I understand, won't be out till the final product is ready.
 
Nov 6, 2013 at 4:49 AM Post #1,641 of 19,652

  Trance, you are having fun aren't you? Since when the specs were so important in our decision? I look at the DX50 thread, every page I see people are complaining about problems, some decided to give up already, nothing seems to be working smooth generally. Some people are happy some people are angry. You can get a better chip like that but offer a problematic product, I rather to use good old one but at least it is stabile.
 
I don't know much about the technical things but I learned that implementation makes the final performance, not just a single chip or something like that.
 
Apart from that, I trust FiiO/James's decision on that chip. Everything was covered in X3 will also be covered. I do believe their optimization is good enough to get a smooth experience. Since the beginning James told us, we want a build a dap, which sounds good and also functions perfectly. Which is a quite rare thing in high resolution daps if you think about it.
 
I think we should have done with specs and trust experiences as well as our hearing.
 
That wasn't directly against you, I am just fed up with people on every thread, discussing numbers on the papers and deciding what is best what is worst. I think this is one of the hobbies that we should trust our experience instead of technical specs.

 
Using the new Ingenics SOC would not alter the stability, it would just increase snappiness of the DAP as well as increase battery life, implementation can increase stability, and a bad implementation can affect speed yes, but were talking a very large gap between the two SOC's I mentioned, the newer being twice as fast and also twice as power efficient, it makes a big difference... For example lets look at the Fiio X3 using the same SOC chip, the X3 has a 3000mah battery, the DX50 only has a 2000mah battery, yet the DX50 can usually better the X3 in battery life, how? The two main reasons are the more power efficient SOC in the DX50 and the amp sections lower power output.
 
Basically the X5's battery life will be on par with most DAPS, eg 12-16 hrs, because it uses a bigger battery than the competition, If it used a newer SOC it would have had battery life better than the competition. Also It would have been better to use the DX50's approach to battery life, eg a newer, faster more efficient SOC with a smaller battery than the reverse used in the X3/5.
 
I am just speaking my mind, I am sure some like reading this stuff, others like you get angry :p
 
   
 
I think fiio decided o go with this soc because it has more support compared to rockchip

 
I'm not saying that Fiio should use rockchip, I'm saying it would have been nice if they used the newer version of the SOC in the X3, X3 uses Ingenics 4760b 600mhz 130nm, the newer version is the Ingenics 4770 1ghz 65nm which is meant to directly compete with the Rockchip RK2926 1ghz 65nm used in the DX50.
 
Nov 6, 2013 at 5:06 AM Post #1,642 of 19,652
   
 
I think fiio decided o go with this soc because it has more support compared to rockchip

 
yes, that is why we can develop USB DAC feature for X3. 
 
FiiO Stay updated on FiiO at their sponsor profile on Head-Fi.
 
https://www.facebook.com/FiiOAUDIO https://twitter.com/FiiO_official https://www.instagram.com/fiioofficial/ https://www.fiio.com support@fiio.com
Nov 6, 2013 at 5:57 AM Post #1,643 of 19,652
  So was just looking at some specs, the SOC in the X5 which is the Ingenics 4760b 600mhz is kinda old scool in performance, plus it uses a 130nm production process, meaning the heat output and battery life will not be great, compared to Ingenics newer chips like the 4770 which are 1ghz and use 65nm process. So its a shame that the X5 cannot get an upgraded SOC at double the price of the X3. Especially since cheaper contenders like the DX50 are using very good SOCs like the RK2926 1ghz on the 55nm process.

 
 
   
Using the new Ingenics SOC would not alter the stability, it would just increase snappiness of the DAP as well as increase battery life, implementation can increase stability, and a bad implementation can affect speed yes, but were talking a very large gap between the two SOC's I mentioned, the newer being twice as fast and also twice as power efficient, it makes a big difference... For example lets look at the Fiio X3 using the same SOC chip, the X3 has a 3000mah battery, the DX50 only has a 2000mah battery, yet the DX50 can usually better the X3 in battery life, how? The two main reasons are the more power efficient SOC in the DX50 and the amp sections lower power output.
 
Basically the X5's battery life will be on par with most DAPS, eg 12-16 hrs, because it uses a bigger battery than the competition, If it used a newer SOC it would have had battery life better than the competition. Also It would have been better to use the DX50's approach to battery life, eg a newer, faster more efficient SOC with a smaller battery than the reverse used in the X3/5.
 
I am just speaking my mind, I am sure some like reading this stuff, others like you get angry :p
 
 
I'm not saying that Fiio should use rockchip, I'm saying it would have been nice if they used the newer version of the SOC in the X3, X3 uses Ingenics 4760b 600mhz 130nm, the newer version is the Ingenics 4770 1ghz 65nm which is meant to directly compete with the Rockchip RK2926 1ghz 65nm used in the DX50.

 
Using a newer, speedier is of course good thing, but keep in mind that with a new SoC means a different firmware. The good point of having the same chip is that the majority of the code between X3 and X5 will be compatible, so X5 already benefit from all the bugfix being done on X3. It also mean X5 will be quite stable on the release, and we don't have to face through the same problem faced by early X3 user (which is relatively minor) or even the ear DX50 adaptor (which is still being criticised by some). I don't know if you have used the X3 before or not, but I haven't seen much complained about the speed of the firmware or SoC. The fact is 600MHz is slow for running Android, but plenty enough for just a music player. If you want to double X3 play time, the easiest thing to do is to give it a much weaker amp section and lower supplier voltage to most of the components, which is what iBasso has done on the DX50. It is not that FiiO can't give you a good number, but they have chosen performance over battery life. The reason why X5 has a even larger battery but still comes in around the same battery life is for the same reason - they picks a much more power hunger top-of-the-line DAC that requires 7 opamps (as compared to X3's 2 opamp) in order to build a better player that has some of the highest power output of its class. If they want to cut some corner by limiting the spec, we can probably see much better battery life.
 
Nov 6, 2013 at 6:34 AM Post #1,645 of 19,652
   
Using the new Ingenics SOC would not alter the stability, it would just increase snappiness of the DAP as well as increase battery life, implementation can increase stability, and a bad implementation can affect speed yes, but were talking a very large gap between the two SOC's I mentioned, the newer being twice as fast and also twice as power efficient, it makes a big difference... For example lets look at the Fiio X3 using the same SOC chip, the X3 has a 3000mah battery, the DX50 only has a 2000mah battery, yet the DX50 can usually better the X3 in battery life, how? The two main reasons are the more power efficient SOC in the DX50 and the amp sections lower power output.
 
Basically the X5's battery life will be on par with most DAPS, eg 12-16 hrs, because it uses a bigger battery than the competition, If it used a newer SOC it would have had battery life better than the competition. Also It would have been better to use the DX50's approach to battery life, eg a newer, faster more efficient SOC with a smaller battery than the reverse used in the X3/5.
 
I am just speaking my mind, I am sure some like reading this stuff, others like you get angry :p
 
 
I'm not saying that Fiio should use rockchip, I'm saying it would have been nice if they used the newer version of the SOC in the X3, X3 uses Ingenics 4760b 600mhz 130nm, the newer version is the Ingenics 4770 1ghz 65nm which is meant to directly compete with the Rockchip RK2926 1ghz 65nm used in the DX50.

Dude, I am not getting angry but -spec wars- is not something that important in this subject. Real life experience is what matters. Apparently, they chose the same chip because of many advantages, not because they don't want to use the newest chip around. I am cool with it honestly. As for the battery life, James said 24bit playback should give more than 10 hours, so probably average music listener with some mp3 and some 16 bit flacs etc should get around 12 13 hours? I think that's fine. As for iBasso, they stated it is 13 or 14 hours but did somebody actually see the unit playing 13 hours? I read comments about saying it is around 8 hours.. I am not an iBasso enemy something, in fact, I do like them, but let's face it, even though DX50 is marvelous product, there are A LOT of things to be fixed and I see so many people who needs help fixing his player or facing a problem and try to save the dap. I want both good SQ and stabile "actually" functioning dap, which is why I am gonna get X5.
 
Nov 6, 2013 at 6:58 AM Post #1,646 of 19,652
  Dude, I am not getting angry but -spec wars- is not something that important in this subject. Real life experience is what matters. Apparently, they chose the same chip because of many advantages, not because they don't want to use the newest chip around. I am cool with it honestly. As for the battery life, James said 24bit playback should give more than 10 hours, so probably average music listener with some mp3 and some 16 bit flacs etc should get around 12 13 hours? I think that's fine. As for iBasso, they stated it is 13 or 14 hours but did somebody actually see the unit playing 13 hours? I read comments about saying it is around 8 hours.. I am not an iBasso enemy something, in fact, I do like them, but let's face it, even though DX50 is marvelous product, there are A LOT of things to be fixed and I see so many people who needs help fixing his player or facing a problem and try to save the dap. I want both good SQ and stabile "actually" functioning dap, which is why I am gonna get X5.

 
I'm pretty sure they used the old chip to save a dollar, both on the chip cost  and on extra design production costs, oh yeah clio makes a good point on the firmware issue though, but that again is mainly to save money on further firmware development costs with the new SOC. I own a DX50, the current firmware is stable, I don't think there are many things to be fixed at all, and I get between 12-16 hour battery life with flac and mp3. I am still looking forward to buying the X5, I'm sure it will sound awesome, just disappointed with the SOC really, I have no other issues with the X5 and still think its awesome, just for me personally if it had the new SOC for once I would have had nothing to complain about with a new Dap :p
 
  Using a newer, speedier is of course good thing, but keep in mind that with a new SoC means a different firmware. The good point of having the same chip is that the majority of the code between X3 and X5 will be compatible, so X5 already benefit from all the bugfix being done on X3. It also mean X5 will be quite stable on the release, and we don't have to face through the same problem faced by early X3 user (which is relatively minor) or even the ear DX50 adaptor (which is still being criticised by some). I don't know if you have used the X3 before or not, but I haven't seen much complained about the speed of the firmware or SoC. The fact is 600MHz is slow for running Android, but plenty enough for just a music player. If you want to double X3 play time, the easiest thing to do is to give it a much weaker amp section and lower supplier voltage to most of the components, which is what iBasso has done on the DX50. It is not that FiiO can't give you a good number, but they have chosen performance over battery life. The reason why X5 has a even larger battery but still comes in around the same battery life is for the same reason - they picks a much more power hunger top-of-the-line DAC that requires 7 opamps (as compared to X3's 2 opamp) in order to build a better player that has some of the highest power output of its class. If they want to cut some corner by limiting the spec, we can probably see much better battery life.
 

 
Well the DX50 supply voltage is 9v and X3 is 11v, not a huge difference, while the X3 only uses 2 opamps the DX50 uses 4 so factor that. I personally think the newer SOC would have increased battery life a decent bit, no need for spec cutting.
 
Nov 6, 2013 at 7:52 AM Post #1,647 of 19,652
   
I'm pretty sure they used the old chip to save a dollar, both on the chip cost  and on extra design production costs...

 
You don't really save much money by not using the newer chip. Buying SoC isn't actually just buying one chip and put it on a PCB of your design. It doesn't work that way in the SoC business. You are actually buying a solution, and the SoC comes pre-soldered on the PCB that you co-designed with the SoC maker. Then you put on whatever extra bits you want. So essentially the most expensive part isn't the chip itself, but the contracting and R&D of the whole solution. It takes months to design a workable PCB and that's where SoC maker makes their money. Having the latest chip or not is not the biggest cost. They are in fact usually fairly inexpensive individually.
 
 
  Well the DX50 supply voltage is 9v and X3 is 11v, not a huge difference, while the X3 only uses 2 opamps the DX50 uses 4 so factor that. I personally think the newer SOC would have increased battery life a decent bit, no need for spec cutting.

 
You don't know that for sure. I read through both SoC datasheet and they didn't even quote a power consumption number.
 
Note that, DX50 runs on +/-4.5 (= combined 9V), but they are not using a rail-to-rail opamp (judging from the line-out voltage and headphone-out voltage output at high gain), and unlikely has a very high current output. Given the supply voltage, the WM8740 might not running on full voltage as well. In contrast, X3 is +/-5.5V. Its WM8740 is running on higher voltage, the AD8397 is one of the highest output power, rail-to-rail single opamp you can use, with a dedicated hardware volume control chip that also double for hardware EQ. Just to show you how much that AD8397 is drawing power - I mod'ed my X3 with AD8620 and immediately notice at least an extra 3~5hours battery life improvement. I used to get around 10~11 hrs for the stock X3 and now I get over 15hrs easily. The downside is of course my mod'ed X3 doesn't have anywhere near the output power of the stock X3.
 
It might seen trivia, but all the little things do add up.
 
Nov 6, 2013 at 8:11 AM Post #1,648 of 19,652
The thread has been a pretty good read so far...
 
it just re-confirmed how much potential X3 has...the mod I have is also using AD8620, but the shop helped me to change the coax out to balance out which sort of compensate a bit of output power (this is done by comparing the normal phone out vs. the new balance phone out)....
 
really eagerly waiting for the release of X5
 
Nov 6, 2013 at 8:36 AM Post #1,649 of 19,652
   
 
 
Using a newer, speedier is of course good thing, but keep in mind that with a new SoC means a different firmware. The good point of having the same chip is that the majority of the code between X3 and X5 will be compatible, so X5 already benefit from all the bugfix being done on X3. It also mean X5 will be quite stable on the release, and we don't have to face through the same problem faced by early X3 user (which is relatively minor) or even the ear DX50 adaptor (which is still being criticised by some). I don't know if you have used the X3 before or not, but I haven't seen much complained about the speed of the firmware or SoC. The fact is 600MHz is slow for running Android, but plenty enough for just a music player. If you want to double X3 play time, the easiest thing to do is to give it a much weaker amp section and lower supplier voltage to most of the components, which is what iBasso has done on the DX50. It is not that FiiO can't give you a good number, but they have chosen performance over battery life. The reason why X5 has a even larger battery but still comes in around the same battery life is for the same reason - they picks a much more power hunger top-of-the-line DAC that requires 7 opamps (as compared to X3's 2 opamp) in order to build a better player that has some of the highest power output of its class. If they want to cut some corner by limiting the spec, we can probably see much better battery life.

The followong isn't a response to ClieOs but a reinforcement regarding Trance's post.
 
Beat me to it and FiiO using more line and gain chips to get the result they're looking for has more effect than the SOC chip difference.  600Mhz with a separate vid processor core is plenty snappy enough for any DAP that doesn't double as tablet or phone and even though a thinner SOC track does save energy, the higher speed (if used as you describe) lowers it so the difference isn't as great as quoted. If they're idling, yes, the newer chip will give a clear advantage in battery time but if you're actually using the device, the difference it would make is almost negligible.
 
There are other considerations as well. Working with the same chip allows them to better tweak the performance. While the X3 works well, it is still effectively the test platform for this product which can adjust board layout/design/topology to get the most from it. A different chip would mean a do over and if there is no usable performance advantage, it would more likely become an overall disadvantage. Familiarity is a big deal as this electronics stuff isn't as cut and dry as pseudo techs seem to think.
 
ClieOs may know more about this but I was under the impression that Rockchip does most of the FW work in house and doesn't release enough code for companies to be fully in control.
 
Nov 6, 2013 at 8:54 AM Post #1,650 of 19,652
  ClieOs may know more about this but I was under the impression that Rockchip does most of the FW work in house and doesn't release enough code for companies to be fully in control.

 
If they release any code at all. One of the reason iBasso has to rely on Android is because Android is free to use and tweak, so iBasso isn't nearly as limited on getting the 24/192 driver working as compared to solely relying on RockChips to find them a proprietary firmware solution. That's the reason why RockChips is not an candidate for FiiO during X3 development and even HifiMan is moving away from RockChips after the HM801 / HM60x. But the situation doesn't just apply to RockChips. Most of the major SoC maker out there don't share their code with their smaller customer as well.
 

Users who are viewing this thread

Back
Top