DSDPlus
- Thread starter redbeard
- Start date
- Status
Mike, yeah I was looking at the reddedit page where the author states that it'll record a MP3 of all voice content unless you use a particular switch.
It looks like that was incorrect -- it just produces the WAV file. Which you can later manually convert.
Recording to mp3 works fine here, although I went back to wav for the better quality.
The MP3 is automatically generated for you or are you doing it manually? I'm only getting a WAV file generated by default.
Boatanchor
Member
- Joined
- Jul 17, 2011
- Messages
- 991
- Reaction score
- 5
DSDPlus audio a bit 'harsh' on the ears - The evidence..
Ok, first off let me say that I get near perfect decode with DSD1.60 and DSDPlus on P25 Phase 1.
My initial feelings were that the DSDPlus audio sounded a bit harsh and not as 'mellow' as the original DSD audio when listening to the same systems. So, in an effort to try and quantify this apparent difference, I used Spectran, an audio spectrum analyzer software package, to look at the recovered audio from both decoder packages.
The results were interesting..
I will try and post some screen grabs later, but the bottom line is:
DSD1.60:
P25 recovered audio frequency response is essentially flat from 0hz - 3,000Hz with a tight roll-off in in high frequency response/amplitude between 3,000Hz to 3,500Hz.
DSDPlus:
P25 audio response is less flat, with an obvious plateau (peak audio amplitude/response) between 500Hz and 4000Hz.
The high frequency roll-off begins at 4,000Hz and ends at 4,500Hz. There is also a noticeable trough in audio response between 0Hz and 500Hz.
Summary:
DSDPlus certainly does produce a wider audio frequency response of 0Hz - 4,000Hz, but the audio response over this bandwidth is not flat. The emphasis on mid to high audio frequencies is what tends to make the DSDPlus audio sound a bit unnatural and a bit harsh to listen.
DSD1.60 on the other hand produces an essentially flat audio response between 0Hz - 3,000Hz.
This flat response makes for a more realistic sounding audio/voice, with excellent bass response.
Basically, the DSDPlus' 0Hz - 500Hz audio spectrum needs addressing and a programmable LPF that allows users to set the point at which the high frequency roll-off begins, would be nice.
A programmable LPF with roll-off beginning at 3Khz, 3.5Khz or 4Khz would be ideal.
Incidentally, in regards to the -dr(#) 'rolloff' switch/function, it does nothing for audio response. I don't know what it does, but it does not affect recovered audio response - In case anyone wonders
Anyway, food for though for someone
DSDPlus is an excellent package, but the recovered audio response could benefit from some software tweaking.
Ok, first off let me say that I get near perfect decode with DSD1.60 and DSDPlus on P25 Phase 1.
My initial feelings were that the DSDPlus audio sounded a bit harsh and not as 'mellow' as the original DSD audio when listening to the same systems. So, in an effort to try and quantify this apparent difference, I used Spectran, an audio spectrum analyzer software package, to look at the recovered audio from both decoder packages.
The results were interesting..
I will try and post some screen grabs later, but the bottom line is:
DSD1.60:
P25 recovered audio frequency response is essentially flat from 0hz - 3,000Hz with a tight roll-off in in high frequency response/amplitude between 3,000Hz to 3,500Hz.
DSDPlus:
P25 audio response is less flat, with an obvious plateau (peak audio amplitude/response) between 500Hz and 4000Hz.
The high frequency roll-off begins at 4,000Hz and ends at 4,500Hz. There is also a noticeable trough in audio response between 0Hz and 500Hz.
Summary:
DSDPlus certainly does produce a wider audio frequency response of 0Hz - 4,000Hz, but the audio response over this bandwidth is not flat. The emphasis on mid to high audio frequencies is what tends to make the DSDPlus audio sound a bit unnatural and a bit harsh to listen.
DSD1.60 on the other hand produces an essentially flat audio response between 0Hz - 3,000Hz.
This flat response makes for a more realistic sounding audio/voice, with excellent bass response.
Basically, the DSDPlus' 0Hz - 500Hz audio spectrum needs addressing and a programmable LPF that allows users to set the point at which the high frequency roll-off begins, would be nice.
A programmable LPF with roll-off beginning at 3Khz, 3.5Khz or 4Khz would be ideal.
Incidentally, in regards to the -dr(#) 'rolloff' switch/function, it does nothing for audio response. I don't know what it does, but it does not affect recovered audio response - In case anyone wonders
Anyway, food for though for someone
DSDPlus is an excellent package, but the recovered audio response could benefit from some software tweaking.
Last edited:
What you are seeing is errors in the decoding... DSD doesn't give any indication of Privacy being used.
br0adband
Member
It's just denoting an error on the voice slot/channel, it's not a designation for encryption from what I can tell. I have multiple systems that I've been monitoring closely the past 2 days since I got DSDPlus and noted that on many of them when there's either a) a data call on one slot and a voice call on the other or b) a voice call on both slots at the same time.
If a comm on MOTOTRBO is actually encrypted, then that would mean it's a purely digital stream of data that would go out as such so there wouldn't be any notice of it, at least as I understand it. In DMRDecode if you see something like "Data Call from 95 to 64250" or something similar that would (more than likely) denote a "private call," again at least as I understand how DMR/MOTOTRBO works - it could also be an actual data transmission to a unit like a short text message or whatever. I'm not fully "briefed" on everything DMR/MOTOTRBO can do but from what I've seen, that's typically the usage.
If a comm on MOTOTRBO is actually encrypted, then that would mean it's a purely digital stream of data that would go out as such so there wouldn't be any notice of it, at least as I understand it. In DMRDecode if you see something like "Data Call from 95 to 64250" or something similar that would (more than likely) denote a "private call," again at least as I understand how DMR/MOTOTRBO works - it could also be an actual data transmission to a unit like a short text message or whatever. I'm not fully "briefed" on everything DMR/MOTOTRBO can do but from what I've seen, that's typically the usage.
br0adband
Member
So what you're saying is you prefer a more bassy/midrange style audio decode, while some of us actually probably prefer what DSDPlus is creating with the default auto settings - myself being one of them. That's been the biggest complaint of mine since I started doing the digital side of things: that all the decodes I've heard, whether they're from scanners that can do P25 natively or using DSD on a computer (not DSDPlus, mind you), have been a bit "mushy" and bassy, muddy in terms of clarity which is the best way I can put it. I'm not talking about the anti-aliasing we almost always end up with because we're dealing with digital transmissions that have to be changed back into analog, I'm talking about the sheer audio quality period in terms of the actual frequency response.
So far for me I'm happy with what DSDPlus is putting out, I find it outstanding overall and clear as a bell with just the defaults. I've toyed around with the tuning somewhat but I haven't noticed any major differences with the modes (and I do most of my monitoring with headphones on). Maybe it has more effect on P25 than DMR/MOTOTRBO or the other modes, I don't know - P25 is few and far between here in my area which seems a bit weird considering how much radio comms are taking place but, the biggest system will soon be adding P25 Phase II pretty hardcore from what I've been reading so, let's hope whoever the talented person that has put DSDPlus together can whip up some more magic and get Phase II support as well.
I'd love to do more experimentation with the P25 and even the X2-TDMA decoding ability of DSDPlus since they ARE used here in my area, but the actual transmissions are so few and far between it's not even funny - the readme/guide says to do a raw recording of 60 seconds of P25 to tune it - I don't think a single transmission I've heard here (when I can FIND them, that is) lasts more than 2-5 seconds at most. The traffic from Nellis AFB is incredibly swift, the people key up the mic and wham, they're done - I feel like I'm a human spread spectrum receiver hopping around with SDR# trying to snatch a snippet here and there. Yes I could set everything up with Unitrunker + both RTL sticks and probably get more to work with but I'm focused on DMR/MOTOTRBO systems for the most almost exclusively.
Anyway, I guess I just prefer my radio comms to have a "high end" even if it's pretty low in the frequency range. Whenever they sound like people talking with their mouths covered not only by their hands but with a handkerchief under their hand it bugs me to no end; as I'm getting older that upper frequency response makes a big difference in the clarity and understanding the speech overall, at least for me.
EDIT:
Having said all that, and because I lack any decent P25 transmissions to work with, would someone out there be so kind as to do a raw recording using DSDPlus (press R to start recording raw traffic, press it again to stop) of some fairly busy and consistent P25 comms taking place, preferably a good 60 to 120 seconds of it so I can play around with the tuning stuff? I simply can't get a good "sample" using the two or three systems here in my area because as noted above the comms are so damned short, just burps lasting a few seconds then minutes before the next one comes around. I'd appreciate it if anyone could help with that, thanks.
Last edited:
Boatanchor
Member
- Joined
- Jul 17, 2011
- Messages
- 991
- Reaction score
- 5
Yep, horses for courses..
I prefer natural/analogue sounding audio, whereas others prefer squeaky/tinny audio that sounds like it's coming out of a 1" speaker in a $2 transistor radio
DSPPlus does tend to accentuate background noise a lot more compared with DSD for some reason too.
I don't know if this is related to the wider audio bandwidth, it is almost as though there is some kind of savage audio compression going on
in the audio section of the software.
Actually, IMHO DSDPlus closely emulates GRE recovered audio, while DSD1.60 provides audio more closely matched to Uniden's audio.
There's plenty of people that say they prefer the GRE recovered audio, but I'm not one of them.
Whatever turns your dial
Last edited:
I've got numerous digital scanners. However, I tend to want to monitor a slew of Smartzone sites that individually have little traffic on them but combined have a lot of traffic. Additionally, there is very little encryption in use on this particular system I monitor.
It's been great to just enter in the alt cc (which also doubles as a voice channel) and all voice channels into one bank on the scanner and then press Scan. That gets fed in via discriminator tap to the PC, and DSDPlus is then used to decode. If were doing this on a digital scanner, I could actually enter in those conventional channels with a NAC for further reduction in interference.
This really allows me to monitor so much more, so much more quickly than having to have 10+ sites programmed in and scanned in order.
Of course it won't take the place of those times when I actually like being able to choose what I want to listen to and see the actual talkgroup / radio iDs by looking over at the scanner. But having the ultrarobust DSD+ to decode, even pretty weak signals, I have a near constrant stream of activity. It's great.
Mike
It's been great to just enter in the alt cc (which also doubles as a voice channel) and all voice channels into one bank on the scanner and then press Scan. That gets fed in via discriminator tap to the PC, and DSDPlus is then used to decode. If were doing this on a digital scanner, I could actually enter in those conventional channels with a NAC for further reduction in interference.
This really allows me to monitor so much more, so much more quickly than having to have 10+ sites programmed in and scanned in order.
Of course it won't take the place of those times when I actually like being able to choose what I want to listen to and see the actual talkgroup / radio iDs by looking over at the scanner. But having the ultrarobust DSD+ to decode, even pretty weak signals, I have a near constrant stream of activity. It's great.
Mike
mmisk
Member
No Luck With DSD+ So Far
Well no luck with DSD+ so far, Have had DSD running 100% for years now with tap.
Will keep on experimenting, lots of fun!
Here is a screen capture:
Well no luck with DSD+ so far, Have had DSD running 100% for years now with tap.
Will keep on experimenting, lots of fun!
Here is a screen capture:
Last edited:
start the program with dsd -i# where # is the input you want. In your case, I think it should be dsd -i3 since you want the virtual cable to be the input. -o# would be for the output. -g### will specify output audio gain.
DSD Plus?
dsdplus is creating WAV files but they have no content. This goes for the auto generated file and when I generate one with my own file name. It is creating mp3's. I am losing audio quality with mp3 though. My
real time audio sounds good
but the mp3 recording has a lot of background.
dsdplus is creating WAV files but they have no content. This goes for the auto generated file and when I generate one with my own file name. It is creating mp3's. I am losing audio quality with mp3 though. My
real time audio sounds good
but the mp3 recording has a lot of background.
Nick,
FWIW, I am seeing similar output for the Disneyland system, which reportedly is using encryption.
Mark
Sent from my Samsung Galaxy S4 using Tapatalk
AZScanner
Member
- Joined
- Dec 19, 2002
- Messages
- 3,342
- Reaction score
- 13
In case anyone is interested, I've set up a live feed of the Phoenix RWC Simulcast F using SDR#, Unitrunker and DSD+ so folks can get an idea how well it works. Enjoy!
-AZ
-AZ
- Status
Similar threads
- Sticky
SDS100/SDS200:
Firmware update available for SDS200/SDS100