What's new

some thoughts about signal/noise ratio (1 Viewer)

Hartwig Hanser

Second Unit
Joined
Oct 9, 1998
Messages
301
I followed with interest the many discussions about pre/pros . One of the arguments seems to be that it should have a high SNR. For example, the Outlaw 950 seems to have a very high SNR (no offical numbers yet, but estimated to be over 100 dB), while the new Rotel 1066 has a SNR of 95 rsp 92 (for DD and DTS) in its specs. Of course, it sounds very reasonable that the higher the SNR, the better.

But I made an experiment just now to ask myself, how relevant this might be.

My tape recorder lists following SNRs: 60dB w/o dolby noise reduction, 68dB with Dolby B, 76dB with Dolby C. I popped in an empty cassette, turned on the volume to my usual setting and listened to the noise of the empty cassette. Of course, w/o Dolby it was clearly present. With Dolby B it was much reduced and very weak, but still detectable. With Dolby C it was practically not detectably.

All this is true for listening from my usual listening position. Of course when I come close to the speakers, the noise is better detectable resp. louder. But I wanted to check it in a normal listening situation.

So that means for me, that with an SNR of 76dB the noise is almost not detectable. Since every 6dB more, the noise level is down by half, I would think that anything over 82dB should be enough for real world listening, even of very quiet music rsp music with a wide dynamic range. With usual movie soundtracks, probably even less would be sufficient.

The conclusion would be that it is a moot point to argue for example a SNR of 92dB versus one of 102 dB, since both would be more than enough for letting noise disappear below the hearing threshold.

I invite discussion and other thoughts.

Hartwig
 

brucek

Second Unit
Joined
Dec 29, 1998
Messages
335
Hartwig,

Well, I think your test doesn't really prove your point too well. I believe the Dolby noise reduction schemes are a record and playback system where if you record in Dolby B and play back in Dolby B then the resultant reduction in noise will be a certain decibel drop. You can't really just pop in a blank tape and make an evaluation of Dolby vs SNR.

That said, lets take a theoretical look at CD's and noise. Since this is a 16 bit medium we can quickly calculate that it can theoretically result in a dynamic range of about 98 db. If your system cannot provide a signal to noise ratio of 98 db you will not be able to resolve all 16 bits of a CD. If your system had a SNR of 92dB (15 bits) then you would be completely missing the last bit of a CD in the noise.

brucek
 

BruceD

Screenwriter
Joined
Apr 12, 1999
Messages
1,220
Another point to remember is that noise is additive.

That means every link (CD player, prepro, amp) in the chain from source to output conspires to worsen (lower) the SNR.

This implies you want to have the highest SNR possible in each component to keep these additive values at a minimum for the entire system.
 

Legairre

Supporting Actor
Joined
Apr 4, 2000
Messages
815
I think S/N is a important spec, but I also think it's just one of many factors that make a unit sound good or great. Many units with S/N in the 90 range sound very good. I'm not sure if I could tell a difference between a 90 S/N and a 100 S/N. Also another thread showed that most of the units posted in that thread with a S/N of 100 or higher also had a price tag $2500 or more.
 

Hartwig Hanser

Second Unit
Joined
Oct 9, 1998
Messages
301
Thanks for the replies.

Another point to remember is that noise is additive.

That means every link in the chain from source to output conspires to worsen the SNR. This implies you want to have the highest SNR possible throughout the system (in each component) to keep these additive values at a minimum.
This is correct and an important point. So in such a chain I have to add the noise of the pre/pro and the noise of the power amp. If the power amp has the same SNR, the total SNR would be 6dB lower. On the other hand, if the Power amp had a much higher SNR (mine has one of 105dB), the total SNR would not be reduced significantly in comparison to the pre/pro SNR.
 

AntonS

Stunt Coordinator
Joined
Dec 18, 2001
Messages
164
You can't really just pop in a blank tape and make an evaluation of Dolby vs SNR.
You can. Essentially that's what Dolby noise reduction schemes do - reduce noise, whatever the source of the noise is. It's just that during the recording the original signal is altered in a way so certain frequencies (depending on Dolby type) are boosted up. Then during playback they are brought down again, taking noise with them. So you can just listen and compare noise reduction from an empty tape.
 

Philip Hamm

Senior HTF Member
Joined
Jan 23, 1999
Messages
6,874
Anton,

It's not quite that simple, but I agree, that's essentially (no where nere completely) what happens with Dolby - B. Dolby C and S are different animals entirely.

Hartwig, do you have any experience with Dolby S on cassette? Holy crap it sounds as good as CD, no shit.

Anyway, S/N is one of many measurements that seems to be getting way too much attention these days.

I get about 50-60 dB s/n on LP records and they still sound frickin' awesome.
 

Philip Hamm

Senior HTF Member
Joined
Jan 23, 1999
Messages
6,874
Ahh yes.. MiniDisc... So nice.... I'm starting my quest to abandon the format, but it has served me well and continues to do so.
 

Users who are viewing this thread

Forum statistics

Threads
356,814
Messages
5,123,734
Members
144,184
Latest member
H-508
Recent bookmarks
0
Top