Rerurn to Romy the Cat's Site

Didital Things
Topic: DSP is a fantastic tool

Page 1 of 1 (13 items)


Posted by noviygera on 05-15-2014
fiogf49gjkf0d
I have been thinking about how nice it would be to have have a good sounding DSP for crossover and delay functionality. I know this subject was covered here before but what was talked about is this DSP is not that good, and this DSP is horrible, etc.. 
But, if it is possible to have a good DAC, why is not possible to have good sounding DSP? The dsp is adc + digital processor + dac, right? So if we apply proper signal level and there is enough processor power, would it not be possible to have good sounding DSP?
Gera   

Posted by rowuk on 05-15-2014
fiogf49gjkf0d
 noviygera wrote:
I have been thinking about how nice it would be to have have a good sounding DSP for crossover and delay functionality. I know this subject was covered here before but what was talked about is this DSP is not that good, and this DSP is horrible, etc.. 
But, if it is possible to have a good DAC, why is not possible to have good sounding DSP? The dsp is adc + digital processor + dac, right? So if we apply proper signal level and there is enough processor power, would it not be possible to have good sounding DSP?
Gera   

Romy has covered this a couple of times. It basically amounts to DSP can delay, but it can't transparently filter. Analog can filter, but it can't delay.

That being said, a DSP is a great tool to test crossover strategies and to help pick good ranges for getting different drivers cooperative. It can even be useful for those that do not demand the most advanced playback. At that point, it just doesn't belong at this site. There are other sites that can really appreciate the "technology".

I used DSP for a long time as I could not decide on how to crossover my horns. When I finally switched to a capacitor and time alignment instead of DSP, a bunch of other stuff became apparent...........

Posted by Paul S on 05-15-2014
fiogf49gjkf0d
Maybe it's not really "funny", but almost all "contemporary" recordings are immediately subject to DSP. Likewise, most amplified sound at contemporary venues. Not saying it doesn't matter anymore, but it matters much more when considering DSP vs end-to-end analog. I remain DSP curious about LF, although I have not cared for the pro LF I have heard yet, even though there can be plenty of it. It seems like total Musical expression always suffers from the DSP, and I have a mental picture of the sound passing through a screen, or maybe the movie "The Fly", where the guy gets freakishly "re-assembled" on the other end of a transporter (or something like that).


Paul S

Posted by Romy the Cat on 05-15-2014
fiogf49gjkf0d
Yes, there is DSP in many DACs but there are different DSPs. When you delay, recalculate, stop and stat digital stream then it is perfectly fine.  The whole notion of quantization of signals and reconstructing it later on is not a big deal. For sure that are problems to do it but they are addressable in one way or another. If digital does only what it should do: process quantization particles then there is no problem with DSP. The problem begin when digital begin to run approximations, adding or removing bites, conduct extrapolations and etc. This is exactly what happens during change of volume, or namely filtration.

Some people would argue that the problems inflicted by DSP are very minor. Dealing with them I can testify that they have not so "evolved" listening culture  and mostly they do not know what they listen while they are listening. True and false in this thing exists only in context of perception. If  the sensitively of your perception. If you feel that the impact that DSP inflicts to sound is not notable to you, or notable but you do not feel bothered, then I do not see why you should not use digital filtration for instance.

Me

Posted by scooter on 05-16-2014
fiogf49gjkf0d
Novi -


I am a big fan of digital processing. . .for experimentation and testing. So flexible, inexpensive, speedy with today's processing power. But once the testing is over, let me know when you find DSP that does not provide a significant negative penalty. And kindly explain why.  


Really wish it was more practical. Please refer to my experience:


http://www.goodsoundclub.com/Forums/ShowPost.aspx?PageIndex=1&postID=19646#19646

Posted by noviygera on 05-17-2014
fiogf49gjkf0d
I will give this one a try and let you know:http://www.waf-audio.com/index.php?lang=en
Sure, it may be another failure but maybe not; the only way is use it and listen.
Gera

Posted by Paul S on 05-17-2014
fiogf49gjkf0d
Gera, what you are listening to might also be a factor; you never did say. The "best" DSP I have heard actually has a sort of smoothing/homogenizing effect, not entirely dissimilar to FM. I have some digitized jazz "classics" that are fine for "easy listening", and I am guessing DSP X/Os would matter less and less as my listening became less and less critical. Otherwise, I agree that the best use of DSP X/Os is for quick, dirty and cheap modeling, and certainly "worth it" for that alone.

Of course, in the end, if you like it for whatever, then you like it, regardless.


Best regards,
Paul S

Posted by noviygera on 05-18-2014
fiogf49gjkf0d
Paul,
I have a slightly different approach, I believe the sound system must not be optimized for the certain kind of music. I listen to a variety, but generally I believe the system must reproduce the sound, not the music. Music is made up of sounds... Paul, you mention your listening is becoming less and less critical. Why?
Gera

Posted by steverino on 05-18-2014
fiogf49gjkf0d
PS: The "best" DSP I have heard actually has a sort of smoothing/homogenizing effect, not entirely dissimilar to FM.

Bingo. But I would quibble with the homogenizing aspects of FM through an excellent vintage tuner because many transients and sounds are quite well defined. What FM is always is rolled off in the treble. If you listen in a concert hall, the very high treble is rolled off too unless you are the oboist in front of the trumpets. I may never get fooled that the FM tuner system is providing a window into the recording but it sure is a glamor photograph of it. Since it is in a modest sized room it also helps that the volume is not set too high. It is a nice contrast with my main system in which I try to optimize clarity and  naturalness.

Posted by Paul S on 05-18-2014
fiogf49gjkf0d
Gera, check back and you will see that I was hypothicizing; I did not say or suggest that my listening is less and less critical, although there are times when I am not "listening in" to the music but I am only in the midst (or, the mist...) of it. And I generally agree that optimizing the sound system for a certain kind of music can be limiting But I do think that "less demanding" listening of "less demanding" or "less loaded" music allows the "listener" "more slack" where the playback is concerned. And conversely, I think that critical listening to "serious", complex and/or loaded music requires "more" from the playback to "deliver" more of the musical goods to the critical listener without screwing things up too much. I posted fairly recently that I also think sound, per se, is an important consideration for "proper" playback, since sound is so important to the music, itself. I am also aware that some people either do not hear or they are simply not bothered by the effects of DSP. And I say, seriously, good for them. In fact, I eagerly await the time when I am not bothered by DSP X/Os, since it would make things a lot easier. I hope you will share any good news on that front.

Best regards,
Paul S

Posted by Paul S on 05-18-2014
fiogf49gjkf0d
I dunno about it's place in this thread, but another thing FM is always is highly compressed. In fact, when I think of all the things that are "wrong" with FM, I am brought back to reconsider the bumble bee. It would be nice to know why it "works". Meanwhile, relax and enjoy what's enjoyable.

Best regards,
Paul S

Posted by steverino on 05-18-2014
fiogf49gjkf0d
FM is Analog Signal Processing vs DSP. The compression aspect is only really intrusive on orchestral works. Pop and jazz are already compressed and chamber music presents a more limited dynamic range if not limited contrast. But on the evidence most listeners have no problem with dynamic compression. To be fair, audio systems even with horns impose a noticeable lag on dynamic changes on top of those in the live performance where they are much smoother (and more differentiated by instrument) anyway

Posted by Romy the Cat on 05-19-2014
fiogf49gjkf0d
 steverino wrote:
FM is Analog Signal Processing vs DSP. The compression aspect is only really intrusive on orchestral works. Pop and jazz are already compressed and chamber music presents a more limited dynamic range if not limited contrast. But on the evidence most listeners have no problem with dynamic compression. To be fair, audio systems even with horns impose a noticeable lag on dynamic changes on top of those in the live performance where they are much smoother (and more differentiated by instrument) anyway

Well, first of all there is no such a thing as Analog Signal Processing in FM, there is no analog broadcasting anymore.  There is nothing wrong in DSP during FM and all stations nowadays send signal from studio to transmitter as digital stream. Again, there is nothing wrong with that. If a signal juts sliced by AD processor and then reconstructed on another side by DAC then it would be fine. The problem is that stations are trying to save money on bandwidth and digitally  compress own stream, that cause math more damage than extra 25 AD to DA straight conversions. There is many other things that they do that fuck up signal. For instance there are regulations that demands that HD Radio shall be synchronous with FM but they never are there are reasons for it). So, what the do is to select one steam and Master , another stream and Slave and they the yank the Slave steam arcos time domain each let say 1 second to assure that the signals are synchronous and if a person loses HDFM signal and his tuner automatically switches to FM then the user would not note it. Well, FM doe not loose signal in binary fashion, HD Radio does, so all station that broadcast in both HD Radio and FM do use HD Radio as master. So, any station that does dual broadcast has FM signal already compromised. I have written about it before. There are many other things but this is not FM thread.

I would like to note one more things. Compression is not inherent evil of FM, it is how digital limiters and compressor are used by station personal. Mostly they used very barbarically and this is the problem. I can only assure you that if we had playback music with dynamic range limited by FM dynamic range we would have much better music then what we have today. FM has 50dB dynamic range, not enough, right? Well, vinyl has 50-60dB and I did not hear people complaining too much about vinyl dynamic range. (Some special pressing vinyl might push 70dB however). The 16bit is 96dB capable but in reality it never goes over 60-65db. Mastering, the stupid one, kills everything. The High Resolution formats of cause are more  dymicly capable. The 24bit recording are effectively 144dB of dynamic range but there are no true 24bit possessors out there and practical range of microphone are 120dB, so the human ears. There are a LOT of compression if concert halls and  to get 80-90dB of dynamic range is practically impossible from the positions where the orchestral sections are properly mixed. So, generally the practical dynamic range would be 60-80dB. Trust me, if you hear live, unmixed, unlimited FM at 60dB it would be all that you need.

Page 1 of 1 (13 items)