Re: [Flexradio] 24 bit soundcards?
>From G3PLX: Lee: Yes, this is what I have discovered the hard way. I was naively expecting the new 24-bit device to be 8 bits better than my old 16-bit card, but I have learnt that it isn't that simple. But this latest discovery is not just about specmanship. There's something odd going on in the hardware which I can't quite work out, and I was hoping someone 'in the know' might explain it. Why is there a low-frequency noise in the lower bits superimposed on the main full-bandwidth signal in the higher bits. By the way, I discovered why the bottom two bits were zero and I hadn't noticed it before. There's a '+12dB gain' button on the control panel. Guess what it does? Guess where I had it set the first time I checked? It's not a 12dB pre-amp, it's a 2-bit good-for-nothing left-shift! 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] 24 bit soundcards?
>From G3PLX: I was talking to a physicist friend about the 24-bit Firebox and the dissapointing performance that I had seen, and he said he had an M-audio Delta 1010LT and would plug it in and see what it was like. I don't know anything about it, but he seemed to think it would be pretty good. His employers probably paid a lot of money for it! He reported slightly lower noise levels than I did, but noticed something odd about the statistics of this noise. Although all 24-bits were 'busy', there was a very high probability (almost 100%) that the difference between two consecutive samples was a multiple of 16. He speculated that the hardware was a 20-bit DAC, and there was some kind of added low-frequency noise or dither added in at the 24-bit resolution. I went back to the Firebox with this in mind. The first thing I found - and I should have seen this much earlier - was that the bottom 2 bits of the 24-bit data from the Firebox were solidly fixed at zero! But on going deeper into the statistics, I found exactly the same characteristic as the Delta 1010. It's a basic 20-bit ADC, but with an extra 2 bits of low-frequency dither rather than 4. I wonder what to make of all this. The discussion on this reflector about 24-bit soundcards revealed that there was a good reason to add-in a bit of dither noise to mask the quantisation effect, but surely that should be at the full sampling bandwidth, not a very low-frequency dither. I estimate it's only a few tens of Hz on the Firebox. Perhaps this isn't a dither mechanism but some kind of DC offset cancellation? In that case it isn't adding LF noise (1/f noise?) but removing it. Whatever, it looks like the Firebox says 24-bits on the box, has 22 bits inside, of which 20 are derived from the ADC, of which 17-18 are useful. That's two weeks in a row I have learnt something new. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] Zero IF: 1/f noise
>From G3PLX: Gerald: Thanks for your response with comments on 1/f noise and image balance. I am really not convinced that 1/f noise is 'physics' in the same sense as, for example, thermal noise is physics which we can't fight. Some circuits exhibit what we term 1/f noise and others don't seem to do so, so I speculate that if we are careful we can design it out. I can see it in my cheapo MP3+, Sami sees it, but Alberto's plot of the Delta 44 spectrum and my results on the Firebox don't show any significant low-frequency noise. There is certainly no sign of it at the input of Gerald's elegant post-QSD amplifier in the SDR1000, which has clearly been designed for the ultimate in low-noise performance, since that's where the signal level is at it's lowest. To me this says we shouldn't be frightened of 1/f noise. I have done a WAV file of an SSB signal received 15dB above noise, received on my SDR1000 into the Firebox using Zero-IF software, and there's no sign of a noise peak in the centre. To do this I 'worked round' the oscillator re-radiation noise by including a unity-gain frequency-conversion (in hardware) before the SDR1000, so it's not a final solution but a demonstration of a possibility. I hoped I could send this to the reflector so that others can listen to it, but this may not be possible. I can email it to anyone who would like to hear it. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Zero IF SDR
>From G3PLX: Sami: Many thanks for taking the time to make those tests. I can see from the results on your website that you get a similar result to Alberto with the Delta 44 card, but there is certainly a lot of noise around zero coming from the SDR1000. If this is with the antenna removed, then it cannot be local oscillator radiation from the antenna. I don't see this here on my SDR1000 set-up, nor on another zero-IF receiver front-end I have here. I can only guess that there are ground loops involved. But if the level of low-frequency noise is likely to be as high as you are seeing, then I agree that zero-IF would not be a popular idea. Maybe it would only work with a dedicated hardware design (like my old commercial GMDSS watch receiver) where the QSD and the ADC are two chips sitting close to each other on the same board. My reason for studying this idea now is because I really believe that the image balance problem could kill the direct-conversion concept. An independent review of the SDR1000 (RadCom June), shows that the image balance, if adjusted to 80dB on 14.2MHz, becomes only 65dB at the band edges and 50dB at 1.8 and 30MHz. I realise that SDR1000 fans will defend this level of performance because they know how to live with it, but if you are honest you must surely agree it's not good. You can poke this problem to move it a little, but it won't go away. If it were possible to solve the low-frequency noise problems, then zero-IF may be one way forward to save the direct-conversion concept. If it cannot be done I think the rest of the world will go back to an analogue superhet with a software-defined I.F. or wait for 100MHz ADCs with 144dB dynamic range. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Tuning Clicks
>From G3PLX: Jack said: >If the cable isn't to blame and this is DDS tuning - as Eric seems to >suggest - then perhaps Peter's comments on linear tuning of the >AD9854 hold an answer. But if DDS tuning is the cause, why is it >frequency dependent? DDS tuning clicks should ONLY occur if there's a big signal nearby. If you do have big birdies on the higher bands (I get them every 1MHz, from other things in the shack), then you may get these clicks worse because of this, especially if the noise level is lower. Try injecting a strong signal lower down the band where you don't normally hear the clicks, and see if the clicks then appear near that. I only just thought of the linear sweep idea. Maybe someone like Eric or Bob might follow it up. Only they could say if it would be easy to do in the supplied software. I have used the linear sweep feature of the AD9854 myself, for following Chirpsounders up the spectrum at 100kHz/sec, but not for tuning steps. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Tuning Clicks
>From G3PLX: Having just read what I just wrote, I realise that there IS a way to reduce the DDS tuning clicks. In the AD9854 DDS chip, as used in the SDR1000, there is a facility to change from one frequency to another by sweeping linearly between the two. There's a 'start frequency' and an 'end frequency' register, and another to set the sweep rate. It should be possible to choose a sweeprate that was slow enough to eliminate the clicks, while still being fast enough to tune across a band manually. You wouldn't need to do it when switching channels or bands. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Tuning Clicks
>From G3PLX: Jack: I have seen tuning clicks on the higher bands with my own software driving the SDR1000 kit. Some of these clicks are the DDS stepping, as Eric described, but these will only be audible if there's a strong signal nearby - in effect the click you hear is the 'keyclick' of this strong carrier stepping sharply, and you will hear it even if the signal in question is outside the passband. If there is nothing but bandnoise for miles either side, you shouldn't here anything when the DDS steps. There can probably never be a fix for this effect when using a DDS direct as a local oscillator (which the SDR1000 does). It didn't happen with the older phase-locked VCO technology because the frequency changes much more slowly. Another source of tuning clicks I found was noise from the parallel port cable between the PC and the SDR1000. When the PC sends a frequency-change to the DDS along this cable there is a burst of fast switching activity here, and this is strong enough to be audible on the higher bands. It was specially noticable if I was scanning across the band, when it's almost a buzzing sound . On my set-up it's not too bad and I have not bothered to try to fix it, but maybe changing to a screened printer cable would help. Of course if the DDS control is done by another method (USB?), then this comment doesn't apply. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Zero IF SDR
>From G3PLX: Alberto: Thanks for your input! The peak in the centre of the spectrum plot you showed at http://sundry.i2phd.com/zeropeak.gif is certainly the small DC offset of the ADC's in the Delta 44. The Zero-IF software would remove this of course. It can be done automatically, it doesn't need a calibration process. Without this null, the DC offset might show as a faint tone, typically at 1700Hz, in the receiver audio. But you didn't hear it anyway on 40m, so there was enough band noise even to drown it. This is a promising result. But it still doesn't show if the new SDR1000 hardware has a low-enough oscillator radiation to be able to do this successfully. Sami's report of 'crud around DC' may say that it will not. Sami: Please could you try an experiment with this crud: First make sure that it is not generated inside the soundcard (see if it stays there with the audio input removed) or in the audio input cable groundloops (unplug the jack from the SDR1000 but touch the jack ground to the SDR100 chassis). If these tests are clear but there is still crud coming through the SDR1000, then see what happens when the antenna is removed and replaced with a 50 ohm termination. If the crud vanishes then it was surely caused by local oscillator radiation intermodulating with LF noise outside the receiver. Or maybe some others could try it. I would really like to know the answer. If it's not going to work with the new SDR1000 hardware then I will take my Zero-IF idea elsewhere. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Zero IF SDR
>From G3PLX: >From what Ahti says, there is not a 1/f problem with the current SDR1000, which is what I found with the earlier version, but the oscillator radiation effect may or may not still be there. In the last topic I raised, concerning 24 bit soundcards, I speculated that the 'ideal' ADC would have the size of the least-significant bit the same as the input thermal noise. Present-day cards seem to have about 48dB (8 bits) more noise than this. This means, conceptually at least, we need 48dB gain before them in order to hear the noise floor. If we put all this gain after the QSD (i.e. at the zero I.F), that would certainly make any 1/f problem show, and I think Ahti was saying that it does show with the Softrock hardware. The answer is to move some of the gain in front of the QSD. Personally I am not even convinced that 1/f noise is real - whenever I encountered noise like this in a circuit I was always able to find a cause and fix it, like the oscillator radiation and power-supply effects I mentioned before. The triple op-amp instrumentation amplifier, used in the SDR1000 for the post-QSD stages, may seem to some people as overkill for what seems to be just an audio pre-amp but it does inherently null the power-supply noise. Notice that this part of the circuit has NO coupling capacitors. I have not seen the softrock circuit diagram. Jim's point about poor (or missing) DC coupling probably crossed with my answer to Frank. It really is not a problem. Jim mentioned clock noise again, and I have to say I haven't studied this at all, but my understanding is that it is a strong-signal effect, not a noise level one. I don't think I have seen any such effects, but I haven't looked for them either. I have not yet heard any convincing arguments to say that zero-IF is a bad idea. If the present SDR1000 software can do it, I would be very interested to hear from anyone who would try some tests with it on the latest hardware. If there is noise around the zero-point, does it change if you insert/remove the pre-amp? Does it vanish if you use the SDR1000 to demodulate a sniff of I.F. signal from another receiver? Are there any nasty strong-signal effects? Can you hear the central notch and does it bother you? How far off can you put the image balance before the wanted signal sounds bad? Do some experiments. Let us know the results. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Zero IF SDR
>From G3PLX: I just checked my two soundcards for the low-frequency roll-off. My new Firebox is 2.4dB down at 1.8Hz and the MP3+ is 1.5dB down at 1.2Hz. And that was done quickly by linking line-out to line-in, so it includes the LF roll-off of the transmit side too. I am quite certain the music business wouldn't touch a soundcard that rolled-off at 200Hz. The LF roll-off is really not a problem for zero-IF anyway. Even if you put the oscillator right in the centre, which theoretically puts a deep narrow null in the passband, I defy anyone to notice it's there on an SSB signal. There are ways to eliminate this null completely, but I really don't think we need to do it. To Ahti: I have never seen 1/f noise in my zero-IF work (I designed such a receiver before I retired, for HF GMDSS working). The local oscillator radiation problem looks just like 1/f noise, but that can be fixed once it is recognised. It's also possible that poor post-mixer design could result in supply-line noise being a problem (this has a 1/f spectrum), but the post-mixer amplifier design of the SDR1000 kit is excellent in this respect. If 1/f noise was present, it would show as a noise peak at the centre of the output spectrum. There is no such peak. If, as Frank says, the SDR1000 software can do zero-IF already, has anyone done any tests with it? What were the results? Were there any problems? Has the local oscillator radiation problem gone now that the RF amplifier is in place? I think it's worth looking at this area again. The 22kHz image problem will be tolerated by SDR1000 fans but this is surely not a proper solution. My GMDSS receiver would not have gained it's approval certificate if the operator had to balance the image rejection each time he changed bands! 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] Zero IF SDR
>From G3PLX: The software that comes with the SDR1000 uses an 11.025kHz intermediate frequency. I understand the reasons for doing it this way, but even before the SDR1000 appeared I was doing software radio with an I.F. of zero. By this I mean that the sine and cosine RF oscillators were set right in the middle of the wanted signal, not offset by 11kHz. This may sound impossible to those who were brought up with analogue RF, but that's because it could never be done with analogue circuitry. With DSP it's actually easier to have the 'IF' frequency down in the audio band than to push it up where you can't hear it. The big advantage of zero IF is that the 22kHz image response problem vanishes. Any slight amplitude or phasing unbalance in the Tayloe sampler just results in an equally-slight amount of in-band distortion. The strongest image-frequency signal you ever need to reject is the wanted signal itself. You don't need to worry about a much stronger unwanted signal 22kHz up the band. When I got the SDR1000 kit (I got a very early one), I used it with this technique, and the results were excellent, except for one thing. It took me a while to trace the problem, but I found it in the end and the cause was a surprise. The problem showed as noise around the centre-frequency of all received signals, but it varied across the bands, and was absent when I unplugged the antenna. It was so bad that it made the receiver unusable on some bands with some antennas. But if I used the SDR1000 to tap-off and demodulate the intermediate-frequency of another receiver, it worked perfectly. The cause was oscillator radiation. The DDS oscillator (right in the middle of the wanted signal) was radiating, intermodulating with all kinds of low-frequency noise sources external to the receiver, and the resulting unwanted products (either side of the oscillator frequency) were re-radiated into the antenna. The effect is well-known to anyone who has ever experimented with home-brew direct-conversion receivers, where it usually shows as a raw power-line buzz in the speaker. It's possible that this effect may well have shown in the early work on SDR and it may have been one reason for offsetting the passband by 11kHz in the present software. The fix is to stop the local oscillator radiation. Screening helps a lot but another way is to add an RF stage, or configure the receiver as a superhet with the Tayloe sampler at the I.F. frequency. My early SDR1000 kit didn't have a pre-amp and I understand the current kits do. The local oscillator radiation is probably considerably lower on the present kits, so the zero-IF technique would probably work a lot better than it does on mine. Has anyone here who is writing his own SDR software tried this on the latest hardware? I can provide more details of the zero-IF technique if required. All the well-known modes can be implemented this way, both for receive and transmit. Maybe the present SDR software could be patched to implement zero-IF, or my own zero-IF software could be run in parallel on another soundcard. Would anyone like to have a go? 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] 16 versus 24 bit audio
>From G3PLX: Sami is right. You CAN recover a signal from noise by narrowing the bandwidth, so if my signal was level with the noise in 24kHz bandwidth, I could filter it to a narrower bandwidth (in software) and improve it's SNR. But that's true regardless of the number of bits in the raw data. So long as we have at least half an lsb of additive noise to dither away the quantisation problem, we can always gain SNR by reducing the downstream software bandwidth. Any more added noise at the front end than this only makes things - well - er - noisier. The noise power calculation I did earlier shows that there is ALREADY about the right amount of noise inherent in the physics to get the dither optimum for a 24-bit ADC at 0dBm. If there are cards with more noise than this, it's probably because the designer couldn't get it any lower for the price, not because he chose to add more noise for some subtle reason. To take Jim's points, I have only so far measured the rms noise with no input, and not yet looked to see if there are any clues in it's spectrum - these things would take a lot more time. Clock jitter wouldn't explain what I see (an output with no input) although it would certainly cause noise in the presence of a large pure tone. I haven't tried that yet either. Let me close this topic before Phil accuses me of cruelty to dead horses. Before I aquired a 24-bit card, I honestly believed that 24-bit cards would be 8 bits better than 16 bit cards. When I did get one recently, I was surprised to find this wasn't the case. Jim is right. 24 bit cards may only be slightly better than 16-bit cards. I have learned something this week. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] 16 versus 24 bit audio
from G3PLX: Hello Sami. Thanks for joining the discussion - I feared I was on my own for a while! I hope some others will contribute too, so I don't get accused of hogging the bandwidth (or perhaps generating too much noise!). You said: >If your low 8 bits are truly random with no input, that's a good >thing! It follows that when you inject a non-random signal, it will >stand out clearly from white noise. Let me be simple-minded and respond:- But if my low 8 bits are truly random and I inject an 8-bit sinewave, isn't it then level with the noise? If I could reduce my 8 bits of added noise to, say, 4 bits, wouldn't my sinewave then be 24dB above the noise? I don't see how adding that much noise can ever be a good thing to do. I am not trying to trap you, I really would like to understand what's going on here. This is all new stuff to me. I had a private email from a broadcast engineer who confirmed what I was saying about dither noise. He said they added 0.5 bits-worth of white noise to the analogue signal before digitisation, and subjectively that was the best result. But that's half a bit of dither, not 8 bits. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] 16 versus 24 bit audio
>From G3PLX: Thanks Paul. Thank goodness I am not alone! I just did a simple calculation. The ultimate rock-bottom noise power is kTB, where k is Boltzmann's constant (1.38e-23), T is room temperature (300 degrees K) and B is the bandwidth in which you measure it - let's say 24kHz for a 48kHz soundcard, so this works out to 1e-16 watts. Let's say the circuit is 600 ohms (just so we can get some idea of the picture), so the r.m.s. noise level across a 600 ohm source connected to a soundcard is going to be the square root of 600e-16, or 0.244uV. Again, just for the sake of it, let's say the soundcard ADC is designed for 2.2 volts peak-to-peak (that's 0dBm in 600 ohms). A 24-bit ADC would then give 2.2/(2^32) volts quantisation level. That comes out to 0.13uV. That's not far off the circuit noise level. That 'feels' about right to me. It says that an 'ideal' 24-bit soundcard should just be able to 'hear' it's own input noise. This is how we have always designed the front-ends of things! I am just throwing numbers in there, and the experts may say that the circuit isn't 600 ohm, or I haven't taken into account the noise factor of the pre-amp, or the typical ADC full-scale isn't 2.2 volts. But I am surely not wrong by a factor of 256? Where is all the noise coming from? Paul's observation that the noise is 100dB down on full-scale could just mean that there is that much noise coming from the SDR1000 RF hardware, but it would be easy to test this by unplugging it and short-circuiting the soundcard input. If the noise stays at the -100dB mark then it says the soundcard is that noisy, which is exactly what I find here with the Firebox. Maybe Paul has already done this. Note that the software measuring the noise needs to be set to the full soundcard bandwidth for this to make sense - it will look a lot cleaner if you select a narrow filter. I am not familiar with the SDR software so I don't know how this works. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] 16 versus 24 bit audio
>From G3PLX: Chris: I wondered about dithering. If I understand the idea, they introduce some deliberate noise in the ANALOGUE part of the ADC in order to spread the spectrum of any quantisation noise. Without it, I can imagine a tiny sinewave input of, say, one bit amplitude, looking to the software like a one-bit squarewave, with 3rd, 5th, ... harmonics. Dithering the analogue input would add a little noise to the fundamental sinewave but push the harmonics down. But, as you say, surely they would only introduce about one bit's worth of dither noise, not eight? 6dB not 48dB? I just did some proper measurements on the Firebox. On runs of 1200 samples at 48kHz, the r.m.s. noise level on the line input is about 400. It's about 450 on the Mike input. I did this by squaring each sample, summing these, dividing by 1200, and taking the square root. At the same time I calculated the DC level, just to make sure that was a lot smaller and didn't confuse things. This is about 9 bits of noise, or more if you suppose the peaks are higher. On the built-in 16-bit card in this laptop the DC level is 15 and I am not sure how to allow for this in the r.m.s calculation, but only the lsb toggles randomly. The MP3+ is the same, unless I select the mike input when the rms noise level is about 10. These figures are what I would have expected intuitively. One would expect the analogue noise-level to be roughly the same as the quantisation noise, in the same way that one would optimise any receiver so that all sources of noise contribute more-or-less the same in order to maximise the dynamic range. Intuitively I would have expected a 24-bit card to do the same, but maybe I am being naive. These measurements, and the figure quoted by Alberto, tells me that the "24-bit" figure in the spec. is to some extent a marketting ploy. I can imagine a scenario where the designer of a new soundcard measured that it only gave 18 bits of useful data, so he decides to mask the data to 18-bits and sell it as an 18-bit card. The marketting manager of the company would surely override him, especially if the competition were all actively marketting so-called 24bit cards! 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] 16 versus 24 bit audio
>From G3PLX: Alberto: Thanks! That's the information I was seeking. It looks as if I was half-right, if there are, at best, 18 good bits in a 24-bit card. As Phil says, if 16-bit cards turn out to be worse that 16-bits, then there is some reason to go for a 24-bit card (which is worse than 24bits!). But for sure, 24 bit cards are not going to be 256-times (48dB, 8-bits) better than 16-bit cards. I will now continue working to convert my own 16-bit SDR code to 24-bits, but I think I will add a button on the front panel which zaps the bottom 8 bits. It will be VERY interesting to see if it makes any difference. To take Frank's point, as far as I can see from both the ASIO and WMME interfaces, there IS an accessible point in the software where this can be done before the audio data becomes floating-point. Thanks to all who replied. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] Fw: 16 versus 24 bit audio
>From G3PLX: > Alan K2WS demonstrated the horrendous broadband noise coming from a very > expensive Audigy 2ZS card by using the spectrum display of his IC-7800 > http://www.n9vv.com/K2WS.html He dumped the Audigy and bought a nice > Delta-66 that has continued to delivery great performance and a clean TX > signal. > de ken n9vv That's a picture of the broadband noise EMITTED from the audio OUTPUT side of the soundcard in question. My query is about the digital noise registered by the input ADC. The tests I have done so far say that there is 8 bits of background noise (1 bit = 6dB) coming from the 24-bit ADC in the Firebox with the input shorted. That says there's only 16-bits of useful signal in there. Why should I pay for 24 bits when 8 of them are below the noise? I say again: Is there anybody out there writing their own SDR software that has actually masked-off the bottom 8 bits or replaced it with random numbers? Phil: I could measure how many bits of noise there are in 16-bit cards, but that would only tell me something about how good or bad 16-bit cards are. What I am trying to figure out is whether we are all buying 24-bit cards because we really need 24 bits, or are we buying 24-bit cards because the manufacturers want us to spend more money. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] 16 versus 24 bit audio
>From G3PLX: Thanks for the pointer Phil. I guess hpsdr.org might be an interesting place to browse anyway, but maybe I didn't make my point clear. I am not complaining that the Firebox I have is NBG, I am wondering if anyone has done any tests which establish whether 24 bit audio gives a better radio than 16-bit audio. OK, I can accept that a good quality soundcard will probably give better results than a cheapo, and the good ones may well be 24-bit while the cheap ones are 16-bit, but has anyone ever masked off the bottom 8 bits of a good 24-bit card and detected the difference? 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] 16 versus 24 bit audio
>From G3PLX: I am just starting to do some serious tests with my new 24-bit Firebox. With the audio input of a channel short-circuit, if I analyse the statistics of the 24-bit data, I find the bottom 8 bits are ALWAYS random. In fact the bottom 9-10 bits are random. This tells me there is no point at all in having 24-bit audio. I might just as well stay with 16-bits. Has anyone here done any rigorous tests comparing 16 and 24-bit audio and found any difference? Like patching the code to mask off the bottom 8 bits completely, or replacing the bottom 8 bits with random data. Maybe someone has added a checkbox to turn this sort of test feature on and off and see what, if any, is the difference in the speaker. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] Soundcard samplerates
>From G3PLX: I don't know if this is common knowledge, but I found a major problem with some soundcards, especially when running SDR software where the input and output sides are running simultaneously. I discovered (the hard way) that you cannot assume that if you set a card to a given samplerate, that it will operate at that rate. Neither can you assume that a given soundcard runs at the same samplerate on receive (record) as it does on transmit (playback). I have three cards here. The first, a Sigma-Tel card built into my Dell laptop, does transmit and receive at the same samplerate always, so SDR software runs well. But the samplerate I get is always rounded-off to the nearest multiple of 300Hz. This means that if I want to run it at 44.1 or 48kHz, I get the exact samplerate, but at NO OTHER common samplerate do I get the right result. For example if I try to set this card to 8kHz, it samples at 8.1kHz. The second card I have is a Creative MP3, a USB device. The receive side behaves the same as the Sigma-Tel card, rounding-off the samplerate to the nearest 300Hz, but the transmit side is different. If I set the transmit side samplerate to every whole Hz value and measure the actual rate, it's clear that it's a much close approximation than 'the nearest multiple of 300Hz'. It's more like the nearest Hz, but not quite. On the face of it, the transmit side of this card is 'better', since it's samplerate is closer to the correct value, but if I use BOTH the receive and transmit sides in the same application, as in an SDR application, there is a terrible problem in that the transmit side (i.e. feeding demodulated receiver audio to the speaker) is running at a different rate from the receive side (i.e. taking received signals from the antenna). At some stage one end will overrun or underrun the other. The software can only handle this by dropping some input data or gapping the output. There's a periodic click in the speaker which can be quite annoying. There isn't an easy solution to this. I found by trial and error that this systematic error between receive and transmit samplerates does NOT occur if I choose a samplerate of 12kHz, 24kHz, or 48kHz. I can see how this might arise from the firmware design. To make sure there are no periodic clicks in the speaker audio, I have re-designed my SDR software to use one of these samplerates. What I don't know is whether this behaviour is peculiar to the MP3 or whether other soundcards do similar nasty tricks. The third 'soundcard' I have now is the Firebox which I have mentioned here before. This behaves exactly like the first card, with the samplerate jumping in 300Hz steps on both receive and transmit. I wonder what other people have found with soundcard samplerate behaviour. 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
Re: [Flexradio] Firebox/ASIO
>From G3PLX: Alberto: Thanks for your answer (and pleased to meet you here). I have not heard of 'WMME' so that's another new word I must throw at Google and learn more. However, if it IS just an extension of the old mmsystem API, then maybe it still suffers from the terrible latency/synchronisation problems. I may not want to go that route. ASIO seems much better in this respect. I could even write AMTOR for an ASIO-based soundcard now! With the Firebox the only way I can find to switch samplerates (and switch the 12dB input boost amplifiers) is via the supplied Control Panel applet. I would have expected to see a way to do this via the ASIO API. The API does have a 'ASIOSetSampleRate' function but it only succeeds if the Control Panel has already set the samplerate I ask for, which seems a bit crazy, so maybe I am doing something wrong. Can anyone tell me if, in the PowerSDR package, the end user who is installing a Firebox is forced to set it up with the Firebox Control Panel, or if PowerSDR knows how to do it itself? If it can be done without the Firebox control panel, then there's hope still. I didn't get any reply from [EMAIL PROTECTED] 73 Peter ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
[Flexradio] Firebox/ASIO
>From G3PLX: Greetings to the group. Can anyone here help me with some questions in the art of driving the Presonus Firebox? The background is that I have been given a Firebox as an unsolicited gift. All the DSP/SDR software I have written has been for 16-bit soundcards, and I found the Firebox runs fine in this mode, especially at the 8kHz samplerate which I use most. What I want to do now is explore the 24-bit capability. I had always had this idea in my head that 24-bits was a sales gimmick and 16 was plenty, but now I have a 24-bit box on the bench, I have no excuse - I must make the effort and actually write some 24-bit code. Although I thought originally that the Windows mmsystem API would do 24-bit (the format would seem to be able to handle it), the Firebox refuses to do it via that route. I had therefore to explore other ways. The 'ASIO' API seemed promising, so I downloaded 'openasio.dll' and the Delphi wrapper (yes I still write in Delphi!), and soon had 24-bit noises emerging from those quaint old 1/4" jack sockets. But only at 44.1, 48, 88.2, and 96kHz samplerate. I don't need these speeds (I do all my SDR at zero I.F) and I really don't want to have to decimate from 48 down to 8kHz just because the hardware won't slow down! So my question is: how can I run the Firebox at 8kHz with 24-bit audio? The only way to change samplerate seems to be via the Firebox control panel, and that only lists 44/48/88/96 in it's drop-down menu. Is there a documented way to change the samplerate programmatically, and can than do 8kHz? I have lot's more questions. If there's someone out there who has already cracked the Firebox and could spare the time to help shorten my learning curve, I would be grateful. 73 Peter G3PLX ___ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/ FlexRadio Homepage: http://www.flex-radio.com
